Abstract

Comparative effectiveness of dried plasma HIV-1 viral load testing in Brazil using ViveST? for sample collection

R.S. Diaz¹, M. Zanoni¹, C. Loveday², M. Holodniy³, R.M. Lloyd Jr.<sup>4

1</sup>Federal University of Sao Paulo, Retrovirology Laboratory, Sao Paulo, Brazil, ²ICVC Charitable Trust HQ, Clinical Director, Buckinghamshire, United Kingdom, ³VA Palo Alto Healthcare System, AIDS Research Center, Palo Alto, United States, ⁴Research Think Tank, Inc., CEO, Buford, United States

Background: ViveST? (ST) is an inexpensive transportation system for HIV-1 virologic assays of dried plasma. Herein we describe the precision, and reproducibility of using ViveST? as a transportation method for shipping specimens and subsequent HIV-1 viral load (VL) testing.
Methods: Ten-fold dilutions from clinical plasma samples with HIV VL > 5 log₁₀/mL were created to generate samples with values of 4 log₁₀, 3 log₁₀ and 2 log₁₀ copies/mL. Thirty singlicate samples at each dilution from ST were compared to frozen (F) (180 samples). Ten samples in triplicate were compared F to ST (60 samples). Finally, 299 samples with HIV VL < 50 copies/mL (99); 1.7 log₁₀ to 3.99 log₁₀ (100); and 4 log₁₀ to 5.99 log₁₀ (100) were applied to ST, dried via driDOC for 12 hours, stored at room temperature for up to 4 days, and reconstituted using 1.175mL of buffer and compared to F using the Siemens VERSANT® HIV-1 RNA 3.0 Assay (bDNA). Significance was analyzed using Student t-test and Pearson correlation.
Results: F vs. ST dried plasma intra-assay mean variance among 3 dried replicates was < 0.15 log₁₀ copies/mL (P = NS). Compared to F plasma, there was a mean reduction in VL of 0.3, 0.27, and 0.35 log₁₀ copies/mL at the 4, 3, 2 log₁₀ copy/mL samples respectively (all comparisons, P < 0.01). 12/99 undetectable F VL were positive with ST, whereas 5/100 F detectable VL were undetectable with ST (mean VL 2.1, 2.3 log₁₀/mL respectively). Overall correlation between F and ST was r =0.97.
Conclusions: Viral load results using SampleTanker were highly reproducible. Quantitation of HIV-1 viral load assays using dried plasma yielded minimally lower values compared to frozen plasma, and were within accepted assay variation for replicate samples. Our comparative data suggests that SampleTanker has promise for use in HIV clinical practice.