Development and validation of a model gene therapy biodistribution assay for AVGN7 using digital droplet polymerase chain reaction

Biodistribution assays are integral to gene therapy commercialization and have traditionally used real-time qPCR. Droplet digital PCR (ddPCR), however, has distinct advantages including higher sensitivity and absolute quantification but is underused because of lacking regulatory guidance and meaningful examples in the literature. We report a fit-for-purpose model process to validate a good laboratory practice (GLP)-compliant ddPCR assay for AVGN7, a Smad7 gene therapeutic for muscle wasting. Duplexed primer/probe sets for Smad7 and mouse TATA-box binding protein were optimized using gBlock DNA over a dynamic range of 10-80,000 copies/reaction in 250 ng mouse gDNA. Linearized plasmid and mouse gDNA were used for validation, which determined precision, accuracy, ruggedness/robustness, selectivity, recovery, specificity, dilution linearity, and stability. Inter-run precision and accuracy met previously established criteria with bias between -5% and 15%, coefficient of variation (CV) less than 19%, and total error within 8%-35%. The limit of detection was 2.5 copies/reaction, linearity was confirmed at 40-80,000 copies/reaction, specificity was demonstrated by single droplet populations and assay stability was demonstrated for benchtop, refrigerated storage, and repeated freeze-thaw cycles. The procedural road map provided exceeds recently established standards. It is also relevant to many IND-enabling processes, as validated ddPCR assays can be used in biodistribution studies and with vector titering and manufacturing quality control.