Integrase Inhibitors and Weight Gain: A Top HIV Clinical Development of 2018

Integrase inhibitors have revolutionized HIV treatment, as a glance at HIV treatment guidelines over the past few years makes clear. And with good reason: They're well tolerated, have a high barrier to resistance, and are potent as all get out. But they are not without their downsides.

Neuro-psych effects are rare, but appear to be a family curse. And there is the issue of neural tube defects, at least for dolutegravir (Tivicay, DTG). These drawbacks have not yet limited the use of these drugs significantly.

This past year, however, a crescendo of data suggest that integrase inhibitors may promote gains in weight beyond those seen with other antiretrovirals, and that could become a threat. Besides being unwelcomed by many patients, excess weight carries health risks. Additionally, in the U.S., most starting HIV therapy are already at least overweight -- especially people in the South.

The "do they or don't they" question of integrase inhibitors and weight gain has been looked at in a few studies presented/published this year -- almost all retrospective. One of the first is a report from Vanderbilt University Medical Center, located in Nashville, Tennessee, the heart of the U.S. obesity belt. There, 136 patients who switched from tenofovir disoproxil fumarate/emtricitabine/efavirenz (Atripla, TDF/FTC/EFV) to an integrase inhibitor experienced a mean gain of 2.9 kg, which was significantly greater than a mean 0.7 kg gain in the 34 changing to a protease inhibitor and a 0.9 kg average gain among the 325 remaining on TDF/FTC/EFV.

Studies of treatment-naive antiretroviral initiators provide a more level playing field to assess differential effects on weight. A retrospective analysis of 3,208 patients at Parkland Hospital in Dallas, Texas, who started HIV therapy between 2009 and 2017 looked at the changes in their body mass indexes (BMI) over time. Again, integrase inhibitor therapy (presumably mostly raltegravir [Isentress, RAL]) was associated with greater BMI increases than treatment with other antiretroviral classes; this was more likely to be seen in black and Hispanic patients than in white patients. Further, both integrase inhibitors and protease inhibitors led to higher BMI gains in women than in men.

At the 2018 International Workshop on Comorbidities and Adverse Drug Reactions in HIV, the NA-ACCORD cohort "weighed" in. Over 21,800 individuals in the U.S. and Canada who started HIV therapy between 2007 to 2015 -- almost 4,100 of whom began with an integrase inhibitor in the mix (51% RAL, 37% elvitegravir [EVG], and 12% DTG) -- were included in analyses that looked at actual and predicted changes in weight over time. Overall, weight increased -- quite a bit -- with all regimens, regardless of composition.

Those initiating with a non-nucleoside saw a median 4.1 kg gain over the first five years of treatment, compared with 5.0 kg with a protease inhibitor and 5.8 kg with an integrase inhibitor. In pairwise comparisons, treatment with either a protease inhibitor or an integrase inhibitor was associated with significantly greater increases in five-year weight gain compared with treatment with a non-nucleoside, with no discernable difference between integrase and protease inhibitors. When examining two-year changes in weight among the integrase inhibitors, EVG was linked to less average weight gain (3.4 kg) than RAL (5.4 kg) or DTG (5.6 kg). In models predicting changes in weight over time, integrase inhibitors were projected to produce the greatest increase in weight, with DTG having a greater effect than other drugs of this class.

The observed NA-ACCORD data are congruent with the findings of a body composition sub-study of the U.S. AIDS Clinical Trials Group (ACTG) study A5257, in which treatment-naïve patients were randomized to start TDF/FTC (Truvada) plus RAL versus atazanavir (Reyataz)/ritonavir (Norvir) versus darunavir (Prezista)/ritonavir. Here, no major differences in regional fat, lean mass, or BMI over 96 weeks of follow-up were observed between those assigned RAL and those assigned to the protease inhibitor arms. In this study, there was no non-nucleoside arm.

The Bottom Line

During the first two decades of the U.S. HIV epidemic, when being gaunt was a commonplace and stigmatizing symptom of advanced disease, most patients appreciated the gain in weight that accompanied the initiation of HIV therapy. Putting on pounds was, and to some extent remains, a welcome indicator of the effectiveness of medication and a reassuring sign of a return to health.

However, once potent therapies came along to stomp viral replication and stop opportunistic conditions, obesity replaced wasting as the major body issue for people living with HIV. Out with growth hormone, and in with food pyramids and low-carb diets! Linked to diabetes mellitus, cardiovascular disease, and even frailty, excessive weight in people living with HIV has become a target for intervention. So, a love-handle-expanding antiretroviral is not what the world needs now.

Collectively, the studies described above may sketch a picture of less weight increase with non-nucleosides compared with other antiretroviral classes. However, this is not an easy thing to study. The majority of non-nucleoside use is with EFV, a medication taken on an empty stomach. Switching from a regimen that has food restrictions to one that does not could, conceivably, have an impact on caloric intake and, therefore, weight. The same effect could be operative in the treatment-naive -- which may account for the finding in some studies of little difference between protease and integrase inhibitors.

However, some of the reports do signal a role for integrase inhibitors in leading to relatively greater weight increases over time, perhaps preferentially in certain groups. It is hard to shake a label: For instance, although true evidence of a direct effect of protease inhibitors on increases in visceral fat does not exist, many still believe in Crix belly and protease paunch. The same could happen with integrase inhibitors if well-conducted studies are not conducted and presented clearly.

At a minimum, the differential effects of integrase inhibitors on weight can be addressed by the grand head-to-head naive and switch trials of these agents -- including studies that have already been conducted. Databases need to be dusted off, and those carefully conducted weights obtained at study visits compared across treatment arms. Future studies should include weight change as an outcome. Only by baring all will we know what is truly happening, and then begin to understand why.

David Alain Wohl, M.D., is a professor of medicine in the Division of Infectious Diseases at the University of North Carolina (UNC). He is site leader of the UNC AIDS Clinical Trials Unit at Chapel Hill, director of the North Carolina AIDS Education and Training Center (AETC), and co-director of HIV services for the North Carolina state prison system. In 2014, he became co-director of the UNC-Duke Clinical RM Ebola Response Consortium.