Detection of low numbers of bacterial cells in a pharmaceutical drug product using Raman spectroscopy and PLS-DA multivariate analysis

Sterility testing is a laborious and slow process to detect contaminants present in drug products. Raman spectroscopy is a promising label-free tool to detect microorganisms and thus gaining relevance as a future alternative culture-free method for sterility testing in the pharmaceutical industry. However, reaching detection limits similar to standard procedures while keeping a high accuracy remains challenging, due to weak bacterial Raman signals. In this work, we show a new non-invasive approach focusing on detection of different bacteria in concentrations below 100 CFU per ml within drug product containers using Raman spectroscopy and multivariate data analysis. Even though Raman spectra from drug product with and without bacteria are similar, a partial least squared discriminant analysis (PLS-DA) model shows great performance to distinguish samples with bacterial contaminants in concentrations down to 10 CFU per ml. We used spiked samples with bacterial spores for model validation achieving a detection accuracy of 99%. Our results indicate the great potential of this rapid, and cost-effective approach to be used in quality control in the pharmaceutical industry. Fast and non-invasive approach to detect drug product (DP) samples with low numbers of bacteria within the primary packaging. The method combines Raman spectroscopy and partial least squared discriminant analysis (RS-PLS-DA).