665 
Evaluation of an Automated Sample Preparation System and Real-time Assay for Quantitation of Genetically Diverse HIV-1 Specimens from Brazil
John Hackett, Jr*1, P Swanson1, V Holzmayer1, J Yamaguchi1, S Huang2, P Bodelle1, C Brennan1, R Badaro3,4, C Brites3, and S Devare1
1Abbott Labs, Abbott Park, IL, US; 2Abbott Molecular, Des Plaines, IL, US; 3Fndn Bahiana de Infectologia, Univ Federal da Bahia, Brazil; and 4Univ of California, San Diego, US
Background: The
combination of automated sample preparation and real-time RT-PCR for
measurement of HIV-1 viral load has the potential to significantly enhance
throughput, reduce operator-associated error, and improve assay sensitivity and
dynamic range. Due to the high rate of evolution and ever-changing distribution
of HIV-1, reliable patient monitoring requires that viral load assays detect
and accurately quantify genetically diverse strains of HIV-1. The aim of this
study was to evaluate performance of the Abbott RealTimeTM HIV-1 assay with
the m2000sp automated sample
preparation instrument (Abbott Molecular, Des Plaines,
IL; not available in the U.S.) on
a panel of genetically diverse specimens.
Methods: The
specimen panel consisted of 91 HIV-1 seropositive
plasmas collected from blood donors in Brazil. Subtype was determined by
sequence/phylogenetic analysis of three independent
genomic regions: gag p24, pol integrase, and env gp41. Specimens for which subtype assignments differed
between the regions analyzed were categorized as mosaics. Viral loads were
measured in the Abbott RealTime HIV-1 assay (m2000sp
sample preparation and m2000rt amplification
and detection instruments) and LCx®
HIV RNA Quantitative assay (LCx HIV; Abbott
Laboratories, Abbott Park, IL; not available in the U.S.).
Results:
Group/subtype was determined for 89 of 91 specimens. The panel included 69
subtype B, 1 C, 2 F, and 17 mosaic strains. Eighty-nine specimens were
quantified by the RealTime
HIV-1 assay, 87 by the LCx HIV assay. The observed
correlation for 86 specimens within the dynamic range of both assays was 0.908
with a slope of 0.843 and intercept of 0.823. Two specimens below the LLD of
the LCx HIV assay (50 copies/ml)
were quantified in the RealTime HIV-1 assay. Notably, two specimens below the LLD of both
assays were also PCR-negative for all three regions used for subtype analysis.
Good agreement was observed between assays with 91% of values within 0.5 log10
copies/ml.
Conclusions: In
the present study, the automated RealTime HIV-1 assay and sample
preparation system accurately quantified this genetically diverse panel. Viral
load determinations were highly correlated with the LCx
HIV assay. The automated RealTime HIV-1
assay offers the advantages of increased throughput and reduced labor while
providing reliable quantification of diverse HIV strains.
|