Pan-cancer prevalence, molecular landscape, and prognostic impact of CCNE1 amplification across 76,046 malignancies.

3125 Background: CCNE1 amplification ( CCNE1 amp) is an oncogenic driver that confers synthetic-lethal vulnerability to agents targeting cell cycle and replication stress. While CCNE1 amp is a poor prognostic marker in some solid tumor types, pan-tumor data are limited. We used the MSK Clinical Sequencing Cohort to characterize its pan-tumor prevalence, molecular landscape, and prognostic impact. Methods: We analyzed data from consented patients with solid tumors sequenced by MSK-IMPACT between July 2014 and August 2025. CCNE1 amp was defined as an amplification or gain per clinical report. Tumor-level prevalence was calculated and reported for histologies with ≥10 CCNE1amp tumors. Co-occurring alterations were evaluated using the first CCNE1 amp sample per tumor type in each patient, with Fisher’s exact test and Benjamini-Hochberg correction. Overall survival was measured from the date of first sequencing to death or last contact. Results: Among 76,046 tumors, the prevalence of CCNE1amp was 2.7% (n = 2,036). Our CCNE1 amp cohort comprised predominantly ovarian (21%, n=435), uterine (15%, n=301), non-small cell lung (11%, n=231), esophagogastric (10%, n=206), and breast cancers (8%, n=168). Prevalence rates were highest in uterine carcinosarcoma (25.6%); ovarian carcinosarcoma (20%); uterine serous carcinoma (16.2%); high-grade serous ovarian carcinoma (15.4%); osteosarcoma (11.9%); gallbladder adenocarcinoma (9.7%); and esophageal adenocarcinoma (9.1%). Prevalence rates in breast cancer were 2.1% and in non-small cell lung cancer 2.6%. Among 67 patients with paired pre- and post-treatment samples of purity ≥30%, CCNE1 amp was reported in all samples in 42% of patients. The most frequently co-mutated genes were TP53 (87%), PIK3CA (12%), KRAS (8%), and RB1 (7%). Most common amplifications included the 19q neighbors CEBPA (31%), KMT2B (22%), and AKT2 (22%); as well as MYC (14%); and ERBB2 (13%). The most significant co-occurrences (q<0.05) were TP53 mutations (OR 11.5) and MYC amplifications (OR 3.5); CCNE1amp was mutually exclusive with PTEN (OR 0.37) and KRAS mutations (OR 0.55). CCNE1amp samples had a higher median fraction of genome altered (median 0.35 vs. 0.12; p<0.001). CCNE1 amp was associated with worsened overall survival in a multivariate Cox model stratified by cancer type and adjusted for age, stage and TP53 status (adjusted HR 1.25; 95% CI 1.17-1.33). Its adverse prognostic impact was more pronounced in TP53 wt (HR 1.62; 95% CI 1.38-1.91) vs. TP53 mut tumors (HR 1.19; 95% CI 1.11-1.27; p for interaction <0.001). Conclusions: CCNE1 amplification defines a pan-cancer genomic subset associated with poor prognosis, genomic instability, and few co-occurring actionable alterations, supporting its relevance as a therapeutic target. Discordance between paired samples suggests that a single negative biopsy may not rule out CCNE1amp.