Increased insulin resistance following switch from efavirenz to cobicistat-boosted elvitegravir

Abstract

Background

Integrase strand transfer inhibitors (INSTIs) have been associated with excess weight gain in people living with HIV compared to other antiretroviral agents. The mechanisms that underlie these effects are not well defined. Thus, we aimed to examine the effects of switching to INSTI-containing regimens on clinical metabolic parameters.

Setting

A secondary analysis of a prospective cohort study in which people living with HIV on a stable efavirenz-based regimen were switched to a cobicistat-boosted elvitegravir or raltegravir-containing regimen. Participants remained on the NRTI backbone of tenofovir disoproxil fumarate and emtricitabine.

Methods

Frozen plasma samples from 19 participants were used to determine concentrations of leptin, adiponectin, insulin and lactate at baseline and 8 weeks post-switch. Fasting lipids and blood glucose not reported in the initial study were obtained to examine metabolic changes. Anthropometric data including height and weight were abstracted from the medical record.

Results

Participants switched from efavirenz to cobicistat-boosted elvitegravir without change in tenofovir disoproxil fumarate/emtricitabine backbone showed a 20% increase in HOMA-IR after 8 weeks (1.84 vs 2.24, p < .05), due mostly to increases in fasting insulin. This increase occurred independent of weight gain in the cohort as whole (83.4 vs 85.9 kg, pre vs post, p = .04), but was linked to increases in circulating lactate.

Conclusions

Participants switched to an INSTI-based regimen tended to gain weight, and those switched to cobicistat-boosted elvitegravir had increases in markers of insulin resistance and elevation in plasma lactic acid compared to raltegravir, suggesting that elvitegravir may promote metabolic perturbations in people living with HIV.

Keywords

integrase inhibitors HIV antiretrovirals weight gain insulin resistance

Introduction

Increasing reports over the past decade of excess weight gain associated with integrase strand transfer inhibitors (INSTIs) compared to other antiretroviral therapies (ART) have raised concern over the possible metabolic consequences of this class of antiretroviral therapy to treat people with HIV infection.¹ While INSTI-based regimens are guideline recommended first-line treatments for people living with HIV-1 (PLWH) due to superior efficacy, potency and high barriers to resistance, the possibility of associated weight gain and cardiometabolic risk raises concern and especially in the high burden of these comorbidities in the population ageing with HIV-1, demands further understanding.²

INSTIs, in particular dolutegravir, have been associated with approximately a 2 kg increase in weight after 18 months, corresponding to 0.5–1 kg/m² increase in body mass index among PLWH switching to a new regimen containing an INSTI.³ In the general population excess weight is a risk factor for several cardiometabolic comorbidities, including type 2 diabetes mellitus.⁴ However, the precise relationship between INSTIs, weight gain and diabetes risk is not yet understood despite data from observational or retrospective studies suggesting an interaction. An increase in insulin resistance related to INSTI-emergent weight gain has been described in some cohorts but not all.^5–8 A strong signal of increased insulin resistance has been linked to switching off efavirenz (EFV)-containing regimens; particularly among slow metabolizers of this non-nucleoside reverse transcriptase inhibitor,⁹ which is thought in itself to be weight suppressive.¹⁰

A major confounder across many observational studies describing INSTI-emergent weight gain has been the simultaneous switch from tenofovir disoproxil fumarate (TDF) to a tenofovir alafenamide (TAF) backbone that has occurred alongside the switch to INSTIs from other anchor agents.¹¹ TDF is a nucleoside reverse transcriptase inhibitor commonly part of combined ART regimen and has been observed to have weight suppressive effects compared to its pro-drug counterpart. Individuals who switch from TDF to TAF due to renal toxicity have also been observed to gain weight independent of other changes in medication regimen.¹²

In this study, we sought to leverage clinical data and biological specimens from a previous prospective cohort study of 20 participants investigating the impact of switch off EFV to an INSTI-containing regimen, either cobicistat-boosted elvitegravir or Raltegravir, on neuropsychologic testing, neurometabolites and systemic inflammation in asymptomatic, virologically suppressed individuals.¹³ We used this existing cohort where participants remained on TDF despite switching the anchor drug from EFV to an INSTI to evaluate for effect of switching to INSTI only, with no other change in the backbone, on change in weight/BMI and associated metabolic parameters, including glycaemic and lipid measures, and correlate them to pre-switch EFV metabolites.

Methods

We used clinical data and specimens from a prior prospective cohort study investigating the impact of switching off efavirenz-containing antiretroviral regimen to INSTI (raltegravir (RAL) or cobicistat-boosted elvitegravir (EVG) while remaining on FTC/TDF backbone on immunologic and neurologic markers on asymptomatic, virally suppressed PLWH (defined as <200 copies/mL) (NCT01978743, NCT01929759).¹³ A total of 19 study participants who were virally suppressed on EFV + TDF/FTC had pre-switch blood samples collected and frozen. Ten participants were switched to RAL and TDF/FTC while ten to EVG/c and TDF/TFC for completion of 8 weeks of treatment, at which time post-switch samples were again collected. This study used pre- and post-switch banked plasma samples from this cohort for further analyses, with additional data collection and analyses approved by the Mass General Brigham IRB.

For this study, we obtained additional clinical data from participants’ medical records, including recorded weight at the date closest to the pre-switch and end of study visit approximately 8 weeks later. Stored frozen plasma from participants from fasting blood draws was thawed, and insulin, adiponectin, and leptin were measured using ELISAs (R&D systems). Plasma lactate was determined using a fluorometric metabolite assay. HOMA-IR was calculated from fasting levels of glucose and insulin using the formula: $\frac{Insulin (pM) xGlucose (mg \cdot {dL}^{‐ 1})}{430}$ .^14,15 A 2-tailed paired t-test was used to compare pre- and post-switch for both combined and individual treatment groups.

Results

Demographic characteristics of the study cohort are detailed in our previous work, and additional characteristics are detailed in Table 1 according to treatment arm.¹³ While we were unable to retrieve complete BMI data for all subjects (n = 13), of those with available weight and height data neither baseline BMI nor weight change (2.0 ± 2.7 kg RAL [p = .09 vs baseline] vs 1.1 ± 2.3 EVG/c [p = .29 vs baseline]) were different between treatment arms during those 8 weeks. However, weight increased in the cohort as a whole (83.3 kg vs 85.9 kg, p = .04) within the 8 weeks switching off EFV to an INSTI. HOMA-IR tended to increase among all participants (1.78 vs 2.29, p = .08, n = 19) after switch to INSTI. Notably, this effect was predominantly driven by those who were switched to EVG/c (1.84 vs 2.24, p = .03 n = 9), whereas those switched to a RAL-containing regimen did not have significant change (1.98 vs 1.96, p = .97, n = 10 Figure 1(A)). This increase in HOMA-IR was largely due to increases in fasting insulin, not glucose (Figure 1(B) and (C)). Notably, circulating lactate increased in those switched to EVG/c (3.77 mM vs 4.75 mM, p = .04, but did not in those switched to RAL (3.72 mM vs 3.77 mM). Total and LDL cholesterol showed decreased levels in the combined population (190 mg/dL ± 33 vs 175 mg/dL ± 23, p = .003 total, 113 mg/dL ± 28 vs 100 ± 22, p = .02 LDL), with no difference between INSTI treatment groups at baseline changed at 8 weeks follow-up. Despite changes in total and LDL cholesterol levels after switch to INSTIs, circulating levels of the adipokines leptin and adiponectin were not different pre- versus post-switch in either group (Figure 1(D) and (E)), or as a whole.

Table 1.

Characteristics pre- and post-switch of participants by treatment arm.

	EVG/c/FTC/TDF^a		RAL + FTC/TDF		p-value RAL vs EVG/c
	Pre	Post	Pre	Post	Pre	Post	Difference
No. Participants	9		10		–
Age in years^b	48.3 ± 9.6		48.0 ± 9.9		0.7
Male gender, n (%)	8 (89)		6 (60)		N/A
Race
Black, n (%)	3 (33)		9 (90)		N/A
White, n (%)	5 (55)		1 (10)		N/A
Unknown, n (%)	1 (11)		0 (0)		N/A
Ethnicity
Non-Hispanic, n (%)	6 (67)		10 (100)		N/A
Hispanic, n (%)	3 (33)		0 (0)		N/A
	Pre	Post	Pre	Post	Pre	Post	Difference
CD4 T Cell (cells/mm³)	649 ± 184	716 ± 334	709 ± 280	766 ± 399	0.36	0.43	0.81
HIV-1 RNA <20 copies/mL, n, (%)	8 (89)	9 (100)	8 (80)	9 (90)	N/A
No. with BMI (%)	7 (78)	6 (67%)	7 (70)	7 (70)	N/A
BMI (kg/m^2)	28.1 ± 4.0	28.3 ± 3.9	27.0 ± 4.4	27. 4 ± 4.4	0.55	0.35	0.52
Date of Weight (Days closest to visit)	59 ± 57	45 ± 25	53 ± 54	75 ± 36	0.81	0.12	N/A
HOMA-IR	1.83 ± 0.73	2.66 ± 1.37	1.98 ± 1.02	1.97 ± 1.52	0.32	0.34	0.067
Lactate (mM)	3.59 ± 0.87	4.57 ± 1.38	3.72 ± 0.71	3.78 ± 0.98	0.61	0.09	0.067
Total Cholestero, (mg/dL)	179 ± 27	169 ± 68	201 ± 37	178 ± 26	0.16	0.31	0.087
HDL Cholesterol (mg/dl)	53 + 12	52 ± 27	59 ± 10	53 ± 13	0.36	0.92	0.20
LDL Cholesterol (mg/dl)	106 ± 25	98 ± 41	119 ± 32	104 ± 24	0.32	0.52	0.53
Triglycerides (mg/dl)	104 ± 34	100 ± 43	116 ± 69	101 ± 55	0.67	0.74	0.40
Adiponectin (μg/mL)	3.47 ± 1.65	3.40 ± 1.31	5.18 ± 1.90	4.43 ± 1.32	0.05	0.12	0.27
Leptin (ng/mL)	9.7 ± 9.5	9.3 ± 5.1	17.5 ± 15.6	22.1 ± 20.1	0.21	0.08	0.31
IL-6 (pg/mL)	1.56 ± 0.62	1.34 ± 0.89	1.49 ± 0.69	1.69 ± 0.75	0.37	0.33	0.76
MCP-1 (pg/mL)	96.0 ± 70.5	91.5 ± 45.6	95.6 ± 34.1	100.0 ± 56.3	0.93	0.93	0.76

^aAbbreviations: EVG-elvitegravir, c-cobicistat, FTC-emtricitabine, TDF-tenofovir disoproxil fumarate, RAL-raltegravir.

^bContinuous variables are presented as mean ± SD.

Figure 1.

Metabolic alterations in individuals switched to EVG/c/FTC/TDF from EFV-based regimens. HOMA- IR(A), Plasma insulin(B), serum glucose(C), adiponectin(D), leptin(E), and plasma lactate(F) were measured in individuals pre- and 8 weeks post-switch from EFV to a RAL (n = 10) or EVG + cobicistat (EVG/c, n = 9) -containing regimen.

Discussion

A growing pool of clinical data indicate that certain PLWH are at higher risk for excess weight gain and diabetes following the initiation or switch to an INSTI-based treatment regimens.¹⁶ The potential metabolic consequences and associated mechanism of this weight gain phenomenon have profound clinical relevance to the health of PLWH. In this report, we show that participants who switched from an EFV-based treatment regimen to EVG/c-containing regimen had increased insulin resistance after 8 weeks of treatment, as measured by HOMA-IR. However, those switched to RAL did not experience this same increase. The changes in HOMA-IR appeared to be independent from measurable weight gain, potentially representing earlier and more subtle metabolic changes of the effect of INSTIs as this study collected samples within 8 weeks of drug switch. Further, we observed that increases in HOMA-IR tended to be accompanied by increases in circulating lactate, a potential indicator of mitochondrial dysfunction, as seen in other disease states¹⁷ or other metabolic perturbations.¹⁸

Although the issue of weight gain in relation to INSTI exposure has been very well described, the potential metabolic consequences remain to be fully elucidated. Recent retrospective studies have identified a potential insulin resistance signal associated with INSTI exposure^5,19; although this has not always been found to correlate with weight gain. Indeed, insulin resistance following a gain in weight would be expected to occur after a substantial time delay. Thus, an early signal of insulin resistance soon after INSTI initiation may suggest a different mechanism driving this adverse effect independent of adipose expansion. In our study we also measured 2 metabolically important adipokines, adiponectin and leptin that did not appreciably change in the 8 weeks following the switch. This lack of change may reflect too early of a time point to detect measurable alterations in the serum levels of adipokines that typically are associated with both fat mass and insulin resistance.^20,21 While little is known about the combined effects of weight changes and different antivirals, in-vitro models indicate that INSTI’s, including elvitegravir, may suppress adiponectin secretion.^22,23

Unlike other early antiretrovirals, especially NRTIs, INSTIs do not appear to result in overt mitochondrial toxicity,²⁴ though previous measures relied on mitochondrial DNA content and not function. However, several studies using more in-depth measures of mitochondrial function have shown that both EVG and DTG result in decreased mitochondrial respiration in CD4+ T-cells and BV4 microglial cells without changes in mitochondrial content.^25,26 Further, INSTIs appear to increase production of reactive oxygen species (ROS) by mitochondria in newly differentiated adipocytes.²⁷ Both mitochondrial dysfunction and subsequent excessive ROS have been proposed as important mediators in development of insulin resistance and type 2 diabetes.²⁸ Our observation of increased circulating lactate after switch to INSTI could be due to decreases in mitochondrial function that likely resulted in a shift from oxidative to glycolytic energy production and provide the potential link to the increases in insulin resistance. Indeed, when exposed to INSTIs, newly differentiated adipocytes have been shown to lower insulin stimulated glucose uptake and phosphorylation of AKT.²⁷

Our study is unique in that the cohort allowed the opportunity to study INSTI effect alone as all participants remained on the same backbone of FTC/TDF unlike many other observational studies which have confounding from simultaneous switch of TDF to TAF.²⁹ This cohort allowed us to investigate the isolated effect of INSTI exposure. Further, in contrast to treatment-naïve individuals, these patients had been virally suppressed for ≥6 months, reducing the potential differences in weight gain related to return-to-health typically seen in ART initiation.

Our study had several limitations. Notably, this study was conducted using a preexisting cohort and protocol and therefore, changes or additions could not be made. Therefore, the preexisting sample set was small, albeit powered for the original study, was not powered to detect the major outcomes we report here. Thus larger future studies will be required to confirm these findings. Because EVG was co-formulated with cobicistat, we were unable to determine if the observed effects on insulin resistance were due solely to EVG, cobicistat booster or the combination of EVG/c. Independent metabolic effects of cobicistat have not been thoroughly examined, though an earlier analysis of individuals switched to either cobicistat-boosted darunavir or cobicistat-boosted EVG observed similar weight gain probabilities at 28 and 56 weeks, yet in this study the initial ART regimens were not well characterized.³⁰ Though both RAL and EVG are INSTIs, it is important to note that the effects of these drugs on metabolically relevant tissues are not common across all ARVs in the INSTI class. EVG strongly inhibits differentiation of adipocytes in-vitro, while RAL does not.²³ Further, while EVG and EFV were both associated with markers of mitochondrial toxicity RAL was not. Weight and height were not measured parameters of the parent study, and BMI measurements were neither consistently collected at baseline and follow-up visits nor in a standardized fashion, potentially adding error to these measurements. Furthermore, some participants’ weight data points were not collected at time points immediately adjacent to the trial start and end dates, although most collected data was at minimum a few months from those timepoints, and some participants did not have close by weight data points collected at all. However, all patients were maintained on INSTI-based treatments following the trial, limiting confounding by a second switch. Additionally, because our sample size was small potential confounding factors related to weight gain and insulin resistance, such as age, sex, race and lifestyle factors, could not be adequately controlled. This is of particular importance as growing evidence, including work from our group, indicate women are at greater risk for severe weight gain on INSTI-based regimens.⁵ Because there were more men in this cohort, we may have underestimated the impact of this switch on weight gain. Due to the time when the original study was done, newer generation of INSTIs (dolutegravir, bictegravir) were not commercially available so these data may not directly translate to newer regimens. We also acknowledge that boosted INSTI regimens with cobicistat are not commonly used in current clinical practice, the observations may still provide insight into the INSTI class specific effects and metabolic changes following switch from EFV. Finally, HOMA-IR is a comparatively crude measure of insulin resistance and glucose homeostasis when compared to clamps or frequently sampled glucose tolerance tests.^14,15 However, HOMA-IR is strongly associated with other known drivers of insulin resistance, such as central adiposity, even in PLWH,³¹ and therefore increases could be used as an early indicator of adverse metabolic trajectories.

To the best of our knowledge, this is the first report of INSTI-associated insulin resistance in such a short period of 8 weeks after switching from EFV use. Even though EFV has been shown to be weight suppressive, and switching from EFV containing regimens tends to result in weight gain,³² this association with increased insulin resistance was potentially independent of weight gain, especially in the case of EVG/c. Other studies examining the link between switch to INSTIs and metabolic parameters are mixed in direction and magnitude. A trial in PLWH with non-alcoholic fatty liver disease who switch from ritonavir-boosted protease inhibitor to raltegravir showed either a slight decrease in insulin resistance or no change after 12 months compared to lifestyle modification only³³; Consistent with our data from switch to raltegravir. However, in a longitudinal matched cohort study of 2437 PWH, people switching to INSTI without adverse metabolic health at baseline, showed a 30% lower risk of insulin resistance compared to those not exposed to INSTI after 1 year.⁶ Further, weight gain was a strong mediator of this effect. However, the specific antiretroviral medications were not reported in this study, which may underlie differences in effects seen here.

Our data highlight the importance of early monitoring of metabolic parameters, such as HOMA-IR, upon initiation of INSTI regimens, especially with the newer generation of INSTIs like dolutegravir, cabotegravir, bictegravir, and in populations already at high risk for these metabolic conditions. These data provide a potential mechanistic pathway, one through mitochondrial toxicity and oxidative stress driving adverse metabolic effects of some INSTI regimens, and a potential target for future strategies to improve metabolic health in people living with HIV on long-term INSTI-based ART.

Footnotes

Acknowledgements

The authors would like to thank the participants for their time and contributions as well as the physicians that referred them for the parent studies.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: RTP receives support from the NIH. AA receives research support from Gilead Sciences, GSK/ViiV, Theratechnologies and DayZero Diagnostics and has received honoraria from Gilead Sciences. DK has received consulting honoraria from AbbVie, GlaxoSmithKline, Gilead, Janssen, Merck, Roche and ViiV, and has received research support from Gilead, Merck and ViiV. NHL is an employee of Moderna Therapeutics since the completion of the study and has received research funding from Gilead Sciences, GSK/ViiV, Merck, and DayZero Diagnostics. This work was funded in part by the Boston University Clinical HIV/AIDS Research Training Program (T32AI052074), and parent studies were supported by the Gilead Sciences and Merck Investigator-initiated studies programs.

ORCID iD

Richard Taylor Pickering

References

Bourgi

Rebeiro

Turner

, et al. Greater weight gain in treatment naive persons starting dolutegravir-based antiretroviral therapy. Clin Infect Dis 2019; 70: 1267–1274. DOI: 10.1093/cid/ciz407.

Data Collection on Adverse Events of Anti-HIV drugs DAD Study Group Smith

Sabin

, et al. Factors associated with specific causes of death amongst HIV-positive individuals in the D:A:D Study. Aids 2010; 24: 1537–1548. DOI: 10.1097/QAD.0b013e32833a0918.

Norwood

Turner

Bofill

, et al. Brief report: weight gain in persons with HIV switched from efavirenz-based to integrase strand transfer inhibitor-based regimens. J Acquir Immune Defic Syndr 2017; 76: 527–531. DOI: 10.1097/qai.0000000000001525.

Smith

Mittendorfer

Klein

. Metabolically healthy obesity: facts and fantasies. J Clin Invest 2019; 129: 3978–3989. DOI: 10.1172/JCI129186.

Asundi

Olson

Jiang

, et al. Integrase inhibitor use associated with weight gain in women and incident diabetes mellitus. AIDS Res Hum Retroviruses 2022; 38: 208–215. DOI: 10.1089/AID.2021.0091.

Milic

Renzetti

Ferrari

, et al. Relationship between weight gain and insulin resistance in people living with HIV switching to integrase strand transfer inhibitors-based regimens. AIDS 2022; 36: 1643–1653. DOI: 10.1097/QAD.0000000000003289.

Randell

Jackson

Milinkovic

, et al. An open-label, randomized study of the impact on insulin sensitivity, lipid profile and vascular inflammation by treatment with lopinavir/ritonavir or raltegravir in HIV-negative male volunteers. Antivir Ther 2017; 22: 145–151. DOI: 10.3851/IMP3098.

Spinner

Kern

Zink

, et al. Neither boosted elvitegravir nor darunavir with emtricitabine/tenofovir disoproxil fumarate increase insulin resistance in healthy volunteers: results from the STRIBILD-IR study. Antivir Ther 2016; 21: 627–631. DOI: 10.3851/IMP3049.

Griesel

Maartens

Chirehwa

, et al. CYP2B6 genotype and weight gain differences between dolutegravir and efavirenz. Clin Infect Dis 2021; 73: e3902–e3909. DOI: 10.1093/cid/ciaa1073.

10.

Wood

Huhn

. Excess weight gain with integrase inhibitors and tenofovir alafenamide: what is the mechanism and does it matter? Open Forum Infect Dis 2021; 8: ofab542. DOI: 10.1093/ofid/ofab542.

11.

Venter

WDF

Moorhouse

Sokhela

, et al. Dolutegravir plus two different prodrugs of tenofovir to treat HIV. N Engl J Med 2019; 381: 803–815. DOI: 10.1056/NEJMoa1902824.

12.

Sax

Erlandson

Lake

, et al. Weight gain following initiation of antiretroviral therapy: risk factors in randomized comparative clinical trials. Clin Infect Dis 2020; 71: 1379–1389. DOI: 10.1093/cid/ciz999.

13.

Asundi

Robles

Starr

, et al. Immunological and neurometabolite changes associated with switch from efavirenz to an integrase inhibitor. J Acquir Immune Defic Syndr 2019; 81: 585–593. DOI: 10.1097/QAI.0000000000002079.

14.

Matthews

Hosker

Rudenski

, et al. Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia 1985; 28: 412–419. DOI: 10.1007/bf00280883.

15.

Wallace

Levy

Matthews

. Use and abuse of HOMA modeling. Diabetes Care 2004; 27: 1487–1495. DOI: 10.2337/diacare.27.6.1487.

16.

Shah

Hill

. Risks of metabolic syndrome and diabetes with integrase inhibitor-based therapy: republication. Curr Opin HIV AIDS 2021; 16: 106–114. DOI: 10.1097/COH.0000000000000671.

17.

Gonzalo

Nogueras

Gil-Sanchez

, et al. Impairment of mitochondrial redox status in peripheral lymphocytes of multiple sclerosis patients. Front Neurosci 2019; 13: 938. DOI: 10.3389/fnins.2019.00938.

18.

Broskey

Zou

Dohm

, et al. Plasma lactate as a marker for metabolic health. Exerc Sport Sci Rev 2020; 48: 119–124. DOI: 10.1249/JES.0000000000000220.

19.

Summers

Lahiri

Angert

, et al. Metabolic changes associated with the use of integrase strand transfer inhibitors among virally controlled women. J Acquir Immune Defic Syndr 2020; 85: 355–362. DOI: 10.1097/QAI.0000000000002447.

20.

Considine

Sinha

Heiman

, et al. Serum immunoreactive-leptin concentrations in normal-weight and obese humans. N Engl J Med 1996; 334: 292–295. DOI: 10.1056/NEJM199602013340503.

21.

Cnop

Havel

Utzschneider

, et al. Relationship of adiponectin to body fat distribution, insulin sensitivity and plasma lipoproteins: evidence for independent roles of age and sex. Diabetologia 2003; 46: 459–469. DOI: 10.1007/s00125-003-1074-z.

22.

Domingo

Quesada-Lopez

Villarroya

, et al. Differential effects of dolutegravir, bictegravir and raltegravir in adipokines and inflammation markers on human adipocytes. Life Sci 2022; 308: 120948. DOI: 10.1016/j.lfs.2022.120948.

23.

Moure

Domingo

Gallego-Escuredo

, et al. Impact of elvitegravir on human adipocytes: alterations in differentiation, gene expression and release of adipokines and cytokines. Antiviral Res 2016; 132: 59–65. DOI: 10.1016/j.antiviral.2016.05.013.

24.

Koethe

. Adipose tissue in HIV infection. Compr Physiol 2017; 7: 1339–1357. DOI: 10.1002/cphy.c160028.

25.

Korencak

Byrne

Richter

, et al. Effect of HIV infection and antiretroviral therapy on immune cellular functions. JCI Insight 2019; 4: e126675. DOI: 10.1172/jci.insight.126675.

26.

George

Mattingly

Roland

, et al. Physiologically relevant concentrations of dolutegravir, emtricitabine, and efavirenz induce distinct metabolic alterations in HeLa epithelial and BV2 microglial cells. Front Immunol 2021; 12: 639378. DOI: 10.3389/fimmu.2021.639378.

27.

Gorwood

Bourgeois

Pourcher

, et al. The integrase inhibitors dolutegravir and raltegravir exert pro-adipogenic and profibrotic effects and induce insulin resistance in human/simian adipose tissue and human adipocytes. Clin Infect Dis 2020; 71: e549–e560. DOI: 10.1093/cid/ciaa259.

28.

Sangwung

Petersen

Shulman

, et al. Mitochondrial dysfunction, insulin resistance, and potential genetic implications. Endocrinology 2020; 161: bqaa017. DOI: 10.1210/endocr/bqaa017.

29.

Mallon

Brunet

Hsu

, et al. Weight gain before and after switch from TDF to TAF in a U.S. cohort study. J Int AIDS Soc 2021; 24: e25702. DOI: 10.1002/jia2.25702.

30.

Hsu

Brunet

Fusco

, et al. Excessive weight gain: current antiretroviral agents in virologically suppressed people with HIV. AIDS Res Hum Retroviruses 2022; 38: 782–791. DOI: 10.1089/AID.2021.0188.

31.

Glesby

Hanna

Hoover

, et al. Abdominal fat depots, insulin resistance, and incident diabetes mellitus in women with and without HIV infection. AIDS 2018; 32: 1643–1650. DOI: 10.1097/QAD.0000000000001873.

32.

Erlandson

Carter

Melbourne

, et al. Weight change following antiretroviral therapy switch in people with viral suppression: pooled data from randomized clinical trials. Clin Infect Dis 2021; 73: 1440–1451. DOI: 10.1093/cid/ciab444.

33.

Calza

Colangeli

Borderi

, et al. Improvement in liver steatosis after the switch from a ritonavir-boosted protease inhibitor to raltegravir in HIV-infected patients with non-alcoholic fatty liver disease. Infect Dis (Lond) 2019; 51: 593–601. DOI: 10.1080/23744235.2019.1629008.