Polymerase chain reaction (PCR) as a diagnostic tool in HIV infection

We set up a PCR laboratory for the diagnosis of HIV-1. Probably due to the variability of the HIV-genome, classical primers that performed well in some laboratories in the past, did not suffice for detection of HIV-1 strains in Belgian hospitals. Two new primer sets amplifying a fragment in the LTR-gag gene and in the env gene, which perform better on strains seen in Belgium, have been developed and evaluated. One primer set, conceived and evaluated on Belgian strains by the "Instituut voor Tropische Geneeskunde" in Antwerp, was also included. These three primer sets performed superior (92% sensitivity and 100% specificity on 24 samples) than the classical primers (83.5% sensitivity and 56% specificity on 21 samples). Together with a well-studied testing algorithm, they allow the reliable identification of the presence of the HIV-1 genome. To detect resistance of HIV-1 to reverse transcriptase (RT) inhibitors, we developed a set of two overlapping nested PCR primer sets and additional sequencing primers to amplify and sequence the total RNA or DNA RT gene using a direct cycle sequencing approach of the amplified fragment. Some clinical isolates were amplified and sequenced. In HIV-1 isolates from TIBO R82913-treated patients we identified two amino acid mutations (V108I and Y188L) involved in resistance (more than 100-fold reduced sensitivity). In an untreated patient we identified an amino acid variant (I/V 179D) involved in a 7-fold reduced sensitivity to TIBO. Several other amino acid variants, not involved in resistance, were detected in treated and untreated patients. Using this sequencing technique on cultured virus isolates we also observed in one TIBO-treated patient a differential selection among the strains of the original HIV-1 pool. From this patient we isolated and sequenced a completely TIBO sensitive HIV-1 strain after extensive cultivation in cord blood lymphocytes of the original TIBO resistant HIV-1 virus pool. We could however identify the resistant genotype after cultivation of this resistant HIV-1 virus pool on CEM cells. Our study revealed that sequencing investigations on emerging resistance should preferentially be done with uncultured patient samples since viral sequences and virus-drug sensitivities obtained from isolates cultured in vitro may not necessarily correspond to the sequences and sensitivities of the dominant strain in vivo.