Switch of protease inhibitor-containing HAART in routine clinical practice: a four-year prospective observational study

Abstract

An evaluation was made of the frequency of outcomes, the features, and one-year outcomes of the substitution, carried out because of failure or toxicity, of protease inhibitor (PI)-containing highly active antiretroviral therapy (HAART). Nine hundred and seventy-two HIV-infected patients were prospectively followed up since 1996, with the condition that they had a minimum 80% adherence to prescribed regimens. Four hundred and fifty-two changes occurred in 397 of the 876 evaluable patients (45.3%). Virological and/or immunological failure was of concern in 245 cases (54.2%). Interest in saquinavir had the greatest incidence and earliest occurrence (although the subsequent switch had a significantly better outcome than that of patients failing with other PIs); nelfinavir benefited from a shorter time to change and a worse long-term outcome (probably attributable to its predominant use in indinavir- and ritonavir-experienced patients); while indinavir showed the lowest overall frequency of substitution. Intolerance occurred in the remaining 207 cases (45.8%); with saquinavir being better tolerated than other PIs. A favourable outcome was obtained more frequently when poor tolerability was of concern, compared with therapeutic failure (P<0.008), while no significant differences were found according to prior antiretroviral experience and the subsequently selected HAART regimen. The overall one-year outcome per single substituted compound proved significantly better for patients who stopped using saquinavir and ritonavir, by contrast with those who stopped using indinavir and nelfinavir (P<0.0008). A significantly shorter mean time to substitution was recognized for nelfinavir and saquinavir than with ritonavir and indinavir (P<0.0001). When analysing the subset of patients experiencing HAART failure, a highly significant reverse relationship was demonstrated between mean time to failure, and rate of subsequent response to a modified antiretroviral regimen (P<0.0001). When considering the different patterns of efficacy, durability, resistance induction, expected adherence, and safety of each antiretroviral drug, initial and subsequent therapeutic choices should be carefully balanced against expected benefits and risks.