Elsevier

Tuberculosis

Volume 92, Issue 4, July 2012, Pages 314-320
Tuberculosis

Diagnostics
Point-of-care breath test for biomarkers of active pulmonary tuberculosis

https://doi.org/10.1016/j.tube.2012.04.002Get rights and content

Summary

Rationale

Volatile organic compounds (VOCs) in breath provide biomarkers of tuberculosis (TB) because Mycobacterium tuberculosis manufactures VOC metabolites that are detectable in the breath of infected patients.

Objectives

We evaluated breath VOC biomarkers in subjects with active pulmonary TB, using an internet-linked rapid point-of-care breath test.

Methods

279 subjects were studied at four centers in three countries, Philippines, UK, and India, and data was analyzed from 251 (130 active pulmonary TB, 121 controls). A point-of-care system collected and concentrated breath and air VOCs, and analyzed them with automated thermal desorption, gas chromatography, and surface acoustic wave detection. A breath test was completed in 6 min. Chromatograms were converted to a series of Kovats Index (KI) windows, and biomarkers of active pulmonary TB were identified by Monte Carlo analysis of KI window alveolar gradients (abundance in breath minus abundance in room air).

Measurements and main results

Multiple Monte Carlo simulations identified eight KI windows as biomarkers with better than random performance. Four KI windows corresponded with KI values of VOCs previously identified as biomarkers of pulmonary TB and metabolic products of M. tuberculosis, principally derivatives of naphthalene, benzene and alkanes. A multivariate predictive algorithm identified active pulmonary TB with 80% accuracy (area under curve of receiver operating characteristic curve), sensitivity = 71.2%, and specificity = 72%. Accuracy increased to 84% in age-matched subgroups. In a population with 5% prevalence, the breath test would identify active pulmonary TB with 98% negative predictive value and 13% positive predictive value.

Conclusions

A six-minute point-of-care breath test for volatile biomarkers accurately identified subjects with active pulmonary TB.

Section snippets

Clinical sites

Four tuberculosis treatment centers participated in the study, in the Philippines (University of Santo-Tomas, Manila, and De La Salle Health Sciences Institute, Cavite), UK (Homerton University Hospital, London) and India (Hinduja Hospital, Mumbai & Sir JJ Group of Hospitals, Mumbai).

IRB approval and informed consent

An Institutional Review Board (IRB) at each collaborating site approved the research. All subjects gave their signed informed consent to participate. Assent from adolescent subjects and consent from a parent or

Human subjects

279 subjects fulfilled recruitment criteria and 251 were entered into data analysis. Characteristics and exclusions are shown in Table 1. No adverse effects of the breath test were reported.

Instrument detection limit requirements were fulfilled by 0.1 μL of 0.1 ppt tridecane solution, which was equivalent to less than 10−12 mol tridecane in a breath sample.

Identification of KI window biomarkers

Multiple Monte Carlo simulations identified eight chromatographic KI windows as biomarkers with better than random performance (Figure 1a).

Discussion

The main finding of this study was that a model based on a point-of-care breath test for volatile biomarkers identified active pulmonary TB with 80% accuracy overall, increasing to 84% accuracy in age-matched subsets. This was consistent with our previous report of a laboratory-based assay for breath biomarkers that identified active pulmonary TB with 85% accuracy.7 Four KI windows corresponded with KI values of VOCs previously identified as breath biomarkers of pulmonary TB and metabolic

Author's contributions

Michael Phillips and Jaime Blais designed and supervised the study and drafted the manuscript. Anirudh Chaturvedi and Urvish Patel constructed and maintained the BreathLink systems. Anirudh Chaturvedi, Urvish Patel & Mauli Pandya performed quality control of the chromatograms and clinical data. Peter Schmitt and Anirudh Chaturvedi developed the software for the BreathLink system. Peter Schmitt analyzed the data statistically. Victoria Basa-Dalay, Graham Bothamley, Kinjal D. Modi, Maria Piedad

References (37)

  • H. Levy

    A reevaluation of sputum microscopy and culture in the diagnosis of pulmonary tuberculosis

    Chest

    (1989)
  • D. Wilkinson et al.

    Diagnosing tuberculosis in a resource-poor setting: the value of a trial of antibiotics

    Transactions of the Royal Society of Tropical Medicine and Hygiene

    (1997)
  • D.J. Horne

    Sputum monitoring during tuberculosis treatment for predicting outcome: systematic review and meta-analysis

    The Lancet Infectious Diseases

    (2010)
  • Anon.

    World TB day – March 24, 2008

    MMWR

    (2008)
  • L.M. Parsons

    Laboratory diagnosis of tuberculosis in resource-poor countries: challenges and opportunities

    Clinical Microbiology Reviews

    (2011)
  • S.L. Bailey

    Missed opportunities for tuberculosis diagnosis

    The International Journal of Tuberculosis and Lung Disease: the Official Journal of the International Union Against Tuberculosis and Lung Disease

    (2011)
  • M.D. Perkins et al.

    Facing the crisis: improving the diagnosis of tuberculosis in the HIV era

    Journal of Infectious Diseases

    (2007)
  • Compendium of chemical terminology

    (1997)
  • Cited by (143)

    View all citing articles on Scopus
    View full text