Stool processing protocols for Xpert MTB/RIF

Various amounts (0.3 g to 1.0 g) of two stool samples with a soft consistency were used to test different stool processing protocols before testing with the Xpert MTB/RIF assay (Cepheid) for the detection of M. tuberculosis and resistance to rifampin. This testing was done on a GeneXpert instrument (Cepheid), following the instructions of the manufacturer. The protocol for the proposed SOS stool method consisted of the standard protocol for sputum processing for Xpert testing (16 (link), 17 (link)) with a few minor modifications (see details below), i.e., a single release/sedimentation step. In addition, various other stool processing protocols were tested, with the following additional steps: (i) addition of glass beads to improve sample homogenization and potentially increase the release of M. tuberculosis, following the method described by Banada et al. (12 (link)); (ii) dilution with PBS, resembling the TS method described by Andriyoko et al. (15 (link)); and (iii) filtration following the method described by Banada et al. (12 (link)). Also, protocols that consisted of various combinations of these additional steps were evaluated. Figure S1 in the supplemental material shows a schematic overview of the series of laboratory experiments and more details on how these were performed.