704P - Exploring the optimal tumour mutational burden (TMB) cutoff for predicting efficacy of immune checkpoint inhibitors (ICI) in Chinese patients with advanced non-small-cell lung cancer (NSCLC): A territory-wide study

Background

Previous literature suggested better efficacy treatment outcomes with ICIs for advanced NSCLC with high TMB (TMB-H, ≥10 mut/Mb). However, the generalizability of TMB-H as a prognostic biomarker in diverse patient populations remains uncertain. This prospective study explored the optimal TMB cutoff for predicting efficacy of ICIs in Chinese stage IV NSCLC patients (NCT05811169).

Methods

Patients with previously untreated stage IV NSCLC in Hong Kong were recruited since January 2021. Tissue-based next-generation sequencing was performed using FoundationOne®CDx prior to first-line treatment. Association between TMB and clinical characteristics was analysed using logistic regression. Patients were categorized into ICI group (treated with ICI ± chemotherapy) and non-ICI groups (treated without ICI or targeted therapy). Similar to the methodology applied in previous studies, TMB cutoff at 10 mut/Mb was compared with the 80^th percentile of TMB in the ICI group for the performance prediction of progression-free survival (PFS) benefits from ICI.

Results

A total of 205 patients with evaluable TMB were accrued. The median TMB was 4.8 mut/Mb (IQR 1.2-10.9 mut/Mb). In the ICI group (n=56), 55 (98.2%) used pembrolizumab ± chemotherapy and the 80^th percentile of TMB was 16.9 mut/Mb. Overall, univariable analysis showed that TMB-H (both ≥10 mut/Mb and ≥16.9 mut/Mb) was significantly associated with males, ever smokers and TP53 mutation (all p<0.05). In the ICI group, there was no significant difference in PFS between TMB-H and TMB-low (TMB-L) groups with 10 mut/Mb as the cutoff (mPFS: 20.6 vs 16.4 months, HR 0.70, 95% CI 0.29-1.66; p=0.41). However, when the cutoff of 16.9 mut/Mb was applied, patients with TMB-H had significantly better PFS than TMB-L (mPFS: 28.0 vs 15.1 months, HR 0.11, 95% CI 0.02-0.83; p=0.03). In the non-ICI group (n=79), no significant difference in PFS was found between TMB-H and TMB-L subgroups with either cutoff.

Conclusions

The 80^th percentile may be considered as better TMB cutoff in predicting PFS benefits from ICI in Chinese patients with stage IV NSCLC. The recruitment is still ongoing to validate our results.

Clinical trial identification

Editorial acknowledgement

Legal entity responsible for the study

The authors.

Funding

Innovation and Technology Commission (ITC)–Hong Kong under the Partnership Research Programme scheme (PRP/067/20FX).

Disclosure

All authors have declared no conflicts of interest.