TRK-inhibitors have demonstrated efficacy across several cancers with NTRK fusions. Their activity in cancers with NTRK overexpression remains unclear.This trial enrolled patients with advanced cancers harboring NTRK fusions or extreme mRNA overexpression, defined as NTRK1/2/3 expression by RNA profiling >5 SDs for a given cancer type. The primary endpoint was objective response rate (ORR), with secondary endpoints including time-to-progression (TTP) ratio [TTP on study to TTP on previous systemic therapy (TTP1)], progression-free survival (PFS), and overall survival (OS). Initially planned for 2 non-comparator groups: primary central nervous system (CNS) and non-CNS tumours with NTRK fusions, the protocol was amended to permit NTRK overexpression.Seventeen patients were treated with larotrectinib: one glioblastoma with a SPECC1L::NTRK2 fusion (group 1), and a peripheral nerve sheath tumor with a TPM3::NTRK1 fusion and 15 patients with overexpression (group 2). The ORR was 6%. An additional 3 of 12 (25%) TTP1-evaluable patients achieved a TTP ratio ≥1.3 and 2 of 5 without an evaluable TTP1 had a PFS >6 months. Median PFS and OS were 3.5 (95% CI, 1.4-6.0) and 15.9 months (95% CI, 6.4-NR), respectively.Unlike its efficacy in NTRK-fusion positive cancers, larotrectinib did not demonstrate a signal of efficacy among tumors with NTRK overexpression.