Clinical and Radiological Features of Atypical Adrenal Masses—A Multicenter Retrospective Study

Abstract

Context

The natural history and malignant potential of cases classified as atypical adrenal masses (AAMs) are unknown.

Objective

To describe the radiological characteristics and clinical outcomes of AAMs.

Design and Participants

A multicenter retrospective study. Patients ≥18 years old with AAMs [diameter of 10-39 mm on first imaging study and pre-contrast attenuation of >10 Hounsfield units (HU) on computed tomography] were studied. We excluded adrenal metastasis, pheochromocytoma, sarcoma, lymphoma, infiltrative lesions, and adrenal hemorrhage, as well as patients with genetic predisposition to adrenal neoplasms. Data are presented as percentages and median values with interquartile ranges (IQRs).

Results

We included 217 patients with 224 adrenal masses (61.3% women); the median age was 58 years (IQR 49-65 years). The median size was 20.5 mm (IQR 15-27 mm), with a median precontrast attenuation of 23.5 HU (IQR 17-30 HU). The median AAM growth rate was 0.3 mm/year (IQR 0-1.8 mm/year). Seventy-one masses (31.7%) underwent pathological evaluation. Adrenal adenoma (n = 38; 17%) and adrenocortical carcinoma (ACC) (n = 25; 11.2%) were the 2 most common diagnoses. For the adenomas, the growth rate was 0.3 mm/year (IQR 0-2.3 mm/year) and for ACCs, the growth rate was 12.9 mm/year (IQR 3.5-22 mm/year). The absolute contrast washout was >60% in 5 out of 7 (71.4%) ACC cases. The best growth rate cutoff for predicting malignancy was 2.68 mm/year (area under the curve 0.939; sensitivity 87.5%, specificity 88.8%).

Conclusion

AAMs carry significant malignant potential, and long-term follow-up is warranted when surgery is not pursued. Contrast washout is not reliable in predicting malignant potential of AAMs.

Adrenal incidentalomas (AIs) are increasingly found on imaging studies performed to evaluate other medical conditions (1). The prevalence of AIs has increased from 0.6% in 1982 to 7.3% in 2020 (2, 3). Over the past 2 decades, the incidence of AIs has increased 10-fold, likely due to the increased use of cross-sectional imaging (4). AIs pose multiple medical challenges including the burden and uncertainties arising from hormonal assessment and imaging surveillance. Most importantly, concerns about the malignant potential of these masses often leads to either surgical resection or long-term surveillance, resulting in potential increased radiation exposure as well as significant costs (1).

Most AIs are lipid-rich adrenal adenomas (LRAAs), which have a high intracytoplasmic fat content (5). Despite the pathological origin of the term LRAAs, these masses are often defined radiologically as lesions with a pre-contrast attenuation of ≤10 Hounsfield units (HU) on computed tomography (CT) scan and demonstrate a loss of signal on out-of-phase imaging on magnetic resonance imaging (MRI). LRAAs are associated with a benign clinical course and have an extremely low risk of malignancy, if any (6-10).

AIs with a precontrast attenuation >10 HU are often labeled atypical adrenal masses (AAMs) (1, 9-12). This label encompasses multiple pathological diagnoses, including lipid-poor adrenal adenomas, which comprise up to 30% of adrenal adenomas (13, 14), pheochromocytomas, and secondary adrenal metastases, among others.

There are limited data about the natural history of AAMs and their malignant potential. In 2 previous retrospective reports, AAMs were associated with a low risk of malignancy; however, these reports had limitations regarding the variability of follow-up and imaging data (15-17). Thus, to fill this important clinical knowledge gap and potentially improve the management of these masses, we explored the growth patterns and the malignant potential of AAMs.

Methods

This was a multicenter retrospective cohort study among member centers of the American-Australian-Asian Adrenal Alliance, which is an international adrenal research consortium. The Institutional Review Board of The University of Texas MD Anderson Cancer Center approved this overall study (PA12-0933), and informed consent was waived. Each center had their own institutional review board approval.

Patients ≥18 years old with AAMs, diagnosed from January 2000 through January 2021, were included in this study to allow for sufficient follow-up time. Adrenal masses 10 to 39 mm in diameter on the first available imaging scan and with precontrast attenuation >10 HU were included. Other inclusion criteria were a minimum follow-up duration of 6 months between the first and last available imaging studies or having a pathological evaluation (surgical resection or biopsy) at any time. We excluded cases of suspected or proven adrenal metastasis, pheochromocytoma, sarcoma, lymphoma, myelolipoma, infiltrative lesions, and adrenal hemorrhage. We also excluded patients with a personal/familial history of genetic predisposition to adrenal neoplasms such as Li-Fraumeni syndrome, Lynch syndrome, and multiple endocrine neoplasia type 2 or familial syndromes associated with hereditary pheochromocytoma/paraganglioma.

Clinical characteristics (patient sex, age, and race/ethnicity, mode of lesion discovery, and clinical presentation), laboratory results (levels of serum potassium, serum glucose, serum cortisol, basal and plasma adrenocorticotropic hormone, salivary cortisol, 1-mg overnight dexamethasone suppression test, 24-hour urinary cortisol, serum dehydroepiandrosterone sulfate, 11-deoxicortisol, testosterone, estradiol, aldosterone, plasma renin activity, plasma metanephrine and normetanephrine, and urinary metanephrine and normetanephrine), and imaging data were collected from the medical records.

We reviewed the imaging characteristics of these adrenal masses, including size, location, attenuation, heterogeneity, calcification, and necrosis on CT, as well as signal drop-out on T1 in and out of phase imaging and T2 hyperintensity on MRI. The absolute contrast enhancement washout and the relative contrast enhancement washout were calculated when available (18-20). The absolute growth of the adrenal mass was defined as the change in size (in mm) between the initial evaluation and the last available imaging study (7). The tumor growth rate was calculated by dividing the absolute growth by the total time (in years) between the first and the last available imaging study. The duration of follow-up was the period between the first imaging study and the last available imaging study or between the first imaging study and the date of pathological evaluation (surgical resection or biopsy).

The definitive histopathological diagnosis was considered when patients underwent pathological evaluation achieved by biopsy or adrenal mass resection.

The patients’ follow-up information was collected by reviewing medical records per the assessment of the treating team.

Each institution collected data using Research Electronic Data Capture software hosted locally and shared the d-identified data with the MD Anderson Cancer Center team (Fig. 1).

Figure 1.

CONSORT flow diagram of the study.

Statistical Analysis

Wilcoxon rank-sum tests were used to compare the distribution of continuous parameters between groups. Fisher's exact tests were used to compare the distribution of categorical variables between groups. Receiver operating curve analysis was used to assess the performance of the tumor growth rate to predict malignancy. Performance was assessed by the area under the curve. The optimal cutoff was chosen by maximizing the sum of sensitivity and specificity. The sensitivity, specificity, positive predictive value, negative predictive value, and accuracy were reported for the analysis. Logistic regression was used to assess the association between malignancy and characteristics of interest. All statistical analyses were performed using R version 4.3.1. All statistical tests used a significant level of 5%. No adjustments for multiple tests were made.

Results

Clinical and demographic characteristics were collected from 217 patients who met the inclusion and exclusion criteria and are summarized in Table 1. Women comprised 61.3% (133 cases) of the cohort, and the median age was 58 years (IQR 49-65 years). The 217 patients had 224 AAMs that met the inclusion and exclusion criteria and were analyzed in this study (Fig. 1). The AAMs were incidentally discovered in 154 (71%) patients and were found during cancer surveillance for a history of extra-adrenal malignancy in 46 (21.2%) patients. These 46 masses did not undergo pathological evaluation; however, according to the local treating team's assessment, based on clinical and radiological evaluation, these masses were not suspected metastasis. Thirty-two (14.7%) patients were initially evaluated due to discovery of hormonal overproduction and 17 (7.8%) due to pain or presumed mass effect. The median follow-up period was 44.4 months (IQR 17.3-76.6 months) for the whole cohort and 52.7 months (IQR 26.2-82.6 months) for those who did not have an adrenal mass resection or biopsy.

Table 1.

Clinical and demographic characteristics of 217 patients with atypical adrenal masses

Variables	n = 217
Sex (%)
Female	133 (61.3)
Male	84 (38.7)
Median age, years (IQR)	58 (49-65)
Race (%)
American Indian/Alaska Native	2 (0.9)
Asian	8 (3.7)
Black	36 (16.6)
Hispanic	16 (7.4)
White	151 (69.6)
NA	4 (1.8)
Mode of discovery (%)
Incidentally	154 (71)
Hormonal overproduction	32 (14.7)
Pain or mass effect	17 (7.8)
NA	14 (6.5)
Type of first imaging study (%)
CT	192 (88.5)
MRI	22 (10.1)
PET CT	3 (1.4)
Unilateral or bilateral (%)
Unilateral	195 (89.9)
Left	111 (56.9)
Right	84 (43.1)
Bilateral	22 (10.1)
Median follow-up, months (IQR)	44.4 (17.3-76.6)

Abbreviations: CT, computed tomography; IQR, interquartile range; MRI, magnetic resonance imaging; NA, not available; PET, positron emission tomography.

Masses were unilateral in 195 (89.9%) cases and on the left side in 111 (56.9%). The median size was 20.5 mm (IQR 15-27 mm), and the median pre-contrast attenuation on CT scan was 23.5 HU (IQR 17-30 HU). Of the 197 masses that were first evaluated with a CT scan, 45 (22.8%) were heterogeneous, 5 (2.5%) contained calcifications, and 2 (1%) had necrosis. Absolute contrast enhancement washout was reported in 104 adrenal masses and was >60% in 74 (71.1%) cases. Fifty-seven masses were evaluated with MRI scan at any time of follow-up; 34 (59.6%) had a loss of T1 signal dropout, and 15 (26.3%) had a hyperintensity on T2. The median growth of AAMs was 2 mm (IQR, 0-6 mm), and the median tumor growth rate was 0.3 mm/year (IQR 0-1.8 mm/year) (Table 2).

Table 2.

Tumor characteristics of 224 atypical adrenal masses and subgroups (imaging surveillance, confirmed adenomas, and adrenocortical carcinomas)

Variables	Total cohort	Imaging surveillance	Pathologically evaluateda
	224 AAMs	153 AAMs currently under surveillance	38 adenomas	25 ACC
Median size, mm (IQR)	20.5 (15-27)	20 (15-25)	21.5 (15-28)	24 (17-30)
Median attenuation, HU (IQR)	23.5 (17-30)	22 (16-29)	24 (18-31.5)	32 (26-37)
Left (%)	125 (55.8)	87 (57.1)	22 (57.9)	9 (36)
Right (%)	99 (44.2)	66 (42.9)	16 (42.1)	16 (64)
Absolute washoutb (%)	104 (100)	74 (100)	19 (100)	7 (100)
=<60%	30 (28.8)	20 (27)	5 (26.3)	2 (28.6)
>60%	74 (71.2)	54 (73)	14 (73.7)	5 (71.4)
Relative washoutb (%)	95 (100)	68 (100)	17 (100)	6 (100)
=<40	23 (24.2)	12 (17.6)	5 (29.4)	4 (66.7)
>40	72 (75.8)	56 (82.4)	12 (70.6)	2 (33.3)
Absolute growth, mm (IQR)	2 (0-6)	1 (0-4)	1 (0-5.8)	32 (14-65)
Median variation in size from baseline, (IQR, %)	8.7 (0-30.8)	5 (−3-20)	3.1 (0-32.3)	170.6 (57.7-291.3)
Growth rate, mm/year (IQR)	0.3 (0.0-1.8)	0.2 (0.0-0.9)	0.3 (0.0-2.3)	12.9 (3.5-22.0)
Median follow-up, months (IQR)	44.4 (17.3-76.6)	52.7 (26.2-82.6)	14.4 (5.2-44.8)	34.3 (16.5-59.7)

Abbreviations: AAM, atypical adrenal masses; ACC, adrenocortical carcinoma; HU, Hounsfield unit.

^a Data from the most frequent diagnoses are shown.

^b Total number of available cases.

Pathological data were available for 71 (32.7%) masses; patients underwent adrenalectomy for 64 (90.1%) of the masses and biopsy for 7 (9.9%) (Table 2). The pathological evaluation was done in 36 cases (50.7%) due to tumor growth during follow-up, in 19 cases (26.8%) due to concerns about the radiological features on imaging, in 13 cases (18.3%) due to tumor growth and development of hormonal overproduction, and in only 3 cases (4.2%) due to hormonal overproduction. The median tumor growth rate in this group was 2.5 mm/year (IQR 0-12.4 mm/year). Adrenal adenoma was the definitive diagnosis in 38 (17%) cases, adrenocortical carcinoma (ACC) in 25 (11.2%) cases, adrenal hyperplasia in 4 (1.8%), adrenal neoplasm of uncertain behavior in 3 (1.3%), and benign anastomosing hemangioma in 1 (0.4%) case. The median tumor growth rate of adrenal adenomas was 0.3 mm/year (IQR 0-2.3 mm/year), whereas the median tumor growth rate of ACCs was 12.9 mm/year (IQR 3.5-22 mm/year) (Fig. 2). The median time between the initial evaluation of AAMs and the surgical resection of ACC cases was 34.3 months (IQR 16.5-59.7 months). Table 3 shows the clinical and radiological characteristics of 25 confirmed ACC cases.

Figure 2.

Illustration of tumor growth rate of subgroups (confirmed adrenocortical carcinomas, adenomas, and adrenal masses in surveillance).

Table 3.

Clinical and radiological characteristics of 25 confirmed ACC cases

Case	Sex	Age	Side	Size (mm)	Hormonal hypersecretion	Precontrast attenuation (HU)	Absolute Washout (%)a	Absolute growth (mm)b	Tumor growth (mm/year)c	Follow-up time (months)d
1	F	44	Left	17	Yes	35	NA	29	3.5	101.2
2	M	75	Right	27	No	15	61.4	13	1.8	87.7
3	F	49	Right	24	No	19	NA	29	2.7	128.8
4	F	33	Right	38	No	35	NA	14	7.8	22.0
5	M	56	Right	17	Yes	34	NA	133	14.3	112.4
6	F	44	Right	10	No	55	NA	10	8.5	25.5
7	M	69	Left	11	No	29	78.1	43	16.9	34.2
8	F	57	Right	19	Yes	17	NA	81	9.0	108.6
9	F	58	Right	34	Yes	18	NA	90	22.0	54.6
10	F	47	Left	25	No	30	NA	65	21.9	35.6
11	M	50	Left	23	Yes	32	NA	67	28.3	2.4
12	M	38	Right	28	No	37	NA	32	48.0	8.6
13	F	56	Right	38	No	59	NA	52	79.0	7.9
14	F	55	Left	26	No	29	63.8	15	3.1	59.5
15	F	72	Left	14	Yes	95	68	44	18.9	29.7
16	F	56	Right	22	Yes	19	NA	58	12.7	59.7
17	F	68	Right	27	No	49	NA	12	17.9	16.5
18	F	58	Right	34	Yes	28	NA	1	0.3	54.6
19	M	56	Right	35	No	58	NA	75	100.0	9.0
20	M	28	Right	30	Yes	44	NA	8	12.7	8.5
21	F	44	Left	17	Yes	37	NA	29	3.5	101.5
22	F	45	Left	16	No	26	15	96	55.4	21.3
23	F	50	Left	19	Yes	29	69.4	3	1.2	49.8
24	F	59	Right	36	No	35	53.4	28	115.9	6.7
25	M	79	Right	12	NA	20	NA	35	13.1	34.3

Abbreviations: HU, Hounsfield unit; NA, not available.

^a The absolute washout was calculated when available, at any time during the follow-up.

^b The absolute growth of the adrenal mass was defined as the change in size (in mm) between the initial evaluation and the last available imaging study.

^c The tumor growth rate was calculated by dividing the absolute growth by the total time (in years) between the first and the last available imaging study.

^d The duration of follow-up was the period between the first imaging study and the date of pathological evaluation.

A total of 153 (68.3%) adrenal masses remained under surveillance at the end of the study period. These masses had a median follow-up time of 52.7 months (IQR 26.2-82.6) and a median tumor growth rate of 0.2 mm/year (IQR 0-0.9 mm/year).

The receiver operating curve analysis of the tumor growth rate for predicting malignancy growth rate had an area under the curve of 0.939, which indicates that the growth rate is an excellent predictor of malignancy. The maximum sum of sensitivity and specificity is achieved at a growth cutoff of 2.68 mm/year (sensitivity 87.5%, specificity 88.8%, positive predictive value 50%, negative predictive value 98.2%, and accuracy 88.6%) (Fig. 3).

Figure 3.

Receiver operating characteristic curve for using the growth rate to predict malignancy.

Laboratory results at baseline evaluation are available in Supplementary Table S1 (21). None of the patients had metanephrine or normetanephrine secretion at baseline. Hormonal overproduction developed during the follow-up period in 23 (10.6%) patients as follows: cortisol in 12 (5.5%) after a median time of 62.2 months of follow-up (IQR 36.5-97.0), aldosterone in 4 (1.8%) after a median time of 76.2 months of follow-up (IQR 57.7-94.8), sex hormones in 3 (1.3%) after a median time of 26.7 months of follow-up (IQR 18.2-43.8), and mixed hormonal secretion in 4 (1.8%) patients after a median time of 65.9 months of follow-up (IQR 53.9-84.5). Of these 23 AAMs that developed hormonal overproduction, 16 were resected or biopsied for pathologic diagnosis (9 were definitive ACCs, 7 were adrenal adenomas), and 7 are still under imaging surveillance.

The univariate logistic regression analysis for malignancy showed that higher attenuation on CT [odds ratio (OR) 1.03, 95% confidence interval (CI): 1.01,1.06; P = .0053), higher tumor growth rate (OR 3.17, 95% CI: 1.62, 6.21; P = .0008), right side laterality (OR 2.51, 95% CI: 1.06, 5.97; P = .0324), and hormonal overproduction (OR 8.51, 95% CI: 3.1, 23.3; P < .0001) were significantly associated with higher probability of malignancy.

Discussion

The challenge in managing AIs in clinical practice is to identify which have a high enough malignant potential to warrant closer follow-up or surgical resection. The optimal imaging study for AIs and the frequency of imaging have not been established, and recommendations for hormonal and imaging evaluations vary between different guidelines (1, 22-24). The major limitation of all these guidelines stems from the lack of strong data to establish the natural history of AIs, especially AAMs. The cost-effectiveness of surveillance imaging is questionable, especially when patients’ clinical and biochemical phenotypes are reassuring and the adrenal mass has precontrast HU < 10.

This study shows that AAMs were associated with significant malignant potential, and long-term follow-up is warranted when resection of the adrenal mass is not pursued. The tumor growth rate is valuable for assessing and following up on these lesions and might help to define their management. Moreover, CT washout is not reliable to predict malignancy in AAMs.

Epidemiological studies of AIs have shown that their etiology includes benign and malignant pathologies. The frequency of etiologies varies across studies due to different inclusion and exclusion criteria, hormonal and imaging assessments, and follow-up periods (1, 4, 9, 10, 16, 25-31). Certain features such as the appearance (homogenous vs heterogenous), attenuation on unenhanced CT, contrast washout characteristics (absolute and relative contrast enhancement washouts), lesion size, and hormonal functionality of these tumors are used to guide the decision between observation and surgical intervention for the treatment of AIs (1, 22, 32-34).

In our cohort, we followed 217 patients with 224 AAMs for a median time of about 45 months. The proportion of lesions with benign behavior was 87.1% (68.3% were adrenal lesions with benign behavior based on the slow tumor growth and stability of the lesion but without pathology, 17% were confirmed adrenal adenomas, and 1.8% were confirmed adrenal hyperplasia).

Change in nodule size has been predicted to be a potential worrisome feature for AIs. In a cohort of 3043 patients with AIs, a change in adenoma size compared to baseline occurred in 6.3% of cases, and an increase ≥10 mm was reported in only 2.5% of those AIs (9). Interestingly, in a review of the follow-up of 4121 patients with AIs, none developed adrenal cancer (9). In our cohort, 140 (62.5%) AAMs had an increase in size, and an increase ≥10 mm occurred in 40 (17.8%) AAMs. The median tumor growth rate of the confirmed adrenal adenomas was similar to that of the 153 adrenal lesions under imaging surveillance.

Malignancy incidence in AIs cohorts varies from 0% to 7.14% (15, 16, 25-27, 29). The evaluation of a cohort with 106 patients with a documented history of cancer (nonadrenal cancer) and an adrenal mass showed a risk of malignancy of 47.2% (35). None of the reported metastatic tumors or adrenal malignancies had an attenuation lower than 10 HU on CT, corroborating the defined threshold attenuation of ≤10 HU used to rule out nonadenomas (6-8, 35). In our study, 71 (31.7%) adrenal masses were pathologically evaluated, and ACC was the definitive diagnosis in 11.2% of all cases. This contrasts with a previous study of 321 atypical AIs in which only 6 (1.8%) cases had histologic evaluation, and none of those masses were malignant (15). In that study, the average adrenal mass size was 21 mm, which was similar to ours; however, no lesion increased during follow-up imaging (15). In our cohort, only 40 (17.8%) lesions did not change in the size compared to the baseline evaluation. One reason for the differences is that the study selected atypical AIs but benign-appearing lesions and no clinical suspicion of hormonal hyperfunctioning, and a minimum of 1 year of stability on imaging or 2 years of stability on clinical follow-up was required, which contrasts with our inclusion criteria.

A retrospective review of a surgical database reported 136 adrenal cases; 111 (91.6%) were benign lesions, 25 were malignant lesions [19 (14%) were metastatic lesions, and 4 (2.9%) were adrenocortical carcinomas], and there was no difference between the benign and malignant masses regarding the initial size at presentation (7). The malignant lesions presented a median tumor growth of 8 mm at 3 to 12 months of follow-up, while benign lesions presented a median growth of 1 mm in the same period, and the tumor growth rate of 16 mm/year had a sensitivity of 50% and specificity of 83.1% for predicting malignancy in this subset group (7). We reported a lower tumor growth rate cutoff for predicting malignancy than previously mentioned, which might result from our homogeneous malignant sample based on confirmed ACC cases (n = 25) and a more extended median follow-up period. In a previous retrospective study, 25 ACC cases with previous imaging studies were evaluated (36). Five (20%) patients had normal adrenal glands, while 20 (80%) had a preexisting adrenal lesion (median size of 28 mm and median precontrast attenuation of 36 HU) (36). The median time from initial imaging of the adrenal lesion to ACC diagnosis was 20 months (36). Another retrospective review of 20 patients with ACC who had preexisting lesions diagnosed as AIs (mean size of 32 mm) showed a mean time from the first CT scan to ACC diagnosis of 44.1 months (37). The mean increase of those lesions was 19 mm/year, and 35% had an average growth rate of <10 mm/year (37). In our cohort, we reported a median time of 34.3 months from the first imaging study evaluation until the pathological diagnosis of ACC. We had access to the absolute contrast enhancement washout of 7 ACC cases, and interestingly, it was >60% in 5 (71.4%) cases. This highlights that contrast washout does not predict the malignant potential of AAM.

The tumor growth rate of the confirmed ACC cases was 12.9 mm/year, which was significantly higher than the tumor growth rate of the benign lesions in our cohort (P < .001) (Fig. 2). It is known that the velocity of tumor growth is a characteristic of malignancy and is related to decreased survival in ACC patients (38). Here, we reported that lesions characterized as AAMs carried a reasonable risk of malignancy.

In a review of 16 158 cases of AIs, the prevalence of hormonal overproduction was 27.5% (31). The most frequent conditions were autonomous cortisol secretion/mild autonomous cortisol secretion detected in 11.7%, primary aldosteronism in 4.4%, pheochromocytoma in 3.8%, and Cushing syndrome in 3.1% of cases (31). The lipid-rich AIs have a low tumor growth rate during follow-up, and the conversion rate to mild autonomous cortisol secretion or clinical hypercortisolism varies from 0% to 22%, depending on the duration of follow-up (9, 10, 29, 30). In our AAMs cohort, we showed that hormonal overproduction developed during follow-up in 10.6% of our cohort (n = 23), and the median time for the development was 4.6 years (IQR 3.1-7.7) after the first imaging study. This highlights the necessity of a long-term follow-up when surgery is not pursued.

The limitations of our study may include but are not limited to its retrospective nature; the description of the images was obtained from medical records and the images were not reviewed, the resected adrenal masses were not independently evaluated, the inability to do secondary pathological review though most cases were seen in major referral centers with known pathological expertise for this report, the potential for differences in CT techniques due to multicenter design, the imaging intervals that were defined based on physicians’ discretion, and a possible selection bias since our institutions are tertiary referral centers. These limitations add to the importance of having future prospective study that can be conducted consistently, allowing for a more accurate calculation of the growth rate of atypical adrenal masses.

Conclusion

Although our cohort might have referral bias, atypical adrenal masses carry significant malignant potential, and long-term follow-up is warranted when surgical resection is not pursued. The tumor growth rate is valuable for assessing and monitoring AAMs and may help define their management. Moreover, contrast washout does not predict malignant potential in AAMs.

This study is the foundation for a multicenter, long-term prospective study and provides insights into the natural history, management, and follow-up of atypical adrenal masses.

Funding

National Institutes of Health/National Cancer Institute Cancer Center Support grant (award number P30 CA016672) and Biostatistics Resource Group.

Disclosures

K.K.-V.: Advisory board, HRA Pharma; H.K.G.: Consultant, Corcept Therapeutics. All other authors declare no competing interests.

Data Availability

Datasets analyzed during the current study are not publicly available but are available from the corresponding author on reasonable request.