angiodysplasia: current conceptsangiodysplasia isa condition ofunknownaetiology in which...

5
Postgraduate Medical Journal (1988) 64, 259-263 Review Arficle Angiodysplasia: current concepts Anne P. Hemingway Professor of Diagnostic Radiology, Sheffield University, Sheffield, UK. Angiodysplasia is a condition of unknown aetiology in which microvascular abnormalities are found in the mucosa and submucosa of the bowel wall. The lesions are found predominantly in the caecum and right side of the colon and are frequently associated with either intermittent acute or continuous chronic intestinal blood loss. There is no family history and no recognized association with vascular abnor- malities of the skin or other organs. The lesions of angiodysplasia which are small (less than 5mm) and usually multiple are diagnosed by selective visceral angiography and/or colonoscopy but cannot be detected on barium enema or by the naked eye at laparotomy. Localization by the histo- pathologist is greatly facilitated by special injection techniques demonstrating the blood vessels of resected colonic specimens prior to fixation and section. The lesions are thought to represent a distinct benign pathological entity characterized in their early stage by dilated tortuous submucosal veins. In the more advanced stages there is further dilatation of the submucosal veins and venules and capillaries.' These characteristic features of right colonic vascular ectasias correspond to the definition given by Gentry et al.2 of telangectasias, i.e. dilatation of pre-existing vascular structures. Galdabini in 19743 first used the term 'angio- dysplasia' to describe the pathological abnormality found in a patient who presented with what is now recognized as a characteristic history' and the angio- graphic features of the condition defined above. There are two main clinical problems related to angiodysplasia; firstly, the fact that true prevalence of the condition in the general population is unknown and secondly the difficulty in detecting the disorder in patients with gastrointestinal bleeding of uncertain origin and, having made the diagnosis of angiodysplasia, establishing that it is the cause of the blood loss. How common? It has been stated that angiodysplasia may represent the commonest single cause of obscure gastrointestinal bleeding in the elderly population.4 Indeed a number of series have established that in patients undergoing visceral angiography in the investigation of obscure gastrointestinal bleeding the most commonly detected abnormality is angio- dysplasia.5 6 In our own published series,5 angio- dysplasia was diagnosed in 40% of patients examined. It must be remembered however that patients undergoing angiography represent a highly selected group since the cause of gastrointestinal haemorrhage will be identified by routine investigations in between 80% and 95% of cases.7 Angiodysplasia without doubt represents the commonest cause of bleeding in the residual 5 to 20% of patients. Extrapolating from these figures it can be seen that angiodysplasia accounts for between 2% and 8% of all patients presenting with gastrointestinal bleeding. This estimate is supported by the work of Richter et al.8 who diagnosed angiodysplasia colonoscopically in 26 patients, in a period when 1044 patients were colonoscoped for bleeding, a prevalence of 2.6%. The same group of workers detected the lesion in 13 patients out of a group of 1400 examined for reasons other than bleeding, a prevalence of 1.4%. Three principal in vitro studies have been performed in an attempt to address the question of the incidence of angio- dysplasia in the general population.4'9"0 In one study employing an injection/radiography technique,9 no lesions were detected in 39 autopsy specimens. The other two studies employed an injection and clearing technique followed by examination using trans-illumination and microscopy and both groups found lesions in up to 50% of specimens examined. We have recently reviewed over 450 visceral © The Fellowship of Postgraduate Medicine, 1988 Correspondence: Professor A.P. Hemingway B.Sc., M.B., B.S., M.R.C.P., D.M.R.D., F.R.C.R., Floor P, Royal Hallamshire Hospital, Glossop Road, Sheffield S1O 2JF, UK Received: 28 October 1987 copyright. on June 19, 2021 by guest. Protected by http://pmj.bmj.com/ Postgrad Med J: first published as 10.1136/pgmj.64.750.259 on 1 April 1988. Downloaded from

Upload: others

Post on 02-Feb-2021

0 views

Category:

Documents


0 download

TRANSCRIPT

  • Postgraduate Medical Journal (1988) 64, 259-263

    Review Arficle

    Angiodysplasia: current concepts

    Anne P. Hemingway

    Professor of Diagnostic Radiology, Sheffield University, Sheffield, UK.

    Angiodysplasia is a condition of unknown aetiologyin which microvascular abnormalities are found inthe mucosa and submucosa of the bowel wall. Thelesions are found predominantly in the caecum andright side of the colon and are frequently associatedwith either intermittent acute or continuous chronicintestinal blood loss. There is no family history andno recognized association with vascular abnor-malities of the skin or other organs. The lesions ofangiodysplasia which are small (less than 5mm)and usually multiple are diagnosed by selectivevisceral angiography and/or colonoscopy butcannot be detected on barium enema or by thenaked eye at laparotomy. Localization by the histo-pathologist is greatly facilitated by special injectiontechniques demonstrating the blood vessels ofresected colonic specimens prior to fixation andsection. The lesions are thought to represent adistinct benign pathological entity characterized intheir early stage by dilated tortuous submucosalveins. In the more advanced stages there is furtherdilatation of the submucosal veins and venulesand capillaries.' These characteristic features of rightcolonic vascular ectasias correspond to thedefinition given by Gentry et al.2 of telangectasias,i.e. dilatation of pre-existing vascular structures.

    Galdabini in 19743 first used the term 'angio-dysplasia' to describe the pathological abnormalityfound in a patient who presented with what is nowrecognized as a characteristic history' and the angio-graphic features of the condition defined above.

    There are two main clinical problems related toangiodysplasia; firstly, the fact that true prevalenceof the condition in the general population isunknown and secondly the difficulty in detectingthe disorder in patients with gastrointestinalbleeding of uncertain origin and, having made the

    diagnosis of angiodysplasia, establishing that it isthe cause of the blood loss.

    How common?

    It has been stated that angiodysplasia mayrepresent the commonest single cause of obscuregastrointestinal bleeding in the elderly population.4Indeed a number of series have established that inpatients undergoing visceral angiography in theinvestigation of obscure gastrointestinal bleedingthe most commonly detected abnormality is angio-dysplasia.5 6 In our own published series,5 angio-dysplasia was diagnosed in 40% of patientsexamined. It must be remembered however thatpatients undergoing angiography represent a highlyselected group since the cause of gastrointestinalhaemorrhage will be identified by routineinvestigations in between 80% and 95% of cases.7Angiodysplasia without doubt represents thecommonest cause of bleeding in the residual 5 to20% of patients. Extrapolating from these figures itcan be seen that angiodysplasia accounts forbetween 2% and 8% of all patients presenting withgastrointestinal bleeding. This estimate is supportedby the work of Richter et al.8 who diagnosedangiodysplasia colonoscopically in 26 patients, in aperiod when 1044 patients were colonoscoped forbleeding, a prevalence of 2.6%. The same group ofworkers detected the lesion in 13 patients out of agroup of 1400 examined for reasons other thanbleeding, a prevalence of 1.4%. Three principal invitro studies have been performed in an attempt toaddress the question of the incidence of angio-dysplasia in the general population.4'9"0 In onestudy employing an injection/radiography technique,9no lesions were detected in 39 autopsy specimens.The other two studies employed an injection andclearing technique followed by examination usingtrans-illumination and microscopy and both groupsfound lesions in up to 50% of specimens examined.We have recently reviewed over 450 visceral

    © The Fellowship of Postgraduate Medicine, 1988

    Correspondence: Professor A.P. Hemingway B.Sc., M.B.,B.S., M.R.C.P., D.M.R.D., F.R.C.R., Floor P, RoyalHallamshire Hospital, Glossop Road, Sheffield S1O 2JF,UKReceived: 28 October 1987

    copyright. on June 19, 2021 by guest. P

    rotected byhttp://pm

    j.bmj.com

    /P

    ostgrad Med J: first published as 10.1136/pgm

    j.64.750.259 on 1 April 1988. D

    ownloaded from

    http://pmj.bmj.com/

  • 260 A.P. HEMINGWAY

    angiograms performed for reasons other thangastrointestinal bleeding. Out of 166 patients inwhom the caecum and the right colon wereadequately visualized the characteristic features ofangiodysplasia were detected in 6 (3.6%). It is clearfrom the wide variation in these figures that furtherresearch is necessary to establish the true prevalenceof angiodysplasia in the non-bleeding population.

    Diagnosis

    The second major clinical problem relating toangiodysplasia is its recognition and this affectsthree groups of medical practitioner: clinicians(physicians and surgeons), radiologists andpathologists. These problems are summarized inTable I.

    It would seem that when both angiography andcolonoscopy are available to a clinician then it isadvantageous to utilize both techniques to make apositive diagnosis of angiodysplasia and excludeother pathology.The radiologist having accepted a patient for

    angiography must perform a 3-vessel study of highquality (i.e. coeliac, superior mesenteric and inferiormesenteric examinations). The characteristic featuresof angiodysplasia include vascular tufts visualizedin the arterial phase, an early filling draining veinand a slowly emptying, dilated, tortuous intramuralvein. In our institution we normally require at leasttwo of these three features to be present before thediagnosis is suggested (Figure 1). Angiodysplasiacan only definitely be implicated as the cause ofblood loss if the lesions are seen to be activelybleeding at the time of the study. Some authorities

    advocate that the colon should be distended withair at the time of angiography and that direct serialmagnification radiography should be used toimprove the chances of detection of these lesions.We have not employed these techniques in ourinstitution, finding that selective catheterization, theuse of adequate volumes of contrast media andhigh quality radiography are the most importantfactors.An important point to note is that the

    angiographic features of angiodysplasia can bemimicked by other conditions including inflam-matory bowel disease, such as Crohn's disease, andmalignancy and the possible existence of thesealternative pathologies should not be ignored. Ifangiography suggests the diagnosis of angio-dysplasia then other disease must be excluded bycolonoscopy, barium enema or even laparotomy.

    Treatment

    Once the diagnosis of angiodysplasia has beenmade the clinician is faced with the question oftreatment. Having excluded malignant disease, andif the blood loss is not severe, simple supportivetreatment for the anaemia may be adequate. Ifthe lesions are small in number colonoscopicfulguration is, in experienced hands, readilyperformed, relatively non-invasive and easilyrepeated. When the disease is more extensive, orother methods are failing to control the anaemiathen surgical resection is the treatment of choice.Laparotomy also allows the surgeon the chance todetect or exclude the presence of other lesionswhich may be the cause of the bleeding. In our own

    Table I Problems relating to the diagnosis and management of angiodysplasia

    The clinician's problems The radiologist's problems The pathologist's problems

    (i) Awareness of condition (i) Awareness of condition (i) Awareness of condition(ii) Mode of presentation (ii) Selection of patients for (ii) Identification of lesion(s) in

    angiography specimen

    (iii) Proof of the presence of (iii) Performing an adequate (iii) Classification of abnormalityangiodysplasia angiogram

    (iv) Proof that angiodysplasia is (iv) Interpreting the angiogramcausative lesion

    (v) Management decisions (v) Post-resection specimenradiography

    (vi) Lack of visibility of lesion atsurgery

    copyright. on June 19, 2021 by guest. P

    rotected byhttp://pm

    j.bmj.com

    /P

    ostgrad Med J: first published as 10.1136/pgm

    j.64.750.259 on 1 April 1988. D

    ownloaded from

    http://pmj.bmj.com/

  • ANGIODYSPLASIA 261

    a

    Figure la & b A selective ileocolic subtraction arteriogram. In the arterial phase (a) a prominent vessel isseen supplying the antimesenteric border of the caecum (arrow). In the venous phase (b) vascular lakes can beseen on the antimesenteric border (arrow); there is a prominent 'intramural' vein (arrow head) and a largedraining vein (curved arrow).

    series, over 20% of patients who underwentlaparotomy were found to have a second lesionwhich could have been responsible for the bloodloss. ' I

    Pathological confirmation of angiodysplasia in aresected specimen may be difficult. The lesions arevery small, not detectable by the naked eye and it isimpossible for a pathologist to section the entirespecimen. Injection of the specimen with either abarium gelatin mixture or latex material allows thelesions to be readily visualized by either directvision (Figure 2) or transillumination. Radiographyof a specimen1 2 which has undergone bariuminjection will often reveal far more extensive diseasethan was originally suspected in vivo (Figure 3).

    Pathological examination of the lesions of angio-dysplasia readily reveals why they bleed so easily.Dilated vascular channels are present in thesubmucosa and mucosa, and the latter areseparated from the bowel lumen by only one ortwo cell layers.

    Non-colonic angiodysplasia

    Traditionally the term angiodysplasia has only beenused to describe lesions found in the colon.Recently similar vascular abnormalities have beendiagnosed by both endoscopy and angiography inthe stomach'13"4 and small intestine.15 Theoccurrence of the condition at these sites is lesscommon and it is important to be certain thathistologically the lesions represent the samecondition and that they fulfil the criteria set out inthe definition at the beginning of this article.

    Aetiology

    There has been a great deal of speculation as to theaetiology of angiodysplasia. The most widelyquoted theory is that proposed by Boley4 which isthat it represents a degenerative condition related toageing and that repeated partial intermittent lowgrade obstruction of submucosal veins occurs. This

    copyright. on June 19, 2021 by guest. P

    rotected byhttp://pm

    j.bmj.com

    /P

    ostgrad Med J: first published as 10.1136/pgm

    j.64.750.259 on 1 April 1988. D

    ownloaded from

    http://pmj.bmj.com/

  • 262 A.P. HEMINGWAY

    Figure 2 The dissecting microscope appearances of angiodysplasia in a resected specimen that has beeninjected with a barium gelatin mixture. The normal honeycomb mucosal pattern is seen to be disrupted byenlarged tortuous vessels which have filled with barium.

    Figure 3 Microradiograph of an area of angio-dysplasia. The cluster of abnormal vessels is clearlyseen and the very large draining veins can beidentified.

    eventually results in dilatation of submucosal veinswhich in turn progresses to incompetence of theprecapillary sphincters and development of smallarteriovenous communications.

    Other theories are based on the supposition thatthe caecal mucosa is subjected, by a variety of

    mechanisms, to chronic ischaemia which damagesthe mucosa and causes arteriovenous shunts todevelop. The often reported association betweenaortic valve disease and angiodysplasia16 has beenused to support the theory of an ischaemicaetiology.None of these theories explains the occurrence of

    angiodysplasia in young people. In our own serieswe have seen the condition in teenagers12 and anumber of patients under the age of 40. We havealso seen the condition in association with Meckel'sdiverticulum"7 and have postulated that thecondition may have a congenital aetiology or thatpatients may have a congenital predisposition to itsdevelopment. More recently we have undertaken adetailed analysis of the distribution of angio-dysplastic lesions found following resection. In allpatients lesions are clustered in the mucosa in theimmediate vicinity of the ileocaecal valve and thepossibility that the small bowel contents exert achemical effect on the mucosa in this region whichpredisposes to the development of arteriovenousshunts has been proposed. This theory has beensupported by the discovery that in three patientsnew angiodysplastic lesions have developed in thecolonic mucosa close to the ileocolic anastomosis inpatients who have previously undergone hemi-colectomy for the condition.

    il

    r-l-

    -Y:: ....

    ..,I -- ..'...n4. ...:4. .,

    copyright. on June 19, 2021 by guest. P

    rotected byhttp://pm

    j.bmj.com

    /P

    ostgrad Med J: first published as 10.1136/pgm

    j.64.750.259 on 1 April 1988. D

    ownloaded from

    http://pmj.bmj.com/

  • ANGIODYSPLASIA 263

    Conclusions

    In conclusion it is clear that colonic angiodysplasiarepresents a major cause of chronic gastrointestinalblood loss and hence morbidity, particularly in theelderly population. Although the lesions ofangiodysplasia occur most commonly in the colon,similar abnormalities occur, but less frequently, inthe stomach and small intestine. The mostappropriate means of diagnosis is dependent onlocal expertise and facilities but is ideally providedby a combination of colonoscopy and angiography.The choice of treatment depends largely on the

    confidence with which other diseases can beexcluded and the severity of blood loss. The exactaetiology of angiodysplasia remains unknown butalmost certainly both congenital and acquiredfactors (possibly multifactorial) are implicated.

    Acknowledgements

    The author wishes to thank the department of MedicalPhotography, Royal Postgraduate Medical School andacknowledges that work described in this article wascarried out in collaboration with Professor D.J. Allison.

    References

    1. Mitsudo, S.M., Boley, S.J., Brandt, L.J., Montefusco,C.M. & Sammartano, R.J. Vascular ectasias of theright colon in the elderly: a distinct pathologicalentity. Hum Pathol 1979, 10: 585-600.

    2. Gentry, R.W., Dockerly, M.B. & Claggett, O.T.Vascular malformations and vascular tumours ofthe gastrointestinal tract. Int Abstr Surg 1949, 88:281-323.

    3. Case records of the Massachusetts General Hospital1974. Case 36. N Engl J Med 1971, 291: 569-575.

    4. Boley, S.J., Sammartano, R., Adams, A., DiBiase, A.,Kleinhaus, S. & Sprayregen, S. On the nature andetiology of vascular ectasias of the colon: degenerativelesions of ageing. Gastroenterology 1977, 72: 650-660.

    5. Allison, D.J., Hemingway, A.P. & Cunningham, D.A.Angiography in gastrointestinal bleeding. Lancet 1982,ii: 30-33.

    6. Sheedy, P.F., Fulton, R.E. & Atwell, D.T.Angiographic evaluation of patients with chronicgastrointestinal bleeding. AJR 1975 123: 338-347.

    7. Spiller, R.C. & Parkins, R.A. Recurrent gastro-intestinal bleeding of obscure origin: report of 17cases and a guide to logical management. Br J Surg1983, 70: 489-493.

    8. Richler, J.M., Hedberg, S.E., Athanasoulis, C.A. &Schapiro, R.H. Angiodysplasia: clinical presentationand colonoscopic diagnosis. Dig Dis Sci 1984, 29:481-485.

    9. Baer, J.W. & Ryan, S. Analysis of caecal vasculaturein the search for vascular malformations. AJR 1976,126: 394-405.

    10. Sabanathan, S. & Nag, S.B. Angiodysplasia of thecolon. J R Coll Surg Edinb 1982 27: 285-291.

    11. Steger, A.C., Galland, R.B., Hemingway, A., Wood,C.B. & Spencer, J. Gastrointestinal haemorrhage froma second source in patients with colonicangiodysplasia. Br J Surg 1987, 74: 726-727.

    12. Allison, D.J. & Hemingway, A.P. Angiodysplasia:does old age begin at nineteen? Lancet 1981, ii:979-980.

    13. Roberts, L.K., Gold, R.E. & Routt, W.E. Gastricangiodysplasia. Radiology 1981, 139: 355-359.

    14. Tung, K.T. & Millar, A.B. Gastric angiodysplasia: amissed cause of gastrointestinal bleeding. PostgradMed J 1987, 63: 865-866.

    15. Scott, S.D. & Royle, G.T. Angiodysplasia of the smallbowel; a 'brilliant' technique for localising thequiescent lesion. Postgrad Med J 1987, 63: 995-997.

    16. Galloway, S.J., Casarella, W.K. & Shimkin, P.M.Vascular malformations of the right colon as a causeof bleeding in patients with aortic stenosis. Radiology1974, 113: 11-15.

    17. Hemingway, A.P. & Allison, D.J. Angiodysplasia andMeckel's diverticulum: a congenital association? Br JSurg 1982, 69: 493-496.

    copyright. on June 19, 2021 by guest. P

    rotected byhttp://pm

    j.bmj.com

    /P

    ostgrad Med J: first published as 10.1136/pgm

    j.64.750.259 on 1 April 1988. D

    ownloaded from

    http://pmj.bmj.com/