Continued innovation in imaging modalities has enhanced sensitivity in identifying abnormalities of the adrenal glands. The prevalence of adrenal adenomas in the general population is estimated at 3-7%, with the overwhelming majority of those discovered incidentally demonstrating benign pathology.7 However, because the risk of primary adrenocortical carcinoma is reported to be as high as 4.7% in patients with incidentally discovered adrenal masses, characterization and appropriate treatment remain crucial.8 The prevalence of clinically unapparent adrenal masses at autopsy is reported at nearly 2.1%. Detection is increasing with the advent of modern imaging technology. Estimates of incidence range from 0.1% for general health ultrasound screening to 0.42% in patients evaluated for non-endocrine symptoms to 4.3% in patients previously diagnosed with cancer. This prevalence also varies with age, at <1% for patients younger than 30 years of age and increasing to 7% in patients 70 years of age or older.9 Overall, incidentally identified adrenal masses occur in up to 9% of CT scans, justifying the need for accurate guidelines to direct their management.10
Advancements in imaging also allow for better characterization of adrenal masses prior to biopsy or surgical resection. Technological developments have improved the anatomic resolution, sensitivity, and specificity of cross-sectional imaging modalities, particularly MRI and contrasted CT (recommended modalities in the setting of AI), over the last 15-20 years.5 This is fueled by the superiority of these modalities over their conventional counterparts in the accurate detection of disease. Similarly, advancements have led to an increase in data generated with more cross-sectional images per scan.11 Improved cross-sectional imaging allows for detection of intralesional fat, water and blood enabling characterization of benign adrenal masses, such as adenomas, angiomyolipomas, cysts, granulomas and hemorrhage. Nieman et al. emphasize the importance of washout in delayed-phase CT scan due to the high prevalence of benign histology in masses with washout >50%.12,13 When inconclusive, or if concerning features are noted, it may be appropriate to establish a tissue diagnosis by performing adrenal biopsy or perform definitive surgical management for potentially malignant lesions. Accurate characterization is imperative due to the potential for metastases to the adrenal gland, which may preclude surgical or radiation therapy to another primary site.10
An additional factor contributing to increased detection is increased utilization of topographic imaging. In 2008, an estimated 60 million CT scans were performed in the United States growing at a rate of 5% per year.14 This is astounding when compared to the 3.8% rate of increased imaging amongst Medicare patients between 1993-1999 and parallels the increase in technology occurring with cross-sectional imaging.15 This widespread use of cross-sectional imaging is thought to have occurred because of ordering physicians’ attempts to address demands of both consumers and referring physicians, while providing superior patient care.14 A significant proportion derives from a somewhat overzealous usage of cross-sectional imaging in emergency departments.16 A 2013 study revealed that CT technology is available in 97% of emergency department (ED) visits in the United States, and is performed during 11.4% of visits, most commonly for complex abdominal pain.17 Similarly, Bellalio et al. report an almost 60% increase in CT utilization in ED’s from 2005-2013 with an order for CT in 17.8% of visits. The only population that saw a decrease in utilization was children, who were not included in our study and are less likely to have adrenal incidentalomas.16 These data are corroborated by Baloescu, who underscores that even with greater than 60 million CT scans ordered in ED’s in the United States in 2005, no decrease in morbidity and mortality was appreciated.18 It is less likely that this increase in cross-sectional imaging is coming from primary care providers in the outpatient setting, as Weilburg et al. demonstrate the ability of a utilization management system to decrease the number of high cost imaging studies (CT, MRI, nuclear imaging, and PET) ordered in the outpatient setting from 0.43 exams per year in 2007 to 0.34 exams per year in 2013, with a cohort of around 100,000 patients.19 Clearly, it is possible to implement systems in which the over-utilization of cross-sectional imaging is curtailed, but it remains prevalent in EDs and other clinical settings, including inpatient admissions, leading to the common discovery of AIs highlighted in our study.
With increased detection comes an increased demand for evidence-based clinical recommendations to appropriately guide health care personnel in the management of AIs. Multiple authors have demonstrated the benign nature of the majority of small adrenal masses. The AACE and AAES published guidelines a decade ago highlighting their recommendations for workup and management of such lesions. In short, they recommend clinical, biochemical, and radiologic evaluation of patients with signs of hypercortisolism, hyperaldosteronism (if hypertensive), pheochromocytoma, or malignancy. Adrenalectomy is recommended after hormonal evaluation in patients with a mass ≥4 cm or with malignant features on CT scan. Adrenalectomy is also recommended in smaller, less suspicious tumors in the presence of hormonal abnormalities as evidenced by aldosterone concentration/renin activity, plasma-free metanephrines and normetanephrines, and overnight 1-mg dexamethasone suppression test. Patients in whom adrenalectomy is not recommended (<4 cm homogenous lesions with regular borders and <10 HU on non-contrast CT) should follow surveillance guidelines, which are discussed below.5
The 4 cm cutoff for adrenalectomy is well-established and perhaps too-readily accepted by those outlining guidelines for the management of AIs. Mantero et al. reported that 4 cm had a greater sensitivity (93%) for detection of ACC when compared to 5 cm (81%) and 6 cm (74%), respectively. However, their 2000 study was limited to an Italian population that included only 1004 patients and the issue of size’s ability to predict benign or malignant pathology has been revisited little in the past two decades.20 Sturgeon et al. evaluated 457 patients with ACC and 47 patients with adrenal adenomas, establishing the sensitivity of tumor size in predicting malignancy to be 96% for tumors ≥4 cm, 90% for tumors ≥6cm, and 77% for tumors ≥8 cm. Based on their data, at a threshold of ≥4 cm, the likelihood of malignancy doubles (to 10%).21 Cutoffs ranging from 4-6 cm have been proposed for surgical excisional of incidentally-discovered adrenal masses. A recent analysis of 2219 patients by Kahramangil et al. shows ACC rates of 0.1%, 2.4%, and 19.5% in patients with masses <4 cm, 4-6 cm, and >6 cm, respectively, with an optimal cutoff of 4.6 cm. In addition, analysis by Hounsfield (HU) density on non-contrast CT showed ACC risk of 0%, 0.5%, and 6.3% for masses of <10, 10-20, and >20 HU, respectively. In addition, male sex and >0.6 cm/year growth rate were independent predictors of ACC.22 Similarly, Birsen et al. (2014) describe development of a scoring system to determine the probability of an adrenal mass representing ACC based on size and HU. Of their 157 patients, seven without hormonal secretion had ACC on final pathology; of these, only a single ACC measured <6 cm.23 Due to increasing risk of malignancy with those >4 cm, authors readily accept 4 cm as the cutoff. However, our data suggests that this cutoff may be overly-cautious and subjects patients to unnecessary morbidity that accompanies adrenalectomy.
Along with a conservative approach to management comes a need to clarify ongoing surveillance imaging and laboratory evaluation intervals. This is due to a 17%, 29%, 47% risk of AIs converting to functional status and a 6%, 14%, 29% risk of increasing in size at 1, 2, and 5 years, respectively.5 Current guidelines vary, with most suggesting follow-up guided by clinical judgment and the presumed cause of the mass based on initial workup. National Institute of Health (NIH) data suggests that for masses that are benign-appearing (<10 HU; washout >50%), small (<3 cm), and completely non-functioning, imaging and biochemical reevaluation at 1–2 years (or more) is appropriate, with subsequent follow-up only if the clinical picture changes; the risk of malignancy or subsequent hyperfunction is almost nonexistent though reported conversion to ACC exists in recent literature.24 For indeterminate lesions, repeat evaluation for growth after 3–12 months is appropriate. Subsequent testing should occur earlier for lesions with increasing size, and later for those with no change.12 The AAES extends the criteria for “small” AIs to include those <4 cm and they better outline surveillance recommendations to include radiographic reevaluation at 3 to 6 months, then annually for 1 to 2 years. Similarly, they suggest hormonal evaluation annually for up to 5 years.5 Repeat screening for hyperaldosteronism is not perceived to be beneficial, but most authors recommend screening for catecholamine and cortisol excess for at least four years due to the similar appearance of pheochromocytomas to lipid-poor adenomas on CT scan. This is one instance in which clinical judgment can be used to guide surveillance as most pheochromocytomas grow over time suggesting that further imaging may not be indicated in the setting of a stable mass.12
The above algorithms include 1-3 radiological assessments in the first two years with reconsideration of surgical excision with 0.5-1 cm of growth during follow up. Corwin et al. (2019) assessed 131 adrenal masses, 26 of which were found to be malignant and 121 adenoma. Of these, all malignant nodules increased in size during follow-up, with a mean growth of 5.8 cm/year. A growth rate of 3 mm/year distinguished adenomas from malignant nodules with sensitivity and specificity of 100%.25 Repeating the same imaging modality (generally CT scan) is encouraged to assess for changes in the mass.6 The American College of Radiology better defines imaging protocols, including measurement of density and contrast washout, as well as the importance of clinical correlation when determining surveillance regimens for AIs <4 cm. Specifically, in 2-4 cm AIs, an adrenal CT protocol should be used at diagnosis to better characterize the lesion, while lesions <2 cm may have an adrenal CT protocol at one year to document stability. Chomsky-Higgins et al. performed a cost-effectiveness analysis of surveillance of adrenal masses 1-4 cm with an initial normal workup including non-contrasted CT showing a mass < 10 HU, no suspicious features, and negative hormonal evaluation. They determined it is most effective to perform a single follow up at one year with non-contrasted CT and biochemical evaluation.26 It may be reasonable to consider excision in those adrenal masses with indeterminate radiologic features that grow at least 0.8 cm during 3-12 month follow-up.13 Post-excision surveillance for AIs >4 cm remains poorly defined and is based on the histological diagnosis at the time of extirpation.27 Although recommendations for radiologic follow-up of nonfunctional adrenal masses vary based on the organization, it is important to consider that frequent adrenal imaging is associated with additional cost, anxiety, and exposure to radiation, which may theoretically induce cancer at an estimated rate similar to the chance of developing adrenal malignancy.13
Limitations of this study include its retrospective nature, which has the potential to introduce biases in data collection and analysis. Although it is a multi-institutional study, both institutions are within the same geographic region with a notably homogenous population, limiting its applicability to wider groups. Also, the lack of surveillance data within the population limits our ability to detect recurrence amongst participants and growth of non-operative AIs. Similarly, surgical patients are the only subjects included, limiting the ability to assess cancer risk in a more specific population of patients on active surveillance who undergo resection following AI growth. However, the above study design achieves the goal of assessing pathological outcomes in patients undergoing adrenalectomy for incidentally detected adrenal masses. Addition of patients to our study data from institutions with more diverse populations in terms of race and age as well as surveillance data provide future directions for this research.