Five-year outcomes of next-generation sequencing implementation at a Brazilian public health system reference centre for rare diseases

IntroductionRare diseases affect approximately 6%–7% of the Brazilian population, representing a significant public health challenge due to diagnostic delays and inequitable access to genomic services. This study evaluates a 5-year implementation of next-generation sequencing (NGS) at a Reference Service for Rare Diseases (RDRS) within the Brazilian Unified Health System (SUS), examining diagnostic performance and barriers to equitable access to genomic medicine.MethodsA cohort of 385 patients with suspected genetic disorders underwent clinical or whole-exome sequencing (CES/WES) between 2019 and 2024. Exome analyses were guided by standardized Human Phenotype Ontology (HPO)–based phenotypic characterization, resulting in an overall diagnostic yield of 38.7% (149/385 cases).ResultsA total of 249 variants were identified across 165 genes, including 75 variants not previously reported in medical literature or public databases. Actionable secondary findings were detected in 3.2% of cases, involving pathogenic variants in genes related to cardiac disease, cancer predisposition, and anesthesia risk, underscoring the preventive potential of genomic testing.DiscussionThe implementation of NGS in the Brazilian public health system is feasible and highly impactful, yielding a 38.7% diagnostic rate that significantly reduces the diagnostic odyssey for rare disease patients. Success was heavily dependent on a multidisciplinary approach, standardized HPO-based phenotypic characterization, and close clinical-laboratory integration. The identification of 75 novel variants highlights the genetic complexity of Brazil’s admixed population and the challenges of interpreting variants of uncertain significance (VUS) in underrepresented groups. Furthermore, the detection of actionable secondary findings demonstrates the broader preventive potential of genomic testing, underscoring the need for expanded local databases, genomic literacy among clinicians, and robust ethical frameworks for patient management.