Weight Gain When Switching From Efavirenz to an Integrase Inhibitor
Virally suppressed people who swapped an efavirenz (Sustiva, Stocrin) regimen for one based on an integrase inhibitor (INI) gained significantly more weight over 18 months than those who stayed with efavirenz or switched to a protease inhibitor (PI), according to results of a single-center retrospective analysis. People switching to single-tablet dolutegravir/abacavir/lamivudine (DTG/ABC/3TC, Triumeq) gained the most weight over 18 months, an average of 5.3 kg.
Clinicians at Nashville's Vanderbilt University HIV clinic noted weight gains in patients taking single-tablet efavirenz/tenofovir/emtricitabine (EFV/TDF/FTC, Atripla). Because rising weight contributes to risk of cardiovascular and metabolic diseases, Vanderbilt researchers compared weight changes in people trading EFV/TDF/3TC for an INI or PI regimen and those staying with EFV/TDF/3TC.
The Vanderbilt team determined antiretroviral regimen, weight change, and other variables in adults with a viral load below 1,000 copies/mL who had taken fixed-dose EFV/TDF/FTC for at least two years. Over the next 18 months, participants either continued that regimen or switched to an INI- or PI-based combination. No one had a viral load rebound to 1,000 copies/mL or higher during follow-up, and no one switched drugs again during that time. The investigators used linear mixed effects models to assess differences in 18-month weight change in the three groups. These models adjusted for age, sex, race, antiretroviral duration, baseline CD4 count, and baseline weight.
The study involved 495 adults, of whom 325 stayed with EFV/TDF/FTC, 136 switched to an INI regimen, and 34 switched to a PI. While 14% of the EFV and INI groups were women, 29% of the PI group were women. Median age and weight at the switch were similar among the three groups (about 39 years and 80 kg). Median CD4 count at the switch was higher in the INI group (662 cells/mm3) than in the PI group (516 cells/mm3) or the EFV/TDF/FTC group (576 cells/mm3) (P = .03).
After adjustment for confounding variables, people who switched to an INI regimen gained significantly more weight through 18 months than those who stayed with EFV/TDF/FTC (mean 2.9 versus 0.9 kg, P = .003). Eighteen-month weight change did not differ significantly between people maintaining EFV/TDF/FTC and those changing to a PI (0.7 kg, P = .81). Subgroup analysis found the greatest 18-month weight gain (mean 5.3 kg) in the 58 people who traded EFV/TDF/FTC for DTG/ABC/3TC (P = .001). The increase with DTG/ABC/3TC was nonsignificantly higher than the mean 2.8-kg gain in the 78 people who switched to an elvitegravir (Vitekta) or raltegravir (Isentress) regimen (P = .19).
Mean HbA1c, a diabetes marker, rose from 6.4% to 6.9% over 18 months in people who switched to an INI regimen while falling from 6.4% to 6.0% in those who maintained EFV/TDF/FTC, but this difference lacked statistical significance (P = .30).
Clinical implications of these findings are not readily apparent. A 2.9-kg weight gain in an 80-kg person amounts to a less than a 4% change. The 5.3-kg gain with DTG/ABC/3TC, representing a 7% jump in an 80-kg person, seems more concerning. But the analysis provides no clues about whether weight gains continue past 18 months. And the authors observe that studies of people starting a first-line INI regimen have yielded inconsistent findings on weight change. In previously untreated people, the STARTMRK trial found lower combined trunk and limb fat gain with a raltegravir regimen than with an efavirenz regimen (19% versus 31%).
A 10-year DXA-scan study of 839 women and 1759 men attending an HIV clinic in Italy independently linked INI use to an average lean mass loss of 0.6 g (0.0006 kg) per year. Lean mass loss with overall weight gain would seem a troubling development. The authors of the switch study do not speculate on mechanisms of weight gain with INI regimens.