Antiretroviral therapy (ART) has revolutionised the management of HIV, from becoming a chronic disease, and increasing life expectancy.1,2 This ageing of the population has brought an increase in the incidence of cardiovascular comorbidities in people living with HIV (PLWHIV), with a higher prevalence of cardiovascular risk factors (hypertension, diabetes, and dyslipidaemia).3,4 Although new drugs with better safety profiles are emerging, they are not exempt from adverse effects, such as changes in the lipid profile.5,6 Further, various studies have demonstrated that PLWHIV have an increased risk of developing heart failure, stroke, or myocardial infarction.7,8

Among the different available drugs, the most appropriate treatment for the patient, considers their individual characteristics, resistance tests, polypharmacy, and risk of interactions. Always urging individualisation, the national and international guidelines are a reference manual and guide for the management of HIV, recommending an early start after diagnosis9,10

The 2 single-tablet treatment regimens most widely used between 2015 and 2019 in our country were dolutegravir/abacavir/lamivudine (DTG/ABC/3TC) and elvitegravir/cobicistat/emtricitabine/tenofovir alafenamide fumarate (EVG/c/FTC/TAF).

Lipid parameters variations in the randomised controlled studies of EVG/c/FTC/TAF and DTG/ABC/3TC were minimal, without clinical or statistical significance.11–16 Furthermore, there are no studies, to the best of our knowledge, directly comparing these regimens.

The primary objective of this study is to compare the lipid profile changes and cardiovascular events in a real-world cohort of naïve and treatment-experienced HIV-infected patients who had been treated with EVG/c/FTC/TAF or DTG/ABC/3TC.

A retrospective observational single-centre study (university hospital in north-western Spain) in adult PLWHIV who received EVG/c/FTC/TAF or DTG/ABC/3TC from March 2015 to June 2019 was performed. The recommendations of the national clinical practice guidelines for each year and characteristics of each patient were decisive in the selected treatment.17 The follow-up time was defined since the beginning of the study ART until its suspension, voluntary abandonment, or loss of follow-up of the patient, death, or the end of the study.

All HIV-infected adults treated with EVG/c/FTC/TAF or DTG/ABC/3TC during the study follow-up time who signed the informed consent and with at least 1 medical visit were included in the study. Those patients transferred from other centres (hospitals or penitentiaries) (N = 93), voluntary withdrawal from treatment without 1 follow-up visit (N = 10), participating in clinical trials (N = 4), or those who did not sign the informed consent (N = 3) were excluded from the study.

This study was approved by the ethics committee of our institution. The study was carried out based on the Declaration of Helsinki of Good Clinical Practices.

Epidemiological, clinical, and immunovirological variables were recorded. Data in naïve and experienced patients of each of the study treatments were analysed.

Cardiovascular risk factors such as arterial hypertension, diabetes mellitus, smoking history, and dyslipidaemia (defined as TC ≥ 200 mg/dL, TG ≥ 150 mg/dL, or LDL-C ≥ 130 mg/dL) were recorded. The data were obtained from the laboratory analysis, the medical record registrations, or the pharmacology therapy (antihypertensive drugs, insulin or oral glucose-lowering agents, or lipid-lowering treatment).

The lipid profile (including total cholesterol (TC), triglycerides (TG), LDL cholesterol (LDL-C), HDL cholesterol (HDL-C), TC:HDL-C, and LDL-C/HDL-C ratios) at 48 and 120 weeks after initiating EVG/c/FTC/TAF or DTG/ABC/3TC was compared with the baseline. For both treatment groups, those patients with lipid-lowering drugs prescription were excluded from the lipid profile analysis. Only those patients who have reached weeks 48 and 120 at the cut-off of the study and with laboratory data in those study weeks were included in the lipid profile statistical analysis, remaining in the rest of analysis.

Cardiovascular events history (including myocardial infarction, heart failure, cerebrovascular accident, deep venous thrombosis, myocardiopathy, non-ST-segment elevation acute coronary syndrome, and ST-segment elevation myocardial infarction) before and during the follow-up were recorded.

While for the lipid profile analysis, those patients with lipid-lowering drugs prescription were excluded (Fig. 1), for the cardiovascular events analysis all patients were included.

Fig. 1.

Design of the study.

(0,39MB).

The statistical analysis was performed with SPSS v. 24 software. The quantitative variables were expressed as mean±standard deviation and the qualitative variables as percentage and frequency. Group differences were compared using the Pearson χ2 or Fisher's exact test and Student's t-test or the Mann–Whitney U-test, respectively, for categorical and continuous variables. Repeated measurements were compared using paired Student's t-test or Wilcoxon signed-rank test. Univariate analysis was performed with all the covariates. To identify the risk of discontinuation because of AEs by treatment a Cox regression analysis was performed. P-values of .05 or less were considered statistically significant.

This is the first real-world observation cohort comparing EVG/c/FTC/TAF and DTG/ABC/3TC in terms of lipid profile variations from baseline values, and cardiovascular events, to the best of our knowledge. The findings from this study showed a higher deterioration of the baseline lipid profile in the EVG/c/FTC/TAF group than that in the DTG/ABC/3TC group after 120 weeks of treatment. However, when we compared both groups, we observed significant differences only between treatment experienced patients, in TC and LDL-C values at 48 weeks, and in TC and HDL-C at 120 weeks, favouring the DTG/ABC/3TC group. While there were no differences in cardiovascular events after initiating the study. Furthermore, no significant differences were observed in the discontinuation of study treatments, with similar risk of interruption due to AEs. It must be taken into account that discontinuation due to hypercholesterolemia was only observed in the EVG/c/FTC/TAF group.

It is known that TAF has lost the lipid-lowering effect of tenofovir disoproxil fumarate (TDF), and has also been associated with the worsening in lipid parameters.18 This was confirmed in our study, where approximately half of our patients had TC and LDL levels above the normal range after 48 and 120 weeks, considering that less than a third of the patients presented those levels at baseline.

Pivotal studies of EVG/c/FTC/TAF in naïve patients found significant increases in TG, TC, HDL-C, and LDL-C compared with EVG/c/FTC/TDF (P < .001), but not in TC:HDL-C ratio (associated with cardiovascular disease risk) at weeks 48 and 144.11,12 These differences were attributed to significant reductions in plasma tenofovir concentrations with TAF. Further, Huhn et al. compared EVG/c/FTC/TAF and EVG/c/FTC/TDF treatment in naïve patients and observed an increase in median fasting lipids (TC, LDL-C, and HDL-C) at week 96 in the EVG/c/FTC/TAF group, with a similar TC:HDL ratio between the groups. Despite the significant increase in fasting lipids in the EVG/c/FTC/TAF group, they concluded that it does not crucially affect the cardiovascular risk profile.19

These results are similar to those we observed in our naïve patients treated with EVG/c/FTC/TAF when we compared the baseline data with those at weeks 48 and 120. As previously found in clinical studies, we also found no differences in the TC:HDL-C ratio between groups, cardiovascular events, or in the proportion of participants initiating lipid-modifying agents in naïve patients, although the proportion in our study (7.7%) was higher than in clinical trials (5.5%) for patients receiving EVG/c/FTC/TAF.

We found significant differences in all lipid values at week 48, including TC:HDL-C, and LDL-C/HDL ratios, but only in TC, LDL-C, and HDL-C at week 120 compared with those of the baseline data in experienced patients in EVG/c/FTC/TAF group. Kuo et al. observed similar results, with an increase in TG, TC, LDL-C, and HDL-C values, but not in the TC:HDL-C ratio at week 48.20 However, Huang et al. also found increases in the TC:HDL-C ratio at week 48 in patients who switched to EVG/c/FTC/TAF, as observed in our study.21

One study showed that patients who switched to DTG/ABC/3TC improved the lipid profile.22 On the other hand, there is a long-standing controversy surrounding the association between ABC and an increased risk of cardiovascular diseases in addition to the hypersensitivity reactions caused by ABC.23

In the DTG/ABC/3TC group, we found significant differences in TC and HDL-C levels in naïve patients at weeks 48 and 120, and in LDL-C only at week 48. However, patients with TG, TC, or LDL-C values above the normal range after 48 and 120 weeks were similar than those at baseline or better in the case of LDL-C. In pivotal studies, no clinically significant changes were observed over time (up to week 144) in the fasting lipid profile in the DTG group. These studies observed a small increase in LDL-C and TC levels and in the TC:HDL-C ratio, with a modest rise in TG, similar to what we observed in our patients.14,15

In pivotal trials of patients who switched to DTG, an increased in TC, LDL-C, HDL-C, and TG were observed at 24 and 48 weeks.16 This is totally different from what was observed in experienced patients, where no statistically significant differences were found in any lipid values or in the TC:HDL-C and LDL-C:HDL ratios during the follow-up. The results described by Bagella et al. in their study of patients who switched to DTG/ABC/3TC suggested an improvement in the lipid profile at 48 weeks, especially in the group of patients who changed from a PI-based regimen (reduction in TC and LDL-C, small increase in HDL-C, and a significant decrease in TG and TG:HDL ratio at weeks 24 and 48).22

Our study confirms the results obtained in previous works, where, as it is well known, the lipid profile worsens after starting ART, although the difference is more notable in the EVG/c/FTC/TAF group. Winston et al. compared patients switching to TAF/FTC or ABC/3TC, while continuing the same third drug. They did not find statistical differences in fasting TC, LDL-C, TG, or TC:HDL ratio at week 48; changes in HDL-C were significantly different yet not clinically relevant. Altogether, they considered both treatments alternatives, effective, and safe.24

No differences were observed in the prescription of lipid-lowering drugs, similar in both treatment groups and between naïve and experienced patients, although in the latter case, up to 26% of patients have been prescribed.

Due to ART and chronic HIV infection, PLWHIV have a long-life expectancy, which can lead to atherosclerotic lesions and cardiovascular disease. In the D:A:D study, they concluded that the use of ABC increased the risk of myocardial infarction.25 Otherwise, no differences were found in our study regarding the number of patients (naïve or experienced) who suffered a cardiovascular adverse event between the 2 drugs studied, despite significant differences in terms of presenting a cardiovascular risk factor.

The observed limitations of this study were the different baseline group characteristics, and the retrospective observational design that might have introduced uncontrolled bias (for instance, prescriber/selection bias is present and probably patients with renal disease were prescribed DTG/ABC/3TC preferentially to avoid tenofovir, yet renal patients probably have a greater chance of having hyperlipidaemia over time). It also could have introduced some bias the fact that have eliminated from the analysis of the lipid profile variation the patients who had been prescribed lipid-lowering treatment. Furthermore, the main limitation was the single-centre study design with a limited sample size. The conflicting results detected between trials and some cohorts, such as our data, could be partially explained by several factors, including study populations heterogeneity, patients´ habits, the follow-up time, and the research design.

In summary, when we compared both study groups, differences were found in TC in experienced patients favouring DTG/ABC/3TC during the follow-up. Accordingly, both regimens can therefore be alternatives in terms of lipid profile, tolerance, and incidence of cardiovascular events, hence future studies should monitor patients receiving TAF-based regimens more closely.

Sandra Rotea-Salvo contributed to the data collection, interpretation and writing of the article. Víctor Giménez-Arufe in data collection, interpretation and critical revision. Alejandro Martínez-Pradeda and Carla Fernández-Oliveira in data collection and critical revision. Álvaro Mena-De-Cea, Luis Margusino-Framiñán and María Isabel Martín-Herranz in the interpretation of results, critical review, and journal selection. And Purificación Cid-Silva, in the conception and design of the paper, and in the critical review of the article. All authors have approved the final version for publication.

This work was supported in part by the Professor Novoa Santos Foundation, A Coruña.

Elvitegravir/cobicistat/emtricitabine/tenofovir alafenamide fumarate and dolutegravir/abacavir/lamivudine are usually safe and well-tolerated drugs used in the management of HIV in clinical practice. However, it has been described that they modify the lipid profile and cardiovascular risk. Although EVG/c/FTC/TAF is no longer recommended as the first line for the management of HIV, the importance of this study lies in the direct comparison of TAF-based vs. ABC-based regimens.

A worsening of the lipid profile was observed in both treatment groups, being more pronounced among treatment experienced patients in TAF-based regimen. Despite these differences, the risk of cardiovascular events was similar in both groups, with no significant differences found. Both regimens continued to be well tolerated, with similar rates of drug-related adverse events or discontinuation. This is the first real-world observation cohort comparing both treatments in terms of lipid profile and cardiovascular events, providing more data to clinicians to choose the best regimen for each patient.