Methods comparison for molecular diagnosis of human herpesvirus 8 infections

Human herpesvirus 8 (HHV-8) virological diagnosis and monitoring relies mainly on real-time PCR.

To evaluate two real-time PCR commercial kit (HHV-8 Premix R-gene^TM and Clonit® HHV-8) and compare with in-house real-time PCR.

Twelve samples (3 undetectable and 9 detectable with viral load ranging from 10¹ to 10⁵ per reaction) were tested for HHV-8 detection and quantification with the 3 methods. Methods comparison was supported with regression curve and diagram presenting difference or ratio between commercial and in-house PCR results and plotted against the in-house PCR results. Statistical analyses, specifically Student tests and Spearman correlation, were performed.

In both cases, qualitative results obtained with commercial kit and in-house PCR were identical and HHV-8 quantitation results were significantly correlated (Clonit®, R_s = 1, p < 0.001 and R-gene^TM R_s = 0.98, p < 0.001). However, Clonit® results were significantly higher compared to the in-house results with an overestimation in median [IQR] of 1.16 log₁₀ copies/10⁶ cells [1.12-1.18] whereas R-GeneTM results were not significantly higher, and an overestimation in median of 0.46 log₁₀ copies/10⁶ cells [0.37-0.52]. Otherwise, repeatability and reproducibility tests of undetectable sample failed with Clonit® technique contrary to the R-Gene^TM.

HHV-8 R-gene^TM assay seems to be the most suitable since it showed consistent qualitative results with in-house HHV-8 PCR, a good quantitative correlation, an overestimation not significantly different and inferior to 0.50 log₁₀ copies/10⁶ cells and a good repeatability.