Background: Allele-specific RT-PCR (ASP) based on subtype consensus sequences has been used to detect low-frequency mutants, including NNRTI-resistant variants in women who have received single-dose nevirapine (sdNVP) for the prevention of MTCT. This approach is limited by genetic variation in the region complementary to the primers, leading to variability in allele detection. The goals of this study were to test the importance of this effect and to improve assay performance.

Methods: Plasma samples from HIV-1 (subtype C) infected women were analyzed at 0, 2, and 6 weeks after sdNVP by ASP using primers that discriminate wildtype and mutant alleles at codon 103 (AAT or AAC). In cases of poor assay performance, primers and DNA quantification standards were modified to match patient-specific HIV consensus sequences. The samples were then reanalyzed using the modified primers and standards. Amplification efficiency is defined as the sum of mutant and wild type PCR products relative to known quantification standards.

Results: In samples from 16/61 women, mutant and/or wildtype 103 alleles could not be amplified efficiently (relative to consensus standards). Population based sequences from a subset of 4 patients with poor amplification revealed a variety of changes in the primer target sequence both near its 3’end and in its center. In all cases, use of modified primers and standards with changes corresponding to patient consensus sequences increased the efficiency of amplification from less than 5% to near 100%. The mismatches responsible for poor amplification efficiency were variable. A single mismatch near the 3’ end of the primer could severely reduce efficiency; some double mismatches in the middle of the primer were tolerated, but triple mismatches were not. Consensus sequences from the 45 patients with good amplification efficiency showed 0-2 mismatches in the middle of the primer target sequence.

Conclusions: Allele-specific PCR can detect NNRTI-resistant variants with high sensitivity, but natural variation in primer target sequence can adversely affect performance. Although mismatches near the 3’ end of the primer appear to have the most influence, other changes can be critical.  Optimization of primers and quantification standards based on patient-specific HIV consensus sequences can improve assay performance. A library of primers and standards can be developed to adjust for common variations in the target sequence.