RR:C19 Evidence Scale rating by reviewer:
Potentially informative. The main claims made are not strongly justified by the methods and data, but may yield some insight. The results and conclusions of the study may resemble those from the hypothetical ideal study, but there is substantial room for doubt. Decision-makers should consider this evidence only with a thorough understanding of its weaknesses, alongside other evidence and theory. Decision-makers should not consider this actionable, unless the weaknesses are clearly understood and there is other theory and evidence to further support it.
***************************************
Review: The paper using a specialized bioaerosol approach shows that a majority of individuals living in TB-endemic communities release Mtb containing bioaerosols during normal respiratory activities like breathing, challenging the current notion that pathology is a prerequisite for TB transmission. The study shows that as high as 80% of individuals randomly selected from a community exhale TB bacterium and they continue to exhale bacteria in similar numbers even when measured after months potentially contributing to TB transmission in the community.
The authors propose a reasonable alternative hypothesis that low-level exhalation of live but metabolically active bacteria for a sufficiently long time by non-sick individuals in the community may explain the invisible missing transmission links in TB transmission. This is also increasingly noted in other airborne diseases, especially viral diseases like COVID-19, Influenza etc. Overall, the current study attempts to link their previous study published in PNAS in 2024.
The results of the study are very interesting and critical, potentially informative in answering unanswered gaps in TB transmission, although the impact of low-level exhalation of bacteria to transmission as hypothesized in this study needs further evaluation and corroboration. Several comments and questions arise from this paper which need to be addressed or require explanation or discussion.
The authors have used DMN-Trehalose staining to detect Mtb and show high positivity rates. Although the authors themselves have highlighted the limitation of potential false positives using such staining to detect Mtb, they have not directly established the Tre specificity in this cohort. Was AFB staining attempted in aerosolized bacteria in this study? They seem to have relied on their earlier studies wherein they had triangulated the DMN-Tre staining with ddPCR, AFB and culture to establish the specificity of their staining method. However this interpretation when extended to community sampling needs caution as the earlier studies were done in confirmed TB patients or symptomatic presumptive TB patients, wherein the likelihood of finding TB is high and bacterial burden is expected to be high if present. The authors’ earlier study showed a higher median bacterial load in symptomatic presumptive patients than the individuals from the community in the present study.
The bacterial burden has an important effect on the specificity of detection. Generally, an increase in sensitivity of any particular test is often accompanied with compromise on specificity. For eg GeneXpert Ultra which can detect a lower number of bacteria has lower specificity than regular GeneXpert MTB-RIF (https://www.nature.com/articles/s41598-019-53086-5). Likewise, it has been noted that in a community screening of high endemic regions with high rates of smear-negative TB patients, the GeneXpert specificity is lower (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9328536/). Therefore, extrapolating the specificity assumption of a particular test in a population with likely high bacterial burden to individuals with very low bacterial burden requires caution.
The authors need to elaborate on the study population further to understand the potential confounding factors that may lead to overestimation of positivity- The description does not give details on the prevalence of TB in the selected erfs and how it compares to other regions in the same city, the TB contact history of the participants, recent history of infections and symptom status at the time of sampling. Some questions that arise are- Could there have been a bias in people who consented to take part in the study? What was the consent rate? Were consenting participants likely to be TB contacts? Were the participants completely devoid of any symptoms? were the participants infection-free for at least three weeks before sampling? It’s possible that any recent infection and associated inflammation of the airway may alter bioaerosol production and hence likely trigger the release of latent bacteria.
The authors in their discussion emphasize the potential role of exhaled bacteria from individuals toward TB transmission, primarily extrapolating their observations from confirmed and presumptive TB patients. Although the authors do recognize that the extent of attribution to transmission requires further research and evidence, the results of this study and their earlier studies do not explain the insights from human to guinea pig transmission studies which describe the rapid impact of treatment on reduction in infectiousness, wherein reduction in transmission was observed long before patients stopped becoming smear negative. (studies of Riley and Ed Nardell group). RNA sequencing of bioaerosols demonstrates that treatment has a profound impact on gene expression patterns of the bacteria especially genes involved in virulence and pathogenicity https://www.nature.com/articles/s41598-021-96902-7. Moreover, CASS studies on transmission rates in TB contacts (https://www.atsjournals.org/doi/full/10.1164/rccm.201208-1422OC) show that contacts of patients with TB who produced high aerosols were more likely to have a new infection compared with contacts from low-aerosol and aerosol-negative cases.
It would be interesting to see if the authors attempt direct inoculation of the bioaerosols on guinea pigs, a sensitive model to estimate the potential of these AFB-negative bacteria on transmission.
How do the authors explain the limiting factor of symptom resolution on Mtb count from Patterson's paper (PNAS 2024) to constant Mtb output across all longitudinal samples in this cohort?
The authors claiming 22% (11/50) discrepant sample as “relatively rare” is an overstatement. This cannot be considered as rare.
How was bioaerosol positivity defined for the cohorts? Was the presence of single DMN +ve stained bacteria considered positive or was a cut-off limit established based on the previous studies? In the first cohort of this study, what was the percentage of bioaerosol-positive participants in all three maneuvers?
What was the result of GXP testing at follow-ups? How many were tested? Was GXP done for all participants at follow-up, if yes what was the sample type?
It would be helpful if the authors revise the manuscript addressing and elaborating on the specific comments mentioned above. Overall, the study findings necessitate deeper investigations into aerosol positivity in the community, and understanding the replicability of the findings in other communities in S. Africa and low or non-endemic settings. More importantly, further studies should focus on measuring its attributability to TB transmission through animal models or environmental monitoring tools. Nevertheless, this study provides a critical hypothesis that questions current thinking and an opportunity to contemplate much-needed newer solutions to TB control.
Creative Commons Attribution 4.0 International License (CC-BY 4.0)
