RESUMO
BACKGROUND: Digital PCR (dPCR) can quantify cell-free viral DNA (DNAemia), a biomarker of active viral infection. To accelerate epidemiologic investigation into low-level viral reactivation in chronic disease, we have evaluated the performance of dPCR to detect cytomegalovirus (CMV) and Epstein-Barr virus (EBV) DNAemia across platforms and blood matrices. METHODS: The droplet-based (BioRad) dPCR platform performance was compared to chip-based (BioMark), and assay validation followed dMIQE guidelines. CMV and EBV DNA reference materials were spiked into known negative plasma and serum samples. In addition, two independent cohorts of ovarian cancer patients were evaluated for viral DNAemia (n = 65 serum and 79 plasma samples). RESULTS: The limit of quantification (LOQ) was at or slightly above 100 copies/mL for both instruments: 105-135 copies/mL for droplet-based detection and 100 copies/mL for chip-based detection. DNAemia in serum had a slightly lower LOQ (105-110 copies/mL) compared to plasma (LOQ; 115-135 copies/mL). The variation (CV) coefficients for each assay and machine were less than 5 %. In patient samples, CVs ranged from 4.5 - 7.4 % and were similar for cell-free DNA derived from serum or plasma. There was good correlation between DNAemia measurements in patient samples across dPCR platforms (r > 0.90 for each assay and matrix). CONCLUSION: dPCR can quantify low-level herpes virus DNAemia with CVs below 8 %. Our results indicate that using serum-derived cell-free DNA and droplet-based dPCR is optimal for quantitating low-level viral DNAemia; however, plasma and chip-based approaches are acceptable alternatives and suitable for epidemiologic investigation.