Dr Stephanie Dominguez of Paris's Hopital Pitie-Salpetriere presented further analyses of the GIGHAART study (ANRS 097) to the Eighth Annual Conference of the British HIV Association last Friday. GIGHAART is a study of salvage therapy with six to eight drugs with or without a treatment interruption prior to starting the new regimen. Click here for previous data on viral load changes in this study.
GIGHAART was first presented at the European Conference on Clinical Aspects and Treatment of HIV last October in Athens. Last Friday’s presentation provided further data, and raised further questions about the strategy of interrupting treatment before switching to a salvage regimen.
Baseline characteristics
Participants in the GIGHAART study had taken an average of eleven drugs prior to joining the study, and had received an average of 6.6 years of therapy when the study began in 1999. New regimens were selected on the basis of treatment history, with 68% of participants receiving six drugs and 31% receiving eight drugs.
At baseline, 79% of patients had three or more thymidine analogue mutations (AZT and d4T-associated mutations) and 10% had nucleoside analogue multidrug resistance mutations such as Q151M or 69SS. Sixty per cent had two or more non-nucleoside reverse transcriptase inhibitor mutations, and 42% had more than three major protease mutations. Eighty two per cent had two or more protease mutations.
Participants commenced a regimen which contained three or four nucleoside analogues, hydroxyurea, one NNRTI and three protease inhibitors (including ritonavir at a virologically suppressive dose of 400mg twice daily). The most commonly prescribed agents were ddI, 3TC and amprenavir, with the option to add lopinavir emerging during 2000 when the drug became available on expanded access in France.
Participants were randomised to start the new regimen immediately or take an eight week break from all antiretroviral therapy, to test whether resistance might be reduced in the absence of drug pressure. Genotypic analysis showed that only 7% of patients lost mutations associated with all three classes of antiretrovirals. Twenty four per cent lost mutations associated with two classes and 15% lost mutations associated with one class. However, 55% of patients did not lose any resistance mutations during the treatment interruption.
Phenotypic resistance analyses are not yet available to show the extent to which people recovered drug sensitivity despite minimal disappearance of mutations, but analysis of drug concentrations showed that where patients had drug concentrations within the normal range and no reversion of mutations, they did better than patients with low drug concentrations and some reversion of mutations. This finding led Stephanie Dominguez to suggest that the genotypic analysis may not have captured drug-sensitive sub-populations of virus that began to emerge during the treatment interruption. Further cloning of isolates is being conducted to pursue this theory.
Although few people discontinued their regimens, hydroxyurea was poorly tolerated, leading 25%-30% in each arm to discontinue the drug. One case of lactic acidosis was reported in each arm of the study, a relatively high incidence within the short follow-up period. Usage of hydroxyurea in these patients was not reported on.
Viral load responses were significantly better in the treatment interruption group; 42% of the treatment interruption group had viral load below 400 copies/ml after 12 weeks of treatment, compared with 16% of the immediate treatment group (p=0.027) by on treatment analysis.
Dr Dominguez reported that the difference between the two arms appeared to diminish somewhat by week 24 (further statistical analysis will be available by the time of the World AIDS Conference in July), leading Dr Adrian Palfreeman of Peterborough Hospital to ask whether it was appropriate to stop recruitment to the study at such an early stage. Dr Anton Pozniak of Chelsea and Westminster Hospital pointed out that a retrospective review of the Frankfurt cohort found that any apparent advantage to a treatment interruption had disappeared after 48 weeks of follow-up. Ultimately, he said, “this strategy needs to be judged by clinical outcome, not virological outcome in patients with very advanced HIV disease.”