|
M. De Wit*, M. Peeters, P. McKenna, and K. Hertogs.
Virco, Mechelen, Belgium.
Background:This program aims to initiate monitoring of laboratory performance in genotypic resistance analysis of recombinant HIV clinical isolates.
Methods:Clinical isolates were selected based upon specific resistance profiles (PRi, NRTi, NNRTi's, MDR). The entire PR region (codons 1—99) and the RT gene up to codon 250 were sequenced in 5-fold by 4 independent laboratories. A consensus sequence for each sample was determined on the basis of the most prevalent nucleotide at each position of the sequences. This reference sequence is used as the comparative standard for analysis of proficiency testing results from participating laboratories. For each participant and for each sample (n = 7), the % discrepancies was calculated. These were scored using weighted factors. Different penalties were given, depending on whether missed calls represented a mixture or a single nucleotide difference compared to reference values. A similar analysis was done for the derived amino acid sequences. Different techniques were used by the different participating labs.
Results:Overall lab nucleotide discordances with the reference sequence ranged between 0.09% and 7.78%. The majority of missed calls were at positions where the lab sequences indicated a mixture but reference indicated a single nucleotide. For the amino acids, the percentage of discrepancy ranged between 0% and 1%. Discrepancies at the level of resistance-associated mutations occurred at positions for RT—41, 74, and 75—in 4 out of 32 sequences returned. In these 4 sequences, it concerned the presence of mixtures versus a WT call in the consensus sequence.
Conclusions:High overall consistency (kvalues ranging from 0.966 to 0.998) exists between the participating laboratories despite individual shortcomings. This initial exercise of testing complex clinical isolates should become routine practice.
© 8th Conference on Retroviruses and Opportunistic Infections