Image credit: Kenny Eliason, Unsplash
People living with HIV have an increased risk of cardiovascular disease (CVD) compared with people not living with HIV, which is attributable to traditional cardiovascular risk factors such as hypertension, hyperlipidemia, tobacco use, and obesity as well the impacts of antiretroviral therapy (ART).1 Worse CVD outcomes have been associated with particular ARTs. In 2007, the Data Collection on Adverse Events of Anti-HIV Drugs study group found that protease inhibitors (PIs) were associated with elevated risk of myocardial infarction (MI); however, this association is not as appreciated in more recent generations of PIs.1,2 Within non-nucleoside reverse transcriptase inhibitors (NNRTIs), abacavir has been most associated with increased risk of MI, but more recent data have been mixed.3 Integrase strand transfer inhibitors (INSTIs) have been associated with weight gain and hypertension, though the impact on cardiovascular outcomes has been unclear.4-6
In 2020, O’Halloran et al from the University of Washington demonstrated a decreased risk of major adverse cardiac events (MACE) in individuals started on INSTIs compared with non–INSTI-based regimens using data from 2008 to 2015.7 The study was limited by a short follow-up period and captured the introduction of INSTI-based regimens into mass adoption. INSTI-based regimens have become more widely utilized since 2015, and the most recent Guidelines for the Use of Antiretroviral Agents in Adults and Adolescents With HIV from the US Department of Health and Human Services recommends mostly INSTI-based regimens for initial treatment of HIV, with PI-based and NNRTI-based regimens generally being considered only in certain clinical scenarios.8Though the study from O’Halloran et al had a promising finding for INSTI-based regimens, it is critical to continue to reassess the cardiovascular impacts of INSTIs as a class as well as individual medications. The same investigatory group from the University of Washington completed an updated analysis to assess the impacts of initiation of different ART regimens on MACE.9
Their retrospective observational study used data from commercial and multistate Medicaid databases from 2008 to 2020. The study population were all adults who were initiated on ART from January 1, 2008, through June 30, 2020, who had at least 6 months of continuous health care enrollment prior to initiation on stable ART and who were consistently in possession of ART. Consistent ART possession was defined as having more than 80% of days with possession of medications for at least 180 days and never having more than 60 days without a part of their ART regimen. Individuals were excluded if they had MACE before the start of first stable ART regimen. Two treatment group classification systems were utilized. The first was a 3-level analysis based on medication class: PI based, NNRTI based, and INSTI based. The second was a 6-level analysis with INSTI medications separated individually (raltegravir, elvitegravir, dolutegravir, and bictegravir) compared with PI-based and NNRTI-based regimens. Using both classification systems, the risk of MACE was compared across either the 3 or 6 groups by propensity score models.
The baseline characteristics of the cohort were a median age of 40 years, 23% female, 15% with hypertension, 6% with diabetes, 2% with hepatitis C, and 10% reported tobacco use. The ART distribution was 16% PI based, 39% NNRTI based, and 45% INSTI based. The distributions within the INSTI-based regimens were 14% raltegravir, 40% elvitegravir, 30% dolutegravir, and 16% bictegravir. Most PI-based regimens utilized either atazanavir (43%) or darunavir (38%). The proportion of INSTI-based regimens increased through the study period from 5% to 97% in 2020. In the 3-level analysis, 14,692 people were in the NNRTI-based cohort, 6136 in the PI-based cohort, and 17,107 in the INSTI-based cohort. The INSTI cohort had the highest percentages of hypertension (17.4%), diabetes (7.0%), obesity (11.4%), tobacco use (16.0%), and use of lipid-lowering medications (16.1%). MACE occurred in 418 individuals (1.1%), which broke down into 199 (1.2%) among INSTI initiators, 87 (1.4%) among PI initiators, and 132 (0.9%) among NNRTI initiators within 48 months of medication start. Compared with NNRTI initiators, patients started on PIs and INSTIs had a higher risk of MACE. PI initiators had significantly higher rates of MACE at 12, 18, and 24 months compared with NNRTI initiators, which equated to 5, 5, and 6 more events per 1000 persons, respectively. INSTI initiators also had significantly higher risk of MACE in comparison with NNRTI initiators at 12 and 18 months, which would account for 2 and 3 more events per 1000 persons, respectively. In the 6-level analysis, there were no significant differences between NNRTI initiators and any particular INSTI initiator group.
There were several limitations with the study. HIV viral load data were not available, which could better help in estimating adherence to medications rather than access to the prescription. There was a lack of data on important CVD risk factors, such as race, body mass index, and family history of CVD, as well as a lack of data on management of established CVD. There was a high potential for confounding due to the observational nature of the data, which could have impacted the INSTI initiators cohort. INSTI initiators had higher rates of several cardiovascular diseases and risk factors for CVD, which likely increased baseline risk. This difference among INSTI initiators is likely multifactorial, with potential explanations being providers selecting INSTIs over PIs in patients with elevated risk for cardiovascular disease but still wanting a high barrier to resistant regimen, or for patients already on medications for CVD, INSTIs may have been chosen to decrease the risk of drug-drug interactions.
From the cardiovascular perspective, MACE occurred in 1.1% of the study population within 48 months after ART initiation, giving an estimated annual risk of 0.28% for MACE. According to the Framingham Risk Score with a composite outcome of coronary death, myocardial infarction, coronary insufficiency, angina, ischemic stroke, hemorrhagic stroke, transient ischemic attack, peripheral artery disease, and heart failure, for people ages 40 to 44 years without risk factors, there is an annual risk of 0.15% and 0.33% for women and men, respectively.10 Prior studies have found that the annual MACE risk in middle-aged adults with established atherosclerotic disease or multiple risk factors is 1.4%.11 With a median age of 40 years, the study population’s risk of MACE is similar to that of a general population not necessarily living with HIV. Of note, the study population was 77% male, so overall annual risk (0.28%) is closer to an annual risk of 0.33% (men) using the Framingham Risk Score and does not reach above that of MACE of 1.4% in a population with established atherosclerotic disease or high risk of cardiovascular disease. Given this, it is unclear whether total MACE occurring in this study is just from background cardiovascular risk rather than due to any of the ART regimens. In light of this and the limitations of this study, further research is warranted.
From the infectious disease clinician perspective, INSTI-based regimens have gained broad adoption as first-line therapy for HIV throughout the study period. Though it is critical that we continue to assess for the impacts of INSTIs as they have gained mass adoption, the benefits from an HIV management perspective outweigh the unclear risks from cardiovascular perspective.
Article reviewed: Luis Parra-Rodriguez, John M Sahrmann, Anne M Butler, Margaret A Olsen, William G Powderly, Jane A O’Halloran, Antiretroviral Therapy and Cardiovascular Risk in People With HIV in the United States—An Updated Analysis, Open Forum Infectious Diseases, Volume 11, Issue 9, September 2024, ofae485, https://doi.org/10.1093/ofid/ofae485
