Minichromosome maintenance protein 2 expression in normal kidney and renal cell carcinomas: relationship to tumor dormancy and potential clinical utility.
PURPOSE: A major problem in the management of patients with renal cell carcinoma is predicting tumor behavior. In the search for more accurate markers of prognosis, tumor cell proliferation has been investigated. These studies have mostly used antibodies directed against Ki-67 or proliferating cell nuclear antigen and have given conflicting results, findings that are likely because of a combination of specificity and methodological differences. Minichromosome maintenance (Mcm) proteins are a series of closely related proteins that are components of the prereplicative complex. The Mcm proteins are essential for initiating eukaryotic DNA replication and serve as useful markers of proliferating cells. EXPERIMENTAL DESIGN: The aims of this study were to determine the frequency and pattern of Mcm2 expression by immunohistochemistry in normal kidney (n = 10) and renal tumors [n = 56; clear cell n = 36; chromophil (papillary) n = 7; oncocytoma n = 5; and transitional cell carcinoma n = 8], compare its sensitivity to the established proliferation marker Ki-67, examine for differences in tumors derived from stable and labile epithelial cell populations in the kidney, and assess the relationship of Mcm2 proliferation to clinicopathological characteristics of kidney tumors. In addition, to additionally investigate the issue of tumor dormancy we wished to assess the relationship between Mcm2 labeling index (LI) and the angiogenic factors angiopoietin-1 (Ang) and Ang-2. RESULTS: In normal tissues, Mcm2 nuclear labeling was identified in both glomeruli (LI median 0.35%; range 0-1.7) and renal tubules (LI median 0.3%; range 0.1-2.9%). In tumors Mcm2 labeling was predominantly at the periphery with LIs ranging from 0.2-91.5%, which was significantly greater than Ki-67 LI (0.2-40.5%; P < 0.001). Mcm2 LI was also significantly higher in tumors derived from a labile epithelium (transitional cell carcinomas) than a stable epithelium (renal cell carcinomas; P = 0.013). A significant association was also demonstrated between Mcm2 LI and tumor grade (P = 0.0006), and angiogenic phenotype (defined by Ang expression; P = 0.03) but not with patient age (P = 0.84), patient sex (P = 0.25), tumor size (P = 0.74), or stage (P = 0.33). Furthermore, although not significant, a survival analysis demonstrated that 100% of patients with a low Mcm2 LI survived compared with 84% of those with a high Mcm2 LI over the follow-up period (up to 53.2 months; P = 0.14). CONCLUSIONS: This is the first study examining Mcm2 protein in normal and tumor kidney samples, and the first to perform histological subgroup analysis. It shows that Mcm2 is a superior marker to Ki-67 in the assessment of cell cycle entry in histological archival material and that normal kidney has a subset of cells within the glomerular and tubular compartments that are in cycle. It demonstrates that the frequency of cells in cycle in tumors formed from stable or labile epithelial populations mirrors that in the nonneoplastic epithelium. This study additionally demonstrates that the number of cells in cycle in tumors is limited by the angiogenic phenotype and supports animal models that show angiogenesis determines the likelihood of tumor dormancy. Additional study to confirm the clinical utility of Mcm2 as a prognostic marker is now indicated.