Clinical relevance of transcriptome profiling for assessing prognosis, platinum insensitivity, and precision...

In endometrial cancer (EC), pretreatment identification of insensitivity to platinum-based chemotherapy (PbCT) may improve survival. We hypothesized that transcriptome profiling would identify seminal genes associated with chemoresistance.

We identified 209 EC cases with robust mRNA data from The Cancer Genome Atlas (TCGA) database. Cases with stage III/IV EC or stage I/II plus myometrial invasion ≥ 66% were categorized as clinicopathologic high risk (CPHR); the remainder were clinicopathologic low risk (CPLR). We assessed expression of genes (excluding POLE variants) associated with chemoresistance. Expression data were normalized, log₂ transformed, and Pearson correlated with 5-year progression-free survival (PFS). PFS rates were compared using log-rank tests; continuous variables were compared using t-tests.

Composite expression scores for chemoresistance-associated genes correlated with 5-year PFS rates: low scores, 84.4%; intermediate, 74.7%; and high, 23.0%. High expression of transcription factor CCNA2/E2F1 significantly correlated with increased expression of DNA damage repair, cell cycle, and antiapoptosis genes (majority, R > 0.80). Independent of most traditional risk factors, low vs high CCNA2 + E2F1 expression stratified patients by 5-year PFS (89% vs 50%), endometrioid and serous histologies, TCGA subgroups, and clinicopathologic risk (all P < .001). Low vs high CCNA2 + E2F1 expression also significantly affected 5-year PFS for CPLR (89% vs 59%) and CPHR (87% vs 39%) cases.

CPHR or CPLR-plus-recurrence cases with low CCNA2 + E2F1 expression likely have PbCT sensitivity; analogous cases with high CCNA2 + E2F1 expression likely have PbCT insensitivity. Pretreatment transcriptome profiling potentially can determine PbCT insensitivity and identify multiple alternative therapeutic targets, thereby sparing patients platinum-associated toxicities and disease progression.