Bleeding complications after percutaneous coronary intervention (PCI) have been associated with higher short and long-term mortality. Bivalirudin has been shown to reduce bleeding complications in patients who underwent PCI; however, the impact of anemia on bleeding complications and long-term mortality has not been studied. A total of 11,991 patients who underwent PCI over a period of 8 years with bivalirudin as the primary antithrombotic agent were included. Anemia was defined according to the World Health Organization definition. Bleeding complications were prospectively collected. Survival analysis was performed using multivariable Cox proportional hazards models. Of the 11,991 patients, 4,815 patients (40%) had baseline anemia. Major bleeding occurred in 3.3% of patients with anemia compared with 0.7% of patients without anemia (p <0.001) driven largely by transfusion events. In the overall study population, major bleeding was a significant predictor of mortality (hazard ratio [HR] 1.4, 95% confidence interval [CI] 1.04 to 1.8, p = 0.027) at a mean follow-up of 2.6 years (interquartile range 1.4 to 3.5). In patients with anemia, major bleeding remained an independent predictor of mortality (HR 1.5, 95% CI 1.1 to 2.0, p = 0.008); however, in patients without anemia, it did not (HR 1.25, 95% CI 0.52 to 3.03, p = 0.62). In patients who underwent PCI with bivalirudin therapy, major bleeding is associated with early and long-term mortality, which is more pronounced in patients with baseline anemia.
It is established that bleeding complications after percutaneous coronary intervention (PCI) are associated with short- and long-term adverse outcomes including mortality. Bivalirudin has been associated with reduced bleeding complications after PCI leading to a reduction in long-term mortality. Nearly a quarter of patients who underwent elective PCI and 40% of elderly patients who underwent PCI for acute myocardial infarction have been shown to have baseline anemia. These patients have a higher incidence of bleeding complications compared with those without anemia. The prognostic importance of anemia in relation to bleeding events among patients treated exclusively with bivalirudin has not been studied.
Methods
For the present study, all patients who underwent PCI from July 2002 to May 2010 with bivalirudin as the primary antithrombotic agent were included. As the aim of the study was to evaluate the long-term effects of bleeding complications after PCI using bivalirudin, 27 in-patient deaths were excluded. For patients who underwent repeat PCI, only the earliest index PCI was included (5,607 repeat procedures were excluded). Patients without a baseline hemoglobin (45) were also excluded. The institutional review board approved the study. Baseline routine laboratory values were measured before patients underwent PCI. After PCI, patients were followed daily by the trained research staff, and all pertinent laboratory values and in-hospital complications collected and recorded prospectively.
All patients were loaded with aspirin 325 mg, clopidogrel 300 to 600 mg, or prasugrel 60 mg before PCI. Bivalirudin was administered as an intravenous bolus of 0.75 mg/kg, followed by an infusion of 1.75 mg/kg/hour and discontinued at the completion of PCI in patients already on thienopyridine. Bivalirudin infusion was continued for 1-hour post-PCI in clopidogrel naïve patients. Administration of glycoprotein IIb/IIIa inhibitors (GPIs) was performed at the discretion of the operator. A bolus with infusion GPI strategy (0.25 mg/kg bolus abciximab followed by 0.125 μg/kg/min for 12 hours or 180 μg/kg × 2 boluses of eptifibatide followed by 2 μg/kg/min infusion for 12 hours) was used until December 2004 after which a bolus-only GPI strategy was adopted based on a study from our institution that showed a reduction in the incidence of bleeding and vascular complications without differences in ischemic outcomes compared with bolus plus infusion. Bivalirudin and GPI doses were adjusted for renal impairment as recommended by the manufacturer. Coronary interventional procedures were performed according to the standardized institutional techniques, through a femoral or radial approach. The choice of drug-eluting stent or bare-metal stent implantation was at the discretion of the operator. Femoral vascular closure devices were used, unless contraindicated according to the femoral arteriogram. Complete blood count and cardiac biomarkers were measured 6 and 12 hours post-PCI. Postprocedure antiplatelet therapy recommendations were standard. Aspirin indefinitely, clopidogrel or prasugrel for at least 1 month for bare-metal stents, and 1 year for drug-eluting stents.
All-cause mortality was ascertained using the National Social Security Death Index and the New York State interventional database by matching patients’ social security numbers to death index records. For the present study, deaths occurring through December 31, 2010, were ascertained.
Anemia was defined as a hemoglobin level <13 g/dl in men and <12 g/dl for women according to the World Health Organization criteria. Chronic kidney disease (CKD) was defined as glomerular filtration rate (GFR) less than 60 ml/min. GFR was calculated using the simplified modification of diet in renal disease formula. Post-PCI myocardial infarction was defined according to the thrombolysis in myocardial infarction criteria. Bleeding complications were defined using the Harmonizing Outcomes with Revascularization and Stents in Acute Myocardial Infarction trial criteria. Major bleeding was defined as intracranial or intraocular hemorrhage, bleeding at the access site with a hematoma 5 cm or larger or that required intervention, a decrease in the hemoglobin level of 4 g/dl or more without an overt bleeding source or 3 g/dl or more with an overt bleeding source, reoperation for bleeding, or blood transfusion.
Continuous variables are presented as mean ± SD. Categorical variables are presented as percentages. The chi-square test was used to compare differences between categorical variables. The independent samples t test was used to compare continuous variables with normal distribution, and the Mann-Whitney test was used to compare continuous variables without a normal distribution. Nonparametric survival analyses with log-rank tests were conducted to find whether the potential covariates had a univariate association with mortality. Of the continuous variables, only left ventricular ejection fraction was categorized using a single cutoff of <40%. Other continuous variables (age, body mass index, and GFR) were introduced in the models as continuous variables due to their normal distribution. Semiparametric (Cox) survival analyses were then conducted to assess the univariate effect of major bleeding on mortality and the extent of confounding or interaction of major bleeding with covariates statistically significant in univariate analysis. Similar analyses were repeated in the anemic and nonanemic subgroups. For in-hospital major bleeding outcome, Mann-Whitney and Fisher’s exact tests were used to find whether predictors of interest and the same set of potential covariates had a univariate association with major bleeding. Age, body mass index, female gender, diabetes mellitus, peripheral vascular disease, CKD, end-stage renal disease, active smoking, acute coronary syndrome (ACS), coumadin use, baseline anemia, and use of closure device were significant univariate predictors of in-hospital major bleeding. Logistic regression analyses were then conducted to assess the univariate effect of anemia and the extent of confounding or interaction of covariates that were statistically significant in univariate analysis. p Values of <0.05 were considered to indicate statistical significance. All statistical analyses are performed using SPSS software version 21.0 (SPSS, Chicago, Illinois).
Results
A total of 11,991 patients underwent PCI and were discharged between July 2002 and May 2010 with bivalirudin as the primary antithrombotic agent. Baseline and procedural characteristics of the total cohort, anemic, and nonanemic subgroups are listed in Tables 1 and 2 . A total of 4,815 patients (40%) had baseline anemia according to the World Health Organization anemia definition. These patients presented more frequently with ACS and tended to be older, female gender, African-American, with lower body mass index and body surface area. There was a higher incidence of hypertension, diabetes mellitus, peripheral vascular disease, congestive heart failure, previous revascularization, and CKD in patients with anemia. In addition, medical therapy with β blocker, angiotensin-converting-enzyme inhibitor, and coumadin was more common in this group, whereas the use of GP IIb/IIIa, aspirin, and statin was less common ( Table 1 ). Procedurally, patients with anemia were more likely to have left main disease and had greater burden of overall disease treated with bare metal stent and were less likely to have use of an access site closure device ( Table 2 ).
| Variable | Anemia | p Value | |
|---|---|---|---|
| No (n = 7,176) | Yes (n = 4,815) | ||
| Age (yrs) | 65 ± 12 | 69 ± 12 | <0.001 |
| Women, n | 2,267 (32) | 1,902 (39) | <0.001 |
| Body mass index (kg/m 2 ) | 29.1 ± 5.9 | 28.5 ± 6.6 | <0.001 |
| Body surface area | 1.95 ± 0.26 | 1.90 ± 0.27 | <0.001 |
| Caucasian | 2,989 (42) | 1,626 (34) | <0.001 |
| African-American | 678 (9.4) | 783 (16) | |
| Hispanics | 1,505 (21) | 1,060 (22) | |
| Southeast Asians and others | 2,004 (28) | 1,346 (28) | |
| Indication | |||
| ST-segment elevation myocardial infarction | 42 (0.6) | 42 (0.9) | <0.001 |
| Non–ST-segment elevation myocardial infarction | 495 (7) | 583 (12) | |
| Unstable angina pectoris | 2,064 (29) | 1,438 (30) | |
| Stable angina or silent ischemia | 4,575 (64) | 2,752 (57) | |
| Hypertension | 6,529 (91) | 4,546 (94) | <0.001 |
| Diabetes mellitus | 2,701 (38) | 2,755 (57) | <0.001 |
| Insulin dependent | 401 (6) | 619 (13) | |
| Hyperlipidemia | 6,604 (92) | 4,352 (90) | 0.002 |
| Peripheral vascular disease | 533 (7.4) | 668 (14) | <0.001 |
| Previous heart failure | 2,641 (37) | 1,901 (40) | 0.003 |
| Active smoker | 1,460 (20) | 685 (14) | <0.001 |
| Previous coronary artery disease | |||
| Previous myocardial infarction | 1,768 (25) | 1,296 (27) | 0.005 |
| Previous coronary artery bypass grafting | 977 (14) | 872 (18) | <0.001 |
| CKD (GFR <60) | 1,590 (22) | 2,199 (46) | <0.001 |
| End-stage renal disease | 183 (2.6) | 406 (8.4) | <0.001 |
| GFR (ml/min/1.73 m 2 ) | 78 ± 33.5 | 64 ± 33.4 | <0.001 |
| GP IIb/IIIa inhibitor | 2,249 (31) | 919 (19) | |
| Abciximab | 833 (12) | 311 (6.5) | <0.001 |
| Eptifibatide | 1,406 (20) | 598 (12) | |
| Tirofiban | 19 (0.3) | 12 (0.2) | |
| Discharge medications | |||
| Aspirin | 7,105 (99) | 4,748 (98) | 0.04 |
| Clopidogrel | 7,043 (98) | 4,716 (98) | 0.46 |
| Prasugrel | 95 (1.3) | 43 (0.9) | 0.04 |
| β Blocker | 5,136 (72) | 3,691 (77) | <0.001 |
| ACE inhibitor | 4,415 (62) | 3,294 (68) | <0.001 |
| Statin | 5,939 (83) | 3,837 (80) | <0.001 |
| Warfarin | 326 (5) | 327 (7) | <0.001 |
| Left ventricle ejection fraction | 55 ± 10 | 53 ± 12 | <0.001 |
| Hemoglobin baseline (g/dl) | 14.1 ± 1.6 | 11.3 ± 1.2 | <0.001 |
| Variable | Anemia | p Value | |
|---|---|---|---|
| No (n = 7,176) | Yes (n = 4,815) | ||
| Target coronary artery | |||
| Left main | 162 (2.3) | 163 (3.4) | <0.001 |
| Left anterior descending | 3,376 (47) | 2,138 (44) | 0.005 |
| Left circumflex | 2,217 (31) | 1,568 (33) | 0.054 |
| Right | 2,020 (28) | 1,373 (29) | 0.66 |
| Saphenous venous graft grafts | 242 (3.4) | 200 (4.2) | 0.029 |
| Left internal mammary artery/right internal mammary artery | 55 (0.8) | 44 (0.9) | 0.60 |
| Number of coronary vessels narrowed | |||
| 1 | 2,841 (40) | 1,613 (34) | <0.001 |
| 2 | 2,333 (33) | 1,612 (34) | |
| 3 | 2,002 (28) | 1,590 (33) | |
| Stents | |||
| Bare metal | 1,430 (20) | 1,425 (30) | <0.001 |
| Drug eluting | 5,565 (78) | 3,258 (68) | <0.001 |
| Mean peak activated clotting time (seconds) | 341 ± 53 | 343 ± 80 | 0.17 |
| PCI successfull | 7,056 (98) | 4,699 (98) | 0.005 |
| Mean length of the stent (mm) | 20 ± 8 | 20 ± 8 | 0.50 |
| Mean diameter of the stent (mm) | 3.0 ± 0.7 | 3.1 ± 0.6 | 0.43 |
| Percent Stenosis of target lesion | 84 ± 9 | 84 ± 9 | 0.25 |
| Closure device use, n | 5,869 (82) | 3,643 (76) | <0.001 |
| Length of stay (days) | 1.06 ± 2.9 | 1.5 ± 2.6 | <0.001 |
There were a total of 322 (2.7%) in-hospital bleeding complications after PCI, of which 210 (1.8%) were major and 112 (0.9%) were nonmajor. Among the major bleeding complications ( Table 3 ), 88 (0.7%) were vascular access site bleeding, 5 (0.03%) gastrointestinal bleeding, and 4 (0.03%) retroperitoneal bleeding. A total of 156 patients (1.3%) received blood transfusions. The incidence of major bleeding was higher in the anemic group 159 (3.3%) compared with the nonanemic group 51 (0.7%; hazard ratio [HR] 3.25, 95% confidence interval [CI] 2.3 to 4.6, p <0.001). The interaction between anemia and major bleeding approached significance (p interaction = 0.066).
| Variable | Anemia | |
|---|---|---|
| No (n = 7,176) | Yes (n = 4,815) | |
| All major bleeding events | 51 (0.7) | 159 (3.3) |
| Access site bleeding | 35 (0.5) | 53 (1.1) |
| Gastrointestinal bleeding | 1 (0.01) | 4 (0.08) |
| Retroperitoneal bleed | 2 (0.03) | 2 (0.04) |
| Blood transfusion | 16 (0.22) | 140 (2.9) |
| Blood transfusion for overt bleed | 8 (0.11) | 39 (0.81) |
Of the 159 bleeding events, 140 (88%) were transfusions. Despite significantly lower drops in hemoglobin level in all patients with anemia (0.96 ± 1.5 vs 0.46 ± 1.0, respectively, p <0.001) and those with major bleed events compared with patients without anemia (3.3 ± 1.5 vs 1.2 ± 2.2, respectively, p <0.001), patients with anemia were significantly more likely to receive transfusion (p <0.001; Table 4 ). Furthermore, in patients who received blood transfusion, change in hemoglobin was significantly lower in the patients with anemia compared with patients without anemia (3.0 ± 1.5 vs 1.0 ± 1.9, respectively, p <0.003; Table 4 ).
| Variable (Mean ± SD) | Anemia | p Value | |
|---|---|---|---|
| No (n = 7,176) | Yes (n = 4,185) | ||
| Baseline hemoglobin (g/dl) | 14.1 ± 1.6 | 11.3 ± 1.2 | <0.001 |
| Post-PCI hemoglobin (g/dl) | 13.2 ± 1.3 | 10.9 ± 1.3 | <0.001 |
| Drop in hemoglobin (g/dl) | 0.96 ± 1.5 | 0.46 ± 1.0 | <0.001 |
| Major bleed | 51 (0.7) | 159 (3.3) | |
| Baseline hemoglobin (g/dl) | 13.9 ± 1.3 | 9.9 ± 1.3 | <0.001 |
| Post-PCI hemoglobin (g/dl) | 10.9 ± 2.1 | 9.0 ± 1.3 | <0.001 |
| Drop in hemoglobin (g/dl) | 3.3 ± 2.3 | 1.2 ± 2.2 | <0.001 |
| Blood transfusion | 16 (0.22) | 140 (2.9) | |
| Baseline hemoglobin (g/dl) | 13.2 ± 0.7 | 9.8 ± 1.3 | <0.001 |
| Post-PCI hemoglobin (g/dl) | 10.2 ± 1.7 | 8.9 ± 1.3 | <0.001 |
| Drop in hemoglobin (g/dl) | 3 ± 1.5 | 1.0 ± 1.9 | 0.003 |
At a median follow-up of 2.6 years (interquartile range 1.4 to 3.5 years), there were a total of 1,157 deaths (9.6%). Survival was significantly lower in patients who suffered major bleeding ( Figure 1 ) and was a predictor of mortality (HR 1.4, 95% CI 1.1 to 1.8, p = 0.027). Similarly in patients with anemia, the survival rate of patients who had a major bleeding event was significantly lower ( Figure 2 ), and major bleeding remained an independent predictor of mortality (HR 1.5, 95% CI 1.1 to 2.0, p = 0.008). However, in patients without baseline anemia, survival rate was similar ( Figure 3 ), and major bleeding was no longer identified as a predictor of mortality (HR 1.25, 95% CI 0.52 to 3.03, p = 0.62).
