The impact of genomic context on outcomes of solid cancer patients treated with genotype-matched targeted therapies: a comprehensive review

Abstract

Background: A critical need in the field of genotype-matched targeted therapy in cancer is to identify patients unlikely to respond to precision medicines. This will manage expectations of individualised therapies and avoid clinical progression to a point where institution of alternative treatments might not be possible. We examined the evidence base of the impact of genomic context on which targeted alterations are inscribed to identify baseline biomarkers distinguishing those obtaining the expected response from those with less benefit from targeted therapies.

Methods: A comprehensive narrative review was conducted: scoping searches were undertaken in PubMed, Cochrane Database of Systematic Reviews, and PROSPERO. Outcomes included in meta-analysis were progression-free and overall survival. Data was extracted from Kaplan-Meier and used to calculate hazard ratios. Studies presenting data on two molecular sub-cohorts (e.g., co-mutation vs no co-mutation) were included in fixed meta-analysis. Other studies were used for descriptive purposes.

Results: The presence of concomitant driver mutations, higher tumour mutational burden(TMB), greater copy number burden and APOBEC signatures significantly reduces benefits of targeted therapy in lung cancers in never smokers(LCINS) and breast cancer; cancers with low TMB. LCINS have significantly poorer outcomes if their cancers harbour p53 co-mutations, an effect also seen in HER2+ breast cancer patients (trastuzumab) and head and neck cancer (PI3K inhibition). PI3K co-alterations have less impact when targeting EGFR-mutations and ALK-fusions, but significantly reduce the impact of targeting HER2- and MET-amplifications. SMARCA4 co-mutations predict for poor outcome in patients treated with osimertinib and sotorasib. In BRAF-mutant melanoma, whilst there are no genomic features distinguishing exceptional responders from primary progressors, there are clear transcriptomic features dichotomising these outcomes.

Conclusion: To our knowledge this is the most comprehensive review to date of the impact of genomic context on outcomes with targeted therapy. It represents a valuable resource informing progress towards contextualised precision medicine.