Striking differences in weight gain after cART initiation depending on early or advanced presentation: Results from the ANRS CO4 FHDH cohort

BACKGROUND
Many studies have reported weight gain in ART-naive people living with HIV (PWH) initiating an integrase strand-transfer inhibitor-based regimen. We studied the impact of early or advanced presentation and that of individual drugs in PWH initiating combined ART (cART) between 2012 and 2018.


METHODS
From the French Hospital Database HIV cohort, we assessed factors associated with a weight gain  ≥10%, weight change after cART initiation or BMI increase  ≥5 kg/m2 up to 30 months. The analyses were conducted overall, and among PWH with early (primary infection or CD4 >350/mm3 and viral load  <100 000 copies/mL, without AIDS) and advanced presentation (AIDS or CD4 <200/mm3, not during primary infection).


RESULTS
At 30 months, 34.5% (95% CI: 33.5-35.6) of the 12 773 PWH had a weight gain ≥10%, with 20.9% (95% CI: 19.6-22.2) among the 5794 with early presentation and 63.1% (95% CI: 60.9-65.3) among the 3106 with advanced presentation. Weight gain was 2.8 kg (95% CI: 2.0-3.7) for those with early presentation and 9.7 kg (95% CI: 8.4-11.1) for those with advanced presentation. Most weight gain occurred in the first 12 months. Underweight and obese PWH were at significantly higher risk of a BMI increase  ≥5 kg/m2 than normal-weight PWH. Results differed within classes and by outcome. Raltegravir and dolutegravir were consistently associated with greater weight gain than the other third agents. Tenofovir alafenamide was also associated with higher weight gain than tenofovir disoproxil or abacavir.


CONCLUSIONS
After initiating cART, PWH with early presentation exhibited a small weight gain, whereas it was large among those with advanced presentation. The choice of ART should account for the risk of weight gain, especially for PWH who present with advanced disease and/or are obese.


SYNOPSIS
Introduction: Many studies have reported weight gain in ART-naïve people living with HIV (PWH) initiating an integrase strand-transfer inhibitor-based regimen.We studied the impact of early or advanced presentation and that of individual drugs in PWH initiating combined ART (cART) between 2012 and 2018.

Methods:
From the French Hospital Database on HIV cohort, we assessed factors associated with a weight gain≥10%, weight change after cART initiation or BMI increase≥5kg/m 2 up to 30 months.The analyses were conducted overall, among PWH with early (primary infection or CD4>350/mm 3 and viral load<100000 copies/mL, without AIDS), and advanced presentation (AIDS or CD4<200/mm 3 , not during primary infection).
Most weight gain occurred in the first 12 months.Underweight and obese PWH were at significantly higher risk of BMI increase≥5kg/m 2 than normal-weight PWH.Results differed within classes and by outcome.Raltegravir and dolutegravir were consistently associated with greater weight gain than the other third agents.Tenofovir-alafenamide was also associated with higher weight gain than tenofovir-disoproxil or abacavir.
Discussion: After initiating cART, PWH with early presentation exhibited a small weight gain, whereas it was large among those with advanced presentation.The choice of ART should account for the risk of weight gain, especially for PWH who present with advanced disease and/or are obese.

INTRODUCTION
In recent recommendations for people living with HIV (PWH), 1,2 the antiretroviral combination prescribed at initiation should include an integrase strand-transfer inhibitor (INSTI).7][8][9] Only a small number of studies have assessed specific drugs. 10][8][9][10][11] Greater weight gain was observed with TAF/FTC than with tenofovir disoproxil and emtricitabine (TDF/FTC) 7,8,10,11 .3][14] Weight gain in PWH raises concerns about the potential associated increased risk of cardiovascular and metabolic diseases and mortality 15,16 .A large meta-analysis in the general population showed that an increase in body-mass index (BMI) of 5 kg/m 2 is associated with a 30% increase in the risk of death. 17ost studies, factors such as sex, race, BMI at the initiation of treatment, CD4 T-cell levels, HIV-1 viral load (VL), and prior acquired immunodeficiency syndrome (AIDS), were shown to be associated with weight gain in treatment-naïve PWH, in addition to the type of drugs.However, the association between weight gain and clinical presentation (early or advanced) was not specifically explored, although the initial signal came from two trials that included many people with advanced HIV disease. 3,4 he return-to-health phenomenon could confuse the effect of treatment on weight gain.
In this context, we aimed to study the factors associated with weight gain in ART-naive PWH initiating combined antiretroviral treatment (cART) in France between 2012 and 2018 according to clinical presentation (early or advanced) and each component of the treatment combination using a large cohort of PWH, the ANRS CO4-French Hospital Database on HIV (FHDH) cohort.

Participants
the present analysis, we selected ART-naïve PWH infected with HIV-1 who initiated cART between 2012 and 2018.Participants had to have had at least one BMI measurement within six months prior to cART initiation (baseline) and one during the follow-up and to have initiated cART at least one year before the last recorded FHDH visit in the centre.We excluded transgender participants because of the interaction of hormone therapy with weight and pregnant women were censored at the time of pregnancy.Only participants with cART combinations taken by at least 140 PWH were considered.

Statistical analysis
In addition to the whole population, we choose to study two contrasting groups, PWH with early or with advanced presentation, to better assess the differential impact of treatment initiation according of HIV disease stages and with the hypothesis that the issue of return to health would be less pronounced in PWH presenting early.PWH presenting early were defined as participants initiating cART at primary infection or with CD4 T cells >350/mm 3 and a VL <100000 copies/mL and without AIDS at cART initiation and PWH presenting with advanced HIV disease as participants with AIDS or CD4 T cells <200/mm 3 not at primary infection at cART initiation.
As primary endpoint, we considered the rates of weight gain of at least 10%, which is often considered to be clinically significant, 20 and as secondary endpoints weight change after cART initiation and BMI increase≥5kg/m 2 .Rates of weight gain of at least 10% and BMI increase≥5kg/m 2 were estimated at 30 months using Kaplan Meier estimates.Follow-up was censored at treatment modification, at the last available weight measurement, or at 30 months, whichever occurred first.
The association between weight change and time since cART initiation was assessed using a mixed linear model with a spatial power law covariance structure to account for irregular duration between clinical visits 22 and adjusted for the same variables as previously described.
All analyses were conducted overall and separately in two subgroups according to early presentation or advanced presentation, except for the factors associated with an increase in BMI of at least 5 kg/m 2 .The change in BMI pattern was assessed in the two subgroups in the first 30 months (±1.5 months).All statistical analyses were conducted using SAS 9.4.A p-value ≤5% denoted statistical significance.

RESULTS
Overall, 12,773 ART-naïve PWH initiating cART between 2012 and 2018 were included, accounting for 28,737 person-years (median 1.76 years (interquartile range (IQR), 0.71-3.32)).Among them, 5,794 presented with an early HIV disease and 3,106 with an advanced HIV disease, with a median followup of 2.05 (IQR, 0.84-3.74)and 1.44 (IQR, 0.56-2.90)years, respectively.The median number of weight measurements was 5 (IQR 3-7), with no difference according to the clinical presentation nor the baseline category of BMI.Participants' characteristics are presented in Table 1.Overall, 75.6% of the population was men and MSM accounted for approximately half of the participants.A quarter of the population originated from SSA.At cART initiation, the median CD4 T-cell level was 377/mm 3 , median VL 4.7 log copies/mL, and median BMI 23.0 kg/m 2 (IQR, 20.7-25.8).At cART initiation, 8.2% of the participants were obese and 22.2% overweight: 20.6% and 30.7% for women originating from SSA, 8.8% and 29.6% for men originating from SSA, 15.3% and 23.2% for other women, and 4.2% and 18.8% for other men, respectively.
Among the total population (Table 1), 41.7% initiated cART with a PI-based regimen, mostly with darunavir, 33.3% with an INSTI-based regimen, mostly with dolutegravir or elvitegravir, and 25.0% with a NNRTI-based regimen, mostly rilpivirine.TDF/FTC was the most frequently used backbone (78.2%), whereas only 1,989 (15.6%)PWH received ABC/3TC and 800 (6.3%) TAF/FTC.The number of participants receiving each specific cART combination is presented in Table S1.
The characteristics differed highly according to advanced or early presentation.PWH with advanced presentation were older (median 42 versus 36 years), less likely to be MSM (29.0%versus 57.1%), more likely to have originated from SAA (36.3% versus 20.3%), more likely to be underweight (12.6% versus 5.0%), and more likely to have received a PI-based regimen (56.2% versus 30.5%) and less likely an INSTI-based regimen (11.6% versus 35.7%).

Weight gain of at least 10%
At 30 months, 34.5% (95%CI 33.5-35.6) of the PWH had a weight gain of at least 10%, 20.9% (95%CI 19.6-22.2) among PWH with early presentation and 63.1% (95%CI 60.9-65.3)among those with advanced presentation (Table 2-4).Higher risks were observed for women, non-MSM, underweight PWH, and those with more profound immunodeficiency, higher VL, and prior AIDS, with no difference between AIDS-defining events.In the overall population, the percentage of PWH with a weight gain of at least 10% at 30 months ranged from 20.9% among those receiving rilpivirine to 44.3% among those receiving raltegravir.In adjusted analyses, PWH initiating cART with dolutegravir or raltegravir had a higher risk of a weight gain of at least 10% than those initiating with other drugs, except PIs, with no significant differences between them.Those initiating with efavirenz had a lower risk of a weight gain of at least 10% than with the other drugs, except rilpivirine (Table S2a).TAF/FTC was associated with a significantly higher risk of weight gain of at least 10% than TDF/FTC (HR=1.52,95%CI 1.29-1.79)or ABC/3TC (HR=1.61,95%CI 1.31-1.96),whereas there was no statistical difference between ABC/3TC and TDF/FTC (HR=0.95,95%CI 0.84-1.06)(Table 2).Among early-presenters, the percentage of weight gain of at least 10% ranged from 11% with efavirenz to 27% with ATV/RTV.In adjusted analyses, there was no statistical difference among the various INSTIs or between PIs, or between INSTIs and PIs.Detailed results are reported in Table 3 and Table S5a.Among PWH with advanced HIV disease, the percentage of weight gain of at least 10% ranged from 37% with rilpivirine to 74% with raltegravir.Detailed results are reported in Tables 4 and     Table S5b.
In the sensitivity analysis excluding participants presenting with AIDS (Table S7), the proportion of participants with a weight gain of at least 10% was 30.9% instead of 34.5% in the whole population.
The results of the Cox model evaluating the factors associated with this weight gain were similar to those of the analysis of the whole sample.

Weight change overtime
The mean weight changes estimated by multivariable mixed regression models at 12, 24 and 30 months are presented in Table 5 for the total population.At 30 months, the adjusted mean weight gain was +7.9 kg (95% confidence interval [CI] 7.1-8.6).Most of the weight gain occurred in the first 12 months.The weight plateaued after 24 months.Weight gains were significantly larger among women, PWH originating from SSA, non-MSM, underweight PWH, and those with more profound immunodeficiency, higher VL, and prior AIDS, with no difference between AIDS-defining events.The results and comparisons according to the various drugs are presented in Table S2b and Figure 1a for third agents and Figure 1b for nucleoside backbones.Over 30 months, the adjusted weight gain was the highest for PWH receiving dolutegravir (+8.7 kg) and raltegravir (+8.5 kg) and the lowest for those receiving efavirenz (+6.5 kg).Weight gain with raltegravir was significantly greater than that with efavirenz, whereas there was no statistical difference with any other third agent.Weight gain with dolutegravir was significantly greater than that with darunavir, atazanavir, rilpivirine, or efavirenz.
Weight gain with TAF/TDF was greater than that with TDF/FTC (p<0.0001) and ABC/3TC (p=0.05), and weight gain with ABC/3TC was greater than that with TDF/FTC (p=0.04)(Figure 1b and Table S2b).
Weight gain varied highly according to whether PWH initiated cART early or at advanced HIV disease.
Among early-presenters, the adjusted mean weight gain at 30 months was +2.8 kg ranging from 1.8 kg for efavirenz to 3.5 kg for dolutegravir.Detailed results are reported in Figure 1c and 1d and Table S3.
Among PWH with advanced HIV disease, the adjusted mean weight gain was +9.7 kg at 30 months, the detailed results are reported in Figure 1e and 1f and Table S4.Over 30 months, the adjusted weight gain was the highest for PWH receiving dolutegravir (+10.9 kg), raltegravir (+10.6 kg), or atazanavir (+10.2 kg) and the lowest for those receiving elvitegravir (+8.4 kg) (Figure 1e and Table S4).

BMI increase and BMI categories over time
Overall, 9.1% of participants (95%CI 8.5-9.8) had a BMI increase of at least 5 kg/m 2 , 3.4% (95%CI 2.8-4.0)among those presenting early and 23.9% (95%CI 21.9-26.1)among those with advanced HIV disease.In adjusted analyses, a higher risk of a BMI increase of at least 5 kg/m 2 was observed with raltegravir than with any other drugs, except dolutegravir, and with TAF/FTC than with TDF/FTC (HR=1.68,95%CI 1.19-2.38)or ABC/3TC than with TDF/FTC (HR=1.28,95%CI 1.02-1.61)(Table S6a and S6b).For the other factors, the association with a BMI increase of at least 5 kg/m 2 was in the same direction as in the other analyses, except for BMI.The risk of a BMI increase of at least 5 kg/m 2 was higher for both underweight (HR=1.33,95%CI 1.07-1.67)and obese participants (HR=1.26,95%CI 0.98-1.61).
The pattern of BMI categories over time was stable among PWH presenting early but changed for PWH with advanced HIV disease, with more obese PWH, especially for those receiving raltegravir, dolutegravir, or atazanavir (Figures S1 and S2).

DISCUSSION
In this observational study, a small weight gain over 30 months was observed among ART-naïve PWH presenting early in the course of HIV disease, whereas it was large among those presenting with advanced HIV disease, with a mean increase of 2.8 kg and 9.7 kg, respectively.At 30 months, 34.5% of the PWH had a weight gain of at least 10% and 9.1% a BMI increase of at least 5k g/m 2 .Among those presenting early, 20.9% had a weight gain of at least 10% and 3.4% a BMI increase of at least 5 kg/m 2 , whereas among those presenting with advanced HIV disease, the corresponding values were 63.1% and 23.9%, respectively.The weight gain trajectories were similar for the early and advanced presenters, with most of the weight gain occurring during the first year.Concerning the drugs received, the results differed within each class and varied by outcome (a weight gain of at least 10% or weight change), making it difficult to draw any general conclusions.However, raltegravir and dolutegravir were consistently associated with higher risk of weight gain than the other third agents and TAF was also associated with higher risk of weight gain than TDF or abacavir.
]23 We show here that the weight gain following treatment initiation is mainly observed in PWH with presenting with advanced HIV disease, whereas the weight gain among PWH presenting early is limited.At treatment initiation, PWH presenting with advanced HIV disease had a lower BMI than those presenting early with median values of 22.3 kg/m 2 (IQR, 19.9-25.1)and 23.2 kg/m 2 (IQR, 21.2-25.7)respectively, and the proportion of participants with a weight gain of at least 10% were significantly larger among underweight PWH (58.4%) than among obese PWH (28.4%) at 30 months.5][26] In PWH with advanced presentation, part of the weight gain may simply be a return to health, and the clinical consequences, if any, could be limited to those who become obese.However, it would be preferable not to be diagnosed late because of the unknown consequences of rapid weight gain associated with treatment initiation and the already known long-term consequences of advanced presentation 27 , which underlines the need of continued efforts to diagnose PWH early.Overall, 9.1% of PWH initiating cART had a BMI increase of at least 5 kg/m 2 , known to be associated with deleterious health outcomes. 17Interestingly, the percentage of those showing such an increase was significantly higher in both underweight (17.0%) and obese (12.4%)PWH, (hazard ratio (HR) 1.33, 95%CI 1.07-1.67 and 1.26, 95%CI 0.98-1.61,respectively relative to PWH with a BMI in the normal range) which may lead to different health consequences in these two groups. 28Additional studies are therefore needed on obese PWH with such weight gain to assess its consequences in terms of metabolic and cardiovascular outcomes. 29n the two initial clinical trials reporting weight gain, we show that, among ART-naïve PWH initiating treatment, weight gain with dolutegravir is greater than that with efavirenz.3,4 Since the publication of these trials, this result has been completed with studies of weight gain after initiating treatment with other third drugs, such as raltegravir, bictegravir, elvitegravir, darunavir, atazanavir, and rilpivirine.In accordance with our results, one large cohort study from northern America showed that raltegravir and dolutegravir were associated with more weight gain than elvitegravir-, NNRTI-or PI-based regimens.6 In pooled analyses that included eight clinical trials comparing weight gain in ART-naïve PWH initiating dolutegravir, bictegravir, elvitegravir, atazanavir, and NNRTIs, the authors found that dolutegravir and bictegravir were associated with more weight gain than elvitegravir, atazanavir, or NNRTIs.8 In another recent American cohort study, elvitegravir was associated with lower weight gain than bictegravir and dolutegravir after six months.11 Thus, more weight gain was observed with INSTIs other than elvitegravir. Although raltegravir has been consistently shown to be associated with the worst weight gain, the current practical consequences are probably limited, because it is no longer frequently used to initiate treatment.In the entire group, efavirenz was associated with less weight gain than all the other drugs, supporting the hypothesis that it could inhibit weight gain.Concerning backbone drugs, we observed a higher risk of weight gain with TAF/FTC than TDF/FTC, regardless of the criteria used to define weight gain, similar to many published studies.4,7,8,10,11 Of note in our study TAF/FTC, was only available in France in multidrug pills, either with rilpivirine or boosted elvitegravir.Overall, the observed effects of the various drugs were not class effects per se.In the INSTI class, elvitegravir was associated with a lower weight gain than the other INSTIs and TAF with a higher weight gain than the other NRTIs.
For clinical practice, our results suggest that it would be important to monitor weight gain in the first year after cART initiation and to carefully select the prescribed regimen, balancing its virological and immunological advantages with its consequences in terms of weight gain, in particular for PWH presenting with advanced HIV disease or who are obese. 29Given the persisting high proportion of people diagnosed with advanced presentation, 29% in France in 2021 30 , this could concern a large proportion of newly diagnosed PWH.
The main strength of our study was its large size, allowing us to study the risk of weight gain according to whether PWH presented early or with advanced HIV disease and to the type of drug.
Distinguishing two subgroups of participants allowed us to reduce the confusion between the effect of the HIV infection itself and the impact of the drugs on weight gain.Studying the association of weight gain by drug rather than by the class of drug allowed us to highlight the differential effects of drugs from the same class on weight gain.Among the limits of our study, bictegravir could not be analyzed due to the study period, which preceded its availability in France.Data on physical activity and diet are not collected in the FHDH and therefore could not be accounted for.The study also lacked pre-HIV measures of weight, which could have helped to assess the return to health phenomenon, and the composition of the body changes following cART initiation.
In conclusion, weight gain was mainly observed among participants presenting with advanced HIV disease.The limited weight gain observed among PWH presenting early emphasizes, once again, the need to be diagnosed early for early immediate treatment.Multiple factors are associated with weight gain and the choice of initial treatment should depend on the characteristics of each PWH.In further studies, it would be important to assess the population of PWH for whom clinical For the purpose of Open Access, a CC-BY public copyright license has been applied by the authors to the present document and will be applied to all subsequent versions up to the Author Accepted Manuscript arising from this submission.

FUNDING
The ANRS CO4 FHDH is supported by the ANRS-MIE (Agence Nationale de Recherche sur le Sida et les hépatites virales-Maladies Infectieuse Emergentes), INSERM (Institut National de la Santé et de la Recherche Médicale) and the French Ministry of Health.The funders had no role in the study design, data collection, data analysis and interpretation, or writing of the report.

AUTHORS AND CONTRIBUTION
DC, SG, JC, and VP designed the study.SA, LB, CA, FC, PT, CD, PE, CK, MAK, OL, SM, GM, HM, JLM, JP, LP, LS, SR, and PT included PWH.VP analyzed the data.SG, VP, and DC drafted the manuscript, had full access to the data, verified the data, and had final responsibility for the decision to submit the study for publication.All authors were involved in the interpretation of the data and critical revision of the manuscript and approved the final version.

TRANSPARENCY DECLARATIONS
CA received personal fees from Gilead, Janssen-Cilag, MSD, and ViiV Healthcare for travel grants and honoraria outside the submitted work.LP reports personal fees from Pfizer (2022).F Caby reports personal fees from Gilead, ViiV Healthcare, and MSD for lectures and serving on an expert board outside the submitted work.PdT received personal fees for workshop participation and travel grants 491 Currently, more than 180 hospitals contribute to the data collection, 206,651 PWH have been included, and 102,030 were being followed in 2018.All participants provided written informed consent for the use of their data for research purposes.The cohort was initially approved by the French data protection authority, the CNIL (Commission Nationale de l'Informatique et des Libertés) on 27 November 1991 (Official Journal, 17 January 1992).The research authorisation was updated to comply with the new regulations, including the General Data Protection Regulation.The ANRS CO4-FHDH cohort was approved by the CEREES (Comité d'Expertise pour les Recherches, les Études et les Évaluations dans le domaine de la Santé) on 20 July 2018 and as a hospital data warehouse by the CNIL on 19 February 2021.The cohort received the authorisation to conduct research projects on the data warehouse by the CNIL on 30 March 2021.

360
consequences of the observed weight gain could occur.361 Part of this work was presented at the Afravih conference 8-11 November 2020 (Abstract number 2000241).