adrenal venous sampling in primary hyperaldosteronism: comparison of radiographic with biochemical...

Adrenal venous sampling in primaryhyperaldosteronism: Comparison ofradiographic with biochemical successand the clinical decision-making with“less than ideal” testingAdrian Harvey, MD,a Gregory Kline, MD, FRCPC,b and Janice L. Pasieka, MD, FRCSC, FACS,a,b

Calgary, Alberta, Canada

Background. Adrenal venous sampling (AVS) is used in the workup of primary hyperaldosteronism(PA). The purpose of this study was to determine the success rate of AVS and to examine the decision-making process after “less than ideal” AVS.Methods. A total of 60 patients underwent 62 AVS for PA. Biochemical evidence of adrenal veincannulization was analyzed with adrenal–peripheral cortisol ratios. Pathology and clinical outcomeswere reviewed in patients undergoing adrenalectomy.Results. Bilateral cannulization was confirmed in only 21% (pre-adrenocorticotropic hormone[ACTH] infusion) and 44% (post-ACTH infusion) AVS. Of 39 patients who underwentadrenalectomy for presumed unilateral disease, only 16 patients had “ideal” AVS, and 18 patients hadonly unilateral cannulization on AVS. Despite this, 11 appeared to lateralize and 7 had imaging tosupport unilateral disease. Postoperatively, 15 (82%) had a significant reduction in their bloodpressure, and 7 (39%) of these were cured. Surgery failed in 2 patients; both were found to havebilateral hyperplasia. Bilaterally unsuccessful cannulization (n � 5) still lateralized in 3 patients, and2 patients had nodules on computed tomography scan. All 5 patients had significant reduction inblood pressure, and 2 were cured.Conclusions. ACTH infusion during AVS enhances the biochemical evidence of adrenal veincannulization. Following “less than ideal” AVS, clinical decisions can still be made using anatomicand partial AVS data. (Surgery 2006;140:847-55.)

From the Division of General Surgerya and the Division of Endocrinology,b University of Calgary, Alberta,
Canada
Primary hyperaldosteronism (PA), as a cause ofhypertension, was first described by Conn1 in 1955.Traditionally, screening for this disease using thealdosterone–renin ratio was reserved for patientswith the combination of hypertension and hypoka-lemia. Using this selective screening approach, the

Presented at the 27th Annual Meeting of the American Associ-ation of Endocrine Surgeons, New York, New York, May, 2006.

Accepted for publication July 28, 2006.

Reprint requests: Janice L. Pasieka, MD, FRCSC, FACS, Depart-ment of Surgery, Foothills Medical Center, 1403 29th Street NW,Calgary Alberta, Canada, T2N 2T9. E-mail: [email protected].

0039-6060/$ - see front matter

© 2006 Mosby, Inc. All rights reserved.

doi:10.1016/j.surg.2006.07.026

prevalence of PA was 0.5% to 2%.2,3 More recentapplication of nonselective screening using aldoste-rone–renin ratios has increased the estimated prev-alence of PA in the hypertensive population to ashigh as 6% to15%.4-6

The patient described by Conn in that original1955 report1 had a 4-cm aldosterone-producingadenoma (APA), and was cured of hypertensionand her hypokalemia by unilateral adrenalectomy.1

However, PA has more than one etiology. In addi-tion to the aldosterone-producing adenoma, thisdisease may result from bilateral adrenal hyperpla-sia (BAH), primary adrenal hyperplasia (PAH),and familial forms of PA. The implementation ofnonselective screening in several large centers hasresulted in a change in the proportion of patientswith surgically correctable unilateral disease. Prior
to universal screening, 68% to 86% of patients
SURGERY 847

848 Harvey et al SurgeryDecember 2006

with hyperaldosteronism were found to have al-dosterone-producing adenomas. In the modernscreening era, this proportion has dropped to ap-proximately 40% (range, 28% to 50%), likely re-flecting a greater detection of milder forms ofhyperaldosteronism associated with BAH.4

The increases in both the incidence of PA andthe proportion of patients with surgically nonreme-diable BAH underscores the importance of preop-erative testing that can distinguish accuratelybetween unilateral and bilateral causes of this dis-ease. Anatomic imaging using computed tomogra-phy (CT) or magnetic resonance imaging (MRI)can detect macroadenomas in the adrenal glands;however, the sensitivity of these modalities for smalladenomas and unilateral hyperplasia is low.7 Inaddition, such modalities have increased the detec-tion of nonfunctioning adrenal tumors that mustbe distinguished from APA. A recent prospectivestudy of 203 patients revealed that operative plan-ning based on anatomic imaging alone would haveresulted in 21.7% of patients inappropriately ex-cluded from adrenalectomy and 24.7% undergoingpotentially unnecessary surgery.8 The addition ofadrenal venous sampling (AVS) allows for the iden-tification of patients with unilateral disease not sug-gested by CT or MRI, as well as patients withmisleading or nonfunctioning unilateral findingson anatomic imaging.

AVS is a technically challenging procedure.Large centers with experienced radiologists reportcannulation of both adrenal veins in up to 95% ofpatients.8-10 However, not all centers have suchhigh success rates, with the average rate of success-ful cannulization in most centers being closer to74%.11 The ability to interpret AVS results whenone or both the adrenal veins are not cannulized isdebatable.8 Outside these large experienced cen-ters, many times the clinician is faced with inter-preting “less than ideal” results from AVS inpatients with PA. Recognizing that we did not al-ways have ideal AVS results, we set out to evaluateour own experience with AVS. The purpose of thisstudy was to compare the radiologist’s perceivedsuccess with biochemical evidence of successfulcannulization of adrenal veins, and to examine thepathologic, clinical outcomes, and decision-makingin patients in whom AVS results were “less thanideal.”

MATERIAL AND METHODSA retrospective review of all patients referred to

an endocrine surgeon or an endocrinologist forevaluation and treatment of PA. The office charts
of all patients undergoing AVS were reviewed. Dic-
tated procedure reports from the radiologist per-forming the venous sampling were graded as“successful” (class 1), “successful but with sometechnical difficulties” (class 2), or “unsuccessful/modified due to technical difficulties” (class 3).The AVS protocol involved simultaneous bilateralcannulization of both adrenal veins. A 250-�g bolusof adrenocorticotropic hormone (ACTH) followedby a slow infusion of 250 �g over 25 min wasadministered. Aldosterone and cortisol levels fromthe left and right adrenal veins and the inferiorvena cava below the renal veins were obtained pre-and post-ACTH bolus. When available, pre- andpost-ACTH infusion values were compared. For thepurposes of this study, biochemical evidence ofsuccessful adrenal vein catheterization was definedas an adrenal vein/IVC cortisol ratio of greaterthan 3:1 for pre-ACTH testing and greater than 5:1for post-ACTH testing.12,13 Those procedures notmeeting the biochemical criteria for cannulizationof one or both adrenal veins were defined as “lessthan ideal” AVS.

Results of additional imaging including CT orMRI imaging and NP-59 iodocholesterol scans wererecorded. Those patients believed to have unilat-eral disease proceeded to surgery; this group wascomposed of patients for whom outcome data andpathology were reviewed. For the purposes of thisstudy, lateralization on AVS was defined as an aldo-sterone–cortisol ratio at least 4 times greater thanthe unaffected side or an absolute aldosterone of10 times greater on the affected side compared tothe unaffected side. For patients undergoing sur-gery despite unilateral or bilateral unsuccessfulAVS (“less than ideal”), clinic notes were reviewedto determine what factors led to the diagnosis ofunilateral disease. Adrenal pathology was classifiedas APA, BAH, and PAH. PAH was defined as asuspected unilateral source of excess aldosteronesecretion on preoperative workup, the presence ofcortical hyperplasia with absence of nodules in theremoved adrenal gland, and a clinical response tounilateral adrenalectomy with a resolution of a hy-peraldosterone state.8

For those patients undergoing surgery, clinicaland laboratory variables were compared from theinitial clinic visit to the last follow-up. Outcomeswere recorded as positive or negative in 5 catego-ries: (1) uncontrolled blood pressure becomes con-trolled without an increase medication dose ornumber; (2) reduction in blood pressure medica-tions without a sacrifice of blood pressure control;(3) normokalemia without supplementation; (4) sys-tolic blood pressure decline of greater than 20 mm
Hg or diastolic blood pressure decline of greater

Surgery Harvey et al 849Volume 140, Number 6

than 10 mm Hg; and (5) cure of hypertension(blood pressure �140/80 mm Hg or no antihyper-tensive medications).

Comparison of the outcomes between successfuland “less than ideal” AVS groups was done with theFisher exact test, with a level of statistical signifi-cance set at P � .05. The University of CalgaryEthics Review Committee approved the study.

RESULTSA total of 60 patients underwent 62 AVS proce-

dures for PA. ACTH stimulation was done in 55 ofthese AVS. Demographic characteristics of the pa-tients are found in Table I. Dictated radiology re-ports were available for 61 procedures. Of theprocedures, 26 were reported as successful withoutdifficulty, 18 procedures had some technical prob-lems but were ultimately deemed successful, and 17procedures were difficult requiring modification ofthe procedure (ie, sampling from outside the ori-fice of the adrenal vein) (Table II).

Using our defined biochemical criteria, bothadrenal veins were cannulated successfully in only21% by pre-ACTH criteria and 44% by post-ACTHcriteria. Biochemical success rates in the 3 catego-ries of radiologic reports are shown in Table I.Despite the radiologist impression of successfulcannulation (class 1), biochemical success usingpost-ACTH criteria was confirmed in only 52% ofthe AVS.

Of the 60 patients, 14 patients have not beensurgically assessed and are currently being followedmedically. Of the group, 8 patients had bilaterallysuccessful AVS, 4 had BAH and 4 lateralized to 1side, of which 3 had normal CTs, and all electedmedical therapy. One patient had a nodule on CTbut refused surgical referral. Six patients had “lessthan ideal” AVS. Five patients had unilateral can-nulization; three of these did not meet lateraliza-tion criteria. Of these 3 patients, 1 had a unilateral

Table I. Demographic data

Patients

No. 60Sex 39 men, 21 womenMedian age 52 yMedian systolic BP 158 mm HgMedian diastolic BP 100 mm HgMedian no. of BP medications 3Median potassium preop 2.9 mEq/LMedian duration of hypertension 5 y

BP, Blood pressure.

mass on CT, 1 had bilateral nodules, and 1 had a

normal CT. All 3 did not want further testing, andthey are being managed medically. Two patientsdid meet lateralization criteria, yet only one had anodule on CT scan; both chose medical manage-ment. One patient had bilaterally unsuccessfulAVS, and normal CT imaging and NP59 scan; thispatient is being managed medically.

The remaining 46 patients were referred forsurgical consideration. Of these, 6 patients werenot offered adrenalectomy because the surgeonthought their AVS indicated bilateral disease (n �3), or the patients refused to undergo adrenalec-tomy (n � 3). The other 40 patients were offeredadrenalectomy. To date, 39 patients have under-gone surgery, and all but 1 of these patients wasbelieved to have unilateral disease. Seventeen pa-tients had biochemically successful catheter place-ment lateralized to one adrenal gland (Group A).Three of these patients had normal imaging. At thepresent time, 16 of these patients have had surgery.Eighteen patients with unilateral unsuccessful ve-nous cannulization were believed to still have uni-lateral disease and underwent adrenalectomy(Group B). Ten patients in Group B met lateraliza-tion criteria biochemically despite “less than ideal”AVS results. The decision to proceed to surgery inthese patients was influenced by CT scan in 8 pa-tients and based solely on AVS in 2. The remaining8 patients in Group B with nonlateralizing “lessthan ideal” AVS results underwent adrenalectomybecause of lateralizing disease that was seen on CTimaging (n � 4) or NP-59 scan (n � 3). An addi-tional patient with known BAH underwent unilat-eral adrenalectomy because of the significantdifference in aldosterone levels from one adrenalcompared to the contralateral side. Group C wascomposed of 5 patients with bilaterally unsuccessfuladrenal vein cannulization, who were believed tostill have unilateral disease. Despite the noncannu-lization of the adrenal veins, 3 patients met lateral-ization criteria on AVS, and 2 patients hadanatomic imaging to support unilateral disease.

Pathology in patients with successful as well asunilateral and bilateral unsuccessful AVS is shownin Table III. The proportion of the 3 clinical-patho-logic entities were similar between those patientswith successful and “less than ideal” AVS. The me-dian follow-up was 16 months (range, 1 to 84months). The majority of patients in all 3 groupsexperienced improved blood pressure control(100% A, 89% B, 100% C), with a significant pro-portion demonstrating a reduction in blood pres-sure medications (87% A, 72% B, 60% C) or adecline of greater than 20 mm Hg systolic or 10
mm Hg diastolic blood pressure (73% A, 83% B,


100% C) (Table IV). Clinical outcomes for GroupA were compared with combined outcomes forGroups B and C. No statistical difference was foundamong the groups on any of the 5 outcomes.

DISCUSSIONAs stated above, AVS is a technologically chal-

lenging procedure, especially cannulization of theright adrenal vein. Several large centers report suc-cess rates of up to 95% for this procedure.8-10 How-ever, a recent review of 47 reports that included384 patients, showed that the overall success rate ofAVS in a variety of centers was only 74%.11 In thispresent series, successful cannulation of both adre-nal veins as defined biochemically was achieved inonly 21% by pre-ACTH criteria and 42% by post-ACTH criteria, and it was not always the right ad-renal vein that failed to meet the biochemicalcriteria for cannulization. Some of the difficulty incomparing each center’s success rates is the lack ofstandardization of what defines biochemical suc-cess. In the literature, biochemical success has beendefined as adrenal–peripheral vein cortisol ratiosranging from 1.1 to 5.4,8-10,13-15 We used a ratio ofgreater than 3 for pre-ACTH testing and greater than5 for post-ACTH testing to compare our data againstlarge, experienced centers.8,9,13,15,16 Interestingly,

Table II. Percentage biochemically successful cannreport

Radiology report

% success (pre-ACTH)

Left Right

Class 1 62 31Class 2 50 33Class 3 35 18Total 51 27

ACTH, Adrenocorticotrophic hormone.

Table III. Pathology of adrenal specimens inpatients with successful (Group A), unilaterallyunsuccessful (Group B), and bilaterallyunsuccessful (Group C) adrenal venous sampling

Pathology

APA PAH BAH

Group A 15 1 0Group B 14 1 3Group C 4 1 0Total 33 3 3

APA, Aldosterone producing adenoma; PAH, primary adrenal hyperpla-sia; BAH, bilateral adrenal hyperplasia.

the radiologists’ perceived success appeared to cor-

relate very poorly with these biochemical criteriaof success. Procedures reported as successful with-out any technical difficulty had cannulation of bothadrenal veins in only 27% by pre-ACTH criteriaand 52% by post-ACTH criteria. This likely is areflection of not only the technically challengingnature of the procedure but also the fact that thebiochemical criteria of success that were used couldbe too stringent.

The use of ACTH greatly enhanced the successof identifying correctly cannulated adrenal veins.ACTH induces both cortisol and aldosterone pro-duction from the adrenal. This stimulus helps toverify catheter placement by minimizing artifactthat can arise from spontaneous steroid secretionfrom the adrenal glands. It also has been shownthat adenomas have a greater sensitivity to ACTHstimulus than hyperplasia. This allows the use ofexaggerated aldosterone–cortisol ratios betweenthe 2 sides for comparison, resulting in a greaterdetection of lateralization in patients with APA.14

Some centers, like ours, use pre- and post-ACTHresults; others use only the ACTH values in theirinterpretations of AVS.8,14

The data presented in this series clearly demon-strates that the success rate for AVS at our facility isbelow that achieved by other centers with an inter-est in this disease.8-10,13 This has prompted achange in our policy regarding AVS. During themajority of the study period, the physician respon-sible for angiographic procedures on the day of thetest performed the AVS. Now we have a dedicatedradiologist and all efforts are made to have thisprocedure done by this single individual. This strat-egy has been credited for the high success ratesreported by centers such as the Mayo Clinic.8 Al-though this change probably will improve the suc-cessful cannulization of AVS studies, there willcontinue to be patients in whom the AVS resultsare “less than ideal.” To date, no one has reportedon the outcomes of patients in whom the AVS wasnot ideal because many of these patients were likely

on of the adrenal veins by class of radiology

% success (post-ACTH)

teral Left Right Bilateral

7 91 52 522 94 59 592 57 21 71 84 44 42

ulati

Bila

2212

not referred for surgery. From this series, however,


it would appear that information from “less thanideal” AVS could still be used in the clinical deci-sion-making for these patients. For example, thefinding of an aldosterone–cortisol ratio suppressedbelow the peripheral ratio in a single cannulatedAVS may well suggest lateralization to the oppo-site side even if the involved side has not beencatheterized. The published criteria that havebeen used to demonstrate lateralization are in-cluded in Table V.4,8,13,15

Bilateral successful cannulation of the adrenalveins was achieved in only 16 of the 39 patients thatunderwent surgery. Despite this, all but 1 of these39 patients was thought to have unilateral diseasebased on multimodality preoperative testing. The sin-gle exception was a patient with very poor medicalcontrol of hypertension and known BAH. The AVSon this patient demonstrated a significant gradientof aldosterone secretion from one side comparedto the other. This adrenal was removed with thegoal of allowing better medical control of his “un-controlled” hypertension, which was achieved. Theremaining 38 patients were thought to have unilat-eral disease on completion of their workup; how-ever, 2 of these patients ultimately proved to haveBAH. These patients’ pathology revealed diffusecortical hyperplasia, with very little improvementin their hypertension on follow-up. Both had per-sistent elevated aldosterone–renin ratios confirm-ing the diagnosis of BAH. An additional 3 patientshad what is believed to be PAH. This entity of PAHconsists of hyperplasia of the adrenal gland andclinical improvement in blood pressure postopera-tively. All but one of these patients had a resolutionof their hyperaldosteronism states postoperatively.This suggests that 1 of these patients may prove,over time, to have BAH. Unilateral PAH has re-cently been reported to occur in as many as 31% ofpatients operated on for PA.17 The diagnosis of

Table IV. Clinical outcomes following adrenalecto

Testing groupBP

controlledDecrease

medication

Group A† (n � 15) 15 (100%) 13 (87%)Group B‡ (n � 18) 16 (89%) 13 (72%)Group C§ (n � 5) 5 (100%) 3 (60%)Total (n � 38) 36 (95%) 29 (76%)

SBP, Systolic blood pressure.*Blood pressure �140/90 mm Hg with no medications �20 mm Hg.†Bilateral cannulization.‡Unilaterally cannulization only.§Unsuccessful biochemical cannulization bilaterally.

PAH, however, depends on long-term follow-up

because the natural history of this condition is notyet clearly understood.17

Some authors have suggested that, because ofthe low incidence of adrenal incidentaloma inyoung patients, one can consider proceeding tosurgery with only anatomic evidence of unilateraldisease in patients under 40 years of age.8 In thisseries, this strategy was applicable in only 4 pa-tients, all of whom lateralized on AVS. However,none of the 8 patients from Group B operated onwith nonlateralizing AVS were under the age of 40years. This leaves the clinician with the dilemma ofhow to interpret the “less than ideal” AVS results inpatients with nodules seen on CT scanning. In the23 patients from Groups B and C, 10 patients whomet lateralization criteria had anatomic imaging tosupport unilateral disease. In 3 additional patients,surgical decisions were based solely on “less thanideal” AVS. In the remaining patients, the diagnosisof unilateral disease was based solely on the ana-tomic imaging and/or scintigraphy. Using thisstrategy, only 2 patients ultimately proved to haveunsuspected BAH and failed to improve followingthe removal of the hyperplastic gland. Despite “lessthan ideal” AVS, the adrenal gland pathology in

Outcome

Normokalemic20 mm Hg

decrease in SBP Cure*

14 (93%) 11 (73%) 2 (13%)14 (78%) 15 (83%) 7 (39%)5 (100%) 5 (100%) 2 (40%)

33 (87%) 31 (82%) 11 (29%)

Table V. Lateralization criteria for adrenalvenous sampling

RatioAffected adrenal tounaffected adrenal

Aldosterone 10:1ACTH-stimulated aldosterone 5:1ACTH-stimulated

aldosterone–cortisol4 or 5:1

ACTH-stimulatedaldosterone–cortisol of IVC

� than unaffected side

ACTH, Adrenocorticotropic hormone; IVC, inferior vena cava.

my

s

Groups B and C revealed a similar portion of uni-


lateral, surgically remediable disease compared toGroup A. These findings suggest that AVS can stillyield some useful data, even in the absence ofbilateral adrenal vein cannulation (as defined byour biochemical criteria), when combined with an-atomic imaging and/or functional scintigraphy.

More important than the histological diagnosisis the patient’s response to adrenalectomy. Theclinical outcomes in Group A were compared withthose in Groups B and C combined. No significantdifferences were noted in any of the outcome vari-ables. Overall, 29% of the patients were cured, 94%experienced improved blood pressure control, and82% had a significant drop in systolic or diastolicblood pressure. Comparison with other series islimited by the ambiguity in the definition of whatconstitutes cure. We used a stringent definitionrequiring a blood pressure of less than 140/80 onno antihypertensive medications. Other series uti-lizing similar criteria have reported similar curerates for APA following adrenalectomy rangingfrom 30% to 50%.18-20 Similar to the results pre-sented here, most surgical series report an improve-ment in blood pressure control or reduction ofantihypertensive medications in the range of 80%to 90% following adrenalectomy.17,18,20-22 Factorsshown to predict success following adrenalectomyinclude young age, short duration of disease, andsingle antihypertensive medication.21

The retrospective nature of this series makes itdifficult to truly assess the prospective applicationof treatment strategies undertaken when “less thanideal” AVS results were obtained. A weakness of ouranalysis lies in the admittedly arbitrary biochemicaldefinitions used to assess catheter placement. Useof more lenient cortisol ratios such as 2:1 may beequally valid and would certainly increase the pro-portion of AVS deemed successful. This may ex-plain our ability to detect useful lateralizationdifference between adrenals despite failing to meetthe predefined cortisol ratio criteria. By using thestrategy of applying all of the published criteria forlateralization in the context of “less than ideal” AVShelped the clinicians in the decision-making forthese patients. The best criterion to use for lateral-ization, however, has yet to be determined in theliterature.4,8,13,15 Until such a determination ismade, it would appear that the utilization of later-alization criteria in combination with anatomic andfunctional imaging when “less than ideal” AVS can-nulization has occurred can provide adequate in-formation in the majority of patients. Failure tomeet the strict cannulization criteria used in this
study should not preclude the clinician from eval-
uating all available information regarding lateral-ization.

In the era of nonselective screening, the numberof hypertensive patients diagnosed with PA has in-creased 5- to 15-fold; however, the proportion ofthese patients with non-surgically correctable BAHalso has increased.4-6 AVS has been proven to be anaccurate method to distinguish unilateral, surgi-cally treatable disease from BAH.8,10,15,23 AVS canbe a technically challenging procedure and success-ful cannulization of the adrenal veins may be elu-sive. Our data demonstrate that patients can bemanaged appropriately when AVS is “less thanideal” with interpretation of biochemical data inconjunction with anatomic and functional imaging.This series illustrates that the current criteria forbiochemical cannulization in AVS requires furtherinvestigation.

REFERENCES1. Conn JW, Louis LH. Primary aldosteronism: a new clinical

entity. Trans Assoc Am Physicians 1955;68:215-31.2. Ganguly A. Primary aldosteronism. N Engl J Med 1998;

339:1828-34.3. Melby JC. Diagnosis of hyperaldosteronism. Endocrinol

Metab Clin North Am 1991;20:247-55.4. Mulatero P, Stowasser M, Loh KC, Fardella CE, Gordon RD,

Mosso L, et al. Increased diagnosis of primary aldosteron-ism, including surgically correctable forms, in centers fromfive continents. J Clin Endocrinol Metab 2004;89:1045-50.

5. Stowasser M, Gordon RD, Gunasekera TG, Cowley DC,Ward G, Archibald C, et al. High rate of detection of pri-mary aldosteronism, including surgically treatable forms,after ‘non-selective’ screening of hypertensive patients. J Hy-pertens 2003;21:2149-57.

6. Omura M, Saito J, Yamaguchi K, Kakuta Y, Nishikawa T.Prospective study on the prevalence of secondary hyperten-sion among hypertensive patients visiting a general outpa-tient clinic in Japan. Hypertens Res 2004;27:193-202.

7. Balkin PW, Hollifield JW, Winn SD, Shaff MI. Primary aldo-steronism: computerized tomography in preoperative eval-uation. South Med J 1985;78:1071-3.

8. Young WF, Stanson AW, Thompson GB, Grant CS, FarleyDR, van Heerden JA. Role for adrenal venous sampling inprimary aldosteronism. Surgery 2004;136:1227-35.

9. Doppman JL, Gill JR Jr. Hyperaldosteronism: sampling theadrenal veins. Radiology 1996;198:309-12.

10. Rossi GP, Sacchetto A, Chiesura-Corona M, De Toni R,Gallina M, Feltrin GP, et al. Identification of the etiology ofprimary aldosteronism with adrenal vein sampling in pa-tients with equivocal computed tomography and magneticresonance findings: results in 104 consecutive cases. J ClinEndocrinol Metab 2001;86:1083-90.

11. Young WF Jr, Klee GG. Primary aldosteronism. Diagnosticevaluation. Endocrinol Metab Clin North Am 1988;17:367-95.

12. Dunnick NR, Doppman JL, Mills SR, Gill JR Jr. Preoperativediagnosis and localization of aldosteronomas by measure-ment of corticosteroids in adrenal venous blood. Radiology1979;133:331-3.

13. Young WF Jr, Stanson AW, Grant CS, Thompson GB, van


Heerden JA. Primary aldosteronism: adrenal venoussampling. Surgery 1996;120:913-9.

14. Carr CE, Cope C, Cohen DL, Fraker DL, Trerotola SO.Comparison of sequential versus simultaneous methods ofadrenal venous sampling. J Vasc Interv Radiol 2004;15:1245-50.

15. Magill SB, Raff H, Shaker JL, Brickner RC, Knechtges TE,Kehoe ME, et al. Comparison of adrenal vein sampling andcomputed tomography in the differentiation of primaryaldosteronism. J Clin Endocrinol Metab 2001;86:1066-71.

16. Dunnick NR, Doppman JL, Gill JR Jr, Strott CA, Keiser HR,Brennan MF. Localization of functional adrenal tumors bycomputed tomography and venous sampling. Radiology1982;142:429-33.

17. Novitsky YW, Kercher KW, Rosen MJ, Cobb WS, Jyothinaga-ram S, Heniford BT. Clinical outcomes of laparoscopicadrenalectomy for lateralizing nodular hyperplasia. Surgery2005;138:1009-16.

18. Blumenfeld JD, Vaughan ED Jr. Diagnosis and treatment ofprimary aldosteronism. World J Urol 1999;17:15-21.

19. Mantero F, Opocher G, Rocco S, Carpene G, Armanini D.Long-term treatment of mineralocorticoid excess syn-dromes. Steroids 1995;60:81-6.

20. Sawka AM, Young WF, Thompson GB, Grant CS, Farley DR,Leibson C, et al. Primary aldosteronism: factors associatedwith normalization of blood pressure after surgery. AnnIntern Med 2001;135:258-61.

21. Lo CY, Tam PC, Kung AW, Lam KS, Wong J. Primaryaldosteronism. Results of surgical treatment. Ann Surg1996;224:125-30.

22. Proye CA, Mulliez EA, Carnaille BM, Lecomte-Houcke M,Decoulx M, Wemeau JL, et al. Essential hypertension: firstreason for persistent hypertension after unilateral adrenal-ectomy for primary aldosteronism? Surgery 1998;124:1128-33.

23. Phillips JL, Walther MM, Pezzullo JC, Rayford W, ChoykePL, Berman AA, et al. Predictive value of preoperative testsin discriminating bilateral adrenal hyperplasia from an al-dosterone-producing adrenal adenoma. J Clin EndocrinolMetab 2000;85:4526-33.

DISCUSSIONDr William F. Young, Jr. (Rochester, Minn): Pa-

tients have become very savvy in looking for exper-tise when it comes to surgical procedures suchas transsphenoidal surgery or laparoscopic adrenal-ectomy. However, they have not become savvy yetwhen they ask for expertise on a radiologic proce-dure such as adrenal venous sampling. I think it isa real issue and important, as you are doing at yourinstitution, to sort out the factors that lead to apoor success rate and to focus the expertise byhaving 1 or 2 radiologists do the procedure.

Dr Harvey: The adrenal vein sampling policy formost of this period was that, if you wanted thisprocedure done, you put in a requisition. Whentime became available in the angiographic suite,the patient would be booked, and the radiologistwho happened to be in the angiographic suite was
the one doing the procedure. In total, there were 4
different radiologists doing a very limited numberof adrenal venous samplings.

Dr Young: I had a second question. You showedus several postoperative criteria for cure of primaryaldosteronism. We all know that finding an ade-noma at the time of surgery does not prove it wasan aldosterone-producing neoplasm; it only provesit was an adenoma. Seeing blood pressure dropafter surgery also does not mean it was an aldoste-rone-producing neoplasm. What were the labora-tory findings following surgery? For example, whatwere the plasma aldosterone–renin ratios or othermeasures of aldosterone secretion?

Dr Harvey: Aldosterone–renin ratio data is notavailable in all patients; therefore, we could notdemonstrate the resolution of hyperaldosteronismin this study. We defined those patients as havingan adenoma-producing aldosterone if they hadpathologic and clinical findings consistent with itand demonstrated improvement postoperatively.

We do follow the aldosterone–renin ratios inthose patients with pathology consistent with hyper-plasia. In those patients, as you recall, there were 3patients we felt had primary adrenal hyperplasia.All of those patients had a decrease in the aldo–renin ratio postop, and on the last follow-up, 2 ofthose patients still had a normal ratio. That ratiohas increased in 1 of those patients. Thus, it ispossible with longer follow-up that this patient maycross over that boundary from primary adrenal hy-perplasia to bilateral hyperplasia. The 2 patientswith what we called bilateral hyperplasia have per-sistent aldo–renin ratios and fail to demonstrateany improvement in their hypertension followingadrenalectomy.

Dr Quan-Yang Duh (San Francisco, Calif): Thatis a great paper, and I agree with you. When wereviewed our own experience, we found, paradox-ically, as the number of venous samplings in-creased, the success rate decreased. We don’t knowhow to explain that. Dr Young just is probably right;you really ought to have only one experiencedperson doing it.

My question is: Have you tried to alter your thresh-old for calling something positive or negative? Youshowed us different ratios. If you lower the threshold,do you get more positive studies and do you thenget more false positives?

Dr Harvey: It is true that a lot of the patients thatwere deemed to be unsuccessful had ratios justoutside the reference range. It is true that if welowered the threshold criteria for biochemical suc-cessful cannulization that more of our patientswould be successful. This likely explains why we are
able to utilize “less than ideal” results.


Currently we are analyzing a larger database tocompare the various published criteria in the liter-ature in the hopes of defining which criteria is thebest predictor of success. The major underlyingmessage here is not that our success rate was only42% but the fact that how we define success isarbitrary. In those patients who were “less thanideal,” information from the AVS was still utilizedand led to what we think was appropriate surgicalmanagement.

Dr Fieumu Nwariaku (Dallas, Tex): I have aquestion about the ACTH. I wonder why you ob-tained blood samples both pre- and post-ACTHadministration. Many endocrinologists now recog-nize that the use of ACTH prevents intraproceduralvariations in serum aldosterone and cortisol levels,and most people now routinely use ACTH stimula-tion during adrenal venous sampling.

And in the group that did not lateralize but wenton to surgery, how did you determine that they hadunilateral aldosterone hypersecretion? In otherwords, were there specific features on CT or MRI,or did the contralateral glands demonstrate sup-pressed aldosterone–renin ratios? Give us an ideaof how you took those patients to surgery eventhough they didn’t lateralize.

Then, lastly, the patients who didn’t go to sur-gery, did you use the selective mineralocorticoidreceptor antagonist, Eplerenone? Given the rela-tively mild side-effect profile of Eplerenone, whatdo you think is the role of Eplerenone in patientswho are not candidates for surgery?

Dr Harvey: To answer your first question, it hasbeen the protocol at the University of Calgary for along period of time to do simultaneous samplingboth pre- and post-ACTH infusion. As I mentioned,we are currently in the process of using our largerdatabase to assess the usefulness of comparing pre-ACTH with post-ACTH results.

For your second question, those patients whodid not meet our predefined lateralization criteriawent on to surgery because either they demon-strated unilateral disease on anatomic imaging, orthe AVS was suggestive of unilateral disease byeither meeting other lateralization criteria includ-ing the one you mentioned or a significant suppres-sion of the contralateral side.

Dr Nwariaku: I just wanted to know what yourthoughts were on the role of Eplerenone. It is aselective mineralocorticoid receptor antagonist. Idon’t know if that is available or if you use it forpatients who don’t go to surgery.

Dr Harvey: In terms of the 14 patients that did
not go to surgery, most of them are on an aldoste-
rone antagonist. However, I do not have the out-come data on them.

Dr Richard A. Prinz (Chicago, Ill): Clearly, ad-renal venous sampling is a tough test to perform. Itis very dependent on the skill of the radiographer.But the surgeon gets the results of the test to inter-pret, and there are published criteria for what is asuccessful test.

In this study, you had the information in approx-imately half the patients from which you couldconclude that the test was not a successful test, orvalid enough for you to interpret.

It doesn’t seem like we, as surgeons, operate onpatients when we have a test that is only 50% valid.So what do you do in that circumstance? I think youdo have the option of repeating the test or sendingthe patient elsewhere, because why are you order-ing a test that you are going to either ignore or notutilize in your decision-making?

Dr Harvey: I think that despite the fact that whatI presented here as a 42% success rate by biochem-ical criteria probably doesn’t reflect the actual suc-cess of these tests. For 10 of the 17 patients inGroup B, the information from the adrenal venoussampling was useful in making a surgical decision.And the information was useful for 3 of the 5patients in Group C. So the point we are trying todemonstrate here is that the actual success rate isprobably higher than 42%, and that the criteriathat we are applying to define success are too strin-gent.

My contention is not that we should ignore ad-renal venous sampling, but that we should betterunderstand how to interpret it and how to definewhat is successful. I agree there are criteria pub-lished in the literature. But I would bring up thepoint that those criteria range anywhere from acortisol ratio as low as 1.1 to 1, all the way up to 5to 1.

Dr Marco Raffaelli (Rome, Italy): I enjoyed yourpresentation, and I congratulate you on your study.I completely agree with you concerning the utilityof morphologic imaging techniques in patientswith aldosteronoma. If I understood correctly, onlya few patients in your series underwent adrenocor-tical scintigraphy. My question is, do you think thata more extensive use of adrenocortical scintigraphyon a routine basis, or even a selective, imagingmethod could be useful before, or even instead of,adrenal venous sampling in your practice, since itgives a morphofunctional imaging of the adrenalsand can provide very useful information for later-alizing the functioning lesion, especially when per-
formed after dexamethasone suppression?


Dr Harvey: Your question, as I understand it, is,should scinitigraphy be used more often?

Dr Raffaelli: My question is: Should adrenocor-tical scintigraphy be used more extensively before,or even instead of, adrenal venous sampling to local-ize the function of the lesion, at least in some cases?

Dr Harvey: I think that the problem with thescintigraphy (NP-59 scans) is that it has the same

sitivity is limited in small lesions. The average sizeof our adrenal adenomas was 1.6 cm (range, 0.5 to4.5 cm). Cost and universal availability are limits toits use in North America.

In these series, the use of functional imaging wasutilized in those patients in whom the adrenal ve-nous sample was less than ideal or unsuccessful andwe were unable to glean any useful information

limitations as the anatomic imaging in that the sen- from their results.

adrenal venous sampling in primary hyperaldosteronism: comparison of radiographic with biochemical...

Documents