142P - Exploring tumor mutational burden and frameshift mutations as predictors of immune checkpoint inhibitor efficacy

Background

The efficacy of immune checkpoint inhibitors (ICIs), such as pembrolizumab and nivolumab, has been established in patients with a high tumor mutational burden (TMB-High). However, the efficacy of ICIs varies widely across different cancer histologies, and additional biomarkers are required. Notably, high TMB cut-offs and the presence or absence of frameshift mutations (fs) have emerged as potential predictors of ICI therapy outcomes. This study investigated the correlation between TMB, fs mutations, and the efficacy of ICI monotherapy.

Methods

Data from 01/06/2019 to 22/04/2024 were extracted from the national caner genomic database (C-CAT database) conducted across Japan. Among the 424 patients with a TMB of 10 or higher who underwent ICI treatment, 368 received ICI monotherapy. We excluded cases with liquid biopsy and patients lacking efficacy data. Time to treatment failure (TTF) was defined as the interval from the start of ICI monotherapy to treatment discontinuation for any reason, including disease progression, treatment toxicity, and death.

Results

A total of 283 patients were identified, with a median age of 61.1 years; 68.1% were male, and 31.8% were female. The most prevalent cancers were lung (23%), urothelial (20%), and esophageal/gastric cancers (18%). The overall response rate (ORR) was 32.2% and the clinical benefit rate was 67.8%. Analysis of the predictive factors revealed significantly better response rates at TMB-20 over cutoff compared to TMB10-20 (54.1% vs. 29.6% for TMB20over/Mb, p = 0.001). However, the ORR and TTF of ICI monotherapy for fs alone were not significantly different (ORR, p = 0.528; TTF, p = 0.648). Examination of the four TMB and fs categories indicated rates of 46.1% for TMB20over/Mb with fs+, 38.1% for TMB20over/Mb with fs-, 20.7% for TMB10-20/Mb with fs+, and 27.9% for TMB10-20/Mb with fs- (p =0.03). High TMB cases (>20) with fs mutations suggest a potential predictive effect for fs.

Conclusions

TMB alone may be insufficient to predict ICI efficacy. However, the combination of TMB-high and frameshift may offer a better prediction of response to ICI monotherapy.

Clinical trial identification

Editorial acknowledgement

Legal entity responsible for the study

T. Shimoi.

Funding

Has not received any funding.

Disclosure

K. Sudo: Financial Interests, Personal, Invited Speaker: AstraZeneca, Pfizer, Eisai, Nihon Medi-Physics, Janssen. T. Koyama: Financial Interests, Personal, Invited Speaker: Sysmex, Chugai, AstraZeneca; Financial Interests, Institutional, Research Grant: PACT Pharma, Boehringer Ingelheim; Financial Interests, Institutional, Local PI: Novartis, Chugai, Lilly, Daiichi Sankyo, Takeda, AstraZeneca, Zymeworks. K. Yonemori: Financial Interests, Personal, Advisory Board: Eisai, AstraZeneca, Sanofi, Genmab, Gliad, OncoXerna, Takeda, Novartis, MSD; Financial Interests, Personal, Invited Speaker: Pfizer, Eisai, AstraZeneca, Eli Lilly, Takeda, Chugai, Fuji Film Pharma, PDR Pharma, MSD, Ono, BMS, Boehringer Ingelheim, Daiichi Sankyo, Bayer, Jansen, Sanofi; Financial Interests, Institutional, Local PI: MSD, Daiichi Sankyo, AstraZeneca, Taiho, Pfizer, Novartis, Takeda, Chugai, Ono, Sanofi, Seagen, Eisai, Eli Lilly, Genmab, Boehringer Ingelheim, Kyowa Hakko Kirrin, Nihon Kayaku, Haihe. All other authors have declared no conflicts of interest.