Accuracy of quantitative HIV‐1 RNA test methods at 1000 copies/mL and the potential impact of differences in assay calibration on therapy monitoring of patients

Nico Lelie,Harry Drimmelen
DOI: https://doi.org/10.1002/jmv.25877
IF: 20.693
2020-04-25
Journal of Medical Virology
Abstract:<p>The World Health Organization (WHO) recommends the clinical use of a HIV‐1 viral load (VL) threshold level of 1000 copies(cp)/mL in patients on antiretroviral therapy (ART) to distinguish between viral control (VL&lt;1000 cp/mL) and viral failure or poor adherence (VL &gt;1000 cp/mL). The accuracy of five quantitative HIV‐1 RNA assays at this level was compared by replicate testing (n=24) of 1000 cp/mL samples prepared from the VQC HIV‐1 subtype B standard, which is in use for validation of NAT methods since 1995. Until 2004 the VL assays reported geometric mean (95% confidence interval (CI)) values ranging between 449 (188‐1067) and 3162 (3057‐2367) cp/mL when using the Siemens bDNA 3.0 assay as reference method for an assigned value of 1000 (962‐1038) cp/mL. In 2018 the following values (95% CI) were found by 24 replicate tests in each of the VL assays on the 1000 cp/mL samples: Abbott RealTime 1084 (784‐1572), BioMerieux EasyQ 1110 (533‐2230), Roche CAP/CTM 1277 (892‐1828), Hologic Aptima 1616 (1324‐1973) and Cepheid GeneXpert 2502 (1713‐3655) cp/mL. Calibration studies involving three consecutive WHO replacement standards showed a significant drift in the amount of RNA copies per International Unit (IU) overtime. Heat‐inactivation of HIV‐1 standards was found to cause a de‐standardizing effect. Our study underlines the limitations in HIV‐1 RNA assay calibration based on frequently replaced WHO international standards. It is therefore proposed that clinicians interpret the recommended 1000 cp/mL alert level in therapy monitoring with an inaccuracy range of 500‐2000 cp/mL.</p><p>This article is protected by copyright. All rights reserved.</p>
virology
What problem does this paper attempt to address?