Abstract LB032: High tumor mutational burden subtends better response in locally advanced rectal adenocarcinoma undergoing neoadjuvant chemoradiotherapy and is associated with mutations in DNA damage repair genes

Tumor mutational burden (TMB), defined as the number of somatic mutations per Megabase (Mb) in a neoplastic specimen measured by next-generation sequencing (NGS), is shown to increase its rate in association with higher genomic instability. Neoadjuvant chemoradiotherapy (NACRT) in locally advanced rectal adenocarcinoma (LARC) may be more active in genomically unstable tumors. However, response to NACRT is highly variable and pathological complete response is observed in only 15-25% of cases. Currently, the association of TMB with response to NACRT in LARC is unknown. The aim of this study was to investigate the role of TMB as a potential predictive biomarker of response in NACRT-treated LARC. We collected a set of pre-treatment biopsies from LARC cases, clinical stage II or III, undergoing NACRT with capecitabine combined with radiotherapy (N = 22). After DNA extraction and quality and concentration evaluation, we performed mutational profiling and assessed TMB through multigene sequencing of 409 cancer-related genes by NGS (OncomineTM Tumor Mutation Load Assay, Thermo Fisher Scientific). The most common mutations were, as expected, in KRAS, TP53 and PIK3CA. We compared our results with those of The Cancer Genome Atlas (TCGA) database and proportion was similar, with no statistically significant differences. We found three cases with TMB over 10 mutations/Mb, a threshold above which neoantigen formation occurs with higher likelihood and response to NACRT may be more efficient due to the stimulation of immune system through radiotherapy. Two cases harbored deleterious mutations in genes involved in DNA damage repair (PMS2 and NBN). Furthermore, cases with mutations in the transcription factor FBXW7 (N = 4/22) and in the receptor tyrosine kinase ERBB3 (N = 3/22) had significantly higher TMB than non-mutated ones (p-value = 0.0090 and 0.0107 respectively, false discovery rate < 0.1 for both genes). We compared TMB in major/complete responders vs. minor/non-responders as evaluated by post-surgical pathological examination (responders if Dworak tumor regression grading score, TRG, was equal to 3-4 and non-responders if TRG was equal to 0-1) and investigated the possible association of higher TMB with better response to NACRT. In our cohort we found a trend for higher TMB in responder vs. non-responder cases (5.08 mutations/Mb, IQR 3.40-9.29 vs. 4.25 mutations/Mb, IQR 2.54-5.5) (p-value = 0.05, one-sided). In conclusion, our study shows that higher TMB is correl