Anything is Better Than The Skin Test’
Abstract & Commentary
Synopsis: An FDA-approved in vitro assay shows promise in the replacement of tuberculin skin testing.
Source: Mazurek G, et al. JAMA. 2001;286:1740-1747.
Mazurek and colleagues evaluated the results obtained with a commercial whole blood IFN-g ELISA assay compared with a 5 TU Mantoux tuberculin skin test (TST) in 1226 adults at 5 US centers. The ELISA measures the amount of interferon gamma (IFN-g) secreted by a whole blood preparation in response to antigens of Mycobacterium tuberculosis and Mycobacterium avium, as well as to phytohemagglutinin and a negative saline control. Subjects were categorized as high risk for latent tuberculous infection (LTBI) (947), low risk for LTBI (98), suspect TB (94), and active TB (87). The induration in response to TST was measured in the transverse diameter at 48-72 hours and interpreted according to the current ATS/CDC risk-stratified criteria.
Overall, 390 (31.8%) had a positive TST and 349 (28.5%) had a positive IFN-g assay, with an overall agreement of 83.1% (k = 0.60*). When the analysis was confined to the target intended for the assay, those undergoing screening for LTBI, the agreement between the 2 tests was 88.1% (k > 0.55). Among this group, agreement was 70.1% (k = 0.41) for those with a history of bacille Calmette-Guérin (BCG) vaccination and 88.1% (k = 0.50) for those not vaccinated. Prior BCG vaccination was associated with an excess of discordances characterized by a positive TST and a negative IFN-g assay. Multivariate analysis found that factors independently associated with a positive TST and negative IFN-g result included BCG vaccination, Asian race, site of study enrollment, and evidence of reactivity to M avium antigens by IFN-g assay. Mazurek et al conclude that "The IFN-g assay was comparable with the TST in its ability to detect LTBI, was less affected by BCG vaccination, discriminated responses due to nontuberculous mycobacteria, and avoided variability and subjectivity associated with placing and reading the TST." However, there was only 69% agreement (k = 0.16) between the tests in patients with culture proven active TB. (Editor’s Note: *k > 0.75 indicates excellent agreement beyond chance; k < 0.4 represents poor agreement beyond chance; k = 0.40.75 represent fair to good agreement beyond chance.)
Comment by Stan Deresinski, MD, FACP
The ancestor of the tuberculin preparations currently used as skin test reagents was Robert Koch’s "Old Tuberculin," which he touted as a therapeutic, not a diagnostic reagent. Mantoux introduced the intradermal injection of tuberculin in 1910, and, thus, was born the eponymous test used for evidence of prior exposure to M tuberculosis that, in perhaps more refined form, continues in widespread use to this day. This test, unfortunately, is a relatively crude one with a large number of pitfalls, such as problems with standardization of the reagent, variability in injection and in determining responses. In the current study, there was strong evidence of "digit preference" in which the observer rounds up to the nearest multiple of 5 mm in reading the skin test reaction. Furthermore, responses to the skin test are affected by the immune status of the host, by exposure to nontuberculous mycobacteria, and by prior vaccination with BCG. Finally, just getting the patient back in 48-72 hours to read the skin test can be a challenge.
In addition to the evidence reviewed here, a study conducted by the Walter Reed Army Institute of Research involving 1961 military recruits found a concordance of 92.1% and an estimated specificity of the IFN-g assay of 95.5% (www.fda.gov). However, since there is no gold standard for the diagnosis of LTBI, interpretation of a study such as this is fraught with uncertainties. Perhaps the strongest evidence of the accuracy of the IFN-g assay used here comes from the bovine form of the assay since cattle can be sacrificed and their tissue cultured for the presence of M bovis. In this setting, cattle that were TST positive but IFN-g negative had positive postmortem cultures in only 3.8% of cases, while cultures were positive in 55.2% of those with only a positive IFN-g assay (www.fda.gov).
The paper reviewed here provides evidence of an effective alternative to the TST. The assay used has been available in Australia since 1995 as the QuantiFERON-TBTM (QFT) assay (Celestis Ltd.) and was determined by the FDA in October 2001, to be "approvable with conditions." Since these conditions (further statistical analyses, additional interlaboratory reproducibility studies, etc) were not onerous, this test should soon be available for use in the United States.
Advantages of the test include a single patient visit, the ability to obtain a result in less than 24 hours, elimination of subjectivity, and lack of a "booster" effect.
Mazurek et al point out that the concordance of this test with TST is similar to that reported when multiple TSTs using different PPD preparations are administered to individuals. The reduced sensitivity in patients with active tuberculosis is, as Mazurek et al point out, consistent with the facts that delayed hypersensitivity reactions are independent of IFN g production (as indicated by experiments in IFN-g-knockout mice), and that active TB is associated with a decrease in IFN-g production in vitro in response to PPD.
As Dr. Valerie Ng, a member of the FDA Microbiology Devices Panel that unanimously approved the test kit stated, "Anything has to be better than the skin test."1
Reference
1. Reuters Health. Oct. 12, 2001.
Dr. Deresinski, Clinical Professor of Medicine, Stanford; Director, AIDS Community Research Consortium; Associate Chief of Infectious Diseases, Santa Clara Valley Medical Center, is Editor of Infectious Disease Alert.
You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content