Article Text

Download PDFPDF

The tuberculin skin test: a hundred, not out?
  1. Delane Shingadia,
  2. Vas Novelli
  1. Great Ormond Street Hospital, London, UK
  1. Dr Delane Shingadia, Department of Infectious Diseases, Great Ormond Street Hospital, Great Ormond Street, London WC1N 3JH, UK; shingd{at}gosh.nhs.uk

Statistics from Altmetric.com

In 1890, at a World Congress of Medicine in Berlin, Robert Koch announced a substance that he claimed would both cure and prevent tuberculosis. Although this substance, a glycerin extract of tubercle bacilli, subsequently failed as a therapy, it has become an important diagnostic tool for tuberculosis. Referred to eventually as “old tuberculin”, this tubercle protein soup was later refined to create the current diagnostic material we now know as purified protein derivative or PPD.

Over the last century, the tuberculin skin test (TST) has become the established screening method for diagnosing latent TB infection (LTBI) in adults and children. The TST exploits the fact that LTBI induces a strong cell-mediated immune response to intradermal inoculation of tuberculin PPD, a crude mixture of >200 Mycobacterium tuberculosis proteins. However, the TST suffers from poor specificity (false positive tests due to prior BCG vaccination or previous exposure to environmental mycobacteria) and poor sensitivity (false negative tests particularly in children and immunocompromised individuals). Furthermore, the TST does not differentiate active disease from LTBI, so additional microbiological, radiological and molecular investigations are used to establish a diagnosis of TB disease. The diagnosis of active disease in children remains difficult with much lower “culture confirmed” disease than in adults. However, in children up to 50% of infants and 15% of older children with LTBI who do not receive chemoprophylaxis, will develop disease within 2 years of being infected, highlighting the importance of identifying children through contact tracing.1

Recently, new tests, based on the production of interferon-γ by sensitised lymphocytes in response to specific M tuberculosis antigens, have been developed. The interferon-γ release assays (IGRA) rely on M tuberculosis specific antigens, early structural antigenic target-6 (ESAT-6) and culture filtrate protein-10 (CFP-10), which are derived from the region of difference-1 (RD-1) of the M tuberculosis genome. These antigens are absent in BCG (deleted during Mycobacterium bovis attenuation) and most environmental mycobacteria. Two assays are available for rapid, convenient measurement of antigen-specific T cell responses in blood. The T.SPOT-TB test (Oxford Immunotec, Oxford, UK) is a rapid ex vivo enzyme-linked immunospot assay which estimates the number of individual antigen specific T cells, while the QuantiFERON-TB Gold test (Cellestis, Carnegie, Australia) is a whole blood enzyme-linked immunosorbent assay which measures the interferon-γ concentration in the supernatant of whole blood after incubation with antigen.

Recently, the US Centers for Disease Control and Prevention have recommended that the TST be completely replaced by the QuantiFERON test.2 In contrast, the National Institute for Health and Clinical Excellence (NICE) in the United Kingdom has recommend the use of IGRA in the following groups: (1) individuals at risk of LTBI who have tested positive by TST and (2) individuals in whom the TST may be unreliable (ie, those with cellular immune suppression who commonly have false negative TST results).3 The NICE guideline development group recognised the lack of evidence on the use of these tests in younger children and recommended that the guideline be updated if significant new evidence emerged.

In this issue, Taylor et al4 present data on IGRA testing in children using QuantiFERON-TB Gold and determine the impact of NICE guidelines on the management of children investigated as part of contact tracing. They report an 85% reduction in the need for chemoprophylaxis using the two-step policy (TST followed by IGRA) because of the exclusion of false positives due to BCG. IGRA tests were more likely to correlate with a negative TST (98%) than with a positive TST (11% agreement). However, the authors also identified two children with probable active TB (although not culture confirmed) who would not have been identified using NICE guidelines (one child with negative TST but with hilar lymphadenopathy on chest x ray, the other with a positive TST and negative IGRA but chest x ray with hilar lymphadenopathy and consolidation). This study highlights the reliance on the superiority of IGRA testing over TST within the NICE guidelines.

Perspective on the paper by Taylor et al (see p 10.1136/adc.2006.106617)

The field of IGRA testing is rapidly changing with emerging data, some conflicting, on the use of these tests in children. Several studies examining the use of IGRA in diagnosing TB disease in children have been published. Liebeschuetz et al showed that in children in South Africa, the ELISPOT assay had a sensitivity of 83% compared with 63% for the TST.5 The sensitivity of TST decreased significantly with younger age, HIV infection and malnutrition, whereas the sensitivity of ELISPOT was not significantly affected. Dogra et al showed complete agreement between TST and QuantiFERON-TB-Gold in rural India with no impact of BCG on either test.6 In a low incidence country, QuantiFERON-TB-Gold, T.SPOT-TB and TST had specificities of 100%, 98% and 58%, respectively, and sensitivities of 93%, 93% and 100%, respectively.7 These studies have found that IGRA are of value in diagnosing active TB but should not replace appropriate microbiological and molecular investigation. The role of IGRA in the diagnosis of extrapulmonary disease has so far not been established. More carefully designed head-to-head studies are needed, particularly for those groups in which TB disease is difficult to diagnose, for example children, immunocompromised individuals and patients with extrapulmonary disease.

The role of IGRA in diagnosing LTBI is less clearly defined because of the lack of a gold standard. Most published reports are based on TB incidents where there is correlation with degree of exposure and comparison with TST.8 Several studies have been published on children investigated as part of contact tracing. Agreement between TST and IGRA appears to be between 70% and 83%, with the suggestion that the sensitivity of IGRA may be lower than that of TST. The issue of discordant IGRA and TST results has been highlighted in several studies, including that of Taylor et al.4 Epidemiological and clinical information suggests that TST may be more accurate than IGRA when there are discordant results. Although it is reasonable to assume that a positive IGRA is predictive of later active TB, as is a positive TST, there is no evidence so far that suggests a higher or lower degree of predictability. Longitudinal studies are needed to establish the real probability of a positive IGRA in predicting future active TB.

The other issue that has been raised is that of indeterminate results. Connell et al reported a 17% failure rate, mainly due to inadequate mitogen control responses.9 Other studies have reported lower failure rates, although there is concern that some children, particularly younger infants, may have higher failure rates due to decreased interferon-γ production.

IGRA are the first major advance in the diagnosis of TB for over a century. While these tests offer hope that we finally have an improved diagnostic tool for the diagnosis of TB in children, we need to be cautious that we are not over-reliant on these tests where the evidence is lacking. Before we completely replace TST with IGRA, we would need to have firm evidence in support of such a recommendation. Other factors, such as cost and technical considerations, may favour the continued use of TST, particularly in resource-limited settings. Ultimately as with TST, clinical judgement should always be exercised when interpreting results to ensure appropriate management of the child.

REFERENCES

View Abstract

Footnotes

  • Competing interests: None.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles