L1-CAM as a predictor of survival in endometrial cancer: An analysis of The Cancer Genome Atlas (TCGA).

Abstract

5599 Background: Despite a good survival rate for early-stage endometrial cancers (ECs), the prognosis for advanced-stage ECs remains poor, with no biomarkers and few therapeutic options currently in existence. L1-cell adhesion molecule (L1-CAM), a glycoprotein which functions in adhesion and migration of tumor cells, has been associated with a poor prognosis in Type I endometrial cancer (ASCO 2011 Abstract #5091). We evaluated the role of L1-CAM among both Type I and II ECs in TCGA. Methods: Partek Genomics Suit was used to define differentially expressed genes with p-values for clinical associations (ANOVA with linear contrast for discrete variables and linear regression for continuous variables). Differences in survival between “high” and “low” expression groups (defined by median expression) were compared using Cox regression analysis, with p-values calculated via log-rank test, using the ‘survival’ package in R. Results: Of 451 downloadable tumor samples, 335 tumors with both clinical and gene expression data were analyzed. Median age was 63 yrs. (range 31-90 yrs.). Stage I, II, III, and IV comprised 65%, 7%, 23%, and 5%, respectively. 82% were endometrioid; 16% (n = 52) serous. Grade 1, 2, and 3 comprised 24%, 27%, and 49%, respectively. Median follow-up was 19.5 months. High L1-CAM expression was found in older (p = 0.0005), suboptimally debulked (p=0.002), and African-American patients (p = 0.0003), and those with high grade (p = 0.008), serous histology (p <0.00001), higher stage (p = 0.0004), positive peritoneal cytology (p = 0.007), deep myometrial invasion (p = 0.02), and positive pelvic (p = 0.003) and para-aortic lymph nodes (p = 0.002). High L1-CAM expression was associated with poor survival with a median overall survival of 17.2 months compared to 21.3 months for low L1-expressing endometrial tumors (HR= 3.1, CI=1.3 -7.3, p = 0.007). Conclusions: L1-CAM expression is associated with poorer survival and high-risk clinicopathologic factors in endometrial cancer. Its prevalence in Type II ECs, such as high-grade, serous ECs, makes it a particularly attractive target for both novel biomarker discovery and therapeutic targeting.