Universal versus guideline-based germline multigene panel testing in solid tumors: Diagnostic yield, variant of uncertain significance burden, and clinical actionability - A systematic review with meta-analysis.

Germline pathogenic variants in cancer predisposition genes, sought by multigene panel in all adults with solid tumours or only in those meeting guideline criteria

This systematic review with meta-analysis compares the diagnostic yield, variant of uncertain significance burden and clinical actionability of universal versus guideline-based germline multigene panel testing in adults with solid tumours. PubMed, Embase, Web of Science and Scopus were searched from inception to 30 January 2026; 144 studies were included, of which 14 formed the comparative evidence set, with risk of bias assessed and proportions pooled using random-effects models. In unselected cohorts, the pooled diagnostic yield was 0.13 (95% CI 0.11-0.17; I² = 98%) and the pooled variant of uncertain significance burden was 0.40 (95% CI 0.33-0.48; I² = 97%). In a two-study synthesis, the sensitivity of National Comprehensive Cancer Network (NCCN) criteria was 0.72 (95% CI 0.67-0.76), meaning that approximately 28% of individuals with pathogenic or likely pathogenic variants would not meet the tested criteria. The authors conclude that universal germline multigene testing detects additional hereditary cancer predisposition beyond phenotype-based criteria but increases uncertain findings, and that standardised actionability definitions and prospective comparative studies are needed to guide implementation.

Synthesis written by Geno'X. For the full original abstract, please refer to the source publication.

The figure that carries weight is the NCCN criteria sensitivity of 0.72: about 28% of carriers would escape a test restricted to the criteria, a concrete argument for broadening testing that must be set against a variant of uncertain significance burden of 0.40, the counterpart of that broadening. But this sensitivity rests on only two studies, and heterogeneity of 97 to 98% (I²) on the pooled estimates means 0.13 and 0.40 cannot be read as values applicable to a given service. The abstract gives neither the number of patients nor a common definition of actionability, which the authors acknowledge by calling for prospective comparative studies: as it stands, this is a synthesis that documents the trade-off, not an argument for deciding between universal and guideline-based testing.

Analysis by Dr Thibaut Benquey