643    

Background:  Enfuvirtide (ENF), the first compound of HIV fusion inhibitors, has been mainly used as a component of rescue treatment in multi-experienced patients. In phase 2 studies, ENF showed dose-related efficacy, but no data are currently available in the clinical setting. Therefore, the aim of our study was to evaluate pharmacokinetic and pharmacodynamic determinants of virologic response to ENF-based regimens.

Methods:  Multi-experienced patients starting an ENF-based regimen were prospectively enrolled. HIV RNA levels and CD4⁺ cell counts were recorded at baseline, week 4 and week 12. HIV-1 genotypic resistance and virtual phenotype were also obtained at baseline. Optimized background score was defined as the number of active drugs associated to ENF given by virtual phenotype. A complete ENF concentration-time curve (area-under-the-curve, AUC) was determined at week 2 and ENF C_trough were measured at week 4 and week 12, A validated HPLC method was used. Predictors of viral load decrease were determined by linear regression analysis. Predictors of viral suppression (HIV RNA < 50 copies/mL) were evaluated by logistic regression analysis (backward method)

Results:  We included 38 patients. At baseline, median (IQR) viral load and CD4 cell count were 5.16 log (4.7 to 5.5), and 49 (19 to 109) cells/µL. Median (IQR) viral load decrease was –0.59 (–2.1 to –0.21), and –0.41 (–1.69 to –0.15) at week 4 and week 12, respectively. Median (IQR) CD4⁺ cell count increase was 61 (12.3 to 86) cells/µL at week 4, and 35.5 (4.75 to 88.5) cells/µL at week 12. Multivariate analysis showed that ENF AUC, number of PI in optimized background score, and the presence of lopinavir in optimized background score were independently associated with viral load decrease at week 4, whereas only optimized background score was associated with viral load decrease at week 12. Viral suppression at week 12 was reached by 15.2% of patients. Optimized background score and week 12 ENF C_trough were independently associated with virologic suppression at week 12 (p = 0.04; and p = 0.07, respectively). Moreover, week 12 ENF C_trough > 2200 ng/mL and optimized background score > 2 were associated to virologic suppression at week 12 (c² = 5.96, p = 0.035, and c² = 6.3, p = 0.012, respectively).

Conclusions:  Efficacy of ENF-based regimens in multi-experienced patients was previously shown to be related to optimized background score. Our data show that ENF plasma exposure is also independently associated to virologic response at week 12. Moreover, a C_trough cut-off of virologic efficacy (> 2200 ng/mL) was found. Therefore, therapeutic drug monitoring could be a useful tool to optimise efficacy of ENF-containing regimen in multi-experienced subjects.

Keywords: Enfuvirtide; PK/PD determinants; efficacy