[Factors predicting the failure of interruption of highly active antiretroviral treatment (HAART) in HIV-infected patients].

Introduction: Treatment interruptions may be an alternative to HAART in the management of chronically infected HIV-patients. We designed this study in an attempt to assess the predictability of this strategy.

Methods: We recruited HIV-infected patients whose treatment had been suspended. Interruption was due to the patient's own decision, or toxicity, or because the patient had started the treatment with more than 350 CD41 cells/microL (immunologic criteria).

Results: Forty-one consecutive patients were included, with a median follow-up of 13 months. Failure was associated with the reason for interruption (p 5 0.0063). Failure occurred in 14.3% of those who interrupted treatment due to immunological criteria and in 40% of those who interrupted treatment due to their own decision or toxicity. The reasons for interruption were: toxicity in 11 patients (26.8%), personal decision in 9 (21.9%) and immunological criteria in 21 (51.2%). In the univariate analysis, the nadir CD41 cell count < 350 cél./microL [OR 16 (p = 0.054)] was statistically significant in the patients who stopped treatment due to immunological criteria, while treatment with protease inhibitors [OR 14 (p = 0.032)] was statistically significant in the remaining patients. In the multivariable analysis only nadir CD41 < 350 cél./microL was independently related with failure.

Conclusions: Failure was related to interruption criteria and was greater in patients who stopped due their own decision or toxicity. When interruption was due to immunological criteria, the factor predicting failure was nadir CD41 cell count < 350 cél./microL. In the remaining patients, none of the variables was related to failure.