787P - A NGS panel for molecular classification of endometrial carcinoma

Background

The conventional classification tools for endometrial carcinoma (EC) is tiresome, including DNA sequencing, IHC or PCR-CE. Large NGS panels are used to simplify test but costly. Here, we propose a concise NGS panel for economically classifying EC.

Methods

Retrospectively enrolled EC cases of hysterectomy with bilateral salpingo-oophorectomy from Fudan University Shanghai Cancer Center from 2020 to 2022 were assessed. A 46-gene NGS Panel was used to identify POLE exonuclease domain mutations, microsatellite instability-high (MSI-H), TP53 mutations, and other clinically valuable targets.

Results

Tumor tissue samples of 331 EC patients were evaluated. The median age of patients was 55 years (range, 49-62), and 284 (85.8%) cases were endometrioid endometrial carcinoma. Patients in stage IA and IB EC were 204 (61.6%) and 52 (15.7%), respectively. The median follow-up time was 18.5 months (n=303). According to WHO classification criteria, patients were stratified into POLE mut (n=47; 14.2%), MMRd (n=79; 23.9%), NSMP (n=148; 44.7%) and p53 abn (n=57; 17.2%), whose prognosis was consistent with previous studies, POLE mut with most favorable prognosis and p53 abn with the worst. NGS reached 91.8% (n=269) concordance with IHC when detecting MMRd. The 24 inconsistent samples were validated by PCR, of which 17 (70.8%) were consistent between PCR and NGS. The concordance of NGS and IHC in detecting p53 was 65.0% (n=201), but patients detected by NGS alone had significantly worse prognosis than those detected by IHC alone, implying higher accuracy of the panel. Given NGS panel performance, particularly in detecting MMRd and p53 abn, our panel provided a reliable EC classification approach. In addition, the most frequent genetic alterations detected in at least 20% of patients were PTEN (83.4%), PIK3CA (59.2%), ARID1A (58.6%), TP53 (29.3%), CTNNB1 (27.8%), and KRAS (23.6%), which might guide the following treatment.

Conclusions

Our study demonstrates that the concise NGS panel is an effective “one-stop” strategy to precisely classify EC with high clinical availability. Further follow-up data are needed to testify to the effect of the classification on the prognosis of different adjuvant therapies or rare pathological subtypes.

Clinical trial identification

Editorial acknowledgement

Legal entity responsible for the study

H. Wen.

Funding

Has not received any funding.

Disclosure

All authors have declared no conflicts of interest.