Individuals still saw TB incidence reduction of two-thirds, but effect was not durable
Preliminary results from the Thibela Project – which looked at whether community-wide provision of isoniazid preventive therapy (IPT) in South African gold mines would improve TB control in these high-incidence sites – were released at the Conference on Retroviruses and Opportunistic Infections Thursday in Seattle.
Dr. Gavin Churchyard of the Aurum Institute in Johannesburg explained that the gold mining population in South Africa has a historically high incidence of TB due to the high prevalence of HIV and silicosis in the mining community. Investigators studied 15 “clusters” comprising all mine workers at a mine shaft and associated hostels that ranged in size from 1,000 up to 12,000 individuals. The clusters were divided into two groups – where consenting miners in the eight intervention clusters were screened for TB symptoms and given a chest Xray. Those with suspected TB had one sputum specimen collected. If active TB was excluded, the miners received nine months of IPT. The seven control clusters received standard of care.
The study population comprised almost 80,000 gold miners, with slightly more in the intervention clusters – the largest trial of IPT ever conducted. The community-wide IPT intervention did not reduce TB incidence (3.04/100 person-years) as compared to the control group (2.96/100 person-years), or TB prevalence, which was 2.34 percent in the intervention arm and 2.14 percent in the control arm. It is important to note that the proportion of participants by cluster who collected at least six months of IPT or more was 35 percent to 79 percent – indicating a high rate of non-adherence to the full nine month IPT regimen. High reported rates of HIV (12 percent), prior TB (12.5 percent) and silicosis (1.9 percent ) in both arms were also listed as reasons the community-wide coverage of IPT did not bring incidence down.
Dr. Katherine Fielding of the London School of Hygiene and Tropical Medicine presented results from looking at individual TB incidence among those enrolled in the intervention arm, and found during the last six months of the nine-month IPT treatment period, those who had collected all monthly IPT packs by the start of month three experienced a 63 percent reduction in TB incidence and mortality; those who missed one or more IPT collection by the start of month three saw a 42 percent reduction. She stressed the benefit was only seen while the population was actively on IPT, and the protection wore off rapidly after treatment stopped, perhaps helping to explain the lack of population level effect of community-wide IPT.
During the question and answer period following the presentations, the presenters explained they will be looking further into the speed of IPT uptake and retention on IPT, as well as cluster size, to see how those affected outcomes. They also collected data on whether or not the individual employees were underground workers or service workers, which they can use to determine if there is a greater impact of IPT in one of those groups. They also stressed that 36-month IPT or life-long therapy might be a consideration for similar high-risk populations. Final results of the study will be released at the South African TB Conference in June.