Multi-methodological quantitative taste assessment of anti-tuberculosis drugs to support the development of palatable paediatric dosage forms

The unpalatability of antituberculosis drugs is often cited as a major cause of non-adherence in children, yet limited quantitative taste assessment data are available. The aim of this research was to quantify the bitterness of isoniazid, rifampicin, pyrazinamide, and ethambutol dihydrochloride using two in vivo (a human taste panel and a rat brief-access taste aversion (BATA) model) and one in vitro (sensor) method. The response of the Insent TS-5000Z electronic tongue was compared to the in vivo drug concentration found to elicit and suppress half the maximum taste response (EC₅₀ in human and IC₅₀ in rats). Using dose-relevant concentrations, an overarching rank order of bitterness was derived (rifampicin > ethambutol > pyrazinamid~isoniazid). In vitro, only ethambutol exhibited a linear response for all sensors/concentrations. Based on the EC₅₀/IC₅₀ generated, a ‘taste index’ was proposed to allow for anticipation of the likelihood of taste issues in practice, taking in account the saturability in the saliva and therapeutic doses; ethambutol and isoniazid were found to be the worst tasting using this measure. The study presents the first quantitative taste analysis of these life-saving drugs and has allowed for a comparison of three methods of obtaining such data. Such information allows the operator to identify and prioritise the drugs requiring taste masking to produce palatable formulations.