Major adverse cardiac event (MACE) and bleeding risks following percutaneous coronary intervention (PCI) for acute coronary syndromes (ACS) are not well defined in individuals with heart failure (HF). We followed 1,145 individuals in the Pharmacogenomic Resource to improve Medication Effectiveness Genotype Guided Antiplatelet Therapy cohort for MACE and bleeding events following PCI for ACS. We constructed Cox proportional hazards models to compare MACE and bleeding in those with versus without HF, adjusting for sociodemographics, comorbidities, and medications. We also determined predictors of MACE and bleeding events in both groups. 370 (32%) individuals did and 775 (68%) did not have HF prior to PCI. Mean age was 61.7 ± 12.2 years, 31% were female, and 24% were African American. After a median follow-up of 0.78 years, individuals with HF had higher rates of MACE compared to those without HF (48 vs. 24 events per 100 person years) which remained significant after multivariable adjustment (hazard ratio [HR] 1.31, 95% confidence interval [CI] 1.00 to 1.72). Similarly, bleeding was higher in those with versus without HF (22 vs. 11 events per 100 person years), although this was no longer statistically significant after multivariable adjustment (HR 1.29, 95% CI 0.86 to 1.93). Diabetes and peripheral vascular disease were predictors of MACE, and end-stage renal disease was a predictor of bleeding among participants with HF. MACE risk is higher in individuals with versus without HF following PCI for ACS. However, the risk of bleeding, especially among those with end-stage renal disease , must be considered when determining post-PCI anticoagulant strategies.
Ischemic heart disease is a leading cause of death among those living in the United States and is associated with both an almost 8-fold increased risk of developing heart failure (HF). It is an important cause of HF hospitalization and death, as patients with HF and coronary artery disease (CAD) have a higher 10 year mortality than HF patients without concomitant CAD. , HF patients who experience an ischemic event have an almost 3-fold increased risk of recurrent ischemic events as compared to HF patients who do not experience an event and preventing these events has been the rationale for long-term dual-antiplatelet therapy (DAPT) regimens. Major bleeding is the most common complication of DAPT therapy with a 3 to 4 times higher risk of all-cause mortality, an increase in the rate of in-hospital mortality, and 30-day major adverse cardiac events (MACE). Understanding the predictors of MACE and major bleeding in individuals receiving PCI for ACS is imperative to mitigate the risks associated with PCI and the medical therapy that follows. MACE events are most often associated with traditional cardiovascular risk factors; however, prior history of HF is an often under-represented cohort in the ACS literature. Advanced age, female sex, and renal insufficiency have been associated with an increased risk of bleeding while taking DAPT, however, the risk of bleeding while on DAPT in patients with HF is unknown. In this study, we sought to compare the risk and characterize predictors of MACE and major bleeding events for patients with and without HF.
Methods
The Pharmacogenomic Resource to improve Medication Effectiveness Genotype Guided Antiplatelet Therapy is a prospective cohort study conducted under the approval of the Institutional Review Board of the University of Alabama at Birmingham (UAB). Patients eligible for enrollment included those who underwent PCI and were subsequently admitted to UAB Hospital. As a component of the study genetic data was obtained and thus participants must be willing and able to provide consent. A total of 1,596 participants were enrolled between April 2014 and November 2019. Of these, 1,145 participants received PCI for an ACS indication and were included in this analysis. ACS, defined as unstable angina, non-ST segment elevation MI, or ST segment elevation MI, was determined by a fellow in the cardiovascular disease fellowship under direct supervision of a board-certified cardiology faculty.
After enrollment, a trained coordinator completed a structured case report form. Demographic information was abstracted from the medical record, including each participant’s age, sex, race, past medical history, and cardiovascular risk factors. An interview was conducted to document the participant’s health insurance status, level of education, smoking history, current employment, home medications, medication adherence, physical activity, marital status, and home ownership.
Procedural characteristics were obtained at the time of PCI. These included indication for PCI, vital signs, access site, contrast volume, number of obstructed coronary arteries, type of stents placed, and number of stents placed. Information about the severity of each participant’s coronary disease was obtained via chart abstraction, with obstructed coronary disease classified as >70% stenosis. Medications given during and after the PCI procedure, including antiplatelet therapy, were documented.
Participants were classified based on the presence or absence of HF prior to undergoing PCI. HF was defined as having an echocardiogram showing a left ventricular ejection fraction of <50% prior to PCI, if there was no EF documented, then HF was defined as a medical history of HF or a brain naturietic peptide level of ≥400 mg/dl prior to PCI.
Participants were followed for up to 1 year after PCI. UAB medical records were reviewed and changes in medication use, laboratory parameters, hospitalizations, MACE, and major hemorrhage (HEM) during follow-up were documented. We requested physical and/or electronic medical records from participants’ primary care physicians, cardiologists, and/or outside hospital records in case of hospitalization. This was done to ensure encounters outside of the UAB health system were obtained and recorded. All potential MACE and HEM events were adjudicated by expert clinician adjudicators after review of all available medical records.
MACE events were defined as the composite of death (both cardiovascular and non-cardiovascular), non-fatal myocardial infarction (MI), non-fatal ischemic stroke or transient ischemic attack (TIA), and stent thrombosis. MI was diagnosed based on an increase in the cardiac troponin above the 99 th percentile upper reference limit in addition to ischemic symptoms or new electrocardiographic findings suggestive of myocardial ischemia. Stent thrombosis was defined as evidence of thrombosis on invasive angiogram. Stroke and TIA were diagnosed based on the presence of a focal neurologic deficit plus imaging evidence of new focal lesion, or new focal deficit defined as TIA by a neurologist. Death was determined through both the inpatient and outpatient medical record.
HEM events were defined as a composite of intracranial HEM, gastrointestinal (GI), or other HEM that resulted in significant hemodynamic compromise requiring treatment, as defined by the Global Utilization of Streptokinase and Tissue Plasminogen Activator for Occluded Arteries (GUSTO) criteria. The GUSTO criteria classified intracerebral HEM and hemodynamic compromise as evidence of severe or life-threatening bleeding, with the identifier of major bleeding being hemodynamic compromise.
The data were summarized using absolute number and percent of total for categorical variables and mean and standard deviation for continuous variables. To compare differences between groups, we used t tests for continuous variables and χ 2 -tests for categorical variables. Next, we calculated incidence rates for MACE and major bleeding events in those with and without HF and compared these rates in those with versus without HF using incident rate ratios (IRR) and 95% confidence intervals (CI).
In order to determine the independent association of HF status with MACE and major bleeding events, we constructed Cox proportional hazards models to calculate hazard ratios (HR) and 95% CI for participants with versus without HF. All covariates listed in Table 1 were added to the model if they varied by HF status in bivariate analyses using a p value of <0.05. In model 1, we added demographics, including age, black race, female sex as well as past medical history, including hypertension, diabetes mellitus, atrial fibrillation, CAD, end-stage renal disease (ESRD), and peripheral vascular disease (PVD), insurance status, and alcohol use. Then, in model 2, we included components of the first model plus medications prescribed at discharge, including DAPT (aspirin with clopidogrel or aspirin with ticagrelor), calcium-channel blocker, aldosterone inhibitor, diuretics, and oral anticoagulation (warfarin, rivaroxaban, or apixaban). Finally, we constructed Cox proportional hazards models, separately in those with and without HF, to determine which factors predicted MACE and major bleeding events. Factors included in the multivariable model were those known to influence MACE and bleeding (age, race, gender, DAPT therapy , ), in addition to factors listed in Table 1 that were associated with the outcome of interest in univariate analyses with a p-value <0.20. All analyses were performed using SAS version 9.4.
| Variable | Heart Failure | ||
|---|---|---|---|
| Yes (N = 370) | No (N = 775) | p Value | |
| Age (years) | 62.7 ± 11.82 | 61.0 ± 12.29 | 0.029 |
| Women | 109 (29%) | 250 (32%) | 0.340 |
| Black Race | 94 (25%) | 181 (23%) | 0.448 |
| Left Ventricular Ejection Fraction (%) | 36.9 ± 9.4 | 58.0 ± 5.7 | <0.001 |
| Systolic Blood Pressure (mm Hg) | 138.6 ± 22 | 144.1 ± 24.3 | <0.001 |
| Diastolic Blood Pressure (mm Hg) | 83.2 ± 14.4 | 84.6± 15.2 | 0.138 |
| Brain Naturietic Peptide (pg/dl) | 1022.9 ± 1902.1 | 294.3 ± 723.3 | <0.001 |
| Hypertension | 326 (88%) | 648 (84%) | 0.046 |
| Dyslipidemia | 271 (73%) | 546 (70%) | 0.329 |
| Diabetes Mellitus | 293 (79%) | 189 (24%) | <0.001 |
| Atrial Fibrillation | 66 (18%) | 71 (9%) | <0.001 |
| Coronary Heart Disease | 279 (75%) | 404 (52%) | <0.001 |
| Chronic Kidney Disease | 89 (24%) | 112 (14%) | <0.001 |
| End-Stage Renal Disease | 32 (9%) | 21 (3%) | <0.001 |
| Peripheral Vascular Disease | 58 (16%) | 74 (10%) | 0.002 |
| Insurance status | |||
|---|---|---|---|
| Private | 139 (38%) | 373 (48%) | <0.001 |
| Medicare | 170 (46%) | 258 (33%) | <0.001 |
| Medicaid | 144 (39%) | 61 (8%) | <0.001 |
| Education level | |||
|---|---|---|---|
| <High-School | 199 (54%) | 404 (52%) | 0.244 |
| >High-School | 142 (38%) | 311 (40%) | 0.571 |
| Alcohol Use | 92 (25%) | 253 (33%) | 0.006 |
| Smoker | 104 (28%) | 223 (29%) | 0.690 |
| Income <$50,000 | 148 (40%) | 301 (39%) | 0.385 |
| Medications at discharge | |||
|---|---|---|---|
| Aspirin | 352 (95%) | 753 (97%) | 0.183 |
| DAPT | 336 (93%) | 732 (96%) | 0.046 |
| DAPT using Clopidogrel | 233 (63%) | 454 (59%) | 0.044 |
| DAPT using Ticagrelor | 103 (28%) | 278 (36%) | 0.009 |
| Statin Therapy | 336 (91%) | 710 (92%) | 0.922 |
| Beta-Blocker | 314 (85%) | 652 (84%) | 0.454 |
| Calcium Channel Blocker | 61 (16%) | 192 (24%) | 0.002 |
| ACEi/ARB | 187 (52%) | 404 (53%) | 0.737 |
| Spironolactone | 51 (14%) | 40 (5%) | <0.001 |
| Diuretic | 183 (51%) | 220 (29%) | <0.001 |
| Insulin at Discharge | 93 (26%) | 127 (16%) | <0.001 |
| Oral Anticoagulant | 57 (16%) | 47 (6%) | <0.001 |
| Warfarin | 36 (10%) | 28 (4%) | <0.001 |
| Apixaban | 21 (6%) | 19 (3%) | 0.005 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
