Abstract Form

EFFICACY OF LOPINAVIR/RITONAVIR IN 104 MULTIPLE PI-EXPERIENCED PATIENTS ACCORDING TO MUTATIONAL PATTERNS : DATA FROM THE FRENCH ATU PROGRAM

Author: V Calvez¹, I Cohen-Codar², AG. Marcelin¹, G. Peytavin³, D Descamps¹, J Ritter¹, C Tamalet¹, M Segondy¹, H Peigue-Lafeuille¹, E. Guillevic², R Rode⁴, J Isaacson⁴, D Kempf⁴, JP Chauvin^2
1HIV resistance group AC11, ANRS, France; ²Abbott France; ³Hopital Bichat, Paris, France; ⁴Abbott Laboratories, Abbott Park, IL, USA

Background of study: Lopinavir/ritonavir (LPV/r) was prescribed in France in a pre-registrational expanded access program (ATU, Autorisation Temporaire d’Utilisation) until April 2001 in 3819 patients. Among this cohort, we identified 104 patients who had had a HIV protease sequence analysis within 3 months before entering the program and a viral load follow-up available through 3 months.

Objective: to study in this subgroup of 104 patients the efficacy and the genetic determinants of the virologic response to LPV/r Design : the virologic response was defined as a HIV viral load < 400 copies/ml or a decrease from baseline of > 1.0 log₁₀ copies/mL. The protease gene mutations considered were those previously described to be associated with a reduced susceptibility to lopinavir (amino acids 10, 20, 24, 46, 53, 54, 63, 71, 82, 84 and 90). Statistical analysis was performed using Fisher’s exact test and univariate logistic regression supported by stepwise logistic regression.

Results: these subjects were extensively PI- and NRTI-experienced (mean 3.22, 4.74 drugs taken, respectively), moderately immunosuppressed (median baseline CD4 : 101 cells/mm³). Median baseline HIV RNA was 4.96 log₁₀ copies/mL. Most patients had been treated with tritherapy (58.7%) and quadritherapy (28.8%). The median baseline lopinavir mutation score was 5 (range 0 to 8). Through 3 months were observed a median HIV-1 RNA decrease of – 1.49 log and a median CD4 increase of + 93. The rate of virologic response was 55/66 (83%) and 17/38 (45%) patients, with baseline lopinavir mutation scores of 0-5 and 6-8, respectively (p<0.0001). The analysis of the other categorical genotypic breakpoints showed a weaker association with the virologic response.

Conclusions: A virologic and immunologic response to LPV/r was obtained, even in these heavily pretreated patients. A lopinavir mutation score of 0-5 vs. ≥ 6 was the clinically relevant genotypic breakpoints for an optimal efficacy of LPV/r observed in this cohort, which is consistent with the findings from other previous trials.

Vincent CALVEZ, Laboratoire de Virologie, CERVI Groupe Hospitalier Pitie-Salpetriere, 75013 PARIS, FRANCE