MI Risk in PWH: New Data From CROI 2022

People with HIV (PWH) have a greater risk of myocardial infarction (MI) than those who are HIV uninfected. Multiple factors contribute to these elevated MI rates, including the traditional risks that drive cardiovascular disease (CVD) in the general population, such as family history, smoking, hypertension, diabetes mellitus, and dyslipidemia. In addition, certain antiretrovirals and persistent immune activation are HIV-specific factors proposed to also heighten MI risk in PWH.

Happily, in the United States, a narrowing in the rates of MI between PWH and people without HIV has been observed, which may be explained by better lipid management, smoking cessation, and use of antiretrovirals with less impact on lipids. However, data presented at the 2022 Conference on Retroviruses and Opportunistic Infections (CROI) suggest that gains made in MI prevention for PWH have stalled in recent years.

MI Risk by HIV Status in San Francisco and Boston
Silverberg and colleagues compared MI rates in PWH and those without HIV using clinical data from Kaiser Permanente Northern California (with ~30,000 PWH) and Mass General Brigham–Partners (with ~7000 PWH). Given imbalances between these groups of patients, propensity score matching used 3 or 4 people without HIV to match to each PWH on demographic characteristics and the Framingham risk calculation components. New MI rates were examined by HIV status over 2 periods: 2005-2009 and 2010-2017.

This was a large study. For the 2005-2009 period, 4280 PWH and 14,059 people without HIV were analyzed; for the 2010-2017 period, 5121 PWH and 15,359 people without HIV were included. They were mostly male and of White race, with a mean age of approximately 44 years. For the PWH, 39% had a baseline plasma HIV-1 RNA level >400 copies/mL between 2005-2009, but only 23% had a baseline HIV-1 RNA above this level between 2010-2017.

During 2005-2009, the cumulative incidence of MI was nearly identical in those with and without HIV, with a 1.1% risk of MI over 5 years. However, during the 2010-2017 period, the MI incidence diverged, with a slight uptick in 5-year risk to 1.2% in PWH and a decline to 0.9% in those without HIV (P = .03). In a multivariable analysis, the HR for PWH for MI in 2010-2017 was 1.6 (95% CI: 1.1-2.4; P = .007), meaning PWH had a 60% greater risk of MI. These trends generally held up in several sensitivity analyses.

Although the Kaplan-Meier curves and multivariable analysis–derived HR for 2010-2017 may look scary, the actual number of MIs provides perspective. Of the 5121 PWH included in the 2010-2017 analysis, there were 43 MIs; among the 15,359 without HIV included during this period, there were 87 MIs.

Interpretation
Overall, the analysis paints a picture of a reduction in MI risk over the most recent years for people without HIV that was not enjoyed by those with HIV. Dr Michael Silverberg speculated that newer HIV therapies, specifically the adoption of integrase inhibitors as an antiretroviral mainstay, could have prevented PWH from realizing the reduction in seen in those without HIV. After all, across the study periods, the use of HIV therapy at baseline increased from 76% to 88%, with a dramatic increase in integrase inhibitor treatment.

However, there is no clear reason integrase inhibitors would be a drag on continued improvements in MI rates for PWH, especially as they largely replaced protease inhibitors, which as a class have been associated with CVD. Weight gain has been associated with integrase inhibitor–based antiretroviral regimens, but neither weight nor BMI data were presented.

Many other confounders could be at work here. Foremost in my mind, the transition to reliance on integrase inhibitors was accompanied by a shift from tenofovir disoproxil fumarate to tenofovir alafenamide. The former is well recognized for reducing lipids. Atazanavir is a protease inhibitor that—via induction of hyperbilirubinemia—is associated with reduced risk of MI; its use fell precipitously over the study periods. Substance use, especially of stimulants, also has been linked to MI, and disproportionate use in PWH could contribute to the finds. These and other potential factors influencing risk of MI should be examined within this robust dataset.

Your Thoughts?
What do you think explains the difference in MI risk by HIV status in the more recent period? What measures should be taken to achieve reductions in MI risk for PWH on par with those without HIV? Join the conversation by posting your comments.