ENDOCRINE SOCIETY
OXFORD
Two Cases of Adrenal Malignancy With Macroscopic Fat
Tom Wilkinson,10 Penny Hunt,1,2 and Alexandra McHaffie3
1Department of Endocrinology, Christchurch Hospital, Christchurch 8011, New Zealand
2Department of Medicine, University of Otago (Christchurch), Christchurch 8011, New Zealand
3Department of Radiology, Christchurch Hospital, Christchurch 8011, New Zealand
Correspondence: Tom Wilkinson, MBChB, Department of Endocrinology, Christchurch Hospital, 2 Riccarton Ave, Christchurch 8011, New Zealand. Email: Thomas.wilkinson@cdhb.health.nz.
Abstract
The presence of macroscopic fat on computed tomography (CT) imaging has been traditionally regarded as an indication that an adrenal lesion is likely to be a benign myelolipoma, for which further investigation is not usually required. Two cases are described where an adrenal lesion was eventually found to be malignant on histology (adrenocortical carcinoma in the first case, undifferentiated sarcoma in the second case), despite the presence of macroscopic fat on CT. In both cases there were other clinical and radiological indicators of potential malignant pathology. These cases add to increasing awareness in the literature that malignant adrenal tumors may rarely contain macroscopic fat, emphasizing a need for clinical vigilance.
Key Words: adrenocortical carcinoma, sarcoma, macroscopic fat Abbreviations: ACTH, adrenocorticotropic hormone; CT, computed tomography.
Adrenal lesions are a common radiological finding. Although the vast majority are benign [1], some represent important malignancies. The presence of macroscopic fat on CT has been traditionally regarded as an indicator of benign path- ology; however, placing too much diagnostic weight on one radiological marker may lead to misleading conclusions. Two recent cases at our institution illustrate this potential diagnostic pitfall, including an instance of a rare malignancy not previously described as manifesting macroscopic fat.
Case 1
A 62-year-old woman, initially presenting with abnormal liver function tests, was found to have a well-circumscribed lobu- lated heterogeneous mass in the right adrenal gland on CT ab- domen measuring 4.9 cm in maximal dimension. Three hypodense foci, measuring up to 10 mm, were identified within the mass and thought to represent macroscopic fat on the basis of being visible to the eye and of similar density to fat seen else- where. The mass was therefore initially reported as consistent with adrenal myelolipoma. Later retrospective review meas- ured density for 2 of these foci as -60 and -90 Hounsfield units, consistent with fat, and calculated this fat as comprising < 2% of the mass (using Vitrea software, Canon Medical Informatics, Inc.). Representative images are presented in Fig. 1. No prior imaging was available for comparison.
Results of functional testing included a significantly ele- vated 24-hour urine cortisol (3997 nmol [1445 mcg]; nr 100-400) and suppressed adrenocorticotropin hormone (ACTH) (< 0.2 pmol/L), consistent with adrenal-based Cushing syndrome. Plasma aldosterone was below the limit
of quantification. 24-hour urine metanephrine, plasma testosterone, and plasma dehydroepiandrosterone sulfate (DHEA-S) were within the normal range.
Past medical history included hypertension and type 2 dia- betes mellitus (possibly secondary to Cushing syndrome). There was no known history of any hematological disorder, nor any abnormality on initial full blood count.
CT abdomen also identified multiple hepatic lesions, unable to be characterized in part due to marked background hepatic steatosis. A subsequent hepatic magnetic resonance imaging (MRI) confirmed the presence of multiple lesions for which the differential diagnosis included metastases, adenomata, or multifocal hepatocellular carcinoma. The adrenal lesion was visualized on this MRI, with no change in size from the CT performed 26 days prior. The areas of interest within the lesion demonstrated high T1 and T2 signal with loss of sig- nal on fat-suppression sequences and chemical shift artifact at their margins on opposed-phase imaging, consistent with macroscopic fat [2].
A hepatic core biopsy was obtained, with morphology and immunohistochemistry consistent with metastatic adrenocorti- cal carcinoma. The patient proceeded to palliative treatment with mitotane (4.5 g daily in divided doses), chemotherapy (3 cycles of doxorubicin, etoposide, and cisplatin), and trans- arterial embolization of hepatic metastases. Cortisol excess re- quired additional management with metyrapone (titrated to 4.5 g daily in divided doses) and ketoconazole (titrated to 1200 mg daily in divided doses). There was an initial response, with normalization of urine cortisol and stability of lesions on surveillance CT; however, the disease subsequently progressed and she died 14 months after the initial diagnosis was made.
A
B
0
A
Case 2
A 72-year-old man presented initially with abdominal pain and was found on CT to have a retroperitoneal lesion, con- tacting and possibly arising from the left adrenal gland. The lesion measured 7.6 cm in maximal dimension; however, there was associated acute retroperitoneal hemorrhage, limit- ing assessment of the true size and contours of the underlying mass. No prior imaging was available for comparison. A sin- gle hypodense focus was identified, measuring 7 mm and on later retrospective review calculated as comprising < 1% of the overall lesion. The focus was visible to the eye and of simi- lar density to fat elsewhere (later retrospectively measured as -20 Hounsfield units) and was therefore thought to represent macroscopic fat. Representative images are presented in Fig. 2. Due to the presence of macroscopic fat the lesion was initially thought to be most compatible with a benign adrenal myelolipoma. Liposarcoma and adrenocortical carcinoma were acknowledged in the initial report as possible, but un- likely, diagnoses.
Plasma normetanephrine was elevated (2716 pmol/L [497 pg/mL]; nr <900) in the context of acute illness. Cushing syndrome and primary aldosteronism were not
A
A
B
0
S
D
0
0
formally excluded. Past medical history included hypertension controlled on candesartan, and hemochromatosis managed with regular blood donation. At presentation blood pressure was 127/75 mmHg and full blood count was normal.
The patient presented again 1 month later with ongoing ab- dominal pain. Repeat CT demonstrated a new small central hyperdense component to the lesion, suspicious for active ex- travasation. He proceeded to embolization of the left inferior adrenal artery. He was seen again 1 month after embolization, at which time a further CT showed the lesion to have enlarged to 9.0 cm.
The patient proceeded to open left adrenalectomy, left nephrectomy, and colonic resection. Histology was consistent with an undifferentiated spindling sarcoma, possibly repre- senting a malignant peripheral nerve sheath tumor. Areas of tumor necrosis were noted. Intra-tumor macroscopic fat was not specifically identified.
Surveillance CT 4 months after surgery unfortunately showed widespread disease recurrence with extensive retro- peritoneal and peritoneal metastases. The patient declined palliative chemotherapy and died 7 months after the mass was first identified.
Discussion
The presence of macroscopic fat on imaging of an adrenal le- sion has been traditionally thought to indicate benign myelo- lipoma. American College of Radiology recommendations for management of incidental adrenal masses regard this finding as essentially diagnostic for myelolipoma, with no additional workup or follow-up imaging generally required, although that recommendation was presented with a caveat that it should not be termed a formal “guideline” and that “Radiologists should feel comfortable deviating from the al- gorithm” [3].
A recent systematic review described 7 cases of adrenocort- ical carcinoma where macroscopic fat was seen on preopera- tive imaging [4]. Four cases had a small proportion of fat (< 5%); percentage fat was not reported for the remaining 3 cases. Three cases reported tumor size > 6 cm; size was not reported for the remaining 4 cases. The authors suggested consideration of follow-up imaging or biopsy as appropriate for patients with large (> 6 cm) symptomatic tumors with < 5% fat, although acknowledging significant uncertainty. Of note, our first case (maximal tumor dimension 4.9 cm) would not meet these criteria, while application to our second case would be problematic due to the presence of acute hemorrhage precluding calculation of the size of the under- lying mass.
A 2009 pathologic case series of 3 patients with adrenocort- ical neoplasms, classified as either adrenocortical carcinoma or of uncertain malignant potential, described foci of myeloli- pomatous and lipomatous metaplasia [5]. This phenomenon may account for the radiologic finding of macroscopic fat in rare cases of adrenocortical carcinoma, particularly noting that in all such described cases (including our case) the propor- tion of fat was < 5% where quantified.
Although adrenocortical carcinoma has not been described with fat > 5%, the yield of using percentage fat to discrimin- ate between myelolipoma and adrenocortical carcinoma is uncertain. Histologically, myelolipoma comprises a mix- ture of mature fat and hematopoietic cells, however the rela- tive proportion of these elements can vary significantly. Consequently, radiological appearances range from tumors that appear to almost entirely consist of macroscopic fat to tumors where visible fat is nonexistent [6].
In both cases described here, macroscopic fat comprised a very small component of the lesion, raising a question of whether, in retrospect, the possibility of malignancy could have been given greater weight at the outset. However, the rarity of adrenocortical carcinoma with macroscopic fat, compared to the well-described phenomenon of myelolipo- ma containing only a small proportion of fat, suggests that myelolipoma remains the more likely diagnosis even where fat is < 5%. Radiologists must weigh providing an exhaust- ive differential diagnosis, including very rare entities, with the potential risks of this approach, including harm and costs generated by overinvestigation. Of note, the initial radiologist report for case 2 suggested malignancy in the dif- ferential diagnosis.
Adrenal tumor size can be used as an indication of the like- lihood of malignancy; however, it is unlikely to be a reliable discriminator when specifically differentiating adrenocortical carcinoma and myelolipoma. One study, using data from pa- tients undergoing adrenalectomy, estimated the population rate of malignancy as 10% in tumors ≥4 cm, 19% in tumors ≥ 6 cm, and 47% in tumors ≥ 8 cm [7]. This is in con- trast to a review estimating the average size of myelolipomata described in the literature as 10.2 cm [8]. Consequently, rec- ommendations to consider adrenalectomy for tumors ≥ 4 cm cannot be readily applied where myelolipoma is thought likely [2].
Our second case was found to have a poorly differentiated sarcoma. We are not aware of previous descriptions in the lit- erature of this manifesting as an apparent adrenal tumor with macroscopic fat. Even if the reporting radiologist had pro- vided an exhaustive differential diagnosis for the initial CT, it would almost certainly have not contained the final diagno- sis in this case.
These factors collectively suggest caution in using isolated radiology findings (macroscopic fat, tumor size) to classify an adrenal tumor as likely benign, without factoring in the broader clinical context. Both cases described had elements to their presentation raising suspicion of malignancy.
Case 1 had evidence of endocrine functionality. This is well- described in adrenocortical carcinoma, with 62% reported to be functional in one review, of which Cushing syndrome was the most common abnormality [9]. Of note, where an adreno- cortical neoplasm contains myelolipomatous metaplasia, it is thought that the myelolipomatous tissue can promote cortical or medullary hyperfunction [5]. Adrenal myelipomata, which comprise adipocytes and hematopoietic cells, are less likely to be associated with endocrine abnormalities, although adrenal hormonal production has been described in 7.5% of cases, while an additional 10% of cases occur in the setting of congenital ad- renal hyperplasia, possibly as a result of ACTH excess [8].
Of more significance, both of our patients had clinical suspicion of malignancy prior to histological diagnosis, with hepatic lesions suspicious for metastases in case 1 and possible growth of the mass in case 2. Although fre- quently large in size, adrenal myelolipomata are typically slow growing. One series of 69 patients with radiographic- ally diagnosed myelolipoma described growth in 11, at a median rate of 0.16 cm/year [10]. In contrast, the mass was seen to change in size by 1.4 cm over 2 months in our second case, although interpreting this finding is com- plicated by hemorrhage.
It is important to note that histology did not identify macro- scopic fat in case 2, and that macroscopic fat was unable to be his- tologically confirmed in case 1 (adrenalectomy was not clinically indicated). There were, however, multiple radiological features present that were consistent with macroscopic fat. Although tu- mor necrosis or cystic change may result in hypodense foci, radio- logically these would be expected to manifest as fluid density rather than fat density. Ultimately, the decision to proceed to adrenalectomy due to suspicion of malignancy rests on clinical and radiological factors, rather than on histological findings which are only known once adrenalectomy has occurred.
Learning Points
. The radiological finding of macroscopic fat in an adrenal lesion is highly suggestive of benign myelolipoma;
however, it does not completely exclude malignancy, par- ticularly when the proportion of fat is < 5%.
· Tumor size is not useful in discriminating between adrenal myelolipoma and adrenocortical carcinoma; however, rapidly growing lesions are more likely to be malignant.
· Ongoing vigilance is important when a clinical presenta- tion has suspicious features for malignancy, even if iso- lated radiology findings are seemingly reassuring.
Contributors
All authors made individual contributions to authorship. P.H .: was involved in the diagnosis and management of case 1. A.M .: retrospectively reviewed radiology images and iden- tified key images.
Funding
No public or commercial funding.
Disclosures
None declared.
Informed Consent for Publication
Case 1: Signed informed consent could not be obtained from the patient or a proxy but has been approved by the treating institution.
Case 2: Signed informed consent obtained directly from the patient’s relatives or guardians.
Data Availability Statement
Data sharing is not applicable to this article as no datasets were generated or analyzed during the current study.
References
1. Song JH, Chaudhry FS, Mayo-Smith WW. The incidental adrenal mass on CT: prevalence of adrenal disease in 1,049 consecutive ad- renal masses in patients with no known malignancy. Am J Roentgenol. 2008;190(5):1163-1168.
2. Schieda N, Davenport MS, Pedrosa I, et al. Renal and adrenal masses containing fat at MRI: proposed nomenclature by the Society of Abdominal Radiology Disease-Focused Panel on Renal Cell Carcinoma. J Magn Reson Imaging. 2019;49(4):917-926.
3. Mayo-Smith WW, Song JH, Boland GL, et al. Management of inci- dental adrenal masses: a white paper of the ACR incidental findings committee. J Am College Radiol. 2017;14(8):1038-1044.
4. Ranathunga DS, Cherpak LA, Schieda N, Flood TA, McInnes MDF. Macroscopic fat in adrenocortical carcinoma: a systematic review. Am J Roentgenol. 2020;214(2):390-394.
5. Montone KT, Rosen M, Siegelman ES, Fogt F, Livolsi VA. Adrenocortical neoplasms with myelolipomatous and lipomatous metaplasia: report of 3 cases. Endocr Pract. 2009;15(2):128-133.
6. Calissendorff J, Juhlin CC, Sundin A, Bancos I, Falhammar H. Adrenal myelolipomas. Lancet Diab Endocrinol. 2021;9(11): 767-775.
7. Sturgeon C, Shen WT, Clark OH, Duh Q, Kebebew E. Risk assess- ment in 457 adrenal cortical carcinomas: how much does tumour size predict the likelihood of malignancy? J Am College Surg. 2006;202(3):423-430.
8. Decmann A, Perge P, Toth M, Igaz P. Adrenal myelolipoma: a com- prehensive review. Endocrine. 2018;59(1):7-15.
9. Ng L, Libertino JM. Adrenocortical carcinoma: diagnosis, evalu- ation and treatment. J Urol. 2003;169(1):5-11.
10. Campbell MJ, Obasi M, Wu B, Corwin MT, Fananapazir G. The radiographically diagnosed adrenal myelolipoma: what do we real- ly know? Endocrine. 2017;58(2):289-294.