Do We Have Evidence for Policy Changes in the Treatment of Children With Latent Tuberculosis Infection?

For more than 50 years, isoniazid has been prescribed to prevent the progression of latent tuberculosis infection (LTBI) to overt TB disease.1,2 Treating LTBI benefits the individual by preventing TB disease, and it benefits public health by diminishing the pool of infection that could become contagious TB disease in the future.3 In this issue of Pediatrics, Finnell et al4 point to rising Mycobacterium tuberculosis drug-resistance rates globally and question the standard regimen of 9 months of isoniazid treatment. From a cost/benefit analysis, they conclude that rifampin should replace isoniazid for treating LTBI in immigrant children from countries where the likelihood of being infected with isoniazid-resistant M tuberculosis exceeds 11%. An estimated 586 000 US children aged 1 to 14 years have LTBI, 70% of whom were born overseas.5 The worldwide epidemic of drug-resistant TB warrants consideration of alternative treatment regimens. For a proposed strategy to be effective, treatment should be efficacious, safe, acceptable to the patient, and cost-effective. Isoniazid fulfills these criteria.Cost/benefit models for treating LTBI are sensitive to several variables: (1) the proportion of infections that are drug resistant; (2) the efficacies of the regimens; and (3) the approximations for factors that rarely have been measured directly and may not be generalizable, such as costs, rates of TB disease after infection, and rates of adverse drug effects. Another factor, reinfection with M tuberculosis after treatment, is regarded as negligible in the United States, because exposure is rare in most communities.The most controversial estimates in the authors’ model are the comparative efficacies of the regimens. The baseline input data were derived from studies among adults. Studies that suggested at least 90% efficacy for isoniazid under good conditions were not incorporated,6,7 and the model underestimates the efficacy of isoniazid in children. After a school-based screening program in San Francisco, California, during the late 1950s and 1960s, the risk of TB disease among children who were tuberculin reactors and who were not treated with isoniazid was 61 times greater than that for treated reactors (case rate: 20.9 and 0.34 per 1000, respectively).8 In a study that followed TB contacts for 30 years, Hsu9 reported efficacy of isoniazid near 100% for preventing TB in children. In the small, randomized trial of rifampin for treating LTBI, rifampin given daily for 3 months to adults who had silicosis had an efficacy of 63%, which was not distinguishable from isoniazid for that population.10 The efficacy of rifampin for preventing TB in children is unknown.The background drug-resistance rates for immigrant children are uncertain. Most countries do not include the culture isolation of M tuberculosis in their standard management of TB. Only a few countries that have large burdens of drug-resistant TB, such as South Korea, have uniform surveillance for drug resistance. The rates for the majority of countries have been derived from periodic surveys.11Finally, overall costs alone are misleading for public-health decisions. Finnell et al show that a major reason the rifampin regimen excelled in their model was the administrative savings of a 6-month regimen compared with the 9-month isoniazid regimen. However, the categorical costs are not exchangeable. Some costs, such as salaries and overhead, are unresponsive to shifts in practice. Savings in person-hours do not translate into funds for purchasing a more expensive drug, and cutting the number of employee positions is unwarranted at a time when some health departments have been pared down to below the capacity they need for investigating contacts of TB patients in their communities.Existing evidence supports the use of isoniazid for treating LTBI in most children and adults, and the encroachment of drug resistance has not altered the benefit of this approach in the United States. However, we must continue to evaluate the effectiveness of isoniazid for LTBI in settings with a growing burden of drug resistance. Alternative regimens are less efficacious (ie, 6 months of isoniazid), or their efficacy is uncertain (ie, 4–6 months of rifampin). Alternative treatment regimens might help for children who cannot be treated for 9 months (eg, children with transient living situations), children who cannot tolerate isoniazid, or children who have been exposed to patients with TB with known drug resistance.12The pivotal weakness of LTBI treatment as a prevention strategy has been the poor treatment-completion rate, commonly attributed to the long duration of available regimens. This is the practical reason why 3 months of isoniazid and rifampin combined, a regimen used in the United Kingdom for 2 decades,13 might be advantageous, as demonstrated in recent controlled trials.14 Another promising approach is a combined regimen of once-weekly isoniazid and rifapentine, 12 doses total, currently under study by Centers for Disease Control and Prevention's TB Trials Consortium. A 2-month regimen of rifampin and pyrazinamide was shown to be as efficacious as 12 months of isoniazid in adults with latent TB who were also coinfected with HIV.15 However, the promise of this relatively short regimen was negated by the occurrence of severe hepatitis and death.16 The missing ingredients for an evidence-based policy change to consistently treat LTBI and prevent TB cases include confirming alternative regimens that are proven effective, have a safety record, and are accepted—and completed—by all patients.17