Background:  Tenofovir (TDF) plasma exposure is increased when co-administered with some ritonavir-boosted protease inhibitors (PI/r). We hypothesized that TDF-treated patients taking a PI/r would have higher TDF plasma exposures and greater declines in renal function than TDF + non-nucleoside reverse transcriptase inhibitor (NNRTI)-treated patients.

Methods:  Subjects from CCTG 578, a completed therapeutic drug monitoring study, were eligible if they received TDF, a PI/r or an NNRTI for 48 weeks. TDF concentrations were measured in stored plasma samples collected at pre- and 2  and 4 hours post a witnessed dose at study week 2 by a validated mass-spectroscopy technique. A population pharmacokinetic model was developed and individual TDF pharmacokinetic parameters were estimated by Bayesian methods. Glomerular filtration rates (GFR) were estimated using Cockcroft-Gault (C-G) and Modification of Diet in Renal Disease (MDRD) equations. Mixed effects models compared GFR decline over time between treatment groups. Covariates such as age, baseline CD4, viral load and treatment history were included in the model if they were significant at α = 0.25. Subjects on non-TDF-containing regimens were analyzed as an additional control group.

Results:  We evaluated 147 subjects:  51 received TDF+PI/r-, 29 TDF+NNRTI- and 67 non-TDF-containing regimens. Baseline mean age (40 years), CD4 (197 cells/mm³) and GFR (C-G: 107 mL/minute and MDRD: 108 mL/min/1.73 m²) were similar between the 3 groups. After 48 weeks of therapy, change in estimated GFR between non-TDF- and TDF+NNRTI-treated patients were not different. In mixed effects models, patients receiving TDF+PI/r had an increased rate of GFR decline compared to the TDF+NNRTI group (C-G 13.9 vs 6.2 mL/minute/year, p = 0.033; and MDRD 14.9 vs 4.3 mL/minute/1.73 m²/year, p = 0.018). While age was associated with GFR (–1.2 mL/minute/year, p = 0.001), baseline CD4, HIV RNA, and treatment history (naïve vs experienced) were not. In the population pharmacokinetic model, week 2 TDF clearance was associated with baseline creatinine clearance. Although TDF+PI/r-treated patients tended to have higher TDF plasma exposures (C_max 235 vs 219 ng/mL, p = 0.39 and C_min 76 vs 58, p = 0.23; TDF clearance 89 vs 100 L/hour, p = 0.26), week 2 TDF pharmacokinetic parameters did not predict renal function over time.

Conclusions:  Combination therapy with TDF+PI/r was strongly associated with greater renal function decline over 48 weeks compared to TDF+NNRTI-based regimens. Differences in week 2 TDF plasma exposure did not explain this observation.