Influence of autonomous cortisol secretion in patients with primary aldosteronism: subtype analysis and postoperative outcome

Abstract

Context

Autonomous cortisol secretion (ACS) has a relatively high prevalence in patients with primary aldosteronism (PA). There is still a lack of relevant studies to analyze the influence of ACS on diagnosing and managing PA.

Objective

To evaluate the influence of ACS on image–adrenal venous sampling (AVS) correlation and the postoperative results.

Methods

This was a retrospective study using the Taiwan Primary Aldosteronism Investigation database from July 2017 to April 2020, with 327 PA patients enrolled. A total of 246 patients were included in the image–AVS analysis. Patients who had undergone unilateral adrenalectomy and a 12-month follow-up were included in the postoperative analysis.

Results

Sixty-five patients (26.4%) had ACS. The image–AVS discordance rate was higher in the ACS group compared to the non-ACS group (75.4% (n = 49) vs 56.4% (n = 102); odds ratio (OR) = 2.37 (CI: 1.26–4.48); P = 0.007). The complete biochemical success rate was higher in the non-ACS group than that in the ACS group (98.1% (n = 51) vs 64.3% (n = 9); OR = 28.333 (CI: 2.954–271.779); P = 0.001). In logistic regression analysis, ACS was the only factor associated with lower biochemical success (OR = 0.035 (CI: 0.004–0.339), P = 0.004).

Conclusion

PA patients with ACS have higher image–AVS discordance rate and worse biochemical outcomes after surgery. ACS was the only negative predictor of postoperative biochemical outcomes. Further studies and novel biomarkers for AVS are crucial for obtaining better postoperative outcomes in PA patients with ACS.

Introduction

Primary aldosteronism (PA) is a disease characterized by excess aldosterone production of the adrenal glands and is considered the most common form of mineralocorticoid excess (1). It is also the most common cause of secondary hypertension, accounting for about 5–15% of patients with newly diagnosed hypertension (2, 3, 4).

PA can be generally subtyped into unilateral disease (including primarily aldosterone-producing adenomas (APAs), and unilateral adrenal hyperplasia in rare cases) and bilateral disease (mainly bilateral hyperplasia) (5, 6). Adrenal venous sampling (AVS) has been regarded as the standard for differentiating between the two disease subtypes (3, 7, 8, 9), in contrast to cross-sectional imaging studies (computed tomography (CT) or magnetic resonance imaging (MRI)). Adrenalectomy has consequently been established as the treatment of choice for unilateral disease.

Recently, there has been a growing focus on a special group with cortisol co-secretion (10) or autonomous cortisol secretion (ACS) (11). The prevalence of ACS is approximately 20% in patients with PA (12, 13). The clinical and laboratory manifestations present as a spectrum from very low cortisol secretion, to subclinical Cushing syndrome, to overt Cushing syndrome (14). Recent studies have shown a significantly increased cardiovascular risk and impaired glucose metabolism in these patients (15, 16). However, hypercortisolism status may be overlooked since most patients do not present with typical signs and symptoms of classic Cushing syndrome. Another significant concern has been raised regarding the potential influence of elevated cortisol levels in these patients, which may alter the lateralization of AVS and increase the possibility of image–AVS disagreement (14, 17, 18). Furthermore, the postoperative clinical and biochemical outcomes associated with these patients have not been well established.

In this study, we focused on how the presence of ACS in PA patients may influence clinical presentation by affecting the concordance between imaging and AVS evaluation methods. Postoperative clinical and biochemical resolution status was also analyzed within different subgroups.

Methods

Study population

This retrospective study was reviewed and approved by the institutional review board of the National Taiwan University Hospital.

From the Taiwan Primary Aldosteronism Investigation (TAIPAI) database, adult (age over 18 years) patients diagnosed with PA who underwent AVS at the National Taiwan University Hospital between July 2017 and April 2020 were enrolled. Clinical characteristics, laboratory data, imaging studies, and postoperative records were carefully reviewed from the clinical database. The diagnosis of PA was confirmed by the following three criteria: i) elevated aldosterone-to-renin ratio (ARR) > 35 (ng/dL)/(ng/mL/h), ii) TAIPAI score > 60% (19), and iii) nonsuppressed aldosterone with a plasma aldosterone concentration (PAC) > 16 (ng/dL) after saline administration (in seated posture) or ARR > 35 (ng/dL)/(ng/mL/h) after captopril administration (20). The TAIPAI score is a probability score developed by Wu et al. to assess the likelihood of PA (19). The equation for calculating the TAIPAI score incorporates various factors, including PAC, plasma renin activity, ARR, body mass index, sex, serum potassium level, and estimated glomerular filtration rate. Patients with clinically overt Cushing syndrome were excluded from the study. None of the included patients received any kind of glucocorticoid medication before testing.

Cross-sectional imaging studies

Before AVS, all patients underwent imaging examination, including CT or MRI. All images were reviewed and confirmed by two radiologists with 15 and 5 years of experience, respectively. An abnormal adrenal gland was identified by nodular thickening, bulkiness, or the presence of any nodular lesion (21). The imaging results were then divided into three groups: bilateral normal appearance, unilateral abnormality, and bilateral abnormalities.

AVS

All AVS procedures were performed by an interventional radiologist with 15 years of experience and in the morning to avoid bias resulting from the circadian rhythms of aldosterone secretion (22). Selectivity index (SI) was defined as the ratio of cortisol in the adrenal vein to that in the inferior vena cava (C_AV/C_IVC). Successful cannulation of the adrenal vein was defined as SI ≥ 2.0 in nonstimulated AVS (8). Patients with cannulation failure on either side were excluded. The lateralization index (LI) was calculated as the aldosterone–cortisol concentration ratio between the dominant and nondominant adrenal veins ((A/C)_{dominant AV}/(A/C)_{nondominant AV}). The lateralization of PA was confirmed by an LI value ≥ 2.0 (20, 23, 24).

Screening for ACS

The 1 mg overnight dexamethasone suppression test (DST) was performed to obtain the hypercortisolism status of the patients. Patients who failed to suppress cortisol secretion (>1.8 µg/dL, (50 nmol/L)) were stratified into the ACS group; otherwise, they were placed into the non-ACS group (13, 25, 26). Patients who did not undergo DST were excluded.

Laboratory measurement

The serum cortisol concentration was measured by a chemiluminescent microparticle immunoassay (Architect, Abbott, VA, USA). Plasma aldosterone levels were measured by radioimmunoassay kits (ALDO-RIACT; Cisbio Bioassays, Codolet, France). The protocols and procedures were performed in accordance with the manufacturer’s instructions.

Discordance and concordance between image and AVS

Discordance of image and AVS results was defined by the presence of bilateral normal adrenal glands on imaging with lateralization on AVS, bilateral abnormal appearance of adrenal glands with lateralization on AVS, or unilateral lesion on imaging with either nonlateralization or contralateral lateralization on AVS. Concordance was defined as a unilateral lesion on imaging with corresponding lateralization on AVS, bilateral adrenal gland abnormality with nonlateralization on AVS, or bilateral normal adrenal imaging with nonlateralization on AVS.

Postoperative analysis

Among patients who met the inclusion criteria, those who underwent unilateral adrenalectomy were included in the postoperative analysis. Patients who had undergone partial adrenalectomy or bilateral adrenal surgery were excluded. The AVS-based group was defined as the adrenalectomy side in agreement with AVS lateralization (no matter what abnormality was shown in imaging studies). The image-based group was defined as the adrenalectomy side determined based on imaging findings (CT/MRI) of unilateral adrenal lesions, regardless of AVS results. The determination of unilateral PA in these patients receiving unilateral adrenalectomy was based on serial board discussions, in which the imaging and biochemical data, including unilateral adrenal lesion, hypokalemia, marked elevation of ARR, and young age (<35 years), were considered (27, 28, 29).

Postoperative clinical and laboratory results were collected from the electronic medical records of the TAIPAI database. Postoperative analysis of clinical or biochemical success was performed in accordance with the International Primary Aldosteronism Surgical Outcome (PASO) international consensus (29). Complete clinical and biochemical success rates were compared between the different groups.

Statistical analysis

All statistical analyses were performed using the IBM SPSS statistical software 23.0 for Windows. Demographic and biochemical data are presented as mean (standard deviation) for normally distributed data, median (interquartile range) for non-normally distributed data, or number (percentage). The independent Student’s t-test or Mann–Whitney U test were used for continuous variables, as appropriate, and the Χ ² test was applied to the categorical variables. Odds ratios (ORs) and 95% CIs were also calculated. We compared the demographic characteristics, AVS lateralization, imaging studies, and image–AVS concordance rates between the ACS and non-ACS groups. Postoperative results of the two groups for complete clinical and biochemical success were examined via Χ ² test, and logistic regression models. Statistical significance was set at P < 0.05.

Results

Demographics

In total, 327 patients were initially enrolled in this study (Fig. 1). Six patients were excluded because of overt Cushing syndrome. Another six patients were excluded because of failed cannulation of either one-sided or bilateral adrenal veins. Sixty-nine patients who did not undergo DST were also excluded. Of the 246 patients remaining, 65 (26.4%) failed to suppress cortisol secretion (>1.8 µg/dL) after a 1 mg overnight DST, indicating ACS (13). The relevant demographics, clinical conditions, biochemical parameters, AVS results, and image analysis are shown in Table 1 (for additional details see the Supplementary Materials, see section on supplementary materials given at the end of this article).

Figure 1.

Flowchart of patient selection and exclusion criteria. DST, dexamethasone suppression test; PA, primary aldosteronism; TAIPAI, Taiwan Primary Aldosteronism Investigation.

Table 1.

Baseline demographics and characteristics of the included patients, including the postoperative results.

	n (%)		P
	ACS (n = 65)	Non-ACS (n = 181)
Age (years)	56.7 ± 10.4	52.6 ± 11.1	0.48
Sex
Female (%)	41 (63.1%)	87 (48.1%)	0.04
Male (%)	24 (36.9%)	94 (51.9%)
BMI	25.7 ± 4.5	27.0 ± 5.5	0.38
BP at diagnosis
SBP (mmHg)	159 ± 26	157 ± 23	0.26
DBP (mmHg)	92 ± 13	96 ± 15	0.32
Biochemistry
Serum potassium (mmol/L)	3.6 ± 0.5	3.6 ± 0.6	0.35
Hypokalemia (<3.5 mmol/L)	23 (35.4%)	66 (36.5%)	0.88
Plasma aldosterone (ng/dL)	37.1 ± 25.4	37.1 ± 27.8	0.73
ARR	285.3 ± 451.3	239.5 ± 515.2	0.73
Cortisol post DST (μg/dL)	4.9 ± 3.5	1.0 ± 0.4	<0.001
AVS lateralization
Lateralized	27 (41.5%)	74 (40.9%)	0.93
Nonlateralized	38 (58.5%)	107 (59.1%)
Image
Bilateral normality	3 (4.6%)	27 (14.9%)	0.006
Single side abnormality	41 (63.1%)	124 (68.5%)
Bilateral abnormalities	21 (32.3%)	30 (16.6%)
Postoperative clinical resolution (n = 69)
Complete	7 (50.0%)	30 (54.5%)	0.52
Partial	5 (35.7%)	22 (40.0%)
Absent	2 (14.3%)	3 (5.5%)
Postoperative biochemical resolution (n = 66)
Complete	9 (64.3%)	51 (98.1%)	0.001
Partial	2 (14.3%)	0 (0%)
Absent	3 (21.4%)	1 (1.9%)
Pathology (n = 84)
Adenoma	19 (90.5%)	44 (69.8%)	0.081
Hyperplasia	2 (9.5%)	19 (30.2%)

ACS, autonomous cortisol secretion; ARR, aldosterone-to-renin ratio; AVS, adrenal venous sampling; BMI, body mass index; BP, blood pressure; DBP, diastolic blood pressure; DST, dexamethasone suppression test; SBP, systolic blood pressure.

The ACS group consisted of more female patients than the non-ACS group (P = 0.04). There were no significant differences in blood pressure, serum potassium level, PAC, and ARR. Cortisol levels after DST were significantly higher in the ACS group (P < 0.001).

Image–AVS concordance

AVS lateralization did not differ between the groups (41.5 vs 40.9%). Most of the patients underwent CT, while only five patients underwent MRI. The ACS group showed a higher percentage of bilateral abnormalities and fewer bilateral normal-appearing adrenal glands than the non-ACS group (32.3 vs 16.6%, P = 0.012 and 4.6 vs 14.9%, P = 0.028, respectively). The comparison of discordance and concordance of the imaging and AVS results between the two groups is shown in Fig. 2. The image–AVS discordance rate was significantly higher in the ACS group than in the non-ACS group ((75.4% (n = 49)) vs 56.4% (n = 102); OR = 2.37 (CI: 1.26–4.48); P = 0.007).

Figure 2.

Comparison of image–adrenal venous sampling (AVS) concordance rates between autonomous cortisol secretion (ACS) and non-ACS groups.

Analysis of the postoperative outcomes

Among the patients who met the inclusion criteria, 111 (45.1%) underwent adrenalectomy, and of those, 101 (91%) underwent unilateral total adrenalectomy. Ten patients (9%) who underwent partial adrenalectomy or bilateral adrenal surgery were excluded from postoperative analysis. Sixty-nine patients (68.3%) had a 1-year follow-up record for clinical outcome evaluation (with 20.6 months mean follow-up duration); sixty-six patients (65.3%) had complete records for biochemical outcome evaluation (with 21.0 months mean follow-up duration).

The international consensus from the PASO study was applied for the postoperative evaluation (29). Clinical, biochemical, and pathological results are shown at the bottom of Table 1. Complete clinical success did not vary between the two groups (50.0% (n = 7) vs 54.5% (n = 30); P = 0.52). Complete biochemical success was attained in 64.3% (n = 9) of patients in the ACS group, which was significantly lower than that in the non-ACS group (98.1% (n = 51)) (OR = 28.333 (CI: 2.954–271.779); P = 0.001) (Fig. 3A).

Figure 3.

Comparison of postoperative clinical and biochemical outcomes between autonomous cortisol secretion (ACS) and non-ACS groups. (A) Outcome of total patient group. (B) Outcome of adrenal venous sampling (AVS)-based unilateral adrenalectomy subgroup. (C) Outcome of image-based unilateral adrenalectomy subgroup.

In single logistic regression analyses, the presence of ACS was the only significant predictor of postoperative complete biochemical success (OR = 0.035 (CI: 0.004–0.339), P = 0.004) (Table 2). No associated factors were found to predict complete postoperative clinical success in our study.

Table 2.

Factors associated with the complete biochemical success after unilateral adrenalectomy.

Characteristics	Univariate
	Odds ratio (95% CI)	P-value
Sex (female)	0.382 (0.065–2.25)	0.288
BMI	0.919 (0.774–1.091)	0.334
SBP	1.011 (0.971–1.052)	0.604
DBP	1.009 (0.946–1.076)	0.784
Hypokalemia (<3.5 mmol/L)	0.244 (0.027–2.22)	0.211
ACS (DST > 1.8 μg/dL)	0.035 (0.004–0.339)	0.004
Lateralized by AVS	2.444 (0.416–14.379)	0.323
Lesion size note on image	1.002 (0.901–1.113)	0.977
Unilateral image abnormality	0.308 (0.048–1.964)	0.213
Bilateral image abnormality	4.5 (0.676–29.948)	0.12

ACS, autonomous cortisol secretion; AVS, adrenal venous sampling; BMI, body mass index; DBP, diastolic blood pressure; DST, dexamethasone suppression test; SBP, systolic blood pressure. Bold indicates statistical significance, P < 0.05.

Patients were divided into two major groups: adrenalectomy based on the image or based on the AVS results. Subgroup analysis of adrenalectomy based on the AVS results is shown in Fig. 3B. All patients in the non-ACS group achieved complete biochemical success (100% (n = 30)), which was significantly higher than that in the ACS group (60% (n = 5), P = 0.017). Among the image-based adrenalectomies, there was a trend of higher complete biochemical results in the non-ACS group, but this was not statistically significant (95.5% (n = 22) vs 66.7% (n = 9); P = 0.063) (Fig. 3C). Complete clinical success in both subgroup analyses did not show a significant difference. In all patients, the postoperative clinical and biochemical outcomes showed no significant differences between the image-based and AVS-based adrenalectomy subgroups (P = 1 and P = 0.41, respectively).

Discussion

In this study, we focused on the influence of ACS in PA patients, including the AVS-image discordance and postoperative outcomes, and revealed unsatisfactory postoperative biochemical outcomes in PA patients with ACS. The ACS rate in our patients was 26.4%, which was within the range reported in previous studies (4–27%) (10, 13, 30, 31, 32). Female patients had a higher ACS prevalence than males (Table 1), consistent with prior research (13, 14, 33). The ACS group had fewer bilateral normal adrenal glands and more bilateral abnormal ones in imaging (4.6 vs 14.4% and 32.3 vs 16.6%, respectively). This aligns with prior studies associating bilateral incidentalomas with a higher ACS likelihood (34). This may be due to the heterogeneity of both aldosterone- and glucocorticoid-producing lesions (35).

A major concern in ACS is that excess cortisol secretion may impact the aldosterone–cortisol ratio and further alter the image–AVS relationship. In the general population of PA patients, image–AVS discordance rates have been reported to range from 37.8% (36) to as high as 67% (37). In our study, the total discordance rate was 61.4% (151/246), with a significantly higher discordance rate in the ACS group than in the non-ACS group (Fig. 2; 75.4 vs 56.4%, P = 0.007). Kline et al. also developed an algorithm for detecting possible ACS in patients without confirmation by DST and stated that the cortisol co-secretion condition was observed in 40% of patients with CT-AVS discordance (14). Thus, the reconsideration of cortisol as a critical parameter in calculating LI has emerged. Ceolotto et al. proposed the use of different laboratory measurements, including androstenedione, metanephrine, and normetanephrine as substitute biomarkers for AVS parameters, and obtained a more sensitive SI than when cortisol was used (38). These biomarkers may also eliminate the effect of cortisol on LI in ACS patients and provide more accurate lateralization. However, further verification is required to confirm this hypothesis.

As for the postoperative analysis, our study further revealed the inferiority of prognosis in the ACS group, especially considering the biochemical results. Without regarding ACS, our study exhibited a slightly higher complete clinical success rate (54 vs 37%) and a slightly lower complete biochemical success rate (91 vs 94%) in comparison to the PASO study (29). However, our results showed that the presence of ACS led to a significantly lower complete biochemical success rate, even after further division into AVS-based or image-based subgroups (Fig. 3). Furthermore, univariate logistic regression analyses showed that among the different factors, ACS was the only negative predictor of biochemical outcome (Table 2). Peng et al. also focused on the prognosis after adrenalectomy in PA patients with ACS (39). They summarized that the presence of hypercortisolism (which was defined as DST > 1.5 µg/dL) had a lower complete clinical success rate. Their biochemical results did not differ significantly. A possible reason for this discordance between our study and theirs might be the different patient groups. Peng et al. focused only on APA patients (39), whereas we did not exclude patients with adrenal hyperplasia. Hacini et al. revealed that patients with bilateral hyperplasia may show asymmetric aldosterone production, resulting in an absent biochemical cure (40). Another possible reason is the difference in the DST criteria applied; we used 1.8 µg/dL as the DST cutoff value in line with other studies (13, 32, 33), while they used 1.5 µg/dL.

The ACS group showed increased image–AVS discordance and a lower biochemical success rate after unilateral adrenalectomy. Choosing adrenalectomy is complex when images suggest a unilateral adrenal lesion, but AVS points to lateralization on the opposite side. This raises the question of whether to rely on imaging or AVS for adrenalectomy decisions.

Recent studies have shown that the clinical and biochemical success of adrenalectomy based on imaging studies alone did not differ from that of adrenalectomy based on AVS in PA patients, especially in younger age groups (23, 41). Our results were similar. Irrespective of the ACS and non-ACS groups, the postoperative outcomes did not differ between the image-based adrenalectomy and AVS-based adrenalectomy subgroups. A randomized controlled trial specifically targeting ACS patients may be necessary to establish a more definitive conclusion.

Our study has some limitations. The total number of ACS patients is still insufficient to support that surgical decisions should be determined by AVS or imaging study. More studies with larger sample sizes are warranted to confirm this result, especially in ACS patients. Another limitation is that we did not investigate the urinary free cortisol, adrenocorticotropic hormone, and dehydroepiandrosterone sulfate levels, postoperative hypothalamic–pituitary–adrenal suppression test, somatic mutation, and pathological demonstration of CYP11B1 in resected specimens. To obtain a panoramic view of the effects of ACS on image–AVS discordance, long-term postoperative cohort and risk stratification should be done, and additional parameters, data, and surgical results should be gathered.

Conclusion

In PA patients with ACS, the image–AVS discordance rate was significantly higher and the postoperative complete biochemical success rate was significantly lower than in those without ACS. In the logistic regression analysis, ACS was the only negative predictor of biochemical outcome. There was no improvement in postoperative outcomes, regardless of whether adrenalectomy was based on AVS or imaging. Further studies with a larger number of ACS patients and novel biomarkers for AVS evaluation are crucial to attain better postoperative outcomes for PA patients with ACS.

Declaration of interest

The authors have no conflict of interest to disclose in relation to this work.

Funding

This study was supported by grants from the MOST-111-2628-B-002-025-MY3 from the Ministry of Science and Technology, Taipei, Taiwan to C.C.C.