Accuracy of Adrenal Imaging and Adrenal Venous Sampling in Predicting Surgical Cure of Primary Aldosteronism

Accuracy of Adrenal Imaging and Adrenal VenousSampling in Predicting Surgical Cure of PrimaryAldosteronism

Vivien Lim, Qinghua Guo, Clive S. Grant, Geoffrey B. Thompson,Melanie L. Richards, David R. Farley, and William F. Young Jr

Divisions of Endocrinology, Diabetes, Metabolism, Nutrition, and Internal Medicine (V.L., W.F.Y.) and theDivision of Gastroenterologic and General Surgery (C.S.G., G.B.T., D.R.F., M.L.R.), Mayo Clinic,Rochester, Minnesota 55905; and the Division of Endocrinology (Q.G.), Chinese PLA General Hospital,Beijing, People’s Republic of China 100853

Context: The accurate distinction between unilateral and bilateral adrenal disease in patients withprimary aldosteronism (PA) guides surgical management. Adrenal venous sampling (AVS), thecriterion standard localization procedure, is not readily available at many centers throughout theworld.

Objective: The objective of the study was to determine factors most consistent with surgicallycurable PA.

Design: This was a retrospective observational study.

Setting: The study was conducted at the Mayo Clinic (Rochester, Minnesota), a tertiary referralcenter.

Patients: All patients who underwent unilateral adrenalectomy for treatment of PA betweenJanuary 1993 and December 2011 participated in the study.

Intervention: The intervention in the study was unilateral adrenalectomy.

Main Outcome Measures: Variables associated with the prediction of unilateral disease weremeasured.

Results: Over 19 years, 263 patients underwent unilateral adrenalectomy for the treatment of PA.Long-term postoperative follow-up was obtained in 143 patients (54.4%). The overall effectivecure rate of PA was 95.5% in those patients sent for adrenalectomy for presumptive unilateraldisease. In patients with cured PA, defined as the resolution of autonomous aldosterone secretion,hypertension was cured in 53 (41.7%) and improved in 59 (46.5%) patients. PA was not cured withunilateral adrenalectomy in six patients (4.2%). Adrenal imaging and AVS were concordant to thesurgically documented side in 58.6% and 97.1% of the patients, respectively. Although there wasno statistically significant difference in mean age between the inaccurate vs the accurate adrenalimaging group, we found that the minimum age in the former was 35.1 years.

Conclusions: Using adrenal imaging and AVS, the effective surgical cure rate for PA was 95.5%.Although the overall accuracy of computed tomography and magnetic resonance imaging indetecting unilateral adrenal disease was poor at 58.6%, adrenal imaging performed well in thosepatients younger than 35 years of age. (J Clin Endocrinol Metab 99: 2712–2719, 2014)

ISSN Print 0021-972X ISSN Online 1945-7197Printed in U.S.A.Copyright © 2014 by the Endocrine SocietyReceived November 19, 2013. Accepted April 22, 2014.First Published Online May 5, 2014

Abbreviations: ALR, aldosterone lateralization ratio; APA, aldosterone-producing adeno-ma; ARR, aldosterone to renin ratio; AVS, adrenal venous sampling; BMI, body mass index;BP, blood pressure; CT, computed tomography; IHA, idiopathic hyperaldosteronism; IVC,inferior vena cava; MRI, magnetic resonance imaging; PA, primary aldosteronism; PAC,plasma aldosterone concentration; PRA, plasma renin activity.

O R I G I N A L A R T I C L E

E n d o c r i n e C a r e

2712 jcem.endojournals.org J Clin Endocrinol Metab, August 2014, 99(8):2712–2719 doi: 10.1210/jc.2013-4146

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 02 September 2014. at 01:39 For personal use only. No other uses without permission. . All rights reserved.

Primary aldosteronism (PA) is one of the most commoncauses of secondary hypertension (1). In a prospec-

tive Italian study, the prevalence of PA was 11.2% in pa-tients with hypertension (2). Jerome W. Conn reported thefirst patient with PA (3) in 1954, and she was cured withunilateral adrenalectomy. However, over the past 6 de-cades, our understanding of this disorder has evolved,with an increasing variety of subtypes being discovered,most of which cannot be cured with unilateral adrenalsurgery (4). The most common cause of PA is bilateralidiopathic hyperaldosteronism (IHA) due to zona glo-merulosa hyperplasia of the adrenal glands, followed byaldosterone-producing adenoma (APA) (2). It is impor-tant to distinguish between IHA and APA because surgicalremoval of the latter cures PA, whereas unilateral adre-nalectomy in patients with IHA simply debulks the diseasebut does not cure it. Bilateral adrenalectomy in patientswith IHA, although curative for PA, results in life-longadrenal replacement therapy and rarely cures the hyper-tension (5).

Through the years, clinicians have studied variousmethods to lateralize the source of aldosterone hyperse-cretion to better identify the subgroup of patients whowould benefit from unilateral adrenalectomy. An exten-sive systematic review reported that adrenal imaging lacksaccuracy in lateralization (6). A study done in our owninstitution found that in almost half of the patients withPA adrenal imaging, findings were discordant with surgi-cal pathology (7). Hence, adrenal venous sampling (AVS)has been advocated as a required step for the lateralizationof aldosterone excess (8). Because of the focused technicalexpertise needed for a successful AVS program (9, 10), itis not universally available and localization of unilateraladrenal PA remains problematic.

Herein we studied the preoperative characteristics ofour patients with definitive unilateral disease to determinethe factors that may aid us in identifying patients with PAwho are most likely to be cured with surgery. Cure of PAin our study was determined with long-term follow-up ofpatients who had undergone unilateral adrenalectomy.

Materials and Methods

This was a retrospective observational study that was re-viewed and approved by our institutional review board.We included all patients who underwent unilateral adre-nalectomy for the treatment of PA at Mayo Clinic (Roch-ester, Minnesota) between January 1993 and December2011. We reviewed the medical records to obtain infor-mation regarding: clinical characteristics, laboratory find-ings for the case detection and diagnosis of PA, computed

imaging scan findings, AVS results, surgical data, pathol-ogy reports, and postoperative follow-up data.

The aldosterone to renin ratio (ARR) was used as a casedetection test for PA and was calculated by dividing theplasma aldosterone concentration (PAC) by the plasmarenin activity (PRA). The lower limit of detection for PRAwas 0.6 ng/mL/h and for statistical analyses, PRA was setat 0.6 ng/mL/h for those PRA values reported as less than0.6 ng/mL/h.

All patients had adrenal-directed computed imagingwith computed tomography (CT) or magnetic resonanceimaging (MRI). In addition, most the patients in this sur-gical cohort underwent AVS. AVS was performed duringcontinuous cosyntropin infusion (50 �g/h) (7).

Catheter placement in the adrenal veins was confirmedbased on an adrenal vein to the inferior vena cava (IVC)cortisol concentration gradient of more than 5:1. We cal-culated the AVS cortisol-corrected aldosterone lateraliza-tion ratio (ALR) by dividing the adrenal vein PAC by thecortisol concentration for each adrenal vein and then di-viding the higher value by the contralateral side (7). Wealso divided the nondominant adrenal vein PAC to cortisolratio by the IVC PAC to cortisol ratio for the contralateralALR (7).

The postoperative PAC was typically obtained the dayafter adrenal surgery. At various times over the 18 years ofthis study, the PAC assay lower limit of detection was 1.0,2.0, or 4.0 ng/dL; for those patients with values belowthese cutoffs, the lower limit of detection was used forstatistical analyses.

We contacted patients via telephone interview to getupdated information regarding their blood pressure (BP),medications, and, where possible, potassium values, po-tassium supplementation needs, and aldosterone values.We studied the group of patients who had follow-up datato ascertain the cure of PA. Cure was adjudicated based ona combination of the postoperative PAC, cure of hypoka-lemia, and improvement or cure of hypertension. Cure inthis group of patients therefore denoted proof of definitiveunilateral disease. The group with definitive unilateral dis-ease was further analyzed based on the adrenal computedimaging findings (ie, unilateral vs bilateral nodularity) inwhich the responsible adrenal gland was confirmed withthe surgery. Unilateral disease on imaging was defined byabnormalities in one adrenal gland with the contralateraladrenal gland being normal. The concordance rates ofcomputed imaging findings with AVS data were alsodetermined.

Hypertension cure was defined as a normal BP of lessthan 140/90 mm Hg without the aid of antihypertensivemedications. Hypertension improvement was defined as

doi: 10.1210/jc.2013-4146 jcem.endojournals.org 2713


similar or improved BP on fewer antihypertensivemedications.

Aldosterone levels were analyzed by using a RIA tomeasure aldosterone liberated by hydrolysis (pH 1) of the18�-glucosiduronic acid conjugate. The aldosterone mea-surement methodology was changed to a liquid chroma-tography-tandem mass spectrometry assay in 2007 (11).Cortisol was measured by a RIA method (12) using a com-mercially available kit (Diagnostic Products Corp). Thecortisol measurement methodology was changed to a liq-uid chromatography-tandem mass spectrometry methodin 2010 (11). PRA was determined by measuring theamount of angiotensin I generated during a 1-hour incu-bation at pH 7.4. PRA less than 0.6 ng/mL/h was the lowerlimit of detection.

All values were reported as mean � SD or median (in-terquartile range) when appropriate for continuous dataand percentages for categorical data. Continuous datawere analyzed by a two-sample t test or a Wilcoxon rank-sum test if the underlying distribution was skewed. Cat-egorical data were analyzed by the Fischer’s exact test orPearson’s �2 test where appropriate. A value of P � .05denotes statistical significance.

Results

Between January 1993 and December 2011, 263 patientsunderwent unilateral adrenalectomy for the treatment ofPA (Figure 1). Table 1 provides a summary of the preop-erative data. Long-term postoperative follow-up was ob-tained in 143 patients (54.4%) (Figure 1). The postoper-ative status of PA was unknown in 10 patients (7%).Except for body mass index (BMI), there were no statis-tically significant differences in the preoperative datawhen comparing the cohorts with and without long-termfollow-up (Table 1). PA was cured in 127 of the 133 pa-tients with known PA status postoperatively (95.5%) (Fig-ure 1). PA was not cured with unilateral adrenalectomy insix patients (4.2%): these patients either had bilateral al-dosterone excess or contralateral adrenal aldosterone ex-cess. For the two subgroups of patients in whom PA waseither cured or not cured, the preoperative data as well asthe follow-up data are summarized in Table 2.

Surgically cured PA subgroupThe overall effective cure rate of PA was 95.5% in pa-

tients sent for adrenalectomy for presumptive unilateraldisease. The mean age of the patients was 50.9 � 10.7years with a minimum age at surgery of 17.3 years, and themaximum age was 76.8 years. The mean BP was 147 � 23mmHgsystolic and88�13mmHgdiastolicwhen treated

with a mean of 2.9 � 1.3 antihypertensive medications.Potassium chloride supplementation was administered to70.9% of the patients with a median supplementation of40 mEq/d (range 0–80). The median PAC was 30 ng/dL(range 19–43). PRA was less than the lower limit of de-tection in 89.5% of the patients. The median ARR was 50(range 31.4–74.2).

Formal confirmatory testing was obtained in 114 of thepatients (89.6%); an oral sodium loading test was used inall but one of these patients. The median 24-hour urinealdosterone excretion was 32 �g (25, 49). Due to the pre-sentation of spontaneous hypokalemia, markedly in-creased PAC, and undetectable PRA, in 10.4% of the pa-tients, formal confirmatory testing was not needed.Localization was performed by adrenal imaging in all thepatients and AVS in 80.3% of the patients. Bilateral AVSwas unsuccessful in four of the patients (3.9%). The me-dian cortisol-corrected ALR was 14.8 (range 7.4–26.3) inwhich the minimum ALR was 2.65. The AVS lateralizedto the left adrenal gland in 60.2% of the patients and to theright adrenal gland in 39.8% of the patients. The medianPAC on the affected side was 7245 ng/dL (range 3740–

Total cohorta

n = 263

Pa�ents with follow-up dataa

n = 143

Pa�ents without follow-up dataa

n = 120

Not cured (bilateral disease)b

n = 6

Unknown status

n = 10

Cured (unilateral disease) b

n = 127

Inaccurate CT or MRIc n = 50 (41%)

Bilateral abnormali�es on scan (n = 48)

Unilateral abnormality on scan but in the opposite adrenal

(n = 2)

Accurate CT or MRIc

n = 77 (59%)

Figure 1. Flow chart showing patients with postoperative follow-up,disease status postoperatively, and findings on computed imaging ofthe adrenal glands. a, Preoperative details of these groups of patientsare shown in Table 1; b, pre- and postoperative details of these groupsof patients are shown in Table 2; c, preoperative details of thesegroups of patients are shown in Table 3.

2714 Lim et al Predicting Surgical Cure of Primary Aldosteronism J Clin Endocrinol Metab, August 2014, 99(8):2712–2719


11 819) with a maximum value of 106 080 ng/dL and aminimum value of 1044 ng/dL.

The pathology in 108 patients (85.0%) was a singleadrenal adenoma, four (3.1%) had more than one ade-noma, and 15 patients (11.8%) had unilateral zona glo-merulosa hyperplasia. The mean diameter of adrenal ad-enomas in those patients with a single tumor was 1.6 � 0.8cm (range 0.3–5.5 cm). An APA less than 1 cm diameterwas found in 21 patients (16.5%). The postoperative PACwas below the assay limit of detection in 72.7% of thepatients.

The mean follow-up was 6.8 � 4.9 years with a max-imum follow-up of 18 years. At the time of the last follow-up, hypertension was cured in 53 patients (41.7%).Whereas 59 patients (46.5%) showed hypertension im-provement, nine (7.1%) showed no improvement, and six(4.7%) showed worsening hypertension on follow-up. Inthose with hypokalemia preoperatively (n � 110), 107patients (97.3%) were cured of their hypokalemia, and thepostoperative potassium status was unknown in the re-maining three patients.

Of the preoperative variables, patients who had a curein PA with surgery and therefore definitive unilateral dis-ease had statistically significantly lower serum potassiumconcentration, higher cortisol-corrected ALR on AVS,and lower contralateral adrenal vein cortisol-correctedALR when compared with those patients who were notcured (Table 2). A multivariate analysis showed the cor-tisol-corrected ALR to be more dominant in effect than thepotassium level.

We also studied the preoperative characteristics of pa-tients with surgically confirmed unilateral disease whohad unilateral adrenal abnormalities on computed imag-

ing, which correlated with the surgical side and comparedthem with those with bilateral adrenal abnormalities onadrenal imaging (Table 3). The PAC and ARR were sta-tistically significantly higher in the patients with unilateralfindings on CT or MRI (Table 3). There were more malesin the imaging group with bilateral adrenal abnormalitiesand a higher number of BP medications required. In thesubgroup of cured patients, 102 patients (80.3%) under-went AVS. Forty-four patients who underwent AVS(44.9%) had discordant findings on scans compared withAVS lateralization: 42 patients had bilateral adrenal ab-normalities on scan but AVS lateralized; the remainingtwo patients showed apparent unilateral adrenal diseaseon scans, but the AVS lateralized to the contralateral ad-renal gland. For these latter two patients, one was aged71.4 years and the other 54.1 years.

In all of the patients with successful AVS, it lateralizedPA with a 100% concordance rate with the surgically doc-umented tumor side.

Persistent PA despite surgery subgroupPA was not cured with unilateral adrenalectomy in six

patients (4.2%). The preoperative data are summarized inTable 2. All six were men with a mean age of 54.7 � 6.9years. The maximum systolic and diastolic BPs recordedwere 170 and 98 mm Hg, respectively. The maximumpotassium chloride supplementation was 60 mEq/d with aminimum serum potassium concentration of 3.8 mEq/L.

Adrenal imaging showed unilateral disease in five pa-tients (83.3%). AVS was performed in five patients(83.3%), with bilateral AVS being unsuccessful in one ofthe patients. Of the four with successful AVS, two hadimaging that was concordant with the AVS lateralization;

Table 1. Preoperative Parameters on 263 Patients With and Without Follow-Up Dataa

Total CohortFollow-Up DataAvailable

Follow-Up DataNot Available

n 263 143 (54.4%) 120 (45.6%)Sex Male: 167 (63.5%) Male: 89 (62.2%) Male: 78 (65.0%)Age at surgery, y 51.4 � 10.9 51.4 � 10.7 51.3 � 11.2BMI, kg/m2 30.7 � 5.6 30.1 � 5.6 31.5 � 5.6b

BP systolic, mm Hg 150 � 22 148 � 22 153 � 22BP diastolic, mm Hg 88 � 12 88 � 12 88 � 12Number of BP medications 3.0 � 1.4 2.9 � 1.3 3.1 � 1.5Number taking potassium supplements, % 184 (70%) 99 (69.2%) 85 (70.8%)Serum potassium, mmol/L 3.7 � 0.5 3.7 � 0.5 3.7 � 0.5Potassium supplement, mEq/d 40 (0, 80) 40 (0, 75) 40 (0, 80)Serum aldosterone, ng/dL 30 (20, 42) 31 (21, 42) 30 (19, 43)Plasma renin activity, ng/ml/h 0.6 � 0.1 0.6 � 0.1 0.6 � 0.1ARR 50.0 (32.0, 75.0) 49.2 (30.3, 72.9) 50.0 (33.3, 75.0)24-hour urine aldosterone, �g 32 (24, 49) 32 (26, 49) 33 (23, 44)CT or MRI scan with unilateral adrenal abnormality, % 173 (65.8%) 91 (63.6%) 82 (68.3%)Number who had AVS, % 213 (81.0%) 116 (81.1%) 97 (80.8%)

a Unless noted otherwise, data shown as mean � SD or median (interquartile range) or as number (percentage of cohort).b P .04.



one patient had bilateral disease on computed imaging butlateralized on AVS. The remaining case showed bilateraldisease with AVS and unilateral disease with adrenal im-aging: unilateral adrenalectomy was performed in this pa-tient with the goal of debulking the source of aldosteroneexcess in an effort to treat the patient’s resistant hyper-tension. The maximum AVS PAC on the affected side inthis group of patients was 11 000 ng/dL. The histologyrevealed a cortical adenoma in five patients and the re-maining patient had hyperplasia. The mean diameter ofadrenal adenomas in those patients with a single tumorwas 2.3 � 1.5 cm (range 1–4.8 cm).

The minimum postoperative PAC was 10 ng/dL. Therewas a difference of two antihypertensive agents betweenthe cured and uncured subgroups postoperatively. Noneof the patients in this persistent PA subgroup had a cure fortheir hypertension, and two (33.3%) had worsening oftheir hypertension. One of the three patients with preop-erative hypokalemia had a postoperative cure in hypoka-

lemia compared with 97.3% in the cured group (P �.0005).

Accuracy of adrenal imaging and AVSThe accuracy of adrenal imaging and AVS was deter-

mined in the 133 patients with long-term follow-up andknown PA cure status (Table 4). Computed imaging andAVS were concordant to the surgically documented side in58.6% and 97.1% of the patients, respectively. The meanof the ages of the groups with inaccurate scans vs accuratescans were 50.2 � 11.3 and 52.4 � 9.2 years, respectively.There were six patients who were younger than 40 yearswith inaccurate scans and 16 patients between the ages of40 and 50 years. In patients younger than 35.1 year of age,the scan results were 100% concordant with AVS results.The relationship between the size of the adenoma and theage of the patient younger than 40 years is illustrated inFigure 2. When comparing the patients who underwentAVS with those who did not, the proportion of males

Table 2. Preoperative and Postoperative Data in Those Patients With Known Surgical Outcomesa

PA Cured PA Not Cured

Preoperative datan 127 6Sex Male: 74 (58.3%) Male:6 (100%)Age at surgery, y 50.9 � 10.7 54.7 � 6.9BMI, kg/m2 30.0 � 5.6 30.9 � 5.6BP systolic, mm Hg 147 � 23 158 � 11BP diastolic, mm Hg 88 � 13 92 � 9Number of BP medications 2.9 � 1.3 3.7 � 1.4Number on potassium supplements, % 90 (70.9%) 4 (66.7%)Serum potassium, mmol/L 3.7 � 0.5 4.1 � 0.3b

Potassium dosage, mEq/d 40 (0, 80) 25 (0, 60)Plasma aldosterone, ng/dL 30 (19, 43) 23 (11, 58)Plasma renin activity, ng/ml/h 0.6 � 0.1 0.7 � 0.1ARR 50.0 (31.4, 74.2) 30.0 (18.9, 92.9)24-hour urine aldosterone, �g per 24 h 32 (25, 49) 41 (30, 81)CT or MRI scan with unilateral adrenal abnormality 79 (62.2%) 5 (83.3%)Number who had AVS, % 102 (80.3%) 5 (83.3%)AVS cortisol-corrected ALR 14.8 (7.4, 26.3) 5.5 (3.2, 7.4)c

Contralateral adrenal-IVC cortisol-corrected lateralization ratio 0.3 (0.2, 0.5) 0.8 (0.5, 0.8)b

Adrenal vein aldosterone from presumptive affected adrenal, ng/dL 7245 (3740, 11819) 10682 (2688, 11000)Postoperative follow-up

Postoperative plasma aldosterone, ng/dL 2 (1, 4) 20 (10, 29)Duration of follow-up of BP, y 6.8 � 4.9 7.2 � 3.3BP systolic, mm Hg 125 � 15 132 � 19BP diastolic, mm Hg 77 � 9 79 � 4Number of BP drugs 1 (0, 2) 3 (2.8, 3.5)d

Change in number of BP drugs 1.7 � 1.4 0.5 � 1.4Number of patients with BP cure, % 53 (41.7%) 0 (0%)Duration of follow-up of serum potassium, y 6.6 (2.6, 11.1) 7.7 (3.9, 10.1)Serum potassium, mmol/L 4.5 � 0.5 4.4 � 0.4Number of patients on potassium supplements, % 4 (3.1%) 2 (33.3%)b

Duration of follow-up of plasma aldosterone, y 3.9 (0.8, 10.3) 5.3 (1.6, 8.9)Plasma aldosterone, ng/dL 7 (4, 11) (n � 16) 24 (12, 36) (n � 2)

a Unless noted otherwise, data are shown as mean � SD or median (interquartile range) or as number (percentage of cohort).b P � .03.c P � .02.d P � .001.



(66.7% vs 50.0%), age at surgery (52.2 � 10.4 vs 48.0 �12.3 y), PAC [28 (range 18–41) vs 36 (range 26–58) ng/dL], ARR [45.8 (range 32.0 –75.0) vs 61.7 (range 41.7–106.1)], and adrenal imaging with unilateral abnormal-ity (59.6% vs 92%, respectively) were statisticallysignificant.

Discussion

Of the 10 known subtypes of PA (4), IHA is the mostcommon (2). Treatment options for patients with PA in-clude mineralocorticoid receptor antagonists (eg, spirono-lactone or eplerenone) and, in selected patients in whomthere is proof of unilateral adrenal disease, unilateral ad-renalectomy. The latter has been shown to be more costeffective in the long-term management of PA for surgically

curable patients (13). Laparoscopic adrenalectomy is thesurgical procedure of choice for these patients with a mor-bidity between 5% and 14% and a mortality of less than1% (14). A recent systematic review of the studies showedthat the cure of hypertension was achievable in 42% of thecases, with the vast majority of the patients at least ob-taining an improvement in BP after unilateral adrenalec-tomy; hypokalemia was cured in more than 95% of cases,whereas left ventricular hypertrophy usually improves orresolves (14).

Given the high prevalence of PA in the hypertensivepopulation and the unique opportunity to cure hyperten-sion, accurate methods to select patients who have thehighest probability of unilateral disease is a key point tothe optimal management of PA. In addition to APA, uni-

Figure 2. Relationship between size of the adenoma and age of thepatients younger than 40 years.

Table 4. Accuracy of Adrenal Imaging and AVS inThose Patients With Long-Term Follow-Up and KnownPrimary Aldosteronism Cure Statusa

Bilateral,accurate

Bilateral,inaccurate

Unilateral,accurate

Unilateral,inaccurate

CT and MRI (n � 133)1 (0.8%) 48 (36.1%) 77 (57.9%) 7 (5.3%)

AVS (n � 102)b

1 (1%) 0 (0%) 98 (96.1%) 3 (2.9%)

a Includes 133 patients, regardless of status of cure.b AVS was not performed in 26 patients (19.5%) and was unsuccessfulin five (3.8%).

Table 3. Computed Imaging Findings in Those Patients With Surgical Cure of Primary Aldosteronisma

Accurate Unilateral AdrenalAbnormality on CT or MRI Scan

Inaccurate AdrenalCT or MRI Scanb

n 77 50Male sex, % 39 (50.6%) 35 (70.0%)c

Age, y 50.1 � 11.4 52.2 � 9.4BMI, kg/m2 29.5 � 5.8 30.7 � 5.3BP systolic, mm Hg 147 � 25 147 � 20BP diastolic, mm Hg 88 � 13 88 � 12Number of BP medications 2.7 � 1.3 3.3 � 1.3d

Number taking potassium supplements, % 59 (76.6%) 31 (64.6%)Serum potassium, mmol/L 3.7 � 0.5 3.7 � 0.5Potassium dosage, mEq/d 40 (10, 80) 40 (0, 60)Serum aldosterone, ng/dL 34 (23, 50) 24 (15, 36)e

Plasma renin activity, ng/ml/h 0.6 � 0.1 0.6 � 0ARR 56.7 (36.7, 89.6) 40 (25.4, 59.6)f

24-hour urine aldosterone, �g 33 (26, 50) 29 (21, 44)a Does not include patients who were not cured after adrenalectomy and hence had bilateral disease. Unless noted otherwise, data are shown asmean � SD or median (interquartile range) or as number (percent of cohort).b This includes patients in the definitive unilateral adrenal group who showed bilateral abnormality on scans or unilateral abnormality, but thesurgically documented side was contralateral.c P � .04.d P � .01.e P � .003.f P � .002.



lateral disease also includes unilateral zona glomerulosahyperplasia, as was present in 11.8% of our cured PApatients. Thus, we studied the preoperative characteristicsof our patients with definitive unilateral disease to deter-mine factors that may aid us in selecting patients who maybenefit from adrenalectomy and might be able to forgoAVS.

Follow-up data were available in 143 of our 263 pa-tients (54.4%). A strength of our study lies in a relativelylarge number of patients who had undergone unilateraladrenalectomy, who had a mean of 6.9 years of follow-up.The preoperative finding of a lower serum potassium con-centration in the subgroup that had a cure in PA comparedwith the uncured subgroup was consistent with the con-cept that patients with APA generally have a more severePA profile than those with IHA. In addition, the curedgroup had a higher cortisol-corrected ALR as well as alower contralateral ALR, in keeping with more markedunilateral disease in the affected side and suppression inthe contralateral side. Similarly, it is not surprising that theunilateral disease group had a smaller number of BP drugsand a smaller proportion of patients requiring potassiumchloride supplementation on follow-up. Although wefound a substantial absolute difference between the twogroups with regard to the PAC and ARR, the small numberof patients (n � 6) in the uncured subgroup lowered thepower of the overall analysis.

On follow-up, six of the patients in the definitive uni-lateral disease group (4.7%) showed worsening of BP vstwo of the patients in the uncured group (33.3%). In theformer group, one had declining renal function and an-other had allograft rejection post renal transplant. Therewas no discernible cause of the worsening hypertension inthe other patients in both subgroups.

Of note is the finding that 15 of patients with definitiveunilateral disease in our study (11.2%) had unilateral zonaglomerulosa hyperplasia. Patients with this subtype of PAwere not distinguishable from APA or IHA on adrenalimaging. Indeed, seven of these patients (46.7%) had ad-renal imaging reported as bilateral, whereas the remainderlateralized accurately on computed imaging.

All the patients in the uncured group were men; how-ever, this did not reach statistical significance and may bea function of our study’s higher male proportion. Ourpredominantly male study population also explains whythere was a higher proportion of men in the scan groupdemonstrating bilateral adrenal abnormality. The PACand ARR were higher in the unilateral scan group, a find-ing consistent with the concept that a larger adenoma, andhence one producing more aldosterone, is more likely to bevisible as a unilateral adenoma on computed imaging.

The PAC after surgery was a good predictor of cure. Allpatients who had postoperative PAC less than the lowerlimit of detection in the assay had a cure. The lowest valuefor the uncured patients was 10 ng/dL. The highest valuefor the cured patients was 6.4 ng/dL.

Overall, adrenal imaging was discordant in 41.4% ofthe patients and in 44.9% of the patients with surgicallyproven unilateral adrenal disease; although slightly higherthan the 37.8% discordance rate in a recent systematicreview (6), it is consistent with an AVS-based study pub-lished previously by our group (7). Although there was nostatistically significant difference in mean age between theoverall inaccurate scan group vs that of the accurate scangroup, we found that the minimum age in the former was35.1 years, a finding that would support the suggestionthat AVS should be considered for any patient who seeksa surgical cure and who is older than 35 years of age,regardless of the adrenal imaging findings. Although weshould note that our study includes only those patientswith PA who went to surgery and not those who may havehad a unilateral adrenal finding on computed imaging butbilateral disease on AVS. In addition, it should be notedthat AVS was accurate in 96.1% of the patients in thisstudy and is the most accurate way of diagnosing unilat-eral forms of PA.

There are several limitations to our study. The group ofpatients that was not cured with surgery was small (n � 6),which limited the power of statistical comparisons withthe cured subgroup. Of these, four had successful AVS, ofwhich three denoted unilateral PA. No follow-up datawere available in 120 of our 263 patients who underwentsurgery to treat PA (45.6%). Because our study extendedover 20 years, in many instances the contact data in themedical records were no longer correct. In addition, ourinstitution is a quaternary care center, and patients re-ferred for the investigation and surgical management ofPA received their follow-up with their local primary carephysicians. Thus, we had only long-term follow-up data ofour patients, and this could be obtained only via telephonecontact by a trained medical professional. Hence, fol-low-up data were absent for those patients who could notbe contacted. A comparison of the patients with and with-out follow-up did not yield any statistically significantlydifferent preoperative characteristics other than a differ-ence in the body mass index (BMI) of 1.4 kg/m2, which issmall and likely of no clinical importance.

Our patients reported herein had well-documented PAand 70% required potassium supplementation, and only16% of the APAs were less than 1 cm in diameter. It isconceivable that, because of our quaternary care setting,our PA cohort overrepresents those with more severe dis-ease and larger APAs. Such a potential referral bias could



lead to an overestimation of the accuracy of the imagingprocedures. In addition, the number of patients who wereyounger than 35 years and operated on was six. Hence,only a small number of patients would have been sparedAVS in this study if an age cutoff combined with a uni-lateral mass on adrenal imaging was used.

In conclusion, the effective surgical cure rate for PA was95.5%. The postoperative PAC may be used to prognos-ticate the cure of PA. Although the overall accuracy of CTor MRI in detecting unilateral adrenal disease was poor at58.6%, adrenal imaging performed well in those patientsyounger than 35 years of age.

Acknowledgments

We thank Dr Ravinder Singh, PhD, director of the Mayo MedicalLaboratories Endocrine Laboratory and the Mayo Clinic Clin-ical and Translational Science Awards statistics consultants.

Address all correspondence and requests for reprints to:Vivien Lim, MBBS, Division of Endocrinology, Khoo TeckPuat Hospital, 90 Yishun Central, Singapore 768828. E-mail:[email protected].

Disclosure Summary: The authors have nothing to disclose.

References

1. Mulatero P, Stowasser M, Loh KC, et al. Increased diagnosis ofprimary aldosteronism, including surgically correctable forms, incenters from five continents. J Clin Endocrinol Metab 2004;89:1045–1050.

2. Rossi GP, Bernini G, Caliumi C, et al. PAPY Study Investigators. Aprospective study of the prevalence of primary aldosteronism in

1,125 hypertensive patients. J Am Coll Cardiol 2006;48:2293–2300.

3. Conn JW. Presidential address. I. Painting background. II. Primaryaldosteronism, a new clinical syndrome. J Lab Clin Med. 1955;45:3–17.

4. Rossi GP. A comprehensive review of the clinical aspects of primaryaldosteronism. Nat Rev Endocrinol. 2011;7:485–495.

5. Young WF Jr, Klee GG. Primary aldosteronism: diagnostic evalua-tion. Endocrinol Metab Clin North Am. 1988;7:367–395.

6. Kempers MJ, Lenders JW, van Outheusden L, et al. Systematic re-view: diagnostic procedures to differentiate unilateral from bilateraladrenal abnormality in primary aldosteronism. Ann Intern Med.2009;151:329–337.

7. Young WF, Stanson AW, Thompson GB, Grant CS, Farley DR, vanHeerden JA. Role for adrenal venous sampling in primary aldoste-ronism. Surgery. 2004;136:1227–1235.

8. Funder JW, Carey RM, Fardella C, et al. The Endocrine Society.Case detection, diagnosis, and treatment of patients with primaryaldosteronism: an endocrine society clinical practice guideline. J ClinEndocrinol Metab. 2008;93:3266–3281.

9. Young WF, Stanson AW. What are the keys to successful adrenalvenous sampling (AVS) in patients with primary aldosteronism?Clin Endocrinol (Oxf). 2009;70:14–17.

10. Harvey A, Pasieka JL, Kline G, So B. Modification of the protocolfor selective adrenal venous sampling results in both a significantincrease in the accuracy and necessity of the procedure in the man-agement of patients with primary hyperaldosteronism. Surgery.2012;152:643–651.

11. Guo T, Taylor RL, Singh RJ, Soldin SJ. Simultaneous determinationof 12 steroids by isotope dilution liquid chromatography-photo-spray ionization tandem mass spectrometry. Clin Chim Acta. 2006;372:76–82.

12. Murphy BEP. Clinical evaluation of urinary cortisol determinationsby competitive protein-binding radioassay. J Clin EndocrinolMetab. 1968;28:343–348.

13. Sywak M, Pasieka JL. Long-term follow-up and cost benefit of ad-renalectomy in patients with primary hyperaldosteronism. Br J Surg.2002;89:1587–1593.

14. Steichen O, Zinzindohoué F, Plouin PF, Amar L. Outcomes of ad-renalectomy in patients with unilateral primary aldosteronism: areview. Horm Metab Res. 2012;44:221–227.



mailto:[email protected]

Documents

Accuracy of Adrenal Imaging and Adrenal Venous Sampling in Predicting Surgical Cure of Primary Aldosteronism