| |
Weight Change Following Antiretroviral Therapy Switch in People With Viral Suppression: Pooled Data from Randomized Clinical Trials
|
| |
| |
Download the PDF here
Clinical Infectious Diseases 14 May 2021 - Kristine M. Erlandson,1 Christoph C. Carter,2 Kathleen Melbourne,2 Todd T. Brown,3 Cal Cohen,2 Moupali Das,2 Stefan Esser,4 Hailin Huang,2 John R. Koethe,5 Hal Martin,2 Grace A. McComsey,6,7 Chloe Orkin,8 Frank A. Post,9 Jürgen K. Rockstroh,10 Paul E. Sax,11 Hans-Jürgen Stellbrink,12 Laura Waters,13 Xuelian Wei,2 and Jordan E. Lake14
1Department of Medicine, University of Colorado Anschutz Medical Campus, Aurora, Colorado, USA; 2Gilead Sciences, Inc, Foster City, California, USA; 3Department of Medicine, Johns Hopkins University, Baltimore, Maryland, USA; 4Department of Dermatology, University Hospital Essen, Essen, Germany; 5Department of Medicine, Vanderbilt University Medical Center, Nashville, Tennessee, USA; 6University Hospitals Health System, Cleveland, Ohio, USA; 7Departments of Medicine and Pediatrics, Case Western Reserve University, Cleveland, Ohio, USA; 8Barts Health National Health Service Trust, London, United Kingdom; 9King's College Hospital National Health Service Foundation Trust, London, United Kingdom; 10Department of Medicine, University Hospital Bonn, Bonn, Germany; 11Department of Medicine, Brigham and Women's Hospital and Harvard Medicine School, Boston, Massachusetts, USA; 12ICH Study Center, Hamburg, Germany; 13Mortimer Market Center, London, United Kingdom; and 14Department of Medicine, University of Texas Health Science Center, Houston, Texas, USA
"In summary, our results demonstrate that modest weight gain is common after ART switch and is correlated more strongly with baseline regimen, especially switch off of TDF or EFV, than with sex-, race-, or HIV-related factors. It remains uncertain whether this is due to the loss of a weight suppressive effect of prior regimens or a weight gain effect of the newer regimen. A better understanding of the underlying biological mechanisms and the clinical implications is needed to fully understand these observations. Close monitoring of weight and counseling to maintain a healthy diet and remain physically active, as well as optimize other lifestyle factors, is imperative for all patients on ART [40].
In this pooled analysis of 12 randomized studies of ART switch, we found that weight gain occurred in both switch and SBR participants, with the magnitude of weight gain generally greater with switches to newer regimens, consistent with observations from other ART-naive and switch studies [2, 4-11]. Additionally, baseline ART regimen was a significant predictor of weight gain after switch. For example, in participants who switched to EVG/c, switch from EFV was associated with weight gain, switch from a PI or NVP to EVG/c was weight-neutral, and switch from RPV to EVG/c was associated with weight loss. Among the NRTIs, switch from TDF to TAF was associated with greater weight gain than switch from ABC to TAF. Younger age and lower baseline BMI were associated with weight gain, while race, ethnicity, sex, and CD4 count were not."
Abstract
Background
We sought to identify factors associated with weight gain in randomized clinical trials of antiretroviral therapy (ART) switch.
Methods
We explored the effects of demographic factors, clinical characteristics, and ART on weight gain in a pooled analysis of 12 prospective clinical trials, wherein virologically suppressed people living with human immunodeficiency virus (PWH) were randomized to switch or remain on a stable baseline regimen (SBR).
Results
Both PWH randomized to switch ART (n = 4166) and those remaining on SBR (n = 3150) gained weight. Median weight gain was greater in those who switched (1.6 kg, interquartile range [IQR], -.05 to 4.0 vs 0.4 kg, [IQR], -1.8 to 2.4 at 48 weeks, P < .0001), with most weight gain occurring in the first 24 weeks after switch.
Among baseline demographic and clinical characteristics, only younger age and lower baseline body mass index were associated with any or ≥10% weight gain.
By week 48, 4.6% gained ≥10% weight (6.4% of switch and 2.2% of SBR),
the greatest risk was with switch from efavirenz (EFV) to rilpivirine (RPV) or elvitegravir/cobicistat and switch from tenofovir disoproxil fumarate (TDF) to tenofovir alafenamide (TAF).
Switch from abacavir to TAF was associated with less weight gain than switch from TDF to TAF and was not associated with increased risk for ≥10% weight gain.
Conclusions
Moderate weight gain after ART switch was common and usually plateaued by 48 weeks. Baseline ART was a predictor of post-switch weight gain; participants who switched off of EFV and TDF had the greatest weight gain. The biological mechanisms that underlie the differential effects of switching ART agents on weight and associated clinical implications require further study.
DISCUSSION
In this pooled analysis of 12 randomized studies of ART switch, we found that weight gain occurred in both switch and SBR participants, with the magnitude of weight gain generally greater with switches to newer regimens, consistent with observations from other ART-naive and switch studies [2, 4-11]. Additionally, baseline ART regimen was a significant predictor of weight gain after switch. For example, in participants who switched to EVG/c, switch from EFV was associated with weight gain, switch from a PI or NVP to EVG/c was weight-neutral, and switch from RPV to EVG/c was associated with weight loss. Among the NRTIs, switch from TDF to TAF was associated with greater weight gain than switch from ABC to TAF. Younger age and lower baseline BMI were associated with weight gain, while race, ethnicity, sex, and CD4 count were not.
Weight gain after switch from TDF to TAF was observed in this study and others [11, 28], but it remains unknown whether these observations result from removal of a weight-suppressive effect of TDF, TAF-associated weight increase, or a combination of both [11, 28]. Preexposure prophylaxis (PrEP) trials suggest an initial weight-suppressive effect of TDF in a setting where confounding effects of HIV and other agents are absent. In iPrEX, participants who took emtricitabine (F)/TDF had initial weight loss followed by a weight gain trajectory similar to the placebo arm. A similar pattern was observed in DISCOVER and HPTN 083, where the F/TDF arm exhibited initial weight loss, followed by a weight gain trajectory similar to the F/TAF and cabotegravir (CAB) arms, resulting in 1-1.5 kg greater weight gain in the F/TAF and CAB arms [29-31]. HPTN 077 found similar weight gain between CAB and placebo, suggesting that initial TDF weight suppression may have contributed to the weight differences in HPTN 083 [32]. The potential for TDF to suppress weight gain is also supported by the GEMINI 1 and 2 studies, where treatment-naive PWH randomized to DTG + lamivudine (3TC) gained more weight than those taking DTG + F/TDF (3.7 vs 2.4 kg at week 144) [33]. Additionally, switch from TAF-containing regimens to DTG/3TC led to similar weight gain as staying on TAF-containing ART in the TANGO study [34], although these findings are confounded by a high proportion of study participants also switching from non-DTG-containing regimens. The TDF weight-suppressive effect does not exclude TAF-associated weight gain and does not entirely explain the weight gain differences among NRTIs we observed in our study; switch from ABC to TAF was also associated with smaller but statistically significant weight gain. ABC/3TC/DTG was associated with more frequent gastrointestinal AEs than BIC/F/TAF in study GS-US-380-1844 (included in our analysis), but the contribution of this observation to weight differences is unknown [24]. Further research including mechanistic studies, healthy volunteer studies, and single-variable ART switch studies may help disentangle these complex observations.
Among the NNRTIs, we observed significant weight gain associated with switch from EFV to RPV; switch off EFV was the only third-agent switch associated with ≥10% weight gain. As with TDF, this observation may be explained, at least in part, by a weight-suppressive effect of EFV. In an observational cohort study, slow metabolizers gained a median 2.0 kg after switch to INSTI, while those with normal EFV metabolizer genotypes gained 0.1 kg after switch [35]. Similar findings were reported from the ADVANCE study, which compared randomized EFV + F/TDF to DTG + F/TDF or F/TAF in ART-naive PWH in southern Africa. While greater weight gain was observed in the DTG-containing arms, a pharmacogenomic substudy found that participants with a slow EFV metabolizer genotype lost weight, while those with an extensive EFV metabolizer genotype gained a similar amount of weight as those in the DTG + F/TDF arm [36]. Further studies focusing on pharmacogenomics of ART metabolism and individual genetic factors associated with weight gain may provide additional insights into the mechanisms of ART-associated weight gain.
Increasing attention has recently been paid to the association between INSTIs and weight gain, an issue of growing relevance as the second-generation INSTIs (DTG and BIC) are first-line recommended therapy [34]. Our findings in the pooled switch data are consistent with the existing literature [8, 34]: weight gain with switch to DTG and BIC is greater than that observed with switch to EVG/c. Potential mechanisms to explain the weight gain observed with newer INSTIs are a topic of intensive study (reviewed in [34]).
An important question is whether the weight changes following ART switch stabilize or continue to increase over time. Across our pooled data, we observed a period of accelerated weight gain that lasted approximately 24 weeks followed by a gradual plateau, although data past 48 weeks were limited. The proportion of those who gained ≥10% continued to increase to week 96 (Figure 2). We and others have observed a similar plateau of weight gain in treatment-naïve participants [4, 5, 37] or following TAF switch [28]. In contrast, other studies showed continued weight gain 2 or more years post-switch [7, 38].
In contrast to our prior pooled analysis in ART-naive persons [5], we did not observe an association between weight gain and Black race or female sex; the reasons for this are unclear. Our pooled dataset included approximately 1600 Black participants (600 women), suggesting the analysis was not underpowered. Sociogeographic characteristics of Black participants could have contributed as well as regional differences in the obesity epidemic. For instance, Black females in the ADVANCE study resided in southern Africa, which has the highest prevalence of female obesity in the world, while Black females in GS-US-380-1961 (who constitute nearly one-third of Black females in our pooled dataset) resided primarily in East Africa, where obesity rates are markedly lower [39]. Similarly, a study of ART switch in Nigeria did not find greater weight gain among women [37].
Similarly, we did not observe an association between baseline CD4 count and weight gain. This is most likely because substantial immune recovery had already occurred for most participants during the pre-ART switch period. However, we cannot exclude the possibility of lingering metabolic impacts of HIV infection that could be reversed with longer ART duration or be differentially impacted by certain ART regimens.
We found no clinically significant changes in metabolic parameters following ≥10% weight increase, albeit with short follow-up. The intrinsic effects of some ART drugs on lipid levels, which we avoided by resetting the lipid baseline to the time of 10% weight gain, makes the impact of weight change on lipids challenging to assess. Long-term follow-up studies are needed to assess metabolic changes associated with ART-associated weight gain.
Our pooled analyses are strengthened by randomized, controlled data from more than 7300 persons enrolled in switch studies, with diversity in geographic location, sex, age, race, and ethnicity. However, several important host factors that may have impacted weight gain were not assessed in these clinical trials, including diet, exercise, alcohol use and smoking, comorbidities including mood disorders, and comedications. Additionally, randomized, clinical trials preclude the assessment of preswitch weight trajectories, lack general population controls, and tend to enroll participants who are younger and healthier than observational cohorts and thus may not be reflective of weight changes in clinical practice. The regimen changes were limited to those studied in these clinical trials and thus did not include other switch regimens of interest (ie, RAL to DTG) and fewer switches to ABC than TAF. Weight and metabolic outcomes were not prespecified in any trial, and imaging was not available to explore distribution of weight gain.
In summary, our results demonstrate that modest weight gain is common after ART switch and is correlated more strongly with baseline regimen, especially switch off of TDF or EFV, than with sex-, race-, or HIV-related factors. It remains uncertain whether this is due to the loss of a weight suppressive effect of prior regimens or a weight gain effect of the newer regimen. A better understanding of the underlying biological mechanisms and the clinical implications is needed to fully understand these observations. Close monitoring of weight and counseling to maintain a healthy diet and remain physically active, as well as optimize other lifestyle factors, is imperative for all patients on ART [40].
|
|
| |
| |
|
|
|
