Stratification of intermediate-risk endometrial cancer patients into groups at high risk or low risk for recurrence based on tumor gene expression profiles

Purpose: Endometrial cancers classified as "intermediate risk" based on clinical and/or pathologic features are associated with a 15% to 20% risk of recurrence. Here, we test whether global gene expression profiling can distinguish intermediate-risk tumors into high-risk and low-risk subgroups. Experimental Design: Tumor specimens were obtained from 75 intermediate-risk endometrial cancer patients, 13 who had recurred and 62 who had not recurred with a median follow-up of 24 months. Gene expression profiles were obtained using the Affymetrix U133A GeneChip oligonucleotide microarray. The genes most associated with risk of recurrence were used to create a risk score using a leave-one-out cross-validation method and the univariate Cox proportional hazards regression model. Time to recurrence curves for the high-risk and low-risk subgroups were estimated using the Kaplan-Meier method, and the difference in time to recurrence between these two subgroups was tested using the log-rank test. Results: There was a significant difference in time to recurrence between high-risk and low-risk patients using risk scores as defined above (P = 0.04). The estimated hazard ratio (95% confidence interval) was 3.07 (1.00-9.43). Conclusions: Patients with intermediate-risk endometrial cancers identified as high-risk for recurrence according to a gene expression-based risk score have a significantly increased risk for recurrence compared with those classified as low risk. These findings suggest that gene expression profiling can potentially contribute to the clinical classification and management of intermediate-risk endometrial cancers.