Dale Kempf, a researcher with Abbott, and Brendan Larder, a researcher with Virco, reported the following data at the recent Retrovirus Conference. This research suggests that using phenotypic resistance tests in the way weíve used them up to now may not be accurate in predicting resistance to ABT-378. Because of the high drug levels of ABT-378 in the blood 4 or 10-fold resistance may not predict non-response to this drug. New "cut-off" levels will have to be identified by Abbott. Because the drug is still in early development the actual new "cut-off" levels are in the process of being identified. Sometime this year we should have more revealing information on this.

Kempf reported in a separate report at Retrovirus:

--ABT-378 is an second generation PI that, co-formulated with a low dose of ritonavir (r), maintains mean plasma levels >30-fold higher than its EC_{50} for wt HIV. The authors also said the Cmin/EC50 ratio for currently available PIs is equal to or less than 4 fold, based on EC50 values determined in the presence of 50% human serumó

Treatment of patients failing a single PI therapy (but naÔve to NNRTIs) with ABT-378/r plus NVP and two NRTIs produced a decline to <400 copies/mL in 84% of patients on study at Week 48 (Study M97-765).

To analyze the virologic response in Study M97-765 after 24 and 48 weeks with respect to baseline viral phenotype and genotype.

Baseline phenotype was measured by the Antivirogram" method. Baseline genotype was determined by population sequencing. Virologic response was analyzed using logistic regression.

Baseline genotype and/or phenotype data were available for 61/70 patients. The incidence of greater than or equal to a 4-fold change in EC_{50} to 1, 2 or all 3 drugs in their existing regimen was 36%, 43% and 17%, respectively. The fold change in EC_{50} to ABT-378 ranged from 0.7 to 26-fold, with 19% of patients displaying a greater than or equal to a 4-fold change. The number of analyzed patients that experienced virologic failure (confirmed increase in HIV RNA to >400 copies/mL) at weeks 24 and 48 was 6/58 and 10/55, respectively. In both univariate and multivariate analyses, neither the presence of PI-associated mutations (p>0.2) nor a greater than or equal to a 4-fold change in EC_{50} to ABT-378 (p>0.6) correlated with virologic response. Instead, baseline HIV RNA was significantly associated with Week 24 (p<0.05) and Week 48 (p<0.10) response.

Statistical models utilizing protease genotype and a >4-fold cutoff in phenotype do not adequately describe the activity of ABT-378/r in patients failing an initial PI-based regimen. These findings and the high response rate are consistent with the high, sustained plasma concentrations of ABT-378 observed in this study.