The tuberculin skin test (TST) for determining if a person is infected with Mycobacterium tuberculosis has been the dominant test for a century. The interferon-γ release assays (IGRAs) have been available for a decade; dozens of studies have compared the performance of the 2 types of tests in a variety of settings. The advantages of the IGRAs have led some experts to recommend the “retirement” of the TST, even in pediatrics.1 However, the best use of the IGRAs in young children remains controversial because of a paucity of data.2
Both test types depend on the cellular immune response to mycobacterial antigens. Purified protein derivative, the solution used in the TST, has dozens of antigens, some of which are also found on many species of nontuberculous mycobacteria and the Bacillus Calmette-Guérin (BCG) vaccines, leading to low test specificity. The IGRAs use only 2 or 3 antigens that do not cross-react with most nontuberculous mycobacteria and BCG and consistently have higher specificity than the TST.3 Unfortunately, both test types have lower sensitivity in individuals with diminished cellular immunity who are at high risk of having tuberculosis infection progress to disease, and neither test type can be used to rule out tuberculosis disease in any child.
Children <5 years old with untreated tuberculosis infection have a high risk of developing tuberculosis disease. Studies in the pretreatment era showed that almost 50% of children <1 year of age with tuberculosis infection developed tuberculosis disease, with 15% developing tuberculous meningitis or disseminated disease; between 1 and 2 years old, 25% developed the disease, with 5% developing meningitis or disseminated disease.4 As a result, there has been a hesitancy to use the IGRAs in children <5 years old because of lack of data about the sensitivity of the tests.5 On the basis of more recent studies, the American Academy of Pediatrics Committee on Infectious Diseases lowered the recommended age for use of IGRAs to ≥2 years in the 2018 Red Book6 and stated that some experts use them in even younger children, but this practice is not yet widespread.
Ahmed et al7 reported the results of the Tuberculosis Epidemiologic Studies Consortium that evaluated 3493 children in 11 states for tuberculosis infection and disease over the course of 4 years. This is the largest US-based study comparing the results of the TST and both IGRAs (the QuantiFERON-TB Gold In-Tube [QFT] and T-SPOT.TB) in the same population. Approximately 92% of the participants were born outside the United States in countries where a BCG vaccine is given routinely. The overall agreement between the TST and IGRAs was 80% (κ = 0.37), and >90% of the discordant results were TST-positive and IGRA-negative. Non–US-born children <2 years old had the lowest concordance and the highest proportion of TST-positive and IGRA-negative results, which is likely due to their recent receipt of the BCG vaccine. Agreement between the 2 IGRA tests was high (93% for US-born and 97% for non–US-born children), supporting the common recommendation that 1 IGRA is not preferred over the other.5,6
Another hesitation to use the IGRAs in young children has been the high reported rates of indeterminate (QFT) or invalid (T-SPOT.TB) results, usually attributable to low reactivity of the positive control.8 In this study <1% of children had an indeterminate QFT or invalid T-SPOT.TB result; for children <2 years old, rates of indeterminate QFT and invalid T-SPOT.TB results were 2.3% and 0.5%, respectively. This issue should no longer limit the testing of young children with an IGRA.
The most important finding of the study is the outcome of the various groups. The study did not control for treatment decisions, which were made by the individual study sites. Only 4 children progressed to culture-confirmed disease; all 4 were either being treated with isoniazid or had received previous treatment. An IGRA result was positive in 3 of the 4 cases; all 3 test results were negative in 1 recently exposed 7-month-old child who later developed tuberculosis disease. However, none of the 533 untreated children who were TST-positive and IGRA-negative, including 54 children <2 years old, developed the disease during a 2-year follow-up period. This mirrors the experience of most pediatric tuberculosis experts. In our Children’s Tuberculosis Clinic at Texas Children’s Hospital, we have chosen to not treat several hundred children who were TST-positive and IGRA-negative whose only risk factor was foreign birth and are unaware of a single case of tuberculosis disease developing among them. The vast majority of positive TST results in this setting are false-positives.
The TST is disappearing from internal medicine and public health, which means that the expertise in placing the test and interpreting the “bump on the arm” is rapidly diminishing. The variation in the reaction size of an individual host varies by 15%; the variability in measuring induration among experienced observers also varies by 15% and is much greater among inexperienced personnel.9,10 One of our patients with inflammatory bowel disease developed severe cavitary tuberculosis after her TST result was misinterpreted as negative. When I asked the resident how many positive test results he had seen, he said, “None,” to which I replied, “Well, you have seen one now.” Considering what we now know about the IGRAs, if they had come first, and we were considering whether to adopt the TST, I doubt we would do so.
Opinions expressed in these commentaries are those of the author and not necessarily those of the American Academy of Pediatrics or its Committees.
FUNDING: No external funding.
COMPANION PAPER: A companion to this article can be found online at www.pediatrics.org/cgi/doi/10.1542/peds.2019-1930.
POTENTIAL CONFLICT OF INTEREST: The author has indicated he has no potential conflicts of interest to disclose.
FINANCIAL DISCLOSURE: Dr Starke is on a data safety monitoring board for Otsuka Pharmaceutical for pediatric studies of delamanid.