Diagnostic Yield of Genetic Testing in Cerebral Palsy: A Systematic Review and Meta-Analysis.

Pooled diagnostic yield of genetic testing in cerebral palsy, by aetiology (cryptogenic, noncryptogenic, mixed or unselected cohorts) and phenotypic features

Genetic aetiologies are increasingly recognised in cerebral palsy (CP), yet diagnostic yields across clinical subgroups and phenotypes remain inconsistent, limiting evidence-based prioritisation of genetic testing. This systematic review and meta-analysis searched PubMed, Embase and the Cochrane Library for studies published between January 2010 and August 25, 2025, selecting peer-reviewed observational studies reporting genetic testing results in cohorts of at least 10 patients with CP, with independent extraction by two reviewers following PRISMA 2020 and a random-effects model. A total of 31 studies met the inclusion criteria, of which 20 were included in the quantitative meta-analysis: the pooled diagnostic yield is 19% (95% CI 11 to 32%; I² = 95.1%) in unselected cohorts, 44% (37 to 50%; I² = 72.2%) in cryptogenic CP versus 13% (7 to 21%; I² = 85.1%) in noncryptogenic CP. Yields are higher with normal brain MRI (25% to 80%), facial dysmorphism (66% to 71%), congenital anomalies (85%) or atypical neurological features (50%), and combined next-generation sequencing and chromosomal microarray gives the highest pooled yield, 53% (95% CI 43 to 62%), in the cryptogenic subgroup.

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

The gap between 44% in cryptogenic cerebral palsy and 13% in noncryptogenic forms is the operational result: it says whom to offer genetic testing first, and the combination of sequencing and chromosomal microarray alone yields over 50% in the cryptogenic subgroup. These figures must nevertheless be read with the heterogeneity the authors themselves measure (I² of 72.2 to 95.1%): they pool studies published between 2010 and 2025, hence different testing methods, and the yields by phenotype (normal MRI, dysmorphism, congenital anomalies) are given as ranges rather than pooled values. The suggestion to integrate genomic evaluation into the standard work-up is plausible, but it rests on observational studies for which the abstract does not allow selection bias to be assessed.

Analysis by Dr Thibaut Benquey