Accepted Manuscript
Pitfalls in the Diagnosis of Adrenocortical Tumors: A Lesson from 300 Consultation Cases
Eleonora Duregon MD, Marco Volante MD, PHD, Enrico Bollito MD, Margherita Goia MD, Consuelo Buttigliero MD, Barbara Zaggia MD, Alfredo Berruti MD, Giorgio Vittorio Scagliotti MD, Mauro Papotti MD
Human PATHOLOGY
W: B. Saunders
| PII: | S0046-8177(15)00318-4 |
| DOI: | doi: 10.1016/j.humpath.2015.08.012 |
| Reference: | YHUPA 3674 |
| To appear in: | Human Pathology |
| Received date: | 3 April 2015 |
| Revised date: | 6 August 2015 |
| Accepted date: | 19 August 2015 |
Please cite this article as: Duregon Eleonora, Volante Marco, Bollito Enrico, Goia Margherita, Buttigliero Consuelo, Zaggia Barbara, Berruti Alfredo, Scagliotti Giorgio Vittorio, Papotti Mauro, Pitfalls in the Diagnosis of Adrenocortical Tu- mors: A Lesson from 300 Consultation Cases, Human Pathology (2015), doi: 10.1016/j.humpath.2015.08.012
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
PITFALLS IN THE DIAGNOSIS OF ADRENOCORTICAL TUMORS: A LESSON FROM 300 CONSULTATION CASES
Eleonora Duregon, MD, Marco Volante, MD, PhD, Enrico Bollito, MD, Margherita Goia, MD, Consuelo Buttigliero, MD, Barbara Zaggia*, MD, Alfredo Berruti^, MD, Giorgio Vittorio Scagliotti, MD, Mauro Papotti, MD
Department of Oncology and of *Clinical and Biological Sciences, University of Turin at San Luigi Hospital, Orbassano, 10043, Turin, Italy; ^Current Affiliation: Medical Oncology, University of Brescia, Brescia, 25100, Italy
Email addresses: eleonora.duregon@unito.it; marco.volante@unito.it; e.bollito@libero.it; margherita.goia@gmail.com; consuelo.buttigliero@gmail.com; barbara.zaggia@gmail.com; alfredo.berruti@gmail.com; giorgio.scagliotti@unito.it; mauro.papotti@unito.it
Conflicts of Interest and Source of Funding: All Authors have no potential conflicts of interest to declare. This study is supported by grants from AIRC, Milan (no. IG/14820/2013 to MP).
Running headline: Pitfalls in ACC diagnosis
Address for correspondence: Marco Volante, MD Department of Oncology, University of Turin at San Luigi Hospital, Regione Gonzole 10, 10043 Orbassano, Torino, Italy Phone +390116705464; Fax: +390119026753; email: marco.volante@unito.it
ACCEPTED MANUSCRIPT
ABSTRACT
The correct pathological classification of adrenocortical carcinoma is relevant to establish an early therapeutic strategy of this rare malignancy. Aim of the study was to assess the most frequent pitfalls in adrenocortical carcinoma diagnosis reviewing a large consecutive series of 300 cases with an original diagnosis or a clinical suspect of adrenocortical carcinoma, that were sent in consultation to our Institution between 2004 and 2014. A major disagreement which significantly modified the clinical management of patients was recorded in 26 cases (9%). The most common pitfall (10 cases) was to distinguish adrenocortical carcinoma from pheochromocytoma and vice versa. Seven other cases diagnosed as adrenocortical carcinoma were reclassified as metastases from other primaries and primary adrenal soft tissue tumors (including 3 angiosarcomas). Finally, five adrenocortical adenomas were reclassified into carcinomas and four adrenocortical carcinomas converted into adenomas. Minor disagreements were mostly related to the identification of adrenocortical carcinoma variants (up to 32% of cases of adrenocortical tumors in the present series). Moreover, more than 50% of adrenocortical carcinoma cases lacked Ki-67. In conclusion, our results indicate that, in the presence of a histologically suspected adrenocortical carcinoma, a special attention should be devoted to exclude metastatic and soft tissue tumors and pheochromocytoma (in this latter case with special reference to the oncocytic variant of adrenocortical tumors). Moreover, pathologists should be aware of the major role of Ki-67 in determining prognosis and in selecting patients to the most appropriate treatment.
KEY WORDS: Adrenocortical carcinoma, adrenal gland, differential diagnosis, pathology consultation
ACCEPTED MANUSCRIPT
1. INTRODUCTION
Adrenocortical carcinoma is a rare malignancy with an incidence of approximately one case (range 0.7-2.0) per million population [1]. Because of its aggressive behavior, distant metastases at the time of diagnosis are relatively common and, in this context, the main diagnostic issue is the exclusion of the reverse condition (i.e. other malignant neoplasms metastatic to the adrenal), which is also relatively common for lung and other cancers. Conversely, in the presence of an adrenal-confined disease, the diagnostic difficulties rely on differentiating adrenocortical from adrenomedullary tumors on the one side, and on taking adrenocortical adenomas apart from carcinomas, on the other. In this regard, the pathological diagnosis of “carcinoma” requires the recognition of multiple morphological parameters combined either in scoring systems [2-4] or in diagnostic algorithms [5,6,7], suggestive but not pathognomonic per se of malignancy. Moreover, some cases lack enough parameters for reaching a proven diagnosis of malignancy. These cases are worrisome from both pathological and clinical standpoints, underlying the need of a more unequivocal pathological classification and posing questions on their appropriate clinical management and follow up [8]. A revision of 387 patients from the German Registry of adrenocortical carcinoma revealed a partial inaccuracy in histopatological diagnoses. In 161 cases (42%), a second opinion on the tumor specimen was requested by the reference pathologist and in 21 patients (13%) the diagnosis of adrenocortical carcinoma was changed to metastases from other malignancies, malignant pheochromocytoma (also termed intraadrenal paraganglioma), renal cell carcinoma or sarcoma [9]. A correct histopathological diagnosis is fundamental to define the proper treatment and to improve outcomes. The rarity of adrenocortical carcinoma and the reported relatively high rate of misdiagnoses underline the importance of referring pathology specimens to specialized centers, particularly in the case of non functioning adrenal tumors. The aim of this study was to assess the most frequent histopathological pitfalls in adrenocortical
ACCEPTED MANUSCRIPT
tumor diagnosis by revising a large series of 300 consultation cases from a single Institution. We here show that: a) a second opinion on surgically resected adrenocortical tumors can determine changes of the pathological diagnosis in 9% of cases, which significantly alter the therapeutic strategies and clinical management of patients; b) major pitfalls are related to the misdiagnosis of pheochromocytoma versus adrenocortical carcinoma, but a high suspicion level is recommended for metastatic and soft tissue tumors, as well; c) among primary adrenocortical tumors, cases with borderline malignant features or the special variants (e.g. myxoid or oncocytic) are the most problematic entities.
2. MATERIALS AND METHODS
2.1 Case series.
The Divisions of Internal Medicine and Medical Oncology of the San Luigi Hospital and University of Turin serve as a referral Center for adrenocortical carcinoma diagnosis and treatment in Italy. Between January 2004 and July 2014, the Pathology Unit of this University Hospital received 300 consecutive consultation cases, related to patients with proven or suspected adrenocortical carcinoma that sought medical advice at the Center. Cases were submitted with a median of nine hematoxylin and eosin stained slides per case (range 3-20) and four immunohistochemical stains per case (range 0-10). Among them, 256 cases (85%) were sent by clinicians (mostly endocrinologists and oncologists internal at San Luigi Hospital) at the time of first visit for advice and/or cure, and 44 (15%) were submitted by pathologists from Italian or European Institutions for the purpose of a second opinion. All cases were reviewed by four of us (ED, MV, EB, MP) having a specific experience in endocrine or urologic pathology, and were classified according to the appropriate diagnostic system, i.e. Weiss Score [2] for conventional and myxoid adrenocortical carcinoma, Lin-Weiss-Bisceglia system [10] for pure oncocytic adrenocortical tumors and Wieneke
ACCEPTED MANUSCRIPT
classification [11] for the two pediatric tumors. In the case of disagreement among the reviewers, each tumor was jointly discussed at a multihead microscope and a consensus was reached. The present study was approved by the Local Ethical Committee at San Luigi Hospital.
2.2 Histochemistry and immunohistochemistry.
Depending on the morphological features and/or the availability of additional unstained slides or paraffin block(s), a panel of reactions, comprising histochemical reticulin stain and immunohistochemical markers, was applied in 221/300 cases. The latter included the adrenocortical-specific molecules melan-A, a-inhibin and steroidogenic factor-1 (SF-1), the proliferation-associated marker Ki-67 (assessed in hot-spots as the percentage of positive nuclei counting at least 2000 tumor cells), as well as organ-specific markers (TTF-1, others) and markers of epithelial or mesenchymal lineages, whenever appropriate. The immunohistochemical markers more frequently added to the available slides are listed in Table 1.
3. RESULTS
3.1 Diagnoses after revision.
Upon revision, the whole series of 300 cases was re-classified as follows: 269 adrenocortical carcinomas, 14 adrenocortical adenomas, seven adrenocortical tumors of uncertain malignant potential, three pheochromocytomas, five soft tissue tumors in the adrenal gland and two metastases from lung and breast carcinomas. An agreement between the original and consultation diagnoses was observed in 274 cases (91%), including minor discrepancies on the Weiss score values and/or the exact tumor variant recognition (see below). In the remaining tumors, a change in the diagnosis upon revision was made, which significantly modified the clinical management of patients (Table 2).
ACCEPTED MANUSCRIPT
The original pathologic report was available in 205 cases (68%). The diagnostic system used (Weiss score, or Lin-Weiss-Bisceglia system for oncocytic tumors) was reported in 41/205 (20%) cases only. However, in 144 cases the microscopic description included morphological parameters which are part of Weiss system (nuclear atypia, 87 cases; necrosis, 84 cases; high mitotic rate, 81 cases; vascular invasion, 70 cases). The exact mitotic count was on record in 68 reports, only. In addition, other parameters not incorporated in the Weiss system were occasionally listed (e.g. hemorrhage, fibrous bands). Moreover, among 269 ACC cases, 14 and 12 only had TNM and status of resection margins reported, respectively. Immunohistochemistry was performed by the original pathologists in 108 cases (36%), using a variable panel of antibodies which most commonly included pan- cytokeratins, vimentin, chromogranin A, synaptophysin, melan A and alpha-inhibin. The Ki-67 proliferation index was originally performed in 100 cases (30%).
3.2 Discordant cases.
A major disagreement was recorded in 26 cases (9%), all but one non-functioning tumors. All four reviewing pathologists (ED, MV, EB, MP) had a 100% agreement in these 26 cases.
The major diagnostic issues were divided into three categories (Table 3): 1) non-endocrine tumors in the adrenal gland, 2) medullary versus cortical-derived adrenal tumors, 3) adrenocortical adenoma versus carcinoma.
3.2.1 Non-endocrine tumors in the adrenal gland. Seven cases were originally classified as adrenocortical carcinoma, but after revision they turned to be soft tissue tumors primary of the adrenal (3 angiosarcomas, 1 PEComa and 1 solitary fibrous tumor) or metastases from other primaries (2 cases). Concerning the three angiosarcomas, they were all submitted for revision by clinicians in the presence of an hemorrhagic adrenal mass (two cases) and synchronous adrenal and lung masses. Histologically, sheets of markedly atypical, large epithelioid cells with abundant
ACCEPTED MANUSCRIPT
amphophilic cytoplasm, vesicular nuclei and prominent nucleoli were arranged in a solid growth with few vasculogenic areas with irregular anastomosing vessel and intracytoplasmic vacuoles containing red blood cells; intratumoral hemorrhage and necrotic areas were predominant. Vascular differentiation markers, such as CD34, CD31 and FLI-1, all stained positive, in the presence of negativity for adrenocortical and medullary markers (Figure 1).
Two other consultation cases were submitted for second opinion by pathologists based on non-
features and consistent immunohistochemical profile, the two cases were diagnosed as a solitary fibrous tumor and a PEComa, this latter with predominant clear cells and focal oncocytic features. The two metastatic tumors (sent one from clinician and one from pathologist) were from lung and breast cancers, and both were undifferentiated carcinomas that mimicked an adrenocortical carcinoma. The patient with metastatic breast cancer had a history of breast cancer operated five years before and developed in the last year signs of hypercortisolism. Indeed, the ultimate responsible for the adrenal hyperfunction was an adrenocortical macronodular hyperplasia, with one of the nodules extensively infiltrated by highly atypical cells that induced the suspicion of a functioning adrenocortical carcinoma. However, estrogen and progesterone receptors were positive whereas adrenocortical markers (SF-1 and Melan A) were negative, pointing to a breast cancer metastasis. The second case was a large cell neuroendocrine lung cancer metastasis to the adrenal (all adrenocortical markers negative with a positivity for pan-cytockeratin and chromogranin A).
3.2.2 Medullary versus cortical tumors. Ten out of 26 cases (38% of discordant cases) were related to the erroneous diagnosis of adrenocortical adenoma or carcinoma versus pheochromocytoma, and vice versa.
ACCEPTED MANUSCRIPT
Two cases with an original diagnosis of adrenocortical carcinoma had a clinical history of hypertension not otherwise specified and a similar morphologic appearance characterized by a solid growth of large pleomorphic cells, prominent vascular invasion, necrosis, hemorrhage and focal cystic degeneration. In one of these patients (Figure 2) a lung lesion was first submitted by the Thoracic Surgery Division of our Hospital. The intra-operative consultation could not confirm the initial diagnosis of adrenocortical carcinoma metastasis and was deferred. At permanent sections, a diagnosis of a lung metastasis from malignant pheochromocytoma was made based on morphology and appropriate immunophenotype (negative adrenocortical markers melan A and SF-1, and positive chromogranin A and synaptophysin) (Figure 2a to 2c). Upon subsequent revision, the primitive tumor was diagnosed as a malignant pheochromocytoma (Figure 2d to 2f). A third case of this group was misdiagnosed as an oncocytic adrenocortical adenoma because of prominent cytoplasmic eosinophilia and minimal atypias, indeed a rare occurrence in classical pheochromocytoma.
Seven other tumors had a microscopic description suggestive of the initial diagnosis of malignant pheochromocytoma. In four cases, sustentacular cells were also described. Ki-67 was performed in all cases, ranging from 2 to 70%. Immunohistochemistry consisted of a variable panel of markers which included synaptophysin and vimentin (always positive), chromogranin A, S100 and epithelial markers (always negative), whereas no adrenocortical markers were tested. All but two cases were submitted for revision by the clinicians, because the diagnosis of pheochromocytoma was questioned in the absence of a consistent clinical picture. Of the remaining two cases, one had a first diagnosis of pheochromocytoma with a recurrence diagnosed as adrenocortical carcinoma, while for the second case there was a diagnosis of pheochromocytoma on the right side and of a controlateral adrenocortical carcinoma ten years later, which gave a lung metastasis after another five years. Upon revision, all cases were classified as malignant adrenocortical tumors with Weiss
ACCEPTED MANUSCRIPT
scores ranging from 3 to 8 points and positive adrenocortical markers. It is worth to notice that in this group of “pheochromocytoma mimickers”, three out of seven cases belonged to a special variant of adrenocortical carcinoma. In particular, two cases were pure oncocytic and the remaining case was a sarcomatoid variant with concurrent heterogeneous patterns resembling pheochromocytoma, conventional adrenocortical carcinoma and spindle cell sarcoma.
3.2.3 Adrenocortical adenoma versus carcinoma. Nine cases were adrenocortical tumors whose interpretation was changed, being four carcinomas reclassified as adenomas and five adenomas converted into carcinomas. All cases initially diagnosed as carcinomas were of large size (from 7 to 12.5 cm). In two cases, the diagnosis of malignancy was supported by the detailed description of morphological features, such as high mitotic rate, capsular and vascular invasion, which were not confirmed upon revision. After revision, one case had a Weiss Score of 0, the remaining three a Weiss score of 1 or 2, based on nuclear atypia alone or with predominant dark cell cytoplasm. The Ki-67 index ranged from 1 to 5%, although “hot spots” of higher proliferation (up to 10%) were focally seen in one case.
Four out of the five adenomas converted into carcinomas were oncocytic tumors, all but one in the pure form (>90% oncocytic cells). They all developed one or more recurrences from 33 to 89 months after initial diagnosis. Upon revision, malignancy was determined by the presence of two out of three major criteria included in the Lin-Weiss-Bisceglia system, i.e. increased mitotic activity and rare atypical mitoses, whereas the third major criterion (presence of vascular invasion) was absent. The last case was an aldosterone-secreting neoplasm, with an original diagnosis of adenoma Weiss Score 1 because of few limited necrotic areas. Upon revision, spotty necrosis was indeed present, but nuclear atypia, diffuse growth pattern, clear cells <25% as well as reticulin framework alterations were also identified, thus supporting a diagnosis of carcinoma with a final Weiss Score of 4.
ACCEPTED MANUSCRIPT
3.3 Minor disagreements.
In 206/274 cases (75%), the diagnosis was not modified although minor disagreements were recorded. These were mainly related to the failure of recognizing an adrenocortical carcinoma morphological variant (49 cases) Myxoid features were originally reported in four cases, only, but upon revision 25 cases of the myxoid variant of adrenocortical carcinoma (9% of the cases overall) were recognized, 13 of which in the pure form. Predominant oncocytic features were originally reported in 22 cases. Nevertheless, the appropriate Lin-Weiss-Bisceglia diagnostic system, specifically designed for adrenocortical oncocytic tumors, was employed in two cases only. Upon revision, the number of pure oncocytic cases increased to 50 (18% of adrenocortical neoplasms, overall; 41 malignant, 7 of uncertain malignant potential and 2 benign), whereas 16 tumors had mixed features (oncocytic component range: 50-90%).
In 174 cases, in the presence of an otherwise confirmed diagnosis of adrenocortical carcinoma, the Ki-67 labeling index was missing. Although it is not recommended for the diagnosis of adrenocortical carcinoma and therefore the lack of Ki-67 data cannot be considered a diagnostic disagreement, it is worth noticing that 124 of such cases were sent by the clinicians exclusively to have this additional information, mandatory to select patients to the most appropriate medical treatment.
3.4 Adrenocortical tumors of uncertain malignant potential.
Eight cases (2.6%) were submitted for revision with a diagnosis of “tumor of uncertain malignant potential” (UMP), an event which can be expected in a large consultation series of adrenocortical tumors. Interestingly, all but one were purely oncocytic neoplasms, and according to the Lin- Weiss-Bisceglia system were classified as UMP. Follow up data were available for two of these
ACCEPTED MANUSCRIPT
patients, only, both alive with no evidence of disease at 48 and 50 months, respectively. The latter case was a adrenocortical tumor of a four months child. This case was classified as benign according to the classification of Wieneke [11,12] due to the presence of focal capsular invasion and necrosis which are not considered in pediatric tumors indicative alone of malignancy.
In this study, the importance of a second opinion on surgical pathology specimens of adrenocortical tumors has been confirmed, at least for those cases having clinical controversial features. In fact, major disagreements were observed in 9% of 300 consecutive consultation cases, with significant prognostic and therapeutic implications.
VENDO laga a. 4. DISCUSSION Three major areas of diagnostic difficulty and discrepancy were identified, which indeed represent the most challenging chapters of diagnostic pathology of the adrenal gland (i.e. adrenocortical tumors versus non endocrine tumors within the adrenal; cortical versus medullary tumors; adrenocortical carcinoma versus adenoma).
As a general comment, the vast majority of cases were sent for consultation by clinicians, either internal at the San Luigi Hospital which serves as national referral center for adrenal cancer or from external hospitals. This aspect influenced the characteristics of cases included in the study, since most of them were submitted not due to difficult or equivocal pathological features but to worrisome clinical findings or missing data (i.e. Ki-67 index) of clinical and not strictly diagnostic relevance. However, the selection criteria of this study are similar to the study based on the data of the German Registry of adrenocortical carcinoma [9], as it was the rate and type of misdiagnosed cases. Another general finding, irrespective or not if affecting the final diagnostic agreement, was the lack in 80% of cases of specific mention (and score) of the diagnostic system employed (Weiss or others), although the use of these systems is highly recommended by the
ACCEPTED MANUSCRIPT
WHO classification [13]. Similarly, the exact mitotic count was recorded in the original reports in 33% of cases only, even though mitotic count represents not only a key diagnostic parameter but also one of the most relevant prognostic features in adrenocortical carcinoma [6,14,15].
Primary non-epithelial adrenal tumors are extremely rare and indeed represent a diagnostic challenge in the absence of clear-cut morphological features, especially in patients lacking signs of hormonal activity. Among the most frequent soft tissue tumors involving the adrenal gland [16-18], epithelioid angiosarcoma is one of the most frequent and problematic [19-21]. In this setting, immunohistochemistry plays a major role. Pan-cytokeratins are not useful, being generally absent in at least half of adrenocortical carcinomas (positivity rate of 16% - 12/77 cases in our series) and conversely focally expressed in some vascular sarcomas. Nor is vimentin, which is commonly produced by adrenocortical carcinomas. As a first line, vascular markers are to be investigated especially in highly vascularized and/or hemorrhagic tumors, to exclude an angiosarcoma of the adrenal gland.
The adrenal gland is also a common site of distant spread of carcinomas of various origin, including lung, breast and others and a high index of suspicion is recommended in the presence of a highly undifferentiated adrenal neoplasm with non-conventional morphological features and/or controversial immunoprofile. In this respect, melan A, alpha-inhibin and calretinin are good markers of adrenocortical cell origin [22]. However, all of them are not adrenocortical specific. In particular, it should emphasized that melan A per se is not sufficient to prove the adrenocortical origin of a given tumor since it is also positive in melanoma whose metastases are not uncommon within the adrenal. In our experience the nuclear factor SF-1 proved highly reliable since it usually presents an intense and diffuse - thus unequivocal - pattern of reactivity, and is restricted to steroidogenic tissues and tumors, only. The two adrenal metastases in our series originated from lung and breast carcinomas. In both cases, the presence of poorly differentiated high grade lesions
ACCEPTED MANUSCRIPT
with solid growth was suggestive but not sufficient to prove a metastatic tumor to the adrenal. The unique morphological aspect supportive of a metastatic lesion was the absence of a tumor capsule, with neoplastic cells merging with the adjacent adrenal cortex, a feature which is invariably absent in adrenocortical carcinoma. The lung primary of the first case could be confirmed by TTF-1 immunohistochemistry, but this marker might assist only in the case of adenocarcinoma or large cell carcinomas with glandular differentiation. With regard to the breast carcinoma metastasis, the classical markers are steroid hormone receptors, but it is to be kept in mind that a fraction of adrenocortical carcinoma and adenomas may also express estrogen receptors [23,24]. Other markers of breast origin, such as the apocrine marker GCDFP-15 and above all GATA-3 [25], seem therefore more appropriate in this context.
As to concern, medullary versus cortical tumors, their correct separation on a pure morphological ground may be difficult in non-conventional cases, and it is worth to notice that this differential diagnosis represented more than 30% of our disagreement cases and 3% of all consultations. The clinical history (e.g. hypertension) may be of help favoring a diagnosis of pheochromocytoma, but only if supported by laboratory data, since cortisol or aldosterone-producing adrenocortical carcinomas are associated to hypertension, as well. In particular, ACTH-secreting pheochromocytomas, although very rare, may represent a diagnostic pitfall both on the clinical and pathological ground [26-28] Moreover, both adrenocortical carcinoma and pheochromocytoma are frequently discovered as incidentalomas, have similar mean size (4 to 5 cm), have usually unspecific imaging findings, and show sometimes overlapping macroscopic (encapsulation, hemorrhage) and microscopic (a solid, diffuse, compact or nested growth of large pleomorphic cells) features. Moreover, mitotic index is a crucial parameter for adrenocortical carcinoma diagnosis, but malignant pheochromocytoma may also have high mitotic and proliferative (Ki-67 index) rates. Therefore, a high level of suspicion should be sustained by the use
ACCEPTED MANUSCRIPT
of appropriate immunohistochemical markers, which easily address the correct diagnosis. In this setting, we recommend to rely on SF-1 and/or Melan A and chromogranin A, which are mutually exclusive in adrenocortical carcinoma and pheochromocytoma, whereas alternative markers for the latter cannot include synaptophysin, that is also expressed by a large fraction of adrenocortical carcinomas. Among other neuroendocrine markers, neurofilaments represent another potential pitfall, since they may be present in pheochromocytoma, but have also been reported in adrenocortical carcinoma, with special reference to the myxoid type [29].
Pitfalls and controversies on the correct application of the diagnostic scores, with specific regard to the individual Weiss parameters, have been extensively analyzed [8,30]. Nevertheless, still a large grey area exists between benign and malignant cases, although from the present series it seems that this diagnostic issue is mostly related to the oncocytic variant of adrenocortical tumors. In fact, 4 out of 9 cases re-classified as benign vs malignant or vice versa were purely oncocytic. Although oncocytic tumors are classified more reliably using the Lin-Weiss-Bisceglia than the Weiss system [10,31], a very small fraction only of oncocytic cases in the present series was originally classified using this system. Moreover, all cases sent for consultation as UMP were oncocytic, except for one pediatric case.
Minor disagreements were mostly related to the identification of adrenocortical carcinoma variants, which are considered a very rare event, assuming that adrenocortical carcinoma is a rare tumor per se. However, they represent overall 32% of cases of adrenocortical tumors in the present series and a growing body of evidence support the concept that carcinoma variants (namely the myxoid and oncocytic ones) differ from conventional adrenocortical carcinoma not only in terms of morphology but also in terms of molecular features [31,32], prognosis [33] and possibly also response to treatment.
ACCEPTED MANUSCRIPT
With regard to Ki-67 index, cut-off values have been reported at around 4% for distinguishing adenomas from carcinomas, but proliferation index is not a major parameter in terms of diagnostic categorization within adrenocortical tumors. However, nearly 50% (124/269) adrenocortical carcinomas in the present series were sent in consultation exclusively since this data was missing. This finding strongly enforce the double role of Ki-67 as a clinically relevant parameter. In fact, recent data from the literature suggest that Ki-67 is a strong and independent prognostic parameter in adrenal cancer [15,34]. Not less important, Ki-67 is among the most relevant parameters used by clinicians to define “high risk” patients to be addressed to adjuvant mitotane treatment [35], or to select the most appropriate treatment (mitotane alone or in association with chemotherapy) in patients with advanced or progressive disease.
5. CONCLUSIONS
In conclusion, the original pathological diagnosis of adrenocortical tumors was changed in up to 9% of cases upon revision, supporting the need of performing adequate immunohistochemical panel (with special reference to markers tracing adrenal cortical and medullary origin) especially in cases of non-secreting adrenal tumors with equivocal morphology, and the essentiality of the availability of specific biological tests and imaging for a correct pathological diagnosis. Key issues include: a) a high index of suspicion for metastatic and soft tissue tumors; b) a special attention for tumors showing a predominantly oncocytic phenotype in order to either assessing malignancy or distinguishing from pheochromocytomas; c) knowledge of the existence and potential clinical relevance of adrenocortical carcinoma histological variants; d) awareness of the major role of Ki- 67 in determining prognosis and in selecting patients to the most appropriate treatment.
ACCEPTED MANUSCRIPT
REFERENCES
[1] Fassnacht M, Kroiss M, Allolio B. Update in adrenocortical carcinoma. J Clin Endocrinol Metab 2013;98:4551-4.
[2] Weiss LM. Comparative histologic study of 43 metastasizing and nonmetastasizing adrenocortical tumors. Am J Surg Pathol 1984;8:163-9.
[3] Weiss LM, Medeiros LJ, Vickery AL, Jr. Pathologic features of prognostic significance in adrenocortical carcinoma. Am J Surg Pathol 1989;13:202-6.
[4] van Slooten H, Schaberg A, Smeenk D, Moolenaar AJ. Morphologic characteristics of benign and malignant adrenocortical tumors. Cancer 1985;55:766-73.
[5] Blanes A, Diaz-Cano SJ. Histologic criteria for adrenocortical proliferative lesions: value of mitotic figure variability. Am J Clin Pathol 2007;127:398-408.
[6] Volante M, Bollito E, Sperone P, Tavaglione V, Daffara F, Porpiglia F, et al. Clinicopathological study of a series of 92 adrenocortical carcinomas: from a proposal of simplified diagnostic algorithm to prognostic stratification. Histopathology 2009;55:535-43.
[7] Duregon E, Fassina A, Volante M, Nesi G, Santi R, Gatti G, et al. The Reticulin Algorithm for Adrenocortical Tumor Diagnosis: A Multicentric Validation Study on 245 Unpublished Cases. Am J Surg Pathol 2013;37:1433-40.
[8] Papotti M, Duregon E, Volante M, McNicol AM. Pathology of the adrenal cortex: a reappraisal of the past 25 years focusing on adrenal cortical tumors. Endocr Pathol 2014;25:35-48.
[9] Johanssen S, Hahner S, Saeger W, Quinkler M, Beuschlein F, Dralle H, et al. Deficits in the management of patients with adrenocortical carcinoma in Germany. Dtsch Arztebl Int 2010;107:885-91.
ACCEPTED MANUSCRIPT
[10] Bisceglia M, Ludovico O, Di Mattia A, Ben-Dor D, Sandbank J, Pasquinelli G, et al. Adrenocortical oncocytic tumors: report of 10 cases and review of the literature. Int J Surg Pathol 2004;12:231-43.
[11] Wieneke JA, Thompson LD, Heffess CS. Adrenal cortical neoplasms in the pediatric population: a clinicopathologic and immunophenotypic analysis of 83 patients. Am J Surg Pathol 2003;27:867- 81.
[12] Magro G, Esposito G, Cecchetto G, Dall’Igna P, Marcato R, Gambini C, et al. Pediatric adrenocortical tumors: morphological diagnostic criteria and immunohistochemical expression of matrix metalloproteinase type 2 and human leucocyte-associated antigen (HLA) class II antigens. Results from the Italian Pediatric Rare Tumor (TREP) Study project. Hum Pathol 2012;43:31-9.
Via [13] DeLellis RA, Lloyd R, Heitz PU, et al. WHO Classification of Tumours of the Endocrine Organs; IARC Press, Lyon, 2004.
[14] Giordano TJ. The Argument for Mitotic Rate-based Grading for the Prognostication of Adrenocortical Carcinoma. Am J Surg Pathol 2011;35:471-3.
[15] Duregon E, Molinaro L, Volante M, Ventura L, Righi L, Bolla S, et al. Comparative diagnostic and prognostic performances of the hematoxylin-eosin and phospho-histone H3 mitotic count and Ki-67 index in adrenocortical carcinoma. Mod Pathol 2014;27:1246-54.
[16] Huwait H, Meneghetti A, Nielsen TO. Kaposi sarcoma of the adrenal gland resembling epithelioid angiosarcoma: a case report. Sarcoma 2011;2011:898257.
[17] Bala M, Ronchi CL, Pichl J, Wild V, Kircher S, Allolio B, et al. Suspected metastatic adrenocortical carcinoma revealing as pulmonary Kaposi sarcoma in adrenal Cushing’s syndrome. BMC Endocr Disord 2014;14:63.
ACCEPTED MANUSCRIPT
[18] Ho YH, Yap WM, Chuah KL. Solitary fibrous tumor of the adrenal gland with unusual immunophenotype: a potential diagnostic problem and a brief review of endocrine organ solitary fibrous tumor. Endocr Pathol 2010;21:125-9.
[19] Hayashi T, Gucer H, Mete O. A Mimic of Sarcomatoid Adrenal Cortical Carcinoma: Epithelioid Angiosarcoma Occurring in Adrenal Cortical Adenoma. Endocr Pathol 2014;25:404-9.
[20] Sung JY, Ahn S, Kim SJ, Park YS, Choi YL. Angiosarcoma arising within a long-standing cystic lesion of the adrenal gland: a case report. J Clin Oncol 2013;31:e132-6.
[21] Lepoutre-Lussey C, Rousseau A, Al Ghuzlan A, Amar L, Hignette C, Cioffi A, et al. Primary adrenal angiosarcoma and functioning adrenocortical adenoma: an exceptional combined tumor. Eur J Endocrinol 2012;166:131-5.
[22] Sangoi AR, Fujiwara M, West RB, Montgomery KD, Bonventre JV, Higgins JP, et al. Immunohistochemical distinction of primary adrenal cortical lesions from metastatic clear cell renal cell carcinoma: a study of 248 cases. Am J Surg Pathol 2011;35:678-86.
[23] Barzon L, Masi G, Pacenti M, Trevisan M, Fallo F, Remo A, et al. Expression of aromatase and estrogen receptors in human adrenocortical tumors. Virchows Arch 2008;452:181-91.
[24] Shen XC, Gu CX, Qiu YQ, Du CJ, Fu YB, Wu JJ. Estrogen receptor expression in adrenocortical carcinoma. J Zhejiang Univ Sci B 2009;10:1-6.
[25] Krings G, Nystrom M, Mehdi I, Vohra P, Chen YY. Diagnostic utility and sensitivity of GATA3 antibodies in triple-negative breast cancer. Human Pathology 2014;45:2225-32.
[26] Folkestad L, Andersen MS, Nielsen AL, Glintborg D. A rare cause of Cushing’s syndrome: an ACTH-secreting phaeochromocytoma. BMJ Case Rep 2014;2014.
[27] Cassarino MF, Ambrogio AG, Pagliardini L, De Martin M, Barresi V, Cavagnini F, et al. ACTH- secreting pheochromocytoma with false-negative ACTH immunohistochemistry. Endocr Pathol 2012;23:191-5.
ACCEPTED MANUSCRIPT
[28] Ballav C, Naziat A, Mihai R, Karavitaki N, Ansorge O, Grossman AB. Mini-review: pheochromocytomas causing the ectopic ACTH syndrome. Endocrine 2012;42:69-73.
[29] Papotti M, Volante M, Duregon E, Delsedime L, Terzolo M, Berruti A, et al. Adrenocortical tumors with myxoid features: a distinct morphologic and phenotypical variant exhibiting malignant behavior. Am J Surg Pathol 2010;34:973-83.
[30] Lau SK, Weiss LM. The Weiss system for evaluating adrenocortical neoplasms: 25 years later. Hum Pathol 2009;40:757-68. a day in ked ne ho vde
[31] Duregon E, Volante M, Cappia S, Cuccurullo A, Bisceglia M, Wong DD, et al. Oncocytic adrenocortical tumors: diagnostic algorithm and mitochondrial DNA profile in 27 cases. Am J Surg Pathol 2011;35:1882-93.
[32] Duregon E, Rapa I, Votta A, Giorcelli J, Daffara F, Terzolo M, et al. MicroRNA expression patterns in adrenocortical carcinoma variants and clinical pathologic correlations. Hum Pathol 2014;45:1555-62.
[33] Wong DD, Spagnolo DV, Bisceglia M, Havlat M, McCallum D, Platten MA. Oncocytic adrenocortical neoplasms-a clinicopathologic study of 13 new cases emphasizing the importance of their recognition. Hum Pathol 2011;42:489-99.
[34] Morimoto R, Satoh F, Murakami O, Suzuki T, Abe T, Tanemoto M, et al. Immunohistochemistry of a proliferation marker Ki67/MIB1 in adrenocortical carcinomas: Ki67/MIB1 labeling index is a predictor for recurrence of adrenocortical carcinomas. Endocr J 2008;55:49-55.
[35] Fassnacht M, Allolio B. What is the best approach to an apparently nonmetastatic adrenocortical carcinoma? Clin Endocrinol (Oxf) 2010;73:561-5.
ACCEPTED MANUSCRIPT
| Marker | Antibody/Reagent | Source aly | # cases |
|---|---|---|---|
| Reticulin stain | silver staining-based kit | Bio Optica, Milan, Italy | 28/300 |
| Melan A | clone A103, diluted 1/350, | aton, le DakoCytomation, Glostrup, DK | 63/300 |
| Alpha-inhibin | clone R1, diluted 1/75 | Diamedix, Miami, USA | 16/300 |
| SF-1 | clone N1665, diluted 1/100 | PRE Perseus Proteomics, Tokyo, Japan | 15*/300 |
| Synaptophysin | clone SY38, diluted 1/100 | Noch DakoCytomation, Glostrup, DK | 14/300 |
| Chromogranin A | clone LK2H10, diluted 1/500 | Abcam, Cambridge, USA | 22/300 |
| Ki-67 | ACCEPTEDASONAJE clone MIB-1, diluted 1/150 | DakoCytomation, Glostrup, DK | 167/300 |
*performed since 2012
ACCEPTED MANUSCRIPT
| Total consultation cases | 300 |
|---|---|
| Sender | Clinicians: 256 |
| Pathologists: 44 | |
| Agreement at revision | Concordance: 274 (91%) |
| 2004-2009: 116/117 (99%) | |
| 2010-2014: 158/183 (86%) | |
| Major discordance: 26 (9%) | |
| Major areas of diagnostic pitfall | Adrenal cortex vs non-endocrine tumors: 7/26 |
| Adrenal cortex vs medulla: 10/26 | |
| ACC vs ACA: 9/26 | |
| Diagnostic classification upon revision | ACA: 14 |
| UMP adrenocortical neoplasms: 7 | |
| ACC: 269 | |
| PCC: 3 | |
| Soft tissue tumors primary of the adrenal: 5 | |
| Metastases: 2 | |
| Adrenocortical tumor variants | Oncocytic: 66/290, 24% (50 pure/16 mixed) |
| Myxoid: 25/290, 9% (13 pure/12 mixed) | |
| Sarcomatoid: 2 |
Abbreviations. ACC: adrenocortical carcinoma; ACA: adrenocortical adenoma; UMP: uncertain malignant potential; PCC: pheochromocytoma.
| Incoming diagnosis | Outgoing diagnosis | Conclusive immunophenotype | |
|---|---|---|---|
| Adrenal cortex | ACC | epithelioid angiosarcoma | MELAN A -, CD31 + |
| vs non-endocrine tumors | ACC | epithelioid angiosarcoma | SF-1 -, MELAN A -, CD31 + |
| ACC | epithelioid angiosarcoma | SF-1 -, MELAN A -, CD31 + | |
| ACC | solitary fibrous tumor | MELAN A -, CD34 +, Bcl-2 + | |
| ACA | PEComa | SF-1 -, MELAN A +, HMB45 +/- | |
| ACC | pulmonary large cell neuroendocrine | MELAN A -, pancytokeratin +, chromogranin A +, TTF-1 +/- | |
| carcinoma metastasis | |||
| ACC | breast cancer metastasis MANUSKRIP. | MELAN A -, pancytokeratin +, estrogen receptor +, progesterone receptor + | |
| Adrenal cortex vs medulla | ACC | malignant PCC | SF1 -, MELAN A -, chromogranin A + |
| ACC | malignant PCC | SF1 -, MELAN A -, chromogranin A + | |
| ACA | PCC | MELAN A -, chromogranin A + | |
| malignant PCC | ACC (low tumor grade) | MELAN A +, Alpha-inhibin +, chromogranin A - | |
| malignant PCC | ACC (high tumor grade) | MELAN A +, Alpha-inhibin +/-, chromogranin A - | |
| malignant PCC | ACC (low tumor grade) | MELAN A +, chromogranin A - | |
| malignant PCC | ACC (high tumor grade) | MELAN A +, chromogranin A - | |
| malignant PCC | oncocytic ACC (low tumor grade) | MELAN A +, chromogranin A - | |
| malignant PCC | oncocytic ACC (low tumor grade) | MELAN A +, chromogranin A - | |
| malignant PCC | sarcomatoid ACC (high tumor grade) | SF1 +, MELAN A +, chromogranin A - | |
| ACC vs ACA | ACC | ACA (Weiss score 0) | Ki-67: 1% |
| ACC | ACA (Weiss score 1) | Ki-67: 6% | |
| ACC | ACA (Weiss score 1) | Ki-67: 1% | |
| ACC | ACA (Weiss score 2) | Ki-67: 3% | |
| ACA | ACC (low tumor grade) | Ki-67: 5% | |
| oncocytic ACA | oncocytic ACC (low tumor grade) | Ki-67: 15% | |
| oncocytic ACA | oncocytic ACC (low tumor grade) | Ki-67: 20% | |
| oncocytic ACA | oncocytic ACC (low tumor grade) | Ki-67: 33% | |
| oncocytic ACA | oncocytic ACC (low tumor grade) | Ki-67: 8% |
Abbreviations. ACC: adrenocortical carcinoma; ACA: adrenocortical adenoma; PCC: pheochromocytoma. Tumor grade in ACC cases was assessed according to reference [14]
ACCEPTED MANUSCRIPT
FIGURE LEGENDS
Figure 1. A case of adrenal tumor with extensive intratumoral hemorrhage (a), necrosis and a predominant solid pattern (b), originally diagnosed as adrenocortical carcinoma. However, few areas showed a vasculogenic pattern with irregular anastomosing vessel, vesicular nuclei and prominent nucleoli (c). These cells were unreactive for adrenocortical specific markers and positive for endothelial markers CD31 (d) and FLY1 (e), which led to a final diagnosis of epithelioidi angiosarcoma. (a, b, c: hematoxylin & eosin; original magnification: 100X, 200X and 400X, respectively; d and e: immunoperoxidase; original magnification: 200X and 400X, respectively).
Figure 2. A case of suspected pulmonary metastasis from an adrenocortical carcinoma diagnosed elsewere. The pulmonary lesion was made of compact nests of atypical eosinophilic cells (a) strongly positive for synaptophysin (b) but also for chromogranin A (c), thus supporting a diagnosis of metastatic neuroendocrine neoplasm. The primary adrenal tumor was subsequently revised and consisted of a proliferation of cells arranged in nests (d) invasive into the periadrenal fat (e) and negative for adrenocortical markers, such as Melan A (positive internal control of normal adrenal at the top of figure) (f). (a, d, e: hematoxylin & eosin; original magnification: a and e, 100X; d, 200X; b, c, f: immunoperoxidase; original magnification: 200X).
ACCEPTED MANUSCRIPT
A
2
C
&
E
AC
ACCEPTED MANUSCRIPT
A
5
E
F
ACC