Challenges in the use of tests to diagnose tuberculosis infection

Dear Editors

Thompson agrees that both the tuberculin skin test (TST) and interferon-gamma release assays (IGRAs) have little application in diagnosing tuberculosis (TB) disease, although they provide important information about likely Mycobacterium tuberculosis infection, as does a careful TB exposure history.^1,2 We acknowledge that choosing a TST induration cut-off represents a sensitivity and specificity trade-off. The use of variable cut-offs (5, 10 or 15 mm) depending on the likelihood of TB exposure (using lower values when the likelihood of M. tuberculosis infection is higher) has been advocated by the American Thoracic Society.³ However, this rationale does not apply to TB-endemic areas where TB exposure risk is near universal. The World Health Organization (WHO) promotes a single cut-off of ≥10 mm (≥5 mm in immunocompromised individuals), using 5 tuberculin units (TU) of purified protein derivative (PPD) or equivalent, such as 2TU RT-23^® (Statens Serum Institute, Denmark).⁴ New South Wales uses 5TU (0.1 ml Tubersol^®; Sanofi Pasteur, Toronto, Canada) with WHO aligned cut-offs.

Since PPD includes multiple peptides that are also found in BCG vaccine and non-tuberculous mycobacteria, TST specificity is compromised. The QuantiFERON-Gold In Tube^® test (QFT-Gold IT) has improved specificity, especially in BCG-vaccinated individuals, but not all discordant results can be ascribed to poor TST specificity. Suboptimal QFT-Gold IT sensitivity is a concern, particularly in young children, while indeterminate results and blood sampling (which requires 3 ml of whole blood) pose additional challenges.^5,6 In the United Kingdom, National Institute for Health and Clinical Excellence (NICE) guidelines previously recommended a two-step process using TST for screening and QFT-Gold IT for confirmation, but revised guidelines encourage clinicians to consider preventive therapy in high-risk individuals who test TST positive and QFT negative.⁷ Since the risk of disease progression, including disseminated forms of disease such as TB meningitis and miliary TB, is highest in young children (<2–3 years of age),⁸ combined testing should be considered, with either positive test indicating likely M. tuberculosis infection. False-positive TST readings in BCG-vaccinated children remain problematic in this context. Research into better immunodiagnostic tests is a key research priority.

Ben J. Marais

Marie Bashir Institute for Infectious Disease and Biosecurity, University of Sydney and the Children’s Hospital at Westmead, Sydney, Australia