Change in Blood Pressure, Weight, and Other Cardiovascular Disease Risk Factors After Switch From Protease Inhibitors to Dolutegravir: Post hoc Analysis of the 48-week Randomised Second-line Switch to Dolutegravir Trial

Abstract

Graphical Abstract

This graphical abstract is also available at Tidbit: https://tidbitapp.io/tidbits/change-in-blood-pressure-weight-and-other-cardiovascular-disease-risk-factors-after-switch-from-protease-inhibitors-to-dolutegravir-post-hoc-analysis-of-the-48-week-randomised-2sd-trial-4ce800fb-ada6-47d5-80fb-c3296274967a?utm_campaign=tidbitlinkshare&=ITP

Open in new tabDownload slide

The use of integrase inhibitors, including dolutegravir (DTG), for the treatment of human immunodeficiency virus (HIV) has been associated with increased weight gain compared to other antiretrovirals (ARVs) drug classes [1]. Along with weight gain, there are conflicting data on potential associations between use of DTG and incident hypertension (HTN), diabetes mellitus, and cardiovascular disease [2]. Considering the broad use of DTG-containing regimens in sub-Saharan Africa, it is important to clarify causation and effect size of potential associations with cardiovascular risks [3].

The Second-line Switch to DTG (2SD) study was an open-label, multicenter, randomized, active-controlled, 48-week noninferiority trial comparing the efficacy and safety of switching from a ritonavir-boosted protease inhibitor (PI/r) to DTG versus continuing PI/r among participants virally suppressed on PI/r plus 2 nucleoside reverse transcriptase inhibitors (NRTIs). Primary 48-week results demonstrated non-inferior efficacy of DTG compared to PI/r [4]. 2SD safety analysis reported all adverse events occurring in ≥5% of participants in either study arm, including investigator-reported hypertension of 10.6% (41/385) in the DTG arm and 5.8% (22/379) in the PI/r arm [4]. With a hypertension signal being reported from other studies, we reviewed 2SD results using a standardized definition of HTN and performed post hoc analysis on incident HTN in participants switching from PI/r to DTG compared to those remaining on PI/r over the 48-week study period.

METHODS

The 2SD trial was conducted at 4 sites in Kenya. Eligible participants were aged ≥18 years and virally suppressed on a second-line antiretroviral regimen containing 2 NRTIs and a PI/r. Exclusion criteria included baseline conditions requiring regimen change, advanced kidney and liver disease, and grade 3–4 lipid abnormalities. Participants were randomized 1:1 to switch to DTG or continue baseline PI/r. The study was approved by the joint ethics review committee of Kenyatta National Hospital and University of Nairobi and that of Jaramogi Oginga Odinga Teaching and Referral Hospital. All participants provided written informed consent. The 2SD study is registered at clinicaltrials.gov, NCT04229290.

Visits were scheduled at enrollment and weeks 4, 12, 24, 36, and 48, during which blood pressure (BP), weight, waist circumference, and hip circumference were measured; height was measured at enrollment. BP was measured using an automated sphygmomanometer with adult standard or adult large cuff size for those with higher body mass index, with the participant seated with both feet flat on the floor and their back and arm supported. If the first measure was ≥140 mm Hg systolic or ≥90 mm Hg diastolic, then a repeat measure was taken after a few minutes of quiet sitting with the lowest of the measures used for analysis. Laboratory parameters included fasting glucose and lipids at screening and weeks 24 and 48.

The primary endpoint of this post hoc analysis is incident HTN, defined as having a systolic BP ≥140 mm Hg or a diastolic BP ≥90 mm Hg during at least 2 study follow-up visits, and excluding any participant classified as having HTN at baseline (on treatment for HTN at enrollment, or a systolic BP ≥140 mm Hg or a diastolic BP ≥90 mm Hg at screening or enrolment and confirmed elevated during at least 1 other study visit).

Secondary outcomes include: incident fasting hyperglycemia (fasting glucose ≥7.0 mmol/L at week 24 or 48, excluding participants on treatment for diabetes at enrollment or with fasting glucose ≥7.0 mmol/L at enrollment); incident dyslipidemia (Division of AIDS [DAIDS] grade 2–4 fasting lipid abnormality at week 28 or 48, excluding participants with DAIDS grade 2–4 fasting lipid abnormality at enrollment or who were on treatment for dyslipidemia at enrollment); change in weight and body mass index; weight gain ≥5% from enrollment; change in waist-to-hip circumference ratio; and atherosclerotic cardiovascular disease (ASCVD) 10-year risk prediction score (calculated for all participants aged 40–79 years using the American College of Cardiology equation with input of age, sex, race, systolic BP, total cholesterol, high-density lipoprotein cholesterol, history of diabetes, smoking status, and HTN treatment status). ASCVD 10-year risk was categorized as low (<5%), borderline (5%–7.4%), intermediate (7.5%–19.9%), or high (≥20%).

Bivariate analysis was performed comparing change in cardiovascular risk parameters between study arms over the 48-week follow-up period. Logistic regression was performed to assess if incident HTN was independently associated with study treatment, age, sex, and weight gain ≥5% from enrollment.

Data used for this analysis are available under the terms of the Creative Commons Zero “No rights reserved” data waiver (CC0 1.0 Public domain dedication) and can be accessed with permission from the Principal Investigator (L.A.O.) through the Harvard Dataverse [5].

RESULTS

Between February and September 2020, 791 participants (397 DTG, 394 PI/r) enrolled into the study; of these, 777 completed 48 weeks of follow-up and are included in the analysis of change in cardiovascular risk factors. Recruitment of participants from the randomized clinical trial into the analysis of incident hypertension and dyslipidemia is shown in Supplementary Figure 1 and S2. All participants were black African, 516 (66.4%) were female, median age was 46 years (range, 19–74). Antiretrovirals regimen at enrollment included 615 (79.2%) on atazanavir/r and 162 (20.8%) on lopinavir/r, with baseline NRTIs consisting of tenofovir disoproxil fumarate (TDF) plus lamivudine (3TC) among 411 (52.9%), zidovudine/3TC among 332 (42.7%) and abacavir/3TC among 34 (4.4%), with median duration on antiretroviral therapy of 10.7 years (interquartile range 7.8 to 13.3). Baseline cardiovascular risk parameters among participants included smoking in 6 (0.8%), obesity in 146 (18.8%), HTN in 190 (24.5%), diabetes mellitus or hyperglycemia in 44 (5.7%), dyslipidemia in 120 (15%), and 85 (14.4%) had intermediate or high 10-year ASCVD risk, with baseline characteristics balanced between arms (Table S1). Information on family history of hypertension, diet, and physical activity was not available. No patients were on weight-lowering treatments, 3.5% of participants on the DTG arm and 2.6% on the PI arm were on medications for management of diabetes, and only 1 participant on the DTG arm and 3 on the PI arm were on medication for dyslipidemia. Prior NRTI and non-NRTI use was well matched between study arms. Concomitant medications at baseline were well balanced between arms and included diuretics in 4.6%, vasodilator agents in 6.7%, antidepressants in 1, and steroids in 3 patients (Supplementary Table 2).

By week 48, there was no difference in incident HTN between study arms (48/292 [16.4%] on DTG; 46/295 [15.6%] on PI/r; P = .868), Table 1. There was also no difference in mean change in systolic BP (+3 mm Hg on DTG; + 5 mm Hg on PI/r, P = .692) or diastolic BP (+5 mm Hg on DTG; + 4 mm Hg on PI/r, P = .099). There was no difference in time to development of hypertension in the 2 arms (Figure S3).

Participants who switched to DTG had higher mean increase in weight than those who remained on PI/r (+1.5 kg on DTG; +1.0 kg on PI/r; P = .019) and were more likely to gain ≥5% of enrollment weight (32% and 21%, respectively; P < .001). There was no correlation between weight gain and increase in systolic or diastolic BP over the 48-week study period (Supplementary Figure 4).

There was no difference between arms in incident fasting hyperglycemia, incident dyslipidemia, change in waist-to-hip circumference ratio, or change in ASCVD risk score (Table 1).

In multivariate analysis, incident HTN was associated with increasing age (adjusted odds ratio [AOR], 1.03; 95% confidence interval [CI]: 1.01–1.05; P = .006), but not with switching to DTG (AOR, 1.02; 95% CI: .66–1.59; P = .921), gaining ≥5% body weight (AOR, 1.25; 95% CI: .75–2.02; P = .378), or being male (AOR, 1.11 [0.69–1.75], P = .661) (Supplementary Table 3).

DISCUSSION

In this post hoc analysis of HTN and other cardiovascular risk factors in the 48-week 2SD trial, there was no difference in incident HTN between participants randomized to switch from PI/r to DTG compared to those who remained on PI/r, and there was no difference in change in systolic or diastolic BP over the study period. These findings are consistent with those of the NEAT-022 trial, which randomized adults with high cardiovascular risk scores who were virally suppressed on PI/r to either switch from PI/r to DTG immediately or after 48 weeks, and found no association between switch to DTG and incidence of HTN [6].

Our results differ from published results of raised BP (≥130 mm Hg systolic or ≥85 mm Hg diastolic) over 192 weeks in the NAMSAL trial, which randomized ARV-naïve participants to DTG + TDF/3TC or efavirenz (EFV)+TDF/3TC, demonstrating higher rates of raised BP with DTG compared to EFV [7]. The RESPOND cohort found higher incidence of HTN with integrase inhibitors when compared to NNRTIs, but not when compared to PIs [8]. Our results also differ from other observational studies that report higher risk of HTN or increase in BP with use of DTG or integrase inhibitors compared to other ARVs [9–11]. The differences could be attributed to different comparator arms (some studies comparing DTG to EFV or other integrase inhibitors rather than to PI), different NRTI components of the regimens, different study populations (particularly those conducted in Europe or North America with a small proportion of black participants), and a weight-gain effect on BP that was not appreciated in our study [12].

Although we found higher weight gain in participants switched to DTG, there was no difference in change in waist-to-hip circumference ratio, incident fasting hyperglycemia, incident dyslipidemia, or change in ASCVD risk score among participants who switched to DTG compared to those who remained on PI/r.

The overall incidence of HTN was high at 16%; this underscores the need to integrate hypertension prevention and treatment services into HIV programs in sub-Saharan Africa.

A limitation of our study is that there were no participants on tenofovir alafenamide, so our results may not apply to people on the combination of DTG plus tenofovir alafenamide plus 3TC or emtricitabine. Another limitation is the relatively short study period of 48 weeks, which may not be long enough to appreciate ARV-related anthropometric and cardiovascular risk changes. The primary strength of our findings is that the randomized clinical trial design minimizes the potential bias that may be seen with observational studies.

In conclusion, there was no difference in development of incident HTN between participants switched to DTG compared to those remaining on PI/r in the 2SD study. We observed weight gain with the switch to DTG, but there was no difference in other cardiovascular risk factors, including incident fasting hyperglycemia, dyslipidemia, change in waist-to-hip circumference ratio, or change in ASCVD risk score.

Supplementary Data

Supplementary materials are available at Clinical Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.

Notes

Acknowledgments. The authors thank the study participants, study staff, staff at participating sites, and the members of the Data and Safety Monitoring Board.

Financial support. This work was supported by ViiV Healthcare through an investigator-initiated grant to L. A. O. [212984] and was sponsored by the University of Nairobi. The Ministry of Health Kenya and ViiV Healthcare provided antiretroviral medications for the study. ViiV Healthcare and the Ministry of Health Kenya had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Data availability. Harvard Dataverse. Replication Data for: Change in blood pressure, weight and other cardiovascular disease risk factors after switch from protease inhibitors to dolutegravir: Post-hoc analysis of the 48-week randomised 2SD trial. DOI: https://doi.org/10.7910/DVN/OFKVQW [5].

This project contains the following underlying data:

This is the analysis dataset from an open-label, randomized, active-controlled, multicenter noninferiority trial conducted by the University of Nairobi. The trial randomized 795 participants ≥18 years old from among people receiving outpatient HIV care at 4 study sites in Kenya, of whom 791 were exposed to treatment and 777 completed 48 weeks of follow up and are included in the analysis dataset. The variables contained in this dataset from screening, enrollment, week 24, and week 48 include: demographics, ARV history, medical history, concomitant medications, BP, weight, height, waist circumference, hip circumference, and laboratory investigations (HIV-1 RNA viral load, creatinine, hepatitis B surface antigen, fasting glucose, fasting lipid profile).

References

Author notes