Evaluation of an immunoassay for determination of plasma efavirenz concentrations in resource-limited settings

Abstract

Introduction: Therapeutic drug monitoring (TDM) may improve antiretroviral efficacy through adjustment of individual drug administration. This could result in reduced toxicity, prevent drug resistance, and aid management of drug-drug interactions. However, most measurement methods are too costly to be implemented in resource-limited settings. This study evaluated a commercially available immunoassay for measurement of plasma efavirenz.

Methods: The immunoassay-based method was applied to measure efavirenz using a readily available Humastar 80 chemistry analyzer. We compared plasma efavirenz concentrations measured by the immunoassay with liquid chromatography tandem mass spectrometry (LC-MS/MS) (reference method) in 315 plasma samples collected from HIV patients on treatment. Concentrations were categorized as suboptimal<1 µg/ml, normal 1-4 µg/ml or high>4 µg/ml. Agreement between results of the methods was assessed via Bland-Altman plot and κ statistic values.

Results: The median Interquartile range (IQR) efavirenz concentration was 2.8 (1.9; 4.5) µg/ml measured by the LC-MS/MS method and 2.5 (1.8; 3.9) µg/ml by the immunoassay and the results were well correlated (ρ=0.94). The limits of agreement assessed by Bland-Altman plots were -2.54; 1.70 µg/ml. Although immunoassay underestimated high concentrations, it had good agreement for classification into low, normal or high concentrations (K=0.74).

Conclusions: The immunoassay is a feasible alternative to determine efavirenz in areas with limited resources. The assay provides a reasonable approximation of efavirenz concentration in the majority of samples with a tendency to underestimate high concentrations. Agreement between tests evaluated in this study was clinically satisfactory for identification of low, normal and high efavirenz concentrations.

Keywords: LC–MS/MS; TDM; antiretroviral therapy; efavirenz; immunoassay; method evaluation.

Figure 1

Bland–Altman plots of the agreement between the immunoassay and the LC–MS/MS. The solid line is the mean difference between the two methods (Immunoassay-LC–MS/MS); the dashed lines are 95% limits of agreement (±2 SD).