MI Risk in Some With HIV Underestimated by ACC/AHA Calculator

Marlene Busko

December 28, 2016

CHICAGO, IL — In a cohort of patients with HIV, the 2013 American College of Cardiology/American Heart Association (ACC/AHA) pooled cohort equations to predict 10-year risk of atherosclerotic cardiovascular disease (ASCVD) were generally adequate in predicting MI risk, especially in white men who made up the majority of the cohort[1]. However, the tool was less effective in predicting MI risk in women or black men and underestimated MI risk in patients with fewer traditional risk factors.

Moreover, two models with added HIV-specific variables did not perform any better, in the study by Dr Matthew J Feinstein (Northwestern University Feinberg School of Medicine, Chicago, IL) and colleagues published online December 21, 2016 in JAMA Cardiology.

Because pooled cohort equations underestimate risk at the lower end of the risk spectrum, this suggests that HIV patients with a lower apparent risk of MI would benefit from statin therapy, Feinstein told heartwire from Medscape in an email. "For instance, if a patient with HIV is estimated as having a 10% 10-year risk for ASCVD using the pooled cohort equation, we know that their actual risk for MI is likely quite a bit higher—maybe 15% to 20%. Thus, this patient would likely benefit from statin therapy, assuming no contraindication and proper attention to drug–drug interactions."

Further studies are needed "specifically in groups in which the number of MIs was consistently underpredicted," Dr Colleen Hadigan (National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, MD) and colleagues urge in an accompanying editorial[2].

With current antiretroviral therapy, patients with HIV have a life expectancy approaching that of the general population, they continue, so these patients are faced with a double jeopardy: HIV infection and an increased risk of dying from CVD. Thus, further research is needed "to develop comprehensive management strategies that appropriately identify and treat those at risk of poor cardiovascular outcomes, thus mitigating the potential for the double jeopardy of HIV infection and increased risk of CVD."

Standard and Two New Tools to Predict MI Risk in HIV Patients

People with HIV have nearly twice the risk for MI compared with those without HIV, Feinstein and colleagues write. However, it is unknown how well the ACC/AHA pooled cohort equations—based on risk factors of sex, race/ethnicity, total cholesterol, HDL-cholesterol, smoking, diabetes, systolic BP, and antihypertensive treatment—predict ASCVD risk in patients with HIV.

The researchers aimed to investigate this and also evaluate and validate two HIV-specific MI risk estimation models, in a large, multicenter HIV cohort. The two models (which use slightly different statistical approaches) included variables from the ASCVD risk score plus statin use and HIV-specific variables (HIV viral load, CD4 lymphocyte count, antiretroviral therapy, and protease-inhibitor use).

Feinstein and colleagues identified 11,288 HIV-infected adults who received care as early as 1995 at one of five Centers for AIDS Research Network of Integrated Clinical Systems sites across the United States who had MI adjudication and complete baseline data.

Patients were mainly white males (6143), followed by black males (3107), black females (1277), and white females (761). They had a mean age of 41 years, and about 77% were receiving antiretroviral therapy.

During a mean follow-up of 4.1 years, 247 patients had an MI. Rates of MI were highest among black women (7.2 per 1000 person-years) and black men (6.9 per 1000 person-years), and lower among white men (4.4 per 1000 person-years) and white women (3.3 per 1000 person-years).

The pooled cohort equations adequately discriminated MI risk (Harrell C statistic 0.75, where <0.70, 0.70– 0.80, and >0.80 are considered inadequate, acceptable, and excellent discrimination, respectively). The two new risk-estimation models did not discriminate MI risk any better than this (Harrell C statistic 0.72 and 0.73).

The study end point was restricted to fatal and nonfatal MI, because there were no adjudicated stroke end points available for analysis. However, "this doesn't change the conclusion that there is substantial mismatch between observed and predicted ASCVD risks, whereby actual risks for HIV-infected persons exceed risks predicted by the pooled cohort equations," Feinstein observed. "In fact, if stroke end points had been included, this would have created even greater mismatch/underprediction."

Plots of predicted vs observed rates of MIs revealed a close match for white men, but a mismatch for women and black men. Risk estimates were likely to be more accurate in white men because the pooled cohort equations used to estimate risk were mostly derived from white male patients, Feinstein suggested.

People with a systemic illness such as HIV are more likely to have a type 2 MI (related to coronary artery spasm, coronary embolism, anemia, arrhythmias, hypertension, or hypotension) as opposed to a type 1 MI (related to plaque erosion or rupture), he added. Moreover, black men and women were more likely than white men and women to have a type 2 MI, which may partly explain why the pooled cohort equations underestimated the MI risk in women and black men.

The proposed models did not improve risk prediction probably because "there simply weren't enough events to get a precise enough estimate of the associations of variables/risk factors with MIs in this cohort even though it is fairly large," Feinstein speculated. "By comparison, the pooled cohort equations—which do pretty well at predicting ASCVD risk in the general population—were based on data from hundreds of thousands of individuals and over a million person-years of follow-up," he noted.

"I am still optimistic that HIV-specific models will be feasible and can improve risk prediction over existing models in the future, but we will need even larger data sources and more person-years of follow-up to obtain more precise model estimates," Feinstein said.

Feinstein had no relevant financial relationships; disclosures for the coauthors are listed in the article. Hadigan and coauthors reported no relevant financial relationships.

For more from theheart.org, follow us on Twitter and Facebook .


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: