The use of prasugrel in patients with coronary artery disease (CAD) has been associated with decreased major adverse cardiac events (MACEs) compared with clopidogrel but with an increased risk of bleeding. However, it remains unclear if the risks of bleeding outweigh those of MACEs in patients on prasugrel treatment. We systematically reviewed randomized controlled trials comparing prasugrel with clopidogrel in patients with CAD. We performed a literature search of PubMed, EMBASE, and Cochrane Central Register of Controlled Trial databases from inception to November 25, 2014, and reviewed the reference lists of retrieved articles. A comparative estimate was made for the combined rates of MACEs and bleeding from the same trials in the framework of this meta-analysis and expressed as odds ratios (ORs) and 95% confidence intervals (CIs) in both random- and fixed-effects models. Nine studies involving 25,214 patients were included in our meta-analysis. In both the random- and fixed-effects models, the risks of MACEs outweighed those of major bleeding (OR 7.48, 95% CI 3.75 to 14.94, p <0.0001, random effects) and of minor bleeding (OR 3.77, 95% CI 1.73 to 8.22, p = 0.009, random effects). Results were corroborated in a standard-dose clopidogrel subgroup analysis (OR 7.46, 95% CI 3.54 to 15.68, p <0.0001, and OR 6.44, 95% CI 2.80 to 14.80, p <0.0001, random effects, respectively). In conclusion, despite the increased risk of bleeding associated with prasugrel treatment compared with clopidogrel, the risk of MACEs far outweighed the risk of bleeding.
Dual antiplatelet therapy (DAT) consisting of aspirin and clopidogrel are indispensable treatments for patients who underwent percutaneous coronary intervention (PCI) to prevent ischemic events and stent thrombosis. However, there remains ∼10% risk of recurrent thrombotic events within 1 year after PCI, even after the use of traditional DAT. Furthermore, over the past years, several drawbacks of clopidogrel therapy have been identified. The search for the ideal antiplatelet agents is still on. Prasugrel showed superior antiplatelet effects compared with clopidogrel, which has been converted to an active form before binding irreversibly to the adenosine diphosphate (ADP) receptor and inhibiting platelet aggregation. Several large-scale clinical trials have confirmed that prasugrel therapy was associated with significantly reduced rates of major adverse cardiac events (MACEs) but with an increased risk of major bleeding. However, to date, it remains unclear whether the risks of bleeding outweigh those of MACEs in patients on prasugrel treatment. Therefore, we aimed to investigate whether the risks of bleeding outweigh those of MACEs in patients on prasugrel treatment.
Methods
We systematically searched all published studies that compared prasugrel with clopidogrel in patients with CAD. We searched PubMed, EMBASE, and the Cochrane Central Register of Controlled Trials from inception to November 25, 2014. The literature search was updated on December 15, 2014, during the final revision process without any language restriction. Major international cardiologic meetings (American College of Cardiology, American Heart Association, European Society of Cardiology, and Transcatheter Cardiovascular Therapeutics) and Internet-based sources of information (Web of Science and clinicaltrial.gov ) were also searched for more trials. In addition to searching databases, the reference lists of all the identified studies were manually searched.
Three investigators (HBC, XLZ, and JX) independently performed study selection and assessed the eligibility of trials identified through search of relevant databases, with disagreements resolved in consultation with another investigator (HBL and XWL). Studies were considered for inclusion based on the following criteria: (1) randomized controlled trials (RCTs) about patients with CAD, (2) patients were randomly assigned to prasugrel and clopidogrel, and (3) all the studies reported data on MACEs and major bleeding or minor bleeding. We defined MACEs as a composite outcome of cardiovascular deaths, myocardial infarction, and ischemic stroke.
The data were extracted and presented according to the Providing Innovative Service Models and Assessment criteria. Four investigators (HBC, XLZ, HBL, and BYH) independently extracted data from all the included studies with disagreements resolved in consultation with another investigator (JX, XYZ, and XWL). The extracted data were first author, the year of publication, country, follow-up duration, number of participants, drug protocol, and the pooled efficacy outcomes including MACEs, major bleeding, and minor bleeding. The risk of bias was assessed according to the guidance in the Cochrane Handbook, version 5.1.0, and RevMan 5.3 (The Cochrane Collaboration, Copenhagen, Denmark) was used to construct summaries. The investigators may be contacted by e-mail for more detailed information if the outcome data were not clear.
A comparative estimate was made for the combined rates of MACEs and bleeding from the same trials in the framework of this meta-analysis and expressed as odds ratios (ORs) and 95% confidence intervals (CIs) in both fixed- and random-effects models. We also determined pooled ORs and corresponding 95% CI for MACEs, major bleeding, and minor bleeding. The Q test and I 2 statistic were used to evaluate heterogeneity between studies. Publication bias was estimated using funnel plots and Begg’s rank correlation test and Egger’s linear regression test. If publication bias seemed to be present through this process, the trim-and-fill procedure by Duval and Tweedie was used to estimate the possible impact of unpublished studies on the pooled estimate. RevMan 5.3 (The Cochrane Collaboration) and R 3.1.2 (R Language and Environment for Statistical Computing; http://www.r-project.org/ ) were used to perform this analysis.
Results
A total of 1,560 potential relevant articles were retrieved by database and manual searching. Eight hundred seventy-five duplicates and 653 irrelevant articles were excluded by primary assessments. Thirty-two articles were excluded for not fulfilling our inclusion criteria after a more detailed assessment. As a result, 9 studies that met the inclusion criteria were included in our meta-analysis, involving 25,214 patients. Figure 1 outlines the search and selection of RCTs for review. The baseline characteristics of each study are described in Table 1 . Overall, the quality of the studies included was judged partially limited by a lack of transparency in the reporting of the random sequence generation, for which 5 trials were at low risk, whereas 4 at unclear risk, and in the reporting of the allocation concealment, 5 trials were at low risk, whereas 4 at high risk. Bias risk assessment of other domains is given in detail in Supplementary Figure A .
| TRITON-TIMI 38 2007 | TRILOGY ACS 2012 | PRASFIT-ACS 2014 | JUMBO-TIMI26 2005 | PRINCIPLE-TIMI 44 2007 | Alexopoulos et al 2011 | OPTIMUS-3 2011 | TAILOR 2014 | ACAPULCO 2010 | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Prasugrel | Clopidogrel | Prasugrel | Control | Prasugrel | Clopidogrel | Prasugrel | Clopidogrel | Prasugrel | Clopidogrel | Prasugrel | Clopidogrel | Prasugrel | Clopidogrel | Prasugrel | Clopidogrel | Prasugrel | Clopidogrel | |
| Study design | RCT | RCT | RCT | RCT | RCT | RCT | RCT | RCT | RCT | |||||||||
| Country | USA | USA | Japan | USA | USA | Greece | USA | Denmark | USA | |||||||||
| Numbers of centers | 707 | 966 | 162 | 1 | 14 | 1 | 1 | 1 | 4 | |||||||||
| Follow-up period (m) | 15 | 17 | 12 | 1 | 1 | 1 | 1 | 24 | 1 | |||||||||
| Population | STEMI and NSTE-ACS | NSTE-ACS | STEMI and NSTE-ACS | CAD | CAD | CAD | DM and CAD | CAD | STEMI and NSTE-ACS | |||||||||
| Numbers of patients (n) | 6813 | 6795 | 4663 | 4663 | 685 | 678 | 200 | 254 | 102 | 99 | 32 | 32 | 18 | 16 | 54 | 52 | 28 | 28 |
| Age (years) | 61 | 61 | 66 | 66 | 65 | 65 | 60 | 59 | 64 | 64 | 62 | 68 | 61 | 63 | 63 | 61 | ||
| Female, n(%) | 1703 (25) | 1699 (27) | 1828 (39.2) | 1823 (39.1) | 149 (21.8) | 140 (20.6) | 50 (25) | 69 (27) | 29 (28.4) | 22 (22.2) | 3 (9.4) | 4 (12.5) | 11 (31.4) | 14 (25.9) | 9 (17.3) | 9 (16) | ||
| White race, n (%) | 6268 (92) | 6319 (93) | NA | NA | 0 | 0 | 180 (90) | 239 (94) | NA | NA | NA | NA | 15 (42.9) | NA | NA | 53 (95) | ||
| Hypertension , n(%) | 4360 (64) | 4349 (64) | 3819 (81.9) | 3825 (82.0) | 495 (72.3) | 491 (72.4) | NA | NA | NA | NA | 21 (65.6) | 22 (68.8) | 33 ((94.3) | 40 (74.1) | 43 (82.7) | 37 (66) | ||
| Smoking, n(%) | 2589 (38) | 2582 (38) | 919 (19.7) | 942 (20.2) | 552 | NA | 50 (25) | 79 (31) | 18 (17.6) | 16 (16.2) | 16 (50.0) | 14 (43.8) | 7 (20.0) | 42 (77.8) | 35 (67.3) | 17 (30) | ||
| Diabetes mellitus , n(%) | 1567 (23) | 1563 (23) | 1758 (37.7) | 1786 (38.3) | 250 (36.5) | 237 (35.0) | 50 (25) | 64 (25) | 33 (32.4) | 29 (29.3) | 10 (31.2) | 13 (40.6) | 35 (100) | 16 (29.6) | 18 (34.6) | 11 (20) | ||
| Dyslipidemia, n (%) | 3815 (56) | 3805 (56) | 2751 (59.0) | 2765 (38.3) | 516 (75.3) | 500 (73.7) | NA | NA | 92 (90.2) | 86 (86.9) | 19 (59.4) | 22 (68.8) | 33 (94.3) | 45 (83.3) | 46 (88.5) | 9 (16) | ||
| BMI (Body mass index, kg/m2) | 28 | 28 | NA | NA | 24.2 | 24.2 | NA | NA | 28.7 | 29.4 | 30.1 | 27.9 | 33.1 | 29 | 28 | 27 | ||
| Pre-PCI, n (%) | NA | NA | 1194 (25.6) | 1245 (26.7) | 34 (5.0) | 41 (6.0) | NA | NA | NA | NA | 3 (9.4) | 2 (6.2) | 20 (57.1) | NA | NA | 24 (43) | ||
| Pre-CABG, n (%) | 545 (8) | 476 (7) | 709 (15.2) | 751 (16.1) | 6 (0.9) | 4 (0.6) | NA | NA | NA | NA | 3 (9.4) | 1 (3.1) | 8 (22.9) | NA | NA | 5 (9) | ||
| Clinical diagnosis | ||||||||||||||||||
| Stable angina (%) | 0 | 0 | 0 | 0 | 0 | 6.2 | 0 | NA | 13 | 21.2 | NA | |||||||
| Unstable angina (%) | 74 | 74 | 29.6 | 30.6 | 22.8 | 18.3 | NA | NA | 28.1 | 18.8 | NA | 87 | 78.8 | 58.9 | ||||
| NSTEMI (%) | 70.4 | 69.4 | 27.3 | 31.4 | NA | NA | 43.8 | 50 | NA | 41.1 | ||||||||
| STEMI (%) | 26 | 26 | NA | 49.6 | 50.3 | NA | NA | NA | 0 | |||||||||
| Loading dose (mg) | 60mg pras vs 300mg clop | 30mg pras vs 300mg clop | 20mg pras vs 300mg clop | 60mg pras vs 300mg clop | 60mg pras vs 600mg clop | NA | 60mg pras vs 600mg clop | 600mg clopidogrel | 900mg clopidogrel | |||||||||
| Maintance dose (mg) | 10 mg pras vs 75 mg clop QD | 10 mg pras vs 75 mg clop QD | 3.25 mg pras vs 75 mg clop QD | 10 mg pras vs 75 mg clop QD | 10 mg pras vs 150 mg clop QD | 10 mg pras vs 150 mg clop QD | 10 mg pras vs 150 mg clop QD | 10 mg pras vs 150 mg clop QD | 10 mg pras vs 150 mg clop QD | |||||||||
In total, 2,981 patients developed MACEs: 1,385 in the prasugrel group and 1,595 in the clopidogrel group. The incidence of MACEs was lower in the prasugrel group than in the clopidogrel group (OR 0.86, 95% CI 0.78 to 0.94, p <0.0001, Supplementary Figure B ). In total, 425 patients developed major bleeding: 242 in the prasugrel group and 183 in the clopidogrel group. The incidence of major bleeding was higher in the prasugrel group than in the clopidogrel group (OR 1.33, 95% CI 1.09 to 1.61, p = 0.004, Supplementary Figure C ). In total, 373 patients developed minor bleeding: 210 in the prasugrel group and 163 in the clopidogrel group. The incidence of minor bleeding was higher in the prasugrel group than in the clopidogrel group (OR 1.29, 95% CI 1.05 to 1.58, p = 0.02, Supplementary Figure D ). The risk of MACEs far outweighed that of major bleeding (OR 7.48, 95% CI 3.75 to 14.94, p <0.0001, random effects, and OR 6.28, 95% CI 5.46 to 7.20, p <0.0001, fixed effects, Figure 2 , respectively) and the risk of MACEs far outweighed that of minor bleeding (OR 3.77, 95% CI 1.73 to 8.22, p = 0.009, random effects, and OR 7.19, 95% CI 6.21 to 8.33, p <0.001, fixed effects, Figure 3 , respectively).
In standard-dose clopidogrel trials, the incidence of MACEs was lower in the prasugrel group than in the clopidogrel group (OR 0.85, 95% CI 0.79 to 0.92, p <0.0001, Supplementary Figure B in appendix), and the incidence of major bleeding and minor bleeding was higher in the prasugrel group than in the clopidogrel group (OR 1.33, 95% CI 1.09 to 1.61, p = 0.004, and OR 1.27, 95% CI 1.03 to 1.57, p = 0.02, Supplementary Figures C and D , respectively). The risk of MACEs far outweighed that of major bleeding (OR 7.46, 95% CI 3.54 to 15.68, p <0.0001, random effects, and OR 6.26, 95% CI 5.45 to 7.19, p <0.0001, fixed effects, Figure 2 , respectively). The risk of MACEs far outweighed that of minor bleeding (OR 6.44, 95% CI 2.80 to 14.80, p <0.0001, random effects, and OR 7.52, 95% CI 6.47 to 8.73, p <0.0001, fixed effects, Figure 3 , respectively). In double-dose clopidogrel trials, there was no difference in the incidence of MACEs, major bleeding, and minor bleeding between prasugrel group and clopidogrel group (OR 0.75, 95% CI 0.27 to 2.03, p = 0.63; OR 0.32, 95% CI 0.01 to 8.23, p = 0.49; and OR 1.72, 95% CI 0.59 to 5.02, p = 0.32; Supplementary Figures B to D , respectively). There was a trend that the risk of MACEs outweighed that of major bleeding (OR 8.01, 95% CI 0.97 to 65.97, p = 0.05, random effects, and OR 8.46, 95% CI 5.68 to 68.28, p = 0.04, fixed effects, Figure 2 , respectively). The risk of minor bleeding was equal to that of MACEs (OR 0.71, 95% CI 0.10 to 4.91, p = 0.73, random effects, and OR 0.89, 95% CI 0.35 to 2.28, p = 0.81, fixed effects, Figure 3 , respectively).
Heterogeneity test showed no significant statistical evidence of heterogeneity for MACEs ( I 2 = 12%, p = 0.34), major bleeding ( I 2 = 0%, p = 0.52), and minor bleeding ( I 2 = 0%, p = 0.48). To test the robustness of results, we performed a sensitivity analysis, and the results were listed in Table 2 . Egger’s linear regression test revealed the presence of publication bias only for major bleeding (p = 0.0401, Table 2 ). Other outcomes seem to be no publication bias through Begg’s rank correlation test and Egger’s linear regression test ( Table 2 ); however, a funnel plot for outcomes seemed to be asymmetric ( Supplementary Figures E to G ). Therefore, the trim-and-fill procedure by Duval and Tweedie, which conservatively imputes hypothetical negative unpublished studies to mirror the positive studies that cause funnel plot asymmetry, was conducted for these end points, and the presence of publication bias may be suggested for major bleeding ( Figure 4 ).
| MACEs | Major bleeding | Minor bleeding | |||||||
|---|---|---|---|---|---|---|---|---|---|
| Pooled OR | 95%CI | P | Pooled OR | 95%CI | P | Pooled OR | 95%CI | P | |
| Fixed effect models | 0.85 | 0.79 to 0.92 | <0.0001 | 1.33 | 1.09 to 1.61 | 0.004 | 1.29 | 1.05 to1.58 | 0.02 |
| Random effect models | 0.86 | 0.78 to 0.94 | 0.002 | 1.33 | 1.09 to 1.61 | 0.004 | 1.28 | 1.04 to 1.57 | 0.02 |
| Subgroup analysis | |||||||||
| In standard-dose clopidogrel | 0.85 | 0.79 to 0.92 | <0.0001 | 1.33 | 1.10 to 1.62 | 0.004 | 1.27 | 1.03 to 1.57 | 0.02 |
| In double-dose clopidogrel | 0.75 | 0.27 to 2.03 | 0.57 | 0.32 | 0.01 to 8.23 | 0.49 | 1.72 | 0.59 to 5.02 | 0.32 |
| Adjusted by Trim-and-fill ∗ | 0.86 | 0.79 to 0.93 | 0.0002 | 1.39 | 1.15 to 1.67 | 0.0006 | 1.26 | 1.03 to 1.55 | 0.028 |
| Omitting TRITON-TIMI 38 | 0.93 | 0.84 to 1.04 | 0.22 | 1.09 | 0.78 to 1.53 | 0.59 | 1.57 | 1.10 to 2.24 | 0.02 |
| Omitting TRILOGY ACS | 0.79 | 0.71 to 0.87 | <0.0001 | 1.37 | 1.09 to 1.71 | 0.006 | 1.28 | 1.02 to 1.60 | 0.07 |
| Omitting PRASFIT-ACS | 0.85 | 0.79 to 0.92 | <0.0001 | 1.37 | 1.12 to 1.62 | 0.002 | 1.24 | 0.99 to 1.54 | 0.07 |
| Omitting JUMBO-TIMI26 | 0.85 | 0.79 to 0.92 | <0.0001 | 1.33 | 1.10 to 1.62 | 0.004 | 1.27 | 1.04 to 1.57 | 0.03 |
| Omitting PRINCIPLE-TIMI 44 | 0.85 | 0.79 to 0.92 | <0.0001 | 1.33 | 1.10 to 1.62 | 0.004 | 1.28 | 1.03 to 1.57 | 0.02 |
| Omitting Alexopoulos et al | 0.85 | 0.79 to 0.92 | <0.0001 | 1.33 | 1.10 to 1.62 | 0.004 | 1.27 | 1.03 to 1.56 | 0.02 |
| Omitting TAILOR | 0.85 | 0.79 to 0.92 | <0.0001 | 1.33 | 1.10 to 1.62 | 0.004 | 1.29 | 1.05 to 1.58 | 0.02 |
| Omitting OPTIMUS-3 | 0.85 | 0.79 to 0.92 | <0.0001 | 1.33 | 1.09 to 1.61 | 0.004 | 1.31 | 1.06 to 1.61 | 0.02 |
| Omitting ACAPULCO | 0.85 | 0.79 to 0.92 | <0.0001 | 1.33 | 1.09 to 1.61 | 0.004 | 1.29 | 1.05 to 1.58 | 0.03 |
| z | t | P | z | t | P | z | t | P | |
| Begg’s rank correlation test | -0.4506 | NA | 0.6523 | -1.4697 | NA | 0.1416 | 0.2474 | NA | 0.8046 |
| Egger’s linear regression test | NA | -0.1648 | 0.8756 | NA | -3.4788 | 0.0401 | NA | 1.3054 | 0.2396 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
