Long-term follow-up andScreening for new aneurysms

after Subarachnoid Hemorrhage

Marieke Wermer

ISBN-10: 90-393-4356-XISBN-13: 978-90-393-4356-2Cover: Roy SandersLay-out: Roy SandersPrinted by: Gildeprint drukkerijen

Publication of this thesis was fi nancially supported by:Boston Scientifi c

Long-term follow-up and Screening for new aneurysms

after Subarachnoid Hemorrhage

Follow-up en Screening op nieuwe aneurysmata na een subarachnoïdale bloeding

(met een samenvatting in het Nederlands)

Proefschrift

ter verkrijging van de graad van doctor aan de Universiteit Utrecht

op gezag van de rector magnifi cus, prof.dr. W.H. Gispen, ingevolge het besluit van het college voor promoties

in het openbaar te verdedigen op vrijdag 29 september 2006 des middags te 14.30 uur

door

Marieke Johanna Hermina WermerGeboren op 26 februari 1973 te Almelo

Promotor: Prof.dr. G.J.E. Rinkel

The research described in this thesis was supported by a grant from the Netherlands Organization for Scientifi c Research (NOW) /ZonMw (grant 945-02-007).

Voor Igor en Juliette

Chapter 1 General introduction 8

Part I. Long-term follow-up after subarachnoid hemorrhage

Chapter 2 Late recurrence of subarachnoid hemorrhage after treatment for ruptured aneurysms: patient characteristics and outcome. Neurosurgery 2005;56:197-204.

14

Chapter 3 Incidence of recurrent subarachnoid hemorrhage after clipping for ruptured intracranial aneurysms. Stroke 2005;36:2394-2399.

30

Chapter 4 Long-term mortality and vascular event risk after aneurysmal subarachnoid hemorrhage. In preparation.

44

Chapter 5 Anosmia after subarachnoid hemorrhage. Submitted. 58

Chapter 6 Long-term effects of subarachnoid hemorrhage on employment, relationships, personality and mood. Neurosurgery, in revision.

70

Part II. Screening for new aneurysms

Screening in familial subarachnoid hemorrhage

Chapter 7 Repeated screening for intracranial aneurysms in familial sub-arachnoid hemorrhage. Stroke 2003;34:2788-2791.

84

Chapter 8 Psychosocial impact of screening for intracranial aneurysms in relatives with familial subarachnoid hemorrhage. Stroke 2005;36:836-840.

98

Content

Screening in patients with a history of subarachnoid hemorrhage

Chapter 9 Yield of screening for new aneurysms after treatment for sub-arachnoid hemorrhage. Neurology 2004;62;369-375.

112

Chapter 10 Follow-up screening after subarachnoid hemorrhage: frequency and determinants of new aneurysms and enlargement of existing aneurysms. Brain 2005;128:2421-2429.

130

Chapter 11 Effectiveness and costs of long-term follow-up screening for new aneurysms in patients with clipped aneurysms after subarachnoid haemorrhage: a cohort study and decision model. Submitted.

146

Chapter 12 Yield of short-term follow-up CT/MR-angiography for small aneu-rysms detected at screening. Stroke 2006;37:414-418.

170

Chapter 13 Risk of rupture of unruptured intracranial aneurysms in relation to patient and aneurysm characteristics: an updated meta-analysis. Stroke, in revision.

182

Chapter 14 General discussionSamenvattingSummaryDankwoordList of publicationsCurriculum vitae

198210216222228234

General introduction

ChapterOne

10

Chap

ter

Subarachnoid hemorrhage (SAH) from a ruptured aneurysm is a subset of stroke that affects mainly young persons; half of the patients are younger than 55 years of age. Despite advances in treatment, short-term outcome is still poor with a case-fatality rate of 50% and a disability rate of 20-30%.1 Patients who recover after SAH and are successfully treated for the ruptured aneurysm generally are believed to have a prognosis similar to that of persons without a history of SAH. There is, however, limited evidence for this widely held opinion.Recently it was shown that new aneurysms can develop in patients after SAH. The incidence of de novo aneurysm formation in previous studies have been described to be between 0.8 and 2.2% per year and the incidence of recurrent aneurysms at the clip-site around 0.5% per year.2-5 Because these series were often small and inhomogeneous the exact incidence of new aneurysm formation after SAH is still uncertain. Several studies suggest that the rupture risk of untreated aneurysms is higher in patients with a history of SAH than in patients without a history of SAH.6, 7 The combination of the ongoing development of aneurysms and the in-creased rupture risk after a previous episode of SAH indicates that patients with a history of SAH might be at considerable risk of a recurrent SAH.Until now, two studies have reported on the incidence of recurrent SAH.3, 4 The fi rst study followed 102 clipped patients over a mean follow-up period of 4.4 years (443 patient years of follow-up) and found a recurrence risk of 0.26% per year.4 The other study followed 220 clipped patients over a mean period of 9.9 years (∼2100 patient years of follow-up) and reported a 2.2% recurrence risk in the fi rst 10 years after the original treatment. 3 Both studies included rela-tively small numbers of patients, had no stringent inclusion criteria and included also patients who had been treated for unruptured aneurysms. Although the exact confi dence intervals of the incidence of a recurrent SAH in the articles are not reported, upon calculation they range somewhere between 0.006 and 1.26% for the one year estimate and between 1.0 and 6.2% for the 10 year estimate. The exact long-term risk of a recurrent SAH remains unclear.Members of families with familial intracranial aneurysms are also at risk for a SAH. Familial clustering (defi ned as more than two fi rst-degree relatives with SAH or unruptured aneu-rysms) occurs in around 10% of patients with SAH.8 The risk of SAH in relatives with familial SAH is higher than the risk of SAH in the general population and the outcome after SAH is probably worse.9

Nowadays screening for intracranial aneurysms can be performed relatively easily with screening techniques that are not or only minimally invasive such as CT-angiography (CTA) or MR- angiography (MRA). Screening for intracranial aneurysms is often recommended in familial SAH.10 Screening patients with a history of SAH is not generally performed but might be offered to patients more frequently when the knowledge on development of new aneurysms disseminates.The fi rst part of this thesis focuses on the long-term follow-up in patients who recover to an independent state after SAH. The second part outlines the effectiveness and consequences of screening for intracranial aneurysms in relatives with familial SAH and in patients with a history of SAH.

11

OnePart I: Long-term Follow-up after subarachnoid hemorrhage

In Chapter 2 we studied the outcome after a recurrent SAH and the characteristics of patients with a recurrence and in Chapter 3 we investigated the incidence of a recurrent episode of SAH in a large cohort of patients who had recovered from the initial episode and in whom the ruptured aneurysm had been successfully clipped.Because hypertension and smoking are risk factors shared between SAH and vascular diseases in general, patients with a history of SAH might not only be at risk for a recurrent SAH but also for other vascular events.11 The risk of recurrent SAH and other vascular events may result in an excess mortality in SAH survivors. Chapter 4 describes the standardized mortality ratio and risk of vascular events other than recurrent SAH in patients who recovered after SAH.Loss of smell (anosmia) is an important aspect of the long-term outcome after SAH because it affects the well-being of patients,12 but it is often neglected by physicians. Chapter 5 describes the frequency, impact and prognosis of anosmia after treatment for a ruptured aneurysm by clipping or coiling.In addition to anosmia, many patients experience psychosocial and cognitive defi cits in the fi rst year after SAH even if they made a good physical recovery and are functionally indepen-dent.13, 14 Because SAH mainly affects relatively young people, the psychosocial effects of SAH may infl uence daily functioning over many years. In Chapter 6 we studied the long-term ef-fects of SAH on employment, relationships, personality and mood in a large group of patients treated by clipping.

Part II: Screening for new aneurysms

Screening in familial subarachnoid hemorrhage

The yield of screening for aneurysms in persons with familial SAH is high; in about 8% an aneurysm is found.15, 16 Because aneurysms develop over time, repeated screening of these persons should be considered. In Chapter 7 we assessed the yield of repeated screening every fi ve years in familial SAH by investigating how many relatives were motivated for repeated screening, the number of newly detected aneurysms and the characteristics of the relatives with new aneurysms. In Chapter 8 we determined the psychosocial effects of familial scree-ning. We compared aspects of social well-being and lifestyle, health-related quality of life and mood symptoms in relatives with positive screens with those of relatives with normal screens, and those of persons in the general population.

12

Chap

ter

Screening in patients with a history of subarachnoid hemorrhage

In Chapter 9 we describe a decision analytic model. We evaluated based on the data available at that time in the literature whether follow-up screening with CTA is benefi cial in patients who have been successfully treated by surgical clipping or endovascular coiling after SAH.In the ASTRA (Aneurysm Screening after Treatment for Ruptured Aneurysms) study we screened 610 patients with CTA who had been admitted for SAH between 1985 and 2001. Chapter 10 reports the results of the ASTRA study on the incidence of and risk factors for new aneurysm formation and enlargement of already existing aneurysms. In Chapter 11 we assess the effec-tiveness and costs of screening for new aneurysms in patients with a history of SAH by using the results of the ASTRA study and data from the literature as input in a decision model.Screening may not only rule in or out the presence of an aneurysm, but may also detect very small aneurysms that are not treated. Such small aneurysms might be followed over time by CTA or MRA. Chapter 12 describes the yield of short-term serial follow-up by means of CTA or MRA for such small aneurysms detected at screening in persons with familial SAH and patients with a history of SAH. In Chapter 13 we describe the results of a meta-analysis to identify risk factors for rupture of intracranial aneurysms.

Finally, in Chapter 14 the main fi ndings of the thesis are summarized and the implications for patient care and future research are discussed.

13

OneReferences

1. Hop JW, Rinkel GJE, Algra A, van Gijn J. Case-fatality rates and functional outcome after subarachnoid hemorrhage: A systematic review. Stroke 1997;28:660-664.2. David CA, Vishteh AG, Spetzler RF, Lemole M, Lawton MT, Partovi S. Late angiographic follow- up review of surgically treated aneurysms. J Neurosurg 1999;91:396-401. 3. Tsutsumi K, Ueki K, Morita A, Usui M, Kirino T. Risk of aneurysm recurrence in patients with clipped cerebral aneurysms. Results of long-term follow-up angiography. Stroke 2001;32:1191- 1194.4. Juvela S, Poussa K, Porras M. Factors affecting formation and growth of intracranial aneurysms: A long-term follow-up study. Stroke 2001;32:485-491.5. Juvela S, Porras M, Heiskanen O. Natural history of unruptured intracranial aneurysms: A long-term follow-up study. J Neurosurg 1993;79:174-182.6. Rinkel GJE, Djibuti M, Algra A, van Gijn J. Prevalence and risk of rupture of intracranial aneurysms: A systematic review. Stroke 1998;29:251-256.7. Wiebers DO, Whisnant JP, Huston J, 3rd, Meissner I, Brown RD, Jr., Piepgras DG, Forbes GS, Thielen K, Nichols D, O’Fallon WM, Peacock J, Jaeger L, Kassell NF, Kongable-Beckman GL, Torner JC. Unruptured intracranial aneurysms: Natural history, clinical outcome, and risks of surgical and endovascular treatment. Lancet 2003;362:103-110.8. van Gijn J, Rinkel GJE. Subarachnoid haemorrhage: Diagnosis, causes and management. Brain 2001;124:249-278.9. Bromberg JE, Rinkel GJE, Algra A, Limburg M, van Gijn J. Outcome in familial subarachnoid hemorrhage. Stroke 1995;26:961-963.10. Bederson JB, Awad IA, Wiebers DO, Piepgras D, Haley EC, Jr., Brott T, Hademenos G, Chyatte D, Rosenwasser R, Caroselli C. Recommendations for the management of patients with unruptured intracranial aneurysms: A statement for healthcare professionals from the stroke council of the American Heart Association. Stroke 2000;31:2742-2750.11. Feigin VL, Rinkel GJE, Lawes CM, Algra A, Bennett DA, van Gijn J, Anderson CS. Risk factors for subarachnoid hemorrhage: An updated systematic review of epidemiological studies. Stroke 2005;36:2773-2780.12. Hummel T, Nordin S. Olfactory disorders and their consequences for quality of life. Acta Otolaryngol 2005;125:116-121.13. Hop JW, Rinkel GJE, Algra A, van Gijn J. Quality of life in patients and partners after aneurysmal subarachnoid hemorrhage. Stroke 1998;29:798-804.14. Hellawell DJ, Taylor R, Pentland B. Persisting symptoms and carers’ views of outcome after subarachnoid haemorrhage. Clin Rehabil 1999;13:333-340.15. Raaymakers TW, Rinkel GJE, Ramos LM. Initial and follow-up screening for aneurysms in families with familial subarachnoid hemorrhage. Neurology 1998;51:1125-1130.16. Ronkainen A, Hernesniemi J, Puranen M, Niemitukia L, Vanninen R, Ryynanen M, Kuivaniemi H, Tromp G. Familial intracranial aneurysms. Lancet 1997;349:380-384.

PartOne

Long-term Follow-up after Subarachnoid hemorrhage

Late recurrence of subarachnoid hemorrhageafter treatment for ruptured aneurysms:

patient characteristics and outcome

M.J.H. Wermer, G.J.E. Rinkel, P. Greebe, K.W. Albrecht, C.M. Dirven and C.A. Tulleken

ChapterTwo

18

Chap

ter

Abstract

Background and purpose

Patients with subarachnoid hemorrhage (SAH) who have been successfully treated for all detected aneurysms are at risk for a recurrent SAH. We assessed the characteristics, compli-cations of re-treatment, and the outcome of patients with a recurrent SAH because these factors are important if screening for new aneurysms is considered.

Methods

We studied patients admitted between 1987 and 2002 to three hospitals in the Netherlands with a new SAH after treatment of all aneurysms identifi ed after the fi rst SAH. We collected data on age, sex, risk factors, site and number of the aneurysm(s), time between the fi rst and the second SAH, complications of re-treatment and outcome after recurrent SAH.

Results

We identifi ed 30 patients; 27 women and three men. Thirty-two aneurysms were documented; 19 were classifi ed as de novo, eight as regrowth and fi ve had been missed in retrospect. The mean time between the fi rst and the second SAH was 7.8 years (range 0.25–17 years for all aneurysms and range 2.8–14 years for de novo aneurysms). Nine patients (30%) had a family history of SAH. No specifi c complications were reported with re-operation in 21 patients. Ten patients (33%) died, four (14%) were severely disabled and 16 patients (53%) had a good outcome.

Conclusions

Among patients admitted with a recurrent SAH, there is a predominance of women and of patients with familial SAH. Re-operation is not associated with specifi c complications. Outcome after recurrent SAH is similar to outcome after initial SAH.

19

Two

Introduction

In patients who have survived an episode of subarachnoid hemorrhage (SAH) the risk of a new SAH is higher than the risk of SAH in the general population, even if all detected aneu-rysms have been treated after the initial SAH. The risk of SAH recurrence is thought to be cumulative over time, being approximately 2% in the fi rst ten years after treatment and 9% after 20 years.5,10

Because of this increased risk, screening for new aneurysms in patients who have been sur-gically or endovascularly treated after a fi rst SAH may be worthwhile. Important factors in determining the effect associated with screening are the outcome after recurrent SAH and the complications associated with re-treatment. For example, the poorer the outcome after a recurrent SAH and the smaller the risk of complications of re-treatment, the more benefi cial screening will be. It would also be important to know which patients are at greatest risk for a new SAH, so that target screening could be targeted to patients who are likely to benefi t the most.We studied the patient characteristics, complications of re-treatment and the fi nal outcome of patients who had been admitted after recurrence of SAH.

Methods

We searched the records at three university hospitals for all patients admitted with aneurysmal SAH who had a previous aneurysmal SAH in their medical history. To identify these patients, we searched an existing database of all patients admitted with SAH between 1987 and 2002 at the University Medical Center Utrecht. In the other two hospitals we searched all discharge letters of patients diagnosed with aneurysmal SAH between 1987 and 1999 (Academic Medical Center Amsterdam) and 1996 and 2002 (VU Medical Center). Patients could also be included if the fi rst admission was in a hospital that was not one of the three study hospitals. Patients were included if they had been admitted for a recurrent SAH after successful treatment of all detected aneurysms at the initial SAH. In addition, patients were included who were not treated for the ruptured aneurysm at the initial SAH and had experienced recurrent SAH from another aneurysm. Patients were excluded if no aneurysm had been found at time of the initial SAH. The diagnosis of the fi rst and recurrent SAH had to have been confi rmed with CT scans, xanthochromia in the cerebrospinal fl uid or autopsy.We assessed clinical characteristics and outcome for all patients. For the fi rst SAH and the recurrent SAH we obtained information from medical records or discharge letters on age, sex, smoking habits, alcohol consumption, hypertension, medication, World Federation Neu-rological Sociaties (WFNS) score on admission, medical history, and family history. In case the patient had been admitted to another hospital than one of the three study hospitals for the

20

Chap

ter

fi rst SAH we retrieved a copy of the medical record from that hospital. We reviewed all available angiograms, CT and MR scans. If original fi lms could not be retrieved, we obtained information from the report of the radiologist. The presence of subdural, intrapa-renchymal or intraventricular blood on the CT scan was recorded. We determined the location and number of aneurysms at the fi rst SAH and at the recurrence. The size of the recurrent aneurysms was measured with (CT)-angiography. The newly documented aneurysms were classifi ed as de novo (aneurysm located at a site remote from the original aneurysm), regrowth (aneurysm located at the same site as the original treated one) or additional (aneurysm visible in retrospect but not identifi ed on the angiogram at the time of the initial SAH). If the patient died from the recurrence before angiography or CTA could be performed and no permission for autopsy was obtained, the aneurysm was classifi ed as unknown. Only if a review of the original fi lms revealed intraparenchymal extension of the hemorrhage, which was suggestive of the site of the aneurysm, aneurysms could be classifi ed as a regrowth or de novo aneurysm in such patients. Predictions of the ruptured aneurysm on the basis of intraparenchymal exten-sion of the SAH is correct in more than 90% of the cases.8,14 We reviewed all surgery reports for complications caused by treatment of the new aneurysm(s). Outcome was assessed with the Glasgow Outcome Scale on the basis of information from discharge letters.9

Results

Patient characteristics

We identifi ed 36 patients with a recurrent SAH. Six were excluded because no aneurysm was documented at the time of the initial SAH. Twenty-seven (90%) of the remaining 30 patients were women. Seventeen patients (57%) had a history of hypertension, and nine patients (30%) had one or more family members with SAH. The baseline characteristics are listed in Table 1. The mean age at the time of the fi rst SAH was 39 years (range 19-58) and 46 years (range 25-65) at the time of the second SAH. The mean time between the fi rst and the second SAH was 7.8 years on average (range three months – 17 years). The recurrent SAH that occurred three months after the initial SAH was caused by an additional aneurysm that had not been identifi ed during treatment of the initial SAH. In the group of de novo aneurysms, the interval between fi rst and second SAH ranged from 33 months to 14 years, with a mean of 7.8 years. Two patients experienced two recurrences, the mean time between the second and the third SAH was 12 years.

Initial SAH

During the fi rst episode of SAH, a single aneurysm was identifi ed in 21 patients, two aneu-rysms were documented in six patients, three aneurysms were identifi ed in two patients,

21

Two

and fi ve aneurysms were found in one patient. Forty-two of the 44 aneurysms in 28 patients were treated by means of clipping. In one patient, a ruptured aneurysm of the middle cere-bral artery (MCA) was wrapped because it was too small to be clipped. In another patient, a ruptured MCA aneurysm was left untreated because extensive secondary ischemia occurred. She was discharged in poor clinical condition but recovered to an independent state within one year. The decision not to operate was never reconsidered. Eleven years later she had a SAH from a newly developed aneurysm of the tip of the basilar artery.

Patients characteristics at recurrent SAH Total (%)N=30

De novo (%)N=17

Mean Age ± SD in years at 2nd episode 46 ± 11 48 ± 11

Sex WomenMen

27 (90%)3 (10%)

16 (94%)1 (6%)

Clinical condition on admission at 2nd episode Good (WFNS I-III)Poor (WFNS IV-V)Unknown

18 (60%)10 (33%)2 (7%)

10 (59%)6 (35%)1 (6%)

Smoking CurrentFormerNon-smokersUnknown

16 (53%)4 (13%)6 (21%)4 (13%)

10 (58%)1 (6%)3 (18%)3 (18%)

Alcohol Current heavy drinkers (≥ 5 drinks per day)Former heavy drinkers (≥ 5 drinks per day)Moderate (1-4 drinks per day)No alcohol useUnknown

02 (7%)9 (30%)11 (37%)8 (26%)

01 (6%)6 (35%)4 (24%)6 (35%)

History of hypertension YesNoUnknown

17 (57%)7 (23%)6 (20%)

11 (64%)3 (18%)3 (18%)

≥ 1 fi rst degree relative with SAH YesNoUnknown

9 (30%)12 (40%)9 (30%)

3 (18%)6 (35%)8 (47%)

Family history of polycystic kidney disease 2 1

Mean time between two episodes of SAH in yearsFrom all treated aneurysmsFrom regrowth aneurysmsFrom additional aneurysms

7.810.84.5

7.8

Table 1 Patients characteristics at the time of the recurrent subarachnoid hemorrhage

22

Chap

ter

Recurrent SAH

At admission for recurrent SAH, aneurysms were found with angiography, CTA or at autopsy in 26 of the 30 patients. One patient had three new aneurysms; the others had a single aneurysm. The other four patients died before CTA or angiography was performed and no permission for autopsy was obtained; in all four, intraparenchymal extension of the hemorrhage in one or both frontal lobes as revealed by CT scans suggested an aneurysm of the anterior communi-cating artery (AcomA). In three of these four patients, the initial ruptured aneurysm was at the AcomA. The recurrent aneurysms in these patients were therefore classifi ed as regrowth. In the other patient, the initial ruptured aneurysm was not located at the AcomA and the aneurysm was classifi ed as de novo. If these four aneurysms are classifi ed as unknown, the results remained essentially the same.All patients with recurrent SAH underwent angiography at the time of their fi rst SAH. In 26 patients, four vessel angiogram was performed; in two patients it was unknown if a complete four vessel angiogram was performed; and in two other patients angiography was performed incompletely. In the latter two patients, the new aneurysm was located at the vessel depicted on the initial angiogram.Overall, in 17 patients 19 aneurysms were classifi ed as de novo; in eight patients aneurysms were classifi ed as regrowth; and in fi ve patients as additional. In four of the 17 patients with de novo aneurysms, an aneurysm of the middle cerebral artery developed at the mirror site of the fi rst SAH. In three other patients with a de novo aneurysm, the new aneurysm was located on branches of the same parent artery. In the two patients with a third SAH, three de novo aneurysms were found. Figure 1 shows the scans of one of the patients with a de novo aneurysm. The location of the aneurysms at the fi rst and recurrent episodes is listed in Table 2. The size of the recurrent aneurysms is shown in Table 3.CT scans were performed in all 30 patients after recurrent SAH. Intraparenchymal extension was found in eight patients, intraventricular extension in 11 patients and subdural hematoma in four patients.

Complications related to re-treatment

Of the 30 patients, 21 patients with 21 aneurysms underwent re-treatment. In three patients with a regrowth aneurysm the aneurysm was successfully clipped. No complications occur-red, although in all patients re-exposure of the aneurysm was more diffi cult owing to the presence of scar tissue, adhesions, or the clip from the previous surgery. In another patient, the regrowth aneurysm was located at the top of the basilar artery. In this patient, an extra-cranial-intracranial bypass was constructed to the posterior cerebral artery to allow coiling as this artery emerged from the aneurysm; but the bypass did not remain patent. Therefore the aneurysm was left untreated.

23

Two

Images obtained in a 58-year-old woman who was admitted with a subarachnoid hemorrhage. Four-vessel angiography (upper row) showed an aneurysm of the left posterior inferior cerebellar artery (which had ruptured), an additional aneurysm of the left middle cerebral artery, but no other aneurysms. Both aneurysms were clipped. Three years later, she was readmitted with a new SAH. The pattern of the hemorrhage on CT (lower row) was highly suggestive for an aneurysm of the right middle cerebral artery, which was confi rmed at autopsy. This aneurysm could not be identifi ed on the angiogram retrospectively (upper row, right panel).

Figure 1

Ten de novo aneurysms and three additional aneurysms were treated by means of surgical clipping. Three de novo aneurysms and one additional aneurysm were treated with coiling. No complications occurred during these 17 procedures, although in one patient, a second coiling procedure was necessary to achieve complete occlusion of the aneurysm.

Outcome

Ten (33%) of the 30 patients died as a result of the second SAH. Nine patients died without operation or coiling of the aneurysm; seven patients died from the initial impact of the hemorrhage and two patients after rebleed. One patient with a surgically treated de novo aneurysm had multiple epidural and subdural hematomas, remained in a vegetative state, and died four months after the recurrent SAH.

24

Chap

ter

Four patients (14%) remained severely disabled after the recurrent SAH (GOS 3) and were discharged to nursing homes. The poor outcome in these patients was attributable mainly to secondary ischemia. Three of these four patients had made a good recovery after the fi rst SAH; the fourth had remained moderately disabled (GOS 4) after the fi rst SAH. The remaining 16 (53%) of the 30 patients were discharged home. Four of these patients had

Type and location aneurysms First SAH Second SAH Third SAH

Site of ruptured aneurysm Carotid / posterior communicating arteryMiddle cerebral arteryAnterior Communicating artery / PericallosaPosterior circulation

106113

7689

---2

Site of unruptured aneurysmsCarotid / posterior communicating arteryMiddle cerebral arteryAnterior Communicating artery / PericallosaPosterior circulation

48-2

-2--

---1

Total number of aneurysms per patients1235unknown

21621-

25-1-4

11--

Treatment of aneurysmClippingCoilingBypassMuscle patchNo treatment

42--11

1641-11

-3---

Type of aneurysm at the new episodeRegrowthDe novoAdditional

---

8195

-3-

Table 2 Aneurysm characteristics

Size (mm) De novo Regrowth Additional< 5 6 1 25-10 4 3 2> 10 3 1 0Unknown* 4 3 1

Table 3 Size of the recurrent aneurysms

* in six of the patients a (CT)-angiogram was not performed and for two patients the images could not be retrieved.

25

Two

been moderately disabled after the fi rst episode (GOS 4) and their handicap was unchanged after the recurrent SAH. Twelve again made a good recovery (GOS 5) although four experien-ced slight cognitive changes. Both patients with three separate episodes had a good recovery after each of them.

Discussion

Patient characteristics

We found that patients with a late recurrent SAH differ in several ways from the general po-pulation of patients with SAH. The patients with a recurrence were predominantly women. This fi nding is in accordance with a recent article that reports an increased risk for aneurysm formation in women after SAH.12 The mean age of our patients at the fi rst SAH was only 39 years. In general SAH occurs most often in patients between 40 and 60 years of age, with a mean of approximately 55 years.1 A familial preponderance occurred more often in patients with a recurrent SAH (30%) than in a series of patients with a fi rst SAH from the same hos-pitals (9%).3 In our series, more than half of the patients (57%) had a history of hypertension, which is also more frequent than previously reported for the general population of patients with SAH.11 The proportion of patients with hypertension and a positive family history in our study is an underestimation, as these risk factors were not known in almost a third of the patients. It has been suggested that the presence of multiple aneurysms is a risk factor for the development of new aneurysms.5 The proportion of patients with multiple aneurysms at fi rst admission did not exceed that usually found in patients with SAH (30%).10

We retrospectively identifi ed all patients with SAH who had a history of SAH, including pa-tients that had been admitted for this fi rst episode to another hospital. Because there is no registry of numbers of patients treated for SAH in these other hospitals, we do not know the total number of patients at risk for a recurrent SAH. In addition, we have no data regarding patients who had been admitted to one of the participating centers and had a recurrent epi-sode afterward from which they died before reaching the hospital or for which they had been admitted to another hospital. Therefore we can not calculate an actual risk of recurrence from our series of patients. In addition, not all information on the characteristics of the patients was complete, so we did not calculate relative risks. Nevertheless, our results suggest an increased risk for recurrent SAH in women and in patients with a family history of SAH.Acute subdural hematoma is thought to be associated with recurrent SAH but also may occur with the initial hemorrhage.16 In a series of 484 patients with fi rst SAH, subdural hematoma was found in 15 patients (3%).13 In our series, subdural hemorrhage was found in four of 30 patients (13%). No other CT characteristics were related to recurrent SAH.

26

Chap

ter

Complications related to re-treatment and outcome

The neurosurgeons determined re-operation technically more challenging than surgery after a fi rst SAH because of surrounding scar tissue and adhesions, and, for regrowth aneurysms the presence of the previously applied clip. Nonetheless, during the re-operations no complications occurred that could be attributed to the previous operation or SAH. In other studies re-opera-tion for incomplete clipped or residual aneurysms also resulted in successful obliteration of 72 to 89% of the aneurysms and a good functional outcome in 84% of the patients.2,6,7

After a second episode of SAH, half the patients died or were left disabled. This outcome is comparable to that in a series of patients with a fi rst SAH during the same study period in the same hospitals.19

Literature on recurrent SAH

We found three Japanese studies on patients with a late recurrence of SAH.17,20,22 Our results were consistent with those reported in these studies for the overall outcome after a recurrence. In one of the Japanese studies, the location of the regrowth aneurysms was at the posterior communicating artery (PcomA) in eight of the nine patients.17 In most of these patients, the PcomA was the parent vessel of the posterior cerebral artery. In our study and in the other two Japanese studies, the new aneurysms were more equally distributed among the com-mon sites for aneurysms. We found no recurrent SAH from de novo or regrowth aneurysms within the fi rst 33 months after operation. This lower limit of approximately three years corresponds with that reported in the other studies. Apparently, time is required for a new aneurysm to develop. This suggests that if screening for new aneurysms after clipping of all detected aneurysms is to be useful, it should not start earlier than 2.5 years after treatment. The predominance of women in our study was not found in the Japanese studies; this might be explained by genetic differences in the study population; in Japan the incidence of SAH is also much higher than in most other parts of the world.15

Pathogenesis of recurrent SAH

A variety of factors determines the development of new aneurysms and new episodes of SAH. In our study, eight episodes of recurrent SAH were caused by regrowth aneurysms. The development of a regrowth aneurysm might be a consequence of incomplete treatment or slippage of the clip after the initial operation. Slippage of the clip was found at re-operation in only one of our eight patients. In two of our eight patients post-operative angiograms had been performed and showed no residual fi lling of the aneurysm. In other words, regrowth aneurysms can develop despite apparently complete obliteration after the fi rst SAH. In the six patients in whom no post-operative angiogram was performed, the neurosurgeon reported complete obliteration of the aneurysm in the surgery report.

27

Two

We cannot exclude the possibility that in some of these patients a small remnant was pre-sent after surgery that enlarged over time and caused the recurrence. Additional aneurysms missed at the time of the fi rst SAH because of the limited sensitivity of angiography are another cause of recurrent SAH. In our study, fi ve patients had an aneurysm that was observed in retrospect on the initial angiogram. In the interval between the fi rst SAH and the recur-rence, three of these fi ve aneurysms had increased in size.The high number of new detected aneurysms located at the posterior circulation in our series cannot be explained by incomplete angiograms after the fi rst SAH, as in eight of these nine patients four vessel angiography had been performed at the fi rst admission. Most of the recurrent episodes of SAH in our study were caused by de novo aneurysms. This predominance of de novo aneurysms as cause for recurrent episodes of SAH indicates that despite successful treatment of the ruptured aneurysm, the underlying vascular condition for the development of aneurysms remains. Aneurysms are usually not present at birth, but develop during life.18 Patients who have had SAH remain prone to new aneurysm formation despite treatment for all detected aneurysms after the fi rst SAH.

Screening

A few studies have reported the magnitude of the risk of recurrence of SAH after surgical treatment of aneurysms. The observed risk is approximately 2% in 10 years and 9% in a period of 20 years.5,20 The risk of SAH in the general population in 10 years is approximately 0.072%.15 This suggests that patients after SAH have a 30 times increased risk for a new episode. If knowledge regarding this increased risk disseminates, screening may be offered routinely to patients who have had SAH. Screening might decrease the rate of recurrent SAH, but it could also lead to an increase in morbidity or mortality from complications of angiography and preventive treatment.4 Because aneurysms develop over time, frequent screening might be necessary, repeatedly reminding patients of the traumatic episode involving their SAH. Also, small aneurysms might be detected for which treatment is not necessary or not possible. The knowledge of having an untreated aneurysm decreases the quality of life.21 Until now no screening program has been properly assessed for its pros and cons. In our opinion screening should not be implemented before its frequency, benefi ts and risks have been properly as-sessed in a clinical study.

28

Chap

ter

References

1. The ACROSS study group. Epidemiology of aneurysmal subarachnoid hemorrhage in Australia and New Zealand: incidence and case fatality from the Australian Cooperative Research on Subarachnoid Hemorrhage Study (ACROSS). Stroke 2000, 31:1843-18502. Boet R, Poon WS, Yu SC: The management of residual and recurrent intracranial aneurysms after previous endovascular or surgical treatment - a report of eighteen cases. ActaNeurochir (Wien) 2001, 143:1093-1101.3. Bromberg JE, Rinkel GJE, Algra A, Greebe P, van Duyn CM, Hasan D, Limburg M, ter Berg HW, Wijdicks EF, van Gijn J: Subarachnoid haemorrhage in first and second degree relatives of patients with subarachnoid haemorrhage. BMJ 1995, 311:288-289.4. Cloft HJ, Joseph GJ, Dion JE: Risk of cerebral angiography in patients with subarachnoid hemorrhage, cerebral aneurysm, and arteriovenous malformation: a meta-analysis. Stroke 1999, 30:317-320.5. David CA, Vishteh AG, Spetzler RF, Lemole M, Lawton MT, Partovi S: Late angiographic follow-up review of surgically treated aneurysms. J Neurosurg 1999, 91:396-401.6. Drake CG, Friedman AH, Peerless SJ: Failed aneurysm surgery. Reoperation in 115 cases. J Neurosurg 1984, 61:848-856.7. Giannotta SL, Litofsky NS: Reoperative management of intracranial aneurysms. J Neurosurg 1995, 83:387- 393.8. Hillman J: Selective angiography for early aneurysm detection in acute subarachnoid haemorrhage. Acta Neurochir 1993, 121:20-25.9. Jennett B, Bond M: Assessment of outcome after severe brain damage. Lancet 1975, 1:480-484.10. Juvela S: Risk factors for multiple intracranial aneurysms. Stroke 2000, 31:392-397.11. Juvela S, Hillbom M, Numminen H, Koskinen P: Cigarette smoking and alcohol consumption as risk factors for aneurysmal subarachnoid hemorrhage. Stroke 1993, 24:639-646.12. Juvela S, Poussa K, Porras M: Factors affecting formation and growth of intracranialaneurysms: a long-term follow-up study. Stroke 2001, 32:485-491.13. Kamiya K, Inagawa T, Yamamoto M, Monden S: Subdural hematoma due to ruptured intra-cranial aneurysm. Neurol Med Chir 1991, 31:82-86.14. Laissy JP, Normand G, Monroc M, Duchateau C, Alibert F, Thiebot J: Spontaneous intrace-rebral hematomas from vascular causes. Predictive value of CT compared with angiography. Neuroradiology 1991, 33:291-295.15. Linn FH, Rinkel GJE, Algra A, van Gijn J: Incidence of subarachnoid hemorrhage: role of region, year, and rate of computed tomography: a meta-analysis. Stroke 1996, 27:625-629.16. Linn FH, Rinkel GJE, Algra A, van Gijn J: The notion of “warning leaks” in subarachnoid haemorrhage: are such patients in fact admitted with a rebleed? J Neurol Neurosurg Psychiatry 2000, 68:332-336.17. Nakase H, Kamada Y, Aoki H, Goda K, Morimoto T, Sakaki T: Clinical study on recurrent intra- cranial aneurysms. Cerebrovasc Dis 2000, 10:255-260.18. Rinkel GJE, Djibuti M, van Gijn J: Prevalence and risk of rupture of intracranial aneurysms: a systematic review. Stroke 1998, 29:251-256.19. Roos EJ, Rinkel GJE, Velthuis BK, Algra A: The relation between aneurysm size and outcome in patients with subarachnoid hemorrhage. Neurology 2000, 54:2334-2336.20. Tsutsumi K, Ueki K, Usui M, Kwak S, Kirino T: Risk of recurrent subarachnoid hemorrhage after complete obliteration of cerebral aneurysms. Stroke 1998, 29:2511-2513.21. van der Schaaf IC, Brilstra EH, Rinkel GJE, Bossuyt PM, van Gijn J: Quality of life, anxiety, and depression in patients with an untreated intracranial aneurysm or arteriovenous malfor-mation. Stroke 2002, 33:440- 443.

29

Two

22. Yamakawa H, Sakai N, Takenaka K, Yoshimura S, Andoh T, Yamada H, Ohkuma A, Takada M, Funakoshi T: Clinical analysis of recurrent subarachnoid hemorrhage after neck clipping surgery. Neurol Med Chir 1997, 37:380-385.

Incidence of recurrent subarachnoid hemorrhage after clipping for ruptured intracranial aneurysms

M.J.H. Wermer, P. Greebe, A. Algra, and G.J.E. Rinkel

ChapterThree

32

Chap

ter

AbstractAbstract

Background and PurposeBackground and Purpose

Because intracranial aneurysms develop during life, patients with subarachnoid hemorrhage (SAH) and successfully occluded aneurysms are at risk for a recurrence. We studied the inci-dence of and risk factors for recurrent SAH in patients who regained independence after SAH and in whom all aneurysms were occluded by means of clipping.

MethodsMethods

From a cohort of patients with SAH admitted between 1985 and 2001, we included those patients who were discharged home or to a rehabilitation facility. We interviewed these pa-tients about new episodes of SAH. We retrieved all medical records and radiographs in case of reported recurrences. If patients had died we retrieved the cause of death. We analyzed the incidence of and risk factors for recurrent SAH by Kaplan-Meier curves and Cox regres-sion analysis.

ResultsResults

Of 752 patients with 6016 follow-up years (mean follow-up 8.0 years), 18 had a recurrence. In the fi rst 10 years after the initial SAH the cumulative incidence of recurrent SAH was 3.2% (95% CI 1.5-4.9%) and the incidence rate 286/100.000 patient-years (95% CI 160-472/100.000). Risk factors were smoking (HR 6.5; 95% CI 1.7-24.0), age (HR 0.5 per 10 years; 95% CI 0.3-0.8) and multiple aneurysms at the time of the initial SAH (HR 5.5; 95% CI 2.2-14.1).

ConclusionsConclusions

After SAH, the incidence of a recurrence within the fi rst 10 years is 22 (12-38) times higher than expected in populations with comparable age and sex. Whether this increased risk justifi es screening for recurrent aneurysms in patients with a history of SAH requires further study.

33

Three

IntroductionIntroduction

The prevalence of intracranial aneurysms under the age of 20 is very low and increases thereafter.1 This suggests that aneurysms are not congenital but develop during life. The development of aneurysms during life and the presence of multiple aneurysms in up to 30% of patients with subarachnoid hemorrhage (SAH) indicate that patients who have had an aneurysm are at risk for developing new aneurysms.2

In patients with SAH and clipping of all detected aneurysms, new aneurysms can develop at a new location (de novo) or at the clip site (regrowth).3-7 In addition, some aneurysms might have been missed at the angiogram at the time of the SAH. The International Study of Unruptured Intracranial Aneurysms (ISUIA) showed that the rupture risk of small unruptured aneurysms in patients with a previous SAH from another aneurysm is higher than in patients with similar aneurysms but without a history of SAH.8 The ongoing development of aneurysms and higher rupture risks after a previous episode of SAH suggest that patients with a history of SAH are at considerable risk of a new SAH.We performed a long-term follow-up study in a large cohort of patients who regained indepen-dence after SAH and in whom all detected aneurysms had been occluded by means of clipping. We determined the incidence of and the risk factors for a recurrent episode of SAH.

MethodsMethods

PatientsPatients

From a database of patients admitted to the University Medical Center Utrecht with SAH, we selected all patients admitted between 1985 and 2001 who met the following inclusion criteria: 1) subarachnoid hemorrhage confi rmed by CT or lumbar puncture; 2) presence of a saccular aneurysm confi rmed by conventional angiography or CT- angiography; 3) clipping of the ruptured aneurysm and all additional aneurysms; 4) age at time of SAH ≥ 18 years and 5) discharge to home or a rehabilitation facility. Patients who were discharged to a nursing home and patients with one or more aneurysms left untreated were not included.

Follow-upFollow-up

After approval of the ethical committee of our hospital, we contacted the general practitioner of all eligible patients to know if the patient was still alive. If patients had died, we asked for the date and the cause of death. If a patient had died in a hospital or other facility, we reviewed the medical records. Subsequently we sent a letter to all patients who were still alive. In this letter, patients 70 years of age or younger were invited to the outpatient clinic.

34

Chap

ter

For patients older than 70 years, we announced a telephone call. If a patient had no phone number or an exdirectory one, we sent another letter asking the patient to contact us. At the outpatient clinic or during the telephone interview, we asked the patients about new episodes of SAH. For all patients, we retrieved data on sex, age, family history of intracranial aneurysms, smoking habits, alcohol use, history of hypertension and number of aneurysms at time or the initial SAH.

Recurrent SAHRecurrent SAH

For patients who reported new episodes of SAH, we retrieved all medical data from the hospital where patients had been treated and reviewed all CT scans of the brain and all angiograms of the intracranial vessels of the initial episode if available and of the recurrence. New episodes of SAH were defi ned as SAH proven by CT, lumbar puncture or autopsy after treatment of all aneurysms that had been found at the time of the initial SAH. On basis of surgery reports, CT scans and angiograms, autopsy reports or – if patients died before (CT)-angiography could be performed – by a pattern of hemorrhage on CT highly suggestive for the site of rupture, we categorized patients into: A) aneurysm located at a site remote from the original clipped aneurysm, B) aneurysm located at the same site as the original clipped aneurysm or as C) unclassifi ed if patients died before (CT)-angiography could be performed and the site of rup-ture could not be derived from the hemorrhage pattern. The aneurysms at other sites than the clip were subdivided in 1. de novo (aneurysm not visible on the initial (CT)-angiogram at the time of the SAH); 2. additional (aneurysm visible in retrospect but not identifi ed on the initial (CT)-angiogram); or 3. possible de novo (when the (CT)-angiogram at time of SAH was not available for review, but the aneurysm was not described in the radiology report). The aneurysms located at the clip were classifi ed as 1. regrowth (post-operative angiogram showed complete clipping of the aneurysm); 2. remnant (post-operative angiogram showed incomplete clipping of the aneurysm) or 3. possible regrowth (when no post-operative angi-ogram was performed). Patients with a history suggestive of aneurysmal rupture but without confi rmation of the diagnosis because they had died before reaching hospital were classifi ed as sudden death probably due to recurrent SAH. A history was classifi ed as suggestive for aneurysmal rupture if we had an eyewitness account of a sudden severe headache before onset of coma. Patients with sudden loss of consciousness not preceeded by acute headache, or patients who were found dead were not classifi ed as suggestive for aneurysmal rupture.

35

Three

Data analysisData analysis

The risk for recurrent SAH was assessed by survival analysis. We calculated the incidence rate per 100.000 patient-years and the cumulative incidence at 5, 10 and 15 years with cor-responding 95% confi dence intervals (CIs) for recurrent SAH and sudden death probable due to recurrent SAH combined and for recurrent SAH separately. If patients had died during the follow-up period, were lost to follow-up, or were admitted to a nursing home, they were censored at that point of time. We calculated incidence rate ratios with corresponding 95% CIs by dividing the incidence rate in our population by a population based incidence rate of SAH. This population incidence rate was derived from a population in Australia and New Zealand with comparable age (45 to 54 years).9 Adjustments were made for differences in sex distribution.We used Cox regression to calculate hazard ratios (HR) and corresponding 95% CIs of baseline characteristics associated with the risk of recurrent SAH. Again, we performed the analyses for recurrent SAH and sudden death probable due to recurrent SAH combined and for recur-rent SAH separately. The following factors were included in the analyses: age at time of SAH (categorical per 10 years), sex, smoking (dichotomous current versus former plus never smokers, and categorical), family history of intracranial aneurysms (defi ned as one or more fi rst-degree relative(s) with a verifi ed aneurysm or a history suggestive of SAH, history of hypertension (dichotomous) and number of aneurysms at the time of SAH (dichotomous: one versus mul-tiple). We performed univariate Cox regression analysis for all risk factors and multivariate Cox regression analysis with forward selection of variables with probability values < 0.20 in the univariate analysis.

ResultsResults

PatientsPatients

Between 1985 and 2001, 930 patients survived after SAH and were successfully treated for the ruptured aneurysm. One hundred fi fty-four did not meet the inclusion criteria (Figure 1). Of the 776 remaining patients, 24 (3%) could not be contacted. The total follow-up of the 752 patients included in the study was 6016 years with a mean period of follow up of 8.0 years (range 0.2-20.1 years). During the follow-up period, 107 patients had died. In 10 patients, the cause of death was a recurrent SAH. Seven patients died suddenly before reaching the hospital, of whom two had a history suggestive of SAH. Other causes of death were cancer (21), cardio-vascular disease (19), infectious diseases (9), other causes, including dementia, trauma, renal failure (14), or unknown (27). In addition, during follow-up, 23 patients were censored because they moved abroad (6) or were admitted to a nursing home (17). The baseline characteristics of the 752 patients are listed in Table 1.

36

Chap

ter

Figure 1 Flow diagram of patients

18

37

Three

Recurrent SAHRecurrent SAH

A new episode of SAH had occurred in 18 (2.4%; 95% CI 1.5-3.8%) of the 752 patients during the follow-up period. The incidence rate in our patients for the fi rst 10 years after SAH was 286/100.000 (95% CI 160-472/100.000). In the population based study, the incidence rate derived from a population in Australia and New Zealand with a comparable age and sex distri-bution was 13/100.000. The risk of a subarachnoid hemorrhage in our population was therefore 22 (95% CI 12-38) times higher compared with the population based incidence (Table 2).9 The cumulative incidence of new episodes in the fi rst 10 years after the index SAH was 3.2% (95% CI 1.5-4.9%) for all certain recurrences (Figure 2 and Table 2). When the two patients

Characteristics Patients N=752 Re-SAH N=18

Mean age in years ± SD (range) 50.3 ± 12.3 (21-83) 42.0 ± 9.5 (20-61)

Women 501 (67%) 15 (83%)

Mean follow-up in years (range) 8.0 (0.2-20.1) 6.5 (0.2-17)

Smoking Current smokerFormerNever smokerUnknown

248 (33%)206 (27%)140 (19%)158 (21%)

10 (61%)1 (5%)2 (11%)5 (28%)

Alcohol use Alcohol use > 5 drinks a dayFormer use > 5 drinks a dayAlcohol use 1-5 drinks a dayNeverUnknown

29 (4%)16 (2%)344 (46%)159 (21%)204 (27%)

--7 (39%)6 (33%)5 (28%)

History of hypertension YesNoUnknown

246 (33%)471 (62%)35 (5%)

7 (39%)11 (61%)-

Familial SAH Medically verifi ed family historyPossible, not medically verifi edNo family historyUnknown

40 (5%)61 (8%)529 (70%)122 (17%)

3 (17%)1 (5%)11 (61%)3 (17%)

No of aneurysms at time of SAH 1234

650 (86%)71 (9%)27 (4%)4 (1%)

10 (56%)6 (33%)2 (11%)-

Table 1 Baseline characteristics of the 752 included patients and the 18 patients with a recurrence

Re-SAH = Recurrent subarachnoid hemorrhage

38

Chap

ter

Time afterinitial SAH

Re-SAH(95%CI)

ReSD-SAH(95%CI)

IR study groupper 100.000 py(95%CI)

IR populationper 100.000 py(95%CI)

Incidence rateratio (95% CI)

5 years 1.0% (0.3-1.8%) 1.2% (0.4-2.1%) 207 (83-426) 13 (10-18) 16 (7-34)

10 years 3.2% (1.5-4.9%) 3.5% (1.8-5.3%) 286 (160-472) 13 (10-18) 22 (12-38)

15 years 4.6% (1.9-7.2%) 4.9% (2.2-7.6%) 287 (167-460) 13 (10-18) 22 (13-38)

Table 2 Cumulative incidence and incidence rates of recurrent SAH 5, 10, and 15 years after the ini-tial subarachnoid hemorrhage for recurrences of SAH and recurrences of SAH including patients with sudden death probable due to recurrent SAH compared with the population risk of SAH

Re-SAH = Cumulative incidence of recurrent subarachnoid hemorrhageReSD-SAH = Cumulative incidence of recurrent subarachnoid hemorrhage including sudden death probable due to SAHIR study group = Incidence rate in the study populationIR population = Incidence rate derived from a population in Australia and New Zealand with a comparable age and sex distributionIncidence rate ratio = the incidence rate observed in the study population divided by the inci-dence rate observed in the population based study py = patient years

Figure 2 Kaplan Meier curve of the cumulative incidence of patients with recurrent SAH

752 519 249 58 number of patients follow-up in years

39

Three

with sudden death probable due to SAH were included the cumulative incidence was 3.5% (95% CI 1.8-5.3%).In the 18 patients with a recurrent SAH after successful treatment, 19 recurrent aneurysms were found: 13 patients had an aneurysm located at a new location, three patients had an aneurysm located at the clip site from the previous operation, one patient had both an aneu-rysm located at a new location and an aneurysm located at the clip site, and one patients had an unclassifi ed aneurysm. Of the 14 patients with aneurysms at a new location, the CTA or angiogram was available for review in eight patients: four aneurysms were certain de novo aneurysms and four were classifi ed as additional. The remaining six aneurysms were clas-sifi ed as possible de novo aneurysms. Of the four patients with an aneurysm at the clip site all were classifi ed as possible regrowths.The mean interval between the initial SAH and the recurrence was 6.5 years (range 0.2-17 years). We found no recurrent SAH from a (possible) de novo or regrowth aneurysm within the fi rst 33 months after operation.

Risk factors for recurrent SAHRisk factors for recurrent SAH

In the univariate Cox regression analysis for recurrent SAH signifi cant risk factors were age (HR 0.6 per 10 years; 95% CI 0.4-0.9), a verifi ed history of familial SAH (HR 3.8; 95% CI 1.1-13.2), current smoking (HR 4.8; 95% CI 1.3-17.4) and the presence of multiple aneurysms at time of the SAH (HR 5.7; 95% CI 2.3-14.5). Gender (HR 0.4 for men; 95% CI 0.1-1.3) was not a statistically signifi cant risk factor but had a p value < 0.2 and was therefore also included in the multivariate analysis. In the multivariate forward Cox regression analyses, current smoking (HR 6.5; 95% CI 1.7-24.0), age (HR 0.5 per 10 years; 95% CI 0.3-0.8) and the presence of multiple aneurysms at time of the SAH (HR 5.5; 95% CI 2.2-14.1) were statistically signifi cant risk factors. In the Cox regression analysis for recurrent SAH, including sudden death probable resulting from SAH, the same risk factors were detected in both the univariate as the multivariate analysis with comparable hazard ratios in the multivariate analysis.

DiscussionDiscussion

We found that in patients who had recovered to an independent state after an episode of aneurysmal SAH and in whom all detected aneurysms were treated by means of clipping, the risk of a recurrence the fi rst 10 years after treatment is 22 times higher than the risk of SAH in a healthy cohort with comparable age and sex. Independent risk factors for subsequent episodes of SAH were current smoking, young age and multiple aneurysms at the time of the initial SAH.

40

Chap

ter

We collected more than 6000 patient years of follow-up with a mean follow-up time of eight years. Two previous studies with much smaller numbers of patients and less stringent inclusion criteria than in our study have reported on the incidence of recurrent SAH.4, 10 These studies found a slightly lower risk of a recurrent SAH in patients treated after SAH. The fi rst study followed 220 clipped patients who survived > 3 years over a mean period of 9.9 years (~2100 patient years of follow-up).10 Six patients had a recurrent SAH and the cumulative incidence of recurrences calculated in that study was 2.2% at 10 years after the original treatment. The other study followed 102 surgically treated patients over a mean follow-up period of 4.4 years (443 patient-years of follow-up).4 Only one recurrent SAH was reported and therefore the overall risk of recurrent SAH was 0.26% per year. We found that the risk of a recurrent SAH is small in the fi rst few years after treatment and increases thereafter.11 The risk for recurrent SAH of 0.26% per year during a follow-up period of 4.4 years can therefore not be extrapola-ted to longer follow-up times. Moreover, more than one-third of the patients included in this study did not have an SAH as initial presentation but had an incidentally detected aneurysm. The ISUIA study showed that the rupture risk of incidentally detected aneurysms in patients without a history of SAH is lower than in patients with additional aneurysms and a history of SAH. Furthermore, loss to follow-up and selection bias occurred more often in this study than in our study. Only patients who underwent late follow-up angiography were included (<10% of the treated population) and therefore patients who died of a recurrent SAH before follow-up could be performed were missed. Risk factors for recurrent SAH were not evaluated in the two studies.We compared the incidence of recurrent SAH in our study with a large population based study on the incidence of SAH in Australia and New Zealand.9 We chose this study as comparison because it is a recent study with over 1.7 million patient-years follow-up and data are reported for men and women separately in different age groups. The SAH incidence in this study was comparable with an incidence study performed between 1978 and 1980 in the Netherlands and with the incidence of SAH found in a meta-analysis of 15 studies from populations other than Finland and Japan.12,13 In Finland and Japan the SAH incidences are higher than in other countries (around 20/100.000 patient-years).13-15 In our study, the incidence of recurrent SAH was 286/100.000. Thus, even if we compare our incidence of recurrences with relatively high incidences of SAH as in Finland or Japan, the difference between the incidence in the popu-lation and the incidence of recurrent SAH is impressive and is included within the limits of the confi dence interval of our estimate (incidence rate ratio 22; 95% CI 12-38).There are some limitations of our study that need to be acknowledged. Although the num-ber of patients lost to follow-up in our study was very small (3%), we did not have follow-up for all patients. Furthermore, in 27 patients, the exact cause of death was unknown, and an additional fi ve patients died suddenly of unknown cause. If some or all of these patients would have had a subarachnoid hemorrhage, the actual recurrence rate is higher than the one observed in our study. Conversely, if none of the patients who were lost to follow-up had

41

Three

a recurrence, we have slightly overestimated the rate of recurrent SAH by not including the follow-up years of these patients.In four patients, the recurrence was caused by an aneurysm at the clip-site. Unfortunately, a post-operative angiogram was not performed in these patients. In some patients the aneurysm might have been incompletely clipped. Even after surgery by experienced neurosurgeons, neck remnants are found in nearly 10% of the cases.16 We therefore cannot exclude that some patients had small remnants of the initial ruptured aneurysm.We excluded patients who were discharged to a nursing home because consent for follow-up in these patients is in general hard to obtain because most of these patients have severe cognitive defi cits. Moreover, from a clinical point of view, the risk of recurrence is less relevant for such patients. Our results therefore apply only to patients discharged home or to a reha-bilitation center which will in general be a younger population than the patients discharged to a nursing home. Whether our results can be generalized to coiled patients is unclear. One can assume that the formation rate and rupture risk of de novo aneurysms in coiled patients is comparable with that in clipped patients. The number of regrowth aneurysms or remnants caused by impaction of coils, however, may be higher than by clipping and therefore these patients are possibly even at a higher risk of a recurrence.The high risk of a recurrent SAH indicates that having an SAH is not a single event in a lifetime. New aneurysms have previously been found in up to 16% of the patients with a history of SAH.3-7 Screening for aneurysms in these patients might be effective, but screening also has disadvantages. Screening carries the risks of angiography and preventive treatment, especially in older patients. In addition not all aneurysms will be treated, and the knowledge of having an untreated aneurysm has a negative effect on the quality of life of these patients.17 Moreover, screening also has psychosocial consequences.18 Thus, whether a regular screening program for detection of new aneurysms is benefi cial for patients who survived after treatment for SAH requires careful weighing of the pros and cons, for example in a decision analysis.

42

Chap

ter

ReferencesReferences

1. Rinkel GJE, Djibuti M, Algra A, van Gijn J. Prevalence and risk of rupture of intracranialaneurysms: a systematic review. Stroke 1998; 29:251-6.2. Juvela S. Risk factors for multiple intracranial aneurysms. Stroke 2000; 31:392-7.3. Wermer MJH, Schaaf van der IC, Velthuis BK, Algra A, Buskens E, Rinkel GJE, for the ASTRA study group. Follow-up screening after subarachnoid hemorrhage: frequency of detected aneurysms and risk factors for aneurysm formation and growth. Brain 2005; 128:2421-2429.4. David CA, Vishteh AG, Spetzler RF, Lemole M, Lawton MT, Partovi S. Late angiographic follow-up review of surgically treated aneurysms. J Neurosurg 1999; 91:396-401.5. Tsutsumi K, Ueki K, Morita A, Usui M, Kirino T. Risk of aneurysm recurrence in patients with clipped cerebral aneurysms. results of long-term follow-up angiography. Stroke 2001; 32:1191-4.6. Juvela S, Porras M, Heiskanen O. Natural history of unruptured intracranial aneurysms: a long-term follow-up study. J Neurosurg 1993; 79:174-82.7. Juvela S, Poussa K, Porras M. Factors affecting formation and growth of intracranialaneurysms: a long-term follow-up study. Stroke 2001; 32:485-91.8. Wiebers DO, Whisnant JP, Huston J, 3rd, Meissner I, Brown RD, Jr., Piepgras DG, Forbes GS, Thielen K, Nichols D, O’Fallon WM, Peacock J, Jaeger L, Kassell NF, Kongable-Beckman GL, Torner JC. Unruptured intracranial aneurysms: natural history, clinical outcome, and risks of surgical and endovascular treatment. Lancet 2003; 362:103-10.9. The ACROSS Group. Epidemiology of aneurysmal subarachnoid hemorrhage in Australia and New Zealand: incidence and case fatality from the Australian Cooperative Research on Subarachnoid Hemorrhage Study (ACROSS). Stroke 2000; 31:1843-50.10. Tsutsumi K, Ueki K, Usui M, Kwak S, Kirino T. Risk of recurrent subarachnoid hemorrhage after complete obliteration of cerebral aneurysms. Stroke 1998; 29:2511-3.11. Wermer MJH, Rinkel GJE, Greebe P, Albrecht KW, Dirven CM, Tulleken CA. Late recurrence of subarachnoid hemorrhage after treatment for ruptured aneurysms: patient characteristics and outcomes. Neurosurgery. 2005; 56:197-204.12. Herman B, Leyten AC, van Luijk JH, Frenken CW, Op de Coul AA, Schulte BP. Epidemiology of stroke in Tilburg, the Netherlands. The population-based stroke incidence register: 2.Incidence, initial clinical picture and medical care, and three-week case fatality. Stroke. 1982; 13:629-34.13. Linn FH, Rinkel GJE, Algra A, van Gijn J. Incidence of subarachnoid hemorrhage: role of region, year, and rate of computed tomography: a meta-analysis. Stroke 1996; 27:625-9.14. Inagawa T. Trends in incidence and case fatality rates of aneurysmal subarachnoid hemorrhage in Izumo City, Japan, between 1980-1989 and 1990-1998. Stroke 2001; 32:1499-507.15. Ohkuma H, Fujita S, Suzuki S. Incidence of aneurysmal subarachnoid hemorrhage in Shimokita, Japan, from 1989 to 1998. Stroke 2002; 33:195-9.16. Kivisaari RP, Porras M, Ohman J, Siironen J, Ishii K, Hernesniemi J. Routine cerebral angiography after surgery for saccular aneurysms: is it worth it? Neurosurgery 2004; 55:1015-24.17. van der Schaaf IC, Brilstra EH, Rinkel GJE, Bossuyt PM, van Gijn J. Quality of life, anxiety, and depression in patients with an untreated intracranial aneurysm or arteriovenous malformation. Stroke 2002; 33:440-3.18. Wermer MJH, van der Schaaf IC, Van Nunen P, Bossuyt PM, Anderson CS, Rinkel GJE. Psycho-social impact of screening for intracranial aneurysms in relatives with familial subarach-noid hemorrhage. Stroke. 2005; 36:836-40.

43

Three

Long-term mortality and vascular event risk after aneurysmal subarachnoid hemorrhage

M.J.H. Wermer, P. Greebe, A. Algra and G.J.E. Rinkel

ChapterFour

46

Chap

ter

AbstractAbstract

Background and PurposeBackground and Purpose

Hypertension and smoking are risk factors shared between aneurysmal subarachnoid hemor-rhage (SAH) and vascular disease in general. Patients with a history of SAH may therefore be at increased risk for vascular events and excess mortality. We studied the overall mortality and long-term risk of vascular events after SAH.

MethodsMethods

We interviewed 752 patients (mean age 50 years, 67% women, mean follow-up 8.1 years) with SAH who had been discharged home or to a rehabilitation facility between 1985 and 2001 about new vascular events (ischemic/hemorrhagic stroke, myocardial infarction or vascular death). If patients had died during follow-up, we retrieved the cause of death. We compared the age and sex-specifi c mortality after SAH with that of the general population by means of standardized mortality ratios. The incidence of vascular events in SAH patients was compared with that in patients with a minor ischemic stroke (including transient ischemic attack) by time to event curves and Cox regression analysis (adjusted for age and sex).

ResultsResults

The standardized mortality ratio for SAH patients was 3.0 (95% CI 2.5-3.7). In the fi rst 10 years after the SAH the cumulative incidence of a vascular event was 10.8% (95% CI 7.8-13.8), which was lower than that in patients with a minor stroke (HR 0.43, 95% CI 0.33-0.56).

ConclusionsConclusions

Patients who have recovered to a functional independent state after SAH have an excess mortality compared with the general population. The risk of vascular events after SAH is lower than after minor stroke, but higher than the population risks reported in the literature, and therefore probably contributes to the excess mortality.

47

Four

IntroductionIntroduction

Subarachnoid hemorrhage (SAH) from a ruptured aneurysm is a subset of stroke that occurs in relatively young persons; half of the patients are younger than 55 years of age.1 In popula-tion-based studies the case-fatality rate is about 50% and another 20% of patients remains disabled and dependent for activities of daily life (ADL).2 Generally, patients who make a good recovery after SAH with successful occlusion of the ruptured aneurysm are believed to have a long-term outcome comparable with that of persons without a history of SAH. However, in a previous study we showed that patients with a SAH have a high risk of a recurrent SAH from new developed aneurysms or from aneurysms that were not detected at the time of the initial SAH.3 Furthermore, hypertension and smoking, two of the main risk factors for SAH, are also main risk factors for vascular diseases in general.4 SAH patients might therefore not only be at risk for recurrent SAH but also for other vascular diseases, which may lead to a reduced life-expectancy compared with persons without a history of SAH.We performed a long-term follow-up study in a large cohort of patients who regained indepen-dence after SAH and in whom all detected aneurysms had been occluded by means of clipping. We studied the overall mortality rate and compared this with the mortality in the general population by means of standardized mortality ratios (SMR). In addition, we determined the incidence of vascular events other than recurrent SAH and compared this with the incidence in patients who had had a transient ischemic attack (TIA) or minor ischemic stroke.

MethodsMethods

PatientsPatients

From a prospectively collected database of patients admitted to the University Medical Center Utrecht with SAH we included all patients admitted between 1985 and 2001 who met the following criteria: 1) subarachnoid hemorrhage confi rmed by CT or lumbar puncture; 2) presence of a saccular aneurysm confi rmed by conventional angiography or CT- angiography; 3) clipping of the ruptured aneurysm and additional aneurysms; 4) age at time of SAH ≥ 18 years and 5) discharge to home or a rehabilitation facility. Patients who were discharged to a nursing home, patients who were treated by means of coiling and patients with one or more aneurysms left untreated were not included.

Follow-upFollow-up

After approval of the ethical committee of our hospital we contacted the general practitioner of all eligible patients to know if the patient was still alive. If a patient had died we asked

48

Chap

ter

for the date and cause of death. If death occurred in a hospital or other facility, we reviewed the medical records. Subsequently, we sent a letter to all patients who were still alive. In this letter patients 70 years of age or younger were invited to the outpatient clinic. For patients older than 70 years we announced a telephone call. If a patient had no phone number or an ex-directory one, we sent another letter asking the patient to contact us. At the outpatient clinic or during the telephone interview we asked the patients about new vascular events other than recurrent SAH by means of a standardized interview. For all patients we retrieved data on gender, age, location and number of aneurysms at time of the SAH, smoking habits, alcohol use and history of hypertension.

Data-analysis Standardized Mortality RatiosData-analysis Standardized Mortality Ratios

We registered the number of patients in our study population who had died during follow-up (all causes). We used Standardized Mortality Ratios (SMR) to investigate possible excess mortality in SAH patients compared with the general population. Population-based statistics of The Netherlands were used as reference for the calculation of the total expected number of deaths (all causes).5 Mortality ratios were standardized in an indirect manner according to age and sex. Person-years of our cohort were calculated for sex and age (5-year) strata. The expected number of deaths in our cohort was calculated by multiplying the age-sex specifi c mortality rates of the reference population by the cumulative number of person-years in each stratum. The sum of all categories yielded the expected number of deaths taking into account variation in national mortality rates according to sex and age. The SMR is the ratio of the observed number of deaths (all causes) in our cohort to that of the expected number of deaths based on the general population. An SMR > 1 means excess mortality in the study cohort compared to the reference population. Ninety-fi ve percent confi dence intervals (CIs) were calculated based on the Poisson distribution.

Data-analysis vascular eventsData-analysis vascular events

The risk of vascular events was assessed by survival analysis. Primary measurement of outcome was the composite event of death from all vascular causes (other than SAH), non-fatal stroke (other than SAH) or non-fatal myocardial infarction, whichever occurred fi rst. All events were classifi ed according to pre-defi ned criteria.Deaths due to ischemic stroke, intracerebral hemorrhage, myocardial infarction, congestive heart failure and sudden death were classifi ed as vascular death. Sudden death was defi ned as an unexpected death of presumed or proven cardiac origin occurring within one hour after onset of symptoms, or within 24 hours given convincing circumstantial evidence. Stroke other than SAH was defi ned as an episode of relevant focal defi cits with acute onset, lasting for more than 24 hours and confi rmed on neurological examination. If CT showed a

49

Four

hypodense lesion compatible with the clinical features, or if CT was performed only in the acute phase and was negative, the stroke was considered a cerebral infarction. If CT showed a hyperdense lesion compatible with the clinical features, the stroke was classifi ed as intracere-bral hemorrhage. Transient ischemic attacks were not included as outcome event. A recurrent SAH was also not included as an outcome event for this analysis.Myocardial infarction had to be documented by at least two of the following characteristics: a history of chest discomfort, specifi c cardiac enzyme levels more than twice the upper limit of normal, or the development of Q waves on the standard 12-lead electrocardiogram.Vascular events that did not fulfi ll the above mentioned criteria, for example because no information from medical records was available, were classifi ed as possible events and were not included in the analysis.We calculated the cumulative incidence of all vascular events at 5, 10 and 15 years after SAH with corresponding 95% CIs. If patients had died during the follow-up period or were admit-ted to a nursing home they were censored at that point of time. Patients who were lost to follow-up were censored at the time of the last known follow-up.We compared the incidence of vascular events in the cohort of SAH patients with the incidence in patients after an ischemic attack (TIA) or minor ischemic stroke. These patients were derived from the LiLAC (Life Long After Cerebral ischaemia) study, in which a cohort of 2473 TIA/minor stroke patients were followed for the occurrence of new vascular events (mean follow-up time 10.1 years).6 The LiLAC cohort was based on participants of the Dutch TIA Trial.7 For comparison of the vascular event risk we used time to event curves. By means of Cox regression analysis we calculated the hazard ratio of the vascular event risk in SAH patients compared with the LiLAC cohort adjusted for age and sex.

ResultsResults

PatientsPatients

Between 1985 and 2001 930 patients with SAH were successfully treated for the ruptured aneurysm and were discharged alive. One hundred fi fty-four did not meet the inclusion criteria (Figure 1). Of the 776 remaining patients 24 (3%) could not be contacted and it was unknown if these patients were dead or alive. Therefore, 752 patients were included in the calculations for the SMR (total FU 6064 years, mean FU 8.1 years, range 0.2-20.1). For another 74 patients (9%) we had information on whether or not they had died but information on vascular events was lacking; these patients were not included in the analyses on vascular events. Therefore, for this part of the study 678 patients were included (total FU 5525 years, mean FU 8.1 years, range 0.2-20.1). During follow-up 23 patients were censored because they moved abroad (6) or were admitted to a nursing home (17).

50

Chap

ter

Figure 1 Flow diagram of patients for the evaluation of vascular event other than SAH

1818

617

participation in researchparticipation in research

593

51

Four

The baseline characteristics of the SAH patients who were included in the vascular event part of the study and the baseline characteristics of the LiLAC patients are listed in Table 1.

Standardized Mortality RatiosStandardized Mortality Ratios

Of the 752 patients, 107 patients died during the 6064 follow-up years. The overall mortality rate was 1.8% (95% CI 1.4-2.1%) per 1000 patient-years of follow-up. Seven patients died suddenly before reaching the hospital. The other causes of death were cancer in 21 patients, recurrent SAH in 10 patients, cardiovascular disease in 19 patients (11 of these 19 events fulfi lled our pre-defi ned

Characteristics TIA/minor stroke patientsN=2473

SAH PatientsN=678

Age mean in years ± SD 65 ± 10.1 50 ± 12.3

Women 987 (40%) 450 (66%)

Follow-up mean in years ± SD 10.1 ± 4.8 8.1 ± 4.6

SmokingCurrent smokerFormer/Never smoker/Unknown

1125 (46%)1348 (54%)

367 (54%)311 (46%)

Alcohol useAlcohol use > 5 drinks a dayFormer use > 5 drinks a dayAlcohol use 1-5 drinks a dayNeverUnknown

Not available16 (2%)28 (4%)331 (49%)146 (22%)157 (23%)

History of hypertension 1484 (60%) 234 (35%)

Location of the ruptured aneurysmAcom / ACAICA (incl Pcom)MCAVertebrobasilar

NANANANA

294 (43%)189 (28%)148 (22%)47 (7%)

No of aneurysms at time of SAH 1234

NANANANA

580 (86%)70 (9%)24 (4%)4 (1%)

Table 1. Baseline characteristics of the 2473 TIA/minor stroke patients and the 678 SAH patients included in the analysis of vascular event rates

Acom / ACA = Anterior communicating artery / Anterior cerebral arteryICA = Internal carotid artery MCA = Middle cerebral arteryVertebrobasilar = Arteries of the vertebrobasilar systemNA= not applicable

52

Chap

ter

Stratum No. patients No. patient-years

No. observeddeath (N)

No. expecteddeath (N)

SMR (95% CI)

Men < 40 years40-60 years> 60 yearsTotal

6213851251

53911843832106

4151938

0.55.79.315.5

7.5 (2.8-20.0)2.6 (1.4-4.5)2.0 (1.3-3.2)2.4 (1.8-3.4)

Women< 40 years40-60 years> 60 yearsTotal

93274134501

853209510103958

5323269

0.55.513.619.7

9.3 (3.9-22.4)5.8 (4.1-8.2)2.3 (1.7-3.3)3.5 (2.8-4.4)

All< 40 years40-60 years> 60 yearsTotal

155412185752

1392327813946064

94751107

1.111.023.035.2

8.4 (4.4-16.2)4.2 (3.2-5.7)2.2 (1.7-2.9)3.0 (2.5-3.7)

Table 2. Standardized Mortality Ratios for SAH patients per different age and sex strata

No. observed death = the number of patients who died during the follow-up time of our cohortNo. expected death = the number of patients expected to die during the follow-up time of our cohort based on population death rates

criteria for vascular event other than SAH), infectious diseases in nine patients, trauma in three patients, dementia in two patients, renal failure in two patients, other causes in seven patients and unknown in 27 patients. The SMRs are listed in Table 2. The overall SMR was 3.0 (95% CI 2.5-3.7). The SMR the highest in women younger than 40 years of age.

Vascular eventsVascular events

In 61 (9.0%, 95% CI 7-12%) of the 678 patients a vascular event had occurred during the 5525 follow-up years: six patients suddenly died, 27 patients had an ischemic stroke, 10 patients had an ICH and 18 patients had a myocardial infarction. In addition, 10 patients had a recur-rent SAH and 14 patients had a possible vascular event (stroke, vascular death or myocardial infarction that did not fulfi ll our pre-specifi ed criteria), but these events were not included in the analysis. The cumulative 5-year risk for vascular events other than SAH was 4.5% (95% CI 2.8-6.1%) and the cumulative 10-year risk was 10.8% (95% CI 7.8-13.8).

53

Four

The cumulative incidences of vascular events in SAH patients are lower (age and gender ad-justed hazard ratio 0.43, 95% CI 0.33-0.56) than in the cohort of patients with a TIA or minor stroke. The age and gender adjusted time to event curves of the cumulative incidence of vascular events in the SAH patients and in the patients with TIA or minor stroke are shown in Figure 2.

DiscussionDiscussion

We found that there is an increased long-term mortality rate in SAH patients compared with the general population. This increased mortality is the most prominent in women and in patients younger than 40 years. In the fi rst 10 years after the SAH one of every nine patients experiences a vascular event other than recurrent SAH. The exact cause of the excess mortality in SAH patients has not been elucidated yet. A fi rst explanation might be recurrent episodes of SAH. In a previous study we found that the risk of a recurrent SAH in patients clipped for ruptured aneurysms is 22 times higher than the risk of SAH in the general population.3 The absolute risk, however, is relatively low (3.2% in the fi rst 10 years after the SAH) and cannot entirely explain the excess in mortality. A second

Figure 2 The age and sex adjusted cumulative incidence of patients with a vascular event after SAH(lower line) and after a TIA or minor stroke (upper line)

Patients at risk SAH 466 226 60Patients at risk LiLAC 1768 1200 254

54

Chap

ter

explanation might be an increased risk of vascular events other than SAH, but currently no direct comparisons are available on the risk of vascular events in patients with SAH versus the general population. One other study described an excess long-term mortality in patients who survived after SAH.8 In this study 1537 Finnish SAH patients (96% treated by clipping and 4% by coiling) were followed over a median time of 7.5 years. The overall SMR was 4.5 for all patients and 2.0 for all patients with a good recovery after 12 months. Cerebrovascular diseases accounted for 24% of all deaths and other vascular diseases for another 17%. In the Finnish study no data are available on the risk of vascular events not leading to death. In both our study and the Finnish study the number of SAH patients was too small to allow calcula-tions for cause specifi c SMRs. When larger data sets of follow-up of SAH patients becomeavailable the calculation of cause specifi c SMRs will help to further elucidate the cause of excess mortality among SAH patientsThe risk of vascular events after SAH is lower than that after TIA or minor stroke. Therefore, SAH cannot be considered such a strong marker for atherosclerosis or a risk factor for atherosclerotic diseases as minor stroke. Unfortunately, at this moment no detailed population-based data on the incidence of stroke or myocardial infarction are available in the Netherlands. Therefore, we cannot directly compare the vascular event risk in our population with the risk in the general population. Some indirect comparisons with data from the literature can, however, be made. The incidence of a stroke (ischemic stroke or intracerebral hemorrhage) in our cohort with a mean age of 50 years and two-thirds of women was 6.7 per 1000 patient-years. The incidence of a fi rst stroke in the Netherlands is only 1.3 per 1000 patient-years for men and 1.0 per 1000 patient-years for women with an age of 50 years.9 The incidence of all vascular events (cerebrovascular events, vascular death and myocardial infarction) in our cohort was around 11 per 1000 patient-years. In a recent study on the occurrence of vascular diseases in a population of 91106 persons in Oxfordshire, the combined incidence of cerebrovascular events other than SAH, vascular death and myocardial infarction was 5 per 1000 patient-years for men and 4 per 1000 patient-years for women.10 These comparisons strongly suggest that the incidence of vascular events in SAH patients is higher than that in the general population with comparable age and gender. Within a few years more detailed population-based data will become available. Until these data have been analyzed, no defi nite conclusions about the extent of the increased risk of vascular diseases in the SAH population can be drawn.There are some indications that atherosclerosis might be involved in the pathophysiology of intracranial aneurysms. Infi ltration of the arterial wall by macrophages and T lymphocytes is an early and persistent feature of atherosclerosis. Cell surface adhesion molecules, comple-ment activation, infl ammatory cytokines and release of metalloproteinases participate in the development of atherosclerotic plaques. In both unruptured and ruptured intracranial aneurysms these signs of vascular infl ammation and immunological reaction related to atherosclerosis have been found. 11, 12

There are some limitations of our study that need discussion. Firstly, although the information on mortality was complete for 752 of the 776 eligible patients, information on vascular events

55

Four

was available for only 678 of the 776 patients. The majority of patients without follow-up on vascular events were patients who had died and patients who were not approached because they pointed out in previous research projects that they did not want to participate in future studies. Because the reason that patients die or do not want to participate in research might be related to a worse health state, we might have underestimated the risk of vascular events. Secondly, we included only vascular events in our analysis that were medically confi rmed. Fourteen patients reported to have had a vascular event but no information about this event could be obtained. Excluding these cases from our analysis might again have resulted in an underestimation of the vascular event risk. Conversely, six of the 61 patients with a vascular event other than SAH died suddenly, presumably of cardiac arrest. Because the cause of death in these patients was not confi rmed by autopsy, we cannot completely rule out that some of them died of a recurrent SAH. Including these patients could have caused a slight overestimation of the risk of vascular events other than recurrent SAH. Thirdly, we focused in our study on patients who recovered to an ADL independent state after SAH and excluded patients from our cohort who were discharged to a nursing home after the SAH. Informed consent for follow-up in patients in nursing homes is in general hard to obtain because most of them have severe cognitive defi cits. Our results therefore apply only to patients discharged home or to a rehabilitation center, which will in general be a younger population than the patients discharged to a nursing home. In our study we included only patients who had been treated by means of clipping. It is, however, likely that the risk of vascular events other than recurrent SAH is the same in coiling as in clipping. Finally, because we compared our study cohort with patients from a different study (LiLAC) one might question if the follow-up of the patients and the outcome events were handled the same in both studies. However, the defi nition of outcome events was comparable and the investigators involved in the follow-up and the ascertainment of outcome events in the present study also participated in the same processes in the LiLAC study.In conclusion, there is an excess mortality after SAH compared with the general population. The risk of vascular events in SAH is lower than in patients after TIA or minor stroke but higher than population-based incidence rates reported in the literature. Because smoking and hypertension are risk factors for both a recurrent SAH and other vascular events, all SAH patients should be stimulated to quit smoking and to have regular checks of their blood pressure. Patients after TIA or minor stroke are generally treated with cholesterol and blood pressure lowering drugs in combination with anti-platelet therapy. Whether similar secondary prevention is warranted in SAH patients depends on the vascular event risk in SAH patients in relation to that of the general population and needs further investigation.

56

Chap

ter

ReferencesReferences

1. The Across study group. Epidemiology of aneurysmal subarachnoid hemorrhage in Australia and New Zealand: Incidence and case fatality from the Australian cooperative research on subarach- noid hemorrhage study (ACROSS). Stroke 2000;31:1843-1850.2. Hop JW, Rinkel GJE, Algra A, van Gijn J. Case-fatality rates and functional outcome after subarach- noid hemorrhage: A systematic review. Stroke 1997;28:660-664.3. Wermer MJH, Greebe P, Algra A, Rinkel GJ. Incidence of recurrent subarachnoid hemorrhage after clipping for ruptured intracranial aneurysms. Stroke 2005;36:2394-2399.4. Feigin VL, Rinkel GJE, Lawes CM, Algra A, Bennett DA, van Gijn J, Anderson CS. Risk factors for sub- arachnoid hemorrhage: An updated systematic review of epidemiological studies. Stroke 2005;36:2773-2780.5. National life tables. CBS offi ce of statistics Netherlands, Voorburg, The Netherlands.6. van Wijk I, Kappelle LJ, van Gijn J, Koudstaal PJ, Franke CL, Vermeulen M, Gorter JW, Algra A. Long- term survival and vascular event risk after transient ischaemic attack or minor ischaemic stroke: A cohort study. Lancet 2005;365:2098-2104.7. The Dutch TIA Trial Study Group. A comparison of two doses of aspirin (30 mg vs. 283 mg a day) in patients after a transient ischemic attack or minor ischemic stroke. N Engl J Med 1991;325:1261- 1266.8. Ronkainen A, Niskanen M, Rinne J, Koivisto T, Hernesniemi J, Vapalahti M. Evidence for excess long- term mortality after treated subarachnoid hemorrhage. Stroke 2001;32:2850-2853.9. Bots ML, Berger-van Sijl M, Jager-Geurts MH, Bos M, Reitsma JB, Breteler MMB, de Bruin A. Incidence of cerebrovascular diseases in the Netherlands in 2000. In Jager-Geurts MH, Peters RJG, van Dis SJ, Bots, ML. Cardiovascular diseases in the Netherlands 2006, disease and death rates. The Hague: Netherlands Heart Foundation 2006.10. Rothwell PM, Coull AJ, Silver LE, Fairhead JF, Giles MF, Lovelock CE, Redgrave JN, Bull LM, Welch SJ, Cuthbertson FC, Binney LE, Gutnikov SA, Anslow P, Banning AP, Mant D, Mehta Z. Population-based study of event-rate, incidence, case fatality, and mortality for all acute vascular events in all arterial territories (Oxford vascular study). Lancet 2005;366:1773-1783.11. Chyatte D, Bruno G, Desai S, Todor DR. Infl ammation and intracranial aneurysms. Neurosurgery 1999;45:1137-1146.12. Kataoka K, Taneda M, Asai T, Kinoshita A, Ito M, Kuroda R. Structural fragility and infl ammatory response of ruptured cerebral aneurysms. A comparative study between ruptured and unruptured cerebral aneurysms. Stroke 1999;30:1396-1401.

57

Four

ChapterFive

Anosmia after aneurysmal subarachnoidhemorrhage

M.J.H. Wermer, M. Donswijk, P. Greebe, B. Verweij and G.J.E. Rinkel

60

Chap

ter

Abstract

Background and purpose

Anosmia has an important impact on well-being, but is often neglected by physicians. In patients with subarachnoid hemorrhage (SAH) anosmia has been reported mainly after sur-gery for aneurysms of the anterior communicating artery (ACA). We studied the prevalence, predisposing factors (aneurysm site and type of treatment), impact and prognosis of anosmia in SAH patients.

Methods

From those SAH patients who resumed independent living, we included all patients treated by coiling between 1997 and 2003 and a sample of patients treated by clipping between 1985 and 2001. Patients underwent structured interviews regarding the presence and duration of anosmia. The impact of anosmia was scored on a visual analog scale (VAS) ranging from zero (no infl uence) to 100 (the worst thing ever). Risk factors for anosmia were assessed by logistic regression analysis.

Results

Overall 89 (28% [95% CI 23-34%]) of the 315 interviewed patients reported anosmia after the SAH (mean follow-up 7.4 years); 10 (15%) of the 67 coiled patients and 79 (32%) of the 248 clipped patients. The median VAS impact score was 53 (range 0-100). In 20 of the 89 patients (23% [95% CI 15-33]) the symptoms had improved over time. Risk factors for anosmia were treatment by clipping (OR 2.7 [95% CI 1.3-5.7]) and ACA aneurysms (OR 2.0 [95% CI 1.2-3.3]).

Conclusions

Anosmia after SAH has a high prevalence, a considerable impact and a poor prognosis. Its oc-currence after coiling suggests that not only damage to the olfactory nerve by clipping but also the SAH itself plays a role in its pathogenesis.

61

Five

Introduction

Loss of smell (anosmia) has an important impact on the quality of life and on nutritional intake. Commonly reported problems from anosmia are safety issues, such as not perceiving potentially toxic environmental substances or spoiled food, and interpersonal relations, such as insecurity about one’s own body odor and impaired sexual life.1

Anosmia has been reported in surgically treated patients who have recovered after aneurys-mal subarachnoid hemorrhage (SAH), especially in patients with a ruptured aneurysm of the anterior communicating artery,2 but also in patients treated for aneurysms at other sites via the frontotemporal approach.3 In patients with anterior communicating artery aneurysms the prevalence of anosmia has been reported to vary with the neurosurgical approach,4 which suggests that brain retraction of the frontal lobe or direct pressure from spatula at the olfactory bulb or nerve during clipping of the aneurysm plays an important role in the development of anosmia. Anosmia has, however, also been reported after neurovascular decompression for trigeminal neuralgia performed in sitting position5, which indicates that not only direct damage but also traction from brain shift can cause anosmia. Whether aneurysmal rupture itself can also cause anosmia and whether anosmia occurs after endovascular treatment of aneurysms is unknown. Also, the prognosis of anosmia after SAH is unclear, because no fol-low-up studies exist that investigated the long-term outcome of anosmia.The objective of our study was to determine the prevalence, the impact and prognosis of anosmia in patients who have been treated for a ruptured aneurysm. By comparing the prevalence in clipped and coiled patients and in aneurysms at different locations we tried to identify causal factors in the development of anosmia.

Methods

Patients

Patients who had been treated by coiling between 1997 and 2003 or by clipping between 1985 and 2001 in the University Medical Center Utrecht (UMCU) were eligible for our study. We included patients who met the following inclusion criteria 1) subarachnoid hemorrhage confi r-med by CT or lumbar puncture; 2) presence of a saccular aneurysm confi rmed by conventional angiography or CT- angiography; 3) treatment of the ruptured aneurysm; 4) age at time of SAH ≥ 18 years and 5) discharge to home or a rehabilitation facility. The clipped patients had been enrolled in a long-term follow-up study on the incidence of recurrent SAH.7 The clipped patients were a sample of participants of this follow-up study; only patients who were enrol-led after January 2004 were asked to participate in the anosmia study. Not eligible for the study were patients who resided in a nursing home, who were not able to communicate well

62

Chap

ter

enough to give informed consent, or who were treated by both clipping and coiling. Patients who reported to have had loss of smell before the SAH were excluded from the analysis.

Data-extraction and questionnaires

We recorded the age and gender of the patients, the mode of treatment, the number of treatments and the location and size of the ruptured and additional aneurysms. Aneurysm locations were categorized into: 1) anterior cerebral artery (ACA) including the anterior com-municating artery and pericallosal arteries; 2) internal carotid artery (ICA) containing the ophthalmic, posterior communicating, anterior choroidal arteries and the carotid T junction; 3) middle cerebral artery (MCA) and the 4) vertebrobasilar system.For the aneurysms located at the anterior communicating artery and the pericallosal artery we retrieved operation reports and categorized the neurosurgical approach into: a) fronto-temporal approach or b) anterior interhemispheric approach.Patients underwent a standardized, semi-structured interview in which a self-developed ques-tionnaire was administered by one of three researchers (MW, PG, MD). The coiled patients were all interviewed by telephone. The clipped patients ≤ 70 years of age were interviewed at the outpatient clinic, the patients older than 70 were interviewed by telephone. The researchers did fi ve of the interviews together at the start of the study to reduce possible interobserver variation. The questionnaire comprised questions on whether patients had noticed any degree of loss of smell and/or taste after the SAH. Patients were asked whether there had been some degree of recovery and if so, after what time recovery had occurred. The impact of anosmia on quality of life was assessed on a visual analogue scale (VAS) by placing a mark on a vertical scale which ranges from zero (no infl uence on their lives at all) to 100 (the worst thing ever happened to them). Patients who were interviewed by telephone were asked to rate the impact on quality of life on a scale from zero to 100 instead of placing a mark.

Data-analysis

Descriptive statistics were performed by SPSS (version 11.5). We used logistic regression ana-lysis to calculate odds ratios (OR) of characteristics associated with anosmia. The following factors were included in the analyses: age at time of SAH (continuous), gender, location of the ruptured aneurysm (categorical: internal carotid artery [ICA], anterior cerebral artery [ACA], medial cerebral artery [MCA], vertebrobasilar and dichotomous: ACA versus other locations), neurosurgical approach (only for anterior and pericallosal aneurysms), size of the ruptured aneurysm (dichotomous ≤ 10 millimeters and > 10 millimeters), number of aneurysms at the time of SAH (dichotomous: one versus multiple) and total number of treatments for the ruptu-red and additional aneurysms. We performed univariate regression analysis for all risk factors and multivariate regression analysis with forward selection of variables with probability values < 0.20 in the univariate analysis. Data were reported with 95% confi dence intervals (CI).

63

Five

Results

Patients

Eighty-two coiled patients were eligible for the study; of these 82 coiled patients 14 were lost to follow-up and one declined participation. Of the 776 eligible clipped patients, 502 were not included because they had been enrolled in the follow-up study of recurrent SAH before 2004 and 24 were lost to follow-up. Of the remaining 250 patients none declined participation, but two had pre-existent anosmia and were therefore excluded from the analysis. Thus, we included 315 patients; 67 (21%) had been treated by coiling and 248 (79%) by clipping. The mean age of the patients at the time of the interview was 60.3 years (range 25 – 91) and 199 (63%) of them were women. The mean follow-up time after the SAH was 7.4 years (median 5.7 range 1.2 – 20.4). The baseline characteristics of the included patients are shown in Table 1.

Incidence, impact and prognosis of anosmia

Eighty-nine (28% [95% CI 23-34]) of the 315 patients reported to have had loss of smell after treatment for SAH; 10 (15% [95% CI 8-26]) of the 67 coiled patients and 79 (32% [95% CI 26-38]) of the 248 clipped patients (Table 2). The mean age of the patients with anosmia was 60.8 years (range 26 - 82) and 50 (56.2%) of them were women. In 48 (54%) of these patients the ruptured aneurysm was located at the ACA, in 17 (19%) at the MCA, in 17 (19%) at the ICA and in seven (8%) at the vertebrobasilar system.Of the 89 patients with anosmia, 15 said they were not able to express the impact of anosmia on a scale from 0 to 100. The remaining 74 patients evaluated the impact on life with a median of 53 on a scale from zero to 100 (mean 55, SD 29.3, range zero-100).Twenty (23% [95% CI 15-33]) of the 89 patients had noticed some degree of recovery over time; improvement had occurred in 13 (16%) of the 79 clipped patients and seven (70%) of the 10 coiled patients (difference between proportions 54% [95% CI 24-83). The median interval bet-ween SAH and onset of recovery was six months (SD 26.0, range 1-80). Three patients reported some degree of recovery more than a year after the SAH, in all other patients improvement occurred within the fi rst year.

64

Chap

ter

Risk factors for anosmia

In the univariate regression analysis treatment by means of clipping (OR 2.7, 95% CI 1.3-5.5) and treatment of a ACA aneurysm (OR 2.2, 95% CI 1.3-3.6) were statistically signifi cant related with anosmia. Gender (OR for men 0.7, 95% CI 0.4-1.1) and the presence of multiple aneurysms at the time of the SAH (OR 1.7, 95% CI 0.9-3.3) were not statistically signifi cant factors in the univariate analyses but had a p-value of < 0.2 and were therefore included in the multivariate analyses. For patients with anterior and pericallosal aneurysms, the risk of anosmia was higher when treated by the frontotemporal approach compared to the anterior interhemispheric approach, but this difference was not statistically signifi cant (OR 1.3; 95% CI 0.2-7.4)In the multivariate regression analysis, treatment by means of clipping (OR 2.7, 95% CI 1.3-5.7) and treatment of a ACA aneurysm (OR 2.0, 95% CI 1.2-3.3) were both statistically signifi cant risk factors for anosmia.

Coiled patientsN=67

Clipped patientsN=248

Women 41 (61.2%) 158 (63.7%)

Mean age at time of SAH ±SD 52.4 ± 11.3 53.2 ± 12.5

Mean time after SAH ± SD 3.5 ± 1.7 8.3 ± 4.8

Location ruptured aneurysm*ACA

-Anterior cerebral artery-Acom-Pericallosa

MCAICAVB

20 (30%)1 17224 (36%)2 (3%)21 (31%)

117 (47%)2108756 (23%)60 (24%)15 (6%)

Size of ruptured aneurysm ≤ 10 mm> 10 mmUnknown

Number of aneurysms at SAH1> 1 with ipsilateral aneurysms> 1 with contralateral aneurysms

24 (36%)16 (24%)27 (40%)

51 (76%)9 (14%)7 (10%)

49 (20%)87 (35%)112 (45%)

222 (90%)13 (5%)13 (5%)

Table 1 Baseline characteristics of the 315 participants

*ACA= anterior cerebral artery including the anterior communicating artery and pericallosal arteriesMCA= medial cerebral artery ICA= internal communicating arteryVB= vertebrobasilar system

65

Five

Presence of anosmia (95% CI)

All patientsclipping coiling

Location of the aneurysm

ACAclippingcoiling

MCAclippingcoiling

ICAclippingcoiling

VBclippingcoiling

Neurosurgical approach

Acom and anterior aneurysms (n=110)frontotemporal (n= 93)anterior interhemispheric (n=0)unkown (n=17)

Pericallosa aneuryms (n=7)frontotemporal (n=0)anterior interhemispheric (n=6)unknown (n=1)

Size of the ruptured aneurysm (n=176)

≤≤ 10 millimetersclippingcoiling

>10 millimetersclippingcoiling

28% (23-34)32% (26-38)15% (8-26)

35% (27-44)38% (29-47)20% (7-44)27% (17-40)28% (18-42)0%21% (13-32)25% (15-39)13% (3-33)19% (9-37)27% (9-55)14% (4-37)

38% (29-48)39% (30-50)0%29% (11-56)

29% (5-70)0%33% (6-76)0%

22% (13-33)31% (19-46)4% (0-23)

28% (20-38)32% (23-43)6% (0-32)

Table 2 Anosmia in relation to type of treatment and location and size of the aneurysm

ACA = Anterior cerebral artery including the anterior communicating artery and the pericallosal arteriesMCA = Middle cerebral arteryICA = Internal carotid arteryVB = Vertebrobasilar arteries

66

Chap

ter

Discussion

This study shows that the prevalence of anosmia is high in patients who have recovered from an aneurysmal SAH and in whom the aneurysm had been treated by means of neurosurgical clipping or endovascular coiling. One out of every three patients treated neurosurgically and one out of every six patients treated endovascularly reported anosmia, and in three out of every four patients anosmia was permanent. Recovery occurred more often after coiling than after clipping. The impact on patients’ well-being was rated as considerable. Risk factors for development of anosmia after endovascular or surgically treated SAH were ACA aneurysms and treatment by means of clipping.Several studies previously reported on the frequency of anosmia following operation for an-eurysms in the anterior circulation. The relation between this aneurysm location and anosmia is understandable due to the close anatomical relation of the anterior circulation and the olfactory tract. The prevalence found in these studies depended on the mode of aneurysm approach. Coming from the front, with the anterior interhemispheric approach (AIA), the frontal lobe is elevated and the olfactory tract must be extensively dissected. In this group, subjective anosmia was reported in 52 (55%) of 94 patients.8 A variation of this approach is the (lower) anterior basal interhemispheric approach (ABIA), which enables a lesser extent of both dissection of the olfactory tract and frontal lobe elevation. In one study 1 (2%) of the 52 patients treated via the ABIA had postoperative anosmia compared to 15 (31%) of the 49 patients treated via the AIA.4 However, drawbacks of these approaches are the lack of control of the circulation proximal from the aneurysm and pressure on the frontal lobes. A more fa-voured approach for aneurysms of the anterior communicating artery, also in our hospital, is the frontotemporal or pterional approach (FTA) (coming mostly from the temporal side) were vascular control is safer and the olfactory tract can be anatomically easier preserved and both sided injury is unlikely. Twenty-two of 25 patients (88%) operated for ruptured intracranial aneurysms by the FTA had postoperatively anosmia ipsilateral to the side of surgery.3 However, in another study on the FTA, the functions of the olfactory nerve could be preserved in 85% of the 100 included patients by placing the retractor only against the frontal lobe and using an aperture of only 1-1.5 centimeters.2 In the majority of these studies anosmia was found with olfactory testing and was not reported by the patients themselves. In our study we found a higher risk of anosmia in aneurysms treated by the AIA (39%) compared to FTA (33%) but this difference was not statistically signifi cant, probably because of the low number of patients treated by AIA.To our knowledge, no other studies reported on the impact and prognosis of anosmia after endovascular or surgically treated aneurysmal SAH or on the prevalence of anosmia after coi-ling. The profound impact of anosmia on quality of life has been described in other studies.1,9 In 17-29% of patients with anosmia signs of depression were found and overall satisfaction

67

Five

with life was reported by only 50% of patients.9 Commonly reported impairments by anosmia patients were inability to recognize spoiled food, gas leaks or fi re and the decreased enjoy-ment of food and cooking. Furthermore, a good sense of smell was found to be important for personal hygiene, social interactions and sex life.1

There are some limitations of our study that need to be acknowledged. The presence of an-osmia after the SAH was assessed by means of a questionnaire. Defi cits in smell could have been measured more objectively by means of standardized olfactory testing. There is, howe-ver, not one standard way of testing olfaction. In two studies that used both questionnaires and formal testing of olfaction, the correlation between subjective and objective test results was high.4,8 The positive predictive value in both studies was 100%. The negative predictive value was 84% in the study that used fi ve odors at eight concentration levels as testing of olfaction8 and 64% in the study that used vanilla essence placement with contralateral oc-clusion as testing of olfaction.4 These fi ndings suggest that the use of questionnaires instead of formal olfactory tests in our study could have led to an underestimation of the prevalence of anosmia. Furthermore, because the nature of our study was retrospective, patients might have had problems to recall anosmia that had subsided during follow up. This also might have resulted in an underestimation of both the proportion of patients with anosmia and the prognosis of anosmia. Finally, we did not systematically exclude other causes of anosmia such as medication, nasal and sinus disease, trauma or early Parkinson’s disease.1 However, given the relation in time with the SAH (we asked for the presence of anosmia right after treatment for the ruptured aneurysm) we consider these other causes unlikely.Anosmia did not only occur in patients who had been treated by clipping for an aneurysm of the anterior cerebral or communication artery, but also in patients treated by coiling and in patients who had ruptured aneurysms at other sites than the anterior circulation. This suggests that direct mechanical damage to the olfactory nerve is not the only causal factor for anosmia. The olfactory nerve and olfactory bulb might also be damaged through contact with subarachnoid blood or by increased intracranial pressure at the time of the aneurysmal rupture. In patients with siderosis of the central nervous system, anosmia occurred in at least 17% of the patients.10 Another possible causal factor for the development of anosmia might be ischemia of parts of the cerebral cortex involved in the processing of olfactory information. In a recent study, 10 (10%) out of 102 stroke patients with a fi rst ever stroke had olfactory dysfunction.11 In SAH patients cortical damage might develop as a consequence of an intra-cerebral hematoma from the aneurysmal rupture or as a consequence of delayed cerebral ischemia. Our study did not include patients with unruptured aneurysms. Assessing smell in treated and untreated patients with unruptured aneurysms could further unravel the role of treatment and subarachnoid blood in the pathogenesis of anosmia.In conclusion, anosmia after endovascular or surgically treated aneurysmal SAH has a wide-spread occurrence and an important impact on quality of life. It is likely that it often goes unrecognised by physicians and therefore more attention should be paid to this defi cit in

68

Chap

ter

the post-operative contact with SAH patients. The results of our study could help physicians in informing patients with anosmia about prognosis. More research is needed on the oc-currence of anosmia in patients treated for unruptured aneurysms to further elucidate its pathophysiology.

69

Five

References

1. Hummel T, Nordin S. Olfactory disorders and their consequences for quality of life. Acta Otolaryngol 2005;125:116-121.2. Aydin IH, Kadioglu HH, Tuzun Y, Kayaoglu CR, Takci E, Ozturk M. Postoperative anosmia after anterior communicating artery aneurysms surgery by the pterional approach. Minim Invasive Neurosurg 1996;39:71-73.3. Eriksen KD, Boge-Rasmussen T, Kruse-Larsen C. Anosmia following operation for cerebral aneurysms in the anterior circulation. J Neurosurg 1990;72:864-865.4. Fujiwara H, Yasui N, Nathal-Vera E, Suzuki A. Anosmia after anterior communicating artery aneurysm surgery: Comparison between the anterior interhemispheric and basal interhemispheric approaches. Neurosurgery 1996;38:325-328.5. Ramsbacher J, Vesper J, Brock M. Permanent postoperative anosmia: A serious complication of neurovascular decompression in the sitting position. Acta Neurochir (Wien) 2000;142:1259-1261.6. Wermer MJH, van der Schaaf IC, Velthuis BK, Algra A, Buskens E, Rinkel GJE. Follow-up screening after subarachnoid haemorrhage: Frequency and determinants of new aneurysms and enlargement of existing aneurysms. Brain 2005;128:2421-2429.7. Wermer MJH, Greebe P, Algra A, Rinkel GJE. Incidence of recurrent subarachnoid hemorrhage after clipping for ruptured intracranial aneurysms. Stroke 2005;36:2394-2399.8. Suzuki J, Yoshimoto T, Mizoi K. Preservation of the olfactory tract in bifrontal craniotomy for anterior communicating artery aneurysms, and the functional prognosis. J Neurosurg 1981;54:342-345.9. Miwa T, Furukawa M, Tsukatani T, Costanzo RM, DiNardo LJ, Reiter ER. Impact of olfactory impairment on quality of life and disability. Arch Otolaryngol Head Neck Surg 2001;127:497-503.10. Fearnley JM, Stevens JM, Rudge P. Superfi cial siderosis of the central nervous system. Brain 1995;118:1051-1066.11. Heckmann JG, Stossel C, Lang CJ, Neundorfer B, Tomandl B, Hummel T. Taste disorders in acute stroke: A prospective observational study on taste disorders in 102 stroke patients. Stroke 2005;36:1690- 1694.

ChapterSix

Long-term effects of subarachnoid hemorrhage onemployment, relationships, personality and mood

M.J.H. Wermer, H. Kool, K.W. Albrecht and G.J.E. Rinkel

72

Chap

ter

Abstract

Background and purpose

Only one third of patients regains functional independence after aneurysmal subarachnoid hemorrhage (SAH). Despite this recovery, many of such patients experience psychosocial pro-blems. We assessed the long-term effects of SAH on employment, relationships, personality and mood.

Methods

We included patients who had been treated by clipping after SAH between 1985 and 2001 and resumed independent living. Patients underwent structured interviews regarding employ-ment, relationships and personality before and after the SAH. Anxiety and depression were assessed by the Hospital Anxiety and Depression Scale (HADS) and scores were compared between the study group and a reference population.

Results

610 patients were interviewed (mean follow-up after SAH 8.9 years). Of the employed patients 26% stopped working and 24% worked shorter hours or had a position with less responsibility. On average, patients returned to work 9.4 months after discharge (range 0-96). Seven percent of patients got divorced because of SAH related problems. Fifty-nine percent of the patients reported changes in personality, most commonly increased irritability (37%) or emotionality (29%). SAH patients had a statistically signifi cant higher mean depression score than the reference population. Approximately 10% of the patients had a HADS score in the range of a probable depression or anxious state. Only 25% reported a complete recovery without psy-chosocial or neurological problems.

Conclusions

The long-term psychosocial effects of SAH are considerable even in patients who regain functional independence. Treating physicians should be aware of these long-term effects of SAH when discussing prognosis and reintegration to work after initial recovery with patients and family.

73

Six

Introduction

Subarachnoid hemorrhage (SAH) from a ruptured aneurysm has a poor prognosis. In popula-tion based studies the case-fatality rate is about 50% and another 10-20% of patients remain disabled with loss of independence.7 Even among survivors who make a good functional re-covery and are independent for activities of daily living (ADL) many experience psychosocial and cognitive defi cits in the fi rst year after SAH.6,8 Because half of the patients with SAH are less than 55 years old at time of the hemorrhage, the psychosocial effects of SAH may affect daily functioning over many years.Several studies have been performed on the psychosocial effects of SAH but most have a mean follow-up time of only 6-24 months after SAH.6,8,9 Until now two studies have been performed that concentrated on the long-term psychosocial outcome.5,10 However, these studies either focused on caregivers of SAH patients or included only a small proportion of eligible patients.We studied the long-term effects of SAH on employment, relationships, personality and mood in a large group of patients that recovered to a functional independent state after treatment for SAH by means of clipping.

Methods

Patients

For this study we included patients who participated in the ASTRA (Aneurysm Screening af-ter Treatment for Ruptured Aneurysms) study. In the ASTRA study we aimed to evaluate the effectiveness of screening with CT-angiography (CTA) in patients with a history of SAH. The inclusion criteria for the ASTRA study were: 1) admission between 1985 and 2001 to the Uni-versity Medical Center Utrecht or admission between 1987 and 1998 to the Academic Medical Center Amsterdam, 2) confi rmation of the SAH by CT and an aneurysm proven by CTA or digital subtraction angiography (DSA), 3) treatment of the ruptured aneurysm by clipping, 4) recovery to a functional independent state after treatment (defi ned as a score of ≤ 3 on the modifi ed Rankin scale1) and 5) age between 18 and 70 years at time of screening. Excluded were patients 1) who were not able to communicate well enough to give informed consent, 2) those with severe co-morbidity or reduced life expectancy, 3) patients in whom one or more aneurysms had been treated by coiling or 4) patients in whom recently a CTA or DSA had been performed. For all patients meeting the inclusion criteria we contacted the general practitioner to see if the patient was still alive and had not moved to a nursing home. Subsequently, we invited all patients to visit our outpatient clinic. The fl ow diagram of the patients in the study is shown in Figure 1. Of the eligible patients 31 were not approached because the intended number of

74

Chap

ter

patients for the ASTRA study had already been reached. We included a total of 610 patients. The study was approved by both hospital ethics committees.

Interview

At the outpatient clinic all patients were interviewed between 2002 and 2004 by one (MW) of the researchers. In all participants we recorded age, sex, Rankin score, the site of the ruptured aneurysm and neurological defi cits that remained after the SAH. The baseline characteristics of the 610 patients are shown in Table 1. The age and sex distribution of non-participants in the study was comparable with the distribution of the participants.

Figure 1 Flow diagram of excluded and included patients

201818

75

Six

By using standardized self-developed semi-structured questionnaires we asked the partici-pants about their employment pre- and post-SAH. If patients had been working before the SAH we recorded the kind of job and the total number of working hours before the SAH and after the SAH at the time of the interview. Also we recorded how long after the SAH patients returned to work. If patients had a relationship before the SAH we asked them if they still had the same spouse and if not if they thought the divorce was related to the SAH. We assessed possible changes in personality and classifi ed these if present in the following fi ve different categories: 1. increased irritability (including agitation / loss of decorum / egocentricity / ag-gression / quick-tempered / impatience) 2. apathy (including passivity / slowness / more intro-vert / poor motivation), 3. increased emotionality (including depression / anxiety / nervousness / lowered self-confi dence), 4. positive personality changes (including decreased irritability / more outgoing / more open minded) or 5. unclassifi ed. We asked patients if and how long they had spent in a rehabilitation facility. The questions on rehabilitation and personality chan-ges were added to the questionnaire after the fi rst 68 patients visited the outpatient clinic. For the fi rst 68 patients these data are not available. In addition we asked patients about

Participants (%) N=610

Mean age ± SD (range) 53.4 ± 9 (24-70)

Women 391 (64%)

Mean time after SAH ± SD (range) 8.9 ± 3.9 (2.3-18.8)

Location of ruptured aneurysm (at time of SAH)*Acom / ACAICAMCAVertebrobasilar

261 (43%)168 (28%)137 (22%)44 (7%)

Modifi ed Rankin score0123

211 (35%)135 (22%)244 (40%)20 (3%)

Neurological defi citsOphtalmoplegiaHemiparesisDysphasia

17 (3%)65 (11%)27 (4%)

Table 1 Characteristics of the 610 participants at time of the interview

* Acom / ACA = Anterior communicating artery / Anterior cerebral arteryICA = Internal carotid artery MCA = Middle cerebral arteryVertebrobasilar = Arteries of the vertebrobasilar system

76

Chap

ter

memory impairment and recorded spontaneous self-reported subjective symptoms such as fatigue, headaches and problems of concentration.To study anxiety and depression levels we administered the Hospital Anxiety and Depression Scale (HADS). The HADS, a simple screening tests for symptoms of anxiety and depression, contains 14 questions on mood that provides separate rating scales for anxiety and depressive symptoms, ranging from 0 to 21, to allow the classifi cations of ‘normal’ (scores <8), ‘possible depression (or anxiety)’ (scores of 8-10) and ‘probable depression (or anxiety)’ (scores >10).12 The HADS data on the patients with a history of SAH were compared with those of a Dutch reference population, which consisted of a representative sample of 199 persons with a same age and sex distribution.11

Statistical methods

Descriptive statistics were performed by SPSS (version 11.5). We presented the prevalence of psychosocial problems for all patients and separately for patients in each category of the modifi ed Rankin Scale and for different anatomical locations of the aneurysm. For changes in employment, personality, memory and mood we compared the proportion of patients affected at different time periods after the SAH (2-5, >5-10 and >10 years after SAH) by means of the γ2 test for trend. Differences between groups in scores on the HADS were assessed using the Students t-test statistic. Data are reported with 95% confi dence intervals (CI).

Results

Rehabilitation

Of the 538 patients in whom data on rehabilitation were available 136 (25%, 95% CI 22-29%) had participated in a rehabilitation program. In 61 (50%) of the patients the program was residential; in 34 (28%) residential followed by outpatient; in 27 (22%) outpatient only; and in 14 patients the rehabilitation setting was unclear. The mean time patients spent in rehabili-tation was 4.8 months (95% CI 3.9-5.8, range 1-24) for the residential and 7.4 months (95% CI 6.0-8.8, range 1-24) for outpatient rehabilitation.

Work, relationship and personality changes

Before the SAH, 565 (93%, 95% CI 90-95%) of the 610 patients were employed (26% housewives and 74% patients with paid employment). After the SAH 146 (26%, 95% CI 22-30%) patients were unable to return to work. The proportion of patients that did not return to work was not statistically different for the patients 2-5, >5-10 and >10 years after SAH (30%, 26% and 25%). The changes in work status are reported in Table 2a and 2b. Before the SAH the patients who

77

Six

Empl

oym

ent

befo

re SA

HFu

ll-tim

eaf

ter S

AHPa

rt-ti

me

afte

r SAH

Hous

ewife

afte

r SAH

Stud

ent

afte

r SAH

Retir

edaf

ter S

AHUn

empl

oyed

/Unfi

taf

ter S

AHTo

tal

befo

re SA

H

Full-

time

129

570

01

9328

0 (4

5.9%

)Pa

rt-ti

me

287

20

049

140

(23.0

%)

Hous

ewife

1

313

30

08

145 (

23.8

%)

Stud

ent

11

01

01

4 (0

.7%)

Retir

ed

00

00

120

12 (2

.0%

)Un

empl

oyed

/Unfi

t2

10

00

2629

(4.8

%)

Tota

l afte

r SAH

135 (

22.1%

)14

9 (2

4.4%

)13

5 (22

.1%)

1 (0.

2%)

177 (

79.1%

)13

(2.1%

)61

0 (10

0%)

Tabl

e 2a

Wor

k sta

tus b

efor

e the

SAH

and

at th

e tim

e of t

he fo

llow

-up

inte

rvie

w

* N=

565

= al

l em

ploy

ed p

atie

nts

(stud

ents

, une

mpl

oyed

and

retir

ed p

atie

nts e

xclu

ded)

** N

=115

= 74

pat

ient

s with

dec

reas

e in

hour

s and

41 p

atie

nts

w

ith b

oth

decr

ease

in h

ours

and

leve

l of e

mpl

oym

ent

Part

icipa

nts (

%)

Chan

ges a

fter S

AH N

=565

*De

crea

se in

hou

rsIn

crea

se in

hou

rsDe

crea

se in

leve

l of e

mpl

oym

ent

Both

dec

reas

e in

hou

rs a

nd le

vel w

ork

Stop

ped

wor

king

No ch

ange

Decr

ease

in w

orki

ng h

ours

N=1

15**

< 25

% le

ss25

-50%

less

50-7

5% le

ss>

75%

less

unkn

own

74 (1

3.1%

)10

(1.8

%)

19 (3

.4%

)41

(7.3%

)14

6 (2

5.8%

)27

5 (48

.7%)

7 (6.

1%)

20 (1

7.4%

)38

(33.0

%)

19 (1

6.5%

)31

(27.0

%)

Tabl

e 2b

Cha

nges

in h

ours

and

leve

l of e

mpl

oym

ent r

elat

ed to

SAH

78

Chap

ter

had paid employment worked on average 35.2 hours per week (95% CI 33.9-36.5, range 4-80). After the SAH the mean number of working hours per week of the patients who returned to work was 30.2 (95% CI 28.7-31.8, range 2-80). On average patients returned to work 9.4 months after discharge from the hospital (95% CI 7.9-10.9, range 0-96).520 (87%, 95% CI 84-90%) of the 596 patients in whom data on relationships were available were married before the SAH. In total 75 patients (14%) got divorced after the SAH of whom 34 (7%, 95% CI 5-9%) reported that this was a result of the consequences of the SAH. In all these instances the main reported reason was that the partner could not cope with the physical or personality changes in the SAH patient. Of the 542 patients in whom data on personality were available 320 (59%, 95% CI 55-63%) reported changes in their personality after the SAH. The reported changes in personality are listed in Table 3. The proportion of patients with changes in personality was not statistically different for the patients 2-5, >5-10 and >10 years after SAH (57%, 64% and 55%). 215 (35%, 95% CI 31-39%) patients reported impairment of memory of which the majority (96%) was classifi ed as mild. There was a statistically signifi cant decrease in the proportion of patients that reported memory impairment over the time periods 2-5, >5-10 and >10 years after SAH (44%, 37% and 30%, p =0.01).The most frequent self-reported subjective symptoms related to the SAH were fatigue (17%), headaches (12%), concentration problems (12%) and dizziness (8%). Only 152 (25%, 95% CI 22-29%) of the 610 patients reported that they had completely recovered without any neurological or psychosocial symptoms.Cognitive changes were not more frequent in patients with anterior cerebral or anterior communicating aneurysms (Table 4a). In patients with a modifi ed Rankin Scale of zero only 2% reported memory changes and the proportion of patients with personality changes was lower compared with patients with higher scores on the Rankin Scale (Table 4b).

Depression and anxiety

The HADS questionnaire was completed by 584 of the 610 patients (96%). The mean HADS score for depression (Figure 2a) in patients with a history of SAH (6.2 ± 3.1) was statistically signifi cantly higher than the score for this measure in the reference population (3.4 ± 3.3) (dif-ference of means 2.8; 95% CI 2.3 to 3.3). Although the mean score for depression in patients with a history of SAH was below the threshold for ‘possible depression’, 55 (9.4%) patients had scores in the range of ‘probable depression’ (Figure 2b). The proportion of patients with scores in the range of ‘probable depression’ was not statistically different for the patients 2-5, >5-10 and >10 years after SAH (8.3%, 10.7% and 8.6%). The mean HADS score for anxiety (Figure 2a) was higher in patients with a history of SAH (5.4 ± 3.9) than in the reference population (5.1 ± 3.6) but the difference in means (0.3; 95% CI –0.3 to 0.9) was not statistically signifi cant. Although the mean score for anxiety in patients

79

Six

Participants (%)

Changes in personality N=542YesNo

320 (59%)222 (41%)

Classifi cation N=3201. Increased agitation2. Apathy3. Increased emotionality4. Positive change in personality5. Unclassifi ed

140* (37%)58 (15%)107 (29%)50 (13%)22 (6%)

Table 3 Changes in personality related to SAH

* the numbers of patients in different personality categories do not add up to 320 because some patients were classifi ed in more than one category

OverallN=610

Acom/ACAN = 261

ICAN = 168

MCAN= 137

VBN= 44

Full-time employment after SAH 135 (22%) 70 (27%) 24 (14%) 36 (26%) 5 (11%)Memory changes 215 (35%) 96 (37%) 52 (31%) 47 (34%) 20 (46%)Personality changes 320 (59%) 140 (59%) 81 (57%) 71 (57%) 28 (72%)Mean HADS anxiety score ± SD 5.4 ± 3.9 5.3 ± 3.7 5.8 ± 4.1 5.2 ± 4.2 5.3 ± 3.8Mean HADS depression score ± SD 6.2 ± 3.1 6.0 ± 2.9 6.2 ±3.1 6.2 ± 3.2 6.4 ± 3.0

Table 4a Changes in work, cognitive function and personality in relation to the location of the ruptured aneurysm

Acom / ACA = Anterior communicating artery / Anterior cerebral arteryICA = Internal carotid artery MCA = Middle cerebral arteryVB = Arteries of the vertebrobasilar system

OverallN=610

Rankin 0N = 211

Rankin 1N = 135

Rankin 2N= 244

Rankin 3N= 20

Full-time employment after SAH 135 (22%) 92 (44%) 31 (23%) 12 (5%) 0 (0%)Memory changes 215 (35%) 5 (2%) 74 (55%) 125 (51%) 11 (55%)Personality changes 320 (59%) 79 (43%) 75 (58%) 157 (72%) 9 (75%)Mean HADS anxiety score ± SD 5.4 ± 3.9 5.3 ± 3.9 5.6 ± 4.1 5.5 ± 3.9 5.9 ± 3.1Mean HADS depression score ± SD 6.2 ± 3.1 6.1 ± 2.9 6.2 ± 3.2 6.2 ± 3.1 3.9 ± 3.0

Table 4b Changes in work, cognitive function, personality and mood in relation to the Modifi ed Rankin Scale

80

Chap

ter

with a history of SAH was below the threshold for ‘possible anxiety’, 65 (11.1%) patients had a score in the range of ‘probable anxiety’ (Figure 2b).The proportion of patients with scores in the range of ‘probable anxiety’ was not statistically different for the patients 2-5, >5-10 and >10 years after SAH (12.0%, 9.4% and 11.5.%).

Scores for depression and anxiety in patients with a history of SAH (N=584) and the reference population.

* = signifi cant difference in mean depression subscore between the reference population and the patients with a history of SAH (difference of means 2.8; 95% CI 2.3 to 3.3)

Figure 2a

Number (inside box) and percentage (y-axis) of patients with a history of SAH (N=584) with a normal, possible or probable depression or anxiety scores

Figure 2b

81

Six

Discussion

Long-term psychosocial effects of SAH

We found that SAH has important long-term psychosocial consequences in patients who recovered to functional independence after SAH. Half of the patients stopped working, wor-ked shorter hours or held a position with less responsibility, one out of every 14 patients got divorced because of the SAH and two-thirds of the patients reported changes in their perso-nality. Moreover, mean depression scores were higher in patients compared with the reference population and one out of 10 patients had scores in the range of a probable depression or an anxious state. These psychosocial consequences seem to be persistent over time because except for memory impairment no signifi cant decrease in affected patients was found over the time periods 2-5, >5-10 and >10 years after SAH.

Literature on long-term follow-up after SAH

Although several studies have been published on the short-term psychosocial consequences after SAH, to our knowledge only two studies addressed psychosocial outcome more than four years after SAH.5,10 One study with a mean follow-up time of 6.5 years after SAH administered the Relative Questionnaire to caregivers of 58 SAH patients.5 Of these 58 patients 22% had a Glasgow Outcome Score (GOS) corresponding with severe disability, 48% with moderate disability and 29% with good recovery. In that study caregivers reported that half of the patients had changes in personality and that from the patients with full-time employment before the SAH half had a decrease in employment status after the SAH. A second study re-ported psychosocial outcome 4-7 years after SAH in 123 patients who survived after SAH, and in whom the mean GOS score was between good recovery and moderate disability. Of these 123 patients 20% had reduced ability to work and 48% reported changes in personality.10 The results of these two studies, however, are not easily to compare with our study because the fi rst mentioned study interviewed only caregivers and both studies included patients with non-aneurysmal causes of SAH (which were excluded from our series).

Methodological considerations

There are some limitations in our study that need to be addressed. The effects of SAH on employment status, relationships, personality and cognition were assessed by means of a self-developed questionnaire. Defi cits in personality and cognition could have been measured more objectively and in more detail by means of neuropsychological testing. We considered this, however, not feasible in this large group of patients. Although we based our results on the patients’ own perception, we think our fi ndings are relevant because they represent the problems that are experienced by patients in daily life. The purpose of our study was to

82

Chap

ter

assess the frequency of long-term psychosocial consequences after SAH. We presented the prevalence of psychosocial problems for patients in each modifi ed Rankin Scale and for dif-ferent aneurysm locations. However, detailed evaluation whether certain variables related to the SAH or treatment were correlated with long-term psychosocial outcome was beyond the scope of our study. The patients included in our study were participants of the ASTRA study. In the ASTRA study the yield of screening for new aneurysms was evaluated with CTA. The CT images of the brain secondarily derived from the CTA are not of enough quality to evaluate ischemic damage. Therefore we could not correlate the cognitive problems of our patients with possible ischemic brain damage at the time of the interview. Furthermore, we restricted our study to patients who regained functional independence and excluded patients who resided in a nursing home after SAH since follow-up in these patients is in general hard to obtain. It is likely that the psychosocial impact of SAH in those patients is much higher than in our study population. In the ASTRA study we only included patients who were treated by clipping since CTA causes large coil-artefacts and is therefore not a useful screenings tool for coiled patients. Whether our results can be generalized to patients who have been treated by coiling is uncertain. Although neuropsychological outcome appears to be primarily caused by the complications of SAH, neuropsychological defi cits seem to appear more often in clipped than in coiled patients.2,3,4 Lastly, because of the age criteria for the ASTRA study (upper age limit of 70 years at time of follow up) the population examined was somewhat younger than expected in an unselected population nine years after SAH. It is possible that the psychosocial effects of SAH are less in older patients because they already have retired from work. If this is indeed the case, we might have overestimated the frequency of the psychosocial effects.

Conclusions

We conclude that the effects of SAH on work, relationships, personality and mood are consi-derable and persist for many years after SAH. The focus of neurosurgical follow-up is often on the physical well-being of the SAH patients. In our opinion more attention should be paid to the psychosocial effects of SAH. The information of this study can help to inform and advise patients about the long-term effects after SAH and about the prognosis of reintegration to work. Appropriate support, professional counseling and treatment with antidepressants, where appropriate, are important in guiding patients with a history of SAH.

83

Six

References

1. Bamford JM, Sandercock PA, Warlow CP et al. Interobserver agreement for the assessment of handicap in stroke patients. Stroke 1989;20:828.2. Bellebaum C, Schafers L, Schoch B et al. Clipping versus coiling: neuropsychological follow up after aneurysmal subarachnoid hemorrhage. J Clin Exp Neuropsychol 2004;26:1081-1092.3. Fontanella M, Perozzo P, Ursone R et al. Neuropsychological assessment after microsurgical clipping or endovascular treatment for anterior communicating artery aneurysm. Acta Neurochir (Wien) 2003;145:867-872.4. Hadjivassiliou M, Tooth CL, Romanowski CA et al. Aneurysmal SAH: cognitive outcome and structural damage after clipping or coiling. Neurology 2001;56:1672-1677.5. Hellawell DJ, Pentland B. Relatives’ reports of long term problems following traumatic brain injury or subarachnoid hemorrhage. Disabil Rehabil 2001;23:300-305.6. Hellawell DJ, Taylor R, Pentland B. Persisting symptoms and carers’ views of outcome after subarachnoid hemorrhage. Clin Rehabil 1999;13:333-340.7. Hop JW, Rinkel GJE, Algra A et al. Case-fatality rates and functional outcome after subarachnoid hemorrhage: a systematic review. Stroke 1997;28:660-664.8. Hop JW, Rinkel GJE, Algra A et al. Quality of life in patients and partners after aneurysmal subarachnoid hemorrhage. Stroke 1998;29:798-804.9. Morris PG, Wilson JT, Dunn L. Anxiety and depression after spontaneous subarachnoid haemorrhage. Neurosurgery 2004;54(1):47-52.10. Ogden JA, Utley T, Mee EW. Neurological and psychosocial outcome 4 to 7 years after subarachnoid hemorrhage. Neurosurgery 1997;41:25-34.11. Spinhoven P, Ormel J, Sloekers PP et al. A validation study of the Hospital Anxiety and Depression Scale (HADS) in different groups of Dutch subjects. Psychol Med 1997;27:363-370.12. Zigmond AS, Snaith RP. The hospital anxiety and depression scale. Acta Psychiatr Scand 1983;67:361- 370.

PartTwo

Screening for new aneurysms

ChapterSeven

Repeated screening for intracranial aneurysms in familial subarachnoid hemorrhage

M.J.H. Wermer, G.J.E. Rinkel and J. van Gijn

88

Chap

ter

Abstract

Background and purpose

In families with two or more fi rst-degree relatives with subarachnoid hemorrhage (SAH) screening for aneurysms is often recommended. The benefi t of repeated screening and the interval at which screening should be performed are unknown. We studied patients’ compli-ance and the yield of repeated screening for familial intracranial aneurysms.

Methods

Relatives with familial SAH screened between 1990 and 1997 were advised to return every fi ve years for follow-up screening with magnetic resonance angiography (MRA). In case neurosurgical clipping had been performed in the past screening was done with computed tomographic angiography (CTA). We analyzed the results for the group as a whole, and for the subgroups of relatives with and without previous aneurysms.

Results

Of 129 relatives who were advised to undergo further screening 27 did not return, 74 had one repeat screening and 28 a second repeat screening. We detected 10 new aneurysms in nine of the 102 screened relatives (9%), three of the 19 (16%) relatives with previous aneurysms and six of the 83 (7%) relatives without previous aneurysms. One of the nine subjects with a new aneurysm and one other relative had an SAH three years after a negative screening procedure.

Conclusions

In persons with familial occurrence of aneurysms the motivation for repeated screening every fi ve years is high and the yield is considerable, particularly in relatives who have been treated for aneurysms in the past. The occurrence of SAH less than fi ve years after a negative screen suggests that screening may have to be repeated at shorter intervals.

89

Seven

Introduction

Familial clustering occurs in around 10% of patients with subarachnoid hemorrhage (SAH).1,

2 The risk of SAH in family members with two or more fi rst-degree relatives with SAH or un-ruptured aneurysms is not exactly known. Presumably, this risk is higher than in fi rst-degree relatives of patients with sporadic SAH, who have already a three to sevenfold higher risk of SAH than the general population.3

The yield of screening in persons with two or more affected relatives is high; in about 8% an aneurysm is found.4, 5 According to the guidelines of the American Heart Association, screening should be considered in such relatives.6 Because aneurysms are not congenital but develop over time, the issue of repeated screening should be considered.6, 7 Repeated screening often is already often performed in clinical practice, but the yield of repeated screening and the interval at which screening should be performed have never been properly assessed.We assessed the yield of repeated screening in members of families with familial SAH by investigating the proportion of relatives who returned for repeated screening, the number of newly detected aneurysms, the time of detection, and the characteristics of the relatives with new aneurysms.

Methods

All patients with SAH and their relatives who attend our hospital for screening are recorded in a database. We retrieved from this database all relatives who were screened for familial intracranial aneurysms between 1990 and 1997; they were all advised to undergo follow-up screening. Familial SAH was defi ned as two or more fi rst-degree relatives (parents, siblings, children) known to have SAH or unruptured aneurysms. The advice for follow-up screening at an interval of fi ve years was given to the relatives after the initial screening and was included in a letter to the family physician about the result of this initial screening. Relatives were told that they would not be invited for follow-up screening but that they had to make an ap-pointment at the outpatient clinic at their own initiative. We did not recommend follow-up screening in relatives from around the age of 65 years because the benefi t of screening at the age of 70 would probably not exceed the risk of preventive treatment. From 1993, relatives who had not previously been treated for aneurysms were screened with magnetic resonance angiography (MRA). If neurosurgical clipping had been performed in the past, screening was done with computed tomographic angiography (CTA). If an aneurysm was detected on CTA or MRA catheter angiography was performed for confi rmation of the CTA or MRA fi ndings and for planning the optimal treatment, i.e. surgical clipping or endovascular coiling. Aneurysms smaller than four millimeters in patients without previous SAH were fol-lowed up with MRA or CTA over time.

90

Chap

ter

We recorded how many relatives visited our outpatient clinic for repeated screening. If relatives had not returned six years after the initial screening, we classifi ed them as nonattendants. Relatives who had not returned after fi ve years but had not passed the 6-year interval were classifi ed as pending because some relatives contacted us between fi ve to six years after the initial screening. For all relatives who had not returned for follow-up screening, we contacted the family physician to fi nd out if the person was still alive, had had an SAH after the initial screening procedure, or had been screened elsewhere. We classifi ed the newly developed aneurysms as de novo (aneurysm located at a site that previously showed no abnormalities), regrowth (aneurysm located at the site of an originally treated aneurysm), or additional (in relatives with previous aneurysms in whom the second aneurysm was seen in retrospect but had not been identifi ed on initial screening with MRA or CTA). We constructed pedigree trees for all families. The characteristics of relatives with new aneurysms in terms of degree of kinship, age, sex, and previous SAH or unruptured aneurysms were compared with those of relatives without aneurysms detected on repeated screening. We calculated proportions and relative risks with corresponding confi dence intervals for the group as a whole, and for the subgroups with and without previous aneurysms (ruptured or unruptured).

Results

In total, 129 relatives in 26 families with familial SAH had been advised to undergo follow-up screening every fi ve years. Of these 129 relatives, 102 (79%) from 19 families actually returned for repeated screening; 42 were men and 60 women. The mean age of the screenees at the time of the initial contact was 37 years (range 18-62 years). The mean number of relatives per family was fi ve (range 1-17). Of the 102 relatives, 57 were siblings of affected relatives, 33 were children, 10 had an affected parent and an affected sibling, one was a parent and one had an affected parent and an affected child. One relative was known to have autosomal domi-nant polycystic kidney disease (ADPK). Nineteen relatives had been treated for aneurysms in the past; 11 after SAH and eight for an asymptomatic aneurysm. Seventy-four relatives had a single follow-up investigation, 28 had two follow-up screens. The total follow-up time between the initial screen and the fi rst follow-up screen was 552.3 years (5.4 years/relative). The total follow-up time between the fi rst and second repeated screens was 104.3 years (3.7 years/relative). Of the 27 relatives who had not returned, eight relatives were pending. All 27 relatives were found to be alive; one of them had been screened in a hospital closer to her residence, reportedly with normal results. Twenty-six had no SAH after the initial screening; for one relative, this information was lacking. The characteristics of the relatives who did not return for screening were similar to those of the patients who did return except that only one of the 27 nonattendants had been treated for an aneurysm in the past.

91

Seven

New aneurysms detected with screening

At follow-up screening, we detected 10 aneurysms in nine (9%) of the 102 repeatedly screened family members. Of the 10 detected aneurysms, three were treated with coiling and three with clipping. No complications occurred during treatment, and all treated patients had a good recovery. The remaining four aneurysms were smaller than four millimeters and are followed up over time. In all patients, the angiogram performed after the positive screening confi rmed the fi ndings on MRA or CTA. In the patient with two new aneurysms, the second one was initially not detected on the CTA but was found on the angiogram performed for the other new aneurysm (Table 1, patient 7). The number of relatives needed to screen per fi ve years to fi nd one aneurysm was 13. The characteristics of the nine subjects and the newly developed aneurysms are shown in Table 1.

Relatives without previous history of SAH or unruptured aneurysms

Six of the 10 new aneurysms were found in six (7%) of the 83 relatives without previous aneu-rysms. In three patients, a new aneurysm was detected at the fi rst follow-up investigation and in three patients at the second. One of the aneurysms detected at the second visit was in retrospect visible on the MRA fi ve years before but not on the initial MRA; in the interval it had grown from 4 to 9 millimeters. In another patient, an aneurysm found on the fi rst fol-low-up screen could in retrospect be identifi ed on the initial MRA; this aneurysm had grown from 5 to 7 millimeters over a period of six years.

Relatives with previous history of SAH or unruptured aneurysms

Four of the 10 new aneurysms were found in three (16%) of the 19 relatives with previous aneurysms (ruptured or unruptured). All aneurysms were detected at the fi rst follow-up screening. One new aneurysm was seen in retrospect on the angiogram that had been performed at the time of the SAH and was therefore classifi ed as additional. This aneurysm had not increased in size in the 5-year interval. One aneurysm developed adjacent to the clip placed at an earlier operation and was classifi ed as a regrowth. The remaining two aneurysms were classifi ed as de novo.

Subarachnoid hemorrhage in the interval between two screening procedures

In one relatives without previous aneurysms, an SAH occurred in the interval between the two screening procedures. A new aneurysm of the anterior communicating artery was found and treated with neurosurgical clipping (Figure 1). Four years after the SAH, a new aneurysm was detected on follow-up screening. This small aneurysm of the middle cerebral artery was clipped, and the patient again made a good recovery (Table 1, patient 4).

92

Chap

ter

NoBi

rth

yrSe

xHi

stor

ySA

H or

UA

Degr

ee o

f ki

nshi

pIn

itial

scre

enRe

sult

First

fo

llow

-up

Resu

lt Se

cond

fo

llow

-up

Resu

ltAn

eury

smcla

ssifi

catio

nSi

ze

mm

Trea

tmen

t

119

45F

NoSi

b+Ch

ild19

93Ne

gativ

e19

97Ne

gativ

e20

02P

com

ADe

Nov

o9

x 5

Coili

ng

219

45F

NoSi

b19

95Ne

gativ

e20

01M

CA A

--

De N

ovo

*7

Coili

ng

319

72F

NoCh

ild19

90Ne

gativ

e19

95Ne

gativ

e20

01Ba

silar

ADe

Nov

o3

Follo

w-u

p

4 **

1957

FNo

Sib

1995

Nega

tive

2000

Nega

tive

2002

MCA

ADe

Nov

o2

Clip

ping

519

50F

NoSi

b19

95Ne

gativ

e20

03M

CA A

--

De N

ovo

3Fo

llow

-up

619

48M

NoSi

b19

95Ne

gativ

e20

00P

com

A-

-De

Nov

o4

Clip

ping

719

56F

UASi

b19

96

4 M

CA A

2001

A co

m A

MCA

A-

-Ad

ditio

nal

De N

ovo

3 x 4

2Fo

llow

-up

819

49F

SAH

Sib

1993

SAH

A co

m A

2001

ICA

A-

-De

Nov

o5

Coili

ng

919

29M

SAH

Sib

1980

SAH

MCA

A19

93M

CA A

1997

Nega

tive

Regr

owth

15Cl

ippi

ng

Tabl

e 1 L

ocat

ion

of a

neur

ysm

s and

char

acte

ristic

s of p

atie

nts w

ith n

ewly

det

ecte

d an

eury

sms

SAH

/ UA

= su

bara

chno

id h

emor

rhag

e / u

nrup

ture

d an

eury

smIn

itial

scre

en/ fi

rst f

ollo

w-u

p/ se

cond

follo

w-u

p= ye

ar o

f ini

tial/

fi rst

follo

w-u

p/ se

cond

follo

w-u

p sc

reen

ing

Size m

m =

size

of t

he a

neur

ysm

in m

illim

eter

sF /

M=

fem

ale /

mal

e4

= nu

mbe

r of a

neur

ysm

sIC

A A

= an

eury

sm o

f the

inte

rnal

caro

tid a

rter

yM

CA A

= a

neur

ysm

of t

he m

iddl

e cer

ebra

l art

ery

A co

m A

= a

neur

ysm

of t

he co

mm

unica

ting

ante

rior c

ereb

ral a

rter

yP

com

A =

ane

urys

m o

f the

com

mun

icatio

n po

ster

ior c

ereb

ral a

rter

yBa

silar

A =

ane

urys

m o

f the

bas

ilar a

rter

y*

this

aneu

rysm

was

in re

trosp

ect v

isibl

e on

the i

nitia

l scr

een

of 19

95**

pa

tient

4 h

ad a

n SA

H in

1998

from

an

aneu

rysm

of t

he a

nter

ior c

omm

unica

ting

arte

ry d

espi

te n

egat

ive s

cree

ning

in 19

95 (F

igur

e 1)

93

SevenOne relative with SAH fi ve years before had a single follow-up screening that showed no abnormalities. Three years later, she was admitted to another hospital with a CT proven SAH. She died, and at autopsy an aneurysm of the vertebral artery was found. In contrast with the other patient who had an SAH in the interval between the two screening procedures this aneurysm was in retrospect visible on the initial angiogram.

Risk factors for development of new familial intracranial aneurysms

A history of ruptured or unruptured aneurysms was associated with a relatively high risk of the development of new aneurysms. A new aneurysm developed in three (16%) of the 19relatives with previous aneurysms and six (7%) of the 83 relatives without previous aneurysms. Women had a higher risk than men for developing aneurysms, and siblings had a higher risk than children and parents (Table 2).

A woman born in 1957 had an initial negative screening with MRA in 1995 (left). Three years later she was admitted to another hospital with an SAH from an aneurysm of the anterior communicating artery (right). The aneurysm was in retrospect not visible on the fi rst MRA. She was treated with clipping and made a good recovery.

Figure 1

94

Chap

ter

Subgroup of relatives Number of relatives Number of relativeswith new aneurysms Relative Risk (95% CI)

No previous aneurysmPrevious aneurysm

8319

6 (7%, CI 3-16%)3 (16%, CI 4-40%)

Reference2.2 (0.6 to 8.0)

Children*Siblings*Parents*

44672

2 (5%)7 (10%)0

Reference2.3 (0.5 to 10.6)-

MenWomen

4260

2 (5%)7 (12%)

Reference2.6 (0.6 to 11.8)

< 40 years**40-60 years> 60 years

60402

4 (7%)5 (13%)0

Reference1.9 (0.5 to 6.6)-

Table 2 Relative risk for the development of new aneurysms or aneurysms detected after growth in subgroups of relatives

CI = Confi dence Interval* = Cumulative number since relatives can be classifi ed in more than one category** = Age at initial screening

A woman born in 1949 was admitted in 1993 with an SAH from an aneurysm of the anterior communicating artery. An angiogram in 1993 did not show any additional aneurysms (left). Screening in 2001 with CTA showed an internal carotid aneurysm (middle). She was treated with coiling and made a full recovery (right).

Figure 2

Discussion

We found that the yield of repeated screening for familial intracranial aneurysms is high; in 16% of the relatives with previous aneurysms and in 7% of the relatives without previous aneurysms new aneurysms were detected, mostly within fi ve years. In addition, two rela-tives had an SAH in the 5-year interval between two screening procedures; one from a newly developed aneurysm and one from an additional aneurysm that was missed on the initial angiogram and the fi rst follow-up screen. These results indicate that persons with two or more

95

Seven

fi rst-degree relatives with SAH are at high risk for the development of new aneurysms.There was a trend for an increased risk of new aneurysm formation in relatives with aprevious aneurysm, women, siblings, and relatives between 40 and 60 years of age but these results were not statistically signifi cant. Because we had no data on smoking habits and hypertension in most of the relatives, we were not able to assess the impact of these risk factors in our study.Seventy-nine percent of the relatives who were advised to return for follow-up screening actually returned. This proportion is high. We could not fi nd attendance rates of comparable screening programs in other medical fi elds. One explanation for the high compliance might be pressure on the relatives from other family members. The motive for the nonattendance remains unknown because we considered it unethical to contact them and ask why they did not return for screening.In our study, we tried to avoid any bias. All relatives were informed and treated in the same way and only one nonattendant was partly lost to follow-up (we knew she was alive but we had no information on whether she had a SAH after the initial screening). In addition, relatives of a large number of families were screened, which increases the generalizability of the results. This is the fi rst study to assess the yield of repeated screening in a large series of subjects with two or more fi rst-degree relatives with SAH or unruptured aneurysms. Three decision analyses have been performed on the issue of screening for familial aneurysms, but none has specifi cally or properly addressed repeated screening in persons with two or more affected fi rst-degree relatives.8-10

Screening for aneurysms carries benefi ts but also risks. Screening can prevent new episodes of SAH but can lead to disability and death from preventive treatment. With the ongoing improvement in MRA and CTA techniques, more small aneurysms will be detected that will often not be treated but followed up over time. The knowledge of having an untreated aneurysm will negatively infl uence the quality of life.11 On the other hand, with the advent of endovascular treatment, many unruptured aneurysms can now be treated with relatively low risk of complications.12 The outcome after sporadic SAH is still very poor and probably is even worse in familial cases.13 Until now no population-based clinical study has assessed the risks and benefi ts or the cost-effectiveness of repeated screening in familial SAH. The high yield of repeated screening in this study is a factor in favor of screening and indicates that repeated screening should be considered for every relative in familial SAH. The appropriate interval at which repeated screening should be performed has not yet been established. It is often assumed that aneurysms need some years to develop. In our study, two relatives were admitted for SAH within three years after a negative screening. Others have reported a member of a family with intracranial aneurysms in whom SAH occurred only two years after a negative angiography.14 These three examples show that screening at an interval of fi ve years is not suffi cient to detect all new familiar aneurysms. In some families, screening may have to be performed at intervals shorter than an arbitrary period

96

Chap

ter

of fi ve years. The follow-up interval may have to be tailored to individual relatives according not only to the family but also to individual risk factors. Smoking, hypertension and female sex are important risk factors for the growth and development of intracranial aneurysms.15 Although we are not yet able to identify all characteristic features indicating high risk of rapid aneurysm formation, individual risk factors should be taken into account in determining the interval of repeated screening.We conclude that repeated screening for familial intracranial aneurysms has a high yield, especially in relatives with previous SAH. The attendance for repeated screening is high. The interval at which screening should be performed remains uncertain and may have to be tailored according to phenotype, genotype and additional risk factors.

97

Seven

References

1. van Gijn J, Rinkel GJE. Subarachnoid haemorrhage: diagnosis, causes and management. Brain 2001; 124:249-78.2. Schievink WI. Genetics of intracranial aneurysms. Neurosurgery 1997; 40:651-62.3. Bromberg JE, Rinkel GJE, Algra A, Greebe P, van Duyn CM, Hasan D, Limburg M, ter Berg HW, Wijdicks EF, van Gijn J. Subarachnoid haemorrhage in fi rst and second degree relatives of patients with subarachnoid haemorrhage. BMJ 1995; 311:288-9.4. Raaymakers TW, Rinkel GJE, Ramos LM. Initial and follow-up screening for aneurysms in families with familial subarachnoid hemorrhage. Neurology 1998; 51:1125-30.5. Ronkainen A, Hernesniemi J, Puranen M, Niemitukia L, Vanninen R, Ryynanen M, Kuivaniemi H, Tromp G. Familial intracranial aneurysms. Lancet 1997; 349:380-4.6. Bederson JB, Awad IA, Wiebers DO, Piepgras D, Haley EC, Jr., Brott T, Hademenos G, Chyatte D, Rosenwasser R, Caroselli C. Recommendations for the management of patients with unruptured intracranial aneurysms: A Statement for healthcare professionals from the Stroke Council of the American Heart Association. Stroke 2000; 31:2742-50.7. Rinkel GJE, Djibuti M, van Gijn J. Prevalence and risk of rupture of intracranial aneurysms: a systematic review. Stroke 1998; 29:251-6.8. Dippel DW, ter Berg JW, Habbema JD. Screening for unruptured familial intracranial aneurysms. A decision analysis. Acta Neurol Scand 1992; 86:381-9.9. Crawley F, Clifton A, Brown MM. Should we screen for familial intracranial aneurysm? Stroke 1999; 30:312-6.10. Leblanc R, Worsley KJ, Melanson D, Tampieri D. Angiographic screening and elective surgery of familial cerebral aneurysms: a decision analysis. Neurosurgery 1994; 35:9-18.11. van der Schaaf IC, Brilstra EH, Rinkel GJE, Bossuyt PM, van Gijn J. Quality of life, anxiety, and depression in patients with an untreated intracranial aneurysm or arteriovenous malformation. Stroke 2002; 33:440-3.12. Brilstra EH, Rinkel GJE, van der Graaf Y, van Rooij WJ, Algra A. Treatment of intracranial aneurysms by embolization with coils: a systematic review. Stroke 1999; 30:470-6.13. Bromberg JE, Rinkel GJE, Algra A, Limburg M, van Gijn J. Outcome in familial subarachnoid hemorrhage. Stroke 1995; 26:961-3.14. Schievink WI, Limburg M, Dreissen JJ, Peeters FL, ter Berg HW. Screening for unruptured familial intracranial aneurysms: subarachnoid hemorrhage 2 years after angiography negative for aneurysms. Neurosurgery 1991; 29:434-7.15. Juvela S, Poussa K, Porras M. Factors affecting formation and growth of intracranial aneurysms: a long-term follow-up study. Stroke 2001; 32:485-91.

Psychosocial impact of screening for intracranial aneurysms in relatives with familial subarachnoid

hemorrhage

M.J.H. Wermer, I.C. van der Schaaf, P. van Nunen, P.M.M. Bossuyt, C.S. Anderson and G.J.E. Rinkel

ChapterEight

100

Chap

ter

Abstract

Background and purpose

In families with two or more relatives with intracranial aneurysms (IAs), screening for IAs in asymptomatic fi rst-degree relatives is often recommended. We assessed the long-term psychosocial impact of such screening.

Methods

We identifi ed all persons with IA (screen-positives) and matched them for age and sex with two controls without IA (screen-negatives) from hospital-based registers of familial IA. Persons underwent telephone interviews using questionnaires that covered the areas of psychosocial impact of screening, health-related quality of life (HRQoL), and mood. Data were compared between screen-positives and screen-negatives, and with reference populations.

Results

Overall, 105 persons from 33 families with IA were included, of whom 35 where screen-positive and 70 were screen-negative. Of the screen-positives, 12 (44%) had reduced their work and 23 (66%) had experienced changes in one or more areas of independence, self-esteem, future outlook, or personal relationships. In contrast, only one (2%) screen-negative person had stop-ped working and 12 (17%) others had experienced changes in their self-esteem, future outlook, or relationships. Screen-positives had lower HRQoL compared with screen-negatives and the reference population, whereas both screen groups had higher mean depression scores than the reference population. Despite these effects, only three persons regretted participating in screening.

Conclusions

Although screening for IA is an important preventative strategy in high-risk individuals, it is associated with considerable psychosocial effects, both positive and negative. Greater awa-reness of such outcomes, and appropriate intervention where necessary, would appear to be a necessary component of IA screening programs.

101

Eight

Introduction

Persons with two or more fi rst-degree family members with a history of intracranial aneurysms (IAs) are at increased risk of subarachnoid hemorrhage (SAH).1 Given that screening of the cerebral circulation may detect asymptomatic IAs in approximately 8% of the relatives,2 and the yield of repeated screening fi ve years after a negative screen is also high in this group,3 various guidelines such as those of the American Heart Association recommend that screening should be considered in relatives with familial SAH.4

The yield of screening for IAs and the risks associated with diagnostic interventions and preventive treatment are well accepted. Less well understood, however, are the psychosocial consequences of screening for IA because no formal assessments of such outcomes have been conducted to date. Indirect evidence from screening programs in breast cancer and abdominal aortic aneurysms indicate the potential benefi ts of a positive result of screening could be offset by depression and reduced quality of life in persons,5 whereas a screen-negative result may be associated with such relief that it can increase the psychosocial well-being of participants.6

We aimed to determine the psychosocial effects of screening for familial IAs by comparing aspects of social well-being and lifestyle, health-related quality of life (HRQoL), and mood symptoms in persons with positive screens with persons with normal screens, and with such measures in the general population.

Methods

We included relatives who had been screened for IAs between 1993 and 2003 because of fa-milial SAH from registers at two institutions; the Academic Medical Center Amsterdam and the University Medical Center Utrecht. Familial SAH was defi ned as two or more fi rst-degree family members with SAH or unruptured IAs. We identifi ed all persons with an IA detected on screening (screen-positives), and selected for each person with a positive screen, two con-trols from the group of relatives in whom no IAs were detected (screen-negatives). The two groups were matched for age (±3 years) and sex. Potential control persons who declined to participate, could not be contacted by telephone after several attempts, or had moved outside the area, were replaced by other controls. In all persons, the screening had to be performed at least one year ago.Some of the participants had participated in another study on screening for familial IA and had provided written informed consent for follow-up contact and administration of ques-tionnaires about their health. The others had consented at their last visit to an outpatient clinic for further contact and participation in a research project regarding the outcome from IA screening.

102

Chap

ter

Participants underwent standardized, semi-structured, telephone interviews with one of two researchers (MW, PvN), who administered four questionnaires: one developed specifi cally for the purposes of the study, two questionnaires assessing HRQoL and the Hospital Anxiety and Depression Scale (HADS). The specifi cally developed questionnaire contained a series of questions on health and lifestyle covering aspects of behavior, work, health insurance, relationships, reproductive decision-making, self-esteem and level of independence, all of wich may have been affected by the screening. Each question could only be answered by “yes” or “no”. If the answer to the question was yes the subsequent question was whether the outcome was attributable to screening (again “yes” or no”). The two researchers did fi ve of the interviews together at the start and in the middle of the study to avoid possible interobserver variation.As HRQoL instruments we used the validated Dutch versions of the Medical Outcome Study 36-item Short Form questionnaire (SF-36) and the EuroQol (EQ-5D).7, 8 The SF-36 comprises 36 questions covering eight domains: physical functioning (PF), role limitations because of physical problems (RP), bodily pain (BP), general health (GH), vitality (VT), social functioning (SF), role limitations because of emotional problems (RE), and mental health (MH). An absence of problems in a domain is indicated by a score of 100 for the PF, RP, BP, SF, and RE domains, and of 50 for the GH, VT, and MH domains. For example, a score of 100 in PF indicates an abi-lity to perform all activities without limitations attributable to ill-health, whereas a score of 50 in MH indicates an ability to function without personal or emotional problems. To obtain scores of greater than 50 for GH, VT, and MH, a subject must rate his/her health positively; for example a score of 100 in MH would indicate that the subject feels ‘peaceful and happy and is calm all the time’. The EuroQol contains fi ve questions addressing mobility, daily activities, pain, self-care, and anxiety and depression, with responses converted to an overall score ran-ging from zero (worst) to 100 (best), and a visual analog scale (EQ-VAS) where persons can rate their overall health state by placing a mark on a vertical scale which ranges from zero (worst imaginable health state) to 100 (best imaginable health state). For the purposes of this study, persons were asked to ‘rate their overall health on a scale from zero to 100’. The HADS9, a simple screening tests for symptoms of anxiety and depression, contains 14 questions on mood that provide separate rating scales for anxiety and depressive symptoms, ranging from zero to 21, to allow the classifi cations of ‘normal’ (scores <8), ‘possible depression or anxiety’ (scores of 8-10) and ‘probable depression or anxiety’ (scores >10).10

Data on screen-positives were compared with those of screen-negatives, and with Dutch reference populations, which consisted of a representative sample of 199 persons for the HADS9, 4172 persons for the SF-367 and 113 persons for the EuroQol (Stolk et al, unpublished data 2004). Responses to the specially developed health and lifestyle questionnaire were only compared between the groups of persons who had undergone IA screening using the γ2

test statistic. Differences between groups in scores on the HADS, the SF-36, and the EuroQol were assessed using the Students t test statistic. All data are reported with 95% confi dence intervals (CI).

103

Eight

Results

Overall, 39 persons with IAs identifi ed through screening for familial SAH were listed on re-gistries. Three of them had moved since the last screening and their address was unknown, and one person declined to participate, leaving 35 persons who were included in this study. Seven of these persons had a small IA and were managed conservatively; the remaining 28 had been treated for the detected IA.There were 184 screen-negative relatives, from which 70 age- and sex-matched controls were selected. For fi ve persons who could not be contacted and two others who declined partici-pation, replacement persons of the same sex and comparable age were found. Thus, a total of 105 persons were included in this study; they came from 33 different families (Table 1).

Screen-positives (n=35) Screen-negatives (n=70)Mean age±SD (range) 51.8 ± 11.7 (26-74) 51.7 ± 11.9 (24-78)Women (%) 23 (66%) 46 (66%)Mean±SD number of screens 1.8 ± 0.9 1.7 ± 0.7Mean±SD time after fi rst screening, years 8.2 ± 1.3 8.5 ± 1.5Mean±SD time after last screening, years 3.7 ± 3.0 4.9 ± 2.9

Table 1. Baseline characteristics of the 105 participants, according to screen status for the presence of familial intracranial aneurysm (IA).

Lifestyle, work, social and emotional outcomes

Of 27 screen-positives who were in full-time employment at the time of screening, 12 (44%) had stopped working or reduced their hours after screening (Table 2). In contrast, only one (2%) of 53 employed screen-negatives had stopped working, but this was done voluntarily because the ‘healthy’ outcome had lead to a re-evaluation of priorities in life. Three (9%) screen-positives reported an increase in health insurance payments after screening.Altogether, two thirds of screen-positives reported changes in their social circumstances and emotional well-being, including a decrease in their level of independence, reduced self-esteem, alteration in their ability to cope with stress, or a change in outlook on the future. At least one of these changes was of a negative nature in 21 (60%) of the respondents. Four subjects reported negative effects on relationships, although another seven said that their relation-ships had improved because the bond with a partner had tightened as a consequence of the

104

Chap

ter

Variable Screen-positives n=35 Screen-negatives n=70

Change in lifestyle, any Smoking (stopped/cut down) Alcohol (stopped/cut down) Diet (less fat, salt) More physical exercise Start anti-hypertensive drugs Start cholesterol lowering drugs

17/35 (49%)1

10/21 (48%)3/20 (15%)6/35 (17%)2/35 (6%)0/35 (0%)0/35 (0%)

9/51 (13%)5/25 (20%)2/37 (5%)2/70 (3%)1/70 (1%)1/70 (1%)1/70 (1%)

Work stopped / working less hours total Quit job voluntarily Retirement / fl exible retirement Fired from job General disability act

Insurance/drivers license Denied medical/life insurance Increased insurance payments Loss/Restriction of drivers license

12/27 (44%)2

7/27 (26%)4/27 (11%)0/27 (0%)1/27 (3%)

0/35 (0%)3/35 (9%)1/35 (3%)

1/53 (2%)1/53 (2%)0/53 (0%)0/53 (0%)0/53 (0%)

0/70 (0%)0/70 (0%)0/70 (0%)

Social-emotional changes, anyNegative total Decrease in level of independence Decrease in self-esteem Negative impact on relationship Worse coping with stress Negative future outlook More worries Impact on reproductive decisions

Positive total Increase in level of independence Increase in self-esteem Positive impact on relationship Improved coping with stress Positive future outlook

23/35 (66%)3

21/35 (60%)7/35 (20%)4/35 (11%)4/35 (11%)10/35 (29%)9/35 (26%)10/35 (29%)1/35 (3%)

13/35 (37%)0/35 (0%)1/35 (3%)7/35 (20%)6/35 (17%)4/35 (11%)

12/70 (17%)7/70 (10%)0/70 (0%)0/70 (0%)1/70 (1%)1/70 (1%)3/70 (4%)7/70 (10%)0/70 (0%)

5/70 (7%)0/70 (0%)1/70 (1%)1/70 (1%)1/70 (1%)3/70 (4%)

Fears in daily activities, any Fear of fl ying in an airplane Fear of driving a car Fear of sexual intercourse Fear of intensive exercise

9/35 (26%)4

2/35 (6%)3/35 (9%)1/35 (3%)1/35 (3%)

0 (0%)0/70 (0%)0/70 (0%)0/70 (0%)0/70 (0%)

Regrets of participation No Yes Neutral

31/35 (88%)2/35 (6%)2/35 (6%)

67/70 (96%)1/70 (1%)2/70 (3%)

Table 2 Changes in lifestyle, behavior, work and psychosocial functioning because of screening for intracranial aneurysms (IA), in screen-positives and screen-negatives

Numerator: number of persons that changed their lifestyle/work/social-emotional behavior because of screeningDenominator: number of persons at risk (number of smokers/alcohol users/ employed persons or total group)1=difference screen-positives and screen-negatives 36% (95% CI 15 to 56%)2=difference screen-positives and screen-negatives 42% (95% CI 24 to 61%)3=difference screen-positives and screen-negatives 49% (95% CI 28 to 69%)4=difference screen-positives and screen-negatives 26% (95% CI 9 to 42%)

105

Eight

fi ndings. One person decided not to have a second child because of the detection of an IA. Of the 70 screen-negatives, 12 (17%) reported changes in their social or emotional well-being but in only seven (10%) was this of a negative nature.There were 17 (49%) of 35 screen-positives and only nine (13%) of 70 screen-negatives who said that they had changed to a healthier lifestyle as a consequence of screening.Nine (26%) of 35 screen-positives reported that they avoided certain daily activities because of concern that they might lead to IA formation or rupture. Despite these changes, only two (6%) screen-positives and one (1%) screen-negative said that they regretted participating in the screening program. Both screen-positives who regretted screening were relatives in whom the aneurysm was left untreated.

Health Related Quality of Life

Screen-positives were similar to the reference population with regard to overall scores on the EuroQol (difference of means 5.7; 95% CI -3.1 to 14.5), but different from screen-negatives (dif-ference of means 13.9; 95% CI 3.5 to 24.3) (Table 3). Screen-negatives had signifi cantly better overall HRQoL than the reference population (difference of means 8.2; 95% CI 1.6 to 14.8). On the EQ-VAS, screen-positives evaluated their overall health state as lower than that of screen-negatives (difference of means 5.6; 95% CI -2.1 to 13.3), and of the reference population (difference of means 8.3; 95% CI 1.7 to 14.8). Screen-negatives also tended to evaluate their overall health lower than the reference population (difference of means 2.7; 95% CI –1.9 to 7.2). Compared with the reference population, screen-positives performed signifi cantly worse on the SF-36 domains of bodily pain (difference of means 14.5; 95% CI 6.3 to 22.6) and vitality (difference of means 6.5; 95% CI 0.1 to 12.9), whereas screen-negatives performed signifi cantly better on vitality (difference of means 6.4; 95% CI 1.8 to 10.9), social functioning (difference of means 7.4; 95% CI 2.3 to 12.5), role emotional (difference of means 10.2; 95% CI 2.5 to 17.9) and mental health (difference of means 9.9; 95% CI 3.7 to 14.0) (Figure 1).There were no differences in quality of life between the seven untreated screen-positives and the 28 treated screen-positives.

Anxiety and Depression

Mean HADS scores for depression (Figure 2a) were similar for screen-positives and screen-negatives (5.6±2.7 versus 5.6±2.2), but they were statistically signifi cantly higher than in the reference population (3.4±3.3; difference of means 2.2; 95% CI 1.1 to 3.4). However, the mean scores for depression in both IA subject groups were below the threshold for ‘possible de-pression’ and only two (5.7%) screen-positives and three (4.3%) screen-negatives had scores in the range of ‘probable depression’ (Figure 2b). The mean HADS score for anxiety (Figure 2a) was lower in the screen-negatives (4.2±3.3) than in the screen-positives (5.0±4.1) and in

106

Chap

ter

EQ-5D index score* EQ-VAS†

Screen-positives (n=35)

Screen-negatives (n=70)

Reference population (n=113)

73.3 (28.4)

87.2 (23.5)

79.0 (21.0)

67.2 (23.6)

72.8 (15.9)

75.5 (14.6)

Table 3 Effects on HRQoL according to the EuroQol summary scores (EQ-ED) and on the visual analogue scale (EQ-VAS).

Scores, ranging from zero (worst health state) to 100 (best health state), are presented as means (±SD)*Signifi cant differences in EQ-5D scores: screen-negatives versus screen-positives (difference of means 13.9, p=0.01), screen-negatives versus reference population (difference of means 8.2, p=0.02)†Signifi cant differences in EQ-VAS scores: screen-positives versus reference population (difference of means 8.3, p=0.01)

Figure 1 Mean SF-36 scores for each domain separately for the relatives with an aneurysm detected at screening (positive screening), relatives with a normal outcome of screening (negative screening) and the reference population. A higher SF-36 score indicates a higher quality of life

PF= Physical Function; RP=Role-Physical; BP=Bodily Pain; GH=General Health; VT=Vitality; SF=Social Functioning; RE=Role-Emotional; MH=Mental Health.* = signifi cant differences in SF-36 domains:BP: screen-positives versus reference population (difference of means 14.5, p=0.0005)VT: screen-positives versus reference population (difference of means 6.5, p=0.05) screen-negatives versus reference population (difference of means 6.4, p=0.01)SF: screen-negatives versus reference population (difference of means 7.4, p=0.004)RE: screen-negatives versus reference population (difference of means 10.2, p=0.01)MH: screen-negatives versus reference population (difference of means 9.9, p<0.0001)

107

Eight

the reference population (5.1±3.6), although the mean difference of 0.9 (95% CI, -0.1 to 1.82) between screen-negatives and the reference population was not statistically signifi cant. Only three screen-negatives (4.3%) had a score in the range of ‘probable’ anxiety, compared with fi ve (14.3%) screen-positives (Figure 2b). The seven untreated screen-positives had comparable depression subscores but slightly higher anxiety subscores (mean 6.1) than the treated screen-positives, but this difference was not statistically signifi cant.

Figure 2a Scores for depression and anxiety in relatives with an aneurysm detected at screening (positive screening), relatives with a normal outcome of screening (negative screening) and the reference population.* = signifi cant difference in mean depression subscore between the reference population and the screen-positives and the screen-negatives (difference of means 2.2, p<0.001)

Figure 2b Number (inside box) and percentage (y-axis) of participants with normal, possible or probable depression or anxiety scores

depr

essi

onpo

sitiv

esc

reen

ing

depr

essi

onne

gativ

esc

reen

ing

anxi

ety

posi

tive

scre

enin

g

anxi

ety

nega

tive

scre

enin

g

108

Chap

ter

Discussion

Screening for IAs can have a signifi cant impact on lifestyle and psychosocial well-being of persons long after the results are known. The most pronounced effects were seen in those persons who screened positive for IA, with many reporting adverse outcomes in aspects of lifestyle, behavior, work and social and emotional functioning. Even so, up a half of these persons also reported changing their behavior toward healthier living. Similar effects were reported in less than one in 10 of the screen-negatives. The HRQol in screen-positive persons was lower than the HRQol in screen-negatives and the reference population.There are no previous published data on the psychosocial effects of screening for IAs. Indirect evidence from other screening programs indicate that a normal result may be reassuring and be associated with an increase in health and well-being of participants.6 The higher scores in certain aspects of HRQoL and lower anxiety levels in screen-negatives compared with a reference population are consistent with fi ndings in other patient groups. With regard to our assessment of work, lifestyle, and social and emotional functioning, we often found negative effects on self-esteem, coping with stress, and future outlook among screen-positives, whereas these areas of life were more often viewed more favorably for screen-negatives. These results are comparable with the psychosocial impact of genetic screening in other neurological di-seases such as Huntington’s disease, ataxias and neuromuscular disorders.11, 12 The psychosocial impact of screening in relatives with a detected but untreated aneurysm may also vary according to the information given about the rupture risk of the aneurysm. All relatives, however, including the seven untreated ones, were seen by one of two neurologists who gave all information on screening and aneurysms as neutrally as possible.The results of the questionnaires did not differ signifi cantly between the untreated and treated screen-positives except that both screen-positives who regretted screening had an untreated aneurysm.The strengths of this study lie in a large and representative sample of persons, with only a few persons lost to follow-up and a high level of participation. Because the baseline characteristics of non-participants were comparable to that of participants, selection bias is likely to be low. Most of the data were obtained using simple and direct questions and with well-validated and reliable instruments. Several limitations must be acknowledged. This was a cross-sectional study undertaken long after the results of screening were known to persons. Therefore, we were unable to assess the early effects of IA screening, when anxiety and depressive symptoms are more likely to occur. In addition, because of the cross-sectional design, we did not have assessments of HRQol and HADS scores from before screening was performed. Another limitation was that the reference population for SF-36 measures was slightly younger (±8 years) and contained fewer women (54% versus 66%) than our IA screen persons, and we did not have age- and sex-specifi c data from the reference population to make adjustments.

109

Eight

The validation study of this reference population showed only modest differences between men and women and between the age categories 41 to 60 and 61 to 70 years.12 Therefore, it seems unlikely, that comparisons of the crude SF-36 data would have had a major impact on the results. Finally, the reference population did not consist of persons with familial SAH that had not participated in an IA screening program. It is possible that asymptomatic persons with fami-lial IA who do not participate in screening have a lower HRQol than a reference population without familial SAH. Knowledge of the presence of an unruptured IA is associated with a decrease in quality of life.13 Although nonparticipants have no such knowledge, they probably realize that they are at increased risk of harboring an aneurysm. Furthermore, the experience of SAH related death or disability in close relatives might have an impact on the quality of life of asymptomatic relatives. Therefore, we recognize that we might have underestimated the positive HRQoL difference for screen-negatives and overestimated the negative HRQoL difference for screen-positives, against the reference population. The psychosocial impact of IA detection among those who participate in a screening pro-gram appears wide-ranging and considerable. Given the major consequences of ruptured IAs, including death in half and disability and reduced HRQoL in many,14 there is no denying the importance of screening to reduce the mortality and morbidity associated with SAH. Yet it needs to be recognized that screening can have a considerable impact on the health and psychological well-being of persons. Appropriate support and professional counseling, where appropriate, may be important components of a screening program. The psychosocial consequences of a positive screening should be discussed with asymptomatic relatives before the decision is made about whether or not to proceed to screening.

110

Chap

ter

References

1. Bromberg JE, Rinkel GJE, Algra A, Greebe P, van Duyn CM, Hasan D, Limburg M, ter Berg HW, Wijdicks EF, van Gijn J. Subarachnoid haemorrhage in fi rst and second degree relatives of patients with subarachnoid haemorrhage. BMJ 1995; 311:288-9.2. Raaymakers TW, Rinkel GJE, Ramos LM. Initial and follow-up screening for aneurysms in families with familial subarachnoid hemorrhage. Neurology 1998; 51:1125-30.3. Wermer MJH, Rinkel GJE, van Gijn J. Repeated screening for intracranial aneurysms in familial subarachnoid hemorrhage. Stroke 2003; 34:2788-91.4. Bederson JB, Awad IA, Wiebers DO, Piepgras D, Haley EC, Jr., Brott T, Hademenos G, Chyatte D, Rosenwasser R, Caroselli C. Recommendations for the management of patients with unruptured intracranial aneurysms: A Statement for healthcare professionals from the Stroke Council of the American Heart Association. Stroke 2000; 31:2742-50.5. Lindholt JS, Vammen S, Fasting H, Henneberg EW. Psychological consequences of screening for abdominal aortic aneurysm and conservative treatment of small abdominal aortic aneurysms. Eur J Vasc Endovasc Surg 2000; 20:79-83.6. Scaf-Klomp W, Sanderman R, van de Wiel HB, Otter R, van den Heuvel WJ. Distressed or relieved? Psychological side effects of breast cancer screening in The Netherlands. J Epidemiol Community Health 1997; 51:705-10.7. Aaronson NK, Muller M, Cohen PD, Essink-Bot ML, Fekkes M, Sanderman R, Sprangers MA, te Velde A, Verrips E. Translation, validation, and norming of the Dutch language version of the SF-36 Health Survey in community and chronic disease populations. J Clin Epidemiol 1998; 51:1055-68.8. Brooks R. EuroQol: the current state of play. Health Policy 1996; 37:53-72.9. Spinhoven P, Ormel J, Sloekers PP, Kempen GI, Speckens AE, Van Hemert AM. A validation study of the Hospital Anxiety and Depression Scale (HADS) in different groups of Dutch subjects. Psychol Med 1997; 27:363-70.10. Zigmond AS, Snaith RP. The hospital anxiety and depression scale. Acta Psychiatr Scand. 1983; 67:361- 70.11. Taylor CA, Myers RH. Long-term impact of Huntington disease linkage testing. Am J Med Genet 1997; 70:365-70.12. Smith CO, Lipe HP, Bird TD. Impact of presymptomatic genetic testing for hereditary ataxia and neuromuscular disorders. Arch Neurol 2004; 61:875-80.13. van der Schaaf IC, Brilstra EH, Rinkel GJE, Bossuyt PM, van Gijn J. Quality of life, anxiety, and depression in patients with an untreated intracranial aneurysm or arteriovenous malformation. Stroke 2002; 33:440-3.14. Hop JW, Rinkel GJE, Algra A, van Gijn J. Quality of life in patients and partners after aneurysmal subarachnoid hemorrhage. Stroke 1998; 29:798-804.

111

Eight

ChapterNine

Yield of screening for new aneurysms after treatment for subarachnoid hemorrhage:

a decision analysis

M.J.H. Wermer, E. Buskens, I.C. van der Schaaf, P.M.M. Bossuyt and G.J.E. Rinkel

114

Chap

ter

Abstract

Background and purpose

Patients who have been successfully treated for subarachnoid hemorrhage (SAH) are at risk for new episodes. We studied the effect of screening with CT angiography (CTA) for new aneurysms.

Methods

In a decision model we compared the strategies “screening” and “no screening” after SAH. A literature review yielded the risks of aneurysm recurrence, complications of CTA and re-treat-ment. We estimated the expected number of quality-adjusted life-years (QALY’s), the number of SAH and the mortality and disability rates for both strategies. We evaluated screening at intervals of two, fi ve or 10 years after SAH, using 10 years and remaining life expectancy as time horizon.

Results

The expected number of QALY’s 10 years after clipping was virtually the same for no scree-ning (8.33), screening once after fi ve years (8.28) and screening every two years (8.27). With screening every two years the expected rate of new SAH decreased from 1.9% to 0.5%, that of mortality from 0.9% to 0.6%, whereas the disability rate increased from 0.5% to 1.9%. Results were comparable with remaining life expectancy as time horizon, and for screening after initial treatment with coils. The key estimates of the analyses were the incidence and rupture rate of new aneurysms, the risk of dying from recurrent SAH, the utility of disability and the risk of complications from DSA and re-treatment.

Conclusions

Presently, screening for new aneurysms after SAH cannot be recommended. Screening may prevent new episodes of SAH but with too high a cost in terms of complications from pre-ventive treatment.

115

Nine

Introduction

Patients who have survived an episode of subarachnoid hemorrhage (SAH) are at risk for a new episode from a newly developed aneurysm even if all detected aneurysms have been obliterated.1, 2 Screening for and treatment of new aneurysms could prevent new episodes of SAH but can also lead to complications from screening and preventive treatment. There is a debate between advocates and opponents of screening for new aneurysms after succes-sful treatment.3, 4 With increasing knowledge about the increased risk for new aneurysms in patients treated after SAH, screening may be offered more frequently even though this approach is not yet justifi ed. We developed a decision analytic model to evaluate whether follow-up screening with CT angiography (CTA) is benefi cial in patients after SAH in whom all aneurysms have been obliterated by surgical clipping or endovascular coiling.

Methods

The model

We used a Markov decision model to compare the strategies “screening” and “no screening” in patients who survived SAH and had all detected aneurysms obliterated (Figure 1). The principle of the Markov model is that it defi nes a number of health states and uses transi-tion probabilities between these states.5 The Markov model differs from ordinary decision models in that instead of modeling uncertain events at chance nodes, the uncertain events are modeled as transitions between health states. The model is useful when events can occur at various points in time, which is the case with SAH. A central premise is that at any point in time a patient is classifi ed into one of these health states. During a period of time (cycles) patients can move from one health state to another as defi ned by the transition probabilities. The transition probabilities have some degree of uncertainty. The impact of this uncertainty is addressed by Monte Carlo simulation. The Monte Carlo simulation is a parametric method in which one assumes some distributional form (Gaussian, binary, triangular) for the transition probabilities.6 For each patient or time period (cycle) the distributions are sampled to account for possible uncertainty around the variables. With the use of Monte Carlo simulation, the clinical courses of 10.000 hypothetical patients were individually modeled. We developed two Markov Monte Carlo (MMC) models. The fi rst assessed the effects of scree-ning after clipping, the second those after coiling. These two models were basically similar, only differing in the transition probabilities for clipping and coiling. CTA was used as screening tool, followed by digital subtraction angiography (DSA) whenever an aneurysm was suspected on CTA. For the “screening” and the “no screening” strategy we calculated the expected number of quality-adjusted life-years (QALY’s), the expected number of prevented episodes of SAH,

116

Chap

ter

Markov Monte Carlo model. A decision node (square) denotes the choice between the “screening” and the “no screening” strategy, the Markov nodes (circles with M) mark the entrance to the Markov decision tree, the branches of the Markov node enumerate the fi ve different health states, the chance nodes (circles) represent the events that follow over a certain period of time (cycle length) and the labels represent a generally agreed action to follow an event. A, B and C refer to structurally identical subtrees.

Figure 1

AB

C

See chance node ASee chance node A

See chance node BSee chance node B

See chance node CSee chance node C

See chance node BSee chance node B

See chance node CSee chance node C

117

Nine

and the mortality and disability rates. We assessed the net effects of screening performed fi ve years and every two years after initial treatment for a 10 year time horizon and those of screening every two, fi ve and 10 years with remaining life expectancy as time horizon. For the “no screening” strategy and the strategy of screening every two years we calculated both the number of QALY’s as well as the life expectancy in years.

Health states

We defi ned the following fi ve health states: 1) death; 2) disabled after rupture of a new aneu-rysm, after complications from DSA or after surgical or endovascular treatment; 3) healthy with a regrowth aneurysm (alive in a good clinical condition with an aneurysm present at the site of the fi rst treated one); 4) healthy with a de novo aneurysm (alive in a good clinical condition with a new aneurysm located at a site remote from the original aneurysm) and 5) healthy with all aneurysms treated.

Model assumptions

We made the following additional assumptions: 1) screening with CTA has no long-term effect on quality of life, since the risk of allergic contrast reactions is very small; 2) after the fi rst SAH all aneurysms present are completely obliterated by coiling or clipping; 3) all aneurysms detec-ted by screening with CTA or detected after rupture of an unknown aneurysm in a recovered patient are suitable for treatment; 4) patients who are disabled after a second SAH end up in such a poor condition that treatment of the ruptured aneurysm is not an option; 5) none of the patients with an aneurysm detected during screening declines treatment.

Utilities

A subjective factor in our model is the valuation of the quality of life or utility of the different health states. We assigned a utility of zero to the state death. The degree of disability varies

Explanation of some key branches:At the start of our model, after SAH all patients in the “no screening” strategy start in the “well treated” state. After one cycle (one year) they may have died from another cause than SAH, they may have developed a de novo or a regrowth aneurysm or they may have had no event at all. If they have died from another cause they are classifi ed in the health state “death” and do not follow another cycle. If they have developed a de novo or regrowth aneurysm they are classifi ed in the health state “well with de novo” or “well with regrowth”. If they had no events they are again classifi ed in the health state “well treated”. In the next cycle patients in the “well treated” state follow the same subtree as in the fi rst cycle. The patients in the “well with de novo” or “well with regrowth” state may die from another cause than SAH, may have a rupture of an aneurysm, or they may have no event at all. If they have died they are classifi ed in the health state “death”. If they had a ruptured aneurysm they are classifi ed in the health state “death”, “disabled” or “well treated” according to the outcome after SAH and treatment. If they had no event at all, they return to the “well with de novo” or “well with regrowth” states and are again at risk for rupture in the next cycle.

118

Chap

ter

among patients who have survived SAH and there are no data on the utility of disability after SAH. We valued the disability from rupture or complication of treatment at 0.25. This utility is comparable with the utility after a major stroke and correlates with a Rankin Score of 4 to 5.19 Since a patent aneurysm in itself does not affect the quality of life of a patient who is not aware of it, all three different health states (healthy with de novo, regrowth or treated aneurysm) were assigned a utility of 1.

Sensitivity analysis

Using Monte Carlo simulation we performed multivariate sensitivity analyses. In the Markov model we performed one-way sensitivity analyses for screening once fi ve years after treat-ment. In one-way sensitivity analysis the sensitivity of the estimated QALY’s to changes in the model assumptions are examined for each estimate one at a time. We evaluated the impact of changes in different transition probabilities and the utility of the health state disabled on the screening effect (does the preferred strategy change by changing the transition pro-babilities) with 10 years as time horizon. In addition, we examined the impact of changes in transition probabilities on the outcome in QALY’s of the “screening” and the “no screening” strategy separately (which estimates are most important for the outcome of the strategies) both with 10 years and with life expectancy as time horizon.

Transition Probabilities

We retrieved the transition probabilities for our model from the literature. A MEDLINE search was performed to identify relevant publications from 1985 to 2001. The following keywords were used in different combinations: subarachnoid hemorrhage, aneurysm, recurrent, re-growth, residual neck, remnant, de novo, complications, outcome, treatment, clipping, coiling, sensitivity, specifi city, CT angiography and digital subtraction angiography. In case clinical data were lacking we estimated the transition probability based on clinical experience. Table 1 lists the estimates used.

Risk of a regrowth or de novo aneurysm after a fi rst episode of SAH:

Only a few articles, all with a short follow-up period, have reported on the risk of developing a regrowth or a de novo aneurysm. After clipping a regrowth rate of 0.3-0.5% per year has been reported.1, 7 Regrowth after coiling was found in 5% after three months, in an additional 9% between three and 18 months and an additional 4% between 18 months and 3-4 years.8 We used a regrowth rate of 0.5% per year after clipping and 5% per year after coiling. The development of de novo aneurysms varied between 0.84-2.2% per year.1, 7, 9, 10 We used a de novo rate of 2% per year.

119

Nine

Risk of rupture of regrowth and de novo aneurysms:

We could not fi nd specifi c rupture rates for regrowth aneurysms in the literature. We computed a yearly rupture risk of 3.2% based on two studies reporting on regrowth after surgically treated aneurysms.1, 2 The reported rates for rupture of de novo and additional aneurysms vary from 0.5-1.4% per year and from 10-25% per 10 years.9, 11, 12 We used a rupture rate for de novo aneurysms of 1.4% per year.

Risk of complications due to re-treatment:

Several case-reports have been published considering re-treatment of regrowth or incom-pletely treated aneurysms. We did not fi nd large series of patients in the literature to deter-

Determinant Probability/yr or mean value (sd)*

Markov chainRange** Distribution***

Regrowth aneurysm after clip 0.005 0.003-0.005 BinaryRegrowth aneurysm after coil 0.05 0.001-0.09 BinaryDe novo aneurysm after clip / coil 0.02 0.009-0.022 BinaryRisk of rupture de novo 0.014 0.005-0.017 BinaryRisk of rupture regrowth 0.032 0.024-0.06 BinaryCTA sensitivity 0.8† 0.75-1• TriangularCTA specifi city 0.9† 0.85-1• TriangularDisability conventional angiography 0.008 0.003-0.085 BinaryMortality conventional angiography 0.0002 0.0002-0.007 BinaryDisability treatment de novo/regrowth 0.062 0.028-0.163 BinaryMortality treatment de novo/regrowth 0.015 0.004-0.026 BinaryDisability after ruptured aneurysm 0.25 0.1-0.25 BinaryMortality after ruptured aneurysm 0.5 0.4-0.5 BinaryUtility death 0 - -Utility disabled 0.25 0.25-0.7 -Utility healthy 1 - -Age 49.1 36.8-61.4 Gaussian (SD 12.3)Men 0.34 - BinaryWomen 0.66 - Binary

Table 1 The transition probabilities, utilities and age and gender distribution used in the Markov Monte Carlo model.

* = Base-line estimate used in analyses** = Range used in univariate sensitivity analyses*** = Distribution used in (Monte Carlo simulation) multivariate sensitivity analyses† = The most likely estimate of the triangular distribution• = The upper and lower limit of the triangular distributionSD = standard deviation

120

Chap

ter

mine the complication risk of re-treatment. Since re-operation on average is considered to be technically more challenging we rated the risk of complications 1.5 times higher than the 4.1% morbidity and the 1.0% mortality reported after clipping of a fi rst aneurysm.13 Although the complications of re-treatment with coiling may be milder than with clipping we did not distinguish between the two procedures in that respect. We used a general treatment related morbidity of 6.2% and a general treatment related mortality of 1.5%.

Sensitivity and specifi city CTA:

The sensitivity of CTA to detect aneurysms in general varies between 76 and 100% and the specifi city between 85 and 100%.14 We expected the specifi city and the sensitivity of CTA to be slightly lower after treatment because of clip-artifacts. In our analyses we used a sensitivity of CTA of 80% and a specifi city of 90%.

Risk of complications from DSA:

A recent meta-analysis reported an overall neurological complication risk of 0.8% in pa-tients presenting with SAH, cerebral aneurysm and arteriovenous malformation.15 We used this percentage as disability rate. We estimated that the risk of dying from DSA would beapproximately 0.02%.

Risk of death and disability after SAH:

We used an overall mortality rate after SAH of 50%, and an overall risk of disability of 25%.16, 17

Risk of dying from causes other than SAH:

We derived the age- and sex-specifi c annual probabilities of dying from national life-tables with SAH related risk of dying excluded.18 These data were used to account for the fact that the cohort in the model ages.

Discount rate:

We used a 3% discount rate to account for time preference. Time preference is the preference to be in a healthy state now as opposed to later since we individually and as a society can benefi t from these healthy years in the interim.

Age and gender distribution:

The age and gender distribution of the patients was based on a recent follow-up study in our

121

Nine

hospital of 455 patients after surgical treatment for SAH (Rinkel GJE, Greebe RN and Algra A, unpublished data, 2002).

Results

New episodes of subarachnoid hemorrhage

Ten years after clipping a new SAH occurred in 1.9% of the 10.000 patients without screening (Table 2). Screening every two years decreased this percentage to 0.5%. With calculations for remaining life expectancy a new episode of SAH occurred in 12% of the patients without screening. Screening every two years decreased the incidence of new SAH to 1.9%. After coiling a new episode of SAH occurred in 6.8% of the 10.000 patients within 10 years (Table 3). Screening every two years decreased this percentage to 2.3%. With calculations for remaining life expectancy, a new episode of SAH occurred in 32.3% after coiling. After screening every two years the incidence of SAH was 7.6%. The results for screening every fi ve and 10 years are shown in Table 2 and 3.

Disability rates

The estimated percentage disabled patients 10 years after clipping for the “no screening” strategy was 0.5% versus 1.9% with screening every two years. With calculations for remaining life expectancy, the disability rate was 2.9% without screening. With screening every two years this percentage was 5.7%. Ten years after coiling the disability rate was 1.8% without screening. The disability rateincreased to 4.0% with screening every two years. With calculations for remaining life expectancy, the disability rate was 7.9% without screening. This rate increased to 12% with screening every two years.

Mortality rates

The overall mortality rate with screening after clipping decreased from 0.9% to 0.6% when screening was performed every two years. With calculations for remaining life expectancy the mortality for the “no screening” strategy was 6%. This percentage decreased to 2% with screening every two years. Ten years after coiling we computed a mortality rate of 3.3%. With screening every two years, mortality rate decreased to 2.0%. With calculations for remaining life expectancy, themortality rate for the “no screening” strategy was 16.2%. Screening every two years resulted in a mortality rate of 5.5%.

122

Chap

ter

Met

hod

of sc

reen

ing

afte

r clip

ping

Tim

e ho

rizon

SAH

regr

owth

(%)

SAH

de n

ovo

(%)

SAH

tota

l (%

) (sd

)

Disa

bSA

H(%

)

Disa

bDS

A(%

)

Disa

bTr

eat

(%)

Disa

bto

tal

(%) (

sd)

Deat

hSA

H(%

)

Deat

hDS

A(%

)

Deat

hTr

eat

(%)

Deat

hto

tal

(%) (

sd)

LE (yea

rs)

QALY

’s

No sc

reen

ing

10 ye

ars

0.7

1.21.9

(13.8

)0.

50

0 0.

5 (7.2

)0.

9 0

0 0.

9 (9

.4)

-8.

33Sc

reen

ing

afte

r 5 ye

ars

10 ye

ars

0.4

0.7

1.1 (1

0.1)

0.3

0.3

0.7

1.3 (1

1.2)

0.6

0.01

0.

1 0.

7 (8.

1)-

8.28

Scre

enin

g ev

ery 2

year

s10

year

s0.

20.

30.

5 (7.3

)0.

1 0.

6 1.2

1.9 (1

3.7)

0.3

0.03

0.

3 0.

6 (7.

6)-

8.27

No sc

reen

ing

LT3.7

8.

3 12

.0 (3

2.9)2

.9

0 0

2.9 (1

6.8)

6.0

0 0

6.0

(23.8

)17

.88

17.72

Scre

enin

g ev

ery1

0yea

rsLT

2.0

3.8

5.8 (2

3.6)

1.20.

62.3

4.1 (

19.8

)2.9

0.01

0.6

3.5 (1

8.6)

-17

.66

Scre

enin

g ev

ery 5

year

sLT

1.3

2.3

3.6 (1

8.8)

0.9

0.9

3.0

4.8

(21.5

)1.9

0.

020.

6 2.5

(15.6

)-

17.6

7Sc

reen

ing

ever

y 2 ye

ars

LT0.

6 1.3

1.9

(13.8

)0.

4 2.0

3.3

5.7

(23.3

)1.0

0.

03

1.0

2.0 (1

4.1)

17.9

417

.54

Tabl

e 2

The

expe

cted

num

ber o

f qua

lity-

adju

sted

life

-yea

rs, n

ew e

piso

des o

f SAH

, mor

talit

y an

d di

sabi

lity

rate

s with

no

scre

enin

g, a

nd sc

reen

ing

at d

iffer

ent

inte

rval

s bas

ed o

n m

ultiv

aria

te M

onte

Car

lo si

mul

atio

ns o

f 10.

000

patie

nts a

fter S

AH tr

eate

d by

clip

ping

.

LT =

rem

aini

ng li

fe ex

pect

ancy

SAH

regr

owth

and

SAH

de n

ovo

= SA

H ca

used

by a

regr

owth

or a

de n

ovo

aneu

rysm

Disa

b SA

H =

disa

bilit

y per

cent

age c

ause

d by

suba

rach

noid

hem

orrh

age

Disa

b DS

A =

disa

bilit

y per

cent

age c

ause

d by

dig

ital s

ubtra

ctio

n an

giog

raph

yDi

sab

treat

= d

isabi

lity p

erce

ntag

e cau

sed

by re

-trea

tmen

t with

neu

rosu

rgica

l clip

ping

Deat

h SA

H =

mor

talit

y per

cent

age c

ause

d by

suba

rach

noid

hem

orrh

age

Deat

h DS

A =

mor

talit

y per

cent

age c

ause

d by

dig

ital s

ubtra

ctio

n an

giog

raph

yDe

ath

treat

= m

orta

lity p

erce

ntag

e cau

sed

by re

-trea

tmen

t with

neu

rosu

rgica

l clip

ping

LE =

life

expe

ctan

cy in

year

sQA

LY’s

= qu

ality

-adj

uste

d lif

e-ye

ars

sd =

stan

dard

dev

iatio

n

123

Nine

Met

hod

of sc

reen

ing

afte

r coi

ling

Tim

e ho

rizon

SAH

regr

owth

(%)

SAH

de n

ovo

(%)

SAH

tota

l (%

) (sd

)

Disa

bSA

H(%

)

Disa

bDS

A(%

)

Disa

bTr

eat

(%)

Disa

bto

tal

(%) (

sd)

Deat

hSA

H(%

)

Deat

hDS

A(%

)

Deat

hTr

eat

(%)

Deat

hto

tal

(%) (

sd)

LE (yea

rs)

QALY

’s

No sc

reen

ing

10 ye

ars

5.7

1.1

6.8

(25.1

)1.8

0

0 1.8

(13.

4)3.3

0

0 3.3

(17.9

)-

8.22

Scre

enin

g af

ter 5

year

s10

year

s3.7

0.

54.

2 (20

.2)1.1

0.

4 1.3

2.8

(16.

1)2.1

0.

01

0.3

2.4

(15.3)

-8.

17Sc

reen

ing

ever

y 2 ye

ars

10 ye

ars

2.1

0.3

2.3 (1

5.1)

0.5

0.7

2.8

4.0

(19.7)

1.20.

04

0.8

2.0 (1

4.1)

-8.

13

No sc

reen

ing

LT26

.7 5.5

32.3

(49.

1)7.9

0

0 7.9

(26.

9)16

.20

0 16

.2 (3

6.9)

17.11

16.6

3Sc

reen

ing

ever

y10

year

sLT

16.6

3.1

19

.7 (4

0.9)

4.8

0.8

4.4

10.0

(29.

9)10

.10.

011.1

11.2 (

31.6

)-

16.79

Scre

enin

g ev

ery 5

year

sLT

10.3

2.1

12.4

(33.5

)3.1

1.4

6.6

11.1 (

31.5)

6.3

0.04

1.6

7.9

(26.

9)-

16.8

8Sc

reen

ing

ever

y 2 ye

ars

LT6.

3 1.3

7.6

(26.

6)1.9

2.

4 7.7

12

.0 (3

2.5)

3.7

0.09

1.8

5.5

(23.1

)17

.81

16.9

6

Tabl

e 3

The

expe

cted

num

ber o

f qua

lity-

adju

sted

life

-yea

rs, n

ew e

piso

des o

f SAH

, mor

talit

y an

d di

sabi

lity

rate

s with

no

scre

enin

g, a

nd sc

reen

ing

at d

iffer

ent

inte

rval

s bas

ed o

n m

ultiv

aria

te M

onte

Car

lo si

mul

atio

ns o

f 10.

000

patie

nts a

fter S

AH tr

eate

d by

coili

ng.

LT =

rem

aini

ng li

fe ex

pect

ancy

SAH

regr

owth

and

SAH

de n

ovo

= SA

H ca

used

by a

regr

owth

or a

de n

ovo

aneu

rysm

Disa

b SA

H =

disa

bilit

y per

cent

age c

ause

d by

suba

rach

noid

hem

orrh

age

Disa

b DS

A =

disa

bilit

y per

cent

age c

ause

d by

dig

ital s

ubtra

ctio

n an

giog

raph

yDi

sab

treat

= d

isabi

lity p

erce

ntag

e cau

sed

by re

-trea

tmen

t with

endo

vasc

ular

coili

ngDe

ath

SAH

= m

orta

lity p

erce

ntag

e cau

sed

by su

bara

chno

id h

emor

rhag

eDe

ath

DSA

= m

orta

lity p

erce

ntag

e cau

sed

by d

igita

l sub

tract

ion

angi

ogra

phy

Deat

h tre

at =

mor

talit

y per

cent

age c

ause

d by

re-tr

eatm

ent w

ith en

dova

scul

ar co

iling

LE =

life

expe

ctan

cy in

year

sQA

LY’s

= qu

ality

-adj

uste

d lif

e-ye

ars

sd =

stan

dard

dev

iatio

n

124

Chap

ter

Life expectancy in years

The total life expectancy in years was 17.88 after initial treatment by clipping for the “no screening” strategy and 17.94 for screening every two years.After coiling the total life expectancy in years was 17.11 when no screening was performed. Screening every two years resulted in an increase of the total number of life-years to 17.81.

Quality-adjusted life-years

Ten years after clipping, the expected number of QALY’s was around 8.3 for all screening stra-tegies. The calculations for remaining life expectancy resulted in 17.72 QALY’s for no screening. Repeated screening every two years reduced the expected number of QALY’s to 17.54. Ten years after coiling, the expected number of QALY’s was 8.22 with no screening and 8.13 with screening every two years. Calculations for remaining life expectancy showed a slight benefi t of screening. In the “no screening” group the expected number of QALY’s was 16.63 which increased to 16.96 with repeated screening every two years.

One-way sensitivity analysis

The one-way sensitivity analyses showed that the majority of estimates, when varied over the range as is shown in Table 1, did not result in a change of preferred strategy. There was no substantial difference in the comparison between the “screening” and the “no screening” strategy (results not shown).The probabilities that had the most impact on the outcome of the strategy screening fi ve years after treatment by coiling or clipping with 10 years as time horizon, were the risk of compli-cations from DSA or treatment, the risks of recurrence and rupture of a de novo or regrowth aneurysm and to a lesser extend the utility of the health state disabled (Figure 2a and 2b). For the “no screening” strategy, the key factors were the risk of recurrence and rupture of a de novo or regrowth aneurysm, the risk of dying from SAH and the utility of the health state disabled (Figure 2c and 2d). With life expectancy as time horizon, the results were comparable although the utility of disability had a higher impact on the outcome than with 10 years as time horizon (results not shown).Figure 2 shows that although some changes in the transition probabilities seem to have a rather large effect, the QALY’s in the strategy “screening” after treatment by clipping (Figure 2a) do not increase to more than 8,31 and do not overlap with the “no screening” strategy after treatment by clipping (Figure 2c). For the strategies “screening” and “no screening” after coiling there is only some overlap for the estimates risk of rupture of a regrowth and regrowth rate (Figure 2b and 2d). However, the effect is in the same direction and does not change the preferred strategy (which still is the “no screening” strategy).

125

Nine

Figure 2 One-way sensitivity analyses to determine the key estimates of the Markov Monte Carlo model. The analyses were performed for the strategy screening fi ve years after treatment by coiling or clipping and for the strategy “no screening” with 10 years as time horizon

de novo aneurysm after clip or coil: 0.009 to 0.022 regrowth aneurysm after coil: 0.001-0.09 regrowth aneurysm after clip: 0.003-0.005 risk of rupture de novo: 0.005 to 0.017 risk of rupture regrowth: 0.024-0.06 mortality after SAH: 0.4-0.5 disability after DSA: 0.003 to 0.085 disability after treatment: 0.028 to 0.163 utility disabled: 0.25 to 0.7

d Key estimates on the outcome of the strategy “no screening” after coiling

a Key estimates on the outcome of the strategy “screening” after clipping

b Key estimates on the outcome of the strategy “screening” after coiling

c Key estimates on the outcome of the strategy “no screening” after clipping

126

Chap

ter

Discussion

Based on the currently available data, our decision analytic model shows that screening for new aneurysms in patients with successful treatment after a fi rst SAH hardly affects the expected number of QALY’s. Screening decreases the percentage of new SAH and the total mortality rate but increases the chances of disability caused by complications of angiography and re-treatment. There is an apparent trade-off between mortality and disability, which limits the QALY effect of screening. The only clear benefi t of screening was obtained with screening every two years after initial treatment with coils when the remaining life expectancy was used as time horizon. For this screening strategy, the high percentage of regrowth aneurysms and their subsequent risk of rupture outweighed the risk of preventive treatment and resulted in an increase in QALY’s from 16.63 to 16.96.In the patients treated with coiling the rates of SAH, morbidity and mortality were considera-bly higher than in the patients treated with clipping. However, the effect on the difference in QALY’s between clipping and coiling, was relatively small because of the rarity of the events and the effect of time preference. In general, for people, healthy years in the near future count for more QALY’s than healthy years later in life. Most death and disability occurred because of SAH, and SAH mostly occurs in the far future. Therefore, it has a smaller impact on the QALY’s than death and disability from screening and preventive treatment, which generally occur in the near future.The results of the sensitivity analyses indicate that the outcome of the analyses is robust over a wide range of assumptions. The parameters with the largest impact on the yield of screening were the incidence and rupture rate of a de novo or a regrowth aneurysm, the risk of dying from SAH, DSA and treatment related complications, and the utility of disability. To our knowledge, no other studies on screening for new aneurysms have been performed. In our analyses, the estimated 10 years cumulative incidence of SAH after clipping was 1.9%. This number is slightly lower than reported after surgical treatment of aneurysms in two follow-up studies. In these two studies, a percentage of new SAH of 2.2% in 10 years and 0.26% per year was reported.1, 2 A follow-up study in our hospital of 455 patients with clipped cerebral aneurysms even showed a 10.5% (95% CI 1.1-19.8%) risk of a new episode of SAH in the fi rst 11 years after the index SAH (Rinkel GJE, Greebe RN and Algra A, unpublished data, 2002). The results of the latter study imply that the true incidence of recurrent aneurysms and the risk of rupture of these aneurysms may be higher than the estimates we used in our model. In that case, the benefi t of screening might be underestimated. Similar comparisons for coiling cannot be made since data on recurrent SAH after this type of procedure are lacking.Some of our assumptions may be responsible for the limited effects of screening. We assumed that all aneurysms were completely obliterated by clipping or coiling after the fi rst SAH. Ho-wever, after coiling and clipping total occlusion of the aneurysm is not always achieved.8, 20, 21,

127

Nine

A remnant could cause a new SAH that might have been prevented by screening. By using in our model estimates of recurrent aneurysms retrieved from the literature based on complete occlusion of the fi rst aneurysm(s), we probably underestimated the benefi t of the screening strategy. Furthermore we did not incorporate the positive effect of reassurance on quality of life when screening for new aneurysms was negative. Other assumptions may have produced a too optimistic outcome of screening. We assumed that all aneurysms were treated. In clinical practice, this is not always true. Aneurysms that are diffi cult to treat are mostly large aneurysms with compression of neighboring structures. Because a screening program would include only asymptomatic patients such aneurysms are unlikely to be found. Very small aneurysms will often not be treated but followed in time by CTA or angiography. The knowledge of having an untreated aneurysm can negatively affect the quality of life in these patients and decrease the perceived benefi t of screening.22 The process of screening itself and the delay between detection of a recurrent aneurysm and re-treatment can cause a considerable amount of stress, which we did not incorporate. We also simplifi ed our model by using only three different utilities; zero for dead, 0.25 for disa-bled and 1 for healthy. After successful operation sometimes minor defi cits like loss of smell or cognitive impairment occur in otherwise healthy patients. We did not account for these defi cits in our model.No study on screening with CTA has been conducted in patients with treated aneurysms. Because CTA is relatively safe compared to angiography it appears suitable for screening. The use of CTA for detection of aneurysms in general has been well established. Data on post-operative CTA after aneurysm clipping are sparse but promising.23, 24 A prospective study of 11 patients with 13 aneurysms clipped with titanium clips showed that the occlusion of the aneurysm and the remnant of the neck could be well evaluated by spiral CT. The titanium clips caused only minor artefacts.23

Our sensitivity analysis showed that the incidence and rupture rate of aneurysms are pivotal to determine the outcome of screening. Factors that have been associated with the growth and development of intracranial aneurysms are smoking and female sex.10 The extent of this increased risk, however, is not well known. Therefore, in our decision analysis we did not in-corporate these individual risk factors. Knowledge on the extent of the increase in risk from smoking and gender and identifi cation of other risk factors for development and rupture of aneurysms may help to delineate subgroups that are at increased risk. Presently, we cannot recommend screening after SAH. However, if in the future high risk groups can be identifi ed with a rupture and formation risk that exceeds the estimates used in our model, screening may be benefi cial when it is tailored to these individuals.

128

Chap

ter

References

1. David CA, Vishteh AG, Spetzler RF, Lemole M, Lawton MT, Partovi S. Late angiographic follow- up review of surgically treated aneurysms. J Neurosurg 1999;91:396-401.2. Tsutsumi K, Ueki K, Usui M, Kwak S, Kirino T. Risk of recurrent subarachnoid hemorrhage after complete obliteration of cerebral aneurysms. Stroke 1998;29:2511-2513.3. Tsutsumi K, Ueki K, Kirino T. Recurrent subarachnoid hemorrhage. J Neurosurg 2001;94:541- 542.4. Juvela S. Risk of subarachnoid hemorrhage from a de novo aneurysm. Stroke 2001;32:1933-1934.5. Beck JR, Pauker SG. The Markov process in medical prognosis. Med Decis Making 1983;3:419-458.6. Raeside DE. Monte Carlo principles and applications. Phys Med Biol 1976;21:181-197.7. Tsutsumi K, Ueki K, Morita A, Usui M, Kirino T. Risk of aneurysm recurrence in patients with clipped cerebral aneurysms. results of long-term follow-up angiography. Stroke 2001;32:1191-1194.8. Cognard C, Weill A, Spelle L et al. Long-term angiographic follow-up of 169 intracranial berry aneurysms occluded with detachable coils. Radiology 1999;212:348-356.9. Juvela S, Porras M, Heiskanen O. Natural history of unruptured intracranial aneurysms: a long-term follow-up study. J Neurosurg 1993;79:174-182.10. Juvela S, Poussa K, Porras M. Factors affecting formation and growth of intracranial aneurysms: a long-term follow-up study. Stroke 2001;32:485-491.11. Rinkel GJE, Djibuti M, van Gijn J. Prevalence and risk of rupture of intracranial aneurysms: a systematic review. Stroke 1998;29:251-256.12. Unruptured intracranial aneurysms--risk of rupture and risks of surgical intervention. International Study of Unruptured Intracranial Aneurysms Investigators. N Engl J Med 1998;339:1725-1733.13. King JT, Jr., Berlin JA, Flamm ES. Morbidity and mortality from elective surgery for asymptomatic, unruptured, intracranial aneurysms: a meta-analysis. J Neurosurg 1994;81:837-842.14. Wardlaw JM, White PM. The detection and management of unruptured intracranial aneurysms. Brain 2000;123:205-221.15. Cloft HJ, Joseph GJ, Dion JE. Risk of cerebral angiography in patients with subarachnoid hemorrhage, cerebral aneurysm, and arteriovenous malformation: a meta-analysis. Stroke 1999;30:317-320.16. Hijdra A, Braakman R, van Gijn J, Vermeulen M, van Crevel H. Aneurysmal subarachnoid hemorrhage. Complications and outcome in a hospital population. Stroke 1987;18:1061-1067.17. Hop JW, Rinkel GJE, Algra A, van Gijn J. Case-fatality rates and functional outcome after subarachnoid hemorrhage: a systematic review. Stroke 1997;28:660-664.18. National life tables. CBS Offi ce of Statistics Netherlands, Voorburg, the Netherlands.19. Post PN, Stiggelbout AM, Wakker PP. The utility of health states after stroke: a systematic review of the literature. Stroke 2001;32:1425-1429.20. Byrne JV, Sohn MJ, Molyneux AJ, Chir B. Five-year experience in using coil embolization for ruptured intracranial aneurysms: outcomes and incidence of late rebleeding. J Neurosurg 1999;90:656-663.21. Vinuela F, Duckwiler G, Mawad M. Guglielmi detachable coil embolization of acute intracranial aneurysm: perioperative anatomical and clinical outcome in 403 patients. J Neurosurg 1997;86:475- 482.22. van der Schaaf IC, Brilstra EH, Rinkel GJE, Bossuyt PM, van Gijn J. Quality of life, anxiety, and depression in patients with an untreated intracranial aneurysm or arteriovenous malformation. Stroke 2002;33:440-443.23. van Loon JJ, Yousry TA, Fink U, Seelos KC, Reulen HJ, Steiger HJ. Postoperative spiral computed tomography and magnetic resonance angiography after aneurysm clipping with titanium clips. Neurosurgery 1997;41:851-856.24. Steiger HJ, van Loon JJ. Virtues and drawbacks of titanium alloy aneurysm clips. Acta Neurochir Suppl 1999;72:81-88.

129

Nine

ChapterTen

Follow-up screening after subarachnoid hemorrhage: frequency and determinants of new aneurysms and

enlargement of existing aneurysms

M.J.H. Wermer, I.C. van der Schaaf, B.K. Velthuis, A. Algra, E. Buskens and G.J.E. RinkelFor the ASTRA study group

132

Chap

ter

Abstract

Background and purpose

Intracranial aneurysms have long been considered a once in a lifetime event. Nevertheless,patients who survive after subarachnoid hemorrhage (SAH) may be at risk for new aneurysms. In a cohort of patients with clipped aneurysms, we studied the yield of screening in the years after the SAH and we tried to identify risk factors for formation of new aneurysms as well as for enlargement of aneurysms that were already present at the time of the SAH.

Methods

We screened 610 patients who had been admitted between 1985 and 2001 for SAH by means of CT-angiography (CTA). Risk factors were evaluated by Cox regression analyses.

Results

With screening we detected 129 aneurysms in 96 (16%) patients, after a mean interval of 8.9 years. Of these, 24 (19%) were located at the site of the previously ruptured and clipped aneurysm and 105 (81%) at a site remote from the clip site. Of the aneurysms at a remote site 59 could be compared with the initial (CT)-angiogram. Of these, 19 were truly de novo (32%) and 40 (68%) were already visible in retrospect. Of the 53 aneurysms that were followed over time 13 (25%) had enlarged. Risk factors for aneurysm formation and growth were presence of multiple aneurysms at time of SAH (HR 3.2, 95% CI 1.2-8.6), current smoking (HR 3.8, 95% CI 1.5-9.4) and hypertension (HR 2.3, 95% CI 1.1-4.9).

Conclusions

These results suggest that intracranial aneurysms should not be considered as a single event in a lifetime but rather as a continuous process. Patients with a previous SAH have a substantial risk for new aneurysm formation and enlargement of untreated aneurysms. Screening these patients might be benefi cial, especially in patients with multiple aneurysms, hypertension and a history of smoking. The risks and benefi ts of screening, however, should be carefully weighed, for example in a decision model.

133

Ten

Introduction

Subarachnoid hemorrhage (SAH) from a ruptured aneurysm is often considered a once in a lifetime event. However, the rarity of intracranial aneurysms under the age of 20 and the presence of multiple aneurysms in up to 30% of patients with SAH suggest that there is an ongoing development of aneurysms during life.1 Patients with a history of SAH may therefore be at risk for new aneurysms despite successful treatment of the ruptured aneurysm at time of the SAH. New aneurysms can develop at a new site (de novo) or adjacent to the clip or coil from the previous treatment (regrowth). Data on the incidence of de novo or regrowth aneurysms after treatment of ruptured aneurysms are sparse. Also the risk factors for formation of new aneurysms or the enlargement of additional (unruptured) aneurysms in patients with a history of SAH are not well known. In previous studies the incidence of de novo aneurysms has been reported as between 0.8 and 2.2% per year and the incidence of regrowth aneurysms after clip-ping around 0.5% per year.2-4 However, these series were small and often inhomogeneous.Nowadays screening can be performed relatively easily with screening techniques that are not or only minimally invasive such as CT-angiography (CTA) and MR- angiography (MRA). We assessed the number of aneurysms detected with screening by means of CTA in a cohort of patients who had been admitted with SAH between 1985 and 2001 and studied risk factors for new aneurysm formation and enlargement of already existing aneurysms.

Methods

Patients

After approval from the hospital ethics committees we selected from our databases of patients admitted with SAH, all patients who met the following inclusion criteria: 1) admission between 1985 and 2001 to the University Medical Center Utrecht (UMC) or admission between 1987 and 1998 to the Academic Medical Center Amsterdam (AMC), 2) confi rmation of the SAH by CT and an aneurysm proven by CTA or conventional intra-arterial angiography (IA-A), 3) treatment of the ruptured aneurysm by clipping, 4) recovery to an at least partly independent state after treatment (defi ned as a score of ≤ 3 on the modifi ed Rankin scale)5 and 5) age between 18 and 70 years at time of screening. Excluded were patients 1) who were not able to communicate well enough to give informed consent, 2) those with severe co-morbidity or reduced life ex-pectancy or 3) patients in whom one or more aneurysms had been treated by coiling.For all patients meeting the inclusion criteria we contacted the general practitioner to see if the patient was still alive and had not moved to a nursing home. Subsequently we invited all patients to visit our outpatient clinic for more detailed information about the study. At

134

Chap

ter

the outpatient clinic we recorded the medical history, actual blood pressure, family history of intracranial aneurysms and possible other risk factors for aneurysm formation.

CT-angiography and conventional intra-arterial angiography

CTA of the brain was performed by means of multi-slice (MS)-CTA. One hospital used a 16-detector MS-CT (UMC), the other hospital used a 4-detector MS-CT (AMC). In patients with an allergy to iodinated contrast agents a 1.5 tesla MRA was performed. Two neuroradiologists in each hospital independently evaluated all CTA’s. In case of disagreement the neuroradio-logists tried to reach consensus in a common reading. If there was still disagreement after the consensus reading a third neuroradiologist was asked to provide a fi nal decision. In all patients in whom, on CTA, an aneurysm was suspected at the clip site, IA-A was performed for confi rmation. In patients with an aneurysm at a site remote from the clip site IA-A was subsequently performed if the estimated size of the aneurysm was at least three millimeters. Patients with an aneurysm smaller than three millimeters were offered a control CTA after one year. IA-A was not performed in patients with a normal CTA.

Classifi cation of aneurysms

If an aneurysm had already been described in the patient discharge letter or in the report of the angiogram at the time of the SAH, but was left untreated, we classifi ed it as already known. If an aneurysm was detected that had not been described before, fi rst, the neuroradiologists were asked to classify it either as an aneurysm at a location remote from the originally clipped aneurysm or as an aneurysm located at the clip site from the previous operation. Secondly, for all detected aneurysms the neuroradiologists had a second look at the angiograms that had been performed at time of the SAH, if still available. The hard copies of the angiogram or the digital datasets of the CTA at time of the SAH were available only in patients who had been admitted after 1992. The aneurysms at sites remote from the clip were subdivided in 1) de novo aneurysm (not visible on the initial CTA/IA-A), 2) additional aneurysm (visible only in retrospect but not identifi ed on the CTA/IA-A at the time of the SAH), 3) probable de novo aneurysm (CTA or four-vessel IA-A performed at time of SAH, no aneurysm reported but CTA or angiogram not longer available for review) or 4) unknown (unknown if a complete four vessel IA-A was performed or if an incomplete IA-A was performed at time of SAH). The aneu-rysms located at the site of the clip were classifi ed as 1) regrowth aneurysm (post-operativeangiogram showed complete clipping of the aneurysm); 2) remnant (post-operative angiogram showed incomplete clipping of the aneurysm); or 3) unknown if post-operative angiography had not been performed or was not available for review.Thirdly, the observers measured the size of all newly detected aneurysms. On the CTA the size of the aneurysm was measured directly. On IA-A the size of the aneurysm was related

135

Ten

to the size of the carotid artery. If an aneurysm had already been present at the time of the original SAH (additional or already known aneurysms), the neuroradiologists measured if the aneurysm had increased in size in the intervening years.

Data-analysis

We recorded the overall number and proportion of aneurysms detected at screening and the number and proportion of different subgroups of aneurysms (de novo, regrowth, additional) with corresponding 95% confi dence intervals (CI). We also calculated the number of aneu-rysms in the different subgroups per patient-year of screening. In addition we assessed the location and the size of the detected aneurysms and the frequency and rate of enlargement of aneurysms in patients with additional or already known aneurysms.We used Cox regression to calculate hazard ratios (HR) and corresponding 95% CIs of factors associated with aneurysm formation and enlargement. For the Cox regression analyses and the calculations of the number of aneurysms per patient-year of screening we assumed that the new aneurysms had developed halfway between the SAH and the screening. Risk factors for aneurysm formation were analyzed by comparing the characteristics of patients with a de novo or a regrowth aneurysm with those of patients with a negative screening. Similarly, risk factors for aneurysm enlargement were analyzed by comparing the characteristics of patients who had an additional or already known aneurysm that had grown in the time after the SAH with those of patients who had an additional or already known aneurysm that had not enlarged in size. In a second analysis we included both the patients with a newly developed aneurysm (de novo or regrowth) and patients with an aneurysm that had enlarged and compared the risk factors in these two groups together with those in patients with a negative screening.The following characteristics were included in the analyses: age at the time of SAH (continuous), sex, smoking (dichotomous: current versus former plus never smokers, and categorical), alcohol use (categorical: never use of alcohol, < 5 drinks a day, former use of > 5 drinks a day and current use of > 5 drinks a day), history of hypertension, blood pressure values at time of screening(categorical: normal (systolic blood pressure [SBP] < 140 mm Hg and/or diastolic bloodpressure [DBP] < 90 mm Hg), borderline (SBP 140-160 mm Hg and/or DBP 90-95) or hyperten-sion (SBP>160 and/or DBP>95), family history of intracranial aneurysms (defi ned as more than one fi rst-degree relative(s) with a verifi ed aneurysm or a history very suggestive of SAH such as relatives who died after sudden very severe headache), location of the ruptured aneurysm at the time of SAH (categorical: internal carotid artery [ICA], anterior cerebral artery [ACA], medial cerebral artery [MCA], vertebrobasilar and dichotomous: MCA versus other locations) and number of aneurysms at the time of SAH (dichotomous: one versus multiple). We per-formed univariate Cox regression analysis for all risk factors and multivariate Cox regression analysis with forward selection of variables with probability values < 0.20 in the univariate analysis. We tried to fi nd a combination of variables that most accurately predicted the oc-

136

Chap

ter

currence of new aneurysm development or enlargement. We used receiver-operator charac-teristic (ROC) curves to evaluate the information content of the predictive model. The more a ROC curve is located in the upper left corner of the graph, the more is the area under the curve and the higher are both sensitivity and specifi city for the prediction of new formation of enlargement of an aneurysm.

Results

Frequency of detected aneurysms

In the period 1985-2001 (UMC Utrecht) and the period 1987-1998 (AMC Amsterdam) a total of 1455 patients were discharged after treatment of the ruptured and all additional aneurysms. From these 1455 patients 396 were excluded because they were too young or too old, 125 be-cause they had died after discharge, 87 because they stayed in a nursing home or had severe co-morbidity, 59 because they were coiled and 13 for other reasons like language problems. In 10 of the 125 patients who died after discharge the cause of death was a recurrent SAH and in 2 of the 87 patients who stayed in a nursing home the cause of admission was disability from a recurrent SAH. In total 775 patients met the inclusion criteria. Of these patients, 99 declined screening, 21 had moved abroad, 14 were lost to follow-up and 31 were not approached (be-cause the study had already ended). We screened a total of 610 patients. Their mean age was 53.5 years (range 24-70); 391 (64%) were women. Table 1 shows the baseline characteristics of all patients. The age and sex distribution of the non-participants was comparable with the distribution of the screened patients. The mean interval after the SAH was 8.9 years (range 2.3-18.8). We found 129 new aneurysms in 96 of the 610 patients (16%; 95% CI 13-19%), that is, aneurysms not identifi ed at the time of the presenting SAH. In addition, 22 aneurysms in 17 patients were already known to be present. In total we found 151 aneurysms in 112 (18%; 95% CI 15-22%) of the 610 patients (Figure 1). Of the 129 aneurysms that were not described at the time of the SAH, 24 (19%) were located at the site of the clip placed at the previous operation and 105 (81%) were located at a site remote from the clip site. For 59 of the 105 aneurysms that were detected at a remote site the CTA/IA-A at the time of the SAH was available for review. Nineteen (32%) of these 59 aneu-rysms were classifi ed as being de novo and 40 (68%) were classifi ed as additional. In three of the 24 patients with an aneurysm at the clip site a post-operative angiogram was available for review, which showed no aneurysm; these three aneurysms were therefore classifi ed as regrowth. The size and location of the different aneurysms are shown in Figure 2 and an example of a regrowth and a de novo aneurysm is shown in Figure 3.Out of 112 patients with aneurysms 84 (75%) had a single aneurysm at screening, 18 (16%) had two aneurysms, nine (8%) had three aneurysms and one (1%) patient had four aneurysms.

137

Ten

Characteristic All (N=610)

No aneurysm(N=498)

De novo (N=14)

Regrowth (N=3)

Enlargement (N=13)

Age at time screen (mean yrs)Age at SAH:

<40 years40-60 years>60 years

53

196 (32%)384 (63%)30 (5%)

53

162 (32%)313 (63%)23 (5%)

50

8 (57%)6 (43%)0 (0%)

55

1 (33%)2 (66%)0 (0%)

55

3 (23%)8 (62%)2 (15%)

Women 391 (64%) 315 (63%) 12 (85%) 2 (66%) 9 (69%)

Current smokerFormer smokerNever smoker

312 (51%)190 (31%)108 (18%)

246 (49%)153 (31%)99 (20%)

11 (79%)3 (21%)0 (0%)

2 (66%)1 (33%)0 (0%)

10 (77%)2 (15%)1 (8%)

Alcohol use > 5drinks a dayFormer use > 5 drinks a dayAlcohol use < 5 drinks a dayNo alcohol use

27 (4%)22 (4%)396 (65%)165 (27%)

22 (4%)20 (4%)323 (65%)133 (27%)

1 (7%)0 (0%)7 (50%)6 (43%)

0 (0%)0 (0%)3 (100%)0 (0%)

1 (8%)0 (0%)7 (54%)5 (38%)

Family history of IA 96 (16%) 83 (17%) 9 (64%) 1 (33%) 2 (15%)

History of hypertension 264 (43%) 208 (42%) 7 (50%) 2 (66%) 9 (69%)

Number aneurysms at SAH*1234

481 (79%)89 (15%)32 (5%)8 (1%)

432 (87%)46 (9%)17 (3%)3 (1%)

9 (64%)3 (21%)2 (14%)0 (0%)

2 (66%01 (33%)0 (0%)0 (0%)

0 (0%)9 (69%)3 (23%)1 (8%)

Location aneurysm at SAH**ACAICAMCAVertebrobasilar

261 (43%)168 (27%)137 (23%)44 (7%)

230 (46%)133 (27%)102 (20%)33 (7%)

7 (50%)2 (14%)3 (21%)2 (14%)

1 (33%)1 (33%)1 (33%)0 (0%)

4 (31%)5 (38%)4 (31%)0 (0%)

*Number of aneurysms at SAH = number of aneurysms at time of SAH including the aneurysms that were not described at time of SAH but could be identifi ed in retrospect.

** ACA= anterior communication artery, ICA=internal communicating artery, MCA=medial cerebral artery.

Table 1 Characteristics of all patients and different subgroups of patients

New aneurysm formation

In the fi rst fi ve years after SAH a de novo aneurysm was found in one of 120 patients (0.8%; 95% CI 0-5.2%), in the fi rst 10 years after SAH in eight of 365 patients (2.2%; 95% CI 1.0-4.4%) and overall in 14 of 610 patients (2.3%; 95% CI 1.3-3.9%). In the total number of 5078 years of follow-up we found 19 de novo aneurysms (in 14 patients) and 42 probable de novo aneurysms (in 34 patients), which corresponds with an annual incidence of de novoaneurysms between 0.37 (95% CI 0.23-0.60%) and 1.20% (95% CI 0.93-1.55%). The mean interval fordetection of the 19 de novo aneurysms was 9.1 years after SAH (range 4.4-13.7).

138

Chap

ter

Figure 1 Different subgroups of aneurysms detected at screening.*The number of patients in the subcategories does not always add up to the number of patients in the main categories because patients are sometimes classifi ed in more than one subcategory.

139

Ten

X-axes: location, Y-axes: number of aneurysmsICA = internal carotid artery including posterior communication arteryMCA = middle cerebral artery,ACA = anterior cerebral artery, including anterior communication and pericallosal arteriesVB = vertebrobasilar system

Figure 2 Size and location of different subgroups of aneurysms

All new aneurysms detected at an-other location than the clip site and all de novo aneurysms.The left bar presents all aneurysms detected at a location remote from the clip site (n=105).The right bars present the subgroup of the de novo aneurysms (n=19).

A

Aneurysms already present at time of SAH (already known or additional).The left bars present all aneurysms that remained stable in size (n=40).The right bars present all aneurysms that enlarged over time (n=13).

B

All aneurysms detected at the clip site (n=24).

C

140

Chap

ter

An aneurysm at the clip site occurred in the fi rst fi ve years after SAH in two of 120 patients (1.7%; 95% CI 0.3-6.5%), in the fi rst 10 years after SAH in 10 of 365 patients (2.7%; 95% CI 1.4-5.1%) and overall in 24 of the 610 (3.9%; 95% CI 2.6-5.9%) patients. The total of 24 aneurysms at the clip site in 24 patients (including three regrowths) in 5078 years of follow-up corresponds to an annual incidence between 0.06% (95% CI 0.02-0.19%) and 0.47% (95% CI 0.31-0.71%). The mean interval after which we detected an aneurysm at the clip site was 10.6 years after SAH (range 4.3-15.5). The three regrowth aneurysms were detected 13, 14 and 15 years after treatment.

3A. IA-A of 1993 (A and B). Left common carotid artery injection with lateral view (A) and anteroposterior view (B) of the internal carotid artery showing an anterior communicating artery aneurysm (arrowhead) and no aneurysm on the pericallosal artery (arrow). IA-A of 2003 (C and D) 10 years after clipping of the anterior communicating artery aneurysm. Lateral (C) and anteroposterior view (D) after selective left internal carotid artery injection demonstrating an anterior communicating artery aneurysm adjacent to the clip site (arrowhead) and a de novo pericallosal aneurysm (arrow).

Figure 3 An example of a de novo and a regrowth aneurysm

3B. IA-A (antero-posterior view) of 1992 (left and middle) and 2003 (right). Left: IA-A at time of SAH with an aneurysm on the anterior communicating artery (Acom) (arrow) and the right middle cerebral artery (MCA) (arrow head). Middle: post operative IA-A with completely clipped aneurysms of the Acom and right MCA.Right: follow up IA-A shows a regrowth Acom aneurysm at the clip site (arrow).

141

Ten

Enlargement of already present aneurysms

For 53 of the 62 aneurysms that were already present at the time of the SAH (40 additional and 13 already known) the initial CTA/IA-A was available for assessment of enlargement. In the fi rst fi ve years after SAH four of 18 aneurysms had enlarged (22.2%; 95% CI 7.4-48.1%), in the fi rst 10 years after SAH nine of 45 aneurysms had enlarged (20.0%; 95% CI 10.1-35.1%) and overall 13 of 53 aneurysms had enlarged (24.5 %; 95% CI 14.2-38.6%) between the SAH and the screening. Four patients had one additional aneurysm that had enlarged and another addi-tional aneurysm that had remained stable. Eleven aneurysms enlarged 1-3 mm, one enlarged 3-5 mm and one aneurysm enlarged > 5 mm. The mean enlargement rate per aneurysm varied between 0.12 and 1.3 millimeters per year.

Risk factors for aneurysm formation and enlargement

In the univariate Cox regression analysis current smoking (HR 3.2, 95% CI 1.0-9.7), a positive family history of intracranial aneurysms (HR 2.7, 95% CI 1.0-7.4), and the presence of multiple aneurysms at time of the SAH (HR 3.3, 95% CI 1.2-8.9) were signifi cantly related with the formation of a de novo or regrowth aneurysm. Sex was not statistically signifi cant in the uni-variate analyses but had a p-value of < 0.2 (HR for men 0.4, 95% CI 0.1-1.3) and was therefore included in the multivariate analyses. In the multivariate forward Cox regression analyses current smoking (HR 3.1, 95% CI 1.0–9.4) and multiple aneurysms (HR 3.2, 95% CI 1.2-8.6) were both statistically signifi cant risk factors. The predictive model with these two characteristics had an ROC area under the curve of 0.68.In the univariate Cox regression analysis for enlargement only current smoking had a p-value < 0.20 (HR 4.5, HR 0.99-20.1), hence we performed no multivariate analyses for risk factors for enlargement.In the univariate Cox regression analysis for formation of new aneurysms and enlargement together, current smoking (HR 3.5, 95% CI 1.4-8.7), sex (HR 0.5 for men, 95% CI 0.2-1.2) and history of hypertension (HR 1.9, 95% CI 0.9-4.4) were related to formation or enlargement of aneurysms and included in the multivariate analysis. In the multivariate analysis current smoking (HR 3.8, 95% CI 1.5-9.4) and a history of hypertension (HR 2.3, 95% CI 1.1-4.9) remained statistically signifi cant. The predictive model with these two characteristics had an ROC area under the curve of 0.69.

Discussion

We found that screening in patients with a history of SAH has a high yield; in one-sixth of the patients an aneurysm was detected that had not been described before. In half of the patients in whom the CTA/IA-A at the time of the SAH was available for review, the detected

142

Chap

ter

aneurysm was de novo or had increased in size. This high risk of recurrent aneurysms indi-cates that development of intracranial aneurysms should not be considered as a once in a lifetime event but rather as a continuous process. Also, small aneurysms cannot always be considered stable lesions. Important risk factors for aneurysm development and enlargement of already existing aneu-rysms were multiple aneurysms, a history of hypertension and current smoking. Other risk factors in the univariate analysis were a positive family history and female sex. These risk factors, except of course the presence of multiple aneurysms, are similar to those for intracra-nial aneurysms and subarachnoid hemorrhage in general. Smoking and female sex have been identifi ed before as an important risk factor for new aneurysm formation and enlargement of already present aneurysms in patients with a history of intracranial aneurysms or SAH.6 Our predictive model for aneurysm formation and enlargement may help to identify those patients at highest risk for recurrent aneurysms and SAH. However, before it can be used in clinical practice it should be validated in a new group of patients. Nevertheless, the present data indicate that hypertension should be closely monitored and treated, and that patients should strongly be discouraged from starting or continuing smoking.To assess risk factors for formation of new aneurysms and enlargement of existing aneurysms we fi rst analyzed these events separately and subsequently together. The mechanisms of formation and enlargement may be rather similar, since for formation of an aneurysm enlar-gement from very small irregularities can be assumed. Moreover at surgery often weak spots in vessels are observed that possibly develop into an aneurysm.For the Cox regression analyses and the calculations of the number of new aneurysms per patient-year screening we assumed that the new aneurysms had developed halfway between the SAH and the date of screening. It is, however, unknown at what time the aneurysms we detected truly developed. In additional analyses we assumed that the new aneurysms had occurred at the date of screening. In this analysis the hazard ratios and the incidences of new aneurysm formation changed no more than 5% compared to the original analyses. We feel, therefore, confi dent that our assumption regarding the timing of aneurysm formation hardly affects the results of our study.MS-CTA is a technique that is suitable for screening since it is minimally invasive and has no vascular complications as in IA-A. The diagnostic properties of MS-CTA in detecting aneurysms in clipped patients are comparable with IA-A for aneurysms at new locations.7 However, for aneurysms adjacent to the clip site the test characteristics are less favorable because of clip artefacts.8 We therefore have probably underestimated the number of aneurysms that deve-loped at the clip site. In all three patients in whom a post-operative angiogram was available for review, the IA-A demonstrated complete clipping of the ruptured aneurysm. This shows that, as in coiling, the long-term durability of clipping is not perfect.Because of the age criteria for our study the population examined was somewhat younger than expected in an unselected population nine years after SAH. We did not include patients

143

Ten

in our study who were older than 70 years at the time of the screening since the risk of com-plications of treatment of aneurysms in this patient group is high and the life expectancy is relatively short. Therefore, if an aneurysm is detected at screening in most of these cases treatment will not be recommended. Our results therefore only apply for patients younger than 70 years at the time of screening.In the study hospitals aneurysms smaller than three millimeters are not treated butfollowed up over time. Since IA-A is not without risks, patients with a small aneurysm on CTA did not undergo IA-A for confi rmation. We cannot exclude the possibility that in some of these patients the fi nding classifi ed as aneurysm will be an infundibulum or a venous overprojection.From the 125 patients who had died and the 87 patients who stayed in a nursing home after discharge the cause of death or disability was a recurrent SAH in 12. By not including these patients in our calculations on the risk of de novo and regrowth aneurysms we haveunderestimated this risk. Conversely, had these 12 patients been included in the analysis of the study we would have overestimated the incidence of de novo and regrowth aneurysms because of confounding by indication. The reason for this confounding is that many other patients died and became disabled after recovery from the SAH from other causes than recurrent SAH. For proper analysis, the follow-up years of these patients and the presence or absence of new aneurysms at time of death or disability from other causes should have been included. Obviously no screening in these patients was performed and therefore, some of them might also have de novo or regrowth aneurysms. Including these patients as having no aneurysm would cause dilution of the true incidence of de novo and regrowth aneurysms. We therefore decided not to include these 12 patients in our calculations. The incidence of recurrent aneurysms in our study is therefore the risk found in a population that is alive and clinically suitable for screening. Our present study was not designed to assess the prevalence of recurrent SAH in patients with a history of SAH.The CTA/IA-A at time of SAH was available only for patients admitted for SAH after 1992. In these patients where the screening CTA could be compared with the initial CTA/IA-A the pro-portion of de novo aneurysms was one third. The remaining aneurysms detected at screening were, in retrospect, already present at time of the SAH. If the same proportion of de novo aneurysms is assumed in the patients for whom the initial CTA/IA-A was no longer available for comparison, the average risk of a de novo aneurysm was 0.65% per person year, with a relatively low risk in the fi rst years after SAH and a relatively high risk thereafter. This risk is slightly lower than in previously published estimates of de novo aneurysms, but in those stu-dies it is unclear if all angiograms at the time of SAH were available for review and therefore, additional aneurysms might have been classifi ed as de novo aneurysms.2-4 In our study we only included patients with aneurysms treated by neurosurgical clipping. It is likely that the number of de novo aneurysms in coiled patients will be comparable with the number of de novo aneurysms we found for the clipped patients. The numbers of regrowths or remnants

144

Chap

ter

caused by impaction of coils, however, might differ from those after clipping.A quarter of the aneurysms that were already present at the time of the SAH subsequently enlarged. Enlarging aneurysms have a relatively high risk of rupture, because size is an impor-tant determinant of the risk of rupture and perhaps also because enlarging aneurysms are unstable.6, 9 Small additional aneurysms that cannot be treated by coiling and that cannot be occluded at the same time as the ruptured aneurysm may need regular monitoring by CTA or angiography, but the yield and frequency of such monitoring requires a separate study.We showed that screening in patients with a history of SAH reveals many previously unde-tected aneurysms. The outcome of a recurrent SAH is as poor as that from a fi rst episode of SAH.10 Routine follow-up screening might, therefore, be effi cient after SAH, especially in patients with additional risk factors. Screening, however, also has disadvantages. Although with the advent of coiling the majority of aneurysms can be treated with relatively low risk of complications, preventive treatment can lead to disability and even death. Furthermore, small aneurysms will often not be treated. The knowledge of having an untreated aneurysm can lead to a decrease in quality of life.11 Before screening after SAH is implemented in clini-cal practice, the risks and benefi ts of screening should therefore be carefully weighted, for example in a decision model.

145

Ten

References

1. Rinkel GJE, Djibuti M, Algra A, van Gijn J. Prevalence and risk of rupture of intracranial aneurysms: A systematic review. Stroke 1998;29:251-256.2. David CA, Vishteh AG, Spetzler RF, Lemole M, Lawton MT, Partovi S. Late angiographic follow-up review of surgically treated aneurysms. J Neurosurg 1999;91:396-401.3. Juvela S, Porras M, Heiskanen O. Natural history of unruptured intracranial aneurysms: A long-term follow-up study. J Neurosurg 1993;79:174-182.4. Tsutsumi K, Ueki K, Morita A, Usui M, Kirino T. Risk of aneurysm recurrence in patients with clipped cerebral aneurysms. Results of long-term follow-up angiography. Stroke 2001;32:1191-1194.5. Bamford JM, Sandercock PA, Warlow CP, Slattery J. Interobserver agreement for the assessment of handicap in stroke patients. Stroke 1989;20:828.6. Juvela S, Poussa K, Porras M. Factors affecting formation and growth of intracranial aneurysms: A long-term follow-up study. Stroke 2001;32:485-491.7. Wintermark M, Uske A, Chalaron M, Regli L, Maeder P, Meuli R, Schnyder P, Binaghi S. Multislice computerized tomography angiography in the evaluation of intracranial aneurysms: A comparison with intra-arterial digital subtraction angiography. J Neurosurg 2003;98:828-836.8. Teksam M, McKinney A, Casey S, Asis M, Kieffer S, Truwit CL. Multi-section ct angiography for detection of cerebral aneurysms. AJNR Am J Neuroradiol 2004;25:1485-14929. Wiebers DO, Whisnant JP, Huston J, 3rd, Meissner I, Brown RD, Jr., Piepgras DG, Forbes GS, Thielen K, Nichols D, O’Fallon WM, Peacock J, Jaeger L, Kassell NF, Kongable-Beckman GL,Torner JC. Unruptured intracranial aneurysms: Natural history, clinical outcome, and risks of surgical and endovascular treatment. Lancet 2003;362:103-11010. Wermer MJH, Rinkel GJ, Greebe P, Albrecht KW, Dirven CM, Tulleken CA. Late recurrence of subarachnoid hemorrhage after treatment for ruptured aneurysms: Patient characteristics and outcomes. Neurosurgery 2005;56:197-20411. van der Schaaf IC, Brilstra EH, Rinkel GJE, Bossuyt PM, van Gijn J. Quality of life, anxiety, and depression in patients with an untreated intracranial aneurysm or arteriovenous malformation. Stroke 2002;33:440-443.

ChapterEleven

Effectiveness and costs of long-term follow-up screening for new aneurysms in patients with

clipped aneurysms after subarachnoid hemorrhage:a cohort study and decision model

M.J.H. Wermer, H. Koffi jberg, I.C. van der SchaafFor the ASTRA study group

148

Chap

ter

Abstract

Background and purpose

Patients who survive after subarachnoid hemorrhage (SAH) are at risk for a recurrence despite successful treatment of the ruptured aneurysm and may therefore benefi t from screening for new aneurysms.

Methods

We screened 610 SAH patients with CT-angiography 2-18 years after clipping of the aneurysm(s). Results of screening were used as input for a Markov decision model. We compared the ex-pected number of recurrent hemorrhages, life-expectancy, quality-adjusted life-years (QALY’s) and costs associated with the strategies “screening every fi ve years” and “no screening”.

Results

Screening individuals with previous SAH prevented almost half of the recurrences, slightly increased life-expectancy (from 21.06 to 21.08 years), but reduced QALY’s (from 12.18 to 12.04) and increased costs (from €2224 to €3369 per patient). Screening was cost-saving without increasing QALY’s in patients with a more than twofold risk above baseline of both aneurysm formation and rupture and it was cost-saving while increasing QALY’s if both risks were at least 4.5 times higher. In patients with a reduced quality of life because of fear for a recurrence, screening increased QALY’s at a maximum cost of €14087 per QALY.

Conclusions

In general, screening patients with previous SAH cannot be recommended. Screening can save costs and increase QALY’s in patients with a relatively high risk of both aneurysm formation and rupture, and increases QALY’s at acceptable costs in patients with fear for a recurrence. More data are needed on risk factors for aneurysm formation and rupture in patients with previous SAH and on management of fear for a recurrence to identify patients who can be-nefi t from screening.

149

Eleven

Introduction

Patients with previous subarachnoid hemorrhage (SAH) and successfully treated aneurysms are at risk for new episodes of SAH from new aneurysms at the original site or at a new location. The risk of bleeding from an unruptured aneurysm is higher in patients with a prior SAH than in patients without prior SAH.1 The risk of SAH after clipping is around 3% in 10 years, which is approximately 22 times higher than the risk in the general population.2, 3

With current non- or minimal invasive screening modalities and with the advent of coiling, aneurysms can now be detected and treated with relatively low risks.4 The case-fatality of a recurrent episode of SAH is about 40%, which is similar to that of a fi rst episode.5 Follow-up screening for aneurysms may therefore be benefi cial in patients with previous SAH. In a Markov decision model we showed that screening could not be recommended on the basis of the data then available.6 However, we also noted that the literature was very sparse on the most essential estimates in the model, especially the occurrence of new aneurysms. Furthermore, the costs of screening, treatment, and SAH were not included in that model.Therefore we systematically undertook CT-angiography (CTA) as a screening method for new aneurysms in patients with previous SAH and successful operative treatment of the aneurysm. By means of a Markov model with Monte Carlo simulations we evaluated the effectiveness, effi ciency and costs of screening.

Methods

Patients

From our databases of patients admitted with SAH we selected all patients who met the following inclusion criteria: 1) admission between 1985 and 2001 to the University Medical Center Utrecht, or between 1987 and 1998 to the Academic Medical Center Amsterdam; 2) confi rmation of the subarachnoid hemorrhage by CT and an aneurysm proven by CTA or conventional intra-arterial angiography (IA-A); 3) treatment of the ruptured aneurysm by clipping; 4) recovery to an at least partly independent state after treatment (defi ned as a score of ≤ 3 on the modifi ed Rankin scale)7 and 5) age between 18 and 70 years at time of screening. Excluded were patients 1) who were not able to communicate well enough to give informed consent; 2) with severe co-morbidity or markedly reduced life expectancy; 3) in whom one or more aneurysms had been treated by coiling or 4) who had been screened with CTA or IA-A within the last 3 years.For each eligible patient we contacted the general practitioner to verify whether the patient was still alive and had not moved to a nursing home. Subsequently we invited patients living independently to our outpatient clinic, where we informed them about the risk of subse-

150

Chap

ter

quent episodes of SAH and the pros and cons of screening and subsequent treatment. We obtained informed consent from all participants. The study was approved by the local ethicscommittees. Before the start of the study a number of 25 patients with newly detected aneu-rysms was considered adequate for a suffi ciently precise data analysis. On the assumption of a prevalence of new aneurysms of 2 to 3% (de novo or regrowth) the required number of participants was calculated at 600.

CT-angiography and conventional intra-arterial angiography

CTA of the brain was performed with multi-slice (MS)-CTA.3 In patients with an allergy to io-dinated contrast agents 1.5 tesla MRA was performed. Two neuroradiologists in each hospital independently evaluated all CTA’s for that hospital.IA-A was performed for confi rmation if CTA showed an aneurysm at the site of an existing clip. In patients with an aneurysm at a new site IA-A was performed only if the size of the aneurysm was at least three millimeters on CTA. In patients with an aneurysm smaller than three millimeters a follow-up CTA was performed after one year. IA-A was not performed in patients with a normal CTA.

Treatment and outcome

For all patients with an aneurysm confi rmed by IA-A we discussed treatment in a multi-disciplinary meeting with neurologists, interventional radiologists and neurosurgeons. The decision whether or not to treat was based on aneurysm and patient characteristics. Surgery and coiling were performed by experienced neurosurgeons and interventional radiologists. We recorded all complications of CTA, IA-A and treatment (for a new detected aneurysm).

(Cost)-effectiveness and effi ciency of screening

We developed a Markov decision model to estimate the long-term effects of screening. The model used in this study was adapted from our previous model.6 The principle of the Markov model is that it defi nes a number of health states and uses transition probabilities between these states. It assumes that at any point in time a patient is in one of these health states. Over time patients can move from one health state to another as defi ned by the transition probabilities. With the use of Monte Carlo simulation, the clinical courses of 100.000 hypo-thetical patients were individually evaluated. For the present study we added a new health state “well with a detected aneurysm” to the fi ve health states of the previous model “death”, “disabled”, “well with an undetected de novo/additional aneurysm”, “well with an undetected regrowth aneurysm” and “well treated after SAH” (Figure 1a). Figure 1b shows an example of one of the branches of the model. Details of the model and the transition probabilities are provided in the Appendix on page 163.

151

Eleven

The six different health states in our Markov model (ellipses). At any point in time a patient is classifi ed into one of these health states. Over a period of time (cycles) patients can move from one health state to another (arrows) as defi ned by the transition probabilities.

Figure 1

A

An example of one of the branches in the “no screening” strategy. At the start of our model, after SAH, patients in the “no screening” strategy start in the “well treated after SAH” state or in the branch of the state shown in Figure 1b: “well with a not detected de novo or additional aneurysm” if at the time of the SAH not all aneurysms present were detected. After one cycle (one year) in the health state “well with a not detected de novo or additional aneurysm” patients may have died from another cause than SAH, they may have had a rupture of an aneurysm or no event at all. If they have died from another cause they are classifi ed in the health state “death” and do not follow another cycle. If they had a ruptured aneurysm they are classifi ed in the health state “death”, “disabled” or “well treated after SAH” according to the outcome after SAH and treatment. If they had no events at all they are again classifi ed in the health state “well with a not detected de novo or additional aneurysm”. The complete Monte Carlo model can be provided by the authors on request.

B

152

Chap

ter

Variable Base-case value Range Source

Age 53.4 SD 9.1 - ASTRAMen 0.36 - ASTRAWomen 0.64 - ASTRATransition probabilities

Regrowth aneurysm after SAH/year 0.004 0.004-0.008 ASTRA,14

De novo/additional aneurysm after SAH/year 0.005 0.005-0.022 ASTRA,11,13-15

Risk of rupture de novo+additional/year 0.014 0.004-0.014 1,12,15

Risk of rupture regrowth/year 0.032 0.016-0.048 2,14

Risk of rupture small aneurysm/year 0.004 0.002-0.006 1

Risk growth small aneurysm/year 0.05 0.05-0.1 ASTRACTA false negative de novo 0.052 0.039-0.07 16

CTA false negative regrowth 0.4 0.15-0.65 17

CTA false positive de novo/regrowth 0.048 0.036-0.059 16

Disability conventional angiography 0.009 0.003-0.009 ASTRA,18,19

Disability preventive treatment by coils 0.014 0-0.02 1, 4

Disability preventive treatment by clips 0.024 0.024-0.058 1,20

Disability after recurrent SAH and treatment 0.14 0.05-0.32 5

Mortality conventional angiography 0.0001 0-0.0002 18,19

Mortality preventive treatment by coils 0.005 0-0.034 1,4

Mortality preventive treatment by clips 0.01 0.006-0.026 1,20

Mortality after recurrent SAH and treatment 0.33 0.18-0.53 5

Mortality recurrent SAH before hospital 0.12 0.05-0.15 21

Chance treatment de novo ≥ 3 mm 0.27 0.2-1 ASTRAChance treatment regrowth ≥ 3 mm 0.25 0.1-1 ASTRAChance of aneurysm < 3 mm 0.27 0.19-0.36 ASTRAChance clipping as treatment for de novo 0.4 0.3-0.5 ASTRAChance clipping as treatment for regrowth 0.7 0.6-0.8 ASTRA

UtilitiesUtility death 0 -Utility disabled 0.25 0.15-0.7 22

Utility known small untreated aneurysm 0.80 SD 0.20 0.7-0.9 ASTRAUtility healthy 0.87 SD 0.13 0.8-1.0 ASTRA

Table 1 The transition probabilities, utilities, age and sex distribution and the costs of screening and treatment used in the decision model

153

Eleven

In the model we compared the strategy “screening every fi ve years from the age of 20 until the age of 70” with the strategy “no screening”. The main outcome measures were the number of prevented SAH episodes, overall disability and mortality, life-expectancy, the number of quality-adjusted life-years (QALY’s) and the costs of both strategies.To account for the uncertainty of the transition probabilities and to investigate the effect of changing the screening interval from fi ve years to two or 10 years we performed sensitivity analyses. The transition probabilities and their sources used in the base-case analysis and the sensitivity analyses are listed in Table 1. With the results of the Monte Carlo simulation we estimated the number of patients needed to prevent one (fatal) episode of SAH and evaluated the cost-effectiveness of screening.The direct costs of CTA, IA-A and occlusive treatment of unruptured aneurysms and the direct costs of SAH including overhead costs were derived from two recent studies from the Netherlands.8, 9 In the cost-effectiveness analysis we used a 4% discount rate to account for time preference.

Results

Patients

Figure 2 shows the study population and the results from screening. Thirty-one patients were not approached because the inclusion number had already reached 600. The baseline characteristics of the 610 participants are shown in Table 2. The age and sex distribution of the non-participants was similar to that of the screened patients.

Costs in euros CTA 220 - 9

Conventional angiography 554 - 9

Preventive treatment coiling 9677 - 9

Preventive treatment clipping 8036 - 9

Visit outpatient clinic 100 23

Coiling after SAH 24435 - 8

Clipping after SAH 24435 - 8

Cost nursing home per day 206 - 23

Variable Base-case value Range Source

Table 1 Continued

154

Chap

ter

Figure 2 The study population and the results from screening

155

Eleven

Characteristics Participants (%)

Mean age at time of screening (years ± SD) 53.4 ± 9.1

Women (%) 391 (64%)

Modifi ed Rankin Scale Grade at time of screening0123

211 (35%)135 (22%)244 (40%)20 (3%)

Behavioral history at time of screeningCurrent smokerFormer smokerCurrent alcohol use > 5 drinks per dayFormer alcohol use > 5 drinks per dayCurrent use of recreational drugsFormer use of recreational drugs

312 (51%)190 (31%)27 (4%)22 (4%)14 (2%)17 (3%)

Medical history at time of screeningHistory of hypertensionIschemic strokeCoronary artery diseasePolycystic kidney disease

264 (43%)31 (5%)32 (5%)5 (1%)

Family history SAH / aneurysm at time of screeningMedically verifi ed family historyPossible family history, not medically verifi edNo family historyUnknown

44 (7%)52 (9%)505 (83%)9 (1%)

Number of aneurysms detected at time of SAH1 aneurysm2 aneurysms >2 aneurysms

511 (84%)64 (10%)35 (6%)

Location ruptured aneurysm at time of SAHAcom / ACAICA / PcomMCAVertebrobasilar

261 (43%)168 (28%)137 (22%)44 (7%)

Table 2 Baseline characteristics of the 610 participants

Acom / ACA = Anterior communicating artery / Anterior cerebral arteryICA = Internal carotid arteryPcom = Posterior communicating artery MCA = Middle cerebral arteryVertebrobasilar = Arteries of the vertebrobasilar system

156

Chap

ter

CT-angiography and conventional intra-arterial angiography

We found 129 new aneurysms in 96 of the 610 patients (16%; 95% CI 13-19%); 105 aneurysms were at a new site and 24 at the clip-site.10 CTA or IA-A at the time of the SAH was available for review in 59 of the 105 aneurysms at a new site; 19 (32%) were really new (de novo) and 40 (68%) were visible in retrospect (‘additional’). Three of the 24 patients with an aneurysm at the clip-site had undergone a control angiogram shortly after the operation, withoutshowing a remnant. Beside these 129 new aneurysms, 22 aneurysms already known to exist were confi rmed in 17 patients. Of 53 aneurysms in which the initial CTA/IA-A was still available, enlargement had occurred in 13 (25%). In total we found 151 aneurysms in 112 (18%; 95% CI 15-22%) of the 610 patients (one patient with a known aneurysm also had a newly detected aneurysm).

Treatment and outcome

Twenty-six patients were treated by coiling or clipping of a new aneurysm. The remaining patients with a newly detected aneurysm were scheduled for follow-up CTA after one year. No permanent complications occurred from CTA, IA-A or preventive treatment except for one patient who had slight cognitive defi cits as a result of a subdural hematoma after treatment by clipping.

(Cost)-effectiveness and effi ciency of screening

In the base-case analysis of the decision model screening resulted in a slightly smaller number of QALY’s at higher costs than the no screening strategy (Table 3). The expected number of prevented recurrences was too small to outweigh the increased disability and death rate from screening and preventive treatment of the aneurysms, and the decrease in QALY’s in patients in whom an aneurysm was detected but not treated. The total costs per patient were in the screening strategy €3369 and in the no screening strategy €2224. In the multivariate sensi-tivity analyses we varied the transition probabilities and utilities over the ranges shown in Table 1. Figure 3a shows that most of the outcomes of these analyses resulted in only a slight gain or loss of QALY’s at increasing costs.In the univariate sensitivity analyses the risks of formation and rupture of new aneurysms and the utility of the state “well treated after SAH” were the estimates that had the largest impact on the outcome of screening. Although the risks of formation and rupture had the largest impact, screening was not benefi cial when either of these two factors was increased from 1 to 10 times the baseline risk. However, screening saved costs without increasing QALY’s when both the risk of formation of aneurysms and the risk of rupture were more than twice the baseline risk (Table 3 and Figure 3b). When these risks (of formation and rupture) were at least 4.5 times the baseline risk, screening saved costs and increased QALY’s (Table 3 and Figure 3b).

157

Eleven

In the base-case analysis we had assumed that the average utility of the health state “well treated after SAH” was identical for screened and non-screened patients (0.87). However, knowledge of the risk of a recurrent SAH without the offer of screening might lead to a de-crease in the utility in the non-screened patients. If we decreased the utility for the state “well treated after SAH” in the non-screened patients with more than 1.5% (from 0.87 to less than 0.86) screening resulted in QALY gain (Figure 3b). The costs for one QALY gained were €14087 for a utility 2% lower than the baseline and decreased to €2477 for a utility 5% lower than the baseline utility (Figure 3b shows the incremental costs per patient). Changing the interval of screening from 5 to 2 or 10 years did not result in a cost-effective strategy.

Strategy NS S NS S NS S

Risk formationRisk rupture

Base-caseBase-case

2x increase2x increase

5x increase5x increase

Life expectancy (years) 21.06 21.08 20.71 20.88 19.38 19.88QALY’s per patient 12.18 12.04 12.00 11.90 11.25 11.28

Death totalDeath SAH/treatment after SAHDeath preventive treatment/IA-A

4159316760

4104198928

4430951610

42721299855

53441177240

491941209079

Disability totalDisability SAH/treatment after SAHDisability preventive treatment/IA-A

5195190

472270202

156515650

1270955315

538153810

41043625479

Number of SAHNumber of treatments after SAHNumber of preventive treatments

414436110

237320833714

12476109160

737664775993

43271379250

29606259809391

Total Number of CTA screens 0 278771 0 285443 0 285160

NNS to prevent 1 SAH (fatal SAH) - 56 (146) - 20 (46) - 7 (18)

Costs per patient in euros €2224 €3369 €6820 €7169 €26301 €22195

Table 3 The number of life-years, QALY’s, mortality and disability, prevented SAH, number needed to screen and costs for the strategy “screening” and the strategy “no screening” based on simulations of 100.000 patients

NS = no screening, S = screening, NNS = number needed to screen

158

Chap

ter

Figure 3 Cost-effectiveness planes: The upper left quadrant shows the strategies that result in a loss of QALY’s at increasing costs. The upper right quadrant shows the strategies that result in a gain of QALY’s at increasing costs. The lower left quadrant shows the strategies that result in a loss of QALY’s and saving of costs. The lower right quadrant shows the strategies that result in both a gain of QALY’s and saving of costs.

Cost-effectiveness plane of the multivariate sensitivity analyses. Each point in the fi gure represents one analysis in which for each of the variables an estimate was drawn from the distribution of these variables given in Table 1.

A

The effects of increasing the risk of both rupture and formation of aneurysms and decreasing the utility of the health state “well treated after SAH” in the not screened patients on the utility (QALY’s) and the costs per patient.

B

159

Eleven

Discussion

Screening all patients with previous SAH for new aneurysms seems not cost-effective. The baseline risks for formation and rupture of aneurysms used in the model were 0.5% per year for formation of de novo aneurysms, 0.4% per year for regrowth of original aneurysms, 1.4% per year for rupture of de novo aneurysms and 3.2% per year for rupture of regrowths. Scree-ning becomes cost-saving without increasing QALY’s in patients with a two-fold risk of both aneurysm formation and rupture, and cost-saving with increased QALY’s when these risks at least 4.5 times increased. Furthermore, screening results in gain of QALY’s at acceptable costs per QALY in patients with a decreased quality of life from fear for a recurrence.Currently, patients with a high enough formation and rupture risk to benefi t from screening cannot be clearly identifi ed. We previously found that smoking (Hazard ratio [HR] 3.8), hy-pertension (HR 2.3) and multiple aneurysms at the time of the initial SAH (HR 3.2) are risk factors for aneurysm formation.10 In a Finnish study, risk factors for aneurysm formation in patients with previous SAH were female sex (Odds ratio [OR] 4.7) and smoking (OR 4.1).11 No data exist on additional risk factors for rupture in patients after a fi rst episode of SAH but it is reasonable to assume that these do not greatly differ from risk factors for rupture in general. Unfortunately, in the only meta-analysis of risk factors for rupture no multivariate analyses could be performed.12 Risk factors for recurrent SAH are smoking (HR 6.5), multiple aneurysms at the time of SAH (HR 5.5) and age (HR 0.5 per 10 years), 3 but recurrent SAH represents a combination of formation and rupture of aneurysms. Thus, currently there is not enough information on separate determinants of development and rupture of new aneurysms in patients after SAH. Further studies are therefore needed to identify patients who may benefi t from screening.Two previous studies described long-term follow-up screening with conventional angiography in patients after SAH, but neither used a decision model to study the (cost)-effectiveness of screening.13,14 In one study 102 patients were screened after a mean follow-up period of 4.4 years (443 patient years of follow-up).14 The other study screened 80 asymptomatic patients (36% of the patients meeting the inclusion criteria) and additionally included 32 patients in whom follow-up angiography had already been performed for other reasons, after a mean interval of 9.3 years after operation (~1043 patient years of follow-up).13 In our study, which included more than 5000 patient-years of follow-up, 80% of the patients who met the inclusion criteria underwent screening. Although a similar number of aneurysms was detected in the two other studies (slightly higher than in our study), the authors reached opposing conclusions on whether or not screening in patients with previous SAH should be performed.We included only patients with clipped aneurysms since CTA is not a useful screening tool for coiled patients, given the extensive coil-artefacts. For coiled aneurysms MRI can be used, whereas this technique produces large artefacts around clipped aneurysms. It is likely that

160

Chap

ter

the number of de novo aneurysms in coiled patients is similar to that we found for clipped patients. The long-term occurrence of regrowth aneurysms or remnants caused by impaction of coils, however, might differ from clipping. In addition, after coiling the long-term outcome without screening in terms of recurrences of SAH is largely unknown. These data are neces-sary before a comparable decision analysis in coiling can be performed.The results from modelling studies such as ours should be interpreted with care, because of the uncertainty surrounding some of the estimates. From a theoretical point of view, a randomized trial would be the preferred method to investigate the effects of screening. However, given the long follow-up time (decades) required to collect suffi cient episodes of recurrent SAH to compare the two strategies, it is very unlikely that such a trial will ever be performed.Two of the most important estimates in our model were the risks of formation and rupture of recurrent aneurysms. The estimates on formation of aneurysms were derived from the pre-sent study. For the base-case analysis we derived the rupture risk for de novo and additional aneurysms from a meta-analysis.12 The rupture risks from the meta-analysis were somewhat higher than the rupture risk for additional aneurysms in patients with previous SAH found in the ISUIA study.1,12 The ISUIA data were used in our sensitivity analysis but not in the base-case analysis because the number of ISUIA patients with previous SAH and a 5-year follow-up is very small. With the estimates used in our model the 10-year cumulative incidence of recurrent SAH was 1.2%. This risk is lower than the actual incidence of recurrent SAH (~3% in 10 years).2,3 This suggests that our estimates of the risks of formation and rupture of new aneurysms are more likely an underestimation than an overestimation. Another important estimate appeared to be the utility that was assigned to the health state “well treated after SAH”. In the base-case analysis we assumed that the utility of this state was the same in the screening and no screening strategy. In a previous study we showed that SAH patients have a 22 times higher risk of a recurrence than the general population.3 Many patients are or will become aware of this increased risk. This insight, coupled with a state of not being screened will probably result in a decreased quality of life. In our model only a slight decrease in quality of life in the non-screened population led to a substantial increase in the benefi t of screening. Unfortunately, little is truly known about the health utility in patients who are aware of an increased health risk but are not offered screening, and about modifying factors. Further research on this subject is therefore needed to assess its effect on screening programs.In conclusion, in patients without known extra risks of new aneurysm formation or rupture, the risks of preventive treatment and the similar outcome for screening and no screening in terms of QALY’s should lead to discourage screening. All patients who smoke should strongly be encouraged to quit, and patients should be advised to have hypertension searched for and treated. More research is needed on determinants of fear for recurrence, impact of this fear on quality of life and on its management. In patients in a good clinical condition and known extra risk factors screening every fi ve years can be considered.

161

Eleven

References

1. Wiebers DO, Whisnant JP, Huston J, 3rd, et al. Unruptured intracranial aneurysms: Natural history, clinical outcome, and risks of surgical and endovascular treatment. Lancet 2003;362:103-110.2. Tsutsumi K, Ueki K, Usui M, Kwak S, Kirino T. Risk of recurrent subarachnoid haemorrhage after complete obliteration of cerebral aneurysms. Stroke 1998;29:2511-2513.3. Wermer MJH, Greebe P, Algra A, Rinkel GJE. Incidence of recurrent subarachnoid haemorrhage after clipping for ruptured intracranial aneurysms. Stroke 2005;36:2394-2399.4. Brilstra EH, Rinkel GJE, van der Graaf Y, van Rooij WJ, Algra A. Treatment of intracranial aneurysms by embolization with coils: A systematic review. Stroke 1999;30:470-476.5. Wermer MJH, Rinkel GJE, Greebe P, Albrecht KW, Dirven CM, Tulleken CA. Late recurrence of subarachnoid haemorrhage after treatment for ruptured aneurysms: Patient characteristics and outcomes. Neurosurgery 2005;56:197-204.6. Wermer MJH, Buskens E, van der Schaaf IC, Bossuyt PM, Rinkel GJE. Yield of screening for new aneurysms after treatment for subarachnoid haemorrhage. Neurology 2004;62:369-375.7. Bamford JM, Sandercock PA, Warlow CP, Slattery J. Interobserver agreement for the assessment of handicap in stroke patients. Stroke 1989;20:828.8. Roos YB, Dijkgraaf MG, Albrecht KW, et al. Direct costs of modern treatment of aneurysmal subarachnoid haemorrhage in the fi rst year after diagnosis. Stroke 2002;33:1595-1599.9. Halkes PHA, Wermer MJH, Buskens E, Rinkel GJE. Direct costs of surgical clipping and endovascular coiling of unruptured intracranial aneurysms. Cerebrovasc Dis 2006;22:40-45. 10. Wermer MJH, van der Schaaf IC, Velthuis BK, Algra A, Buskens E,Rinkel GJE. Follow-up screening after subarachnoid haemorrhage: Frequency and determinants of new aneurysms and enlargement of existing aneurysms. Brain 2005;128:2421-2429.11. Juvela S, Poussa K, Porras M. Factors affecting formation and growth of intracranial aneurysms: A long-term follow-up study. Stroke 2001;32:485-491.12. Rinkel GJE, Djibuti M, Algra A, van Gijn J. Prevalence and risk of rupture of intracranial aneurysms: A systematic review. Stroke 1998;29:251-256.13. Tsutsumi K, Ueki K, Morita A, Usui M, Kirino T. Risk of aneurysm recurrence in patients with clipped cerebral aneurysms. Results of long-term follow-up angiography. Stroke 2001;32:1191-1194.14. David CA, Vishteh AG, Spetzler RF, Lemole M, Lawton MT, Partovi S. Late angiographic follow-up review of surgically treated aneurysms. J Neurosurg 1999;91:396-401.15. Juvela S, Porras M, Heiskanen O. Natural history of unruptured intracranial aneurysms: A long-term follow-up study. J Neurosurg 1993;79:174-182.16. Wintermark M, Uske A, Chalaron M, et al. Multislice computerized tomography angiography in the evaluation of intracranial aneurysms: A comparison with intraarterial digital subtraction angiography. J Neurosurg 2003;98:828-836.17. Teksam M, McKinney A, Casey S, Asis M, Kieffer S, Truwit CL. Multi-section ct angiography for detection of cerebral aneurysms. AJNR 2004;25:1485-1492.18. Willinsky RA, Taylor SM, TerBrugge K, Farb RI, Tomlinson G, Montanera W. Neurologic complications of cerebral angiography: Prospective analysis of 2,899 procedures and review of the literature. Radiology 2003;227:522-528.19. Cloft HJ, Joseph GJ, Dion JE. Risk of cerebral angiography in patients with subarachnoid haemorrhage, cerebral aneurysm, and arteriovenous malformation: A meta-analysis. Stroke 1999;30:317-320.20. Raaymakers TW, Rinkel GJE, Limburg M, Algra A. Mortality and morbidity of surgery for unruptured intracranial aneurysms: A meta-analysis. Stroke 1998;29:1531-1538.21. Schievink WI, Wijdicks EF, Parisi JE, Piepgras DG, Whisnant JP. Sudden death from aneurysmal subarachnoid haemorrhage. Neurology 1995;45:871-874.

162

Chap

ter

22. Post PN, Stiggelbout AM, Wakker PP. The utility of health states after stroke: A systematic review of the literature. Stroke 2001;32:1425-1429.23. Oostenbrink JB, Bouwmans CAM, Koopmanschap MA, Rutten FFH. Manual for costing research 2004 (in Dutch). Health Care Insurance Board, Amstelveen, The Netherlands.

163

Eleven

Appendix

The Model

Classifi cation of aneurysms

The aneurysms were classifi ed as de novo (aneurysm located at a site remote from the original aneurysm), regrowth (aneurysm located at the same site as the originally treated one) or ad-ditional (aneurysm that is visible in retrospect but was not identifi ed on the CTA or DSA at the time of the initial SAH).

Health states

We defi ned the following six health states: 1) death; 2) disabled [after rupture of a new aneurysm, after complications from DSA or after surgical or endovascular treatment]; 3) well with an un-detected regrowth aneurysm [alive in a good clinical condition with an unidentifi ed aneurysm present at the site of the fi rst treated one]; 4) well with an undetected de novo/additional aneu-rysm [alive in a good clinical condition with a new aneurysm or a additional aneurysm located at a site remote from the original aneurysm]; 5) well with a detected [but untreated] aneurysm and 6) well treated after SAH (Figure 1a). At the start of the analysis we assumed that 3% of the patients had an additional aneurysm that was not detected at the time of screening and therefore these patients started in the health state “well with an undetected de novo/additional aneurysm”. The remaining patients started in the health state “well treated after SAH”.

Model assumptions

We made the following additional assumptions: 1) screening with CTA has no long-term effect on quality of life, since the risk of allergic contrast reactions is very small; 2) after the fi rst SAH all detected aneurysms present are completely obliterated; 3) patients who are disabled after a recurrent SAH and have another recurrence end up in such a poor condition that treatment of the ruptured aneurysm is not an option.

Utilities

A subjective factor in the model is the valuation of the quality of life or utility of the different health states. We assigned a utility of zero to the state death. Disability from rupture or com-plication of treatment was valued at 0.25. This utility is comparable with the utility after a major stroke and correlates with a Modifi ed Rankin Scale of 4 to 5.1 The health state well with a detected aneurysm was valued 0.8.2 Since a patent aneurysm in itself does not affect the

164

Chap

ter

quality of life of a patient who is not aware of it, the three remaining health states (well with an undetected de novo/additional or regrowth aneurysm and well treated after SAH) were as-signed a utility of 0.87. This is the utility of patients with a history of SAH who participated in the ASTRA study but had a normal outcome of screening.2 These utilities were modelled with a normal distribution with a maximum set at 1.

Discount rate

We used a 4% discount rate to account for time preference. Time preference is the preference to be in a healthy state now as opposed to later since individuals and society can benefi t from these healthy years in the interim.

Sensitivity analysis

With Monte Carlo simulation we performed one-way, two-way and multivariate sensitivity analyses. We evaluated the impact of changes in different transition probabilities and the utility of the health state disabled and the health state well with a detected aneurysm withlife-expectancy as time horizon.

Transition probabilities and utilities

We retrieved the transition probabilities for our model from the ASTRA study and the literature. A MEDLINE search was performed to identify relevant publications from 1985 to 2005. The fol-lowing keywords were used in different combinations: subarachnoid hemorrhage, aneurysm, recurrent, regrowth, residual neck, remnant, de novo, complications, outcome, treatment, clipping, coiling, sensitivity, specifi city, CT angiography and digital subtraction angiography.Table 1 in the article lists the estimates used.

Risk of development of a regrowth or de novo aneurysm after a fi rst episode of SAH

In the ASTRA study de novo aneurysms were found in 14 patients and possible de novo aneurysms were found in another 34 patients. One-third of the possible de novo aneurysms is probably a truly de novo aneurysm. Therefore, we estimated the risk of a de novo to be 25 (14 + 1/3*34) per 5078 follow-up years = 0.49% per year. The de novo’s in ASTRA were found after screening. Since the estimate for the de novo risk in the model is at the start of the tree before screening is performed we adjusted it for the sensitivity of the CTA for detecting de novo aneurysms by multiplying it by ((1 / 1- false negative risk of CTA for de novo’s). Therefore we used a de novo risk of 0.49/0.95= 0.5% per year. This is slightly lower than the de novo risks reported in the literature that vary between 0.84-2.2% per year.3-6

165

Eleven

In the ASTRA study we found three defi nite regrowth aneurysms (post-operative angiogram performed) and 21 possible regrowths (no post-operative angiogram performed) in 24 patients. We assumed that half of the possible regrowth aneurysms was a defi nite regrowth and not a remnant from the operation. Regrowth aneurysms in 13 patients (3 defi nite regrowths and 0.5*21 possible regrowths) in 5078 follow-up years resulted in 0.26% regrowths per year. We adjusted for the sensitivity of the CTA for regrowth aneurysms by multiplying 0.26% by ((1 / 1- false ne-gative risk of CTA for regrowths) = 0.26%*1.7=0.4%. This is comparable with the regrowth rates found in the literature that vary between 0.3-0.5%.3, 4 The upper limit of the range we used for the sensitivity analysis was calculated by adjusting a regrowth rate of 0.5% by the sensitivity of CTA for regrowth aneurysms.

Risk of rupture of regrowth and de novo/additional aneurysms

The reported rates for rupture of de novo and additional aneurysms vary from 0.5-1.4% per year and from 10-25% per 10 years.5, 7, 8 We used a rupture rate for de novo/additional aneurysms of 1.4% per year. We could not fi nd specifi c rupture rates for regrowth aneurysms in the literature. We computed a yearly rupture risk of 3.2% based on two studies reporting on regrowth after surgically treated aneurysms.3, 9

For aneurysms smaller than three millimeters in size we used a rupture risk of 0.4% per year derived from the ISUIA study.10 The risk of 0.4% was also used as lower limit for the rupture risk of de novo and additional aneurysms in the sensitivity analyses.

Risk of complications due to retreatment

For preventive clipping and coiling of de novo/additional aneurysms we used the ISUIA data and reviews on complications of preventive aneurysm treatment in general. 10-12 Retreatment of a regrowth aneurysm was considered to be technically more challenging. We therefore rated the risk of complications higher by incorporation of an additional factor of 1.5 times increased risk (range 1-2) for retreatment of a regrowth aneurysm.

Sensitivity and specifi city CTA

The false negative rate of MS-CTA for de novo/additional aneurysms was considered to be comparable with the false negative rate of MS-CTA for aneurysms in general. The false negative rate was calculated by 1 - the sensitivity. In the model a sensitivity of 94.8% with corresponding confi dence intervals was used.13 The false negative rate of MS-CTA for detecting regrowth aneurysms was derived from a study on the test characteristics of multi-slice CTA for regrowth aneurysms.14 For the calculation a sensitivity of 60% was used.

166

Chap

ter

The false positive rate of a regrowth and de novo/additional aneurysms was considered to be equal and was calculated by 1 - the specifi city of MS-CTA. For the model we used a specifi city of 95.2% with corresponding confi dence intervals.13

Risk of complications from DSA

In the ASTRA study none of the 86 patients died or were permanently disabled after DSA. In another screening study for intracranial aneurysms in our hospitals one of 25 patients was disabled after conventional angiography.15 We therefore estimated the disability rate of DSA to be 0.9%. This rate is slightly higher compared with two meta-analyses that reported an overall neurological complication risk 0.3%-0.5%.16, 17 We used a risk of 0.3% as lower limit and a risk of 0.9% as upper limit. In the meta-analyses no patient died from angiography. We estimated that the risk of dying due to DSA would be about 0.01% with a lower limit of zero and an upper limit of 0.02%

Risk of death and disability after recurrent SAH

We used a risk of dying of a recurrence before reaching the hospital of 12% with a range of 5-15%.18 When patients reached the hospital with a recurrent SAH an overall mortality rate of 33%, and an overall risk of disability rate of 14% was used with the corresponding confi dence intervals as range.19

Risk of detection and growth of a small aneurysm

In the ASTRA study an aneurysm smaller than three millimeters was found in 27% of the patients with a recurrent aneurysm. These aneurysms were not treated but followed over time by yearly CTA. The risk of growth of a small aneurysm was estimated to be 5% per year.

Treatment of recurrent aneurysms

In the ASTRA study 30% of the de novo/additional aneurysms with a size of at least three milli-meters and 25% of the regrowth aneurysms with a size of at least three millimeters were treated. These proportions were used in our model with a lower limit of 0.1-0.2 and an upper limit of 1. The chance that clipping was the preferred treatment for a de novo/additional aneurysm was 40% and for a regrowth 70% in the ASTRA study.

Risk of dying from causes other than SAH

We derived the age and sex specifi c annual probabilities of dying from national life-tables.20 These data were used to account for the fact that the cohort in the model ages.

167

Eleven

Age and gender distribution

The age and gender distribution of the patients was based on the 610 participants of the ASTRA study.

Costs

The costs of CTA, DSA and preventive treatment by means of clipping and coiling were derived from a series of patients who were treated for an unruptured aneurysm between 1997 and 2003 in the University Medical Center Utrecht.21 In this study 23 patients coiled in this period were matched with 23 clipped patients according to year of treatment, age and gender. All pre-admission costs for diagnostic procedures, all costs for treatment, and costs during follow-up including standard DSA control examinations at six and 18 months after coiling were compared for clipping and coiling. Costs were calculated as the product of the used resources and the costs of these resources including overhead costs. The direct costs of SAH were derived from a recent study from the Academic Medical Center Amsterdam.22 During a one-year period the direct costs of treatment, which included the costs of medical and nursing care and the related travel expenses of all admitted patients with SAH were calculated. No costs were assigned to the health states “death” and the four “well” health states. The costs of the health state “disabled” were estimated based on the costs of staying in a nursing home derived from the Dutch Manual for Costing: Methods and Standard Costs for Economic Evalu-ations in Health Care.23

168

Chap

ter

References Appendix

1. Post PN, Stiggelbout AM, Wakker PP. The utility of health states after stroke: A systematic review of the literature. Stroke 2001;32:1425-1429.2. Van der Schaaf IC, Wermer MJH, Velthuis BK, Buskens E, Rinkel GJE. Quality of life in patients with an untreated aneurysm detected by screening in patients with previously clipped aneurysms after subarachnoid hemorrhage. J Neurol Neurosurg Psychiatry 2006; 77:748-752.3. David CA, Vishteh AG, Spetzler RF, Lemole M, Lawton MT, Partovi S. Late angiographic follow-up review of surgically treated aneurysms. J Neurosurg 1999;91:396-401.4. Tsutsumi K, Ueki K, Morita A, Usui M, Kirino T. Risk of aneurysm recurrence in patients with clipped cerebral aneurysms. Results of long-term follow-up angiography. Stroke 2001;32:1191-1194.5. Juvela S, Porras M, Heiskanen O. Natural history of unruptured intracranial aneurysms: A long-term follow-up study . J Neurosurg 1993;79:174-182.6. Juvela S, Poussa K, Porras M. Factors affecting formation and growth of intracranial aneurysms: A long-term follow-up study. Stroke 2001;32:485-491.7. Rinkel GJE, Djibuti M, Algra A, van Gijn J. Prevalence and risk of rupture of intracranial aneurysms: A systematic review. Stroke 1998;29:251-256.8. International study of unruptured intracranial aneurysms (ISUIA) investigators. Unruptured intracranial aneurysms risk of rupture and risks of surgical intervention. N Engl J Med 1998;339:1725- 1733.9. Tsutsumi K, Ueki K, Usui M, Kwak S, Kirino T. Risk of recurrent subarachnoid haemorrhage after complete obliteration of cerebral aneurysms. Stroke 1998;29:2511-2513.10. Wiebers DO, Whisnant JP, Huston J, 3rd, et al. Unruptured intracranial aneurysms: Natural history, clinical outcome, and risks of surgical and endovascular treatment. Lancet 2003;362:103-110.11. Brilstra EH, Rinkel GJE, van der Graaf Y, van Rooij WJ, Algra A. Treatment of intracranial aneurysms by embolization with coils: A systematic review. Stroke 1999;30:470-476.12. Raaymakers TW, Rinkel GJE, Limburg M, Algra A. Mortality and morbidity of surgery for unruptured intracranial aneurysms: A meta-analysis. Stroke 1998;29:1531-1538.13. Wintermark M, Uske A, Chalaron M, et al. Multislice Computerized Tomography Angiography in the evaluation of intracranial aneurysms: A comparison with intraarterial digital subtraction angiography. J Neurosurg 2003;98:828-836.14. Teksam M, McKinney A, Casey S, Asis M, Kieffer S, Truwit CL. Multi-section CT angiography for detection of cerebral aneurysms. AJNR 2004;25:1485-1492.15. The magnetic resonance angiography in relatives of patients with subarachnoid haemorrhage (MARS) study group. Risks and benefi ts of screening for intracranial aneurysms in fi rst- degree relatives of patients with sporadic subarachnoid haemorrhage. N Engl J Med 1999;341:1344-1350.16. Cloft HJ, Joseph GJ, Dion JE. Risk of cerebral angiography in patients with subarachnoid haemorrhage, cerebral aneurysm, and arteriovenous malformation: A meta-analysis. Stroke 1999;30:317-320.17. Willinsky RA, Taylor SM, TerBrugge K, Farb RI, Tomlinson G, Montanera W. Neurologic complications of cerebral angiography: Prospective analysis of 2,899 procedures and review of the literature. Radiology 2003;227:522-528.18. Schievink WI, Wijdicks EF, Parisi JE, Piepgras DG, Whisnant JP. Sudden death from aneurysmal subarachnoid haemorrhage. Neurology 1995;45:871-874.19. Wermer MJH, Rinkel GJE, Greebe P, Albrecht KW, Dirven CM, Tulleken CA. Late recurrence of subarachnoid haemorrhage after treatment for ruptured aneurysms: Patient characteristics and outcomes. Neurosurgery 2005;56:197-204.20. National life tables. CBS offi ce of statistics Netherlands, Voorburg, The Netherlands.

169

Eleven

21. Halkes PHA, Wermer MJH, Buskens E,Rinkel GJE. Direct costs of surgical clipping and endovascular coiling of unruptured intracranial aneurysms. Cerebrovasc Dis 2006;22:40-45.22. Roos YB, Dijkgraaf MG, Albrecht KW, et al. Direct costs of modern treatment of aneurysmal subarachnoid haemorrhage in the fi rst year after diagnosis. Stroke 2002;33:1595-1599.23. Oostenbrink JB, Bouwmans CAM, Koopmanschap MA, Rutten FFH. Manual for costing research 2004 (in Dutch). Health Care Insurance Board, Amstelveen, The Netherlands.

ChapterTwelve

Yield of short-term follow-up CT/MR-angiography for small aneurysms detected at screening

M.J.H. Wermer, I.C. van der Schaaf, B.K. Velthuis, C.B. Majoie, K.W. Albrecht and G.J.E. Rinkel

172

Chap

ter

Abstract

Background and purpose

Patients with a history of subarachnoid hemorrhage (SAH) or familial intracranial aneurysms (FIA) are at increased risk for aneurysm formation and rupture. Small aneurysms detected at screening may be left untreated and followed over time. The yield of follow-up CT/MR-angi-ography (CTA/MRA) one or two years after detection to evaluate growth of these aneurysms is unknown.

Methods

We prospectively followed patients with small aneurysms detected at screening at a one-year interval using CTA or MRA. We assessed size, site and number of the aneurysms, and risk factors such as smoking, alcohol use and hypertension. We evaluated the short-term growth and rupture rate and possible risk factors for growth and rupture.

Results

Ninety-three patients, (67 with a history of SAH, 16 with FIA and 10 with both a history of SAH and FIA), with 125 aneurysms underwent CTA/MRA follow-up. Sixty-fi ve patients were followed-up once and 28 patients were followed-up twice (median follow-up time 1.3 years). In three of the 93 (3.2%) patients an aneurysm slightly enlarged (0.5 to 1.5 mm). Two patients (2.2%) had an SAH; one from an aneurysm at the clip-site from a previous operation that ruptured without enlargement and the other from a newly developed dissecting aneurysm. The only statistically signifi cant risk factor for growth and rupture was a history of both SAH and FIA (RR 10.1 95% CI 1.3-81.9).

Conclusions

The yield of early follow-up of small aneurysms in patients with a history of SAH or FIA is small and does not eliminate the risk of rupture. Whether follow-up at larger intervals than one year is useful requires further study.

173

Twelve

Introduction

Subarachnoid hemorrhage (SAH) from a ruptured intracranial aneurysm has a poor prognosis.1 Screening is often recommended in people with an increased risk of aneurysm formation and rupture, such as persons with familial intracranial aneurysms (FIA [more than two relatives with SAH or unruptured aneurysms]).2 The risk of new aneurysms and recurrent SAH is also increased in patients with a history of SAH. Screening these patients has a high yield with an aneurysm detected in 16% of the patients.3

If a very small aneurysm is found with screening, it is not always treated. Small aneurysms, however, can increase in size over several years. Enlarging aneurysms have a relatively high risk of rupture because size is an important determinant of the risk of rupture and possibly also because enlarging aneurysms are unstable.4, 5 Small aneurysms that are left untreated may be followed over time by noninvasive methods such as CT or MR angiography (CTA/MRA). Because the growth rate of aneurysms is largely unknown, it is unclear how frequent CTA or MRA should be performed. Frequent follow-up might increase the detection of growing aneurysms and subsequent treatment may prevent SAH. Conversely, frequent follow-up is also costly, has psychosocial consequences and in case of CTA might be harmful because of radiation effects or contrast allergic reactions.6, 7

We studied the yield of short-term serial follow-up by means of CTA or MRA for small aneu-rysms detected at screening in persons with FIA and patients with a history of SAH. We assessed the frequency and rate of growth and tried to identify risk factors for growth and rupture.

Methods

We prospectively performed short-term clinical and radiological follow-up in patients with a history of SAH and persons with FIA in whom an aneurysm smaller than fi ve millimeters was detected at screening but not treated. We identifi ed patients with a history of SAH from a database of patients from the University Medical Center Utrecht (UMCU) and the Academic Medical Center Amsterdam (AMC) who had participated in a study on screening for new aneurysms after SAH (the ASTRA study). In this study 610 patients, who had been admitted between 1985 en 2001 for SAH to the UMCU or AMC and in whom the ruptured aneurysm was treated by means of clipping were screened with CTA between 2002 and 2004. All patients were between 18 and 70 years of age at the time of the fi rst screening. The persons with FIA (defi ned as more than two fi rst-degree relatives with SAH or unruptured intracranial aneu-rysms) were retrieved from a hospital based screening register on familial SAH in the UMCU. In the UMCU, screening with MRA is performed every fi ve years in asymptomatic relatives with FIA. All patients with aneurysms smaller than fi ve millimeters detected at familial scree-

174

Chap

ter

ning that were left untreated were included. Excluded from the study were patients who had untreated fusiform aneurysms. At baseline we recorded risk factors such as size, site and number of the aneurysms that were detected, history of hypertension, smoking and alcohol use, and number of fi rst-degree relatives with intracranial aneurysms. In patients with a history of SAH, the small aneurysm detected at screening was fi rst classifi ed as A) aneurysm located at the clip from the previ-ous operation (regrowth) or B) aneurysm at a new location remote from the clip-site. For the aneurysms at a new location, the initial CTA or conventional intra-arterial angiogram (DSA) at the time of the SAH was reviewed, when available, to evaluate if the aneurysm was inretrospect visible. If the aneurysm was not present at the time of the SAH, it was classifi ed as “de novo” and if the aneurysm was present in retrospect, it was classifi ed as “additional”. For the additional aneurysms, growth in the time between the SAH and the detection at screening was assessed. This classifi cation (de novo or enlarged additional aneurysms as opposed to stable additional aneurysms) was evaluated as a potential risk factor for growth or rupture at short-term follow-up in patients with a history of SAH.For the additional aneurysms, not only “prospective” follow-up time (time from detection of the aneurysm at screening to the time of the follow-up CTA or MRA) but also “retrospective” follow-up time (time from the SAH to detection of the aneurysm) is available. Because the purpose of the study was to assess the yield of short-term CTA or MRA after detection of the aneurysm the follow-up time reported is the time from detection of the aneurysm, at screening to the time of the follow-up CTA or MRA. Follow-up CTA or MRA was scheduled at a one-year interval. In patients with a history of SAH, CTA was performed because MRA is associated with more artefacts in patients with ruptured aneurysms treated by means of clipping. In the AMC, CTA was performed on a 4-detector multi-slice CT (Philips Picker Mx8000); in the UMCU, CTA was performed on a 16-detector multi-slice CT (Philips Mx8000 LDT). The CT scans were performed with a fi eld of view of 160 and a slice thickness of 1 mm reconstructed at 0.5mm, resulting in a voxel size of 0.3x0.3x0.5mm. In persons with FIA a 1.5 Tesla MRA was performed (Philips Medical Systems). Both CTA and MRA have a high sensitivity in the evaluation of small intracranial aneurysms.8-10. Aneurysm size was measured in two directions (maximum length [neck to fundus] and maximum width). Enlargement was defi ned as increase in size of at least 0.5 millimeters in at least one direction. To avoid interobserver variation, the size of the aneurysm at follow-up CTA/MRA was assessed by the same neuroradiologist who measured the size of the aneurysm at detection. If patients did not show up for the follow-up CT/MR-angiography, we tried to contact the patient or the general practioner to ensure that the patient was still alive and had not had a SAH. We assessed the frequency of growth and rupture including corresponding 95% confi dence intervals (CIs). We evaluated possible risk factors for growth and rupture of aneurysms at fol-low-up combined by comparing the characteristics of the patients with growth or rupture at short-term follow-up with those of the patients with stable aneurysms at short-term follow-up. Continuous variables were compared by t-tests and categorical variables by γ2 tests.

175

Twelve

Results

We found 104 patients with an aneurysm smaller than fi ve millimeters detected at screening. In seven of the 104 (7%) patients the aneurysm was treated. Treatment in these patients was performed at their own request or was advised because rupture of aneurysms smaller than fi ve millimeters within two years of formation had occurred in the family. Four patients declined further follow-up. In total we performed follow up CTA or MRA in 93 patients; 67 patients with a history of SAH, 16 patients with FIA and 10 patients with both a history of SAH and FIA, with a total of 125 aneurysms. Eight (6%) of the 125 aneurysms were located at the clip-site from previous surgery for a ruptured aneurysm; the other aneurysms were located at different sites. The mean age of the patients was 51 years (range 20 to 69 years) and 75% of them were women (Table 1). The mean size of the aneurysms was three millimeters. Sixty-fi ve patients were followed-up once after detection of the aneurysm (median interval between detection and follow-up 1.1 years, range 0.7 to 2.2 years), and 28 patients were followed-up twice (median interval between detection and second follow-up 2.2 years, range 1.4-3.8 years). The median follow-up time of the 93 patients was 1.3 years (range 0.7-3.8 years).

Aneurysm growth

Enlargement of the aneurysm on the follow-up CTA occurred in three of the 93 patients (3.2 %; 95 % CI 0.8-9.8). All three patients had a history of SAH and two of them had also FIA. Enlarge-ment did not occur in any of the 16 patients who were screened for FIA but had no history of SAH themselves. None of the eight regrowth aneurysms enlarged. In two of the three patients enlargement was found at the fi rst follow-up and in one at the second follow-up. The fi rst patient, a 56-year old woman, had been treated in 1996 for a ruptured aneurysm of the left middle cerebral artery (MCA). Screening in 2003 revealed three new right MCA aneurysms that were not visible in retrospect on the DSA of 1996 and were classifi ed as de novo. On the follow-up CTA one year later, one of these aneurysms had enlarged from 2.8 x 2 mm to 3.5 x 2 mm. The second patient, a 38-year old woman, had been treated in 1993 after SAH from a left MCA aneurysm. Screening at the patient’s request in 1999 showed no new aneurysms. Screening in 2003 showed a right MCA aneurysm that could in retrospect be identifi ed on the CTA of 1999. The DSA of 1993 was no longer available for review in 2003. In retrospect the aneurysm had enlarged in size between 1999 and 2003 from 2 x 1 mm to 2 x 3 mm. A prospectivefollow-up CTA was performed in 2004, and the aneurysm had marginally enlarged from 2 x 3 mm to 2.5 x 3 mm. The third patient, a 58-year old woman, had been treated for a SAH in 1996. A 2.5x3 mm aneurysm of the anterior communicating artery (Acom) was detected on screening in 2003. The second follow-up CTA in 2005 showed an increase in size of this aneurysm to 4x4 mm (Figure 1). Because the enlargement was only modest in all three patients no treatment wasperformed.

176

Chap

ter

Characteristics patients Screened patients N=93 Growth/rupture N=4*

Age mean in years (± SD) 51 (10) 47 (8)

Women (%) 70 (75%) 4 (100%)

Mean follow-up in years (range) 1.6 (0.5-3.8) 1.4 (1.4-1.9)

Smoking (%)Current smokerFormer smokerNever smokerUnknown

53 (57%)27 (29%)7 (8%)6 (7%)

3 (75%)1 (25%)--

Alcohol use (%)Alcohol use > 5drinks a dayFormer use > 5 drinks a dayAlcohol use < 5 drinks a dayNeverUnknown

2 (2%)3 (3%)56 (60%)24 (26%)8 (9%)

--3 (75%)1 (25%)-

History of hypertensionYesNoUnknown

46 (49%)42 (45%)5 (5%)

2 (50%)2 (50%)-

Familial intracranial aneurysms (FIA)Medically verifi ed family historyPossible, not medically verifi edNo family historyUnknown

26 (28%)2 (2%)63 (68%)2 (2%)

2 (50%)-2 (50%)-

History of both SAH and FIAyesno

10 (11%)83 (89%)

2 (50%)2 (50%)

Number of untreated aneurysms123

69 (74%)16 (17%)8 (9%)

2 (50%)-2 (50%)

Size of aneurysm at detection**≥ 1 – 2≥ 2 - 3≥ 3 - 4≥ 4 – 5

8 (6%)61 (49%)43 (34%)13 (10%)

1 (13%)5 (62%)2 (25%)-

Location of aneurysm***Acom/ACAICAMCAVertebrobasilar

16 (13%)39 (31%)53 (42%)17 (14%)

3 (38%)-5 (62%)-

Classifi cation of aneurysm (n=55)****De novo or additional with enlargementAdditional stable in size

22 (40%)33 (60%)

3 (75%)1 (25%)

Table 1 Baseline characteristics of the participating patients and their aneurysms and the 4 patients with growth or rupture of the aneurysm on short-term follow-up

177

Twelve

Legend Table 1

* One patient with a recurrent SAH was not included in the risk-factor analysis because the aneurysm followed did not enlarge or rupture but a newly developed dissecting aneurysm caused the SAH.** maximum size in one of the two directions measured.*** Acom / ACA = Anterior communicating artery / Anterior cerebral arteryICA = Internal carotid artery MCA = Middle cerebral arteryVertebrobasilar = Arteries of the vertebrobasilar system**** the CTA or DSA at the time of the subarachnoid hemorrhage was available for review for 55 of the aneurysms in the patients with a history of SAH.

Figure 1 A 2.5x3 mm anterior communicating artery aneurysm was detected with CTA screening in 2003 in a 58-year old woman with a history of SAH in 1996. This aneurysm could be identifi ed in retrospect on the DSA at the time of the SAH but did not enlarge in the intervening years.a. Follow-up CTA in 2004 did not show enlargement of the aneurysm (arrowhead on axial maximum intensity projection slab of CTA).b. A second follow-up CTA in 2005 showed a small increase in size to 4x4 mm (arrowhead).

Aneurysm rupture

Rupture occurred in two of the 93 patients (2.2 %, 95% CI 0.4-8.3%). One patient had a recur-rent SAH from an aneurysm at the clip-site from a previous operation that ruptured without enlargement (Figure 2). The other had a SAH from a newly developed dissecting aneurysm of the A1 segment of the right anterior cerebral artery (Figure 3). Both patients had a history of SAH and one also had FIA. Rupture did not occur in any of the 16 patients who were screened for FIA but had no history of SAH themselves.

Risk factors for growth and rupture

Since one patient had a SAH from a new developed aneurysm instead from an aneurysm that was followed over time, she was not included in the risk factor analyses. The four remaining patients with growth or rupture of an aneurysm were all women (Fisher exact p= 0.57), were

178

Chap

ter

Figure 2 A 46-year old woman had been successfully treated for a ruptured anterior communicating artery (Acom) aneurysm in 1993. Postoperative angiography did not show an aneurysm remnant. In 2003 three aneurysms were detected (a de novo aneurysm at the right middle cerebral artery (MCA) of 3 mm, a de novo aneurysm of the left pericallosal artery of 2.8 mm and a regrowth aneurysm of the right Acom of 4.3 mm). Ia shows the CTA with axial maximum intensity projection slab, and Ib shows the conventional angiography of the left carotid artery with the regrowth (arrowhead) and the pericallosal aneurysm (arrow) performed in 2003. Although all three aneurysms were stable in size (IIa) on follow-up CTA one year later, she had a recurrent SAH six months after this CTA. Distribution of SAH on unenhanced CT (arrowhead in IIIa) suggests that the regrowth Acom aneurysm was the most likely cause of the SAH. All three aneurysms were still stable in size on DSA at the time of the recurrence (IIIb shows the stable regrowth aneurysm and the pericallosal aneurysm). The right MCA aneurysm is not shown.

on average younger (difference of means 4 years; 95% CI –6.1 to 14.1), more often had mul-tiple aneurysms at the time of the detection of the aneurysm (RR 3.0; 95% 0.4 to 22.9) and aneurysms located at the MCA (RR 3.5; 95% CI 0.7 to 9.6), more often were current smokers (RR 2.0; 95% 0.2 to 19.9), and had a de novo aneurysm or an additional aneurysm that had enlarged between the time of the SAH and detection at the fi rst screening (RR 4.5; 95% CI 0.5 to 40.5), but these risk factors did not reach statistical signifi cance. The only risk factor that was statistically signifi cant was a history of both SAH and FIA (RR 10.1; 95% CI 1.3 to 81.9).

179

Twelve

Discussion

We found that the yield of early follow-up of small untreated aneurysms in patients with a history of SAH or FIA is very small. Only a few aneurysms enlarged, and this enlargement was very small without consequences for treatment. The low rate of growth suggests that follow-up CTA one year after detection in patients with a high risk of SAH is not useful. This study also once again showed that rupture of small aneurysms remains unpredictable and enlargement is only one factor that is involved in the process of rupture. Despite absence of growth, an SAH occurred in one patient with a small stable regrowth aneurysm and in one other patient from a newly developed dissecting aneurysm. The only risk factor we found for growth or rupture at short-term follow-up was a history of both SAH and FIA.Two other studies have addressed the short-term radiological follow-up of unruptured aneu-rysms.11, 12 Neither specifi cally addressed small aneurysms, both included different categories of patients, and both had a wide range of follow-up. In a Japanese study, 166 aneurysms in

Figure 3 A 65-year old woman was treated in 2001 for a ruptured aneurysm of the left vertebral artery (arrow in a). A second additional 3 mm aneurysm was present on the origin of the left ophthalmic artery (arrowhead in b). No aneurysm was present on the right anterior cerebral artery (ACA) (arrow in c). At follow-up CTA in 2003 no aneurysm was present on the ACA (d), and the aneurysm of the ophthalmic artery was stable in size (not shown in fi gure d). In 2005 she presented with a recurrent SAH just before her second follow-up CTA was scheduled from a dissecting aneurysm on the right ACA (f). The ophthalmic artery aneurysm was still stable in size (arrowhead in e).

180

Chap

ter

140 patients were followed by means of CTA over a mean period of 17.7 months (range 3 to 84 months).11 Enlargement was found overall in 10 aneurysms (6.4%) and in three of 125 (2.4%) aneurysms smaller than fi ve millimeters. Another, retrospective study, that used serial MRA found that in four out of 57 patients (7%) with 62 unruptured aneurysms an aneurysm had enlarged during a mean follow-up time of 50 months (range 17-90 months). The median aneu-rysm size in this study was fi ve millimeters. However, none of the aneurysms that enlarged were smaller than nine millimeters in diameter and no enlargement occurred in the fi rst 23 months of follow-up.12 In both studies no aneurysm ruptured during the follow-up period. The results of our study cannot be easily be compared with the other two studies because of dif-ferences in patient populations. Only seven (3.5%) of the total 197 patients investigated in the other two studies had a history of SAH and 34 (17%) had a positive family history. In addition, the mean age of our patients was 10 years younger. Because we only included aneurysms of patients with a history of SAH or a positive family history, our results cannot be extrapolated to small aneurysms that are detected incidentally in persons without a history of SAH or a family history. However, because the rupture risk of incidental aneurysms is smaller than in patients with a history of SAH or a positive family history, it is likely that for these aneurysms short-term follow-up is not useful.5

In our study all patients had aneurysms smaller than fi ve millimeters. In our clinic, patients with a history of SAH or FIA with aneurysms of fi ve millimeters or larger are usually treated. The advantage of our study is that the proportion of treated patients with an aneurysm smaller than fi ve millimeters was known. Only a small number of patients were treated and therefore selection bias because of treatment of more aggressive or larger aneurysms is likely to be low. In most other studies on rupture risk or growth rate of aneurysms these data are unknown.5, 11, 12

Because only three of the aneurysms followed over time enlarged and one ruptured, the sample size for the univariate analysis was small and multivariate analysis for risk factors could not be performed. The risk factor we found should therefore be interpreted with some caution because it might be confounded with other risk factors. In addition, in our study some risk factors might not have become evident because of the small number of aneurysms with growth or rupture.If, for whatever reason, it is decided not to treat an unruptured aneurysm, radiographicfollow-up is an option. Our study shows that short-term follow-up by CTA or MRA of aneurysms smaller than fi ve millimeters is not helpful. The yield in terms of growing aneurysms is small and it does not eliminate the risk of rupture. These data should be discussed on an individual basis in patients with small aneurysms when decisions on management are made. Given de lack of effi ciency and safety, some patients might therefore choose treatment of the aneurysm. It remains unclear whether follow-up at longer intervals is useful. Long-term prospectivefollow-up data on large cohorts of patients with unruptured aneurysms that undergo frequent serial CTA or MRA are needed to study the effectiveness of such a follow-up and to provide more information on the mechanism and risk factors of aneurysm growth and rupture.

181

Twelve

References

1. Hop JW, Rinkel GJE, Algra A, van Gijn J. Case-fatality rates and functional outcome after subarachnoid hemorrhage: a systematic review. Stroke 1997; 28:660-4.2. Bederson JB, Awad IA, Wiebers DO, Piepgras D, Haley EC, Jr., Brott T, Hademenos G, Chyatte D, Rosenwasser R, Caroselli C. Recommendations for the management of patients with unruptured intracranial aneurysms: A Statement for healthcare professionals from the Stroke Council of the American Heart Association. Stroke 2000; 31:2742-50.3. Wermer MJH, van der Schaaf IC, Velthuis BK, Algra A, Buskens E, Rinkel GJE. Follow-up screening after subarachnoid haemorrhage: frequency and determinants of new aneurysms and enlargement of existing aneurysms. Brain 2005;128:2421-2429.4. Juvela S, Poussa K, Porras M. Factors affecting formation and growth of intracranial aneurysms: a long-term follow-up study. Stroke 2001; 32:485-91.5. Wiebers DO, Whisnant JP, Huston J, 3rd, Meissner I, Brown RD, Jr., Piepgras DG, Forbes GS, Thielen K, Nichols D, O’Fallon WM, Peacock J, Jaeger L, Kassell NF, Kongable-Beckman GL, Torner JC. Unruptured intracranial aneurysms: natural history, clinical outcome, and risks of surgical and endovascular treatment. Lancet 2003; 362:103-10.6. Wermer MJH, van der Schaaf IC, Van Nunen P, Bossuyt PM, Anderson CS, Rinkel GJE. Psychosocial impact of screening for intracranial aneurysms in relatives with familial subarachnoid hemorrhage. Stroke 2005; 36:836-40.7. Brenner DJ, Elliston CD. Estimated radiation risks potentially associated with full-body CT screening. Radiology 2004; 232:735-8.8. Prokop M, Galanski M. Computed Tomography of the body. Stuttgart: Georg Thieme Verlag, 2003:46- 48.9. Kato Y, Nair S, Sano H, Sanjaykumar MS, Katada K, Hayakawa M, Kanno T. Multi-slice 3D- CTA - an improvement over single slice helical CTA for cerebral aneurysms. Acta Neurochir (Wien) 2002; 144:715-22.10. Kouskouras C, Charitanti A, Giavroglou C, Foroglou N, Selviaridis P, Kontopoulos V, Dimitriadis AS. Intracranial aneurysms: evaluation using CTA and MRA. Correlation with DSA and intraoperative fi ndings. Neuroradiology 2004; 46:842-50.11. Matsubara S, Hadeishi H, Suzuki A, Yasui N, Nishimura H. Incidence and risk factors for the growth of unruptured cerebral aneurysms: observation using serial computerized tomography angiography. J Neurosurg 2004; 101:908-14.12. Phan TG, Huston J, 3rd, Brown RD, Jr., Wiebers DO, Piepgras DG. Intracranial saccular aneurysm enlargement determined using serial Magnetic Resonance Angiography. J Neurosurg 2002; 97:1023-8.

ChapterThirteen

Risk of rupture of unruptured intracranial aneurysms in relation to patient and aneurysm characteristics:

an updated meta-analysis

M.J.H. Wermer, I.C. van der Schaaf, A. Algra and G.J.E. Rinkel

184

Chap

ter

Abstract

Background and purpose

We updated our previous review from 1996 on the risk of rupture of intracranial aneurysms, aiming to include newly published papers.

Methods

We reviewed all studies from our former meta-analysis and performed a Medline search for new studies published after 1996. We calculated overall risks of rupture for studies with a mean follow-up time of < 5, 5-10 and > 10 years. Relative risks (RR) were calculated by comparing the risk of rupture in patients with and without potential risk factors. We aimed to perform multivariable analyses of the different risk factors with meta-regression analysis.

Results

We included 19 studies (10 new) with 4705 patients and 6556 aneurysms (follow-up 26122 patient-years). The overall rupture risks were 1.2% (FU <5 years), 0.6% (FU 5-10 years) and 1.3% (FU >10 years). In the univariable analysis, statistically signifi cant risk factors for rupture were age > 60 years (RR 2.0, 95% CI 1.1-3.7), female gender (RR 1.6, 95% CI 1.1-2.4), Japanese or Finnish descent (RR 3.4, 95% CI 2.6-4.4), size more than fi ve mm (RR 2.3, 95% CI 1.0-5.2), a posterior circulation aneurysm (RR 2.5, 95% CI 1.6-4.1) and a symptomatic aneurysm (RR 4.4, 95% CI 2.8-6.8). Meta-regression analysis yielded implausible results.

Conclusions

Age, gender, population, size, site and type of aneurysm should be considered in the decision whether or not to treat an unruptured aneurysm. Pooled multivariable analyses of individual data are needed to identify independent risk factors and to provide more reliable risk esti-mates for individual patients.

185

Thirteen

Introduction

Intracranial aneurysms are relatively common; approximately 2% of the adults harbor an un-ruptured aneurysm.1 With the ongoing improvement of imaging techniques, the chance that an asymptomatic aneurysm is detected has increased. In patients with unruptured aneurysms the decision whether or not to treat is often not straightforward. The risk of treatment has to be carefully balanced against the risk of rupture. Although the morbidity and mortality rates associated with clipping and coiling are relatively well known, the natural course of unruptured aneurysms remains controversial.2, 3

In 1996 our group performed a meta-analysis on the risk of rupture of unruptured intracranial aneurysms.1 In this meta-analysis, however, no multivariate analysis was performed. Moreover, since 1996 several new studies on the risk of rupture of aneurysms have been published.We updated our former meta-analysis with all relevant articles on the follow-up of unrup-tured aneurysms. Our aims were fi rstly to incorporate the new information in the existing pooled data, secondly to increase the amount of data in subgroups of patients according to location of the aneurysm, size of the aneurysm, and to clinical risk factors such as age, gender, smoking, a history of SAH or familial intracranial aneurysms, thirdly to perform multivariable analyses with meta-regression analysis, and fi nally to incorporate new insights on growth of aneurysms in the review.

Methods

We reviewed all publications on the risk of rupture of unruptured aneurysms used in the former meta-analysis.1 This meta-analysis included studies published from 1955 until 1996. We performed a new MEDLINE search to retrieve all articles on risk of rupture of unruptured aneurysms published between July 1996 and March 2006. The following key words were used in different combinations: unruptured, untreated, incidental, additional, symptomatic, risk of rupture, subarachnoid hemorrhage, intracranial aneurysm(s), intracerebral aneurysm(s), growth and follow-up. We searched the reference lists of all relevant publications for addi-tional studies. In addition, we checked the Web of Science for articles that cited our former meta-analysis.Studies were included 1) if the presentation of data included crude numbers or allowed recalculation into crude numbers; 2) if the type of aneurysm was identifi able (aneurysms were classifi ed as incidental if they were found with screening in asymptomatic individuals or with examination for symptoms unrelated to the aneurysms, as additional if they were found in patients with a history of SAH and as symptomatic if they caused symptoms other than SAH); 3) if in patients with a history of SAH and additional unruptured aneurysms the ruptured (“index”) aneurysm had been treated by clipping or coiling and 4) if in patients

186

Chap

ter

with previously treated aneurysms the source of subsequent bleeding was identifi ed by CT, surgery or autopsy (to exclude re-rupture of the previously treated aneurysm as cause for the hemorrhage).In some studies only subsets of patients met the inclusion criteria and therefore only these patients were included in the review. Studies that primarily evaluated growth of untreated aneurysms were only included if all patients were studied in whom follow-up was intended and the report was not restricted to those patients who had two or more follow-up scans. Case reports and papers published in another language than English were excluded. In case multiple publications reported on the same study population the most recent publication was used.

Data-extraction

Two reviewers (M.W./I.S. or M.W./G.R.) independently extracted data from the studies that met the inclusion criteria. Information was extracted on patient and aneurysm characteristics. In case of disagreement between the two reviewers, consensus was reached by joint review.The location of the aneurysms was classifi ed as follows: 1. posterior communicating artery (Pcom), 2. internal carotid artery (ICA) other than Pcom, 3. anterior circulation (anterior cerebral artery, anterior communicating artery and the pericallosal artery), 4. middle cerebral artery (MCA), 5. posterior circulation (vertebral artery, basilar artery, posterior cerebral artery) and 6. cavernous sinus. In most studies the Pcom was considered to be part of the ICA. Therefore, we calculated the risk of rupture of Pcom aneurysms in combination with the other ICA aneurysms and the risk of rupture of Pcom aneurysms alone and other ICA aneurysms alone. Because in the studies different cut points were used for aneurysm size we made the following categories: < 5 millimeters (mm), < 7 mm, 5-10 mm, > 10 mm, > 12 mm and >15 mm. No strict defi nition for familial intracranial aneurysms was used; aneurysms were classifi ed familial if the authors of the article under review reported them as familial. We assessed methodolo-gical quality of all included studies. The quality of a study was rated high when it fulfi lled allfollowing three criteria: 1. prospective study-design, 2. loss to follow-up less than 3% and 3. if a distinction was made between certain SAH (confi rmed by CT, MRI, autopsy or xanthochro-mia in the cerebrospinal fl uid) and possible SAH (from history or medical records) duringfollow-up. Finally, because the incidence of SAH is higher in Japan and Finland than in other western countries,4 we classifi ed the studies according to origin of study population.

Data-analysis

For data-analysis we prespecifi ed the following subgroups according to: age (decades), gender, family history of intracranial aneurysms, smoking (current versus former/never), hypertension, excessive alcohol use (> 5 glasses per day), location of the aneurysm, size of the aneurysm,

187

Thirteen

type of aneurysm (incidental, additional or symptomatic), prospective or retrospective study design, high quality studies versus studies of less quality, and origin of study population (Japanese/Finnish versus other populations).The risk of rupture was reported for studies with a mean follow-up time of < 5 years, with a mean follow-up time between 5-10 years and with a mean follow-up time of > 10 years. First, we used the “SAH per patient-years at risk” method to calculate the risk of rupture in the prespecifi ed subgroups. With this method we divided the number of SAH (in each subgroup) by the number of person-years or aneurysm-years of follow-up (in that subgroup), yielding the risk of SAH per patient-year. When the specifi c follow-up time in a certain subgroup could not be extracted from the article we multiplied the number of patients by the average period of follow-up of all patients to obtain the total number of person-years. Data were reported for those studies that reported the specifi c follow-up time for the prespecifi ed subgroups and for all studies combined (studies with specifi c follow-up times for subgroups and studies in which the average follow-up time for calculations was used).Second, we used Poisson meta-regression analysis to evaluate the infl uence of patient, aneurysm and study characteristics on the risk of rupture. In this analysis we used the same prespecifi ed subgroups as in the “SAH per patient-year at risk” method. Age of the patients was analyzed as continuous variable (mean age). The characteristics gender, family history of intracranial aneurysms, smoking, hypertension, excessive alcohol use, location of the aneu-rysm, size and type of aneurysm were incorporated in the analysis as proportion of patients with this particular characteristic. The size of the aneurysm was analyzed both as continuous variable (mean size) and by proportion of patients with an aneurysm of a certain size. De-sign of the study, study quality and population of the study were analyzed as dichotomous variables. Finally we assessed the infl uence of the mean follow-up time of the studies on the risk of rupture.

Results

Included studies

We found 23 studies (nine from the previous meta-analyses from 1996 and before and 14 new studies between 1996 and 2006) that fulfi lled the inclusion criteria. Three studies reported on patients who were also included in later publications and were therefore combined with these later studies,5-7 and one study was excluded because patients were selected on basis of availability of follow-up scans.8 The 19 included studies are listed in Table 1. The median year of publication was 1998 (range 1966-2005). If rupture of an aneurysm had occurred, the diagnosis SAH was established only by taking history of patients or their relatives in two studies,9,10 by review of medical records in two,11,12 by CT, MR, surgery or autopsy in eight7,13-19 or not specifi ed

188

Chap

ter

in seven studies.12,20-25 The follow-up of the patients was done by telephone in combination with reviewing medical records in four studies,11,15,25, 26 by annual questionnaires in one,14 by questionnaires in combination with review of medical records in two,16,27 by outpatient clinic visits, telephone calls and letters in seven7,9,10,13,17,18,20 or was not specifi ed in another fi ve stu-dies.12,21-24 Eleven studies reported the proportion of patients lost to follow up; this proportion was 0% in seven studies, and 0.2%, 5%, 6% and 35% in the other four studies (Table 1).

Patients

The 19 studies included a total of 4705 patients with 6556 aneurysms with a mean follow-up of 5.6 years (26122 patient-years). Seventeen studies provided data on the age of the patients; the weighted mean age was 55.6 years. Fourteen studies with 4148 patients provided data on the gender of the patients; 2891 (70%) were women.

Risk of rupture by the “SAH per patient-year at risk” method

The overall risk of rupture of untreated aneurysms in the studies with a mean follow-up < 5 years was 1.2% (95% CI 1.0-1.5), in the studies with a mean follow-up between the 5 and 10 years 0.6% (0.5-0.7%) and in the studies with a mean follow-up time > 10 years 1.3% (0.9-1.8). The patient characteristics that had a statistically signifi cant association with an increased risk of rupture of intracranial aneurysms were age > 60 years, female gender, and Japanese or Finnish descent (Table 2). In addition, smoking increased the risk of rupture but this factor was not statistically signifi cant. There were not enough data to evaluate the effects of excessive alcohol use or a family history of SAH on the risk of rupture of intracranial aneurysms. The aneurysm characteristics that were related to an increased risk of rupture were site at the posterior circulation, size larger than fi ve millimeters and symptoms caused by the aneurysm other than SAH (Table 3). The risk of rupture was lower in high quality studies than in studies with limited quality (Table 4). The relative risks found in studies that reported the specifi c follow-up time of the subgroups were mostly comparable with those in all studies combined but their CI was wider because of less data (left columns Tables 2 and 3).

189

Thirteen

First

aut

hor

Year

of

publ

icatio

nM

ean

FU ti

me

(rang

e)St

udy

desig

nCo

untr

yLo

ssto

FUNo

. of

patie

nts

No. o

f PY

No. o

f SA

HSA

H/PY

Lock

sley23

1966

3.4

(0-12

.0)

R?US

6%32

108

92.5

%Za

cks22

1980

2.8 (0

.1-7.5

)R

Cana

da?

1028

00%

Prze

lom

ski18

1986

6.4

(1.0-

12.0

)R

US0

958

00%

Eske

sen10

1987

2.1 (2

.0-2

.2)P

Denm

ark

5%22

464

8.7%

Wie

bers

2719

878.

3 (5.0

-?)

RUS

013

010

7915

1.4%

Inag

awa9

1992

5.2 (0

.5-10

.9)

R?Ja

pan

?47

244

10.

4%As

ari25

1993

3.6 (0

-9.7)

R?Ja

pan

?54

197

115.6

%M

izoi15

1995

4.3 (

0.4-

10.0

)R

Japa

n?

4921

18

3.8%

Yasu

i1119

976.

3 (0.

3-22

.5)R

Japa

n35

%23

414

6534

2.3%

ISUI

A I16

1998

8.3 (

?)R

US/C

an/E

ur0

1449

1202

332

0.3%

Kam

itani

3119

998.

6 (1.

3-20

.0)

RJa

pan

?11

953

3.2%

Tsut

umi13

2000

4.3 (

0.5-

17.0

)R

Japa

n0

6226

67

2.6%

Juve

la (c

ombi

ned)

6, 7,

3220

0018

.1 (0

.8-3

9.9)

P an

d R

Finl

and

014

225

7533

1.3%

Tsuk

ahar

a/In

oue5,

2020

022.0

(1.0

-4.9

)P

Japa

n?

11021

87

3.2%

Mat

sum

oto26

2003

2.6 (?

)R?

Japa

n?

9123

75

2.1%

ISUI

A II14

2003

4.1 (

0-6.

0)P

US/C

an/E

ur0.

2%16

9265

4451

0.8%

Yone

kura

2420

041.2

(0.5-

3.0)

PJa

pan

?32

137

84

1.1%

Mat

suba

ra12

2004

1.5 (0

.3-7.0

)P

Japa

n0

140

207

00%

Wer

mer

1720

051.6

(0.7-

3.8)

PNe

ther

land

s0

9214

30

0%

Tabl

e 1 O

verv

iew

of t

he 19

inclu

ded

stud

ies

?= u

nkno

wn/

unce

rtai

n, FU

= fo

lluw

-up,

R= re

trosp

ectiv

e, P=

pro

spec

tive,

No.=

num

ber, P

Y= p

atie

nt-y

ear

190

Chap

ter

Stud

ies w

ith sp

ecifi

ed FU

tim

e pe

r sub

grou

pAl

l stu

dies

with

dat

a

Varia

ble

No. o

fst

udie

sM

ean

FUTi

me

(rang

e)No

. of

PYNo

. of

SAH

Rela

tive

Risk

(9

5% C

I)No

. of

stud

ies

Rang

e of

FUNo

. of

PYNo

. of

SAH

Rela

tive

Risk

(9

5% C

I)

Age <

20 ye

ars

20-2

9 ye

ars

40-5

9 ye

ars

60-7

9 ye

ars

> 80

year

s

0 4 5 5 0

- 8.2 (

0-39

.9)

7.9 (0

-39.

9)5.3

(0-2

0.0)

-

- 815

1523

209

-

- 12 21 3 -

- 1.1 (0

.5-2.2

)Re

f1.0

(0.3-

3.5)

-

0 6 8 9 1

- 6.5 (

0-39

.9)

5.8 (0

-39.

9)4.

1 (0-

20.0

)1.2

(0.5-

3.0)

- 848

1830

709

12

- 12 24 19 0

- 1.1

(0.5-

2.2)

Ref

2.0 (1

.1-3.7

)-

Gend

erM

enW

omen

4 45.3

(0-2

0.0)

5.3 (0

-20.

0)72 21

81 11

Ref

3.6 (0

.5-28

.1)10 10

5.7 (0

-39.

9)5.7

(0-3

9.9)

2255

2885

32 65Re

f1.6

(1.1-

2.4)

Hype

rten

sion

No Yes

0 0- -

- -- -

- -4 4

6.5 (

0-39

.9)

6.5 (

0-39

.9)

2357

572-

35 9Re

f1.1

(0.5-

2.2)

Smok

ing

No Yes

1 118

.1 (0

.8-3

9.9)

18.1

(0.8

-39.

9)13

5212

2313 20

Ref

1.7 (0

.9-3

.4)

1 19.

8 (0

.7-39

.9)

9.8

(0.7-

39.9

)14

0413

0413 20

Ref

1.7 (0

.8-3

.3)

Popu

latio

nNo

n Ja

pane

se /F

inni

shJa

pane

se o

r Fin

nish

11 84.

6 (0

.3-39

.9)

5.3 (0

-12)

2042

260

93111 113

Ref

3.4

(2.6

-4.4

)11 8

4.6

(0.3-

39.9

)5.3

(0-12

)20

422

6093

111 113Re

f3.

4 (2

.6-4

.4)

Tabl

e 2 R

elat

ive r

isk o

f rup

ture

acc

ordi

ng to

pat

ient

char

acte

ristic

s

FU=

follu

w-u

p, No

.= n

umbe

r, PY=

pat

ient

-yea

r, Ref

= re

fere

nce

191

Thirteen

Stud

ies w

ith sp

ecifi

ed FU

tim

e pe

r sub

grou

pAl

l stu

dies

with

dat

a

Varia

ble

No. o

fst

udie

sM

ean

FUtim

e (ra

nge)

No. o

f PY

No. o

fSA

HRe

lativ

e Ri

sk

(95%

CI)

No. o

fst

udie

sM

ean

FUtim

e (ra

nge)

No. o

f PY

No. o

fSA

HRe

lativ

e Ri

sk

(95%

CI)

Site

of a

neur

ysm

ACA

ICA

inclu

ding

Pco

mIC

A w

ithou

t Pco

mPc

omM

CAVB Ca

vern

ous s

inus

4 4 3 3 4 4 1

4.3 (

0.1-2

2.5)

4.3 (

0.1-2

2.5)

4.9

(0.1-

20.0

)4.

9 (0

.1-20

.0)

4.3 (

0.1-2

2.5)

3.0 (0

.1-22

.5)2.8

(0.1-

7.5)

343

455

82 76 471

213

3

11 20 3 6 9 6 0

0.7 (

0.4-

1.5)

Ref

0.8

(0.3-

2.8)

1.8 (0

.7-4.

5)0.

4 (0

.2-1.

0)0.

8 (0

.3-2.8

)-

14 14 6 5 14 11 5

5.1 (0

-39.

9)5.1

(0-3

9.9)

5.5 (0

-20.

0)5.3

(0-2

0.0)

5.1 (0

-39.

9)4.

9 (0

-39.

9)6.

1 (0-

20.0

)

1083

3558

813

317

2734

791

2159

19 46 8 7 33 26 2

1.4 (0

.8-2

.3)Re

f0.

7 (0.

4-1.6

)1.7

(0.8

-3.8

)0.

9 (0

.6-1.

5)2.5

(1.6

-4.1)

0.1 (

0-0.

3)

Size

of a

neur

ysm

< 5 m

m<

7 mm

5-10

mm

>10

mm

>12 m

mgi

ant (

>15 m

m)

4 3 4 3 1 2

3.7 (0

.1-20

.0)

7.7 (0

.1-39

.9)

7.9 (0

.1-39

.9)

8.1 (

0-39

.9)

8.6

(1.3-

20.0

)6

(0-2

0.0)

565

2249

329

216

5 22

5 23 8 10 0 3

Ref

* 2.8 (0

.9-8

.4)

5.2 (1

.8-15

.3)- 15

.4 (3

.7-64

.5)

10 5 9 9 3 8

3.9 (0

.1-20

.0)

7.1 (0

.1-39

.9)

6.1 (

0.1-3

9.9)

6.2 (

0-39

.9)

5.7 (0

-20.

0)5.0

(0-3

9.9)

1939

7206

1187

3670

1089

293

10 32 14 55 42 18

Ref

* 2.3 (1

.0-5

.2)2.9

(1.5-

5.7)

7.5 (3

.8-14

.9)

11.9

(5.5-

25.8

)

Type

of a

neur

ysm

(%) I

ncid

enta

l (%

) Add

ition

al(%

) Sym

ptom

atic

4 2 2

5.5 (0

.1-22

.5)7.5

(0.3-

22.5)

5.9 (0

-22.5

)

1439

526

110

31 13 9

Ref

1.2 (0

.6-2

.2)3.8

(1.8

-8.0

)

12 8 8

5.5 (0

-39.

9)5.5

(0-3

9.9)

5.9 (0

-39.

9)

3315

3158

472

50 46 31

Ref

1.0 (0

.7-1.4

)4.

4 (2

.8-6

.8)

*bec

ause

size

< 5

is pa

rt o

f the

cate

gory

< 7

no re

lativ

e risk

for t

he su

bgro

up <

7 is

give

n.

Tabl

e 3 R

elat

ive

risk o

f rup

ture

acc

ordi

ng to

ane

urys

m ch

arac

teris

tics

FU=

follu

w-u

p, No

.= n

umbe

r, PY=

pat

ient

-yea

r, Ref

= re

fere

nce

192

Chap

ter

Variable No. ofstudies

Mean FU time (range)

No. of PY

No. ofSAH

Relative Risk (95% CI)

Study design Retrospectiveprospective

136

6.3 (0-39.9)2.1 (0-7.0)

185867929

15866

Ref1.0 (0.7-1.3)

Quality studies limited qualityhigh quality

172

5.6 (0-39.9)2.6 (0-6.0)

194357080

17351

Ref0.8 (0.6-1.1)

Table 4 Relative risk of rupture according to study design

Risk of rupture by the meta-regression analysis

In the univariable Poisson regression analysis the relative risk of the dichotomous variables study design (RR 1.0, 95% CI 0.7-1.3) and study quality (RR 0.8, 95% CI 0.6-1.1 for a high quality study) and Japanese or Finnish study population (RR 3.4, 95% CI 2.6-4.4) were identical to the relative risks found by the “SAH per patient-year at risk method” (results not shown in the tables). The continuous variable size had an RR of 1.05 (per 1 mm increase in size, 95% CI 0.93-1.18) and age of 1.06 (per 1 year increase in age, 95% CI 1.03-1.08). The RR for the mean follow-up time of a study was 0.97 (95% CI 0.94-1.01), meaning that the risk of rupture decreased with 3% for each additional year of follow-up in a study. The variables with proportions of patients with a certain characteristic showed a RR that was in the opposite direction compared with the “SAH per patient-year at risk method”. For example the RR for percentage women was 0.94 (95% CI 0.93-0.97) and the RR of aneurysms smaller than fi ve mm was 1.03 (95% CI 1.02-1.04), meaning that women had a lower risk of rupture than men and aneurysms smaller than fi ve mm had a higher risk of rupture than aneurysms of a larger size. Because the latter results of the univariable analysis were not considered plausible and most studies did not report enough data for all our prespecifi ed subgroups to allow multivariable analysis, no further regression analysis was performed.

Discussion

We found that patient characteristics increasing the risk of rupture are higher age, female gender, Japanese or Finnish descent and smoking, although this last factor was not statisti-cally signifi cant. Aneurysms characteristics that increase the risk of rupture are location at the posterior circulation, increasing size, and symptoms caused by the aneurysm other than SAH. In prospective studies the risk of rupture was similar to that in retrospective studies. In high quality studies the risk tended to be lower than the risk in studies of limited quality. We were not able to perform multivariable analysis because meta-regression analysis yielded

FU= folluw-up, No.= number, PY= patient-year, Ref= reference

193

Thirteen

implausible results.The addition of new papers to the previous review resulted in an increase in patient-years from 3906 to 26122 with narrowing of the confi dence intervals surrounding the estimates. Furthermore, in the present meta-analysis additional risk factors such as smoking, hyperten-sion and Japanese or Finish origin were assessed. In Japan and Finland the incidence of SAH is much higher than that in other western countries.4 In Finland the prevalence of intracranial aneurysms is similar to that in other countries.28 To our knowledge, comparable Japanese data on the prevalence of aneurysms are lacking. Our results suggest that the increased risk of rupture of intracranial aneurysms is an important reason for the high SAH risk in the Finnish and Japanese population.Because individual risk factors for rupture might be infl uenced by other risk factors we aimed to perform a multivariable analysis by means of meta-regression analysis. Unfortunately, this method appeared to be not suitable for analysis of several risk factors in our study. We found in the meta-regression analysis a statistically signifi cant higher risk of rupture in aneurysms smaller than fi ve millimeters (RR > 1) compared with large aneurysms when size was incorpo-rated as proportion of aneurysms within a certain size category. The most likely explanation for these contradictive results is that the meta-regression analysis is not based on crude data. Size can be incorporated as the proportion of aneurysms smaller than fi ve millimeters in a study as a risk factor for aneurysmal rupture. The outcome of such an analysis is that when the proportion of aneurysms smaller than fi ve millimeters increases with one percent, the risk of rupture changes with a certain factor X. When the meta-analysis includes a study with a relatively low overall rupture risk and a percentage of aneurysms smaller than fi ve millimeters of 20% and a study with a higher overall rupture risk and a percentage of aneurysms smaller than fi ve millimeters of 40%, the conclusion of the analysis will be that when the proportion of aneurysms smaller than fi ve millimeters increases the risk of rupture also increases (RR for aneurysms smaller than fi ve mm > 1). However, the meta-regression method ignores the fact that all SAHs in the study with 40% might have appeared in the large aneurysms. Furthermore, even if the univariable meta-regression analysis had shown plausible results, multivariable analyses could only have been performed for very few variables because many studies did not report data for all of the subgroups and would have been excluded from the multivariable analysis. Multivariable analysis is therefore only possible by pooling the crude patient data of multiple studies.Because we could not perform multivariable analysis with the data presently available in the literature, we could not assess the independent contribution of patient and aneurysm characteristics to the risk of aneurysm rupture. We found that both large aneurysms and symptomatic aneurysms had a high risk of rupture. It is, however, unlikely that these risk fac-tors are independent because aneurysms that cause cranial nerve palsies are often large.29 We could not confi rm a higher risk of rupture in additional aneurysms compared with incidental aneurysms as suggested in both ISUIA studies. However, the lack of a difference in rupture risk between incidental and additional aneurysms in our study should not be considered as proof

194

Chap

ter

of absence of such a difference. A potential explanation is that the additional aneurysms were smaller than the incidental aneurysms; a higher risk for additional aneurysms may therefore have been masked by smaller size with inherently lower rupture risk.Unfortunately most studies did not provide specifi c data on all subgroups of patients. The number of patients, the number of SAHs or both the number of patients and SAHs were frequently not reported for the subgroups of our interest. In six studies not even complete data on age and gender of the patients were reported. Although the ISUIA studies together included 3141 patients and have therefore a great impact on the overall risk of rupture in our study, most of their data could not be used for the subgroup analysis because of lack of detailed information. Furthermore, in most studies limited information was provided on study-design, methods, completeness of follow-up and data-analysis and only two studies fulfi lled our criteria for high quality.14, 17

Except for size, other aneurysm characteristics may be involved in the risk of rupture. Aneu-rysms of irregular shape or with nipples might have higher risks of rupture and trombosed or calcifi ed aneurysms lower risks, but these factors have not been taken into account in the parent studies. Moreover, bias may have been introduced through selection of patients for treatment. For example, unruptured aneurysms in old and sick patients with cerebrovascular diseases might be left untreated. It is unclear how these factors involved in the treatment decision have infl uenced the results of the studies. In addition, in some studies patients in the initially conservative group were treated during follow-up of the study. In the ISUIA II study this proportion of patients was almost one-third.14 Although the reasons for treatment were not specifi ed, it is likely that the most frequent reason is growth of the aneurysm at serial follow-up. Because enlarging aneurysms have a higher risk of rupture, treating patients with growing aneurysms probable resulted in an underestimation of the risk of rupture.30

The “SAH per patient-year at risk method” assumes constant rupture rates of aneurysms over the years. In a mathematical model we recently found that growth of intracranial aneurysms is probably not constant and time independent but rather an irregular and discontinuous process with periods with and without growth (H. Koffi jberg et al. 2006; unpublished data). In our opinion, it is not correct to assume that the average rupture risks per year calculated by the “SAH per patient-year at risk“ method hold true for the rest of a patients life. In our regression analysis, we found that the risk of rupture tends to decrease for every (mean) year increase in follow-up. It would be better to calculate rupture risks in relation to the follow-up time of patients (for example the rupture risk in all patients followed during the fi rst year after aneurysm detection, during the second year etcetera). However, for this calculation again the crude patient data are needed. Because these data are not available we pooled the rupture risks of the studies based on mean follow-up time and range of follow-up. Thus, we think that more reliable risks of rupture are reported for defi ned periods of time.We conclude that the main patient and aneurysm risk factors for rupture of intracranial aneurysms are higher age, female gender, Japanese of Finnish descent, larger size, location

195

Thirteen

of the aneurysm at the posterior circulation and symptoms caused by the aneurysm. It is not known to what extent these risk factors are independent of each other. Although over the last 40 years 6556 aneurysms have been followed for over 26515 years, it is still not possible to perform multivariable analysis with the data that are currently available. Therefore, un-certainty still abounds for individual risk calculation. New follow-up studies on intracranial aneurysms should have a prospective study design, provide detailed information on follow-up of patients and data-analysis, and report the number of SAH and the number of follow-up years for all subgroups of patients. Because meta-regression analysis is not a suitable method for multivariable analysis of risk factors for rupture, collaborative efforts with pooled analysis of individual data are needed to identify independent risk factors for aneurysm rupture. In this pooled analysis also the follow-up time should be taken into account because the growth of aneurysms is probably not constant over time. Only in this way more reliable risk estimates will be available to enable physicians and patients to make a sound decision on whether or not to treat an unruptured aneurysm.

196

Chap

ter

References

1. Rinkel GJE, Djibuti M, Algra A, van Gijn J. Prevalence and risk of rupture of intracranial aneurysms: A systematic review. Stroke 1998;29:251-256.2. Brilstra EH, Rinkel GJE, van der Graaf Y, van Rooij WJ, Algra A. Treatment of intracranial aneurysms by embolization with coils: A systematic review. Stroke 1999;30:470-476.3. Raaymakers TW, Rinkel GJE, Limburg M, Algra A. Mortality and morbidity of surgery for unruptured intracranial aneurysms: A meta-analysis. Stroke 1998;29:1531-1538.4. Linn FH, Rinkel GJE, Algra A, van Gijn J. Incidence of subarachnoid hemorrhage: Role of region, year, and rate of computed tomography: A meta-analysis. Stroke 1996;27:625-629.5. Inoue T. Treatment of incidental unruptured aneurysms. Acta Neurochir Suppl 2002;82:11-15.6. Juvela S, Porras M, Heiskanen O. Natural history of unruptured intracranial aneurysms: A long-term follow-up study. J Neurosurg 1993;79:174-182.7. Juvela S, Porras M, Poussa K. Natural history of unruptured intracranial aneurysms: Probability of and risk factors for aneurysm rupture. J Neurosurg 2000;93:379-387.8. Phan TG, Huston J, 3rd, Brown RD, Jr., Wiebers DO, Piepgras DG. Intracranial saccular aneurysm enlargement determined using serial Magnetic Resonance Angiography. J Neurosurg 2002;97:1023- 1028.9. Inagawa T, Hada H, Katoh Y. Unruptured intracranial aneurysms in elderly patients. Surg Neurol 1992;38:364-370.10. Eskesen V, Rosenorn J, Schmidt K, Espersen JO, Haase J, Harmsen A, Hein O, Knudsen V, Marcussen E, Midholm S, et al. Clinical features and outcome in 48 patients with unruptured intracranial saccular aneurysms: A prospective consecutive study. Br J Neurosurg 1987;1:47-52.11. Yasui N, Suzuki A, Nishimura H, Suzuki K, Abe T. Long-term follow-up study of unruptured intracranial aneurysms. Neurosurgery 1997;40:1155-1159.12. Matsubara S, Hadeishi H, Suzuki A, Yasui N, Nishimura H. Incidence and risk factors for the growth of unruptured cerebral aneurysms: Observation using serial Computerized Tomography Angiography. J Neurosurg 2004;101:908-914.13. Tsutsumi K, Ueki K, Morita A, Kirino T. Risk of rupture from incidental cerebral aneurysms. J Neurosurg 2000;93:550-553.14. Wiebers DO, Whisnant JP, Huston J, 3rd, Meissner I, Brown RD, Jr., Piepgras DG, Forbes GS, Thielen K, Nichols D, O’Fallon WM, Peacock J, Jaeger L, Kassell NF, Kongable-Beckman GL, Torner JC. Unruptured intracranial aneurysms: Natural history, clinical outcome, and risks of surgical and endovascular treatment. Lancet 2003;362:103-110.15. Mizoi K, Yoshimoto T, Nagamine Y, Kayama T, Koshu K. How to treat incidental cerebral aneurysms: A review of 139 consecutive cases. Surg Neurol 1995;44:114-120.16. International study of unruptured intracranial aneurysms investigators. Unruptured intracranial aneurysms--risk of rupture and risks of surgical intervention. N Engl J Med 1998;339:1725-1733.17. Wermer MJH, van der Schaaf IC, Velthuis BK, Majoie CB, Albrecht KW, Rinkel GJ. Yield of short-term follow-up CT/MR-angiography for small aneurysms detected at screening. Stroke 2006;37:414-41818. Przelomski MM, Fisher M, Davidson RI, Jones HR, Marcus EM. Unruptured intracranial aneurysm and transient focal cerebral ischemia: A follow-up study. Neurology 1986;36:584-587.19. Wiebers DO. Patients with small, asymptomatic, unruptured intracranial aneurysms and no history of subarachnoid hemorrhage should generally be treated conservatively: For. Stroke 2005;36:408-409.20. Tsukahara T, Murakami N, Sakurai Y, Yonekura M, Takahashi T, Inoue T. Treatment of unruptured cerebral aneurysms--a multi-center study of Japanese national hospitals. Acta Neurochir Suppl 2002;82:3-10.

197

Thirteen

21. Kamitani H, Masuzawa H, Kanazawa I, Kubo T. Bleeding risk in unruptured and residual cerebral aneurysms--angiographic annual growth rate in nineteen patients. Acta Neurochir (Wien) 1999;141:153-159.22. Zacks DJ, Russell DB, Miller JD. Fortuitously discovered intracranial aneurysms. Arch Neurol 1980;37:39-41.23. Locksley HB. Natural history of subarachnoid hemorrhage, intracranial aneurysms and arteriovenous malformations. J Neurosurg 1966;25:321-368.24. Yonekura M. Small unruptured aneurysm verifi cation (Suave study, Japan)-interim report. Neurol Med Chir (Tokyo) 2004;44:213-214.25. Asari S, Ohmoto T. Natural history and risk factors of unruptured cerebral aneurysms. Clin Neurol Neurosurg 1993;95:205-214.26. Matsumoto K, Akagi K, Abekura M, Nakajima Y, Yoshiminie T. Investigation of the surgically treated and untreated unruptured cerebral aneurysms of the anterior circulation. Surg Neurol 2003;60:516- 522.27. Wiebers DO, Whisnant JP, Sundt TM, Jr., O’Fallon WM. The signifi cance of unruptured intracranial saccular aneurysms. J Neurosurg 1987;66:23-29.28. Ronkainen A, Miettinen H, Karkola K, Papinaho S, Vanninen R, Puranen M, Hernesniemi J. Risk of harboring an unruptured intracranial aneurysm. Stroke 1998;29:359-362.29. Teasdale E, Statham P, Straiton J, Macpherson P. Non-invasive radiological investigation for oculomotor palsy. J Neurol Neurosurg Psychiatry 1990;53:549-553.30. Juvela S, Poussa K, Porras M. Factors affecting formation and growth of intracranial aneurysms: A long-term follow-up study. Stroke 2001;32:485-491.31. Kamitani H, Masuzawa H, Kanazawa I, Kubo T. Bleeding risk in unruptured and residual cerebral aneurysms-- angiographic annual growth rate in nineteen patients. Acta Neurochir 1999;141:153- 159.32. Juvela S. Natural history of unruptured intracranial aneurysms: Risks for aneurysm formation, growth, and rupture. Acta Neurochir Suppl 2002;82:27-30.

ChapterFourteen

General DiscussionSamenvatting

SummaryDankwoord

List of publicationsCurriculum Vitae

200

Chap

ter

Part I: Long-Term follow-up after SAHPart I: Long-Term follow-up after SAH

Main fi ndingsMain fi ndings

Patients who make a good functional recovery and are independent for activities of daily living (ADL) after SAH may encounter several problems in the years following the SAH. Firstly, patients have a 22 times higher risk of a new episode of SAH than the general population. This risk is especially increased in patients who are young, who smoke and who have multiple aneurysms at the time of the initial SAH. Around 90% of patients with a recurrent episode are women. The outcome of such a recurrent SAH is as poor as the initial episode. Secondly, the mortality rate among SAH patients who recovered to an independent state is three times that of the general population. The excess mortality is most prominent in women and younger people. In the fi rst 10 years after the SAH one out of every nine patients experiences avascular event other than recurrent SAH. This risk, however, is lower than that in patients after TIA or minor stroke. Thirdly, anosmia is common in SAH survivors: one out of every three pa-tients treated neurosurgically and one out of every six patients treated endovascularly reports a decreased or loss of smell after the SAH. In the majority of patients anosmia is permanent and has a considerable impact on quality of life. Finally, patients who recover to independence for ADL after SAH often notice long-term psychosocial effects on employment, relationships, personality and mood. Half of the SAH patients stop working or work shorter hours or hold a position with less responsibility. One out of every 14 patients gets a divorce for which they blame the sequelae of the SAH and two-thirds of the patients notice changes in their perso-nality. Most frequent reported personality changes are increased agitation or emotionality. SAH patients have higher mean depression scores compared with the reference population and more often have scores in the range of a probable depression or an anxious state.

Implications for patient care and future researchImplications for patient care and future research

Even for patients who make a good recovery, SAH has important long-term implications for health and well-being. Because the risk of a recurrent SAH or other vascular diseases is considerable, taking measures to prevent these events is important. Patients who smoke should be strongly encouraged to quit. Hypertension is another known risk factor for SAH and should be searched for and treated in every patient.1 It is unknown whether secondary prevention of vascular events by means of cholesterol or blood pressure lowering drugs or anti-platelet therapy is indicated in patients who recover after SAH. Although we found that the risk of vascular events is lower in SAH patients than in patients after a TIA or minor stroke, it may still be higher than in the general population. Unfortunately, at this moment no detailed population based data (e.g. age- and gender- specifi c) on the incidence of stroke or myocardial infarction are available in the Netherlands. Therefore, we cannot directly compare

201

Fourteen

the vascular event risk in our population with the risk in the general population. However, some indirect comparisons with data from the literature can be made. The incidence of a stroke (ischemic stroke or intracerebral hemorrhage) in our cohort with a mean age of 50 years and 67% women was 6.7 per 1000 patient-years. The incidence of a fi rst stroke in theNetherlands is only 1.3 per 1000 patient-years for 50-year old men and 1.0 per 1000 patient-years for 50-year old women.2 The incidence of all vascular events other than recurrent SAH in our cohort was around 11 per 1000 patient-years. In a recently published study on the oc-currence of vascular diseases in a population of 91106 persons in Oxfordshire, the combined incidence of cerebrovascular events (other than SAH), sudden death and myocardial infarc-tion was 5 per 1000 patient-years for men and 4 per 1000 patient-years for women.3 These comparisons strongly suggest that the incidence of vascular events in SAH patients is higher than that in the general population with comparable age and gender. Within a few years, more detailed population-based data will become available. Future studies should investigate a possible higher risk of vascular diseases in the SAH population by comparing the incidence with that in the general population.The long-term effects of SAH on smell, work, relationships, personality and mood are consi-derable and persist for many years. Currently, the focus of follow-up at the outpatient clinic is mostly on the physical condition of SAH patients. The fi ndings of this thesis can help to inform and advise patients about the long-term effects after SAH and about the prognosis of reintegration to work. Personal support, professional counseling and treatment with antide-pressants, where appropriate, are important in the care for patients with a history of SAH.

202

Chap

ter

Part II: Screening for new aneurysmsPart II: Screening for new aneurysms

Main fi ndingsMain fi ndings

Screening in familial subarachnoid hemorrhageScreening in familial subarachnoid hemorrhage

The yield of repeated screening for familial intracranial aneurysms is high; in one out of six relatives with previous aneurysms and in one out of fourteen relatives without previous aneurysms new aneurysms are detected, mostly within fi ve years. Relatives are highly moti-vated to undergo repeated screening; four out of fi ve of relatives who are advised to return for follow-up screening actually return. The result of screening appears to have a signifi cant impact on lifestyle and psychosocial well-being of relatives long after the results are known. Of the screen-positives, almost half had reduced their work and two-thirds had experienced changes in one or more areas of independence, self-esteem, future outlook, or personal rela-tionships. In contrast, only one out of fi fty screen-negatives had stopped working and one out of six had experienced changes in their self-esteem, future outlook, or relationships. Even so, up to half of screen-positive relatives also reported changing their behavior towards healthier living. Screen-positives had lower health related quality of life (HRQoL) compared with both screen-negatives and the reference population, while both screen groups had higher mean depression scores than the reference population. Despite these effects, only three percent of the relatives regretted participating in screening.

Screening in patients with a history of subarachnoid hemorrhageScreening in patients with a history of subarachnoid hemorrhage

Screening for new aneurysms in patients with a history of SAH also has a high yield; in one out of six patients an aneurysm is detected that had not been observed before. This high risk of recurrent aneurysms indicates that development of intracranial aneurysms is a con-tinuous process and not a once in a lifetime event. Risk factors for aneurysm formation and growth are presence of multiple aneurysms at the time of SAH, smoking and hypertension.According to our decision model, screening individuals with previous SAH prevents almost half of the recurrences, slightly increases life-expectancy, but reduces QALY’s and increases costs.Screening is cost-saving without increasing QALY’s in patients with a more than twofold risk above baseline of both aneurysm formation and rupture. It is cost-saving while increasing QALY’s if both risks are at least 4.5 times higher. In patients with reduced quality of life because of fear for a recurrence, screening increases QALY’s at acceptable costs.Only about a quarter of SAH patients with newly detected aneurysms are treated with clipping or coiling. The remaining patients have small aneurysms that are left untreated. The yield of early CT/MR- angiographic follow-up one or two years after detection of such small untreated aneurysms in patients with a history of SAH or with familial SAH is small and does not eliminate the risk of new episodes of SAH. Only three percent of aneurysms enlarge and this enlarge-ment is only moderate (0.5-1.5 millimeters), without consequences for treatment.

203

Fourteen

Implications for patient care and future researchImplications for patient care and future research

Screening in familial subarachnoid hemorrhageScreening in familial subarachnoid hemorrhage

We found that the yield of repeated screening is high and that relatives are highly motivated for repeated screening. Almost none of the screened relatives regretted participation in the screening program. One of the main reasons for this high motivation is probably that the participants have seen the devastating consequences of SAH among their close relatives. The (cost)-effectiveness of screening for aneurysms in patients with familial SAH has not yet been properly investigated. In the nineties, three decision analyses have been performed on the issue of screening for familial aneurysms.4-6 None has specifi cally or properly addressed repeated screening in persons with more than two affected fi rst-degree relatives. In addition, none incorporated coiling as a treatment option. Although, in general, strategies that are not based on evidence should not be advocated, it is hard to deny screening in a relative that re-cently lost two siblings after SAH. The high prevalence of aneurysms in relatives with familial SAH, the relatively young age of occurrence and the possible poorer outcome after familial SAH compared with sporadic cases are strong factors in favor of screening. Even though the effectiveness, optimal interval and age limits have not yet been studied in a decision analysis, screening in relatives with familial SAH with MRA or CTA at a 5-year interval seems justifi ed. The psychosocial impact of screening, however, should be discussed and appropriate support and professional counseling should be offered when necessary. If in the future a decision model will be developed for screening for familial SAH, it should be restricted to relatives with two or more fi rst-degree relatives with intracranial aneurysms, it should include coiling as possible treatment and it should take the emotional aspects of familial screening into account.

Screening patients with a history of subarachnoid hemorrhageScreening patients with a history of subarachnoid hemorrhage

In patients without known extra risks factors for new aneurysm formation or rupture, the risks associated with preventive treatment and the similar outcome for screening and no screening in terms of QALY’s should lead to discourage screening. Screening can save costs and increase QALY’s in patients with a more than twofold risk above baseline of both aneurysm formation and rupture. Unfortunately, it is not yet possible to clearly identify such patients. Firstly, there is inconsistency about the risk facors for new aneurysm formation. In the ASTRA study we found that smoking, hypertension and multiple aneurysms at the time of the initial SAH each at least double the risk of new aneurysm formation. The only other study on risk factors for aneurysm formation, however, confi rmed the increased risk in smokers but did not identify hypertension and multiple aneurysms as risk factors.7 Furthermore, in contrast with our study, an increased risk of aneurysm formation in women was found. Secondly, there are no reliable data for independent predictors of rupture of aneurysms in patients with a history of SAH. It is reasonable to assume that risk factors for aneurysm rupture in patients with a

204

Chap

ter

history of SAH do not greatly differ from risk factors for rupture in general. Unfortunately, in our meta-analysis on risk factors for rupture no multivariable analyses could be performed. Thus, currently there is not enough information on separate determinants for development and rupture of new aneurysms in patients after SAH. Therefore, at this time, patients who will benefi t from screening can not be identifi ed. More studies are needed to confi rm the reported risk factors for new aneurysm formation. Collaborative efforts with pooled analysis of individual data are needed to identify independent risk factors for aneurysm rupture. Also, new opportunities to identify patients with an increased risk of aneurysm formation and rup-ture should be investigated, such as diagnostic tests for genetic risk factors. Genetic factors probably play an important role in the development, growth and rupture of intracranial aneu-rysms. Several candidate genes have been found, most of them involved in the maintenance of the integrity of the extracellular matrix of the aneurysm wall.8 When in the future these tests might become clinically available, screening for new aneurysms can be further targeted to those patients at the highest risk of aneurysm formation and rupture. Besides patient characteristics, the confi guration of the circle of Willis might also be related to formation and rupture of aneurysms. Asymmetry of the A1 segments of the anterior cerebral artery is more often found in patients with aneurysms of the anterior communicating artery than in patients with aneurysms at other locations.9,10 Similar fi ndings have been done for patients with aneurysms of the posterior communicating artery.11 Patients with an incomplete circle of Willis or an abnormal shape or angel of the arterial vessels might be more prone to aneurysm formation and rupture.10, 12 The motions and the thickness of the aneurysm wall might also be related to the risk of rupture. We are more and more able to detect these features by means of high resolution MR or 4D-CTA.13 The prognostic value of these possible anatomical risk factors needs further investigation. Until more data on separate risk factors for aneurysm formation and rupture become available, screening every fi ve years can be considered in patients in a good clinical condition and with extra risk factors.Coiling increasingly is being used as an alternative to neurosurgical clipping. A randomized trial showed that in patients with ruptured intracranial aneurysms suitable for both treat-ments, coiling is more likely to result in independent survival at one year than clipping.14 In the ASTRA study we included only patients with clipped aneurysms, since CTA is not a useful screening tool for coiled patients, given the extensive coil-artefacts. For coiled aneurysms MRI can be used, whereas this technique produces large artefacts around clipped aneurysms. It is likely that the number of de novo aneurysms in coiled patients is similar to the number of de novo’s we found for clipped patients. The long-term occurrence of regrowth aneurysms or remnants caused by impaction of coils, however, might differ from clipping. In addition, after coiling the long-term outcome without screening in terms of recurrences of SAH is largely unknown. These data are necessary before a comparable decision analysis on the effectiveness of screening for new aneurysms in coiled patients can be performed.

205

Fourteen

Decision model or randomized trialDecision model or randomized trial

It is important to realize that decisions based on modeling have limitations largely because of the uncertainty surrounding some of the estimates. A randomized trial would be the prefer-red way of investigating the effects of screening in both familial screening and in screening patients with a history of SAH. However, at least 20-30 years of follow-up are required to col-lect enough episodes of recurrent SAH to compare a no screening and screening strategy. In addition, for familial screening most relatives are so motivated for screening and preventive treatment that they probable will not participate in a randomized trial. Therefore, it is very unlikely that such a trial will ever be performed. And even if such a trial would start today, the results would be available only after decades.

Psychosocial effects of screeningPsychosocial effects of screening

Screening might have effects on HRQoL by reassurance or by feelings of guilt towards affected relatives when no aneurysm is found and by inducing fear and depression when an aneurysm is detected, especially when this aneurysm is not treated.15-17 In a recent study we found no major differences on HRQoL and mood between patients with a new small aneurysm detected at screening that is left untreated and patients with a negative screening result.18

When screening is not performed, this might also affect HRQoL in a negative way. Many patients with a history of SAH are or will become aware of the risk of new aneurysm develop-ment and the subsequent risk of recurrent SAH. This insight, coupled with a state of not being screened might result in a decreased HRQoL. In our decision model only a slight decrease in HRQoL in the non-screened population led to a substantial increase in the benefi t of screening. Unfortunately, little is truly known about the health utility in patients who are aware of an increased health risk but are not offered screening, and about modifying factors. Fear for a recurrence might also decrease by providing adequate information to patients (although the relative risk of a recurrence is high, the absolute risk is low) or by psychosocial counseling. Further research on determinants of fear for recurrence, impact of this fear on HRQoL, the management of fear and the effect of reassurance of a normal screen is needed to assess its effect on aneurysm screening programs.

Follow-up of small aneurysmsFollow-up of small aneurysms

If one decides not to treat an unruptured aneurysm, radiographic follow-up is an option because growth of an aneurysm is an important predictor of rupture. 7, 19 We found that the yield of short-time follow-up for untreated aneurysms is small and does not eliminate the risk of new episodes of SAH. Only few aneurysms enlarged in the fi rst one or two years after detection. A few studies have reported on the growth rate at slightly larger follow-up intervals but their results are confl icting. Frequent enlargement (8% of the aneurysms smaller than

206

Chap

ter

fi ve millimeters over a mean follow-up time of 29 months) was found in one Japanese study21 whereas in two other studies 0% or 3% of the small aneurysms enlarged over a mean follow-up time of 18 and 50 months.21, 22 In a mathematical model we recently showed that the growth of intracranial aneurysms is probably not constant and time independent but rather an irregular and discontinuous process with periods with and without growth (H. Koffi jberg et. al 2006; unpublished data). It is important to verify this fi nding in long-term follow-up studies of in-tracranial aneurysms since it has consequences for the usefulness of radiological follow-up of unruptured aneurysms and for the evaluation of the cost-effectiveness of aneurysm screening programs. Until studies on the effectiveness of long-term follow-up have been performed, follow-up at intervals longer than two years should be strongly considered. However, the pros, the cons (such as radiation associated with CT and psychosocial consequences of follow-up) and the exact frequency of follow-up have to be carefully weighted and discussed with each individual patients before radiographic follow-up is performed.

207

Fourteen

ReferencesReferences

1. Feigin VL, Rinkel GJE, Lawes CM, Algra A, Bennett DA, van Gijn J, Anderson CS. Risk factors for subarachnoid hemorrhage: An updated systematic review of epidemiological studies. Stroke 2005;36:2773-2780.2. Bots ML, Berger-van Sijl M, Jager-Geurts MH, Bos M, Reitsma JB, Breteler MMB, de Bruin A. Incidence of cerebrovascular diseases in The Netherlands in 2000. In Jager-Geurts MH, Peters RJG, van Dis SJ, Bots, ML Cardiovascular diseases in the Netherlands 2006, disease and death rates The Hague: Netherlands Heart Foundation 2006.3. Rothwell PM, Coull AJ, Silver LE, Fairhead JF, Giles MF, Lovelock CE, Redgrave JN, Bull LM, Welch SJ, Cuthbertson FC, Binney LE, Gutnikov SA, Anslow P, Banning AP, Mant D, Mehta Z. Population-based study of event-rate, incidence, case fatality, and mortality for all acute vascular events in all arterial territories (Oxford vascular study). Lancet 2005;366:1773-1783.4. Crawley F, Clifton A, Brown MM. Should we screen for familial intracranial aneurysm? Stroke 1999;30:312-316.5. Dippel DW, ter Berg JW, Habbema JD. Screening for unruptured familial intracranial aneurysms. A decision analysis. Acta Neurol Scand 1992;86:381-389.6. Leblanc R, Worsley KJ, Melanson D, Tampieri D. Angiographic screening and elective surgery of familial cerebral aneurysms: A decision analysis. Neurosurgery 1994;35:9-18.7. Juvela S, Poussa K, Porras M. Factors affecting formation and growth of intracranial aneurysms: A long-term follow-up study. Stroke 2001;32:485-491.8. Ruigrok YM, Rinkel GJE, Wijmenga C. Genetics of intracranial aneurysms. Lancet Neurol 2005;4:179- 189.9. Velthuis BK, van Leeuwen MS, Witkamp TD, Ramos LM, Berkelbach van der Sprenkel JW, Rinkel GJE. Surgical anatomy of the cerebral arteries in patients with subarachnoid hemorrhage: Comparison of computerized tomography angiography and digital subtraction angiography. J Neurosurg 2001;95:206-212.10. Kasuya H, Shimizu T, Nakaya K, Sasahara A, Hori T, Takakura K. Angeles between A1 and A2 segments of the anterior cerebral artery visualized by three-dimensional computed tomographic angiography and association of anterior communicating artery aneurysms. Neurosurgery 1999;45:89-93.11. Horikoshi T, Akiyama I, Yamagata Z, Sugita M, Nukui H. Magnetic resonance angiographic evidence of sex-linked variations in the circle of willis and the occurrence of cerebral aneurysms. J Neurosurg 2002;96:697-703.12. Rossitti S, Lofgren J. Vascular dimensions of the cerebral arteries follow the principle of minimum work. Stroke 1993;24:371-377.13. Ishida F, Ogawa H, Simizu T, Kojima T, Taki W. Visualizing the dynamics of cerebral aneurysms with four-dimensional computed tomographic angiography. Neurosurgery 2005;57:460-471.14. Molyneux A, Kerr R, Stratton I, Sandercock P, Clarke M, Shrimpton J, Holman R. International subarachnoid aneurysm trial (isat) of neurosurgical clipping versus endovascular coiling in 2143 patients with ruptured intracranial aneurysms: A randomised trial. Lancet 2002;360:1267-1274.15. Scaf-Klomp W, Sanderman R, van de Wiel HB, Otter R, van den Heuvel WJ. Distressed or relieved? Psychological side effects of breast cancer screening in the netherlands. J Epidemiol Community Health 1997;51:705-710.16. van der Schaaf IC, Brilstra EH, Rinkel GJ, Bossuyt PM, van Gijn J. Quality of life, anxiety, and depression in patients with an untreated intracranial aneurysm or arteriovenous malformation. Stroke 2002;33:440-443.17. Sobel SK, Cowan DB. Impact of genetic testing for Huntington disease on the family system. Am J Med Genet 2000;90:49-59.

208

Chap

ter

18. van der Schaaf IC, Wermer MJH, Velthuis BK, Buskens E, Bossuyt PM, Rinkel GJE. Psychosocial impact of fi nding small aneurysms that are left untreated in patients previously operated on for ruptured aneurysms. J Neurol Neurosurg Psychiatry 2006;77:748-752.19. Wiebers DO, Whisnant JP, Huston J, 3rd, Meissner I, Brown RD, Jr., Piepgras DG, Forbes GS, Thielen K, Nichols D, O’Fallon WM, Peacock J, Jaeger L, Kassell NF, Kongable-Beckman GL, Torner JC. Unruptured intracranial aneurysms: Natural history, clinical outcome, and risks of surgical and endovascular treatment. Lancet 2003;362:103-110.20. Miyazawa N, Akiyama I, Yamagata Z. Risk factors for growth of unruptured intracranial aneurysms: Follow-up study by serial 0.5-t magnetic resonance angiography. Neurosurgery 2006;58:1047-1053.21. Phan TG, Huston J, 3rd, Brown RD, Jr., Wiebers DO, Piepgras DG. Intracranial saccular aneurysm enlargement determined using serial magnetic resonance angiography. J Neurosurg 2002;97:1023- 1028.22. Matsubara S, Hadeishi H, Suzuki A, Yasui N, Nishimura H. Incidence and risk factors for the growth of unruptured cerebral aneurysms: Observation using serial Computerized Tomography Angiography. J Neurosurg 2004;101:908-914.

209

Fourteen

Addendum ASTRA study groupAddendum ASTRA study group

Executive committeeA. Algra MD1,2; E. Buskens MD2; P. Greebe RN1; H. Koffi jberg MSc2, G.J.E. Rinkel (principal inves-tigator) MD1; I.C. van der Schaaf MD3; B.K. Velthuis MD3; M.J.H. Wermer MD1.

Steering committeeK.W. Albrecht MD7, A. Algra MD1,2; P.M.M. Bossuyt PhD4, E. Buskens MD2; J. van Gijn MD, FRCP, FRCPE1; P. Greebe RN1; A. Gorissen RN6; G.J. den Heeten, MD5; H. Koffi jberg MSc2; C.B. Majoie MD5; G.J.E. Rinkel (chair) MD1; I.C. van der Schaaf MD3; B.K. Velthuis MD3; M. Vermeulen MD6; M.J.H. Wermer MD1.

Writing committeeA. Algra MD1,2; P.M.M. Bossuyt PhD4, E. Buskens MD2; J. van Gijn MD, FRCP, FRCPE1; P. Greebe RN1; H. Koffi jberg MSc2; C.B. Majoie MD5; G.J.E. Rinkel MD1; I.C. van der Schaaf MD3; B.K. Velthuis MD3; M.J.H. Wermer MD1.

Statistical analysisA. Algra MD1,2; E. Buskens MD2; H. Koffi jberg MSc2; I.C. van der Schaaf MD3; M.J.H. Wermer MD1.

Patient accrual and outpatient careK.W. Albrecht MD7; P. Greebe RN1; A. Gorissen RN6; M.J.H. Wermer MD1

Radiological assessmentsN Freling MD5; G. de Kort MD3; C.B. Majoie MD5; B.K. Velthuis MD3, T. Witkamp MD3

Radiological angiographic evaluation and interventionsG. de Kort MD3; TH Lo MD3; C.B. Majoie MD5; WJJ van Rooij MD5

Neurosurgical evaluation and interventionsK.W. Albrecht MD7, C.A.F. Tulleken MD8, A. van der Zwan MD8

From the 1) department of neurology, Rudolf Magnus Institute of Neuroscience, University Medical Center Utrecht 2) Julius Center for Health Sciences and Primary Care Utrecht 3) department of Radiology, Univer-sity Medical Center Utrecht 4) department of clinical epidemiology and biostatistics, Academic Medical Cent-er Amsterdam 5) department of radiology, Academic edical Center, Amsterdam 6) department of neurology, Academic Medical Center, Amsterdam 7) department of neurosurgery, Academic Medical Center, Amsterdam 8) department of neurosurgery, Rudolf Magnus Institute of Neuroscience, University Medical Center Utrechtall in the Netherlands.

ChapterFourteen

General discussionSamenvatting

SummaryDankwoord

List of publicationsCurriculum Vitae

212

Chap

ter

Het eerste deel van dit proefschrift beschrijft de gevolgen op de lange termijn bij een sub-arachnoïdale bloeding mensen die goed hersteld zijn na het doormaken van een dergelijke bloeding (SAB) en bij wie het gebarsten aneurysma behandeld is door middel van chirurgisch afklemmen van het aneurysma (‘clippen’). Het tweede deel van dit proefschrift beschrijft de effectiviteit en de psychosociale gevolgen van screening op nieuwe aneurysmata bij mensen die in het verleden een SAB hebben doorgemaakt en bij mensen met een familiaire vorm van SAB.

In Hoofdstuk 1 wordt een korte inleiding gegeven over SABs. Een SAB is een bloeding tussen de hersenen en de schedel. Een dergelijke bloeding onstaat meestal uit een geruptureerde vaatafwijking (aneurysma). Het is een bloeding met een slechte prognose die vaak bij relatief jonge mensen voorkomt. Het aneurysma kan behandeld worden door de neurochirurg door middel van clips of door de radioloog door middel van coils. Lang is gedacht dat mensen die hersteld waren na een SAB en bij wie het gebarsten aneurysma succesvol behandeld was eenzelfde levensverwachting hadden als mensen die nooit een SAB hadden doorgemaakt. Er zijn echter steeds meer aanwijzingen dat bij deze mensen nieuwe aneurysmata kunnen ontstaan en dat de kans op een nieuwe SAB of een andere vasculaire aandoening verhoogd is. Tien procent van de mensen met een SAB heeft een familiaire vorm waarbij minstens twee eerstegraads familieleden bekend zijn met aneurysmata van de hersenvaten. Eerste-graads familieleden hebben een verhoogde kans op het krijgen van een SAB. Tegenwoordig kan screening op nieuwe aneurysmata relatief eenvoudig worden uitgevoerd door middel van CT-angiografi e (CTA) en MR-angiografi e (MRA). De opbrengst van screening op nieuweaneurysmata bij mensen met een voorgeschiedenis van een SAB en de opbrengst van her-haaldelijk screenen van mensen met een familiaire vorm van SAB is onbekend.

Deel 1. De gevolgen op lange termijn van een subarachnoïdale bloedingDeel 1. De gevolgen op lange termijn van een subarachnoïdale bloeding

Hoofdstuk 2 beschrijft de karakteristieken van mensen die jaren na hun eerste SAB een nieuwe SAB hebben doorgemaakt en de afl oop na een dergelijk recidief bloeding. Van de 30 patiënten met een nieuwe bloeding waren er relatief veel vrouwen (90%) en mensen met een familiaire vorm van SAB (30%). De afl oop na een nieuwe bloeding is hetzelfde als na een eerste bloeding: van alle patiënten die niet waren overleden voordat ze werden opgenomen in het ziekenhuis, overleed éénderde van de patiënten alsnog, was één op de zes niet ADL zelfstandig (zodanig gehandicapt dat ze afhankelijk waren van hulp bij het uitvoeren van dagelijkse bezigheden zoals wassen en aankleden) en herstelde de helft tot een zelfstandig bestaan.In Hoofdstuk 3 wordt de incidentie van een nieuwe SAB onderzocht bij 752 patiënten die tus-sen 1985 en 2001 een eerste SAB hadden doorgemaakt en bij wie het aneurysma succesvol behandeld was door middel van clippen. De eerste 10 jaar na de eerste SAB was de cumulatieve

213

Fourteen

incidentie van een recidief SAB 3.2%. Dit is ongeveer 22x hoger dan de kans op een SAB in de algemene bevolking met een vergelijkbare leeftijd en geslachtsverdeling. Risicofactoren voor een nieuwe SAB zijn roken, jonge leeftijd ten tijde van de eerste SAB en meerdere aneurysmata ten tijde van de eerste SAB.In Hoofdstuk 4 wordt de kans op overlijden en de kans op een beroerte of hartinfarct op de lange termijn onderzocht bij dezelfde groep van 752 SAB patiënten. De kans op sterfte in de SAB populatie was 3x hoger dan de kans op sterfte in de algemene bevolking. De cumulatieve incidentie van een beroerte of hartinfarct in de eerste 10 jaar na de SAB is 10.8%. Deze inci-dentie is lager dan de kans op een beroerte of hartinfarct bij mensen die een TIA of een klein herseninfarct hebben doorgemaakt maar hoger dan het in de literatuur beschreven risico in de algemene bevolking.In Hoofdstuk 5 bestudeerden we de incidentie, de invloed op de kwaliteit van leven en de prognose van reukverlies (anosmie) na een SAB. Anosmie kwam voor bij één op de drie geclipte patiënten en bij één op de zes gecoilde patiënten. De invloed op de kwaliteit van leven werd door de patiënten aangegeven als aanzienlijk. Slechts bij één op de vier patiënten trad er na verloop van tijd verbetering op van de klachten. Risicofactoren voor het ontstaan van anosmie na een SAB waren behandeling door middel van clippen en een gebarsten aneurysma ter plaatste van de arteria cerebri anterior, communicans anterior of de arteria pericallosa.Hoofdstuk 6 beschrijft de lange termijn effecten van een SAB op het gebied van werk, relaties, karakter en stemming. Van de 610 geinterviewde patiënten die na een SAB waren hersteld tot een zelfstandig functionerend bestaan was de helft gestopt met werken of was minder uren gaan werken. Veertien procent van de mensen was gescheiden van wie zeven procent aangaf dat dit kwam vanwege de gevolgen van de SAB. Tweederde van de mensen rapporteerden veranderingen in karakter. SAB patiënten hadden een hogere gemiddelde depressie score op de Hospital Anxiety and Depression Scale dan de algemene bevolking. Slechts een kwart van de patiënten was hersteld zonder psychosociale of neurologische problemen.

Deel 2. Screening op nieuwe aneurysmataDeel 2. Screening op nieuwe aneurysmata

Screening van mensen met een familaire vorm van subarachnoïdale bloedingScreening van mensen met een familaire vorm van subarachnoïdale bloeding

In Hoofdstuk 7 onderzochten we de opbrengst van herhaaldelijke screening bij mensen met twee of meer eerstegraads familieleden met een SAB. De motivatie voor herhaaldelijk screenen bleek hoog; vier op de vijf familieleden kwam vijf jaar na de eerste screening terug voor een nieuw screeningsonderzoek. Nieuwe aneurysmata werden gevonden bij 9% van de 102 mensen die terugkwamen voor een nieuwe screening; bij 16% van de mensen met een familiare SAB die zelf ooit een SAB hadden gehad en bij 7% van de mensen die zelf nooit een SAB hadden gehad.

214

Chap

ter

Hoofdstuk 8 beschrijft de psychosociale gevolgen van screening bij 105 mensen met een familiaire vorm van SAB. De gescreende familieleden werden ingedeeld in mensen bij wie een aneurysma werd gevonden bij familiaire screening (positieve screening) en in mensen bij wie geen aneurysma werd gevonden (negatieve screening). Van de mensen met een positieve screening was 44% minder gaan werken en had 66% veranderingen gemerkt op één of meer-dere psychosociale gebieden zoals zelfstandigheid, relaties, zelfvertrouwen en vertrouwen in de toekomst. Van de mensen met een negatieve screening was slechts 2% minder gaan werken en had 17% verandering gemerkt op dezelfde psychosociale gebieden. De mensen met een positieve screening hadden een lagere kwaliteit van leven dan mensen met een negatieve screening en de algemene bevolking. Zowel de mensen met een positieve als de mensen met een negatieve screening hadden een hogere depressie score dan de algemene bevolking. Ondanks deze psychosociale gevolgen hadden slechts drie van de 105 personen spijt van deelname aan de screening.

Screening bij mensen met een subarachnoïdale bloeding in de voorgeschiedenisScreening bij mensen met een subarachnoïdale bloeding in de voorgeschiedenis

In Hoofdstuk 9 onderzochten we met behulp van een beslismodel waarbij gebruik werd gemaakt van gegevens uit de literatuur of screening op nieuwe aneurysmata effectief was bij mensen die een SAB hebben doorgemaakt. Zowel bij geclipte als bij gecoilde patiënten bleek screening een groot aantal recidief SABs te voorkomen. Het aantal jaren in goede kwa-liteit van leven (QALY’s) was echter vrijwel gelijk in de gescreende en niet gescreende groep, met name vanwege de toename van mensen die gehandicapt waren geraakt als gevolg van screening en preventieve behandeling. Er bleken weinig literatuur gegevens beschikbaar te zijn over de belangrijkste factoren in het beslismodel zoals de incidentie en ruptuurkans van nieuwe aneurysmata.In Hoofdstuk 10 beschrijven we de resultaten van de ASTRA studie. In de ASTRA studie on-derzochten we 610 mensen die in het verleden een SAB hadden doorgemaakt op nieuwe aneurysmata met behulp van CTA. In totaal werden 129 aneurysmata gevonden bij 96 (16%) van de 610 patiënten na een gemiddelde follow-up tijd van 8.9 jaar. Van de 129 aneurysmata bevonden zich er 24 (19%) ter plaatste van de clip en 105 (81%) op een andere locatie. Eénderde van de aneurysmata was nieuw, tweederde kon retrospectief geïndentifi ceerd worden op de beeldvorming ten tijde van de eerste SAB. Risicofactoren voor nieuwvorming of groei van aneurysmata waren roken, hypertensie en meerdere aneurysmata ten tijde van de SAB.In Hoofdstuk 11 werden de resultaten van de ASTRA studie en gegevens uit de literatuur gebruikt als input voor een beslismodel om de kosten en de effectiviteit van screening op nieuwe aneurysmata te onderzoeken bij mensen die geclipt zijn voor een geruptureerdaneurysma. In dit beslismodel voorkwam screening de helft van het aantal recidief bloedingen, gaf het een geringe verhoging van de levensverwachting, maar veroorzaakte het een daling in het aantal QALY’s en een stijging van de kosten. Bij mensen die echter een minimaal twee

215

Fourteen

keer verhoogd risico hadden op zowel ruptuur en vorming van aneurysmata dan baseline was screening kostenbesparend. Indien het risico op vorming en ruptuur 4.5 keer of meer verhoogd was resulteerde screening ook in een toename in QALY’s. Ook verhoogde screening het aantal QALY’s tegen acceptable kosten bij mensen die erg bang zijn voor een recidief SAB.Hoofdstuk 12 beschrijft het nut van follow-up na één of twee jaar van kleine onbehandelde aneurysmata gevonden bij screening door middel van CTA of MRA. Voor dit onderzoek werden 125 aneurysmata gevolgd bij 93 mensen met familiaire aneurysmata of een voorgeschiedenis van SAB. Groei van een aneurysma trad op na een mediane follow-up van 1.3 jaar bij drie (3.2%) van deze 93 mensen. De groei was echter gering en had geen behandelingsconsequenties. Bovendien voorkwam de follow-up niet dat twee mensen een SAB kregen; één uit een nieuw dissectie aneurysma en één uit een klein aneurysma ter plaatste van de clip van de vorige SAB.In Hoofdstuk 13 voerden we een meta-analyse uit naar risicofactoren voor ruptuur vananeurysmata. Negentien studies met 4705 patiënten en 6556 aneurysmata werden geïn-cludeerd. Risicofactoren voor ruptuur waren leeftijd jonger dan 60 jaar, vrouwelijk geslacht, Japanse of Finse afkomst, afmeting van het aneurysma groter dan vijf millimeter, een aneu-rysma van de posterieure circulatie en een symptomatische aneurysma. Multivariate analyse met behulp van meta-regressie bleek niet mogelijk te zijn.De resultaten van dit proefschrift en hun implicaties voor de patiëntenzorg en verder weten-schappelijk onderzoek worden besproken in Hoofdstuk 14.

ChapterFourteen

General discussionSamenvatting

SummaryDankwoord

List of publicationsCurriculum Vitae

218

Chap

ter

The fi rst part of this thesis addresses the long-term follow-up of patients with a history of subarachnoid hemorrhage (SAH) in whom the ruptured aneurysm was treated by means of clipping. The second part describes the effectiveness and consequences of screening for new intracranial aneurysms in patients with a history of SAH and in relatives with familial SAH.

Chapter 1 gives a short introduction on subarachnoid hemorrhage (SAH). An SAH is most often caused by rupture of an intracranial aneurysm. The prognosis of an SAH is poor and it affects relatively young persons. Intracranial aneurysms can be treated by neurosurgical clipping or endovascular coiling. Patients who recover after SAH and are successfully treated for the rup-tured aneurysm generally are believed to have a prognosis similar to that of persons without a history of SAH. More recently it was shown that new aneurysms can develop in patients with a history of SAH. Patients with a history of SAH might be at risk for recurrent SAH or for vascular diseases other than SAH. Familial clustering (defi ned as two or more relatives with intracranial aneurysms) occurs in around 10% of the patients with SAH. First-degree relatives have a higher risk of SAH than the general population. Nowadays screening for intracranial aneurysms can be performed relatively easily with screening techniques that are not or only minimally invasive such as CT-angiography (CTA) or MR-angiography (MRA). However, the yield of screening in patients with a history of SAH and the yield of repeated screening in relatives with familial SAH is unknown.

Part 1. Long-term follow-up after subarachnoid hemorrhagePart 1. Long-term follow-up after subarachnoid hemorrhage

Chapter 2 describes the patient characteristics and the outcome of patients with a late recur-rent SAH. In the 30 included patients admitted with a recurrent SAH there was a predominance of women (90%) and patients with familial SAH (30%). The outcome of a recurrence was similar to the outcome after SAH in general: one-third of the admitted patients died during the clinical course, one out of six remained disabled and were independent for activities of daily life (ADL) and half of the patients had a good outcome.In Chapter 3 the incidence of a late recurrent SAH is investigated in 752 patients admitted between 1985 and 2001 for SAH. In these patients the ruptured aneurysm and all the other detected aneurysms at the time of the fi rst SAH had been successfully treated by means of clipping. In the fi rst 10 years after the initial SAH the cumulative incidence of a recurrence was 3.2%. This incidence is 22 times higher than expected in populations with comparable age and sex. Risk factors for a recurrence were smoking, young age and multiple aneurysms at time of the initial SAH.In Chapter 4 we investigated the overall mortality and long-term risk of vascular events other than recurrent SAH in the same group of 752 patients. The standardized mortality ratio was 3. The cumulative incidence of a vascular event in the fi rst 10 years after SAH was 10.8%. The risk of vascular events after SAH is lower than that after a minor stroke, but higher than the

219

Fourteen

population risks reported in the literature.In Chapter 5 we investigated the incidence, impact and prognosis of loss of smell (anosmia) after a SAH. Anosmia was reported by one of every three clipped patients and by one of every six coiled patients. The impact of anosmia was considerable. In only a quarter of patients the symptoms had improved over time. Risk factors for anosmia were treatment by means of clipping and SAH caused by an aneurysm located at the anterior cerebral, anterior com-municating or pericallosal artery.Chapter 6 describes the long-term effects of SAH on employment, relationships, personality and mood. Of the 610 interviewed patients who resumed independent living after the SAH half stopped working or worked shorter hours. Fourteen percent of patients got divorced, of whom 7% reported that this was a result of the consequences of the SAH. Two-thirds of patients reported changes in personality. SAH patients had a statistically signifi cant higher mean depression score than the reference population. Only a quarter of all patients reported a complete recovery without psychosocial or neurological problems.

Part 2. Screening for new aneurysmsPart 2. Screening for new aneurysms

Screening in familial subarachnoid hemorrhageScreening in familial subarachnoid hemorrhage

In Chapter 7 we investigated the yield of repeated screening in families with two or more fi rst-degree relatives with SAH. The motivation for repeated screening was high; four out of every fi ve patients returned for screening fi ve years after the initial screening. New aneurysms were detected in 9% of the 102 relatives who returned for screening; 16% of the relatives with previous aneurysms and 7% of the relatives without previous aneurysms.Chapter 8 describes the psychosocial impact of screening for aneurysms in 105 relatives with familial SAH. Of the screen-positives 44% had reduced their work and 66% had experienced changes in one or more areas of indepence, self-esteem, future outlook, or personal relation-ships. In contrast, only 2% of the screen-negatives reduced their work and 17% had experienced changes on the above mentioned psychosocial areas. Screen-positives had lower quality of life compared with both screen-negatives and the reference population, while both screen groups had higher mean depression scores than the reference population. Despite these effects only three of the 105 relatives regretted participation in the screening program.

Screening in patients with a history of subarachnoid hemorrhageScreening in patients with a history of subarachnoid hemorrhage

In Chapter 9 we studied, by means of a decision model using data from the literature, whether screening for new aneurysms was effective in patients with a history of SAH. In both clipped and coiled patients screening appeared to prevent many recurrent SAH’s. However, the number of quality adjusted life-years (QALY’s) was comparable in the screened and the not-screened population because of the complications from screening and preventive treatment. Another

220

Chap

ter

important fi nding was that the literature on the key estimates of the model such as the incidence and risk of rupture of new aneurysms was very sparse.In Chapter 10 we describe the results of the ASTRA study. In the ASTRA study we screened 610 patients, who had made a functional good recovery after SAH, for new aneurysms by means of CTA. In total 129 aneurysms were detected in 96 (16%) of the 610 patients after a mean fol-low-up of 8.9 years. Of these, 24 (19%) were located at the site of the previously ruptured and clipped aneurysm and 105 (81%) at a site remote from the clip site. One-third of the aneurysms were truly de novo, two-third were already visible in retrospect on the CTA or angiography at the time of the initial SAH. Risk factors for aneurysms formation and enlargement were smoking, hypertension and the presence of multiple aneurysms at the time of the SAH.In Chapter 11 the results of the ASTRA study combined with data from the literature were used as input for a new decision model to study the costs and effectiveness of screening for new aneurysms in patients with clipped aneurysms after SAH. Screening prevented almost half of the recurrences, slightly increased life-expectancy, but reduced QALY’s and increased costs. However, screening was cost-saving without increasing QALY’s in patients with a more than twofold risk above baseline of both aneurysms formation and rupture. It was cost saving while increasing QALY’s if both risks were at least 4.5 times higher. In addition, screening increased QALY’s at acceptable costs in patients with reduced quality of life because of fear for a recurrence.Chapter 12 describes the yield of short-term follow-up CTA/MRA for small aneurysms de-tected at screening. We included for this study 93 persons with a history of SAH or with a familial SAH who had in total 125 aneurysms. After a median follow-up time of 1.3 years ananeurysm enlarged in three of the 93 persons. The growth was, however, only moderate wit-hout consequences of treatment. Despite radiological follow-up, two patients had an SAH; one from an aneurysm at the clip-site that ruptured without enlargement and one from a newly developed dissecting aneurysm.In Chapter 13 we performed a meta-analysis on the risk of rupture of intracranial aneurysms. Nineteen studies with 4705 patients and 6556 aneurysms were included. Risk factors forrupture were age younger than 60 years, female gender, Japanese or Finnish descent, size longer than fi ve millimeters, location of the aneurysm at the posterior circulation and a symptomatic aneurysm. Multivariate meta-regression analysis was not feasible.The results of all studies and their implications for patient care and future research are dis-cussed in Chapter 14.

221

Fourteen

ChapterFourteen

General discussionSamenvatting

SummaryDankwoord

List of publicationsCurriculum Vitae

224

Chap

ter

Graag wil ik een aantal mensen bedanken die door hun steun op wetenschappelijk en/of sociaal gebied hebben bijgedragen aan het tot stand komen van dit proefschrift.

Alle SAB patiënten wil ik bedanken voor hun deelname aan de diverse studies.

Prof. Rinkel, beste Gabriel, het is al weer tien jaar geleden dat ik als student nerveus bij je binnen kwam lopen om onderzoek te doen. Ik heb er geen moment spijt van gehad. Ik heb veel bewondering voor de directe manier waarop je al je promovendi begeleidt. Er is altijd tijd om problemen te bespreken, manuscripten komen razendsnel retour, je kunt dingenbijzonder helder op een rijtje zetten en je houdt de vaart er goed in. Toen ik begin dit jaar met een overzicht van 13 hoofdstukken voor het proefschrift bij je kwam was ik stiekum toch best een beetje tevreden met mezelf. Helaas duurde dit gevoel niet lang want je merkte fi jntjes op dat ik toch echt nog een hoofdstuk vergeten was. Ziehier het resultaat. Ontzettend bedankt voor alles!

Dr. van der Schaaf, beste Irene, wat een geluk zo’n slimme, enthousiaste, ambitieuze enbovenal leuke onderzoekspartner te hebben! Het is erg prettig samen te werken met iemand die dezelfde ideeën over onderzoek heeft. Zelfs in de sportschool konden we nog even de laatste plannen doornemen. Ik hoop dat we in de toekomst nog veel samen zullen doen.

Paut Greebe, Anick Gorissen en Dorien Slabbers, bedankt voor jullie geweldige hulp. Dankzij jullie is de inclusie van het ASTRA onderzoek bijzonder soepel verlopen. Paut, je bent je ge-wicht in goud waard. Bedankt voor je enorme inzet voor mijn onderzoek en de gesprekken over alle andere dingen in het leven. Anick, de poli’s in het AMC waren niet alleen vlekkeloos georganiseerd maar ook erg gezellig. Je oprechte belangstelling voor mensen en je empathie vind ik erg bijzonder. Ik was ervan onder de indruk dat jij je vrijwel alle 200 SAB patiënten uit de jaren ‘85-’97 nog precies wist te herinneren....en zij jou! Dorien, de ASTRA database zag er tiptop uit. Bedankt voor al je inspanningen voor diverse onderdelen van mijn proefschrift.

Prof. Algra, beste Ale, als beginnend epidemioloog kon ik altijd bij je terecht voor de nodige feedback. Als jij naar een manuscript gekeken had kon ik het met een gerust hart submitten wetende dat alle puntjes op de i stonden en alle tabellen uitgebreid met de hand nagerekend waren. Door het volgen van de NIHES opleiding klinische epidemiologie hoopte ik toch stiekum een beetje “een Ale“ te worden maar helaas kom ik nog niet echt bij je in de buurt. Bedankt ook voor je bijdrage op de dansvloer ten tijde van de diverse congressen.

Dr. Buskens, beste Erik, een Markov model als begin van een promotie onderzoek beschouw ik achteraf toch wel als een pittige ontgroening. Je liet je niet beïnvloeden door mijn vele depres-sieve buien maar bleef stoïcijns geloven in een goede afl oop. Gelukkig kreeg je uiteindelijk gelijk. Erg jammer dat je naar Groningen vertrekt maar dit betekent vast niet het einde van onze samenwerking.

225

Fourteen

Drs. Koffi jberg, beste Erik. Ik weet nog goed dat we in september 2004 vol enthousiasme be-gonnen aan Markov model nr 2. Dit moest toch echt “een eitje” zijn. Het eerste Markov model moest slechts aangepast worden dus voor de kerstdagen zouden we zeker klaar zijn. Helaas waren we na 1.5 jaar gezweet en geploeter toch wat minder vrolijk gestemd. Zeker toen het software programma onze uit de kluiten gewassen beslisboom niet meer trok. Als een echte wizzkid programmeerde je binnen no time je eigen programma waardoor we de analyses toch nog konden doen. Door je humor bleef mijn Markov depressie deze keer uit. Succes bij het afronden van je eigen proefschrift. Het lijkt me leuk weer samen te werken aan het familiaire beslismodel (dit moet voor ons experts nu toch echt een makkie zijn!).

Dr. Velthuis, beste Birgitta, bedankt voor al je hulp bij het radiologische deel van het ASTRA onderzoek en je nuttige en heldere commentaar op vorm en inhoud van de diverse manus-cripten.

Dr. Lo, de Kort en Witkamp, beste Rob, Gerard en Theo, ontzettend bedankt voor de bergen ASTRA CTA’s die jullie in korte tijd beoordeeld hebben. CT laboranten bedankt voor jullie inzet bij het maken van alle CTA’s.

Het ASTRA onderzoek werd uitgevoerd in nauwe samenwerking met het AMC. Graag wil ik hierbij een aantal AMC-ers bedanken. Dr. Albrecht, beste Kees, erg bedankt voor je inzet voor de ASTRA studie in het algemeen en je bereidheid ASTRA patiënten met een nieuw aneurysma op je poli te zien. Dr. Majoie, beste Charles bedankt voor je hulp bij de radiologische studies, je waardevolle inbreng bij de ASTRA vergaderingen en je commentaar op de manuscripten. Dr. Frehling, beste Nicole, bedankt voor de beoordeling van alle CTA’s. Prof. Bossuyt, beste Patrick, bedankt voor het scherpe en zeer nuttige commentaar op de diverse manuscripten. Ik heb er veel van geleerd.

Prof. van Gijn en prof. Wokke wil ik bedanken voor het stimulerende werkklimaat van de vakgroep. Ik ga na zes jaar opleiding nog steeds met plezier naar mijn werk. Prof. van Gijn, tevens bedankt voor uw hulp bij hoofdstuk 7 van dit proefschrift.

Prof. Bär, het is al weer enige tijd geleden maar toch wil ik u bedanken voor het begeleiden van mijn eerste stappen in de wetenschappelijke wereld en de stimulans om naast geneeskunde ook farmacie af te maken.

Mede-onderzoekers Kuan Kho, Janneke van Beijnum, Peter Woerdeman en Peter Willems, bedankt voor de lunches, gezellige gesprekken en verhitte discussies. Jammer dat onzegezamelijke tijd in kamer 03.121 voorbij is.

226

Chap

ter

Studenten Maarten Donswijk, Pieter van Nunen en oud-studenten inmiddels collega’s Hieke Kool en Patricia Halkes wil ik bedanken voor hun hulp bij de diverse hoofdstukken van mijn proefschrift.

Mijn neurologie collega’s wil ik bedanken voor de prettige sfeer en de gezellige borrels. Jikke-Mien, het is al weer zes jaar geleden dat we onze eerste stappen hebben gezet bij de neurologie. De tijd is omgevlogen. Ik ben blij dat je met wat extra onderzoekstijd en zwan-gerschapsverlof je verblijf nog wat gaat rekken. Walter (Appelbolletje voor intimi), bedankt voor je humor waarmee je me altijd weer aan het lachen krijgt. Ynte, bedankt voor je vriend-schap. Hopelijk spreken we elkaar binnenkort weer wat vaker als alle proefschrift- en verhuis ellende voorbij is.

Lieve vriendinnen Lien, Kat, Marieke, Andrea, Dees, Fem, Kirsten en Suzanne, bedankt voor alle gezelligheid de afgelopen jaren.

Het is fi jn te weten dat ik de angstige minuten voor mijn promotie zal worden bijgestaan door de twee meest nuchtere personen uit mijn omgeving. Mirella, je eerlijkheid en humor waardeer ik erg. Nicolien, vanaf dag 1 van geneeskunde mijn beste vriendin. Altijd met alles gelijk op, zelfs met bevallen. Hopelijk keer je snel weer terug uit het verre Alkmaar naar het warme UMC nest.

Lieve pap, mam en Sas. Bedankt voor een fi jne jeugd en de steun die jullie me altijd geven. Pap en mam, hoewel ik niet zo vaak meer thuis kom is het altijd heerlijk om weer onder jullie vleugels te zijn. Geweldig dat jullie een dag per week helemaal uit Twente komen om op ons honingbeertje te passen.

Lieve Igor en Juliette, zonnetjes in mijn leven. Bedankt voor al jullie vrolijkheid en liefde.Ik houd intens veel van jullie. Le Boulac, here we come!

227

Fourteen

ChapterFourteen

General discussionSamenvatting

SummaryDankwoord

List of publicationsCurriculum Vitae

230

Chap

ter

This thesisThis thesis

1. Wermer MJH, Rinkel GJE, Greebe P, Albrecht KW, Dirven CM and Tulleken CA. Late recurrence of subarachnoid hemorrhage after treatment for ruptured aneurysms: Patient characteristics and outcome. Neurosurgery 2005;56:197-204.

2. Wermer MJH, Greebe P, Algra A and Rinkel GJE. Incidence of recurrent subarachnoid hemorrhage after clipping for ruptured intracranial aneurysms. Stroke 2005;36:2394- 2399.

3. Wermer MJH, Greebe P, Algra A and Rinkel GJE. Long-term mortality and vascular event risk after aneurysmal subarachnoid hemorrhage. In preparation.

4. Wermer MJH, Donswijk M, Verwey B and Rinkel GJE. Anosmia after subarachnoid hemorrhage. Submitted.

5. Wermer MJH, Kool H, Albrecht KW, Rinkel GJE. Long-term effects of subarachnoid hemorrhage on employment, relationships, personality and mood. Neurosurgery, in revision.

6. Wermer MJH, Rinkel GJE and van Gijn J. Repeated screening for intracranial aneurysms in familial subarachnoid hemorrhage. Stroke 2003;34:2788-2791.

7. Wermer MJH, van der Schaaf IC, Van Nunen P, Bossuyt PM, Anderson CS and Rinkel GJE. Psychosocial impact of screening for intracranial aneurysms in relatives with familial subarachnoid hemorrhage. Stroke 2005;36:836-840.

8. Wermer MJH, Buskens E, van der Schaaf IC, Bossuyt PM and Rinkel GJE. Yield of screening for new aneurysms after treatment for subarachnoid hemorrhage. Neurology 2004;62:369- 375.

9. Wermer MJH, van der Schaaf IC, Velthuis BK, Algra A, Buskens E and Rinkel GJE. Follow-up screening after subarachnoid haemorrhage: Frequency and determinants of new aneurysms and enlargement of existing aneurysms. Brain 2005;128:2421-2429.

10. Wermer MJH, Koffi jberg H and van der Schaaf IC for the ASTRA study group. Effectiveness and costs of long-term follow-up screening for new aneurysms in patients with clipped aneurysms after subarachnoid haemorrhage: a cohort study and decision model. Submitted.

11. Wermer MJH, van der Schaaf IC, Velthuis BK, Majoie CB, Albrecht KW and Rinkel GJE. Yield of short-term follow-up CT/MR angiography for small aneurysms detected at screening. Stroke 2006;37:414-418.

231

Fourteen

12. Wermer MJH, Algra A and Rinkel GJE. Risk of rupture of unruptured intracranial aneurysms in relation to patient and aneurysm characteristics: an updated meta-analysis. Stroke, in revision.

Other publicationsOther publications

1. Klivenyi P, St Clair D, Wermer MJH, Yen HC, Oberley T, Yang L, Flint Beal M. Manganese superoxide dismutase overexpression attenuates MPTP toxicity. Neurobiol Dis 1998;5:253- 258.

2. Klivenyi P, Beal MF, Ferrante RJ, Andreassen OA, Wermer MJH, Chin MR, Bonventre JV. Mice defi cient in group IV cytosolic phospholipase A2 are resistant to MPTP neurotoxicity. J Neurochem 1998;71:2634-2637.

3. Klivenyi P, Matthews RT, Wermer MJH, Yang L, MacGarvey U, Becker DA, Natero R, Beal MF. Azulenyl nitrone spin traps protect against MPTP neurotoxicity. Exp Neurol 1998;152:163- 166.

4. Matthews RT, Klivenyi P, Mueller G, Yang L, Wermer MJH, Thomas CE, Beal MF. Novel free radical spin traps protect against malonate and MPTP neurotoxicity. Exp Neurol 1999;157:120-126.

5. Klivenyi P, Ferrante RJ, Matthews RT, Bogdanov MB, Klein AM, Andreassen OA, Mueller G, Wermer MJH, Kaddurah-Daouk R, Beal MF. Neuroprotective effects of creatine in a transgenic animal model of amyotrophic lateral sclerosis. Nat Med 1999;5:347-350.

6. Reijneveld JC, Wermer MJH, Boonman Z, van Gijn J, Rinkel GJE. Acute confusional state as presenting feature in aneurysmal subarachnoid hemorrhage: Frequency and characteristics. J Neurol 2000;247:112-116.

7. Wermer MJH, Rinkel GJE, Van Rooij WJ, Witkamp TD, Ziedses Des Plantes BG, Algra A. Interobserver agreement in the assessment of lobar versus deep location of intracerebral haematomas on CT. J Neuroradiol 2002;29:271-274.

8. Van der Schaaf IC, Velthuis BK, Wermer MJH, Majoie C, Witkamp T, de Kort G, Freling NJ, Rinkel GJE. New detected aneurysms on follow-up screening in patients with previously clipped intracranial aneurysms: Comparison with DSA or CTA at the time of SAH. Stroke 2005;36:1753-1758.

232

Chap

ter

9. Van der Schaaf I, Algra A, Wermer MJH, Molyneux A, Clarke M, van Gijn J, Rinkel GJE. Endovascular coiling versus neurosurgical clipping for patients with aneurysmal subarachnoid haemorrhage. Cochrane Database Syst Rev. 2005:CD003085.

10. Wermer MJH, de Gans J, van Dijk GW. Reversible paralysis of the legs after acute bacterial meningitis. J Neurol 2006;253:114-115.

11. Van der Schaaf IC, Wermer MJH, van der Graaf Y, Velthuis BK, van de Kraats CI, Rinkel GJE. Prognostic value of cerebral perfusion Computed Tomography in the acute stage after subarachnoid hemorrhage for the development of delayed cerebral ischemia. Stroke 2006;37:409-413.

12. Van der Schaaf IC, Algra A, Wermer MJH, Molyneux A, Clarke M, van Gijn J, Rinkel GJE. Endovascular coiling versus neurosurgical clipping for patients with aneurysmal subarachnoid haemorrhage. Stroke 2006;37:572-573.

13. Wermer MJH, Van der Schaaf IC, Velthuis BK, Majoie CB, Albrecht KW, Rinkel GJE. How frequently should we follow small asymptomatic intracranial aneurysms? Stroke, 2006;37:1651.

14. Van der Schaaf IC, Wermer MJH, van der Graaf Y, Hoff R, Rinkel GJE, Velthuis BK. CT-Perfusion in the acute stage after SAH: relation of quantitative perfusion parameters and perfusion asymmetries with the development of delayed cerebral ischemia. Neurology, 2006;66;1533- 1538.

15. Van der Schaaf IC, Wermer MJH, Velthuis BK, Rinkel GJE. Psychosocial impact of fi nding but not treating aneurysms in screening of clipped patients. JNNP, 2006;77:748-752.

16. Halkes PH, Wermer MJH, Rinkel GJE, Buskens E. Direct costs of surgical clipping and endovascular coiling of unruptured intracranial aneurysms. Cerebrovasc Dis 2006;22:40- 45.

17. Van der Schaaf IC, Velthuis BK, Wermer MJH, Frenkel N, Witkamp T, de Kort G, Frehling N. Multislice CT-angiography in screening for new aneurysms in patients with previously clipped intracranial aneurysms: Feasibility, Positive Predictive Value and Interobserver Agreement, J Neurosurg, accepted for publication.

18. Van der Schaaf IC, Velthuis BK, Wermer MJH, Rinkel GJE. Perfusion in patients with a peri- mesencephalic hemorrhage and patients with an aneurysmal subarachnoid hemorrhage. Submitted.

19. Koffi jberg H, Buskens E, Algra A, Wermer MJH, Rinkel GJE. Growth rates of intracranial aneurysms: assessing constancy and linearity. Submitted.

233

Fourteen

ChapterFourteen

General discussionSamenvatting

SummaryDankwoord

List of publicationsCurriculum Vitae

236

Chap

ter

237

Fourteen

Marieke Wermer werd op 26 februari 1973 geboren te Almelo. Zij behaalde in 1991 haar gymnasium ß diploma aan het Bataafse Kamp te Hengelo. In hetzelfde jaar ging zij farmacie studeren aan de Universiteit Utrecht. In 1993 begon zij tevens aan de studie geneeskunde. Tijdens deze studie was ze student-assistent neuroanatomie en pathologie en nam ze deel aan de medische summerschool aan de Universiteit van Lund in Zweden. In 1996 deed zij een wetenschappelijke stage naar atypische presentaties van subarachnoidale bloedingen (begeleiding professor Rinkel). In 1997-1998 verrichte zij negen maanden wetenschappelijk onderzoek naar oxidatieve schade en neuroprotectie bij transgene muismodellen van neuro-degeneratieve ziekten in het Massachussetts General Hospital van de Harvard Medical School in Boston (begeleiding professor Beal). In 1998 behaalde ze haar doctoraal examen farmacie en haar doctoraal examen geneeskunde. Na een keuze co-schap neuropathologie, een keuze co-schap extra-pyramidale ziekten in het Massachussetts General Hospital in Boston en het behalen van het arts-examen in 2000 startte ze als assistent in opleiding op de afdeling neurologie van het Universitair Medisch Centrum Utrecht (opleider professor van Gijn). In 2002 ving zij aan met het wetenschappelijke werk dat in dit proefschrift wordt beschreven. In 2004 behaalde ze haar Master of Science diploma in de klinische epidemiologie aan “The Netherlands Institute of Health Sciences” (NIHES) in Rotterdam. Zij is sinds 1993 samen met Igor Schillevoort en in 2005 werd hun dochter Juliette geboren.