Role of gene expression profiling in distinguishing biologically and clinically distinct subclasses of endometrial carcinoma.

摘要

This presentation will highlight a series on ongoing studies designed to exploit the power of comprehensive, genome-wide, gene expression profiling in distinguishing biologically and clinically relevant subclasses of endometrial carcinoma. Two distinct studies, one directed toward a biological problem and one toward a clinical problem will be presented here. The type I/type II paradigm for endometrial carcinoma is now widely accepted as representing a set clinicopathologic criteria that define two distinct types of endometrial cancer with different outcomes. Although a small number of specific genetic alterations have been attributed to each (e.g., MLH1 methylation and PTEN mutation in type I tumors and TP53 mutation in type II tumors), this candidate gene approach has only begun to define the complex genetic basis for these two classes of endometrial cancers. In the first study, we hypothesized that a supervised class comparison between tamoxifen-associated endometrial cancers and matched cases not associated with tamoxifen exposure may provide insights into the molecular mechanism through which estrogen contributes to endometrial tumorigenesis. No statistically significant difference between the two groups was evident (P = 0.48). However, unsupervised hierarchical clustering of the entire sample of tumors revealed two groups with extremely diverse molecular profiles (P < 10(8)) that were independent of tamoxifen exposure. Of recognized clinicopathologic factors, histologic grade correlated best with these two molecular classes. The first finding suggests that tamoxifen may function to stimulate the progression of clinically occult endometrial carcinomas, such that the spectrum of tumors that is observed in both groups is essentially the same. The second finding provides strong support for the definition of two biologically distinct classes of endometrial carcinoma, and that these two molecular classes may underlie the frequently cited but ill-defined type I and type II distinction. In the second study,we tested the hypothesis that women with intermediate risk endometrial carcinomas could be stratified into subgroups at high-risk (and thus appropriately treated with adjuvant therapy) or low-risk (and thus not likely to benefit from adjuvant therapy) for recurrence. Despite there being no single gene significantly associated with time to recurrence after correcting for multiple comparisons, there was a significant difference in time to recurrence between high-risk and low-risk patients, as defined using a statistical risk score (the basis of which will be presented). The estimated hazard ratio was 3.07 (95% CI, 1.00-9.43, P = 0.04). We conclude that women with intermediate risk endometrial cancers identified as high-risk for recurrence according to a gene expression-based risk score have a significantly increased risk of recurrence compared to those classified as low-risk. This suggests that gene expression profiling may contribute to the clinical management of intermediate risk endometrial cancer patients.