Evaluating the Dual-Target Aptima HIV-1 Quant Dx Assay: Comparison between Viral Loads Measured with pol and LTR...

For effective management of HIV-1 patients, accurate measurement of HIV-1-RNA viral load (VL) is fundamental. The latest generation molecular assays for monitoring VL perform simultaneous detection of two regions of the viral genome, but without specifying the target used for VL quantitation. By using the "open" software (research use only [RUO]) of Aptima HIV-1 Quant Dx Assay (Aptima) which provides both results obtained with the pol and LTR targets, we were able to compare n = 500 plasma samples results from chronically HIV-1-infected patients under antiretroviral therapy (ART). Correlation and concordance were analyzed. By stratifying VL into two groups (<30 and ≥30 copies/mL HIV-1-RNA) according to pol-based results and matching them with their respective LTR values, concordance was substantial (κ = 0.635; 95%CI = 0.569 to 0.700) as expected. Considering the specimens (n = 224) with VL exactly quantified (i.e., ≥30 copies/mL) with both targets, an optimal correlation subsisted (r = 0.8882; P < 0.0001) and Bland-Altman plot showed no significant mean difference between them. However, by stratifying all these data in three ranges (30 to 200, 201 to 1,000, and >1,000 copies/mL) according to pol-based results, concordance analysis showed fair agreement (κ = 0.344; 95%CI = 0.257 to 0.432). Indeed, after excluding mutually concordant VL values in each range (n = 134), the remaining discordant samples (n = 90; 40.1%) showed significant (P < 0.05) difference between VL measured with the two targets. With the Aptima "open" software, samples with pol-based VL <1,000 copies (cp)/mL HIV-1-RNA, the corresponding LTR values were on average 0.5 log₁₀ cp/mL higher. Further studies on these discrepancies and the nature of viral RNA elements detected only with the LTR despite efficient ART are in progress. IMPORTANCE The last generation dual-target platforms for quantification of HIV-1 RNA return a single value of viral load (VL) derived from a combined reading of two HIV-1 genome targets. By using a modified version of Aptima software, providing both the VL results obtained from pol and LTR amplification separately, we observed discordant VL results in some samples from HIV-1-infected patients on antiretroviral therapy. In particular, some samples with pol-based quantified <1,000 copies/mL VL showed the LTR-based value on average 0.5 log₁₀ copies/mL higher, and other samples, always by treated patients, showed VL exclusively quantified with LTR target while the corresponding pol-based VL results were completely undetected. Standard software of double-target based diagnostic systems does not allow recognizing discrepant VL values in these particular, but not rare, clinical specimens. This issue could have implications for clinical management by leading physicians to consider changing antiretroviral regimen based on presumed failure of antiretroviral therapy.