Another reason, as Dr Castro mentioned, is the potential for
reinfection, particularly for people who are more susceptible to TB (due to
poor health, nutrition and HIV), in places where they are more likely to be
repeatedly re-exposed to TB. In such settings, anything short of continuous
therapy might appear to be wrong-headed.
Or is it?
Although clearly people do become reinfected, even in high-burden
settings, the risk varies from setting to setting, person to person, difference
in CD4 cell count and tuberculin skin test status. Data to support the benefits
of continuous IPT as being worth the
costs (side effects, financial and other health system resources) are, at
present, far from unequivocal.
Only a few studies have looked at the durability of TB
preventive therapy, with variable results. Most of the older IPT studies in
high TB burden countries suggest that HIV-positive people who are put on six
months of IPT continue to have a reduced incidence of TB compared to those who
were untreated for awhile after discontinuing treatment, though the effect
diminishes gradually over time.23 One
study in Zambia suggested that the protective effect could last up to three
years.24
while a study from Uganda found the effect only lasted one year.25
These studies have to be interpreted with caution however as there was
substantial loss to follow-up.26
On the basis of these data, one of the ‘old questions’ about
IPT was how often HIV-positive people in regions with a higher risk of TB
exposure should have to repeat a course.
This changed somewhat when the
BOTUSA study results came out. Again, everyone in that study got six months of
IPT and was then randomised to either placebo (stopping IPT) or continuing IPT
out to 36 months. Much to their surprise, the investigators found that the
protection provided by the short course of IPT in their study wore off more
quickly than they expected in those randomised to placebo — within six months of
completing treatment. The researchers theorised this was probably because
people were at such a high risk of re-exposure to, and reinfection with, TB in
Botswana. They may be right, or at least partly right, but there are a number
of possible explanations. For instance, in the one of the studies mentioned
above. Quigley et al wrote that “the diminishing effect of preventive therapy
over time in HIV-infected Zambian adults may be explained by the high risk of
new infection or by inadequate
sterilization of dormant tubercle bacilli in the absence of immunity.”
It is odd that the BOTUSA
investigators don’t consider that this last idea, that six months of IPT
doesn’t entirely clear the infection in all individuals, is at least part of
the explanation for what happened in their study. After all, most studies and a
meta-analysis suggest that in tuberculin skin test positive patients, six months
of IPT reduces the risk of TB by only around 60%. The remaining cases will
still happen — without more preventive therapy.
Instead, the BOTUSA
investigators give two primary reasons why they think reinfection is what is
occurring. One, they note that
there have been high rates of clustering and reinfection of successfully
treated cases of tuberculosis reported in Botswana — and there is no
denying this. Then they claim that short-courses of isoniazid preventive treatment have a durable
protective effect for people with HIV and positive tuberculin skin tests in
areas of low tuberculosis incidence.” But the study that they cite to support
their argument involved 12 months of isoniazid — not six months — and 12 months
of IPT may have a better chance of clearing the bacilli if there is good
adherence.
This was very clearly suggested
in an on-treatment analysis of the pivotal IPT study conducted by the
International Union Against Tuberculosis Committee on Prophylaxis including
over 28,000 people in the late 1970s.27
The preventive effect became evident after the first 12 weeks on treatment with
a 31% reduction in incidence of active disease, but the effect increased to 69%
after 24 weeks of IPT. But in adherent patients, the benefit increased to over 93% at 52 weeks.
People may be putting too much
faith in early studies of IPT in people living with HIV that found there were
no differences in efficacy between six versus 12 months in HIV-positive
patients (in intent to treat analyses). This could be due to the rampant
adherence problems or other issues. It is difficult for this writer to ignore
the weight of the Union’s study. Furthermore, it seems rather paradoxical that
the WHO guidelines review committee concluded that twelve months isn’t more
effective than 6 months, and yet gave a conditional recommendation of
continuous IPT, on the basis of a study that reported the loss of benefit so
shortly after discontinuing six months of IPT.
One of the BOTUSA study’s
findings undermines their argument somewhat: continued IPT only reduced TB
incidence in people who, when they first entered the study, were TST positive
(the clearest measure of having a pre-existing latent TB infection). Note, that
only 23% of the study participants were TST positive and therefore showing
clear evidence of previous exposure to TB.
If ongoing exposure to TB is
so widespread, why didn’t the study see a big increase in the number of TB
cases in the rest of the study population on placebo? One would have expected to see some slight trend of difference in the
remaining participants, who made up roughly 77% of the study population. The
fact that breakthrough cases only occurred in someone who was TST positive at
baseline is suggestive of an infection that wasn’t cleared, although it is
possible that people who were already latently infected (as the TST shows) are
somehow more susceptible to becoming infected again.
But there is limited evidence,
such as the study previously mentioned in Uganda —that the benefit of 6 months
of IPT is relatively short-lived. However, in the same study, the benefit of
shorter more potent combination regimens is sustained out to three years.28
Similarly, in the TRC study in Chennai, there did not appear to be any
statistically significant rebound of TB after the six-month
ethambutol/isoniazid regimen concluded, (though the event rate was higher than
in those randomised to three years of IPT).
If it does take a longer course
of isoniazid to clear the infection (if indeed it is possible some of the more
advanced people living with HIV), then it may be possible that a more potent
regimen may better clear latent TB in less time. If this is true, then there
will be less loss of effect — a longer window period free from the risk of
TB — even in the higher burden settings.
This will need to be demonstrated by further clinical research but it is
certainly worth investigating.
If rifapentine/isoniazid, or
another short course regimen is effective and safe, then the question of when
to give people living with HIV a repeat course may become pertinent again. Dr
Martinson doesn’t believe it will be necessary.
“In HIV-positive people the "real sterilising"
effect is getting ARVs and restoring immunity soon after receipt of
preventive treatment,” he said. “I think few people would be diagnosed soon
enough, with a high CD4 count, to receive a short course twice prior to ART
initiation. ARVs are the key. IPT has no impact on mortality and has a low
durability — get people on ARVs!”