Aerosol Sampling Outperforms Sputum to Predict Mycobacterium tuberculosis Transmission Free

To theEditor —The study by Williams et al evaluating the use of face mask sampling (FMS) to predict incident Mycobacterium tuberculosis infection among household contacts (HHCs) is an important advance [1]. HHCs exposed to high FMS producers had a 3-fold increase in infection risk, measured by interferon gamma release assay (IGRA) conversion or an absolute increase (>1 IU/mL) between baseline and 6 months. These findings are similar to those we reported in Brazil and Uganda using a cough aerosol sampling system (CASS). The authors explain most of the similarities between the 2 methods, so we thought it would be informative to focus on the differences in methodology and interpretation.

First, there is significant difference in sensitivity between FMS and CASS. In addition to the reasons mentioned by the authors (free DNA and differentially culturable organisms), another potential explanation is the effect of antimycobacterial treatment [2]. A major strength on Williams' study was the exclusion of patients on any treatment; whereas the overall CASS positivity increased from 25% to 65% as we moved from partially to untreated pulmonary tuberculosis (TB) patients, the proportion of high aerosols producers (≥10 colony-forming units [CFU]) remained stable (∼30%–40% of aerosol positive cases). In Uganda, we also observed an increase of ≥1 in IGRA readouts in 21/83 (25%) of HHCs exposed to low or CASS negative versus 14/22 (64%) in those exposed to high aerosol TB cases (P = .001, odds ratio [OR]: 5.2 95% confidence interval [CI] 1.8–14.9) irrespective of time of TB treatment [3–4]. We suspect there is significant overlap between high producers FMS and high CASS phenotypes, which we posit represent disease superspreaders. Second, we used tuberculin skin tests (TST) rather that IGRA conversion as primary outcome and used both TST and IGRA rather than IGRA alone. This avoided the widely reported issues with IGRA specificity when using the manufacturer's cutoff and identified important yet poorly understood differences in TST and IGRA responses that we have hypothesized are related to variable intensities of exposure [5]. Third, the Williams and colleagues' study excluded children which eliminated the contact population most likely to reflect new infection.

An important limitation to our understanding of FMS is the lack of particle size data. In the supplementary information, the authors suggest they collected data on coughing and talking, both of which create larger size aerosols than exhaled breath [6] and could predict if the aerosol particles would remain airborne and to deposit in the alveolar region of the lung as shown with other respiratory pathogens [7]. We would appreciate them sharing such data, which could suggest the origin of the collected particles and help us better understand the pathogenesis and transmission of TB.

In conclusion, the study by Williams et al confirms that aerosol sampling is a better, more refined marker of infectiousness that the widely used sputum AFB smear; the adaptation of CASS and the more simplified, easier to use FMS could also work as a measure of the inhaled infectious inoculum and allow targeted preventive therapy strategies focused on contacts of high aerosol producers.

Notes

Financial support. K. P. F. reports support for this work from National Institutes of Health/National Heart, Lung, and Blood Institute (NHLBI) Division of Intramural Research (DIR) (100% funded by DIR).

References

Author notes