At the 25th Conference on Retroviruses and Opportunistic Infections (CROI)
Diana Gibb, MD, professor of epidemiology, program leader of the Pediatric Program of trials and cohorts at the Medical Research Council Clinical Trials Unit at UCL, London, shared results from a REALITY trial
(ISRCTN43622374) which explored the impact of raltegravir intensification of first-line antiretroviral therapy (ART) on immune reconstitution inflammatory syndrome (IRIS).
Dr. Gibb shared how 12-week raltegravir (RAL)-intensified quadruple therapy resulted in significantly faster viral load (VL) declines through 24 weeks among HIV-infected adults and older children but did not reduce overall mortality or WHO 3/4 events compared to standard triple-drug antiretroviral therapy (ART).
“As we know, integrase inhibitors are replacing non-nucleoside reverse-transcriptase inhibitors (NNRTIs) in first-line and there has been concern that rapid viral load declines may increase rates of IRIS in those starting ART with very low CD4,” Dr. Gibb said. “In fact, at this conference last year, there were a number of presentations of observational studies which found higher rates of IRIS in those starting integrase inhibitor combinations.”
However, according to Dr. Gibb, there have not been previous trials which were powered on clinical endpoints in Africa.
In the trial, a total of 1,805 HIV-infected adults, adolescents, and children 5 years of age or older with CD4 less than 100, were randomized into 2 arms: those who started (2NRTI+NNRTI) ART with an additional 12 weeks of raltegravir and those who started standard of care ART (2NRTI+NNRTI alone). The follow-up was at 48 weeks. The trial sites included 8 centers across 4 different countries in Africa—Uganda, Kenya, Zimbabwe, and Malawi.
“Patients all had CD4s done 3-monthly after baseline to 48 weeks, but of note, viral loads were done retrospectively and were not available in real-time,” Dr. Gibb said. “The primary endpoint was at 24 weeks, and it was mortality.” In addition, there were 2 other randomizations, which investigated 12-weeks enhanced prophylaxis (published last year in the New England Journal of Medicine
) and 12-weeks supplementary food.
Dr. Gibb stressed that for the trial, they had a blinded Endpoint Review Committee (ERC) who did not have any knowledge of which arm patients were in. The Committee adjudicated serious adverse events, grade 3 or 4 events, WHO stage 3 and 4 events, as well as causes of death and compatibility with IRIS.
“As you know, IRIS is an atypical or exaggerated presentation of an opportunistic infection or tumor soon after ART initiation. Of note, the Committee did not have viral loads available in real-time, and the first CD4 was done 4 weeks after starting treatment,” Dr. Gibb reported. “Diagnostic testing was quite limited, so the ERC relied heavily on the description of the clinical presentation—things like the timing and evolution in relation to ART initiation. They also adjudicated the type of IRIS, such as TB or cryptococcal.”
Using “backward elimination,” the team also looked at predictors of the first fatal and non-fatal IRIS compatible event, treating death from other causes as a competing risk. Dr. Gibb discussed the baseline characteristics; about half of participants were male, and only 4% were children. “CD4 was very low, only 37 cells in this trial,” Dr. Gibb noted. “Three-quarters had viral load over 100,000, but despite that, nearly half were either asymptomatic or only had mild symptoms with stage 1 or 2.”
When looking at viral load, they found that 40% and 70% of the RAL group had viral load under 50 at 4 and 12 weeks respectively. “This was significantly different from the standard of care ART arm, where it was just over 13% and just over 50%” at 4 and 12 weeks,
Dr. Gibb said.
The researchers found that the all cause mortality at 24 and 48 weeks was similar in both arms. As for IRIS events, there were 36 in the RAL group and 31 in the standard ART group; they occurred around 4 weeks after ART initiation. “As expected, TB IRIS was most common, but almost identical in both arms, similarly with cryptococcal IRIS and others of known etiology,” she noted.
When looking at fatal and non-fatal IRIS-compatible events combined, the researchers noted 89 in the RAL-intensified arm compared with 86 in the standard ART arm. “Again, very similar in both arms, with TB IRIS being the most common,” Dr. Gibb said. Additionally, there was no difference seen between the 2 arms when it came to predictors. Current TB disease and ART initiation were predictors. Log10
viral load, current Cryptococcus at ART initiation, current steroid treatment at ART initiation, and gender, were not independent predictors.
However, in another randomization that looked at enhanced prophylaxis (isoniazid B6 given as a fixed dose combination, 12 weeks of fluconazole, 100 mg, 5 days of azithromycin, and a single-dose of albendazole), they found that the enhanced prophylaxis package did
have an effect on IRIS events. “A big difference between the enhanced prophylaxis with only 67 virus-compatible events, and 108 in the standard arm. And this mirrors the difference that we saw of 27% reduction in all cause mortality with this intervention,” Dr. Gibbs stressed.
Dr. Gibbs concluded that the trial did not find any evidence that supported higher rates of IRIS despite faster viral load decline on ART with 12 weeks of raltegravir.
“We feel that our data provide reassurance that the current move to first-line integrase inhibitor-based ART will not increase IRIS. You have similar viral load reductions with all integrase inhibitors. And so, feel that the results can be extrapolated to other integrase inhibitors including dolutegravir,” she concluded. “CD4 before ART start is required to identify those at high-risk of both death and IRIS in whom this prophylaxis package would be helpful.”
To stay informed on the latest in infectious disease news and developments, please sign up for our weekly newsletter.