Therapeutic Drug Monitoring
Friday, 1:30 - 3:30 pm
Background: The use of therapeutic drug monitoring (TDM) in clinical practice is controversial since many TDM studies have found that few patients required dose changes. A primary objective of CCTG 578 was to determine the rate and predictors of TDM-based recommendations for regimen changes.
Methods: CCTG 578 was a 5-center, randomized, 3x2 factorial study of TDM vs SOC crossed with 3 adherence interventions. All patients had protease inhibitor (PI) and non-nucleoside reverse transcriptase inhibitor (NNRTI) drug levels drawn before and 2 and 4 hours post a witnessed dose at week 2. Pharmacokinetic modeling was done in real time; an expert committee, masked to randomization, reviewed data (TDM, HIV RNA, CD4, toxicity) and recommended regimen changes as needed. Site investigators received the recommendations only for patients in the TDM group. Rates and factors associated (by logistic regression) with a TDM recommendation (TDMR) to change drug exposure were evaluated from these week-2 recommendations.
Results: In 199 randomized patients, the mean baseline HIV RNA was 5.2 log copies, CD4 was 189, age was 40, and weight was 164 lb. Patients were ARV naïve (29%), experienced on a current regimen (33%), or experienced on a treatment interruption (38%). Baseline regimens included NRTI plus: NNRTI in 43%, PI in 61% (59% ritonaviar [RTV] boosted) and both in 5%. Lopinavir (LPV) (47%) and efavirenz (EFV) (32%) were the most commonly used drugs. Of the 177 week 2 TDM evaluations, 67 (38%, 95% CI 31% to 45%) were to change drug exposure of which 3 (1.6%) were to decrease exposure. Increased weight (OR 1.16 per 10 lb, p = 0.003), body mass index (OR 1.06, p = 0.05), EFV use (OR 1.8, p = 0.07) and LPV use (OR 1.9, p = 0.04) were associated with greater likelihood of having a TDMR to change drug exposure while patients with Hispanic ethnicity (OR 0.5, p = 0.05) were less likely. Factors not associated with a TDMR to change included: age, gender, baseline HIV RNA and CD4, prior ARV, and adherence (by MEMS). Independent predictors in a multivariate model were: weight (OR 1.16, p = 0.003) and use of EFV (OR 4.6, p = 0.001) and LPV (OR 4.6, p = 0.008).
Conclusions: In this cohort, a surprising 38% of patients had a TDMR to alter (mostly increase) drug exposure following careful evaluation by the committee. Greater baseline weight and use of LPV or EFV were significant independent predictors of the need for TDM. Since the majority of patients did not need TDM, future research should be aimed at refining the identification of candidates for TDM.
Keywords: Therapeutic Drug Monitoring; antiretroviral therapy; drug concentrations