Search

A reappraisal of the major advances in the diagnostic pathology of adrenal cortical lesions and tumors in the last 25 years is presented, with special reference to the definition of malignancy in primary adrenal cancer and its variants. Slightly more than 25 years ago, Weiss proposed his diagnostic scoring system for adrenal cortical carcinoma. This represented a milestone for adrenal pathologists and the starting point for further modifications of the system, either through minor changes in the scoring procedure itself or concentrating on some particular Weiss criterion such as mitotic index, integrated into alternative scoring schemes or algorithms that are currently under validation. Improvements in diagnostic immunohistochemistry have led to the identification of markers of cortical origin, such as Melan-A, alpha-inhibin, and SF-1 and of prognostic factors in carcinoma, such as the Ki-67 proliferation index and SF-1 itself. With regard to hyperplastic conditions, genetic investigations have allowed the association of the majority of cases of primary pigmented nodular adrenocortical disease (PPNAD) in Carney complex to mutations in the gene encoding the regulatory subunit 1A of protein kinase A (PRKAR1A). Other hereditary conditions are also associated with adrenal cortical tumors, including the Li-Fraumeni, Beckwith-Wiedemann, Gardner, multiple endocrine neoplasia type 1, and neurofibromatosis type 1 syndromes. Moreover, several advances have been made in the knowledge of the molecular background of sporadic tumors, and a number of molecules/genes are of particular interest as potential diagnostic and prognostic biomarkers.

We report a follicular carcinoma of thyroid that showed a range of histologic appearances, with microfollicular, macrofollicular/pseudopapillary, oncocytic, and poorly differentiated areas. We used comparative genomic hybridization to detect the major DNA copy number changes in each component, in order to study the inter-relationships among them. All showed gains in 11q and 17q, suggesting that these were early events in the development of the tumor, and these were the only changes in the follicular component. The other components each showed additional gains and losses, some unique to one component. The oncocytic component showed most changes, including loss on 16q in the region of the E-cadherin gene. This was associated with reduced intensity of immunostaining for E-cadherin specifically in that component. No mutations in the E-cadherin gene were detected in this component. The demonstration that some DNA copy number changes are consistent across each component suggests that they are all clonally related. The additional chromosomal and immunohistochemical heterogeneity across the macrofollicular/pseudopapillary, oncocytic, and poorly differentiated components would be consistent with the emergence of subclones, possibly as part of tumor progression.

The adrenal gland is not a common specimen in surgical pathology practice as, until recently, adrenal tumors were recognized in life only if associated with hypersecretion of hormones or evidence of malignancy. However, adrenal nodules are not uncommon at autopsy, and the number of these found in life is now increasing as they are identified when the abdomen is scanned for the investigation of other diseases using computed tomography or magnetic resonance imaging. It is therefore becoming increasingly important for the surgical pathologist to be aware of the range of pathology in the gland and to understand how to approach the specimens. This short review will deal with lesions of the adrenal cortex.

Paragangliomas arise from sympathetic or parasympathetic paraganglia and should now be defined by their site and type. The term pheochromocytoma is reserved for intra-adrenal tumors. This short review discusses the gross and microscopic features, the immunohistochemical profile, the problem of recognizing malignant potential, and the rare instances where a differential diagnosis has to be considered.

The roles of the pathologist in assessing adrenal cortical tumors are, first, to differentiate adenoma from carcinoma, and, second, to assess prognosis when the diagnosis of malignancy is made. How easy is it to achieve these goals with standard histology and immunohistochemistry, and to what extent can molecular analysis help?

Six pathologists from Japan and the United Kingdom evaluated four different cases of adrenocortical disorders independently. These adrenocortical disorders included an adrenal tumor in a 45-yr-old female without any endocrine abnormalities, bilateral adrenocortical lesions in a 55-yr-old female with Cushing syndrome, an adrenocortical mass in a 44-yr-old man with hypertension, and an adrenocortical lesion in a 62-yr-old female with chronic hypertension for 30 yr. In this article, we provide a clinical summary, macroscopic and histologic findings, and histologic diagnosis of these four adrenocortical cases.

There appears to be a relationship between mitotic activity and malignant behavior in adrenocortical tumors, and carcinomas with a high mitotic index may have a poorer prognosis. This has been investigated further by quantifying and comparing the Ki-67 index using antibody MIB-1 in a series of 14 adrenocortical adenomas and 40 carcinomas. The levels have been correlated with survival and disease-free survival in carcinomas and with evidence of abnormal p53 expression as detected by immunohistochemistry. Nevertheless, many carcinomas have a low level of proliferation that may reflect varying abnormalities within the regulation of both cell division and apoptosis. Expression of bcl-2 protein, an inhibitor of apoptosis has therefore also been examined. The Ki-67 index in carcinomas was significantly higher than in adenomas, but below 4% there was overlap. There was no significant difference in survival between carcinomas with MIB-1 index <3% and those greater, but the lower group had significantly longer disease-free survival (p = 0.02). There was no significant difference between p53 immunopositive and p53 immunonegative carcinomas. No tumor showed immunopositivity for bcl-2 protein. It is concluded that MIB-1 index may contribute additional prognostic information in adrenocortical tumors. Inhibition of apoptosis by bcl-2 does not appear to play a role in tumor growth.

The hormonal immunoreactivity and vascularity of pituitary adenomas containing folliculostellate (FS) cells have been compared with those of tumors in which such cells were not identified. FS cells were present in variable numbers in 36 of 92 tumors. Adenomas immunoreactive for growth hormone (GH), adrenocorticotropic hormone (ACTH), or prolactin (PRL) contained FS cells in 40-50% of cases. Those immunoreactive for glycoprotein hormones and alphasubunit contained FS cells in 67-85% of cases, a statistically significant correlation. When alpha-subunit was also present in GH-, GH/PRL-, and ACTH-immunoreactive tumors, a higher proportion contained FS cells (57-91%). These data suggest a correlation between the presence of FS cells and glycoprotein immunoreactivity in pituitary adenomas. Vascular channels identified by the binding of the lectinUlex europaeus were quantified in the two types of tumors. Those containing FS cells were not more vascular than those without FS cells, which suggests that FS cells do not play a significant role in the regulation of intratumoraf vascularization in human pituitary adenomas.

The immunohistochemical distribution of the protease inhibitors alpha 1 -antichymotrypsin (alpha 1 -ACT) and alpha 1 -antitrypsin (alpha 1 -AT) has been documented in the normal human pituitary gland and in a series of pituitary tumors. In normal gland, alpha 1 -ACT was localized mainly in the dendritic folliculostellate cells, identified by immunopositivity for S 100 protein. A minority of endocrine cells also stained in 3 of 10 autopsy glands. Folliculostellate cells were identified in 11 of 28 tumors, and again, the distribution of alpha 1 -ACT positivity corresponded to these cells. In 4 cases, there was staining of a small minority of tumor cells. Alpha 1 -AT was localized to colloid in the microfollicles of the anterior lobe. In I normal gland, there was granular staining of endocrine cells. Alpha 1 -AT was present in 5 tumors, in microfollicles and in scattered endocrine cells in 2 adenomas. These data would support a physiological role for alpha 1 -ACT and alpha 1 -AT in the pituitary gland. Their differing distribution might reflect different functions.