Instability of 8E5 calibration standard revealed by digital PCR risks inaccurate quantification of HIV DNA in clinical samples by qPCR.

Establishing a cure for HIV is hindered by the persistence of latently infected cells which constitute the viral reservoir. Real-time qPCR, used for quantification of this reservoir by measuring HIV DNA, requires external calibration; a common choice of calibrator is the 8E5 cell line, which is assumed to be stable and to contain one HIV provirus per cell. In contrast, digital PCR requires no external calibration and potentially provides 'absolute' quantification. We compared the performance of qPCR and dPCR in quantifying HIV DNA in 18 patient samples. HIV DNA was detected in 18 by qPCR and in 15 by dPCR, the difference being due to the smaller sample volume analysed by dPCR. There was good quantitative correlation (R = 0.86) between the techniques but on average dPCR values were only 60% of qPCR values. Surprisingly, investigation revealed that this discrepancy was due to loss of HIV DNA from the 8E5 cell calibrant. 8E5 extracts from two other sources were also shown to have significantly less than one HIV DNA copy per cell and progressive loss of HIV from 8E5 cells during culture was demonstrated. We therefore suggest that the copy number of HIV in 8E5 extracts be established by dPCR prior to use as calibrator.