Primary Bilateral Macronodular Adrenal Hyperplasia: A Case Series from a Tertiary Hospital

Primary bilateral macronodular adrenal hyperplasia (PBMAH) is a heterogeneous and rare disease characterized by the presence of bilateral adrenocortical benign macronodules (larger than 1 cm), which can lead to varying levels of excess cortisol secretion [1]. Mild autonomous cortisol secretion (MACS) is more frequent in patients with bilateral hyperplasia or nodules than in those with unilateral nodules [2]. PBMAH accounts for less than 2% of endogenous Cushing’s syndrome (CS) cases. However, it may be responsible for up to one-third of bilateral adrenal incidentalomas with MACS [3].

The pathogenesis is complex, and various mechanisms have been identified, such as (i) aberrant adrenal expression of hormone receptors that stimulate the cAMP/PKA pathway, including vasopressin, luteinizing hormone/human chorionic gonadotropin, glucagon, serotonin, gastric inhibitory polypeptide (GIP), angiotensin, and beta-adrenergic receptors; (ii) paracrine production of adrenocorticotropic hormone (ACTH) in adrenal tissues; and (iii) germline pathogenic variants of ARMC5 and KDM1A genes, as well as in the context of syndromes such as multiple endocrine neoplasia syndrome type 1, familial adenomatous polyposis, hereditary leiomyomatosis and renal cell cancer, and McCune Albright syndrome [3, 4]. The clinical presentation can range from asymptomatic to overt CS, as a consequence of the variable degree of cortisol secretion [5]. Hormonal workup is similar to that of other causes of ACTH-independent hypercortisolism (MACS or overt CS) [2]. It is important to assess aldosterone secretion considering the possibility of the co-secretion of both hormones [3]. The European Society of Endocrinology clinical practice guidelines on the management of adrenal incidentalomas recommend genetic testing for germline pathogenic ARMC5 variants in patients with multiple nodules and ACTH-independent hypercortisolism. It also advises against routine testing of aberrant receptor expression [2].

Available evidence for patient management is limited. Therefore, an individualized strategy that considers age, degree of cortisol excess, general condition, comorbidity status, and patient preference is reasonable [2]. Active surveillance, treatment of comorbidities, medical therapy with steroidogenesis inhibitors, and adrenal surgery are all options depending on the patient’s condition [2, 3]. Bilateral adrenalectomy (BA) should be avoided as first-line option [2]. The aim of our study was to describe the clinical, hormonal, and radiological phenotypes and treatment of patients diagnosed with PBMAH who are being followed at our tertiary center.

We performed a retrospective observational study evaluating patients with PBMAH who are currently or were previously followed up from June 2010 to May 2024 in the Endocrinology Department of our tertiary center. The inclusion criteria were (i) age at diagnosis greater than or equal to 18 years, (ii) PBMAH diagnosed by adrenal histology or computed tomography (CT) scan, and (iii) evidence of endogenous ACTH-independent hypercortisolism. The exclusion criteria were (i) chronic treatment with glucocorticoids or drugs that interfere with dexamethasone metabolism, (ii) use of estrogens (which had to be withheld for at least 6 weeks before performing hormonal workup), (iii) pheochromocytoma, and (iv) missing information in the hormonal or radiological workup.

Demographic, clinical, analytical, and radiological data were collected from the medical records at the time of diagnosis and during the last evaluation. Clinical data included age at diagnosis, sex, BMI, context of initial investigation, presence or absence of symptoms and signs of CS, hypertension, diabetes or prediabetes, dyslipidemia, osteoporosis, history of venous thromboembolism, first-line treatment chosen, and follow-up time. The biological assessment of cortisol excess included 8:00 a.m. serum cortisol after 1 mg dexamethasone overnight suppression test (1 mg DST), late-night salivary cortisol (LNSC), 24 h urinary-free cortisol (UFC), 7:00 a.m. serum cortisol after two-day, low-dose dexamethasone suppression test (2d-LDST). Other analytical parameters recorded included serum potassium, 8:00 a.m. plasma ACTH, serum dehydroepiandrosterone sulfate, plasma direct renin concentration, and plasma aldosterone concentration. Additionally, data of screening for aberrant receptors and genetic tests for germline variants of ARMC5 or KDM1A genes were recorded. KDM1A gene variants were analyzed if no pathogenic or likely pathogenic variants were identified in ARMC5 gene. CT scans or adrenal magnetic resonance imaging was reviewed by two radiologists, who measured the total adrenal size as the sum of the largest diameters of both adrenal glands [6], the size and density (Hounsfield units [HU]) of the largest adrenal nodule, and the presence or absence of bilateral adrenal involvement.

The histological or radiological classification of PBMAH was made according to the 2022 WHO Classification of Adrenal Cortical Tumors, which is defined by the presence of at least two adrenal nodules greater than 1 cm with benign characteristics [1]. To hypercortisolism diagnosis, we considered the presence of at least two unequivocally abnormal first-line tests (same or different) [7]. The first-line tests included the dexamethasone suppression test (DST), LNSC, and 24 h-UFC [7]. Modest cortisol excess was defined as a 24 h-UFC level less than three times the upper limit of normal (ULN), while severe cortisol excess was defined as a 24 h-UFC level greater than three times the ULN [3]. ACTH independency was established if the ACTH plasma concentration was below 1.1 pmol/L (5 pg/mL); ACTH levels between 1.1 and 4.4 pmol/L (5–20 pg/mL) were also considered indicative of an adrenal cause for hypercortisolism due to the clinical context [8]. CS was established when at least two first-line tests were unequivocally abnormal in the presence of signs and symptoms of overt CS [9]. MACS was defined as the presence of cortisol after 1 mg DST above 50 nmol/L (1.8 μg/dL) on at least two tests in the absence of overt CS signs and symptoms, without any further stratification based on the degree of cortisol nonsuppressibility [2]. Hypertension, diabetes, prediabetes, dyslipidemia, and osteoporosis were defined according to the most recent international guidelines [10‒13]. Bilateral adrenal involvement was defined as the presence of at least one nodule greater than 1 cm in diameter in each adrenal gland. Unilateral involvement was defined as the presence of one or more nodules greater than 1 cm in diameter on one adrenal gland, with the contralateral gland appearing normal.

The 1 mg DST involved measuring cortisol in the morning (8:00 a.m.) after the administration of 1 mg of dexamethasone the night before the blood sampling, between 11:00 p.m. and 00:00 a.m. [14]. The 2d-LDST consisted of measuring cortisol levels at 7:00 a.m. after the administration of dexamethasone (0.5 mg) every 6 h for 48 h (last administration at 1:00 a.m.) [14]. Screening for aberrant receptors was performed according to the protocol published by Lacroix et al. [15]. The receptors tested include angiotensin II, vasopressin, β-adrenergic, ANP, GIP, luteinizing hormone, FSH, GnRH, TRH, TSH, prolactin, serotine (5-HT4). An increment of 50% or greater in cortisol levels was considered a positive response to the test, between 25 and 49% a partial response, and less than 25% no response [15]. We calculated the aldosterone-to-renin ratio (ARR), and a diagnosis of primary aldosteronism was considered unlikely if the ARR was under 91 [16]. Table 1 provides information about laboratory assays and the normal range of each parameter for our laboratory.

Data analysis was performed using SPSS Statistics 27^® software. Continuous variables with a normal distribution are presented as mean ± standard deviation, and continuous variables with a non-normal distribution are expressed as median (25th−75th percentiles). The normality of the data distribution was tested using the Kolmogorov-Smirnov test of normality. Categorical variables are displayed as frequencies and percentages. Spearman’s correlation coefficients were used to assess the correlation between the different screening tests for CS, and between these tests and radiological features. All reported p values are two-tailed, with a p value below 0.05, indicating statistical significance. All procedures performed in this study were in accordance with the ethical standards of the institution.

A total of 20 patients with hypercortisolism due to PBMAH were included. The clinical, hormonal, and imaging characteristics are summarized in Table 2 and are detailed in online supplementary Table 1 (for all online suppl. material, see https://doi.org/10.1159/000541914).

In our cohort, the male-to-female ratio was 1:1. The mean age at diagnosis was 56.1 ± 11.1 years. At the first evaluation, the mean BMI was 27.6 ± 3.6 kg/m², with 14 (77.8%) patients having overweight or obesity.

Patients were most frequently investigated for adrenal incidentalomas identified on abdominal imaging (n = 15, 78.9%). Three (15.8%) patients were investigated for overt CS, and one (5.30%) was diagnosed during familial screening of first-degree relative carriers of an ARMC5 pathogenic variant.

Truncal obesity (n = 13, 65.0%) was the most frequent clinical sign present at diagnosis, followed by moon facies (n = 8, 40.0%) and facial plethora (n = 7, 35.0%). Proximal muscle weakness and violaceous striae were less frequent symptoms, each presenting in one (5.0%). Five (25.0%) patients did not exhibit any clinical manifestations of CS.

All patients had one or more comorbidities potentially attributable to hypercortisolism, with hypertension (n = 19, 95.0%) and dyslipidemia (n = 16, 80.0%) being the most common. Eleven (55.0%) patients had diabetes mellitus, and five (25.0%) had prediabetes. The prevalence of osteoporosis and venous thromboembolism was 25.0% and 5.0%, respectively. At diagnosis, five (25.0%) patients had hypokalemia, with a mean serum potassium of 3.04 ± 0.11 mmol/L.

Sixteen (80.0%) patients had at least two different abnormal first-line tests, while four (20.0%) patients showed abnormal DST alone. The median level of 8:00 a.m. serum cortisol after 1 mg DST was 223.5 (73.1–698.6) nmol/L (8.11 [2.65–25.3] µg/dL), and after 2d-LDST was 397.2 (52.4–888.0) nmol/L (14.4 [1.90–32.2] µg/dL). There was a strong correlation between the results of these two tests (r = 0.85, p = 0.004). The median LNSC level was 1.48 (0.765–3.21) times the ULN, and the 24 h-UFC level was 1.61 (0.42–3.64) times the ULN. LNSC levels showed a moderate correlation with 24 h-UFC levels (r = 0.636, p = 0.048). In 17 (85.0%) patients, ACHT was lower than 1.1 pmol/L (5 pg/mL), and in the remaining three cases, the ACTH level was within the range of 1.1 to 2.2 pmol/L (5–10 pg/mL). The median serum dehydroepiandrosterone sulfate level was 1.35 (0.48–2.01) µmol/L (49.7 [17.7–74.1] µg/dL).

One patient (number 8) was diagnosed with primary aldosteronism, with an ARR of 201, and a confirmatory test demonstrating inappropriate aldosterone secretion. Aberrant receptor expression was screened in 12 (60.0%) patients. It was negative in five (41.7%) patients. Four (33.3%) patients had a positive response in at least one test, and three (25.0%) had at least one partial response. The glucagon and upright tests were the most frequently abnormal, with the glucagon test showing two positive responses and one partial response, whereas the upright test showed one positive response and three partial responses (online suppl. Table 1).

Bilateral adrenal macronodules were present in 19 (95.0%) patients. One patient harbored several nodules in the left adrenal gland, with an apparently normal contralateral gland. The mean total adrenal size was 98.6 ± 24.8 mm. The mean size of the largest adrenal nodule was 32.4 ± 8.92 mm and had a mean attenuation of 6.8 ± 13.5 HU. In six (31.6%) patients, the largest nodule showed an attenuation higher than 10 HU. The total adrenal size showed a moderate correlation with cortisol after the 1 mg DST (r = 0.633, p = 0.006) and 24 h-UFC (r = 0.644, p = 0.005).

Genetic study for the ARMC5 gene was performed in eight (40.0%) patients, and in four of them (50.0%) a pathogenic or likely pathogenic germline variant was identified. In one family (cases 13 and 15), the likely pathogenic variant c.888_907dup (Gln303Argfs*4) and a variant of uncertain significance c.2277C>G p.(Pro826Arg) were identified. Case 14 was a sibling of these patients, but did not undergo genetic testing because she died due to postoperative complications. In other cases, heterozygous pathogenic variants c.1379T < C p.(Leu460Pro) [17] and c.1379T>C p.(Leu460Pro) were identified.

One patient underwent a genetic study for KDM1A gene, which did not identify pathogenic variants. ARMC5-mutated patients showed a more distinct phenotype compared to noncarrier of ARMC5 germline variants for hypercortisolism (8:00 a.m. cortisol after 1 mg DST: 750.3 [715.9–755.9] vs. 78.7 [71.8–160.0] nmol/L [27.2 {26.0–27.4} vs. 2.85 {2.60–5.80} µg/dL]; 24 h-UFC: 3.80 [3.21–4.52] vs. 0.31 [0.15–0.96] times the ULN), hypokalemia (n = 3 [60.0%] vs. n = 0 [0.00%]), and adrenal morphology (total adrenal size: 133 [127–143] vs. 81.5 [78.5–84.5] mm; density of the largest adrenal nodule 21.1 [14.9–22.0] vs. 1.17 [−9.59–5.34] HU) (online suppl. Table 2).

First-line treatment modality was chosen individually based on the degree of cortisol excess, general condition, comorbidity status, and patient preference. The criteria used for selecting the adrenalectomy side included the size of the largest nodule and/or asymmetrical uptake on norcholesterol scintigraphy.

Thirteen (68.4%) patients were treated with surgery or cortisol-lowering drugs as the first-line treatment. Among the 12 operated patients, nine (75.0%) underwent unilateral adrenalectomy (UA), and three (25.0%) underwent BA.

Three patients who underwent BA presented with severe CS at diagnosis, with 24 h-UFC levels more than three times above the ULN (range: 3.08–9.86 times the ULN). One patient died due to surgical complications, and the other 2 patients remain on follow-up in our department (mean follow-up time of 120 months).

Nine patients underwent UA. Among these, six (66.7%) presented with modest cortisol excess, two had severe SC, and one had MACS.

After a median follow-up time of 92.5 (72.8–151.5) months, contralateral adrenalectomy was performed in three (37.5%) patients (one is waiting contralateral adrenal-sparing surgery). Cortisol levels were normalized in three (37.5%) patients, and two (25.0%) patients exhibited MACS with controlled cardiometabolic comorbidities (one with improved cortisol secretion). One patient underwent UA recently, and a subsequent cortisol secretion reassessment was scheduled soon.

Patients with severe CS underwent contralateral adrenalectomy, while those with MACS continued to exhibit mild cortisol excess. The patient with primary aldosteronism achieved complete biochemical remission but continued to require antihypertensive medication.

One patient is undergoing lifelong steroidogenesis inhibitor treatment because she refused adrenal surgery. She was treated with ketoconazole 400 mg/day for 159 months. During the last visit, despite the normal 24 h-UFC levels (0.55 times the ULN), the LNSC levels were elevated (2.37 times the ULN), and morning cortisol level was above the target (513.1 nmol/L). Therefore, the ketoconazole dose was increased to 600 mg/day. Serum potassium levels were within the normal range. During follow-up, the patient was diagnosed with osteoporosis and denosumab was prescribed. Currently, diabetes mellitus is under control with two antidiabetic drugs, but hypertension remains uncontrolled despite treatment with four drugs.

In six (30.0%) patients, active surveillance and control of comorbidities potentially attributable to hypercortisolism were chosen. Four (66.7%) patients presented with MACS. After a median follow-up of 34.0 (10.5–83.0) months, these patients continued to exhibit MACS, with their cardiometabolic comorbidities controlled.

Two (33.3%) patients showed modest excess cortisol excess secretion at the time of diagnosis. One patient refused adrenalectomy and opted for active surveillance. Currently, after a mean follow-up of 117 months, 1 patient has MACS, while the other did not show progression of cortisol secretion (LNSC levels were 1.75 times the ULN, and 24 h-UFC was within the normal range). Both patients were newly diagnosed with diabetes mellitus, but other cardiometabolic comorbidities were controlled.

This study provides a comprehensive analysis of 20 patients with hypercortisolism due to PBMAH followed in a tertiary center. Despite the small sample size, it offers essential data on the clinical, hormonal, radiological, treatment, and follow-up outcomes of this rare condition, addressing a gap in the current literature.

Contrarily to previous epidemiological data, which reported a higher prevalence in females (approximately 60% of cases) [3‒6, 18], our series showed an equal ratio of males to females, possibly due to small sample size. However, other sociodemographic and clinical characteristics in our study were consistent with those reported in previous studies, showing that the typical age at diagnosis was 50–60 years [3‒6, 18] and that most cases were discovered incidentally (approximately 70%) [6]. Clinical presentation was variable, ranging from asymptomatic patients to overt CS. The prevalence of comorbidities was similar to previously reported data, with hypertension diagnosed in 65–85% of cases [3, 5, 6] (vs. 95% in our cohort), diabetes in 20–45% [3, 5, 6] (vs. 40%), and osteoporosis in 20–27% [3, 5, 6] (vs. 25%). Hypokalemia was more prevalent in our study (25 vs. 10% [3]).

To date, LNSC has not been assessed specifically in large cohorts of patients with PBMAH [3]. In our sample, the median LNSC level was 1.48 times the ULN and showed a moderate correlation with 24 h-UFC levels. Further larger multicenter studies should be carried out to confirm these findings. Unlike 24 h-UFC, LNSC is less cumbersome for patients to perform [7]. However, 24 h-UFC measurement may be useful to validate CS diagnosis [7, 19] and guide some decisions on patient management, such as the need for Pneumocystis jirovecii pneumonia prophylaxis [19].

On unenhanced CT scan, adrenal adenomas are typically characterized by an attenuation lower than 10 HU with 100% specificity and 58% sensitivity [2]. Approximately 40% of lipid-poor adenomas are present with an unenhanced density greater than 10 HU. The largest published series of patients with PBMAH did not report CT attenuation values [3]. Available data come from a small series, and some case reports indicate density values between 5 and 26 HU [3, 20]. A series with 46 PBMAH patients reported a mean CT attenuation of 4.58 ± 6.98 HU, with a range of 8–24 HU [20]. In our series, the radiological features were similar, with a mean CT attenuation of 6.8 ± 13.5 HU and a range of −16.9 to 32.3 HU. Therefore, imaging evaluation should be approached with caution in PBMAH patients, as sizes larger than 4 cm or an unenhanced density above 10 or 20 HU are common and not unequivocally associated with malignancy [3]. Large prospective studies with patients with PBMAH are necessary to clarify these findings.

In our cohort, we observed a correlation between larger adrenal size and excess cortisol secretion. This finding is in line with that of a larger retrospective study involving 352 patients with PBMAH, 14.8% of whom carried ARMC5 germline pathogenic variants [6]. In ARMC5-mutated patients, the total adrenal size correlated with cortisol after 1 mg DST (r = 0.66, p < 0.001) and 24 h-UFC (r = 0.39, p = 0.01). Conversely, for patients without ARMC5 pathogenic variants, adrenal size showed a weak correlation with cortisol after a 1 mg DST (r = 0.16, p = 0.017) and no correlation with 24 h-UFC (r = −0.06, p = 0.36) [6].

In our cohort, only 40% of the patients underwent genetic testing, and 50% of those were carriers of ARMC5 germline pathogenic variants. The limited number of tests performed in past decades can likely be attributed to the high cost and limited accessibility of genetic testing. However, it is now recognized that genetic studies are essential and should be performed in all patients with PBMAH, as recommended by current guidelines [2]. ARMC5 pathogenic variants account for 20 to 25% of sporadic cases [3, 6, 21, 22]. Some studies, along with our data, suggest that ARMC5-mutated patients exhibit a more pronounced phenotype characterized by more severe cortisol excess, larger adrenal glands, and more frequent metabolic complications [3, 6, 21, 22]. Recently, germline-inactivating pathogenic variants of KDM1A with loss of heterozygosity of the second KDM1A locus in adrenal lesions were identified as the cause of GIP-dependent PBMAH in 90–100% of cases [3, 23, 24]. Identifying these mutations within families could facilitate the earlier diagnosis of hypercortisolism and optimize its management.

Treatment presents several challenges. In our cohort, patients with severe CS were typically advised to undergo BA. Those with overt CS and modest cortisol excess were generally recommended for UA, while patients with MACS were advised to undergo active vigilance. Current recommendations suggest against BA, but if deemed necessary, adrenal-sparing surgery might be considered [2]. In the past, this procedure was recommended for patients with severe hypercortisolism. However, it carries higher perioperative morbidity and mortality risks, and the patient becomes dependent on lifelong gluco- and mineralocorticoid replacement therapy and is at risk for life-threatening adrenal crises [2, 3]. Despite the limited evidence, UA is considered a viable alternative [3]. In our cohort, 83.3% of patients with modest excess cortisol who underwent UA did not exhibit hypercortisolism recurrence or only had MACS after long-term follow-up. In a recent review encompassing 31 published cases or series, among 286 patients treated with UA, 222 (77.6%) achieved initial CS remission and 83 (29.0%) experienced effective control of hypercortisolism or avoided recurrence from the residual adrenal tissue [3]. A cohort with 124 patients, followed by a median of 28.5 months, identified a contralateral adrenal volume >33.54 mL and preoperative 24 h-UFC levels 5.8 times above the ULN as predictors of recurrence [25]. Most recent European guidelines indicate that there is no published evidence suggesting that individuals with PBMAH and MACS should be managed differently than other patients [2]. Therefore, surveillance of comorbidities and appropriate treatment may be adequate. Our findings support this approach, demonstrating stable cardiometabolic outcomes and no progression of excess cortisol secretion over a median follow-up period of 34 months. In a study with 98 patients with adrenal adenomas and MACS, which included 31 patients with PBMAH, no differences in the risk of developing comorbidities were observed between PBMAH and non-PBMAH patients after a median follow-up of 33.7 months [5].

Our study has some limitations that must be acknowledged. First, not all patients underwent confirmatory histological diagnosis, which may have led to an incorrect diagnosis. However, careful imaging analysis may be helpful in distinguishing PBMAH from bilateral adenomas. In PBMAH, internodular hyperplasia is typically visible and can aid in diagnosis [3]. Owing to its retrospective nature, there is heterogeneity in the tests performed to diagnose hypercortisolism, screening of comorbidities attributable to cortisol excess, and management. Additionally, not all patients underwent genetic testing. Finally, the small sample size limits the statistics power to evaluate correlations. The strength of our study lies in the comprehensive description of patients with PBMAH over a longer follow-up period. Despite the small sample size, we addressed gaps in the literature by providing insights into LNSC levels, nodule attenuation, and an approach for managing MACS.

PBMAH is a rare and complex disease that presents with significant challenges. Managing it requires personalized patient care, long-term follow-up, and comprehensive multidisciplinary evaluation. Further large-scale studies are warranted to validate our observations and guide more precise management strategies for PBMAH.

This study protocol was reviewed and approved by Ethics Committee of Unidade Local de Saúde de São João (CES 219/24). Written informed consent was obtained from participants to participate in the study and publication of the details of their medical case. Due to death during follow-up, two participants were unable to provide consent for themselves. Written informed consent was obtained from the next of participants prior to the study.

The authors have no conflicts of interest to declare.

The results were presented at ECE2023 and supported by a scientific presentation grant from the Portuguese Society of Endocrinology, Diabetes and Metabolism (SPEDM). The design, data collection, data analysis, and reporting of this study were not supported by any sponsor or funding sources.

Conceptualization, J.G. and J.P.; methodology, J.G., M.B.C., and J.P.; software and writing original draft preparation J.G.; validation, D.F.S., M.B.C., J.P., and J.Q.; formal analysis, J.G. and M.B.C.; investigation, J.G., T.M., and J.D.; data curation, J.G. and J.D.; writing, review, and editing, T.M., H.U.F., J.M., J.D., D.F.S., M.B.C., J.P., and J.Q.; visualization, all authors; supervision, D.F.S., M.B.C., J.P., and J.Q. All authors have read and agreed to the published version of the manuscript.

The data that support the findings of this study are not publicly available due to the presence of information that could compromise the privacy of research participants but are available from the corresponding author (J.G.) upon reasonable request.