A Pan-Gene Analysis of Hereditary Cancer Syndromes Between Admixed Cohorts and Global Databases
Fernando Augusto Siuta dos Santos, Pedro Dytz Marin, Giulia Scain, Isabela Alicia Fink, Julia Loureiro Chaves Cé, Stéphanie Klein Brum, Mariana Weremeier Weber, Martina de Sá, Rafaela Sumienski Girardi, Solana de MeloPURPOSE
Global genomic databases underrepresent admixed populations, creating a systemic diagnostic bias. This study integrates the Brazilian ABraOM repository with ClinVar to quantify the interpretability gap across a five-gene panel (BRCA1, BRCA2, TP53, MLH1, MSH2), establishing the risk of inconclusive results in the Global South.
METHODS
A multi-database cross-referencing protocol was executed. Germline variant data for the BRCA1, BRCA2, TP53, MLH1, and MSH2 genes were extracted from the Brazilian ABraOM cohort, totaling 1,204 distinct variants. These variants were mapped against NCBI ClinVar pathogenic records for the same genes. We calculated the Ancestry-Driven Odds Ratio (OR) to determine the probability of a variant lacking a definitive clinical classification compared to European baselines. Statistical significance for Allele Frequency (AF) divergence was established using chi-square tests against global reference baselines.
RESULTS
The pan-gene integration exposed a profound and reproducible diagnostic bottleneck for the Brazilian population. Across the 1,204 variants identified, the aggregate failure rate in clinical classification reached 80.5%, representing variants that are either classified as Variants of Uncertain Significance or are entirely absent from global pathogenic registries. The highest gap was observed in TP53, where 83.5% (n = 158) of the 189 native variants identified lacked a pathogenic correlate in ClinVar. For the BRCA1/2 complex, 80.3% (n = 447) of the 557 identified variants remained clinically unclassified. Statistical modeling established a 3.24-fold Diagnostic Penalty (OR, 3.24 [95% CI, 2.18 to 4.81];
CONCLUSION
This analysis provides irrefutable quantitative evidence that current precision medicine guidelines are systematically biased. The 80% interpretability gap across critical hereditary cancer genes proves that Eurocentric reliance nullifies the clinical utility of multi-gene panels in Brazil. Integrating native repositories like ABraOM into interpretation algorithms is mandatory to ensure equitable screening and prevent misclassified clinical interventions.