Background. ABT-378/r is a new protease inhibitor (PI) that, by virtue of high trough plasma levels, has shown activity in PI-experienced patients whose isolates contain multiple mutations in protease. In order to properly interpret the results of HIV genotypic resistance assays with respect to ABT-378/r, the response to therapy with ABT-378/r + EFV + NRTIs in 52 multiple PI-experienced, NNRTI-naïve subjects with respect to baseline genotype was examined.

Methods. The binary response variable (HIV RNA > or <400 copies/mL) at Week 24 was analyzed by univariate and stepwise logistic regression.

Results. Virologic response was associated (p=0.019) with the number of baseline mutations at positions 10, 20, 24, 46, 53, 54, 63, 71, 82, 84, and 90 in HIV protease. Response point estimates were 96, 88, 70, and 43% for subjects with 3, 5, 7 and 9 baseline mutations, respectively. In addition to the number of baseline mutations, the final model of a stepwise logistic regression procedure (p<0.15 for entry and exit) included time since HIV diagnosis, subject weight, the number of new NRTIs and the NRTI susceptibility (>2.5-fold). The F53L mutation, which was only present with 5-9 other mutations, was the best predictor of failure among the 11 individual mutations, but was less highly associated with response than the total number of mutations.

Conclusions. ABT-378/r exerts significant antiviral activity in vivo even in patients with up to seven baseline mutations associated with reduced phenotypic susceptibility in vitro. The number of mutations is the best genotypic predictor of response. These results may be useful to guide the interpretation of genotypic resistance testing with ABT-378/r and suggest that the genetic barrier to resistance to ABT-378/r is large.

¹Abbott Laboratories, Abbott Park, IL, ²ViroLogic, South San Francisco, CA

---

Main New/Newsworthy

Kaletra Conference Data

Home

---