Simultaneous quantification of artesunate and mefloquine in fixed-dose combination tablets by multivariate calibration with middle infrared spectroscopy and partial least squares regression.

Background: Malaria is one of the most lethal and life-threatening infectious diseases in the world, causing more than half a million deaths annually. Treatment with mefloquine and artesunate is currently recommended by the World Health Organization, and was historically the first artemisinin-based combination therapy used clinically for treatment of Plasmodium falciparum. The problem of poor-quality medicines, such as counterfeit and sub-standard anti-malarials, is a worldwide issue; therefore, it is essential to develop rapid, low cost, solvent-free, and reliable methods for routine quality control for these drugs. The aim of this study was to develop and validate a novel multivariate method for direct simultaneous quantification of mefloquine and artesunate in tablets by diffuse reflectance, middle infrared spectroscopy and partial least squares regression (MIR-PLS).

Methods: Diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) and partial least squares regression were applied for simultaneous quantification of artesunate and mefloquine in tablets provided by the Brazilian Government. The model was obtained with full spectra (4000-400 cm(-1)) preprocessed by first derivative and Savitzky-Golay smoothing followed by mean centring, and built with three latent variables. The method was validated according to Brazilian and international guidelines through the measuring of figures of merit, such as trueness, precision, linearity, analytical sensitivity, selectivity, bias, and residual prediction deviation. The results were compared to an HPLC-MS/MS method.

Results: The MIR-PLS method provided root mean square errors of prediction lower than 2.0 mg per 100 mg of powder for the two analytes, and proved to be valid according to guidelines for analytical methods that use infrared (IR) spectroscopy with multivariate calibration. For the samples obtained from Brazilian healthcare units, the method provided results statistically similar to those obtained by HPLC-MS/MS.

Conclusion: MIR-PLS was found to be suitable for the quality control of these drugs. It is fast, does not use solvents, and does not generate chemical waste. Furthermore, the proposed method may be transferred and developed for use in portable equipment, increasing the possibilities for assessing the quality of these drugs.