Compliance is significantly better in HAART responders compared to non-responders

A longitudinal prospective study was performed to verify the impact of compliance on response to HAART and the development of drug resistance during therapy. In 43 mainly drug-experienced patients on HAART, taking compliance (TC), dosing compliance (DC), and the number of drug holidays > 24 h (DH) were monitored for protease inhibitors using electronic event monitoring (EEM). Genotypic resistance was scored using LiPA HIV-1 RT (M41L, T69D, K70R, L74V, M184V, T215Y/F), LiPA HIV-1 protease (D30N, M46I, G48V, 150V, 154V/A, V82A/F/T/I, 184V, L90M), ARMS-151 (Q151M) and sequencing. Patients had been changed to HAART including a PI for 9.7 plus or minus 6.6 months when EEM was performed for 112 plus or minus 24 days. Overall median compliance rates were high: TC 98.1% (Inter Quartile Range, IQR: 5.3%), DC 91.8% (IQR: 18%), number of DH I (IQR: 5). Thirty-seven patients responded (R) to the therapy by a drop in VL below 50 c/ml over almost the entire period of HAART monitored. Six patients either never reached undetectable or had a rising VL above 500 c/ml after the initial response and were considered non-responders (NR). All compliance parameters were better in R as compared to NR, but only drug holidays were significantly different: 98.4% (IQR: 4.28%) TC in R compared to 93.1% (IQR: 16.76%) in NR (P = 0.16); DC was 92.2% (IQR: 15.1%) and 83.0% (IQR: 46.78%), respectively (P = 0.292); number of DH was 0 (IQR: 4) and 6.5 (IQR: 13.5), respectively (P = 0.029). The prevalence of genotypic resistance, as measured by LiPA and ARMS, was higher in both the reverse transcriptase and the protease gene in the non responding patients (100% and 50% respectively), compared to the responding patients (85% and 6.5%), only baseline protease mutations were significantly different between both groups. For the six poor responders, samples at baseline and at the end of the study were genotypically characterized using LiPA, ARMS and sequencing. Four of the six poor responding patients showed an increase in genotypic resistance towards the therapy that was monitored. Perfectly compliant patients can control virus replication for a prolonged period, even in the presence of baseline resistance mutations. The patients that had a poor response to HAART were also those with a significantly lower compliance. In our patient population, reduced compliance resulted in therapy failure, mostly associated with a build-up of resistance mutations.