Comprehensive DNA profiling strategy for patient-derived xenograft model authentication in the J-PDX library

Patient-derived xenograft models are crucial in cancer research, although authentication methods remain limited. This study developed a comprehensive DNA profiling strategy for patient-derived xenograft quality control in the J-PDX Library by comparing short tandem repeat and single nucleotide polymorphism analyses. Overall, 325 patient-derived xenograft models were analyzed using both short tandem repeat and single nucleotide polymorphism profiling. Concordance rates between patient-derived samples and patient-derived xenograft models were assessed. Single nucleotide polymorphism-based analysis cutoffs classified models as matched, undeterminable, or unmatched. Additionally, the impact of microsatellite instability and loss of heterozygosity on concordance were evaluated. Short tandem repeat analysis showed greater variability than single nucleotide polymorphism analysis. Among the 325 models, 296 (91.1%) were matched, 24 (7.4%) were undeterminable, and five (1.5%) were unmatched. Notably, undeterminable models showed significantly higher microsatellite instability, with single nucleotide polymorphism analysis providing reliable authentication, even when short tandem repeat analysis failed. This study established an efficient patient-derived xenograft model quality control workflow using short tandem repeat analysis for primary screening and confirmatory single nucleotide polymorphism analysis for challenging cases, particularly microsatellite instability tumors. This workflow enables robust quality control of the J-PDX Library, enhancing its reliability for drug development and preclinical research.