It’s a clinical trialist’s dream: A prevalent disease, an effective intervention, and an implementation strategy that’s practical in a community setting. How could the result be anything but dramatically positive? In this week’s NEJM, the Thibela TB study investigators report on a cluster-randomized trial with a setting precisely this promising – and discuss their unexpected results.
The overwhelming majority of workers within South African gold mines have latent tuberculosis (TB), and reactivation contributes to a remarkably high prevalence of active TB (3000 cases per 100,000 miners; compare this to 859 cases per 100,000 among the general population of South Africa or 3 cases per 100,000 among the general population of the United States). Latent TB can be treated by a prolonged course of isoniazid, preventing reactivation and interrupting TB transmission. In the Thibela TB study, Dr. Gavin J. Churchyard (University of the Witwatersrand, South Africa), Dr. Katherine L. Fielding (London School of Hygiene and Tropical Medicine, United Kingdom) and colleagues hoped to use this strategy to reduce the incidence of active TB among miners.
The investigators identified fifteen South African gold mines with a total of 78,744 workers. All workers were offered screening and treatment for active TB. In the eight mines randomized to the intervention arm, workers without active TB were offered a nine-month course of isoniazid. In the seven mines randomized to the control arm, workers without active TB received standard diagnostic and treatment services but were not offered isoniazid.
As expected, the incidence of active TB was lower in intervention-arm mines during the period in which workers were receiving isoniazid. Unexpectedly, however, this difference disappeared in the year following isoniazid completion (3.02 vs 2.95 cases per 100 person-years; P=0.98). The investigators concluded that the transmission of tuberculosis between miners wasn’t interrupted by community-level isoniazid dispensation.
In an accompanying editorial, Dr. Eric J. Rubin (Harvard School of Public Health, Boston) characterizes the results as “disappointing and a bit surprising.” Possible explanations include the high prevalence of HIV and silicosis among South African gold miners, the low proportion of miners taking isoniazid simultaneously, and the possibility of post-treatment reinfection in the setting of extraordinarily high TB transmission rates.
“Although negative, these results don’t signify failure,” says Infectious Diseases specialist and NEJM Deputy Editor Dr. Lindsey R. Baden, “rather we learn that in an environment with a high TB transmission force additional TB prevention strategies are required.”
In the search for the ideal strategy to control TB, the research community will have to follow the example of hard-working miners everywhere: When you’re searching for the motherload, sometimes you have to dig a little deeper.