TDM of nelfinavir and indinavir in treatment-naive patients improves treatment response.

Background: Adherence to protease inhibitor-containing antiretroviral therapy is crucial, but difficult to measure. Objective: To compare and combine various methods of measuring adherence to the strict protease inhibitor-containing regimens. Methods: The following methods were used: medication event monitoring system (MEMS) caps (electronic monitoring), therapeutic drug monitoring, pill count, pharmacy refill data, questionnaires, diaries (for registration of food patterns and special events related to the use of MEMS), adherence assessment by the physician and clinical nurse specialist, and in-depth interviews. In addition, ultrasensitive viral load and resistance testing was performed. Results: Twenty-eight patients were included; data could be evaluated in 26. According to MEMS data, 25% of the patients took fewer than 95% of all doses, and two thirds of the patients took fewer than 95% of the doses on time. Only 43% of the patients showed good adherence with food restrictions. Methods that showed significant correlations with MEMS results were patients' self-reported adherence; therapeutic drug monitoring, indicating plasma levels outside predefined ranges; and estimation of adherence by a clinical nurse specialist, especially by in-depth interview. Conclusion: Diary-corrected MEMS data gave a detailed insight into patients' adherence patterns. Patients' self-report and therapeutic drug monitoring were significantly correlated with the MEMS data, and the clinical nurse specialist may also play a role in identifying patients who are imperfectly adherent.

The relationship between plasma concentrations of nelfinavir and virologic treatment failure was investigated to determine the minimum effective concentration of nelfinavir. Plasma samples were prospectively collected from treatment-naive patients who began taking nelfinavir, 1,250 mg BID + two nucleoside reverse transcription inhibitors (NRTIs). Nelfinavir concentration ratios were calculated by dividing each individual nelfinavir level by the time-adjusted population value. Virologic failure was defined as either no response (a detectable viral load after 6 months) or a relapse (detectable viral load after being undetectable, or an increase in viral load >1 log above nadir). Forty-eight patients were included with a median follow-up period of 8 months. The median concentration ratio of nelfinavir was 0.98 (interquartile range, 0.76-1.47). Virologic failure was observed in 29% of the patients. In a univariate analysis, the nelfinavir concentration ratio appeared to be the single determinant that was related to virologic failure (P = 0.039). Patients with a median ratio <0.90 had a relative risk of 3.0 (95% CI, 1.2-7.6) for virologic failure. Using this threshold, virologic failures were detected with 64% sensitivity and 74% specificity (P = 0.014). Virologic failure of nelfinavir-containing triple therapy can be explained, to a large extent, by low plasma levels of nelfinavir.