Recent advances of imaging technologies and the widespread use of high-resolution cross-sectional imaging in medical practice has increased the frequency of detection of adrenal disorders (1). “Adrenal Incidentalomas” (AI) are defined as adrenal masses larger than 1 cm found incidentally on abdominal imaging performed initially for another purpose (2). This definition excludes lesions discovered during cancer staging or suspicion of adrenal diseases (3). The incidence of AI has increased 10-fold during the last two decades (4). This increase parallels the technological advances of imaging modalities. AI are thus considered the pathology of modern technology (1). AI are mainly unilateral and are bilateral in only 10-15% of cases (1, 2). Their prevalence increases with age, from less than 1% in people less than 30 years old in autopsy series to 7% in people more than 70 years old (5). They are particularly more frequent in Caucasians patients with obesity, diabetes, and hypertension (1).

Discovery of an AI raises two questions whether it is malignant or if it is secreting excess amounts of hormones. Imaging and biochemical testing are crucial tools to answer these questions. Malignancy is found in 5 to 8% of adrenal masses while 15 to 20% are functional (5 to 30% for hypercortisolism whether clinical or subclinical, 3-7% for pheochromocytomas, and 1 to 10% for primary aldosteronism (PA)) depending on the studied population (1). This highlights the need to evaluate and appropriately treat these patients because of the potential detrimental consequences and complications (6, 7).

Unenhanced computer tomography (CT) offers a valuable tool for evaluating the lipid content within adrenal masses. For homogeneous lesions, a Hounsfield Unit (HU) value below 10 on a non-contrast CT scan strongly suggests a lipid-rich composition. This finding can be highly reassuring, as it reliably excludes malignancy, even in larger adrenal masses. Only lesions larger than 4 cm or with a density greater than 20 HU are typically managed surgically due to their increased risk of malignancy (7). All other cases should be discussed using a multidisciplinary approach (8), including the referring physician, the endocrinologist, the radiologist, and the surgeon.

To find out if the adrenal lesion is functional, a hormonal workup should be performed for all patients with AI (4, 9). This workup includes excluding hypercortisolism in all patients, pheochromocytomas in AI with a density more than10 HU, and PA in patients with AI and hypertension (4). Moreover, in cases of bilateral adrenal masses, congenital adrenal hyperplasia should be ruled out (4).

To date, studies evaluating the prevalence of AI in the Middle East are lacking (10). Furthermore, reports investigating the rate of referral to endocrinologists are scarce (10–13) Our study aims to determine the prevalence and characteristics of AI in a tertiary care center in Lebanon, as well as the proportion of AI that were appropriately followed for hormonal testing.

The prevalence of AI in our population is 7%. The prevalence of AI reported in the literature is extremely variable, ranging from 0.4% to 7.3% ( Table 3 ) (10, 14–26) This variability could be attributed to several factors, including the characteristics of the study population (age, health status, geographic location, exclusion criteria), sensitivity of the imaging procedures, nature of the study (retrospective versus prospective studies), and sample size. Overall, our prevalence is one of the highest among published studies, along with the one reported by Reimondo et al. in Italy (7%) (25). The high-resolution of our CT scan and the prospective nature of our study might partly account for this higher prevalence. Indeed, studies have proved that the prevalence of AI increases from approximately 0.64% to 4-5% with the use of more powerful CTs (1, 27). In addition, retrospective studies tend to demonstrate a lower incidence of AI (as low as 2.5%), probably because AI are often not mentioned on the final report (27). Finally, the fact that our hospital is a tertiary center that handles more complex medical conditions might also explains the higher prevalence observed in our study. Notably, we found a high prevalence of AI despite having excluded all oncologic patients. Our prevalence may have been even higher had we included subjects undergoing scans for tumor or metastasis detection.

We did not find any statistically significant difference in the distribution of AI between men and women. In the literature, AI were found to be more prevalent in women in several radiological series (1, 18, 21, 28). Conversely, two published studies demonstrate a male predominance (25, 29). Male sex has also been associated with a greater risk of malignancy (30). Another study showed that patients with AI frequently develop hypertension and hyperglycemia regardless of sex (31).

Our study showed the expected increase in the prevalence of AI with age. This is in accordance with all previously published studies (1, 32–34) Increased adrenal tumorigenesis with age is likely driven by both an increase in genetic mutations as well as remodeling of the tissue microenvironment (35). Increased use of imaging in older subjects could also explain their higher prevalence of AI.

The median size of AI was 1.7 cm (1.2-2.5), with the majority (59%) ranging between 1 and 2 cm, and most of them (83%) being unilateral, mainly located on the left side (65%). This is similar to the findings in the literature, with studies reporting an average size of 3 cm, with a range of 0.8 to 2.3 cm, and left sided predominance (27, 29, 36, 37) The predominance on the left size might be due to a detection bias, since left-sided adenomas are more easily detectable to radiologists. Indeed, several CT series have shown an equal distribution between the right and left adrenal glands, similarly to autopsy series (27, 29, 36). Furthermore. The density of our incidentalomas is mostly either lower than 10 HU (45%) or higher than 20 HU (45%). However, in a large retrospective study conducted by Ebbehoj et al., the majority of AI had a density inferior to 10 HU (54.7%), and only 21.6% had a density of 20 or more (38). The high proportion of AI with a density > 20 HU in our study underlies the importance of recognizing and investigating AI and could be explained by the fact that our patients were recruited from a tertiary care hospital, with most of them being possibly more ill than patients in primary care settings.

Moreover, AI in our study are significantly more frequent in patients with a higher BMI, hypertension, dyslipidemia, and a history of CVD. According to our multivariable analysis, however, the only statistically significant clinical predictive factor for AI was BMI. This finding suggests that, after adjusting for potential confounders, each one unit increase in BMI is associated with a 10.7% increase in the odds of having an AI. In the study published by Reimondo et al., higher BMI was also significantly associated with the presence of AI, but hypertension, dyslipidemia and CVD were not. Type 2 diabetes however, was significantly more frequent in the group with AI (25) in contrast to our study where diabetes was frequent but didn’t reach statistical significance. This is most probably because patients with autonomous cortisol secretion were not included in our study. The association between AI and the presence of co-morbidities is mainly due to the fact that some AI are functional and secrete either cortisol or aldosterone or catecholamines, which highlights the importance of investigating all AI.

Finally, the referral rate of AI to endocrinology in our study is very low (5.4%). The few published studies analyzing this issue also revealed that clinicians rarely prescribe the appropriate medical assessments and/or refer their patients to an endocrinologist (11, 12). Indeed, a study conducted in a large UK university teaching hospital reported that only 9.6% of patients with AI were referred to an endocrinologist (11). Another review of 804 AI showed that 30% were investigated and followed-up (13). Our study has the lowest referral rate, even though 45% of our AI have a density higher than 20 HU and 7% are larger than 4 cm. Lebanon’s ongoing political and financial crisis might contribute to this lower referral rate, as it could deter healthcare providers from pursuing investigations and many cases were lost to follow-up. Keeping in mind that 37% of AI might increase in size on follow-up (39), and 29% of patients with nonfunctioning AI could develop mild autonomous cortisol secretion during follow-up (40), AI should be followed-up regularly by an endocrinologist to rule out a change in size or aspect on imaging or a new hormonal secretion. The European Society of Endocrinology clinical practice guidelines state that patients with an adrenal incidentaloma should undergo a 1mg overnight dexamethasone suppression test to rule out hypercortisolism. They also recommend no further imaging for benign appearing adrenal masses (homogeneous appearance and a density ≤ 10 HU). As for non-secreting homogenous adrenal masses with a size inferior to 4 cm and a density of 11–20 HU, they suggest either an immediate additional imaging or a follow-up CT scan or MRI in 12 months. To improve patient outcomes, it is therefore essential to raise awareness among physicians of all specialties about the significance of evaluating and monitoring AI, in line with the latest European Society of Endocrinology guidelines (7).

The large sample size of 3168 patients increases the accuracy of our results as well as the validity of our conclusions. Moreover, the prospective design of this study minimizes the likelihood of missed diagnoses of AI by the radiologist. Furthermore, our study is the first conducted in Lebanon and the second in the Middle East regarding the prevalence, clinical and radiological characteristics, and referral rate of AI.

Nevertheless, some limitations should be considered. Patients were selected from a single center, which could induce a selection bias and limit the generalizability of our results. Additionally, the fact that every oncologic patient was excluded might underestimate the prevalence of AI in our population. Moreover, our data collection relies on medical records and patient questionnaires, which may be subject to missing or inaccurate information. Finally, it is important to remember that the evolving nature of imaging technology can influence the detection of AIs, necessitating a continuous reevaluation of findings.

Our study revealed a prevalence of AI of 7%. AI were found to be more frequent in older people as well as patients with higher BMI, hypertension, dyslipidemia, and a history of CVD. The very low referral rate to endocrinologists in our study may be attributed to Lebanon’s ongoing crisis and developing healthcare system. To optimize patient outcomes, raising awareness among healthcare providers about the significance of AI and the need for timely referral is crucial.