Long-Term Follow-Up of Adrenal Incidentalomas: A Portuguese Single-Center Study Open Access

Adrenal incidentaloma is an asymptomatic adrenal mass detected in imaging studies not performed for suspected adrenal disease [1]. This strict definition, recognized by the European Society of Endocrinology and European Network for the Study of Adrenal Tumors (ESE/ENSAT), excludes adrenal lesions discovered during the screening of patients with hereditary syndromes or extra-adrenal tumors [2].

The use of cross-sectional abdominal imaging has increased the incidental detection of adrenal tumors. Adrenal incidentalomas are found in about 1–8% of the population, and incidence increases with age [2, 3].

Current guidelines recommend that incidentally discovered adrenal lesions with a diameter <1 cm should not undergo further investigation unless clinically indicated [2]. A personalized approach for patients with adrenal incidentalomas is designed to detect malignant and hypersecreting tumors [4]. Whereas most incidentalomas are benign nonfunctioning adrenal tumors, about 15–30% are associated with hormone excess [5]. Autonomous cortisol secretion is the most common endocrine hypersecretion, and it has been associated with several cardiovascular risk factors, in particular obesity, arterial hypertension, type 2 diabetes, and osteoporosis [6‒9]. Moreover, hypercortisolism has been associated with an increased risk of major cardiovascular events [10].

The working group of the adrenal incidentalomas guidelines produced by the ESE did a systematic review that focused on malignant transformation in adrenal incidentalomas and found that once the benign nature of an adrenal mass has been established, further radiologic follow-up is unnecessary [2]. The risk of malignant transformation of a benign-appearing adrenal mass (0.2%) seems to be the same as the risk of developing cancer from the amount of radiation in computerized tomography (CT) [11].

The updated ESE/ENSAT guidelines characterize homogeneous lesions as those with a uniform texture and consistent appearance on imaging, often indicating a benign etiology. In contrast, heterogeneous lesions are described as having variable densities or irregular appearances, which may suggest malignancy or other complex pathologies. Pre-contrast Hounsfield unit (HU) values are central to evaluating adrenal masses, with HU ≤10 strongly indicating a benign lesion and HU >20 raising suspicion of malignancy. Although previous studies identified 6 cm as a critical cut-off for surgical consideration, recent updates suggest that masses ≥4 cm and concerning imaging features, such as heterogeneity or elevated HU values, warrant further evaluation or surgical removal [2, 12].

Current guidelines suggest a pre-contrast HU value >10, heterogeneous appearance, and mass size ≥4 cm as criteria for discriminating suspicious from benign lesions and considering surgical removal [2, 12]. However, in a large cohort study of 1,149 patients, the optimal cut-off values of mass size and pre-contrast HU to distinguish malignant from benign lesions were 3.4 cm and 19.9, respectively [13].

Also, there is still controversy regarding the optimal follow-up strategy of non-operated patients. Available guidelines diverge on indications for surveillance of benign adrenal masses [14]. ESE/ENSAT and the American College of Radiology (ACR) guidelines propose that no follow-up imaging be obtained for a patient with a mass that appears benign on initial CT [2, 15]. The American Association of Clinical Endocrinologists (AACE) guidelines recommend reimaging the patient in 6 months and then annually for 1–2 years [12]. For adrenal masses with indeterminate characteristics (unilateral mass ≥4 cm but HU <20 and nodules with HU >20 or heterogenous appearance but <4 cm), the ESE/ENSAT guidelines propose a single follow-up exam after 6–12 months, to exclude significant growth [2].

The ESE/ENSAT guidelines do not recommend additional hormonal testing if the initial assessment falls within normal limits [2]. In contrast, the AACE, the French Endocrinology Society, and the Spanish Society of Endocrinology and Nutrition advocate for an annual hormonal panel over a period of 5 years following diagnosis [5]. Additionally, an Italian multicenter study identified a higher risk of subclinical hypercortisolism in individuals with adenomas measuring ≥2.4 cm, supporting the need for long-term biochemical follow-up in this population [16].

In relation to this point, Chomsky-Higgins et al. developed a cost-effectiveness model, which evaluates different surveillance strategies for management of small (<4 cm), benign-appearing adrenal incidentalomas [17]. Four possible strategies were compared: no surveillance, single surveillance at 12 months, annual surveillance for 2 years, and annual surveillance for 5 years. The results showed that the two strategies with more surveillance were less effective and more costly because of false positives, unnecessary surgery, and exposure to radiation. The single surveillance strategy was cost-effective in younger patients (<60 years), whereas in older patients (>60 years) the no-surveillance strategy was more effective. The authors conclude by suggesting a management strategy similar to what is indicated by the ESE/ENSAT guidelines but adding a single radiological and biochemical follow-up in patients aged <60 years [17].

A meta-analysis revealed that new development of overt hormone secretion almost never occurs (<0.1%) [3]. Moreover, they found that adenomas with a larger mean size at baseline were less likely to grow during follow-up, suggesting a maximum growth potential. The objective of this study was to assess tumor growth and the development of hormone secretion during follow-up in nonfunctioning adrenal incidentalomas followed at a single center.

Retrospective cohort study of the medical records of patients with adrenal incidentaloma, evaluated between 2014 and 2024, at the Endocrinology Department of Hospital das Forças Armadas in Portugal.

Inclusion criteria included patients over 18 years with adrenal incidentalomas, with at least 5 years of follow-up. Exclusion criteria included patients with Cushing’s syndrome, primary hyperaldosteronism, pheochromocytoma, or adrenal carcinoma.

CT was used as the preferred diagnostic technique. The largest transverse diameter of an adrenal lesion was used to represent the lesion size; in bilateral adenomas, both diameters were reported. The HU value in unenhanced CT was measured and a unit cut-off of <10 was used to indicate a benign imaging phenotype. In patients with higher values, a contrast CT scan or a magnetic resonance imaging (MRI) was conducted in 6–12 months, according to our department protocol. Nodules ≥4 cm, with HU >20 or heterogeneous appearance were considered having high malignancy risk and were considered for surgery.

For benign nodules and suspicious adrenal masses that did not fulfill the criteria for surgery (unilateral mass ≥4 cm but HU <20 and nodules with HU >20 or heterogenous appearance but <4 cm), a control CT re-evaluation was performed at least at 1, 2, and 5 years. Homogeneous lesions were defined as having a uniform texture and appearance throughout imaging studies, and heterogeneous lesions were defined as presenting varied textures or densities within the same lesion, indicating a non-uniform composition. Significant tumor growth was defined as an increase of more than 20% in maximum diameter and more than 5 mm. Dimensional stability was considered if <5 mm variation between successive imaging exams.

All patients were studied as outpatients in a non-stressed environment. The following parameters were measured at baseline and at each annual evaluation for at least 5 years: 1-mg overnight dexamethasone suppression test (DST); urinary fractionated metanephrines; and plasma aldosterone, renin activity, and aldosterone/renin ratio if the patient had hypertension. A cut-off value of serum cortisol levels post dexamethasone ≤1.8 µg/dL was used to exclude cortisol hypersecretion. According to ESE guidelines, autonomous cortisol secretion was defined as DST serum cortisol concentrations >1.8 µg/dL, in the absence of typical Cushingoid features. When autonomous cortisol secretion was diagnosed, ACTH was measured to exclude ACTH-dependency, and salivary cortisol and 24 h urinary cortisol were measured to assess the degree of cortisol secretion.

Continuous data are presented as mean and standard deviations or median and interquartile ranges, according to their adaptation to a normal distribution, analyzed using the Shapiro-Wilk test. Differences between variables along the follow-up period were assessed with the paired T test or Wilcoxon signed-rank test.

Categorical variables are presented as absolute and relative frequencies and their associations studied using Pearson’s chi-square test, Fisher’s exact test, and McNemar test as appropriate. All analyses were performed using SPSS software version 23^®, with a significance level set to 5%.

A total of 136 patients (44% female), with a median age at diagnosis of 74 [66; 81] years were evaluated. The median follow-up was 7 [5; 8] years, with a maximum of 10 years. At baseline, hypertension (59%) and dyslipidemia (43%) were the most common comorbidities, but a significant number of patients also had type 2 diabetes (30%) and obesity (20%). The median tumor size at the time of diagnosis was 20 [15; 26] mm, with a median HU of 0 [−7; 10] and 24.3% of patients presented bilateral tumors (shown in Table 1).

Nine patients (6.6%) presented with suspicious image criteria at diagnosis: one nodule had more than 40 mm (45 mm) but had 10 HU and was homogeneous; two had HU >20 (HU 35 and HU 44) but measured 23 and 12 mm and were both homogenous; and six were heterogeneous nodules with less than 40 mm. Of these, only two were submitted to surgery: a hemorrhagic adrenal cyst due to a significant growth, and a pheochromocytoma. All 11 nodules with indeterminate HU were studied with enhanced CT scan or MRI, revealing benign features.

Each patient underwent a median of 3.4 imaging studies during follow-up. Overall, there was no difference in size, HU, or homogeneity (shown in Table 1).

In 111 patients (81.6%), the size of the tumor did not change along the follow-up (p = 0.278, shown in Table 1; Fig. 1). A significant reduction in size of −19 [−20; −18] mm (p = 0.005, shown in Table 1; Fig. 1) was observed in 10 patients (7.4%) and a significant growth of 10 [7; 20] mm (p = 0.001, shown in Table 1; Fig. 1) was reported in 15 patients (11%). None of the last registered a change in HU or homogeneity.

Heterogenous pattern was associated with significant growth (50% vs. 12%, p = 0.018). There was no difference between patients with significant nodular growth and those without it, regarding age (p = 0.317), gender (p = 1.000), baseline nodular size (p = 1.000), and HU (p = 1.000). At baseline, 10 (7.4%) patients presented with hyperfunction, of which nine with autonomous cortisol secretion (median DST 2.1 [2.0; 2.5]), without overt Cushing’s syndrome along the follow-up period, and one with a pheochromocytoma (shown in Table 2).

All 9 patients with autonomous cortisol secretion had ACTH-independency and normal salivary and urinary cortisol levels. There was no difference at baseline between the prevalence of cortisol associated comorbidities to the rest of the population: type 2 diabetes (55.6% vs 28.3%, p = 0.086); hypertension (88.9% vs. 56.7%, p = 0.058); dyslipidemia (66.7% vs. 40.9%, p = 0.132); and obesity (44.4% vs. 18.1%, p = 0.056). Only 1 patient with autonomous cortisol secretion developed T2D along the follow-up, considered statistically non-significant (p = 0.347). During follow-up, none of the remaining patients developed hyperfunction (shown in Table 2).

In agreement with previous investigations, we found that most adrenal incidentalomas remained stable in size (p = 0.278), HU (p = 1.000) and homogeneity (p = 1.000). Significant tumor growth (>20% and >5 mm) was observed in 11 patients, of which only one required surgery, being diagnosed as a hemorrhagic adrenal cyst, while the others remained under surveillance. In none of these was observed a change in HU or homogeneity.

Adrenal cysts are very rare lesions, usually asymptomatic [18]. Ultrasonography, CT, and MRI studies have been very effective in recognizing cystic lesions; however, radiologic findings are typically inadequate for the definitive subtyping of a cystic adrenal lesion [19]. The other only nodule submitted to surgery was a pheochromocytoma, diagnosed at the initial workup.

A recent Spanish multicenter study reported female sex as a risk factor for adrenal tumor growth [5]. In the present study, no difference was found between genders (p = 0.001), tumor size, or HU at baseline. The only variable associated with significant growth was heterogeneity, but the small number of patients and the high disparity between groups (6 vs. 133 with homogeneous pattern) makes this comparison biased.

Regarding imaging studies, 13 patients with benign tumors presented HU >10. Subsequently, they underwent enhanced CT scan or MRI, excluding malignancy. These results support data revealing that 30% of benign adenomas have an attenuation value of HU >10 on unenhanced CT and are considered lipid poor, overlapping in density with malignant lesions and pheochromocytomas [20].

As for the long follow-up of indeterminate adrenal nodules, the ESE/ENSAT guidelines have no recommendations regarding imaging beyond the first initial revaluation after 6–12 months [2]. Most previous cohort studies revealed no malignant transformation in typical benign nonfunctional adrenal tumors [2]. In our study, only one pheochromocytoma was diagnosed at baseline and most adrenal incidentalomas remained stable in size and hormonal function. Our results show that although there may be an initial nodular growth in a minority of patients in the first 2 years, size stability is achieved from there on.

All patients who experienced tumor growth were over 65 years, which contradicts the single surveillance approach for those under 60 years by Chomsky-Higgins [17]. As so, we support ESE/ENSAT recommendations that propose no follow-up imaging should be obtained for a patient with a mass that appears benign on initial CT [2]. We also suggest that in indeterminate adrenal nodules, following confirmation of other benign features either through enhanced CT scan or MRI, no more follow-up imaging is required.

Despite 9 patients having been diagnosed with autonomous cortisol secretion at baseline, there was no difference regarding comorbidities compared to the remaining population and no new cases were observed during the follow-up 7-year period. This supports the recommendation of ESE/ENSAT, based on studies with a total of more than 3,000 patients with nonfunctioning adrenal incidentalomas, confirming the extremely low risk of developing clinically relevant overt hormonal excess [2].

Also, it contradicts a study of Araujo-Castro et al. [5] in which 10.5% of patients with nonfunctioning adrenal tumors at diagnosis progressed to autonomous cortisol secretion during a median follow-up of 3 years. Despite this progression, no patient developed overt Cushing’s syndrome. Interestingly, they proposed a follow-up for incidentalomas based on serum cortisol post-DST levels at diagnosis: <0.9 µg/dL, one DST 5 years after diagnosis; 0.9–1.45 µg/dL, performing DST every 2.5 years for 5 years; and 1.45–1.8 µg/dL, performing DST yearly for 5 years. This approach seems reasonable, particularly in tumors >2.5 cm [16]. Since most cases developing autonomous cortisol secretion do so during the initial 5 years of follow-up, it is rational not to repeat hormonal workup unless new clinical signs of endocrine activity appear or there is worsening of comorbidities, as recommended in recent guidelines [2]. However, we cannot infer any conclusions from our results regarding this approach.

While we recognize the importance of evaluating bone health, including bone mineral density and vertebral fractures, in patients with autonomous cortisol secretion, this was not included in our study due to logistical and resource constraints. Specifically, our center does not have routine access to dual-energy X-ray absorptiometry scans or vertebral imaging as part of the standardized follow-up protocol for adrenal incidentalomas. Additionally, the primary objective of our study was to focus on the longitudinal assessment of tumor growth and hormonal activity, and expanding the scope to include bone health parameters would have required a significantly larger sample size and extended follow-up, which were beyond the scope of our available resources. We acknowledge this as a limitation and suggest that future research incorporates bone mineral density assessments to provide a more comprehensive evaluation of comorbidities associated with autonomous cortisol secretion.

Regarding the potential for false-positive results in DST when diagnosing autonomous cortisol secretion, we acknowledge this as an important consideration. False positives may arise due to accelerated dexamethasone metabolism, influenced by factors such as certain medications (e.g., enzyme-inducing drugs) or individual metabolic variability, as well as poor compliance with the test protocol. This could lead to an overestimation of autonomous cortisol secretion prevalence in clinical and research settings. Measuring serum dexamethasone levels during the DST could help identify inadequate suppression attributable to these factors, providing a more accurate assessment of cortisol autonomy. While this approach was not implemented in our study, it represents a promising area for future research to refine the diagnostic accuracy of DST and better delineate true cases of autonomous cortisol secretion. Incorporating dexamethasonemia levels into routine protocols could help standardize the test and mitigate variability, ultimately improving patient care and enhancing study findings.

We are aware that the retrospective design and the small number of patients included in this study are limitations. However, there are several strengths that should be mentioned. All the patients were evaluated in the same hospital with the same protocol, with consequent homogenous diagnosis criteria and follow-up avoiding bias. We excluded patients with less than 5 years of follow-up, and the median follow-up time of the study was over 7 years, longer than previous cohort studies. All imaging exams in our study were evaluated by the same dedicated Imagiology team at our institution, ensuring consistency in protocols and interpretation criteria. However, the assessments were not performed by a single radiologist but rather by different members of the team, all of whom followed standardized guidelines for measuring tumor size, density, and other relevant characteristics. While inter-operator variability cannot be completely excluded, this approach minimizes inconsistencies by leveraging a team with specialized expertise and a uniform methodology. We acknowledge this as a potential limitation.

In conclusion, our study reinforces the prevailing knowledge that the majority of adrenal incidentalomas exhibit stability in size, density, and homogeneity over time, and that the risk of malignant transformation in typical benign adrenal tumors is low. Furthermore, the absence of new cases of autonomous adrenal hormone secretion during the follow-up period supports existing evidence suggesting a low risk of developing clinically relevant hormonal excess in patients with nonfunctioning adrenal incidentalomas. We advocate for adherence to established guidelines, particularly those proposed by ESE/ENSAT, in determining the necessity for continued surveillance based on initial imaging and biochemical characteristics.

This study has been granted an exemption from requiring ethics approval by the Hospital das Forças Armadas Ethics Committee and complies with the Declaration of Helsinki as it is a retrospective cohort study. This study has been granted an exemption from requiring written informed consent by the Hospital das Forças Armadas Ethics Committee and complies with the Declaration of Helsinki as it is a retrospective cohort study.

The authors have no conflicts of interest to declare.

This study was not supported by any sponsor or funder.

David Veríssimo, Beatriz Pereira, and Joana Rodrigues performed material preparation, data collection, and data analysis. David Veríssimo wrote the first draft of the manuscript. Catarina Ivo, Ana Cláudia Martins, João Nunes e Silva, Dolores Passos, Luís Lopes, João Jácome de Castro, and Mafalda Marcelino reviewed and commented on previous versions of the manuscript. All authors contributed to the study conception and design and read and approved the final manuscript.

The data that support the findings of this study are available upon request from the corresponding author. Due to privacy regarding restrictions, it is not publicly available.