Molecular-Based Classification for Endometrial Cancer Could Help Predict Survival

Researchers were able to molecularly classify endometrial cancers with distinct survival differences using a new classification tool called ProMisE (Proactive Molecular Risk Classifier for Endometrial Cancer) that uses clinically applicable methods.  

Details of this tool were presented as a poster by Aline Talhouk, PhD, of the University of British Columbia in Vancouver, Canada, and colleagues at the 2016 American Society of Clinical Oncology (ASCO) Annual Meeting, held earlier this month in Chicago (abstract 5518).

“Success of ProMisE in diagnostic samples would enable informed surgical and adjuvant treatment decision-making from the earliest time point,” wrote Talhouk and colleagues in their poster presentation. “Immediate applications include consistent categorization, stratification of trials, identifying women who may have Lynch syndrome, and targeted therapy.”

To test this new tool, Talhouk and colleagues evaluated 460 women with endometrial cancer hysterectomy specimens. Samples underwent immunohistochemistry looking for the presence or absence of mismatch repair (MMR) proteins, sequencing for POLE exonuclease domain mutations (POLE EDM), and for p53 abnormalities.

Among the samples, 22.8% of patients were MMR deficient (MMR-D), 9.1% were POLE EDM, 43.9% were p53 wild-type, and 24.1% had p53 abnormalities.

Fifty-seven of these women had matched pre-operative biopsy and post-operative hysterectomy samples. The researchers analysis showed that the molecular classification used in the model identified four prognostic groups with distinct differences in overall survival and progression-free survival.

Those patients classified as the POLE group consistently had the best prognosis. In contrast, those patients with the p53 abnormality had­­­­ the worst prognosis. Using this new model, the researchers were able to distinguish between two intermediate outcome survival curves; MMR abnormalities subgroup had worse outcomes than patients with wild-type p53 disease.

Talhouk and colleagues also compared the prognostic ability of the ProMisE tool to a current clinicopathological-based risk stratification system from the European Society for Medical Oncology (ESMO). In contrast to the ProMisE  tool, which showed distinct survival curves for each group, in the ESMO system low-, intermediate-, and high-risk groups had near overlap in low- and intermediate-risk group outcomes for overall survival, disease-specific survival, and progression-free survival.

“ESMO risk groups are cross tabulated with molecular subgroups revealing substantial overlap between the p53 abnormality group and high risk but otherwise no predictable associations,” the researchers wrote. “Possibility of both overtreatment or undertreatment exist in these subgroups with diverse clinical outcomes.”

The researchers have currently completed the discovery and confirmation phase testing of the new model and will next work on final validation in clinical trials.