Objectives

NTRK gene fusions are rare mutations found across cancer types, detectable through various molecular tests. Identifying patients with these mutations for targeted treatments presents a costly challenge for the healthcare system, with population rates unknown. We performed a systematic literature review of NTRK fusion prevalence across cancer types to understand the variation in rates, to inform effective population diagnostic screening and scale of potential therapeutic uptake.

Methods

We searched Medline, Embase and Cochrane databases on April 1, 2021. Studies reporting NTRK fusion rates in solid tumour cohorts were included. Further criteria were English language, publication post-2010 and minimum sample size. Cohorts selected based on fusion status and abstracts were excluded. Study bias and quality were assessed using two adapted checklists for evaluating prevalence studies.

Results

A total of 159 studies were included to inform NTRK fusion prevalence, with 17 pan-cancer estimates and 429 estimates for specific cancer types (including 36 paediatric cohorts). Adult pan-cancer estimates ranged from 0.03% to 0.70%, with higher rates reported based on RNA assays. Across common cancers, rates were consistently <1% for studies with sufficient sample size. Higher rates were seen in common cancers for cohorts with driver mutation exclusion, colorectal cancers with microsatellite instability, and rare morphological subtypes. Overall, NTRK fusions were identified across 79 different cancers. Zero NTRK fusions were found for 187 (43.5%) cancer type estimates, but only a minority of these had sufficient sample size and high sensitivity testing. No studies were free from external validity bias; reporting of cohort demographics was lacking in 39% of studies.

Conclusions

Current evidence affirms NTRK fusion positive cancers are rare and widely distributed. Rarity increases the complexity of implementing efficient population screening, while small-scale, heterogeneous data confound prediction of health system impact. Further large-scale, standardised genomic data are needed to characterise NTRK fusion epidemiology.