Vaccination with bacillus Calmette-Guérin (BCG) has been used for the prevention of tuberculosis (TB) in humans since 1921; approximately 100 million doses per year are administered to children in 170 countries of the world (1). BCG vaccine is thought to be effective in preventing or ameliorating severe complications of TB in children, i.e., disseminated disease or meningitis, but is of marginal efficacy in preventing adult forms of TB (2, 3). Because it does little to prevent the most infectious forms of TB, i.e., smear-positive cases, the epidemiologic impact of BCG for global TB control is limited. Industrialized countries with a low incidence of TB do not use BCG or have diminished its use, and revaccination of schoolchildren or young adults has been discontinued in most places (3).

Therefore in industrialized countries, the major effect of BCG vaccination in TB control programs involves tuberculin skin testing, the predominant method of detecting latent TB infection (LTBI) in persons at risk and identifying candidates for treatment of LTBI (4). BCG vaccination results in reactive tuberculin skin tests (TSTs) in many recipients, thereby creating the potential for confusion in interpretation; there is no reliable method to distinguish a true-positive TST due to infection versus a false-positive test from prior vaccination (5). This is most relevant in two situations: (1) in foreign-born persons from high-incidence countries who have emigrated to low-incidence countries but have usually received prior BCG vaccination, and (2) in TST programs that use two-step “boosting” to distinguish prior and recent infection, i.e., in health care settings.

Two articles published in this issue of the Journal help to inform us on issues related to the effect of BCG vaccination on TST (6, 7). Chee and colleagues (pp. 958–961) evaluated the interpretation of skin tests in Singapore schoolchildren, all of whom had received BCG at birth and many of whom, 83% and 36%, were revaccinated at ages 12 or 16, respectively, based on results of their nonreactive (< 10 mm) TSTs (6). The risk of developing TB in the 4 yr after TST reading or revaccination in these two cohorts of children was determined by cross-matching the results of TSTs with notifications of TB in the national database. The study found that in teenagers who had received one BCG vaccination at birth, 10 mm of induration was the most sensitive and specific cutoff for the development of TB, whereas in those who had been revaccinated, 16 mm was more predictive.

Singh and colleagues (pp. 962–964) evaluated two-step boosting of TSTs in health care workers without recent TB exposure in the United Kingdom, who were 21–54 yr old and had received BCG vaccination as teenagers or adults 8 to 40 yr previously (7). The subjects had TSTs read at 48, 72, and 96 h, and then repeat testing was done 1 wk later. A booster effect was noted with a mean increase of 7.8 mm induration at 48 h, but minimal or no increase at 72 or 96 h. This differential in boosting resulted in more positive results (> 15 mm) after the second TST (19 versus 9) when the reading was done at 48 h than at 72 h (10 versus 11).

What are the implications of these studies? The findings of Chee and colleagues corroborate prior studies which have shown that BCG vaccination can confound interpretation of TST in 0–80% of individuals, and that vaccination or revaccination given later in life is more likely to do so than that which is remote in time (8). The study nicely demonstrates that in those who had received one vaccination early in life, a 10-mm cutoff was predictive of TB, which pertains to most foreign-born persons arriving in industrialized countries. Also, the study suggests possibly using a cutoff greater than 10 mm for decisions about treatment of LTBI among individuals who have received more than one BCG vaccination; this may apply in those countries that have used BCG revaccination, with current low incidence and programs for treatment of LTBI. However, many countries with low or decreasing incidence of TB, such as in Singapore and several Eastern European countries, have discontinued BCG revaccination (3), and therefore this will be of diminishing relevance in the future. In developing countries where BCG vaccination is common (1), treatment of LTBI is not widely used and where it is being implemented, focuses on childhood contacts of infectious cases of TB and on human immunodeficiency virus (HIV)-infected persons (9).

For decisions about treatment of LTBI in foreign-born persons arriving in industrialized countries, should we use a different cutoff than the 10 mm recommended in most situations? History of BCG vaccination and revaccination in recent immigrants is not always accurate or available. More importantly, most positive TSTs in foreign-born persons are the result of true TB infection (8, 10), and risk–benefit for treatment of LTBI is favorable in young adults, the group evaluated in Singapore. Hence, we should continue to follow current recommendations which use 10 mm for most foreign-persons, and 5 mm for those who are recent contacts of persons with TB or who are HIV-positive (4). History of BCG vaccination should not be a factor in decision-making about LTBI for foreign-born persons from high-incidence countries.

The study by Singh and colleagues suggests that when using two-step TST to evaluate for boosting, as is done in many health care settings, 48 h may be the optimal time to read TSTs to enhance sensitivity, at least in persons with prior BCG vaccination (7). However, this may not be pertinent in places where BCG has not been used frequently, such as the United States, and therefore is of limited relevance for these situations. Of note was the difference in results when evaluating 15 mm as the cutoff, which reiterates that in recent guidelines, for health care workers who have no other known risks for LTBI, 15 mm rather than 10 mm was chosen in an attempt to enhance specificity of TSTs for this population (4). In contrast, for health care workers who have recently immigrated from high-incidence countries, the 10 mm cutoff applies and as previously noted, remote BCG vaccination should not be a consideration (5, 6).

The issues raised in these studies also remind us of the imperfections of TSTs and BCG vaccination as tools for global TB control. We need better diagnostic tests than TST to identify persons with LTBI and ideally, those at greatest risk for the development of TB (11). Certainly, a vaccine with greater efficacy than BCG to prevent TB is highly desirable (12). A decade ago we might have considered either goal elusive, but now with major advances in technology and a rejuvenation in TB research, our dreams may someday be a reality.

References

Section:

ChooseTop of pageReferences <<CITING ARTICLES