Quantification of isoniazid in tablets for tuberculosis treatment by thin layer chromatography with smartphone image capture and ImageJ analysis

Abstract

Isoniazid (INH) oral dosage tablets are an antibacterial medication commonly used to treat tuberculosis (TB). While this medication can be highly effective in managing and curing TB, manufacturing of poor-quality INH products can lead to the proliferation of drug resistant TB strains – particularly in low- and middle-income countries (LMICs). It is therefore imperative that reliable methods of quality screening of INH tablets are available for areas that have limited access to traditional testing resources. In this study, a thin layer chromatography (TLC) procedure for INH identification is validated in conjunction with smartphone image capture and open-source image analysis software to provide an alternative quantitative quality screening method for INH tablets, primarily intended for applications in LMICs. The method assessed two individual tablet formulations: 100 mg and 300 mg INH, each from different manufacturers. The modified TLC procedure used for the study was based on a validated high-performance thin layer chromatography method for the determination of rifampicin, INH, and pyrazinamide in a fixed dosage tablet. Quantitation was performed by visualizing and photographing TLC plates using a UV lamp, 3D-printed light box, and Apple iPhone SE 2nd Generation (2020) and Google Pixel 4a (5G) smartphones, and images were then analyzed using ImageJ Fiji software. The processed images’ pixel data were used to create a calibration curve from the INH standard spots to determine the concentration of an INH sample spot. Linearity, range, accuracy, repeatability and intermediate precision, specificity, and robustness were evaluated. The method was linear in the range of 0.150–0.450 mg mL−1, with coefficients of determination ≥0.98. The overall accuracy of the method was 99.2 % with an RSD of 2.4 % for image sets captured using the Apple iPhone, and 99.7 % with an RSD of 3.3 % for image sets captured using the Google Pixel. The pooled standard deviations for repeatability and intermediate precision were 2.27 % and 3.25 %, respectively, for the iPhone and 2.62 % and 3.74 %, respectively, for the Pixel. The method demonstrated sufficient specificity and robustness. The results of this method validation suggest that this procedure may provide a feasible alternative to traditional testing as a low-cost means of quality screening in LMICs.