Comparison of tumor-agnostic and tumor-specific clinical oncology trial designs: a systematic review and meta-analysis.

To examine whether tumor-specific and tumor-agnostic oncology trials produce comparable estimates of objective response rate (ORR) in -altered cancers. Electronic database searches were performed to identify phase I-III clinical trials testing tyrosine kinase inhibitors from 2000 to 2021. A random-effects model was used to pool ORRs. A total of 22 cohorts from five tumor-agnostic trials and 41 cohorts from 27 tumor-specific trials had published ORRs. There was no significant difference between pooled ORRs from either trial design for multitumor analyses (37 vs 50%; p = 0.05); thyroid cancer (57 vs 33%; p = 0.10); non-small-cell lung cancer (39 vs 53%; p = 0.18); or melanoma (55 vs 51%; p = 0.58). For -altered advanced cancers, tumor-agnostic trials do not yield substantially different results from tumor-specific trials.