CDC changing prophylaxis position on HIV-positive patients

Studies show preventive therapy benefit too small

Tuberculosis researchers and clinicians have been at odds over the need to offer isoniazid prophylaxis to HIV-positive patients, particularly if they are anergic. New evidence supports the notion that preventive therapy in this population is unwarranted, but health officials say it may have its place in settings where HIV-positive patients are at higher risk of TB exposure, particularly correctional facilities.

Since 1991, the Centers for Disease Control and Prevention in Atlanta has recommended that anergic HIV-positive patients be considered for preventive therapy if they belong to groups in which the prevalence of tuberculosis is greater than 10%. While anergic HIV-positive patients are deemed high-risk for developing TB, no study has investigated their benefit in receiving prophylaxis.

The rationale for preventive therapy in this population is based on several observational studies that found HIV-positive patients who had positive PPD tests but received no preventive therapy had a much higher rate of active TB than those who did receive prophylaxis. At the same time, the inability to diagnose latent TB infection in some HIV-positive patients (i.e., being anergic) makes preventive therapy more attractive in this group.

In recent years, however, studies have questioned the diagnostic benefit of anergy testing (see TB Monitor, July 1997, p. 75) and the cost benefit of prophylaxis in high-risk HIV-positive patients.

The latest and most compelling evidence against the recommendation to prophylax anergic patients comes from a large controlled trial conducted from 1991 to 1996.1 The study followed 517 patients with two or more risk factors for TB and who were treated with six months of isoniazid (INH) or placebo. The patients, three-quarters of whom were from New York City, were similar in both groups, with 23% diagnosed with AIDS and a mean CD4 count of 240. The only difference was that the placebo group had more patients who were unemployed for a year or more, which could have put them at higher risk of exposure. In both groups, 63% of the patients completed therapy.

The study found that the treatment was safe but offered no benefit in preventing TB or improving survival. In the treatment group, three of 260 patients developed confirmed TB, compared with six of 257 in the placebo group.

"The impact of INH was so light that it was really not statistically significant," lead author Fred Gordon, MD, chief of infectious diseases at the VA Medical Center in Washington, DC, tells TB Monitor.

Gordon points out, however, that the incidence of TB in the placebo group was only 0.9 per 100 person years. That is a much lower rate than expected for this population, suggesting that the risk of acquiring new TB infections has decreased because of improved public health controls. The authors conclude that isoniazid therapy in this group of patients is unwarranted and that it should be limited to specific, high-risk situations in which people have recently come into contact with active TB patients.

"The population has so little TB that you really shouldn’t invest your time and effort in providing prophylaxis," Gordon says.

Indeed, the studies the CDC based its recommendation on were conducted nearly seven years ago and recruiting mostly New York City patients — ensuring the rate of TB infection in anergic patients was quite high. Today, after four years of declining TB cases, epidemiologists extrapolate from the general population that anergic patients also are at decreased risk of TB, says Daniel Chin, MD, MPH, chief of surveillance and epidemiology at the TB control branch of the California Department of Health Services in Sacramento.

"The populations of anergic patients who are infected are probably lower now," he says. "That’s maybe why the study isn’t showing that great a difference."

Published in the New England Journal of Medicine, the study comes on the heels of a similar study conducted on anergic HIV-patients in Spain. That study, published in the August issue of the Archives of Internal Medicine, found that preventive therapy for 12 months did decrease TB cases and increase survival. The authors recommended that HIV-positive patients receive daily INH therapy for 12 months.

Responding to the Spain study, Kenneth Castro, MD, director of the CCD’s Division for Tuberculosis Elimination, notes that because of the country’s high background rate of TB infection, anergic patients often go on to develop TB disease. A positive anergy test in U.S. patients doesn’t portend the same clinical course it does in other countries, even in high-incidence states like New York and New Jersey, TB experts note. Nonetheless, other TB researchers say U.S. physicians should still consider giving isoniazid to anergic patients because other data show anergics’ risk for developing active TB falls approximately midway between that of non-anergics and people who are PPD positive.

Another recent unpublished study of preventive therapy in high-risk anergic patients was presented at the American Thoracic Society meeting earlier this year. That study followed about 190 HIV-positive patients, many of whom had been incarcerated, lived in homeless shelters, and injected drugs. The patients had not completed preventive therapy for various reasons and were followed to see what their risk was for developing TB.

"Their progression rate over a follow-up time of over 27 months was extremely low," says Anita Barry, MD, TB control officer for the Boston Department of Health. "At least in our study it would not have been cost-effective to treat them."

As a result of these and other studies, the CDC is revising its guidelines on anergy testing and expects to publish them in September.

"We concluded that anergy testing, at least in the United States, was not very helpful in predicting which people would be at risk of tuberculosis," says Rick O’Brien, MD, chief of the CDC’s research and evaluation branch in the division of TB elimination. "Our revised guidelines are saying that anergy testing in program settings is no longer routinely recommended, but there are individual instances where anergy testing may be useful."

Clinicians who treated patients in areas at high risk for TB, particularly in New York, argue that INH preventive therapy is reasonable and will probably continue to recommend it, O’Brien says. The new guidelines don’t rule out the use of preventive therapy in anergic patients and will continue to recommend that any HIV-positive patient exposed to TB should receive preventive therapy, he tells TB Monitor.

"The guideline doesn’t come out explicitly as saying we shouldn’t give preventive therapy [to anergic patients]," he says, adding that a meeting in early September on TB prevention issues in HIV infection will consider that question more definitively.

Where preventive therapy may be useful

Gordon’s study raises the question of where, if not among the general population of anergic patients, preventive therapy is beneficial. The answer most experts give is in correctional facilities. But even in that setting the use of preventive therapy remains controversial.

The most resounding argument for preventive therapy in prisons comes from the experience of Arkansas prisons, where INH prophylaxis has been given to HIV-positive patients, regardless of TB status, since 1993. William Stead, MD, director of the Arkansas Health Department’s TB Program, adopted the policy based on the results of a primary prophylaxis study of calves in the 1950s. (See TB Monitor, October 1996, pp. 114-116).

Such a non-judicious use of preventive therapy is justified by a high rate of TB in the state’s prisons, says Kashef Ijaz, MD, MPH, a TB researcher at the Arkansas Department of Health. During the four-year period there were 46 TB cases and 300 PPD test conversions, yielding an average case rate of 129 per 100,000 and a conversion rate of 3,529 per 100,000.

"It’s very simple and easy to see that the prison is a high-risk setting for exposure," he says. "The folks who are HIV-positive are at added risk, so they are put on INH indefinitely."

An observational study of 62 HIV-positive patients found that only one of the 46 TB cases was the result of co-infection with HIV. In that case, the patient did not complete therapy because of side effects. Such a low rate of disease indicates that INH therapy is beneficial in such a setting, Ijaz says.

Despite the effectiveness found in the prisons, the Arkansas prison system is an exception among state prisons, where preventive therapy usually is not administered for both clinical and operational reasons.

At Rikers Island prison system, where 25% of women and 12% to 15% of men are infected with HIV, offering INH preventive therapy would create an overwhelming operational and financial burden, says Eran Bellin, MD, director of infectious diseases for the Montefiore Rikers Island Health Service in East Elmhurst, NY.

"If I commit to INH prophylaxis, what do I give up?" he asks. "The people who are known to be PPD positive? It’s a question of resources. I can’t even get INH reliably into prisoners known to be PPD positive."

Another barrier is the cost of determining who is anergic. Using a multiple antigen test is prohibitive for a large populations, and a single antigen test — Candida, for example — would result in an unacceptably high number of false positives, he says.

Because prisons have captive audiences, preventive therapy can be administered reliably, says Chin. Nonetheless, the issues of side effects, cost, and administrative burden make it unattractive to most prisons, including those in California.

"There are so many factors, you can’t just say everyone should take it," says Madim Khoury, MD, deputy director for healthcare policy for the California Department of Corrections. "Each case will have its own recommendations based on the stage of disease."

California corrections officials have discussed the possibility of INH prophylaxis in HIV-positive inmates but have decided against it, even in light of two TB outbreaks. In a 1995 outbreak at a unit holding HIV-positive prisoners, 14 inmates were infected after an HIV-positive prisoner developed active TB, most likely from a visit by his TB-infected spouse. In a second outbreak in 1996, another 14 HIV-positive patients became infected from an index patient who later died of TB. Despite these outbreaks, the cases of TB in state facilities has remained relatively low through other controls, such as mandatory PPD testing of inmates and staff, says Sarah Royce, MD, MPH, the state’s TB controller. In 1995, there were 13 reported cases of TB among HIV-positive prisoners, a rate of 160 per 100,000. Last year, only five or six cases were reported, she adds.

"I don’t think it’s pre-emptive INH that is driving the numbers down," she tells TB Monitor. "It’s probably because we are being more aggressive in TB control, such as testing upon entry, doing annual PPD — those kinds of things."


1. Gordin G, Matts J, Miller C, et al. A controlled trial of isoniazid in persons with anergy and HIV infection who are at high risk for tuberculosis. N Engl J Med 1997; 337:315-320.

2. Kashef I, Stead W. Primary prophylaxis in HIV-positive persons before exposure to tuberculosis. Presented at the American Lung Association’s 1997 International Conference, May 16-21, San Francisco.