The trusted source for
healthcare information and
Loss of isolates precludes RFLP testing
HIV-positive TB patients who have finished treatment and show signs of immunocompromise should be considered for post-treatment TB prophylaxis, says Warren Johnson, MD, the B.H. Kean professor of tropical medicine and chief of the division of international medicine at Cornell University.
In a country like Haiti, where TB is endemic, such patients should be certainly considered for lifelong post-treatment prophylaxis — perhaps even lifelong prophylaxis, he adds. In Haiti, Warren and other researchers found that giving 12 months of isoniazid to such patients dropped the incidence of recurring TB in HIV-infected patients significantly, from 7.8 per 100 person-years to 1.4 per 100 person-years.
The trouble is that Johnson (whose research was published recently in The Lancet) doesn’t know whether the recurrences reflected relapses of old disease or a new infection, because some of the strains that were to be tested were lost. "So we don’t have RFLP [restriction-fragment-length polymorphism] or other markers to indicate whether the recurring TB was exogenous or endogenous," he says. "I wish we had the information."
From a practical standpoint, it didn’t make any difference to the group in Haiti, he points out. "They had TB, so they had a problem," he says. For patients in other areas where TB is endemic and there’s lots of HIV infection, the same logic probably applies. "But from the standpoint of extrapolating to parts of the world like the United States, it does raise questions," he adds. "If it was new TB, the probability of acquiring a new bug in Haiti certainly would be much greater in Haiti than in the U.S."
Still, if what researchers in Haiti were seeing was reactivation of old disease, "then the premise is that it’s going to recur when you’re immunocompromised. I don’t know the answer, but if you had someone here with HIV and they begin to deteriorate, then maybe you need to go back on INH. Whether it’s your own bug that will reactivate or whether you’ll contract someone else’s, you may be able to control a new infection. And you’ll get disease, whether it’s exogenous or endogenous."
Relapses occur in HIV patients here, too
Indeed, American patients with HIV who kept either relapsing or succumbing to new TB infections provided part of the impetus for the study, Johnson says. "We started seeing occasional patients here, and they were also seeing them in Haiti," he says. "These were patients who developed TB, who’d been treated and were doing well, but who developed TB again a couple of years later." There have been other studies where researchers looked for the same effect but didn’t find it, perhaps because they didn’t follow patients long enough, Johnson adds. "We looked for 24 months after treatment," he says. "We followed some patients even longer."
Well, yes and no, says Rick O’Brien, MD, chief of research and evaluation branch of the Division for TB Elimination at the Centers for Disease Control and Prevention in Atlanta. "In a Zairian study that was done five or six years ago, they saw the same thing," O’Brien says. That study, structured along the same lines as Johnson’s, randomly assigned HIV-positive patients who’d finished their TB treatment to continue therapy with rifampin, isonizaid, or placebo. Comparing the entire HIV-positive group with a group of HIV-negative patients, those researchers found "no significance in relapse rates, and so they concluded that standard therapy [without any post-treatment prophylaxis] was fine," explains O’Brien.
But an analysis of all the arms of that study turns up something else, he continues. Relapse rates among the HIV-positives on extended therapy turned out to 1%; among the HIV-positive on standard therapy, 9%; and among the HIV-negative group on standard therapy, 5%, he says.
That means there’s no statistically significant difference between the HIV-positive and HIV-negative groups, with thjeir respective relapse rates of 9% and 5%, he explains. But by adding extended therapy, relapse rates dropped from 9% to 1% — a big difference.