Search

Purpose: The majority of cases of Cushing's disease are due to the presence of a corticotroph microadenoma. Less frequently no adenoma is found and histology shows either corticotroph hyperplasia, or apparently normal pituitary. In this study we have used molecular pathology to determine whether the tissue labeled histologically as "normal" is indeed abnormal., Experimental Design: Tissue from 31 corticotroph adenomas and 16 nonadenomatous pituitaries were subject to methylation-sensitive PCR to determine the methylation status of the p16 gene CpG island. The proportion of methylated versus unmethylated CpG island was determined using combined bisulphite restriction analysis. Methylation status was correlated with immunohistochemical detection of p16., Results: Seventeen of 31 adenomas (54.8%), 4 of 6 cases of corticotroph hyperplasia, and 7 of 10 apparently normal pituitaries showed p16 methylation. Ten of 14 (71%; P = 0.01) adenomas and 2 of 3 cases of corticotroph hyperplasia, which were methylated, failed to express p16 protein. However, only 2 of 7 apparently normal pituitaries that were methylated failed to express p16 protein. Quantitative analysis of methylation using combined bisulphite restriction analysis showed only unmethylated CpG islands in postmortem normal pituitaries; however, in adenomas 80-90% of the cells within a specimen were methylated. The reverse was true for corticotroph hyperplasia and apparently normal pituitaries where only 10-20% of the cells were methylated. Thus, the decreased proportion of cells that were methylated, particularly in those cases of apparently normal pituitary, is the most likely explanation for the lack of association between this change and loss of cognate protein in these cases., Conclusions: To our knowledge this is the first report that describes an intrinsic molecular change, namely methylation of the p16 gene CpG island, common to all three histological patterns associated with Cushing's disease. Thus, the use of molecular pathology reveals abnormalities undetected by routine pathological investigation. In cases of "apparently" normal pituitaries it is not possible to determine whether the change is associated with adenoma cells "scattered" throughout the gland, albeit few in number, or with the ancestor-clonal origin of these tumor cells.

We report a 63-year-old woman who presented with clinical and biochemical features of ACTH dependent Cushing's syndrome secondary to a primary ovarian carcinoma. The tumour produced very high levels of ACTH precursors, consistent with defective POMC processing.

Pro-opiomelanocortin (POMC) mRNA was demonstrated in pituitary adenomas from 16 patients with Cushing's disease and 10 with Nelson's syndrome. The intensity of signal was significantly greater in Nelson's syndrome than in Cushing's disease and there was a trend towards a greater proportion of positive cells. This probably reflects inhibition of POMC gene expression by the high circulating levels of cortisol in Cushing's disease. In the para-adenomatous gland, the intensity of signal was variable in cells showing Crooke's hyaline change, ranging from negative to strongly positive, in keeping with the functional heterogeneity of corticotrophs. In one case, junctional corticotrophs were present and these were more intensely stained than anterior lobe corticotrophs in the same gland. This supports the concept that these cells are subject to different regulatory influences from corticotrophs in the anterior lobe. Whether this is related to differences in embryological origins or to local factors is at present unclear.

Cushing's disease may originate from either the anterior pituitary lobe or the neurointermediate lobe, a major characteristic of the latter group being bromocriptine responsiveness. This study of two patients with Cushing's disease demonstrates that bromocriptine responsiveness also may be associated with anterior pituitary corticotroph hyperplasia or a normal pituitary gland. The two patients were a 14-yr-old boy (patient 1) and a 29-yr-old woman (patient 2); their cortisol production rates were 121 and 234 mumol/24 h (normal values, less than 80 mumol/24 h), respectively. A single oral dose of 2.5 mg bromocriptine resulted in a gradual decrease in plasma cortisol from 680 to 130 nmol/L after 6 h in patient 1 and from 640 to 170 nmol/L after 4 h in patient 2. Both patients then received medical treatment for a period of 2 yr. Whereas sodium valproate was ineffective, bromocriptine (5 mg/day) abruptly decreased the cortisol production rate to 60 mumol/24 h in patient 1 and to 138 mumol/24 h in patient 2, and both patients had a partial clinical remission. Despite an increase in bromocriptine dosage to 30 mg daily and 24 mg/day cyproheptadine, the clinical and biochemical remission was not sustained in patient 1, and no further improvement occurred in patient 2. Total hypophysectomy then was performed in both patients. Sections of the pituitary from patient 1 showed diffuse anterior pituitary corticotroph hyperplasia, with early nodule formation in some areas. The sections from patient 2 showed normal numbers and distribution of corticotrophs. We conclude that the heterogeneous nature of Cushing's disease cannot be explained on the basis simply of anterior vs. intermediate lobe origin of the disease.

The application of modern investigative techniques, particularly electron microscopy and immunohistochemistry, to the pituitary gland in Cushing's disease have confirmed that in the majority of cases (up to 90% in some series) the disease is due to a corticotroph microadenoma. It has also been shown that the tumours may produce not only ACTH, but also other peptides derived from the same precursor molecule, pro-opiomelanocortin and, in a small minority of cases, other pituitary hormones (e.g. prolactin). Since these peptides are known to have physiological actions they may account for some of the varied symptoms and signs of Cushing's disease. Because of the high incidence of single tumours the treatment of choice in many centres has become selective adenomectomy by the transsphenoidal route. However, a minority of cases appear to be the result of primary hypothalamic or central abnormalities and this may account for the identification of a normal pituitary gland or of corticotroph hyperplasia (with or without tumour formation). It is not possible at the present to identify these groups of patients on the basis of biochemical testing. It is hoped that detailed prospective studies correlating hormone secretion, responses to biochemical testing and detailed investigation of pathological tissue will provide further insight into the pathogenesis of particular variants of the disease.

The overall anatomical distribution of specifically stained corticotropic cells has been studied in pituitaries obtained at autopsy from 9 patients with Cushing's disease. Three patterns have been demonstrated consistent with the theory that it is not a single entity, but that some cases are of primary pituitary aetiology, whereas others are the result of hypothalamic or central nervous system dysfunction. The junctional corticotropic cells appear to react differently from those in the anterior pituitary to some stimuli, supporting the hypothesis that these cells may represent the remnants of the pars intermedia in the human adult, and suggesting that they may have some as yet unidentified physiological function.

There is little evidence for a separate functional or anatomical intermediate lobe in the adult human pituitary gland. Nevertheless, Lamberts et al. (1982) proposed that a subgroup of corticotroph adenomas in Cushing's disease arise in that lobe and can be identified by the presence of argyrophil (? neural) fibres, and that these tumours are more often associated with corticotroph hyperplasia and hyperprolactinaemia than those arising in the anterior lobe. We have examined a series of corticotroph adenomas from patients with Cushing's disease for evidence of argyrophil fibres, and have correlated this with tumour site, corticotroph distribution in the para-adenomatous gland, serum PRL levels and PRL immunoreactive cells in the tumour. Argyrophil fibres were identified not only in tumours adjacent to the posterior lobe, but also in tumours situated deep in the anterior lobe. There was no correlation between the presence of fibres or the site of the tumour and corticotroph hyperplasia. Whilst the two patients with the highest serum PRL levels did have argyrophil fibres they also had a subpopulation of PRL immunoreactive cells in the tumour. On the basis of these results, we propose that the 'intermediate lobe' hypothesis as outlined above should not be accepted.