Genomic findings in non-cryptogenic cerebral palsy: a systematic review and meta-analysis
Paloma Arana-Rivera1,2,3, Myriam Martín-Bermejo2, Diana Marcela Nova-Díaz3,4
1Pediatric Neurology Unit, Department of Pediatrics, Hospital Universitario de Navarra, Pamplona, Spain.
Background:
The contribution of genomic variants to non-cryptogenic cerebral palsy (CP), defined by identifiable perinatal or acquired risk factors, remains incompletely characterized. This study aimed to estimate the frequency of reported pathogenic or likely pathogenic (P/LP) genomic findings in non-cryptogenic CP and compare it with cryptogenic cohorts.
Methods:
We conducted a systematic review and meta-analysis searching PubMed and Scopus to May 30, 2026, for sequencing-based CP studies with extractable non-cryptogenic data. Eligible approaches included whole-exome sequencing, whole-genome sequencing, and targeted next-generation sequencing panels. Risk of bias was assessed using an adapted JBI prevalence checklist. Pooled frequencies were calculated using random-effects models with logit transformation.
Results:
Eleven studies were included in the qualitative synthesis, and eight contributed to the meta-analysis. The primary non-cryptogenic analysis included 1,885 individuals, of whom 325 had reported P/LP genomic findings. The pooled frequency was 12.6% (95% CI 8.9-17.6; I 2 = 80.6%), approximately one in eight tested individuals. In seven studies with cryptogenic subgroup data, the pooled frequency was 32.3% (95% CI 21.0-46.1; I 2 = 69.9%). Cryptogenic cases were more than twice as likely to have a reported P/LP genomic finding as non-cryptogenic cases (risk ratio 2.21, 95% CI 1.56-3.14). Across all analyzed cohorts, the overall pooled frequency was 19% (95% CI 12-28). Prematurity was generally associated with lower frequencies, whereas selected hemorrhagic or cerebrovascular phenotypes showed enrichment for COL4A1/COL4A2-related findings.
Interpretation:
Reported P/LP genomic findings occur in a clinically meaningful subset of individuals with non-cryptogenic CP, although less frequently than in cryptogenic CP. Interpretation is limited by heterogeneous definitions of non-cryptogenic CP and incomplete genotype-phenotype adjudication across studies. These findings support careful assessment of perinatal risk factors, neuroimaging patterns, and genotype-phenotype concordance when interpreting genomic results.
Systematic Review Registration:
CRD420251169588, https://www.crd.york.ac.uk/PROSPERO/view/CRD420251169588.
