Quantification of bactericidal activity using the PATHFAST TB LAM Ag assay during the first 14 days of pulmonary tuberculosis treatment

Background: Tuberculosis (TB) remains a global health challenge. Culture-free, rapid, and quantitative biomarkers to monitor treatment response are critical to accelerate development of better TB treatments. The PATHFAST TB LAM Ag assay (PATHFAST-LAM), a desktop chemiluminescent enzyme immunoassay that measures mycobacterial lipoarabinomannan (LAM) in sputum within 1 hour, is a promising candidate for this purpose. This study aimed to assess whether the PATHFAST-LAM can serve as a rapid, reliable biomarker for monitoring early treatment response in pulmonary TB. Methods: We conducted a retrospective longitudinal repository study using stored sputum samples from 14-day early bactericidal activity trials involving 75 pulmonary TB patients who received one of five different regimens. The results were compared with those from the TB LAM ELISA "Otsuka" (LAM-ELISA), which was previously shown to measure early bactericidal activity but takes more than 5 hours to obtain results, and conventional culture-based methods. Results: The LAM concentrations in both raw and decontaminated sputum showed strong correlations between the PATHFAST-LAM and the LAM-ELISA, with Spearman's correlation coefficients of 0.975 (95% CI: 0.971 - 0.979) and 0.987 (95% CI: 0.984 - 0.989), respectively. LAM concentrations in raw and decontaminated sputum by the PATHFAST-LAM were also highly correlated with a Spearman coefficient of 0.957 (95% CI: 0.950 - 0.964). Importantly, the LAM concentrations by the PATHFAST-LAM correlated with bacterial loads determined by culture-based methods in all five treatment arms (Spearman's coefficients: 0.723 - 0.947). Furthermore, the change in LAM levels over the treatment period mirrored the changes in bacterial load. Additionally, culture-based methods often result in missing data due to contamination: in our study, we observed a missing data rate of 9.6% (62/649) on quantifying CFU counts and 4.2% (27/649) on obtaining a valid MGIT TTD, while we obtained a valid LAM value with the PATHFAST-LAM (0/634 in raw samples and 0/637 in decontaminated samples). Conclusion: Our findings suggest that the PATHFAST-LAM can quantify bactericidal activity in the first 14 days of treatment with a quick turnaround time. The test's utility to monitor conversion from positive to negative and to predict relapse-free cure compared to culture-based methods should be determined in longer trials.