The retinoblastoma-related gene Rb2/p130 encodes a protein that is a negative cell-cycle regulator normally expressed in a number of adult tissues. This protein shares many structural and functional features with the product of the retinoblastoma gene, one of the best-studied tumor-suppressor genes, and plays a fundamental role in growth control. The Rb2/p130 gene product associates with specific members of the E2F family and various cyclins, displaying a growth-suppressive activity specific for the G(0)/G(1) phases. It has been reported that Rb2/p130 is involved in the pathogenesis and progression of lung cancer and mesothelioma. We previously demonstrated for the first time that reduced immunohistochemical expression of Rb2/p130 was a strong independent predictor of poor outcome in endometrial cancer. The aim of the present study was to evaluate Rb2/p130 expression in normal, hyperplastic, and neoplastic endometrial lesions to determine whether the protein plays a significant role in endometrial carcinogenesis. We evaluated Rb2/p130 expression by immunohistochemistry staining in 102 specimens chosen to represent a spectrum of endometrial changes, including proliferative endometrium (n = 18), secretory endometrium (n = 18), simple or complex hyperplasia without atypia (n = 18), atypical hyperplasia (n = 18), and invasive carcinoma (n = 30). We found that Rb2/p130 was highly expressed in proliferative endometrium and in hyperplasia without atypia, the mean percentage of stained nuclei being 66% and 60%, respectively, but was downregulated in secretory endometrium, atypical hyperplasia, and carcinoma, with mean scores of 38%, 25%, and 22%, respectively. When categorized on a semiquantitative scale (negative v 1% to 50% v >50% positivity), endometrial cancer displayed significantly less staining than all other endometrial samples (P <.001). Poorly differentiated carcinomas (n = 9) showed a significantly lower immunoreactivity for Rb2/p130 than did well-differentiated carcinomas (n = 11; P =.005) and moderately differentiated carcinomas (n = 10; P =.03). In addition, atypical hyperplasia showed a significantly lower immunoreactivity than either proliferative endometrium (P =.003) or hyperplasia without atypia (P = 0.02). Our findings of a progressive decrease in Rb2/p130 expression from hyperplastic endometrium through atypical hyperplasia to poorly differentiated carcinomas suggest the involvement of this negative cell-cycle regulator in endometrial carcinogenesis. Furthermore, immunostaining for Rb2/p130 may prove diagnostically useful in the often difficult distinction between hyperplastic and atypical hyperplastic endometrium. HUM PATHOL 32:360-367.
Copyright 2001 by W.B. Saunders Company