Peripheral Plasma 18-Oxocortisol Can Discriminate Unilateral Adenoma from Bilateral Diseases in Primary Aldosteronism Patients

Abstract

Adrenal venous sampling is currently the only reliable method to distinguish unilateral from bilateral diseases in primary aldosteronism. In this study, we attempted to determine whether peripheral plasma levels of 18-oxocortisol and 18-hydroxycortisol could contribute to the clinical differentiation between aldosteronoma and bilateral hyperaldosteronism in 234 patients with primary aldosteronism, including CT-detectable aldosteronoma (n=113) and bilateral hyperaldosteronism (n=121), all of whom underwent CT and adrenal venous sampling. All aldosteronomas were surgically resected and the accuracy of diagnosis was clinically and histopathologically confirmed. 18-oxocortisol and 18-hydroxycortisol were measured using liquid chromatography tandem mass spectrometry. ROC analysis of 18-oxocortisol discrimination of adenoma from hyperplasia demonstrated sensitivity/specificity of 0.83/0.99 at a cutoff value of 4.7ng/dL, compared to that based upon 18-hydroxycortisol (sensitivity/specificity: 0.62/0.96). 18-oxocortisol levels above 6.1ng/dL and/or of aldosterone above 32.7ng/dL were found in 95 of 113 aldosteronoma patients (84%) but in none of 121 bilateral hyperaldosteronism, 30 of whom harbored CT-detectable unilateral nonfunctioning nodules in their adrenals. In addition, 18-oxocortisol levels below 1.2ng/dL, the lowest in aldosteronoma, were found 52 out of the 121 (43%) patients with bilateral hyperaldosteronism. Further analysis of 27 patients with CT-undetectable micro aldosteronomas revealed that eight of these 27 patients had CT-detectable contralateral adrenal nodules, the highest values of 18-oxocortisol and aldosterone were 4.8 and 24.5ng/dL, respectively, both below their cutoff levels indicated above. The peripheral plasma 18-oxocortisol concentrations served not only to differentiate aldosteronoma, but also could serve to avoid unnecessary surgery for nonfunctioning adrenocortical nodules concurrent with hyperplasia or microadenoma.

Primary aldosteronism (PA) is the most frequent form of secondary hypertension.^1–2 Patients with PA present cardiovascular and cerebrovascular complications more frequently than those with essential hypertension.^3–4 Therefore, appropriate diagnosis and treatment of PA has become important for individual patients. In addition, results of our recent study indicated that PA should be detected and treated as early as possible to prevent chronic kidney disease.⁵ However, it is also true that its final diagnosis requires relatively long procedures, such as detection, confirmation testing and subtype diagnosis comprising computed tomography (CT) scans and adrenal venous sampling (AVS)].^6–7 AVS has been clearly established as the only reliable method for differential diagnosis between surgically curable unilateral aldosterone-producing adenoma (APA) and bilateral hyperaldosteronism (BHA).^8–12 Nowaday, an increased number of specialized centers perform AVS in the world,^10,11 but this test is still time-consuming, labor-intensive and costly, which unfortunately prevents a wider application. Therefore, a much simpler, non-invasive and less expensive diagnostic method has been in demand by clinicians. 18-hydroxycortisol (18OHF) and 18-oxocortisol (18oxoF), which have structural characteristics of both aldosterone and cortisol in their chemical structures,¹³ are both synthesized from 11-deoxycortisol as substrate by aldosterone synthase (CYP11B2), although 18OHF can be also produced by 11-β-hydroxylase (CYP11B1).^14,15 Their production is known to be highly elevated in glucocorticoid remediable aldosteronism (GRA).^16,17 Furthermore, production of 18OHF and 18oxoF is elevated in patients with PA, especially in APA.¹³ The 24-hour urine excretion of these steroids is higher in PA patients than that in normal subjects and patients with essential hypertension (EH).^13.17–20 Urinary 18OHF is considered more reliable in terms of subclassification of PA than urinary 18oxoF provided that they have been accurately evaluated by enzyme-linked immunoassay.²⁰ We recently developed a highly sensitive test to measure 18oxoF based upon liquid chromatography tandem mass spectrometry (LC/MS/MS).²¹ In the present study, we examined the clinical significance of peripheral plasma concentrations of 18OHF and 18oxoF (p18OHF and p18oxoF) by LC/MS/MS in subtype classification of PA.

Methods

Patients and Imaging

From October 2010 to September 2013, 234 patients consisting of 113 patients with CT-detectable macro APA (APA) and 121 patients with BHA, underwent computed tomography scanning and successful cosyntropin-stimulated AVS. All of the 113 APA patients underwent adrenalectomy, and diagnostic accuracy was further confirmed by postoperative measurement of plasma aldosterone concentration (PAC) and histopathological evaluation, including immunohistochemical analysis of steroidogenic enzymes.^22–24 27 patients with CT-undetectable micro APA (microAPA) were also diagnosed by AVS and underwent unilateral adrenalectomy with their diagnosis being confirmed histopathologically after surgery. Our preliminary study revealed that p18oxoF and 18OHF of microAPA patients were as low as those of BHA patients (data not shown) and therefore, we evaluated mainly patients with APA and BHA in this study. The present study was approved by the ethics committee of Tohoku University School of Medicine (#2010-359 and #2010-360) and written informed consent was obtained from all participants. The patients with a plasma aldosterone concentration (PAC)/plasma renin activity (PRA) ratio (ARR) of >20 (ng/dLper ng/mL/h) after challenge with captopril 50mg were diagnosed with PA as previously reported.⁹ Patients were treated with calcium channel blockers and α₁-blockers during the PA workup. None of the 234 enrolled patients showed autonomous cortisol secretion which was confirmed by cortisol concentrations > 3.0 μg/dL following an overnight 1mg dexamethasone suppression test (1-mg DST). Among the 249 (234 plus 15) PA patients studied, 15 (6.0%) with 1-mg DST-positive cortisol concentrations (6.2 ± 0.54 μg/dL, mean ± SEM) were excluded from this study. The ratio of such PA patients who could harbor subclinical cortisol hypersecretion is considered to fall between 21% and 3.9% according to previously reported studies.^22–23 Blood pressure was measured with Omron Hem 907 (Omron Heathcare Co. Ltd, Kyoto, Japan) after >15-minutes rest in a sedentary position, and the average of 3 consecutive measurements was recorded. Peripheral blood samples were collected after the patient stayed in the recumbent position for 30 minutes between 8 and 10 am.

The imaging procedure for CT scanning and its detailed interpretation is described in “Supplemental Methods.”

Reagents and Measurement

Fusaric acid was obtained from Sigma-Aldrich (St. Louis, MO). 18oxoF was kindly provided by Dr. Gomez-Sanchez. 18-OHF was purchased from Steraloids Inc. (Newport, RI). Cortisol-²H₄ (F-d4) and aldosterone-²H₇ (Aldo-d7) were purchased from Isosciences (King of Prussia, PA) and C/D/N Isotopes (Pointe-Claire, Canada), respectively. InertSep Pharma cartridges and InertSep SI cartridges were obtained from GL Sciences (Tokyo, Japan). 4-Dimethylaminopyridine and 2-methyl-6-nitrobenzoic anhydride were purchased from Tokyo Chemical Industry (Tokyo). LC/MS/MS grade acetonitrile was obtained from Wako Pure Chemicals (Osaka, Japan). All other reagents and solvents were of analytical grade. PRA and PAC were measured using commercially available kits as previously reported.⁹ We used SPAC-S Aldosterone RIA Kit (TFB, INC; Japan) for measuring PAC in the present study. The intra-assay variability of this assay was 1.8–8.3%. This radioimmunoassay (RIA) of PAC had been standardized (Supplemental Figure S1, Spearman’s r = 0.9055, P < 0.0001) using 90 plasma samples with different aldosterone concentrations confirmed by liquid chromatography tandem mass spectrometry (LC-MS/MS), as we previously reported.²⁴ 18oxoF was measured by LC/MS/MS as previously reported.²¹ Sample preparation for LC-MS/MS measurement of 18OHF was described in “Supplemental Methods.”

The intraassay accuracy and precision were 90–111% and 3–9%, respectively, for both 18OHF and 18oxoF. Those of the interassay of their steroid metabolites were 87–108% and 1–10%, respectively. The lower limits of quantification for 18OHF and 18oxoF were 2.5 and 0.25 ng/dL, respectively.

Statistical analysis

Normality of the collected data was analyzed by Kolmogorov-Smirnov test, and variables between groups were analyzed by Kruskal-Wallis test with Dunn’s multiple comparison test as a post-hoc test. Receiver operating characteristic (ROC) analysis was performed to evaluate diagnostic ability. ROC curves were compared with the area under the curve (AUC). A linear regression model was employed to analyze the correlation between two numerical variables, and their correlation was evaluated by Spearman’s correlation coefficient. Statistical significance was set at P ≤ 0.05.

Results

Clinical characteristics, CT imaging and adrenal venous sampling in PA patients

We studied 113 patients with APA and 121 with BHA. As demonstrated in Table 1, baseline aldosterone concentration, baseline ARR and captopril-challenged ARR were all significantly higher in those with APA (46.6 ng/dL, 363 ng/dLper ng/mL/h and 233 ng/dL per ng/mL/h) than those with BHA (18.3 ng/dL, 87.1 ng/dL per ng/mL/h and 63.0ng/dL per ng/mL/h). CT scanning detected 15 APA cases with bilateral adrenal nodules and 30 BHA cases with unilateral non-functioning adrenocortical nodules, which also did confirm the superior diagnostic ability of AVS compared to imaging modalities (Table 1). Furthermore, AVS findings enabled us to diagnose 27 additional patients with microAPA (CT-undetectable) during the present study. Eight of them turned out to harbor unilateral nonfunctioning nodules in the contralateral adrenal. Thus, CT imaging findings were in agreement with those of AVS in 189 (i.e., 261-15-30-27) of 261 PA patients, that is a little more than 72% of the study population. This discriminatory value of only CT imaging was not so different from that found in previous studies.^8–12 We also performed ROC analysis to compare the discriminating ability of serum potassium between APA (CT-detectable) and BHA. Those with BHA were tentatively regarded as control and those with APA as the unilateral tumor group. The value of serum potassium had a significant discriminating ability with an AUC of 0.78 using a cutoff value of 3.85 mmol/L, associated with a sensitivity of 0.785 and specificity of 0.785. The serum potassium concentrations in microAPA patients were not significantly different from those in BHA patients.

TABLE 1.

Clinical Characteristics

	APA	BHA	P
Number	113	121
Age (years)	52.4 ± 1.1	54.0 ± 1.0	NS
Sex (M, F)	66, 47	40, 81	NS
Body mass index (kg/m2)	24.3 ± 0.43	24.9 ± 0.41	NS
SBP (mmHg)	163.3 ± 5.4	160.8 ± 7.0	NS
DBP (mmHg)	99.1 ± 6.3	97.2 ± 4.5	NS
Serum Na (mM)*	144.0 ± 0.21	142.3 ± 0.18	<0.05
Serum K (mM)*	3.4 ± 0.058	4.1 ± 0.033	<0.05
Proportion of diabetes (%)	11.5	11.6	NS
18-oxo-cortisol (ng/dl)*	23.6 ± 3.4	1.89 ± 0.14	<0.05
18-hydroxy-cortisol (ng/dl)*	357.1 ± 34.6	128.6 ± 7.3	<0.05
Aldosterone (ng/dl)*	46.6 ± 3.6	18.3 ± 0.60	<0.05
PRA (ng·ml−1·h−1)	0.17 ± 0.022	0.23 ± 0.021	NS
ARR (ng·dl−1 per ng·ml−1·h−1)*	363.3 ± 38.7	87.1 ± 4.50	<0.05
Cortisol after 1mg DST (μg·dL−1)	1.19 ± 0.059	0.94 ± 0.035	<0.05
Captopril-challenged ARR (ng·dl−1 per ng·ml−1·h−1) *	232.5 ± 33.4	63.0 ± 3.70	<0.05
CT-detected adrenal nodules (Uni, Bi, None)	98, 15, 0	30, 6, 85	NS
Size of CT-detected adrenal nodules (mm)	15.3 ± 2.9	19.6 ± 6.1	NS

Data are shown as mean ± SEM, unless otherwise stated. Proportion of diabetes shows percentage of patients with glycated hemoglobin of 6.5% or above. APA; aldosterone-producing adenoma, BHA; bilateral hyperaldosteronism SBP; systolic blood pressure, DBP; diastolic blood pressure, PRA; plasma renin activity, ARR; aldosterone over renin activity ratio, Uni; unilateral, Bi; bilateral, DST ; dexamethasone suppression test, NS; not significant,

^*P<0.05

Peripheral levels of 18-oxocortisol and 18-hydroxycortisol

Both p18oxoF and p18OHF were significantly elevated in patients with APA (23.6 ng/dL and 357 ng/dL) compared to those with BHA (1.89 ng/dl and 129 ng/dL) (Table 1). When comparing the averaged peripheral plasma concentrations of these two steroids between APA and BHA, the APA/BHA ratio of p18oxoF (12.5) was five times higher than that of p18OHF (2.77). Comparison between PAC and p18oxoF levels subsequently demonstrated a statistically significant correlation between the above two groups, and those with APA showed the most marked correlation (Spearman’s r = 0.5336, P <0.05) compared to those with BHA (Spearman’s r = 0.1987, P <0.05) (Supplemental Figure 2SA and 2SB). In addition, linear regression analysis between peripheral aldosterone and 18oxoF levels also revealed the model was most fitted in those with APA (R² = 0.6488) compared to those with BHA (R² = 0.0367) (Supplemental Fig S2A and S2B). In contrast, comparison between peripheral PAC and p18OHF demonstrated a significant correlation in the APA group (Spearman’s r = 0.4886) (Supplemental Figure S2C) but not in the BHA group (Supplemental Figure S2D).

ROC analyses using p18oxoF, p18OHF, aldosterone and ARR

ROC analyses were performed to compare the diagnostic abilities of p18oxoF, p18OHF, PAC and ARR (Figure 1A, 1B, 1C and 1D) in terms of differentiation between unilateral neoplastic lesions and BHA. Those with BHA were regarded as control and those with APA as a unilateral tumor group. The value of p18oxoF was demonstrated to have the highest diagnostic ability with an AUC of 0.956 at a cutoff value of 4.7ng/dL showing a sensitivity of 0.83 and specificity of 0.99 (Figure 1A). The value of PAC had the second highest ability with an AUC of 0.917 based upon a cutoff level of 21.5 ng/dL showing a sensitivity of 0.81 and specificity of 0.93 (Figure 1C). Similarly, the value of p18OHF was third and that of ARR was fourth, with the AUC areas being 0.85 and 0.82 at cutoff values of 234ng/dL and 152ng/dL per ng/mL/h and showing a sensitivity of 0.62 and 0.67 and specificity of 0.96 and 0.90, respectively (Figure 1B and 1D).

Figure 1.

Receiver operating characteristic (ROC) analysis of patients with APA compared to those with BHA as control and distribution plot analysis. A, B, C and D depict ROC curves to analyze the diagnostic value of respectively peripheral 18-oxocortisol, 18-hydroxycortisol, aldosterone and aldosterone-renin activity ratio, to discriminate APA from BHA. E, F, G and H show respectively the distribution of peripheral 18-oxo-cortisol, 18-hydroxycortisol, aldosterone and aldosterone-renin activity ratio in APA and BHA. 18oxoF, 18-oxocortisol ; 18OHF, 18-hydroxycortisol ; ARR, aldosterone-over-renin activity ratio; AUC, area under the curve; APA, aldosterone-producing adenoma; BHA, bilateral hyperaldosteronism.

The distribution of p18oxoF showed that the highest level in the BHA group was 6.1 ng/dL, and the minimal value of p18oxoF in the APA group was 1.2 ng/dL (Figure 1E). The highest levels of p18OHF, PAC and ARR in the BHA group were 345 ng/dL, 32.7 ng/dL and 254 ng/dL per ng/mL/h, respectively, and their minimal values were 5.7 ng/dL, 8.6 ng/dL and 20 ng/dL per ng/mL/h, respectively (Figure 1F, 1G and 1H). In 86 of the 113 APA patients (76%), p18oxoF concentration was above the maximum level (6.1 ng/dL) observed in BHA patients. In 61 (54%), 56 (50%), and 47 (42%) of APA patients, PAC, ARR and p18OHF were higher than the maximum levels observed in BHA patients. Therefore, p18oxoF was considered to be of the highest diagnostic value for PA subtyping. Furthermore, we examined whether p18oxoF concentration served to differentiate 52 APA patients whose PAC were equal to or below 32.7 ng/dL, which was the maximum level in BHA patients. Of particular interest, 34 (65%) of these 52 APA patients who had p18oxoF concentrations above 6.1 ng/dL were considered to require surgery (Figure 2). Therefore, 95 (84%) of the 113 APA patients could be differentiated by the values of PAC and p18oxoF.

Figure 2.

Distribution plot of patients with PAC ≤32.7 ng/dL subdivided into three groups depending on 18-oxo-cortisol levels. Those of groups 1, 2 and 3 (G1, G2 and G3) are considered to require adrenal CT scan (and AVS when necessary) with subsequent adrenalectomy, CT and AVS to determine disease laterality, and pharmacological treatment, respectively. 18oxoF, 18-oxo-cortisol; APA, aldosterone-producing adenoma; BHA, bilateral hyperaldosteronism.

Moreover, we measured p18oxoF in patients with essential hypertension (EH) patients by excluding PA and any other cause of secondary hypertension. The p18oxoF (1.3 ± 0.2 ng/dL) in 79 EH patients was significantly lower than that in APA patients but not significantly different from that in BHA patients. The highest concentration of p18oxoF in EH patients was 5.6 ng/dL, which was lower than the highest value in BHA patients (Supplemental Figure S3). Interestingly, there were 52 (43%) of 121 BHA patients (Figure 2) and 52 (66%) of 79 EH patients (Supplemental Figure S3) whose p18oxoF was less than 1.2ng/dL, the lowest 18oxoF concentration in APA patients.

In addition, we measured p18oxoF and p18OHF in 27 patients with microAPA who underwent adrenalectomy based on their AVS findings and who were clinically and histopathologically confirmed. The levels of both p18oxoF (1.80 ± 0.26 ng/dL) and p18OHF (147 ± 21.4 ng/dL) in these 27 microAPA patients were significantly lower than those patients with CT-detectable APA but not significantly different from those in BHA patients. The highest levels of p18oxoF, p18OHF, PAC and ARR in microAPA patients were 4.8 ng/dL, 483 ng/dL, 24.5 ng/dL and 245 ng/dL per ng/mL/h, respectively, and their distributions were almost overlapped with those in BHA patients (Supplemental Figure S4). None of these patients with microAPA with or without contralateral unilateral nonfunctioning adrenocortical nodules had a p18oxoF level higher than 6.1 ng/dL, which has been considered as the cutoff value to recommend adrenalectomy.

Furthermore, we added analyses regarding the relationship among p18oxoF, AVS lateralization index (LI) and somatic KCNJ5 mutations, and we summarized these results in “Supplemental results.”

Discussion

In the present study, we measured p18oxoF and p18OHF in 261 (234 plus 27) PA patients, who were subsequently classified based upon the results of AVS into 113 APA, 121 BHA and 27 with microAPA. In all of the 113 APA and 27 microAPA patients who underwent adrenalectomy, the diagnosis was further confirmed postoperatively based on PAC and immunohistochemical analysis of steroidogenic enzymes in resected adenomas.^25–27 We found that p18oxoF was a more reliable diagnostic parameter to differentiate between APA and BHA than p18OHF. The possibility of measuring 18oxoF by LC/MS/MS (lower limit: 0.25 ng/dL) has enabled us to detect a very low concentration in peripheral plasma,²¹ and to differentiate APA from BHA (or EH) without the need to have patients to collect the 24-hour urine. Thus, this is the first study to demonstrate that the peripheral blood concentrations of 18oxoF measured by LC/MS/MS can clinically differentiate between APA and BHA with a reasonable precision. In particular, 84% of APA patients had a p18oxoF concentration above 6.1 ng/dL and/or PAC of more than 32.7 ng/dL. In this study, APA patients represented only those with a CT-detectable aldosterenoma and did not include any of the microAPA patients. Moreover, this accuracy rate of 84% was obtained based on p18oxoF and CT findings. This is the primary reason why this accuracy rate of 84% was much higher than that found in previous studies.^8–12 The discriminatory value of CT imaging alone was 72% as described above, which was not so different from that obtained in previous studies.^8–12 Besides, in 43% of BHA patients, p18oxoF was below 1.2ng/dL, which was the lowest value found in APA patients. In addition, further analysis of the 27 microAPA patients revealed that eight (30%) of them harbored contralateral adrenocortical nodules on CT images and the highest p18oxoF and PAC were 4.8 and 24.5 ng/dL, respectively. These values were significantly lower than their respective cutoff values of 6.1 and 32.7 ng/dL. Also, 30 of 121 BHA patients had CT-detectable unilateral nonfunctioning adrenocortical nodules. Consequently, the evaluation of p18oxoF in such patients might contribute to avoid an unnecessary surgery in an institution where AVS is not available for preoperative routine evaluation of PA patients with unilateral adrenal nodules.

Based upon the results of the present study, we recommend adding p18oxoF to the diagnostic workup for PA (Figure 3). Throughout this systematic clinical workup, we might have been able to limit the use of AVS in 87 PA patients whose p18oxoF was between 1.2 and 6.1 ng/dL, a range in which APA and BHA patients were found to overlap after selecting the patients highly suspected of having APA and those with BHA (Figure 2 and Figure 3).

Figure 3.

Diagnostic flow model based upon peripheral aldosterone and 18-oxo-cortisol levels. Those with confirmed PA and both an aldosterone level >32.7 ng/dL and a unilateral nodule by CT undergo adrenalectomy without AVS. Those with an aldosterone level ≤32.7 ng/dL are divided into three subgroups depending on their peripheral 18-oxocortisol level to determine subsequent diagnostic workup steps. PA, primary aldosteronism; PAC, plasma aldosterone concentration; p18oxoF, peripheral 18-oxocortisol; CT, computed tomography; PCR, polymerase chain reaction; GRA, glucocorticoid-remediable aldosteronism; AVS, adrenal venous sampling; ADX, adrenalectomy; MRA, mineralocorticoid receptor antagonist.

In this study, the LC-MS/MS intraassay accuracy and precision were 90–111% and 3–9% for 18OHF and 18oxoF, respectively. Those of interassay were 87–108% and 1–10% for 18OHF and 18oxoF, respectively. The lower limits of quantification of 18OHF and 18oxoF were 2.5 and 0.25 ng/dL, respectively. The 3–9% precision of LC-MS/MS indicated therefore that this method was considered as a reliable mode of measurement in this setting. Further to this, the cutoff value was higher than the lower limit of quantification. Thus, we believe that precision of the measurement method was demonstrated by the results obtained. Likewise, the sensitivity and the specificity of the cutoff value were influenced by the precision of the measurement method. Nevertheless, it is entirely true that the 3–9% precision of this method indicated that the clinical diagnosis of AP should be carefully performed in patients whose 18O HF and 18oxoF values are around the cutoff value.

18-hydroxycorticosterone (18OHB) was also reported to help discriminate APA from BHA, especially after posture test.²⁸ However, Mulatero et al. reported in 2012 that basal levels of 18OHB did not help differentiate PA subtypes in their study²⁰ and since we wanted to find surrogate markers to discriminate PA subtypes by their basal peripheral blood levels, we selected 18OHF and 18oxoF. Because 18oxoF production was reported to be small but could be localized with CYP11B2, we thought that if we developed a very sensitive LC-MS/MS method, we could expect to measure even basal p18oxoF and be able to differentiate APA from BHA. In contrast, 18OHF can be produced by both CYP11B2 and CYP11B1 in the zona glomerulosa and the zona fasciculata, respectively, resulting in higher secretion of 18OHF compared to 18oxoF.¹⁵ In the present study, p18OHF was also higher than p18oxoF. However, the ratio of p18OHF between APA and BHA was merely one fifth that of p18oxoF, indicating that the sensitivity of differentiation by p18OHF (0.62) is lower than that by p18oxoF (0.83). Specificity of p18OHF (0.96) regarding subtyping was reasonably high the same as that by p18oxoF (0.99), and therefore, p18OHF can also be a reliable diagnostic tool to discriminate APA from BHA.

A major drawback of measuring p18oxoF by LC/MS/MS is the cost but this (approximately 150 U.S. dollars per one sample) is far less expensive than that of AVS, which amounts to approximately 10,000 U.S. dollars in the U.S.A. Moreover, we propose a diagnostic strategy for APA by precisely measuring p18oxoF after selecting APA patients according to PAC from among confirmed PA patients (Figure 3). Through this approach, we should be able to reduce the medical cost by not only omitting an unnecessary AVS procedure, but also by limiting p18oxoF measurement. Therefore, we could reconfirm the usefulness of this strategy, which might have enabled 95 (i.e., 61 plus 34) of the 113 APA patients (84%) to undergo surgery following the detection of unilateral adrenal nodules by CT scanning without subjecting them to AVS, and 52 of the 121 BHA patients (43%) (p18oxoF <1.2 ng/dL) to receive pharmacological treatment with mineralocorticoid receptor antagonists without performing AVS. Furthermore, the highest concentration of p18oxoF in EH patients was 5.6 ng/dL (<6.1 ng/dL), and in 52 (66%) of 79 EH patients p18oxoF was less than 1.2ng/dL, the lowest p18oxoF in APA patients. Therefore, these results indicated that higher or lower levels of 18oxoF could serve to predict the presence or absence APA without further confirmatory evaluation in the patients with high ARR, but a further larger study will be required to confirm this interesting hypothesis.

Perspectives

The results of this large study revealed that measurement of 18oxoF in peripheral blood by LC/MS/MS should enable us to improve the clinical algorithm for PA, potentially omitting unnecessary AVS, which is a very expensive and invasive procedure, and requires the dedicated work of experienced radiologists. Yet, AVS is still the only reliable method for determining the cause of hyperaldosteronism at specialized institutions where skillful radiologists make efforts to perform successful catheterizations.^11,29 In particular, AVS is the only way to distinguish a CT-undetectable microAPA from BHA,^8,30–32 even though no standardization of protocols has been performed yet.^11,33 In this study, p18oxF levels in microAPA patients were completely overlapped with those in BHA patients, and AVS is therefore considered critical in differentiating these two subtypes. This was a monocentric, retrospective study in Japanese patients. A prospective, multi centric study in centers treating different ethnicities is necessary to confirm the results.

Novelty and Significance.

What is New?

This is the first study to demonstrate that the peripheral blood concentration of 18-oxocortisol measured by LC/MS/MS can be a clinically significant discriminating marker between aldosterone-producing adenoma and bilateral hyperaldosteronism.

What is relevant?

This much simpler, non-invasive and less expensive measurement of 18-oxocortisol by LC/MS/MS might allow us to assign aldosteronoma patients for surgery without undergoing adrenal venous sampling.

Summary

Discrimination by the peripheral plasma level of 18-oxocortisol > 6.1 ng/dL and/or aldosterone > 32.7 ng/dL allowed classification of 95 of 113 (84%) aldosteronoma patients, and 18-oxocortisol < 1.2 ng/dL allowed differentiation of 52 of 121 (43%) patients with bilateral hyperaldosteronism. This non-invasive measurement of peripheral plasma 18-oxocortisol concentration by LC/MS/MS might contribute to further subclassification of primary aldosteronism.