New co-infection TB guidelines blast short-course, twice-weekly regimen

Gold standard for isoniazid prophylaxis is nine months, not six

New guidelines for treatment and prevention of TB in HIV- positive patients hold a stern reminder: What’s convenient for programs isn’t always what’s best for the patient. The recent official pronouncement on preventive therapy for HIV-positive patients comes as a disappointment to some program administrators, who had hoped for the go-ahead to provide the new short-course, two-month preventive regimen of pyrazinamide and rifampin (PZA/RIF) on a twice-weekly basis.

No way, the guidelines plainly state. Another unpleasant surprise to some is the reminder that when it comes to high-risk patients (including those infected with HIV), the duration of isoniazid prophylaxis isn’t six months; it’s nine months.

The recommendations reflect a conscious decision to drag people back from the brink of wishful thinking and to clarify what the data do and don’t support, explains Larry Geiter, MPH, PhD, a consultant with Sequela Research Foundation in Rockville, MD. "We wanted to make the point that recommendations for clinical decision making shouldn’t be made on a cost-effectiveness basis," says Geiter, an expert whose testimony is said to have swayed the roomful of other national experts gathered to mull over the new guidelines.

"I can understand people’s disappointment," says Rick O’Brien, MD, chief of the research and evaluation branch at the division of tuber culosis elimination at the Centers for Disease Control and Pre vention (CDC). Purely from a programmatic perspective, he says, the less onerous recommendations "would have been great." The problem is that the data don’t support such recommendations, he explains.

Reaction runs across the board. To some program administrators already hit with a new round of funding cuts, the guidelines offer powerful evidence that more study on short-course therapy is needed urgently.

"I think [rifampin and pyrazinamide] on a daily basis severely limit the advantages of short-course preventive therapy," says David Ashkin, MD, medical director for Florida’s TB control program. "We’ve had enormous amounts of difficulty getting people through preventive therapy."

Preventive therapy only works if it’s in the patient’s stomach, Ashkin argues; thus, what an intermittent course of PZA/RIF might lose in efficacy, it more than makes up for in adherence. "Even if RIF/PZA isn’t as effective by, say 3%, you more than offset that difference if you im prove adherence by 30% to 40%," he explains. Besides, "in the study that was already done, the data on twice-weekly looks as if it performed as well as daily," he adds.

Partly in the hope of gathering more evidence in support of twice-weekly PZA/RIF, Ashkin is considering a trial protocol of the twice-weekly regimen.

One investigator associated with the intermittent trial of PZA/RIF also expresses disappointment with the CDC guidelines, and with the decision to discount the results of that trial, which was conducted in Haiti. "It’s unfortunate," says Jacqueline Coberly, PhD, assistant scientist at the department of international health at the school of hygiene and public health at Johns Hopkins University in Baltimore.

The reasons the CDC discounted the results of the Haitian trial are complicated, says Geiter. Most importantly, the Haitian trial used as its standard for comparison a six-month course of self-administered isoniazid. However, nine to 12 months of INH, not six months, is the gold standard for HIV-infected patients and should have been the standard for comparison in the trial, he notes.

True enough, says Coberly. "There’s no question that nine months of INH is more efficacious than six," she says. "But when it comes to the probability of someone actually taking the nine months, we found only 55% of people actually finish; whereas 74% of people taking the [intermittent] PZA/RIF finish."

The trouble comes when the two kinds of regimens — the daily PZA/RIF and the intermittent — are stacked against one another, says Geiter.

Two separate trials — the Haitian study and another in Zambia, the results of which haven’t yet been published — compared two months of intermittent PZA/RIF to six months of INH, Geiter says. The Haitian trial put the level of protection for intermittent PZA/RIF at 60% and at 65% for six months of INH. Thus, that trial suggested intermittent PZA/RIF might perform as well as six months of INH, but it also might not. Point estimates in the Zambian trial suggested the same slightly ambiguous conclusions.

Third study more certain

On the other hand, a third study comparing 12 months of daily INH to two months of daily PZA/RIF left no such room for doubt, says Geiter. That trial found an 80% reduction from 12 months’ INH and an 84% reduction from two months of daily PZA/RIF. That suggests two months’ daily PZA/RIF might offer even better protection than 12 months of INH, he says.

"So we’re saying that PZA/RIF on an intermittent basis is probably just as good as six months of INH — but may actually be worse," says Geiter. "Plus, we already know six months of INH isn’t as good as 12." That, and the way both the Haitian and Zambian trials seemed to tilt against the intermittent regimen of PZA/RIF, left intermittent PZA/RIF saddled with too many doubts, he says.

As the dust settles, other experts are taking a more philosophical stance, declaring they intend to cut their losses by replacing nine months of INH with two months of daily PZA/RIF, for both HIV-negative and HIV-positive patients. That regimen "has been shown to be state of the art, and I say people have no business using nine months of anything if they can use RIF/PZA," says Lee Reichmann, MD, MPH, director of the National Tuberculosis Center at New Jersey Medical School in Newark. "It’s there; it works; I’ll use it."

To keep the debate interesting, at least one expert is blasting the CDC recommendation of daily PZA/RIF as misguidedly liberal rather than overly conservative. "PZA plus rifampin? You tell my buddies at the CDC that I love them all, but that I said this is going to be the biggest disaster they’ve ever given us," says John Sbarbaro, MD, professor of medicine at the University of Colorado Health Sciences Center in Denver.

"It goes totally contrary to everything we know about PZA in the continuation phase," which, he adds, is for all practical purposes the same as latency. "So why on earth add it? It merely adds hepatotoxicity. You’re going to see a ton of hepatitis with this new regimen," he says. "This whole thing will set us back for 20 years."

Minor side effects cause concern

One of Sbarbaro’s colleagues and neighbors is more sanguine on the subject. "I don’t think PZA is exactly dangerous," says William J. Burman, MD, infectious disease expert with the Denver TB control program. "But I do think there are some concerns about its tolerability. People may not want to put up with the side effects, which are fairly common, though seldom serious: the upset stomach, the itching after doses, the myalgia."

Back to that other line in the sand — nine months of INH as opposed to six months — Sbarbaro as well as experts on the more pragmatic side were more understanding.

"Nine months of INH? That’s just being conservative, but it makes sense," Sbarbaro notes. "What’s better? Eighty percent of the drugs for a few months, or 30% of the drugs for a year? It’s counter-intuitive, but the right answer is the second. That’s because in preventive therapy, you’re going after bugs that are dormant, and the longer you continue the therapy, the better the chance that the drugs will be there when the bugs wake up."

The nine-month rule also is important to emphasize because it helps explain why risk-benefit equations break down in the face of the higher risks for reactivation posed by HIV-positive patients, Geiter says.

"Back in 1985, people began talking about a six-month regimen of IPT being more cost-effective," he says. "What kind of got lost over time is the fact that the individual patient would still benefit from taking another six months of treatment. And with HIV-positive patients, the overall risk for progressing to TB is so high that even on a cost-benefit basis, going from six to 12 months is still worthwhile."

From the standpoint of what’s doable, Geiter adds that he doesn’t dispute appeals to practical arguments; he simply believes it’s important not to mix the nitty-gritty of real-life decision making with the practices dictated by the data.

Plus, there are ways programs can make regimens easier without sacrificing efficacy altogether, he adds. "If a public health program wanted to say, OK, we’re going to give people PZA/RIF, but we’ll only supervise them five times a week, or three times a week, and let them self-administer on the weekend’ — and then maybe give them nine weeks of the drugs instead of eight weeks — in practice that makes sense, even though it’s not supported by clinical trial data," Geiter says. "Or you could DOT [directly observed therapy] them five times a week, and give them a drug holiday on the weekend. There are all kinds of things you can do within reason to be creative."

In the same way, it’s acceptable to aim for getting six months’ worth of INH into "your garden-variety, HIV-negative patient," Geiter says. "But what we’ve tended to lose sight of is that the data don’t support that. What they actually show is that maximum benefit is achieved right about at nine months, and I think over the years we’ve lost sight of that fact."

With an HIV-negative patient, the risk for reactivation after six months is small to start with — about 10%. Six months of INH cuts it by 65%, so adding another three months only buys an additional 3.5%, which is not cost-effective in most cases, he says.

Practical considerations

There’s a practical reason to keep the bar high, he explains. If a program formally adopts a six-month course of INH as the ideal, it’s hard to go back to state legislatures after the fact and beg for more money for preventive care. On a smaller scale, a primary care physician who thinks mistakenly that six months is the ideal will probably be happy with five months’ worth of treatment, and that would surely be a mistake, says Geiter.

In sum: "If you’re going to make these decisions on a cost-effective basis, you need to be clear that’s what you’re doing," he says. "I understand the pressure people are under. I’m just saying, don’t mix clinical guidelines with cost considerations."