ESTRO 2021 Abstract Book

S989

ESTRO 2021

PO-1193 The impact of driver mutations on the pathologic response to Chemoradiation (CRT) in LA- NSCLC S. Appel 1 , J. Bar 2 , Y.R. Lawrence 1 1 Sheba Medical Center, Radiation Oncology, Ramat Gan, Israel; 2 Sheba Medical Center, Medical Oncology, Ramat Gan, Israel Purpose or Objective Pathologic regression after CRT for LA-NSCLC to is important prognostic factor. We aimed to evaluate pathologic regression after chemoradiation in patients undergoing triple modality treatment, stratified by molecular biomarkers. We hypothesized that cases with driver mutations will have reduced pathologic regression. Materials and Methods We conducted a retrospective analysis of patients with LA-NSCLC treated at our department between 2012- 2021 by trimodality strategy: platinum-based chemoradiation (CRT) to 60 Gy followed by surgery. We collected data about driver mutations (defined here as EFGR, ALK rearrangement, RET, ROS1 and BRAF), FGFR, KRAS mutations and PDL1 IHS (negative vs. positive ≥1% and percent). Other variables collected were gender, smoking history, stage, GTV volume, histology, tumor location, and treatment variables: dose (Gy), and chemotherapy type. Endpoints analyzed were and pathologic regression grade, converted to a binary variable: Major Pathologic regression (MPR, < 10% residual tumor cells) vs. residual tumor (> 10% tumor cells remained). We also recorded the percent of residual tumor cells and the rate of complete responses (CR). Pathological assessment was performed blinded to biomarkers. The presence/ absence of a driver mutation was analyzed as a binary variable. Results A total of 45 patients with LA-NSCLC that met the eligibility criteria. Mean dose 60.7Gy (SD 4.6), mean GTV was 138ml (SD 121 ml). The drivers mutations identified were EGFR (13; 29% of 45 patients included), BRAF (1; 2.2%), RET (1; 2.2%), and ROS1 (1; 2.2%), and were analyzed as one group. ALK was not detected. Cases with driver mutations (EGFR/RET/ROS/BRAF) had a significantly reduced MPR: 19% (3/16) vs. for no driver mutation: 72% (21/29) (odds ratio (OR)-11, p=0.001) and had no case of complete response (CR) (0/13) compared with 20% (6/29) amongst those lacking a mutation (figure1). Percent of residual tumor cells was 44.3% (SD 7.9%) for driver mutations vs 16.4% (SD 4.2%) for those lacking a driver mutation (p=0.002). Both cases with a FGFR mutation demonstrated a complete response (CR). In cases with a K-Ras mutation, the MPR was 78% (7/9). MPR was 50% (6/12) for PDL1 negative vs. 46% (12/26) for PDL1 positive (OR-0.83, p=0.66). Cases with negative PDL1 did not have CR: (0/12) vs 23% (6/26) for positive PDL1 (p=0.3) (figure1). The percent of residual tumor cells did not correlate with PDL1 staining (p=0.3) (figure2). Histology and smoking correlated significantly with MPR but also with co-variate driver mutations (p<0.05). Other variables did not correlate with MPR (p>0.5). PDL1 was negative in 12 (34% of 35 with available data) and positive in 23/35 (66%). MPR was noted in 53% (24/45) and the mean residual tumor cells was 26% (SD 29.4%).