782P - Correlation between chemotherapy response score (CRS) and germline BRCA1/2 (gBRCA) status in women diagnosed with FIGO stage IIIC/IV high-grade serous ovarian cancer (HGSOC)

Background

CRS after neoadjuvant chemotherapy (NACT) and gBRCA status predict survival outcomes of women diagnosed with advanced HGSOC. We investigated the correlation between CRS and gBRCA to understand the potential of CRS as an inexpensive biomarker for PARPi therapy.

Methods

Analysis of the OC gBRCA testing database from a tertiary referral centre. Eligible criteria included women with FIGO stage IIIC/IV HGSOC who received NACT and had gBRCA tested between 01/09/2017 and 31/12/2022. Univariable and Multivariable (MV) Cox Proportional-Hazards Models were applied to assess the association between PFS, OS and clinical factors. Two-Proportions Z Test was applied to assess the proportions of CRS1+2 or CRS3 between gBRCA.

Results

536 women were eligible. Mean age 66 years (range 36–89). 139 had FIGO stage IVB HGSOC. 71% received 3-weekly carboplatin with paclitaxel. Median PFS was 14.4 months (95%CI 13.6–15.4) and OS 38.0 months (95%CI 34.8–42.6) in the overall population. Performance status (0–1 vs. 2–4), cytoreductive surgery (Yes vs. No) and gBRCA status (mutant[MT] vs. wild type[WT]) were predictive (PFS) and prognostic (OS) in the overall population (all P<0.05). 363 women underwent delayed primary surgery (DPS) and had a CRS. Of these, 90% received 3–4 cycles of NACT and 96% had optimal (≤1 cm) cytoreduction. In the NACT-DPS group, MV analysis showed that residual disease (≤1 vs. >1 cm), CRS (3 vs. 1+2) and gBRCA status (MT vs. WT) were predictive (PFS) (all P≤0.02). Proportion testing showed women with gBRCA2 were more likely to have CRS3 vs. CRS1+2 than those with gBRCA1 (P=0.008). Women with gBRCA1 (P=0.075) and gBRCA2 (P=0.078) were not more likely to have CRS3 vs. CRS1+2 then those with WT.

Conclusions

We found a correlation between CRS3 and gBRCA2, but not gBRCA1, in women with FIGO stage IIIC/IV HGSOC treated with NACT-DPS. CRS should be evaluated in a prospective study as a predictor of PARPi efficacy.

Table: 782P

Clinical trial identification

Editorial acknowledgement

Legal entity responsible for the study

R.D. Morgan.

Funding

Has not received any funding.

Disclosure

All authors have declared no conflicts of interest.