International Journal of Surgical Pathology http://ijs.sagepub.com/
Adrenocortical Oncocytic Tumors: Report of 10 Cases and Review of the Literature Michele Bisceglia, Ornella Ludovico, Antonio Di Mattia, David Ben-Dor, Judith Sandbank, Gianandrea Pasquinelli, Sean K. Lau and Lawrence M. Weiss INT J SURG PATHOL 2004 12: 231 DOI: 10.1177/106689690401200304
The online version of this article can be found at: http://ijs.sagepub.com/content/12/3/231
Published by: SSAGE http://www.sagepublications.com
Additional services and information for International Journal of Surgical Pathology can be found at:
Email Alerts: http://ijs.sagepub.com/cgi/alerts Subscriptions: http://ijs.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav
Permissions: http://www.sagepub.com/journalsPermissions.nav Citations: http://ijs.sagepub.com/content/12/3/231.refs.html
>> Version of Record - Jul 1, 2004 What is This?
Adrenocortical Oncocytic Tumors: Report of 10 Cases and Review of the Literature
Michele Bisceglia, MD,* Ornella Ludovico, MD,+ Antonio Di Mattia, MD,+ David Ben-Dor, MD,§ Judith Sandbank, MD, ** Gianandrea Pasquinelli, MD,*tt Sean K. Lau, MD,## and Lawrence M. Weiss, MD##
Ten additional adrenocortical oncocytic tumors are presented: 2 benign oncocytomas, 4 borderline oncocytomas of uncertain malignant potential, and 4 oncocytic carcino- mas. Histologically all tumors were entirely or predominantly composed of onco- cytes. Immunohistochemically all tumors were immunoreactive for mitochondrial antigen mES-13. Electron microscopy was performed in 8 cases and was confirma- tory of the oncocytic cell change. The morphologic parameters of the Weiss system, considered to be predictive of the biologic behavior of conventional (nononcocytic) adrenocortical tumors, are reviewed in the context of their possible application to the oncocytic tumor variant. Proposed major criteria (high mitotic rate, atypical mitoses, venous invasion) and minor criteria (large size and huge weight, necrosis, capsular in- vasion, sinusoidal invasion) in distinguishing malignant tumors are discussed, and de- finitional criteria (predominantly cells with eosinophilic and granular cytoplasm, high nuclear grade, diffuse architectural pattern) in common with all types of oncocytic tumors are outlined. The authors’ proposed working rules for diagnostic categoriza- tion of oncocytic adrenocortical tumors are defined, with the presence of 1 major cri- terion indicating malignancy, 1 to 4 minor criteria indicating uncertain malignant po- tential (borderline), and the absence of all major and minor criteria indicative of benignancy. Using these criteria, the diagnosis of malignancy was straightforward in 3 of the 4 cases designated as oncocytic carcinoma (presence of at least 2 major crite- ria and all the minor criteria), while in 1 case the original diagnosis of benign onco- cytoma was reversed to malignant following critical review of the original pathologic material after local tumor recurrence. Tumor recurrence occurred in 2 carcinomas at 8 and 20 months, respectively, and was followed in 1 case by the patient’s death. The third patient expired at 6 months from unrelated causes, and the fourth patient is free of disease at the relatively short follow-up interval of 6 months. Regarding the 4 pa- tients with borderline tumors, all are alive with no evidence of disease, with follow- up ranging from 10 to 61 months (mean 38.7 months). The 2 benign tumors have a follow-up of 25 and 30 months, respectively. Diagnostic difficulties are delineated and a complete review of the literature on this topic has also been performed. Int J Surg Pathol 12(3):231-243, 2004
Key words: oncocytoma, oncocytic carcinoma, adrenocortical tumor, endocrine tu- mor, electron microscopy, immunohistochemistry.
*Department of Pathology and +Operative Unit of Endocrinology, IRCCS “Casa Sollievo della Sofferenza” Hospital, San Giovanni Rotondo (FG), Italy; * Division of General Surgery, “Umberto I” Hospital, San Marco in Lamis (FG), Italy; §Department of Pathology, Barzilai Medical Center, Ben-Gurion University, Ashkelon, Israel; ** Department of Pathology, Asaf Harofe Hospital, Zerifin, Israel; ++Department of Experimental Pathology, Clinical Pathology Section, Policlinico “S. Orsola”, Bologna, Italy; and Department of Pathol- ogy, City of Hope National Medical Center, Duarte, California, USA.
Reprint requests: Michele Bisceglia, MD, Department of Pathology, Division of Anatomic Pathology, IRCCS “Casa Sollievo della Sof- ferenza” Hospital, I-71013 San Giovanni Rotondo (FG), Italy.
Oncocytic tumors are rare neoplasms, histologi- cally composed of epithelial cells with abundant aci- dophilic, granular cytoplasm that can be arranged in alveolar, tubular, trabecular, solid, and diffuse pat- terns. They are usually defined as tumors predomi- nantly or exclusively composed of oncocytes. Elec- tron microscopic studies of these tumors have shown that the cytoplasm of oncocytes is packed with mitochondria [1]. Oncocytic tumors have been described in many organs, such as the kidney, the salivary gland, and endocrine glands, as well as in a variety of other anatomic sites [1,2]. Although the first description of a case of an oncocytic tumor of the adrenal gland dates back 11/2 decades [3], the concept of a true oncocytic tumor in this site, irre- spective of its biologic behavior, has not yet been popularized. Herein we report on 10 such cases of oncocytic adrenocortical tumors, all of which were studied immunohistochemically, with 8 also inves- tigated by electron microscopy. Emphasis is laid on morphologic predictors of biologic behavior and di- agnostic difficulties.
Materials and Methods
Ten cases of adrenocortical oncocytic tumors are presented. Five cases (cases 1, 3, 4, 9, and 10) were retrieved from the files of the Anatomic Pathology Division of IRCCS “Casa Sollievo della Sofferenza” Hospital, San Giovanni Rotondo (FG), Italy; 1 case (case 2) was first seen in the Division of Anatomic Pathology of the Barzilai Medical Center, Ben-Gu- rion University, Ashkelon, Israel, and then at the Division of Anatomic Pathology of the Asaf Harofe Hospital, Zerifin, Israel; and 4 cases (cases 5, 6, 7, and 8) were from the consultation files of 1 of the authors (LMW) at the Department of Pathology, City of Hope National Medical Center, Duarte, Cali- fornia, USA.
Light Microscopy
Tissue samples from all cases were embedded in paraffin blocks, and sections measuring 4 um were cut and stained with hematoxylin and eosin, Mas- son’s trichrome, and Gomori’s silver stain for retic- ulin. In order to assess tumor behavior and clinical prognosis, each case was morphologically evaluated according to the 9 histologic parameters of the Weiss system used for discriminating benign from malignant conventional adrenocortical tumors [4].
Immunohistochemistry
For the purpose of this study all cases were eval- uated immunohistochemically by using the avidin-
biotin-peroxidase complex technique, after em- ploying a preliminary heating-based antigen re- trieval step (either microwave or pressure cooking treatments).
The immunohistochemical panel included anti- bodies to wide spectrum (w.s.) keratin (monoclonal, clone AE1-AE3, prediluted, NeoMarkers, Fremont, CA, USA); keratin peptide 8 for cases 1, 2, 3, 4, 9, and 10 (monoclonal, clone Cam5.2, prediluted, Becton Dickinson, San Jose, CA, USA) or keratin peptides 8/18 for cases 5, 6, 7, and 8 (monoclonal, Zym5.2, UCD/PR-10.11, diluted 1:100, Zymed Lab- oratories, Inc., South San Francisco, CA, USA); ep- ithelial membrane antigen (EMA) (monoclonal, clone E29, prediluted, Dako); vimentin (mono- clonal, clone V9, dilution 1:10, Dako); S-100 pro- tein (polyclonal, prediluted, Dako); neuron-specific enolase (NSE) (polyclonal, diluted 1:100, Dako); chromogranin-A (monoclonal, clone PHE5, predi- luted, Enzo Diagnostics, Farmingdale, NY, USA); synaptophysin (monoclonal, clone SY38, diluted 1:10, Dako); HMB-45 (monoclonal, clone HMB-45, prediluted, Dako); Melan-A (monoclonal, clone A103, prediluted, Biogenex, San Ramon, CA, USA); inhibin alpha (monoclonal, clone R1, diluted 1:30, Serotec, Kidlington/Oxford, United Kingdom); CD10 (monoclonal, clone 56C6, diluted 1:80, Novo- castra Laboratories, Newcastle upon Tyne, UK); and mitochondrial antigen mES-13 (monoclonal, clone A-113-1, diluted 1:75, Biogenex [cases 1, 2, 3, 4, 9, and 10] as well as monoclonal, diluted 1:1000, courtesy of E. Santos, National Institutes of Health, Bethesda, Maryland, USA [cases 5, 6, 7, and 8]). The mitochondrial antigen mES-13 was specifically detected even following employment of a biotin- blocking procedure (biotin-blocking system, Dako).
The tumor-proliferating fraction (TPF) was evalu- ated with MIB-1 MoAb (dilution 1:50, clone MIB-1, Dako) against Ki-67 human antigen and deter- mined by counting the positive tumor nuclei in the areas of highest labeling density of each individual tumor. About 1,000 tumor cells were counted over a minimum of 10 high power fields (HPF).
Electron Microscopy
Fragments of paraffin-embedded tissue were sub- mitted for ultrastructural study in order to confirm the oncocytic nature of the tumors. The electron microscopical analysis, which was confirmatory, was performed either at the Electron Microscopy Unit of the Department of Clinical Pathology, Poli- clinico “S. Orsola”, Bologna, Italy (cases 1, 2, 3, and 4) and at the Department of Pathology, IRCCS “Casa Sollievo della Sofferenza” Hospital, San Giovanni Rotondo, Italy (cases 9 and 10), or at the referring
| Clinical Data | Case 1a | Case 1bt | Case 2a | Case 2bt | Case 3 | Case 4 |
|---|---|---|---|---|---|---|
| Sex | M | M | F | F | F | F |
| Age, years | 46 | 46 plus 8 months | 32 | 32 plus 20 months | 22 | 62 |
| Side | R | R | R | R | L | R |
| Clinical syndrome | Gynecomastia | — | None | — | None | None |
| Hormone serum & urinary assay | Normal | — | Serum 17-OH cortisolÎ; urinary free cortisolt | — | serum ACTH4; urinary free cortisolÎ | Normal |
| Size, cm and weight, g | 17; 1,900 | — | 11; 252 | — | 5; 40 | 8; 260 |
| WSS Parameters Clear cell composition (≤25%) | 0% | 20% | 20% | 0% | <10% | 0% |
| Nuclear grade | IV | III | IV | III | IV | III |
| Diffuse architecture | Present | Present | Present | Absent | Present | (mainly) Absent |
| Mitotic rate (per 50 HPF) | 16 | 10 | 5 (focally) | 10 | 0 | 20 |
| Atypical mitoses | Present | Present | Absent | Absent | Absent | Present |
| Necrosis | Present | Present | Absent | Present | Absent | Present |
| Venous invasion | Present | Present | Absent | Present | Absent | Present |
| Capsular invasion | Present | — | Absent | — | Present | Present |
| Sinusoidal invasion | Absent | Absent | Absent | Absent | Absent | Present |
| Final Diagnosis | Malignant* | — | Malignant ** | — | Borderline | Malignant |
| Follow-up | DOD at 58 months | Recurrence of case 1 at 8 months | AED at 35 months | Recurrence of case 2 at 20 months | ANED at 26 months | DNOD at 6 months |
| Clinical Data | Case 5 | Case 6 | Case 7 | Case 8 | Case 9 | Case10 |
|---|---|---|---|---|---|---|
| Sex | M | F | F | F | F | F |
| Age, years | 50 | 26 | 46 | 24 | 25 | 77 |
| Side | R | L | L | R | L | L |
| Clinical syndrome | None | None | None | Virilization | None | None |
| Hormone serum | Normal | Normal | Normal | Normal | Normal | |
| & urinary assay | Normal | |||||
| Size, cm and weight, g | 12.5; 430 | 8; 186 | 6; 85 | 20; 1400 | 11; 165 | 10; 120 |
| WSS Parameters | ||||||
| Clear cell composition (≤25%) | 0% | 0% | 0% | 0% | 0% | 15% |
| Nuclear grade | IV | IV | IV | IV | III | III |
| Diffuse architecture | Present | Present | Present | Present | Present | Present |
| Mitotic rate (per 50 HPF) | 0 | <1 | 0 | 0 | 0 | 20 |
| Atypical mitoses | Absent | Absent | Absent | Absent | Absent | Present |
| Necrosis | Absent | Absent | Absent | Absent | Absent | Present |
| Venous invasion | Absent | Absent | Absent | Absent | Absent | Absent |
| Capsular invasion | Absent | Absent | Absent | Absent | Present | Present |
| Sinusoidal invasion | Absent | Absent | Absent | Absent | Present | Present |
| Final Diagnosis | Borderline | Benign | Benign | Borderline | Borderline | Malignant |
| Follow-up | ANED at 61 | ANED at | ANED at | ANED at | ANED at | ANED at |
| months | 30 months | 25 months | 58 months | 10 months | 6 months |
*Initially diagnosed as (nononcocytic) adrenocortical carcinoma of compact cell type; t(1b and 2b) = recurrences of case 1 and case 2 (primaries are la and 2a, respectively); M = Male; F = Female; R = Right; L = Left; DOD = Dead of disease; Initially diagnosed as benign oncocytoma; DNOD = Dead not of disease. AED = Alive with evidence of disease. ANED = Alive with no evidence of disease.
institutions (cases 5 and 8). The small pieces of tis- sue were dewaxed, osmicated, dehydrated, and em- bedded in araldite. Thin sections were stained with uranyl acetate and lead citrate and observed in a transmission electron microscope.
Results
Clinical Data and Pathological Features
The clinical data, size and weight of the tumors, and the histologic features of the 10 cases are sum- marized in Table 1. Of the 10 patients, 2 were men and 8 were women. The age range was 22 to 77 years (average: 41 years). The tumors were equally distributed per each side (right/left = 5:5). Most tu- mors (8 of 10) were hormonally nonfunctioning. Size and weight ranges were 5 to 20 cm and 40 to 1,900 g, respectively. Of the 9 histologic parameters of the Weiss system, 2 features were observed in all tumors (clear cell composition ≤ 25% and high nu- clear grade) and 1 feature (diffuse architecture) was present in all but 1 case (Figs. 1, 2). Trichrome stain and Gomori’s stain were helpful in detecting focal trabecular and/or nesting patterns in 2 cases, which otherwise would have been overlooked, and served to confirm the mainly trabecular and nesting pat- tern observed on H&E in an additional case (case 4). The remaining 6 parameters were variably present, with necrosis observed in 4 cases, high mitotic rate in 4 (Fig. 3), atypical mitoses in 3 (Fig. 3), capsular invasion in 5 (Fig. 4), venous invasion in 3 (Fig. 5), and sinusoidal invasion in 3 (Fig. 6). Of the 10 cases, 2 were diagnosed as benign, 4 as borderline, and 4 as malignant based on criteria discussed below.
Immunohistochemical Findings and Tumor-Proliferating Fraction
The immunohistochemical findings and the TPF values are detailed in Table 2. All tumors were strongly immunoreactive for mitochondrial antigen mES-13 (Fig. 7). Seven tumors were immunoreac- tive for Melan-A, 7 for inhibin alpha, 5 for vi- mentin, 5 of 8 for NSE, 4 for synaptophysin, and 1 of 9 for CD10 in only scattered cells. Six of the neo- plasms were positive for pankeratin, while 1 of 6 cases studied were positive for keratin 8, and none of 4 for keratin 18. EMA, chromogranin-A, HMB- 45 (9 of 10), and S-100 protein were consistently negative. The TPF with MIB-1/Ki67 was in a value range of 0-4% for tumors classified as benign and borderline, while it was ≥5% in 3 of 4 and < 2% in 1 (case 2a) of the malignant cases.
A
B
Ultrastructural Findings
The ultrastructural analysis was confirmatory of the oncocytic change of tumor cells in all 8 cases studied. The tumor cells exhibited innumerable mi- tochondria as well as smooth endoplasmic reticulum and few lipid droplets, which were more represented in those cases with a clear cell component (Fig. 8). Mitochondria exhibited either typical lamellar cristae as seen in zona glomerulosa cells of normal adrenal cortex (cases 1, 2, 5, and 8) or tubulovesicu- lar cristae as seen in zona fasciculata cells (cases 3, 4, 9, and 10) (Figs. 9, 10). Further, some amorphous intramitochondrial inclusions were observed in 2 cases (cases 2 and 9) (Fig. 11). Tumor cells exhibited focal rudimentary cell junctions and attenuated desmosomes (cases 3 and 9) (Fig. 12).
Discussion
Oncocytic tumors originating from the adrenal gland cortex are exceedingly rare, with 35 total cases so far reported [3,5-7], some appearing only in ab- stract form [10,17,20] (Table 3). Two cases of onco- cytic carcinoma were subtracted, as both had previ- ously appeared in the literature more than once in abstract form and as full reports [20,21], with 1 of the 2 also already reported in a cytopathology jour- nal [22]. The two cases of adrenal cortical adenomas with oncocytic features partly illustrated by Erland- son and Reuter [5], which were also included in the series of cases reported by Lin et al [13], were not added up as well. Another fully reported case by Segal et al [23] has not been included in this tabula- tion as it refers to case 2 of the present series.
| Immunomarker | Case 1a | Case 1bt | Case 2a | Case 2bt | Case 3 | Case 4 |
|---|---|---|---|---|---|---|
| Keratin w.s | +(dot-like) | - | - | - | +(focal) | - |
| Keratin 8 or 8/18* | - | - | - | - | - | - |
| EMA | - | - | - | - | - | - |
| Vimentin | +(focal) | - | + | + | - | - |
| S-100 protein | - | - | - | - | - | - |
| NSE | - | - | +(focal) | - | - | +(focal) |
| Chromogranin-A | - | - | - | - | - | - |
| Synaptophysin | - | - | +(focal) | +(focal) | - | +(focal) |
| HMB-45 | - | - | - | - | - | - |
| Melan-A | - | - | + | + | + | - |
| Inhibin alpha | + | - | + | +(focal) | + | +(weak) |
| CD10 | - | - | - | - | - | - |
| mES-13 | + | + | + | + | + | + |
| MIB-1 | 5-20% | 5-20% | <2% | 15% | <2% | 5% |
| Final diagnosis | Malignant | Malignant | Borderline | Malignant | ||
| Immunomarker | Case 5 | Case 6 | Case7 | Case 8 | Case 9 | Case 10 |
| Keratin w.s | - | + | + | Rare cells + | + | - |
| Keratin 8 or 8/18* | - | - | - | Not done | + | - |
| EMA | - | - | - | - | - | - |
| Vimentin | - | - | - | + | + (focal) | + (focal) |
| S-100 protein | - | - | - | - | - | - |
| NSE | Not done | + | + | Not done | + (focal) | - |
| Chromogranin-A | - | - | - | - | - | - |
| Synaptophysin | - | - | Rare cells + | - | + | - |
| HMB-45 | - | - | - | Not done | - | - |
| Melan-A | + | + | + | + | + | - |
| Inhibin alpha | - | + | + | - | - | + |
| CD10 | - | - | Rare cells + | Not done | - | - |
| mES-13 | + | + | + | + | + | + |
| MIB-1 | 0% | 4% | 1% | 0% | 1-2% | 5-20% |
| Final diagnosis | Borderline | Benign | Benign | Borderline | Borderline | Malignant |
*Cases 1, 2, 3, 4, 9, and 10 were tested for keratin 8, while cases 2, 5, 6, and 7 were tested for keratins 8 and 18. +(1b and 2b) = recurrences of case 1 and case 2 (primaries are la and 2a, respectively); w.s. = wide spectrum.
A
B
·
&
8
The oncocytic tumors thus far reported have been of variable size (3-20 cm) and weight (12-2,415 g), with significant female-to-male predominance (2.5:1), wide age range (27-74 years), and a strong left-to-right side prevalence (3.5:1). Of a total of 35 tumors, 20 have been designated as benign [3,5,7-13,15,17,19], 8 borderline [13,14,20], and 7 malignant [6,16,18,20,21] (Table 3). Tumors were for the most part nonfunctioning, with only 3 being
hormonally active (1 case with virilization [7], 1 case with feminization [19], 1 case with Cushing’s syndrome [16]), thus probably explaining the large size commonly attained by these neoplasms (in 20 cases: median size 8 cm; median weight 200 g). With immunohistochemistry [5,13,17,18,20] the oncocytic tumors were consistently positive for mi- tochondrial antigen mES-13, and often exhibited positivity for keratin 18 and vimentin as well as for
B
A
Melan-A, synaptophysin, NSE, inhibin alpha, and CD10. The tumors were typically negative for ker- atin 20, S-100 protein, HMB-45, EMA, p53, and steroidogenic enzymes. Sixteen of a total of 29 fully reported cases were also studied by electron mi- croscopy [5-8,13-15,17,19,21], all of which demon- strated ultrastructural cell findings of oncocytes.
We have collected and described an additional 10 cases of adrenocortical oncocytic neoplasms in or- der to further characterize the light microscopic, im- munophenotypic, and ultrastructural features of this class of tumors and to correlate morphologic parameters with their biological behavior.
Immunohistochemically (Table 2), all tumors in the present study were diffusely immunoreactive for mitochondrial antigen mES-13 in a cytoplasmic distribution. Similar to conventional nononcocytic adrenocortical tumors, Melan-A and inhibin alpha positivity were also observed in the majority of the oncocytic tumors in the present series. CD10 was negative in 8 of 9 tested cases. HMB-45 and S-100 protein were likewise negative in all cases studied. Pankeratin (w.s.) was positive in 6 cases with 2 cases exhibiting only focal or rare immunoreactivity, and 1 case displaying cytoplasmic dot-like staining often seen in some neuroendocrine tumors. Only 1 case of 6 tested for keratin 8 was positive, while none of 4 cases examined was positive for keratin 18. Vi- mentin staining was observed in 5 tumors (focally in 3). In regard to neuroendocrine-associated mark- ers, synaptophysin positivity was observed in 5
cases and largely focal NSE immunoreactivity was noted in 5 of 8 cases investigated. However, all tu- mors were negative for chromogranin A.
Ultrastructurally, all tumors, as expected, were composed of cells exhibiting numerous mitochon- dria as well as smooth endoplasmic reticulum and few lipid droplets, which were more represented in those cases with a clear cell component (cases 1b, 2a, 3, and 10). Mitochondria exhibited either typi- cal lamellar cristae as seen in zona glomerulosa cells of the normal adrenal cortex or tubulovesicular cristae as seen in zona fasciculata cells. In 2 cases mitochondrial inclusions of the amorphous elec- tron-dense type were also observed, similar to those usually found in adenomatous hyperplasia and in conventional adrenocortical neoplasms [24] and previously documented in published cases of adrenal oncocytic tumors [5,6].
Obviously, the most important practical problem regarding adrenocortical oncocytic neoplasms con- cerns their biologic behavior. Aside from those ob- viously malignant cases in which the diagnosis is based on common clinicopathologic features uni- versally valid for tumors of any site, such as distant metastases, surgical unresectability, and/or invasion of adjacent organs, the biologic behavior and clini- cal outcome for adrenocortical tumors are other- wise difficult to predict.
Tumor size and weight have long been empha- sized as predictors of biological behavior, with large tumors (>50 g and >5 cm) likely to have a lethal
| Reference Number | Authors | Publication Year | Number of Cases | Diagnosis |
|---|---|---|---|---|
| 3 | Kakimoto S et al | 1986 | 1 | Benign oncocytoma |
| 5 | Erlandson RA and Reuter VE | 1991 | 1+ | Benign oncocytoma |
| 6 | El-Naggar AK et al | 1991 | 1 | Malignant oncocytoma |
| 7 | Sasano H et al | 1991 | 3 | Benign oncocytoma |
| 8 | Begin LR | 1992 | 1 | Benign oncocytoma+ |
| 9 | Nguyen G-K et al | 1992 | 1 | Benign oncocytoma |
| 10 | Muir TE et al (abstract) | 1996 | 1 | Benign oncocytoma |
| 11 | Gandras EJ et al | 1996 | 1 | Benign oncocytoma+ |
| 12 | Waters PR et al | 1997 | 2 | Benign oncocytoma |
| 13 | Lin BT-Y et al | 1998 | 7 | 5 Benign oncocytomas; 2 Borderline oncocytomas |
| 14 | Macchi C et al | 1998 | 1 | Borderline oncocytoma |
| 15 | Kitching PA et al | 1999 | 1 | Benign oncocytoma+ |
| 16 | Alexander A et al | 1999 | 1 | Malignant oncocytoma |
| 17 | Scheibel M et al (abstract) | 2001 | 1 | Benign oncocytoma |
| 18 | Kurek R et al | 2001 | 1 | Malignant oncocytoma |
| 19 | Icard P et al | 2001 | 2 | Benign oncocytoma |
| 20 | Wu SL et al (abstract) | 2002 | 7 | 5 Borderline oncocytoma; 2 Malignant oncocytomas |
| 21 | Hoang MP et al | 2002 | 2* | 2 Malignant oncocytomas |
tIn the same article the authors had also illustrated 2 other cases acknowledging their respective original sources. These other 2 cases subse- quently made part of the series of Lin et al (ref. 13).
*After subtraction of 2 other cases, which we believe are included in ref. 20.
From the description provided in the original report, the tumor appears morphologically benign.
outcome and, conversely, small tumors (<50 g and <5 cm) expected to behave favorably. However, since this is not always the rule [4,25-29], a num- ber of systems have been devised to distinguish be- nign from malignant conventional (non-oncocytic) adrenocortical tumors [4,26,30-32], among which the most widely used and reliably accepted is the Weiss system [4,33]. In this particular system, a combination of 9 histologic features are evaluated (Fürhman nuclear grade III-IV, mitotic rate >5 per 50 HPF, atypical mitoses, clear cell composition ≤25% of the entire tumor mass, diffuse architec- ture, necrosis, venous invasion, sinusoidal invasion, capsular invasion), with each parameter scored 0 when absent and 1 when present (total Weiss score in order to grade a tumor ranging from 0 to 9). The presence of 2 of these criteria is considered allow- able in benign tumors while the presence of ≥3 cri- teria is suggestive of malignancy [32,34]. There has also been suggestion to give precedence to some of the above-cited histologic features over others [34]. In particular, a diagnosis of malignancy would prob- ably be warranted by the presence of high mitotic activity, atypical mitoses, and venous invasion even in isolation. These latter criteria would merit desig- nation as major criteria, although those authors [34] suggest looking carefully for additional histo- logic features of malignancy when only a single ma- jor criterion is evident. Further, there are some other authors [33] who recommend a modified Weiss scoring system by retaining only the most re-
liable criteria in their experience and assigning a value of 2 to some features (mitotic rate, eosinophilic cytoplasm) and a value of 1 to others (abnormal mitoses, necrosis, capsular invasion) with a total score ranging from 0 to 7 and a thresh- old of malignancy of ≥3, after excluding considera- tion of nuclear grade, diffuse architecture, venous invasion, and sinusoidal invasion owing to their lack of reliability.
The use of MIB-1/Ki-67 has also proven helpful in the assessment of benign versus borderline and malignant cases of adrenocortical tumors [33,35,36]. In 1 referenced source [36] adenomas showed a mean TPF of around 1.5% (with no case >8%) while carcinomas showed a TPF of around 20% (no case <8%). In another report [33] a TPF of ≥4% was significant for malignancy with a mean of 2.4 ± 1.3% in benign tumors. The TPF determined with MIB-1 significantly correlated with mitotic rate [33].
Around a decade ago Erlandson and Reuter [5] and Sasano et al [7] described a small series of adrenocortical oncocytic tumors and, despite the short available follow-up, first raised the question regarding the biological behavior of this subset of adrenocortical neoplasms, the latter authors specif- ically questioning whether the Weiss system was applicable to this particular subset of adrenocortical tumors. Subsequently, Lin et al [13], including 1 of the present authors (LMW), suggested that the Weiss system be modified for the assessment of on-
cocytic tumors of the adrenal cortex. Since most on- cocytic tumors already possess 3 positive criteria (composition of eosinophilic cells [i.e., less than 25% clear cells out of the entire tumor mass], high nuclear grade, diffuse architecture), they would all be automatically designated as malignant, which is not supported by the available follow-up of some of the previously reported cases (up to 5-8 years). As such, for adrenocortical oncocytic neoplasms, the morphologic (gross and microscopic) criteria of ma- lignancy employed for conventional cortical neo- plasms of the adrenal do not apply. In the paper by Lin et al [13], which reported on 7 cases of onco- cytic tumors (5 benign oncocytomas and 2 of un- certain malignant potential), the suggestion is given that benign oncocytomas be distinguished from possible malignant oncocytomas on the basis of mi- totic activity, necrosis, and vascular and capsular in- vasion. However, it was not stated how many of these criteria need to be present to warrant a diag- nosis of malignancy. The 2 tumors of uncertain ma- lignant potential were so designated owing to the presence of necrosis and/or some mitotic activity.
Recently, Wu et al [20], presented another series of adrenocortical oncocytic tumors (5 oncocytic tu- mors with uncertain malignant potential and 2 on- cocytic carcinomas) in abstract form, but similarly did not explain which combination of criteria were used to classify their tumors as oncocytomas, onco- cytic tumors with uncertain malignant potential, and oncocytic carcinomas. One of us (MB) person- ally contacted Dr. J. Y. Ro, the lead author of the Wu group [20] concerning their guidelines for diagnos- ing oncocytic tumors. The classification rules of adrenocortical oncocytic tumors according to Ro and Wu (MD Anderson Cancer Center guidelines) are as follows: oncocytoma-absence of mitotic activity (<5/50 HPF), necrosis, capsular invasion, extratu- moral venous invasion, metastasis; oncocytic adrenocortical neoplasm of uncertain biologic be- havior-increased mitotic activity (>5/50 HPF) or microscopic foci of necrosis; oncocytic carcinoma- nuclear pleomorphism along with any of the fol- lowing: atypical mitoses, capsular invasion, extratu- moral venous invasion, or metastasis.
The authors of the latest article on this particular subject [21], including a representative from the MD Anderson Cancer Center, in dealing with 4 adrenocortical oncocytic carcinomas, obviously stated that “large size, vascular invasion, capsular extension and surgical unresectability are features of malignancy” but did not go into analysis of how many (i.e., 1 or more than 1) of those criteria are needed to diagnose malignancy, although each of their cases appeared to have at least 2 of them.
Moreover, this report did not address the issue of uncertain malignant potential, a point that should merit clarification since large size or weight alone seems not to be sufficient to identify clear-cut ma- lignancy in view of the fact that the size range of (benign) oncocytomas overlaps with that of onco- cytic carcinomas [13].
For diagnostic categorization of oncocytic adrenocortical tumors, we propose the following criteria, modified from the Weiss system for nononcocytic adrenocortical tumors: if an oncocytic tumor exhibits a mitotic rate of more than 5 mitoses per 50 high power fields, any atypi- cal mitoses, or venous invasion (defined as major cri- teria), it is considered malignant; if the tumor ex- hibits 1 of several other worrisome features (large size [>10 cm and/or >200 g], necrosis, capsular in- vasion, or sinusoidal invasion) (defined as minor cri- teria), the tumor is considered of uncertain malig- nant potential (borderline); and conversely, if none of the above-listed features are present it is reason- able to call the tumor benign. In summary, for on- cocytic adrenocortical neoplasms, the presence of 1 major criterion indicates malignancy (oncocytic adrenocortical carcinoma), 1 to 4 minor criteria pre- sent indicates uncertain potential (borderline), and the absence of all major and minor criteria is indica- tive of benignancy (adrenocortical oncocytoma). Three of the original histopathologic parameters of the Weiss system, which can be considered accept- able for an oncocytic tumor (predominantly cells with eosinophilic [and granular] cytoplasm, high nuclear grade, and diffuse architectural pattern), are not taken into account for purposes of classification and determination of biologic behavior and there- fore are considered as definitional criteria of this pecu- liar class of adrenocortical tumors.
The series herein presented (Table 1) includes 10 additional cases of oncocytic adrenocortical tumors. Applying the criteria described above, 2 were diag- nosed as benign oncocytomas, 4 as borderline onco- cytic tumors of uncertain malignant potential, and 4 as malignant oncocytic tumors (oncocytic carcino- mas). The diagnosis of malignancy was relatively straightforward (presence of at least 2 major criteria and all the minor criteria) in 3 of the 4 total cases of oncocytic carcinoma, while in 1 case (case 2) the original rendered diagnosis of benign oncocytoma was reversed and changed to malignant following critical review and reevaluation of additional tumor slides prompted by subsequent local recurrence of the tumor. Tumor recurrence occurred in 2 of the oncocytic carcinomas and was followed in 1 by the patient’s death. Death supervened from unrelated causes in another case, and the follow-up is still too short in the fourth patient for meaningful comment
(case 10). With regard to the 4 patients with bor- derline oncocytic tumors of uncertain malignant potential, all are free of disease with follow-up rang- ing from 10 to 61 months (mean 38.7 months). The 2 oncocytomas have behaved in a benign fashion, with a follow-up of 25 and 30 months, respectively.
Of obvious critical importance in the successful application of this proposed system are extensive tissue sampling, meticulous mitotic rate evaluation (since mitotic activity might also be only focal or dif- ficult to distinguish from nuclear pyknotic changes), and a careful search for tumor cell necro- sis which can also be difficult to discriminate from ischemic necrosis secondary to surgical arterial liga- ture procedures.
These points are borne out by the rather convo- luted trajectory taken by case 2 in the present series. Two to 3 slides of the original tumor in this patient were seen in consultation individually by several se- nior pathologists, including some known to be ex- perts in endocrine pathology, who were unanimous in the diagnosis of oncocytoma. Following appear- ance of the recurrence, slides taken from different blocks of the original tumor (other than those ini- tially reviewed) were submitted to 1 of the original consultants, who identified sufficient mitotic activ- ity in the additional material to warrant the diagno- sis of malignancy, which in this sense could have been made at the beginning if this material had been made available initially.
In the present study, 2 cases (5 and 8) were con- sidered to be of uncertain malignant potential only because of size considerations (12.5 cm/430 g and 20 cm/1,400 g, respectively) in the seeming absence of other histologic features indicative of aggressive- ness. Tumor size and weight were included among our minor criteria for classification, as 3 of 4 cases categorized as malignant were observed to be greater than 200 g, and may thus be a feature asso- ciated with aggressive behavior. For example, we were wrong for initially classifying as benign the above-mentioned case 2, which had a positive weight parameter (over 200 g), and which, in retro- spect, also had an accompanying focal high mitotic rate, the latter overlooked at first examination. A large oncocytic tumor of the adrenal should there- fore be approached cautiously and prompt a careful search for other histologic features of malignancy.
The presence of clear cells in an otherwise onco- cytic tumor may have an important impact on clin- ical prognosis. Whether a tumor should be entirely or predominantly oncocytic to be designated as an oncocytic neoplasm remains unclear; however, most previous series have required that such tumors be composed exclusively of oncocytes. Four cases of
adrenocortical oncocytic tumors included in the pre- sent series contained a component of clear cells (Table 1). Case 1 presented initially as an entirely on- cocytic neoplasm. The patient subsequently devel- oped tumor recurrence, which histologically con- tained 20% clear cells, and eventually died of disease at 58 months. Case 2, classified as malignant based on the presence of high mitotic rate, was composed of 20% clear cells; however, the recurrent tumor was interestingly composed only of oncocytes.
Case 3, diagnosed as borderline, was mostly com- posed of highly pleomorphic oncocytic cells but contained less than 10% clear cells. The patient is still alive and without evidence of disease at 26 months follow-up. Case 10, categorized as malig- nant based on increased mitotic activity and the presence of atypical mitoses, contained 15% clear cells, and has no evidence of disease after a short follow-up period of 6 months. Based on the clinical behavior of this rather small number of cases, it seems that the presence of clear cells in a predomi- nantly oncocytic adrenocortical tumor could well be associated with an adverse clinical outcome. Since information concerning predominantly oncocytic adrenocortical neoplasms with a clear cell compo- nent is limited, it would be best to regard and iden- tify such tumors as an entity separate from purely oncocytic tumors of the adrenal cortex until the prognostic significance of the presence of clear cells can be determined.
In regard to the use of calculated TPF as a means of distinguishing benign and malignant oncocytic adrenocortical tumors, TPF as measured by staining with MIB-1 was in accordance with the previous experience of other authors with conventional adrenocortical neoplasms [33,35,36]: a value of ≥5% and a good correlation with mitotic activity was found in all malignant cases (including recur- rences), except for the TPF <2% in the primary of case 2, which also showed only focal increased mi- totic activity, while a TPF value range of 0-4% was found in borderline and benign tumors.
The differential diagnosis of adrenocortical onco- cytic tumors includes the following entities: onco- cytic pheochromocytoma [37], conventional adrenocortical carcinoma exclusively composed of compact cells (case 1 of our series was so initially di- agnosed), conventional adrenocortical carcinoma with focal oncocytic changes [39a], conventional renal carcinoma with oncocytic features or eosinophilic variant of chromophobe renal carci- noma [40] involving the adrenal, and hepatocellu- lar carcinoma [21]. The use of immunohistochem- istry [13,38] and electron microscopy [5-7,13,17,21,41] can provide support for the diag-
nosis of an oncocytic neoplasm, especially in dubi- ous cases.
In summary, from the analysis and the interpre- tation of the available published data, in addition to the present cases, we propose that if an adrenocorti- cal oncocytic tumor exhibits 1 major criterion (high rate of mitotic activity, atypical mitoses, or venous in- vasion), the tumor should be regarded as “malig- nant”; if it has 1 minor criterion (large size, necrosis, or capsular invasion), then it should be considered “bor- derline”; and if it possesses none of the major or mi- nor criteria, then it should be called “benign.” High nuclear grade, predominantly nonclear cell composi- tion, and diffuse architecture are histologic features (definitional criteria) that can be observed in both be- nign and malignant adrenocortical oncocytic neo- plasms, and thus likely have no influence on clinical outcome in this context. The reliability of these crite- ria is dependent on the pathologists’ making the cor- rect assessment of an oncocytic neoplasm at the on- set. Disregarding these features in a nononcocytic adrenocortical tumor might result in misdiagnosis.
Finally, although it seems apparent that there cer- tainly are many examples of benign oncocytic neo- plasms of the adrenal (refs. 13 and 19, and cases 6 and 7 of this series), caution is still suggested in clas- sifying a particular tumor as such, since the follow- up of these reported cases is generally short and the overall number of reported cases is relatively small. Further and larger studies with extended clinical follow-up are needed before definitive conclusions can be drawn regarding the clinical behaviour of these particular tumors [20,39b].
Additional Case
After the submission of this paper but before its publication, one of us (MB) had the opportunity to see an additional case of pure oncocytic adrenocor- tical tumor, mainly epithelioid with some micro- scopic foci of spindling, which, based on the above histological predictive parameters, met the criteria for carcinoma by virtue of the presence of 2 major criteria (high rate of mitotic activity of 8M:50 HPF and atypical mitoses) and 3 minor criteria (large size of 12 cm & 400 g of weight; necrosis; capsular inva- sion). Obviously the definitional criteria were all pre- sent (clear cell composition of 0%; nuclear grade IV; diffuse architecture). This tumor also showed frank invasion into the periadrenal fatty tissue.
Follow-up
In case 10 an abdominal recurrence occurred 10 months after the initial diagnosis of adrenocortical oncocytic carcinoma was made. The recurrence has
been histologically proven, showing exclusive on- cocytic features.
Acknowledgments
We gratefully acknowledge the helpful corre- spondence and the critical discussion with Drs. J. Y. Ro and S. L. Wu (Department of Pathology, MD An- derson Cancer Center, Houston, Texas, USA) con- cerning their own practical usage of prognostic morphologic parameters for adrenocortical onco- cytic tumors. The authors also wish to thank the fol- lowing pathologists who kindly provided case mate- rial for this study: Dr. J. C. Watts, William Beaumont Hospital, Royal Oak, Michigan, USA; Dr. J. M. Odell, Black Hills Pathology, Olympia, Washington, USA; Dr. J. Z. Sickel, El Camino Hospital, Mountain View, California, USA; and Dr. H. Safaii, New Eng- land Medical Center, Boston, Massachusetts, USA.
References
1. Chang A, Harawi SJ. Oncocytes, oncocytosis, and on- cocytic tumors. Pathol Annu 27:263-304, 1992
2. Baloch Z, LiVolsi VA. Oncocytic lesions of the neu- roendocrine system. Semin Diagn Pathol 16:190-199, 1999
3. Kakimoto S, Yushita Y, Sanefuji T, Kondo A, Fujishima N, Kishikawa M, Matsumoto K. Nonhormonal adrenocortical adenoma with oncocytoma-like ap- pearances. Hinyokika Kiyo 32:757-763, 1986
4. Weiss LM. Comparative histologic study of 43 metas- tasizing and nonmetastasizing adrenocortical tu- mors. Am J Surg Pathol 8:163-169, 1984
5. Erlandson RA, Reuter VE. Oncocytic adrenal cortical adenoma. Ultrastruct Pathol 15:539-547, 1991
6. El-Naggar AK, Evans DB, Mackay B. Oncocytic adrenal cortical carcinoma. Ultrastruct Pathol 15:549-556, 1991
7. Sasano H, Suzuki T, Sano T, Kameya T, Sasano N, Nagura H. Adrenocortical oncocytoma. A true non- functioning adrenocortical tumor. Am J Surg Pathol 15:949-956, 1991
8. Begin LR. Adrenocortical oncocytoma: Case report with immunocytochemical and ultrastructural study. Virchows Arch A Pathol Anat 421:533-537, 1992
9. Nguyen G-K, Vriend R, Ronaghan D, Lakey WH. Het- erotopic adrenocortical oncocytoma. A case report with light and electron microscopic studies. Cancer 70:2681-2684, 1992
10. Muir TE, Ferreiro JA, Carney JA. Oncocytoma of the adrenal gland. Mod Pathol 9:50A (Abstract), 1996
11. Gandras EJ, Schwartz LH, Panicek DM, Levi G. Adrenocortical oncocytoma: CT and MRI findings. J Comput Assist Tomogr 20:407-409, 1996
12. Waters PR, Haselhuhn GD, Gunig WT, Philips ER, Sel- man SH. Adrenocortical oncocytoma: Two cases re- ports and review of literature. Urology 49:624-628, 1997
13. Lin BT-Y, Bonsib SM, Mierau GW, Weiss LM, Medeiros LJ. Oncocytic adrenocortical neoplasms. A report of seven cases and review of the literature. Am J Surg Pathol 22:603-614, 1998
14. Macchi C, Rebuffat P, Blandamura S, Piazza M, Mac- chi V, Fiore D, Nussdorfer GG. Adrenocortical on- cocytoma: Case report and review of the literature. Tumori 84:403-407, 1998
15. Kitching PA, Patyel V, Harach HR. Adrenocortical on- cocytoma. J Clin Pathol 52:151-153, 1999
16. Alexander A, Paulose KP. Oncocytic variant of adrenal carcinoma presenting as Cushing’s syn- drome. J Assoc Physicians India 47:351-352, 1999
17. Scheibel M, Cavaliere A, Sidoni A, Bellezza G, Bartoli G, Bucciarelli E. Adenoma oncocitario del surrene. Riferimento di un caso e revisione della letteratura. Pathologica 93:407 (Abstract), 2001
18. Kurek R, Von Knobloch R, Feek U, Heidenreich A, Hofmann R. Local recurrence of an oncocytic adrenocortical carcinoma with ovary metastasis. J Urol 166:985, 2001
19. Icard P, Louvel A, Le Charpentier M, Chapuis Y. Tumeurs surrénaliennes à cellules oncocytaires: Bénignité ou malignité? Ann Chir 126:249-253, 2001
20. Wu SL, Biddle DA, Ordonez NG, Krishnamurthy S, Ayala AG, Ro JY. Oncocytic adrenocortical neo- plasms: A clinicopathologic and immunohisto- chemical study of seven cases. Mod Pathol 15:121A (Abstract), 2002
21. Hoang MP, Ayala AG, Albores-Saavedra J. Oncocytic adrenocortical carcinoma: A morphologic, im- munohistochemical and ultrastructural study of four cases. Mod Pathol 15:973-978, 2002
22. Krishnamurthy S, Ordonez NG, Shelton TO, Ayala AG, Sneige N. Fine-needle aspiration cytology of a case of oncocytic adrenocortical carcinoma. Diagn Cytopathol 24:222-223, 2001
23. Segal S, Cytron S, Shenav S, Gemer O. Adrenocorti- cal oncocytoma in pregnancy. Obstet Gynecol 98:916-918, 2001
24. Erlandson RA. Diagnostic electron microscopy of tu- mors with clinicopathological, immunohistochem- ical, and cytogenetic correlations. Raven Press, New York, pp. 97-99, 1994
25. Kay S. Hyperplasia and neoplasia of the adrenal gland. Pathol Ann 11:103-139, 1976
26. Hough AJ, Hollifield JW, Page DL, Hartmann WH. Prognostic factors in adrenal cortical tumors. A mathematical analysis of clinical and morphologic data. Am J Clin Pathol 72:390-399, 1979
27. Gandour MJ, Grizzle WE. A small adrenocortical car- cinoma with aggressive behaviour. An evaluation
of criteria for malignancy. Arch Pathol Lab Med 110:1076-1079, 1986
28. Nakano M. Adrenal cortical carcinoma. A clinico- pathologic and immunohistochemical study of 91 autopsy cases. Acta Pathol Jpn 38:163-180, 1988
29. Luton J-P, Cerdas S, Billaud L, et al. Clinical features of adrenocortical carcinomas, prognostic factors, and the effect or mitotane therapy. N Engl J Med 322:1195-1201, 1990
30. Medeiros LJ, Weiss LM. New developments in the pathologic diagnosis of adrenal cortical neoplasm. A review. Am J Clin Pathol 97:73-83, 1992
31. van Slooten H, Schaberg A, Smeenk D, Moolenaar AJ. Morphologic characteristics of benign and ma- lignant adrenocortical tumors. Cancer 55:766-773, 1985
32. Weiss LM, Medeiros LJ, Vickery AL. Pathologic fea- tures of prognostic significance in adrenocortical carcinoma. Am J Surg Pathol 13:202-206, 1989
33. Aubert S, Wacrenier A, Leroy X, Devos P, Carnaille B, Proye C, Wemeau JL, Lecomte-Houcke M, Le- teurtre E. Weiss system revisited. A clinicopatho- logic and immunohistochemical study of 49 adrenocortical tumors. Am J Surg Pathol 26:1612-1619, 2002
34. Medeiros LJ, Weiss LM. Adrenal gland: Tumors and tumor-like lesions. In Weidner N (ed): The difficult diagnosis in surgical pathology. WB Saunders, Philadelphia, pp. 377-407, 1996
35. Vargas MP, Vargas HI, Kleiner DE, Merino MJ. Adrenocortical neoplasms: Role of prognostic markers MIB-1, p53 and RB. Am J Surg Pathol 21:556-562, 1997
36. Goldblum JR, Shannon R, Kaldjian EP, Thiny M, Daven- port R, Thompson N, Lloyd RV. Immunohistochemical assessment of proliferative activity in adrenocortical neoplasms. Mod Pathol 6:663-668, 1993
37. Li M, Wenig BM. Adrenal oncocytic pheochromocy- toma. Am J Surg Pathol 24:1552-1557, 2000
38. Weiss LM, Gaffey MJ, Warhol MJ, Mehta P, Bonsib SM, Bruder E, Santos E, Medeiros LJ. Immunocy- tochemical characterization of a monoclonal anti- body directed against mitochondria reactive in paraffin-embedded sections. Mod Pathol 4:596-601, 1991
39. Lack EE. Tumors of the adrenal gland and extra- adrenal paraganglia. In Atlas of tumor pathology. Fascicle 19, 3rd Series. Armed Forces Institute of Pathology, Washington DC (a) p. 147; (b) 118-119, 1995
40. Bonsib SM, Lager DJ. Chromophobe cell carcinoma: Analysis of five cases. Am J Surg Pathol 14:260-267, 1990
41. Mackay B, el-Naggar A, Ordonez NG. Ultrastructure of adrenal cortical carcinoma. Ultrastruct Pathol 18:181-190, 1994