645    

Background:  Outcome of atazanavir (ATV)-containing regimens showed to be related to a number of protease inhibitor (PI)-associated mutations and to plasma drug exposure, expressed as genotypic inhibitory quotient (gIQ). Moreover, ATV C_trough appeared to drive total and unconjugated bilirubin elevations. However, no ATV plasma thresholds of efficacy and safety have been yet determined. Therefore, the aim of our study was to define the ATV therapeutic range in the clinical setting.

Methods:  Patients from the ATV Expanded Access Program were prospectively evaluated. HIV RNA levels (viral load), T CD4⁺ cell counts, and total and unconjugated bilirubin levels were recorded at baseline, week 4, and week 12. Baseline genotypic resistance was assessed. ATV C_trough was measured between week 4 and week 12 by a validated SPE-HPLC method. gIQ was calculated as C_trough/number of PI-associated mutations ratio. Virological response at week 12 was defined as viral load of < 50 copies/mL or viral load decrease > 2 logs. Increase of unconjugated bilirubin > 2 mg/dL was considered as safety cut-off.

Results:  Of 38 patients evaluated (65.7% with boosted ATV), 32 had detectable viral load at baseline (median 4.43 log) and showed a median (IQR) viral load decrease of –2.2 log (–3.2 to –0.65). Overall, 80% of patients showed a virologic response. Median (IQR) of ATV C_trough and number of PI-associated mutations were 402 (78 to 934) ng/mL and 3 (2 to 4), respectively. ATV C_trough, number of PI-associated mutations and gIQ were found to be associated to an higher decrease of viral load at week 12 (R = 0.38, p = 0.04; R = 0.640, p = 0.008; and R = 0.5, p = 0.04, respectively). ATV C_trough > 150 ng/mL, PI-associated mutations < 5, and gIQ > 60 were associated to virologic response (c² = 8.8, p = 0.007; and, c² = 9.9, p = 0.025; and c² = 6.15, p = 0.03, respectively). Median (IQR) increases on total and unconjugated bilirubin at week 12 were 0.8 (0.26 to 1.69) and 0.61 (0.26 to 1.52) mg/dL, respectively. Moreover, 25% of subjects showed unconjugated bilirubin levels > 2 mg/dL. ATV C_trough was associated with 12-week serum levels of both total and unconjugated bilirubin (R = 0.47, p = 0.008; and R = 0.46, p = 0.009, respectively). ATV C_trough > 850 ng/mL was found in 75% of patients with unconjugated bilirubin > 2 mg/dL and in 12,5% of subjects below this unconjugated bilirubin threshold at week 12 (c² = 11.59, p = 0.002).

Conclusions:  In our study population we found that ATV C_trough ranging from150 to 850 ng/mL showed the highest probability of virologic response associated with the lowest probability of unconjugated bilirubin increase. Moreover, in populations with different previous PI experience and selection of PI-associated mutations, the use of gIQ > 60 could optimise the efficacy threshold by individualising the lower limit of this range. Therefore, ATV therapeutic range could be a tool for TDM.

Keywords: Atazanavir; therapeutic range; TDM