Infectious Disease · Medical Literature
Evaluation of microscopically quantitative loads of bacilli as early predictors of TB treatment outcome.
Abstract
BACKGROUND Two-month smear conversion has been adopted instead of cultures to predict unfavourable bacteriological outcomes in TB treatment. However, this approach is limited by lower sensitivity and specificity. OBJECTIVE In absence of early-stage biomarkers of treatment response, we investigated the association between bacteriological outcome and microscopy data with auramine and fluorescein diacetate stains at baseline and during the first 2 weeks of treatment in the context of the OneRIF trial. This is a secondary analysis of data collected during the OneRIF trial. RESULTS Failure or relapse occurred in 15 patients of 321 treated with double-dose R. Only quantitative baseline auramine significantly predicted failure or relapse: for every 100 bacilli per field increase, there was 5% higher odds of having an unfavourable bacteriological outcome. CONCLUSION While quantitative baseline auramine predicted having an unfavourable bacteriological outcome, considering our small data we could not identify any added value in implementing quantitative bacillary load in the first 2 weeks of treatment. These findings should be interpreted with caution as we were limited by the relatively low number of failure and relapse. .
Clinical relevance
Two-month smear conversion remains the standard surrogate for predicting unfavourable bacteriological outcomes in TB treatment, yet it carries well-recognized limitations in sensitivity and specificity. This secondary analysis of the OneRIF trial investigates whether earlier, quantitative microscopy data — using auramine and fluorescein diacetate stains at baseline and within the first two weeks of treatment — can serve as actionable early biomarkers of treatment response. Identifying reliable early predictors is clinically urgent given the global burden of TB and the consequences of undetected treatment failure or relapse.
Key findings
1. Among 321 patients treated with double-dose rifampicin (double-dose R) in the OneRIF trial, failure or relapse occurred in 15 patients, representing a relatively low event rate that limits statistical power. 2. Quantitative baseline auramine smear load was the only microscopy parameter that significantly predicted failure or relapse: for every increase of 100 bacilli per field, there was a 5% higher odds of an unfavourable bacteriological outcome (specific confidence intervals and p-values not reported in the abstract). 3. Fluorescein diacetate staining at baseline did not demonstrate a statistically significant independent predictive value for treatment outcome based on the data presented. 4. Quantitative bacillary load measurements obtained during the first two weeks of treatment did not add predictive value beyond baseline auramine quantification in this dataset. 5. The two-month smear conversion endpoint, the current standard, was implicitly compared as a reference, with the authors noting its limitations of lower sensitivity and specificity relative to culture-based endpoints. 6. The findings must be interpreted with caution given the small number of failure and relapse events (n=15), which substantially constrains the ability to detect additional significant predictors or to generalize conclusions.
Practice implications
Clinicians managing pulmonary TB should be aware that a high quantitative bacillary burden on baseline auramine smear — specifically, elevated bacilli-per-field counts — may flag patients at incrementally higher risk of treatment failure or relapse, warranting heightened follow-up even within standard treatment regimens. However, given the limited event numbers and the absence of added predictive value from serial quantitative smears during the first two weeks, routine implementation of early serial quantitative microscopy as a monitoring tool is not yet supported by this evidence and should not replace standard culture-based endpoints or clinical judgment. Future adequately powered prospective studies are needed before quantitative early smear data can be formally integrated into TB treatment monitoring protocols or risk-stratification algorithms.