It is unclear whether intracoronary (IC) bolus administration of glycoprotein IIb/IIIa inhibitors (GPIs) during percutaneous coronary intervention (PCI) in patients with acute coronary syndromes is superior to intravenous (IV) administration. We conducted a meta-analysis of randomized controlled trials (RCTs) to compare the effects of IC and IV administrations of GPIs in patients with acute coronary syndromes. We systematically searched the Cochrane Library, EMBASE, and MEDLINE databases for RCTs comparing IC to IV administration of GPIs (abciximab, eptifibatide, tirofiban) during PCI. Data were pooled and stratified into short (1 month to 3 months) and mid-/long-term (≥6 months) follow-up durations. Ten RCTs involving 1,590 patients met our inclusion criteria. Compared to the IV group the IC group was more likely to have complete perfusion (Thrombolysis In Myocardial Infarction grade 3 flow) after PCI (risk ratio [RR] 1.08, 95% confidence interval [CI] 1.02 to 1.15). IC administration was associated with similar bleeding rates as IV (RR 0.92, 95% CI 0.68 to 1.24) but with a significant decrease in short-term target vessel revascularization (RR 0.54, 95% CI 0.30 to 0.96). IC administration was also associated with a significant decrease in short-term mortality (RR 0.45, 95% CI 0.23 to 0.90) but this decrease was no longer significant in mid-/long-term RCTs. In conclusion, compared to IV administration IC administration of GPIs has favorable effects on Thrombolysis In Myocardial Infarction flow, target vessel revascularization, and short-term mortality after PCI, with no difference in rates of bleeding. Data regarding mid-/long-term outcomes were limited and inconclusive. Large RCTs with longer follow-up are required to determine long-term safety and efficacy.
Glycoprotein IIb/IIIa inhibitors (GPIs) are usually administered as an initial large intravenous (IV) bolus followed by ≥12-hour maintenance infusion at doses calculated to inhibit platelet aggregation. However, intracoronary (IC) administration of the initial bolus could potentially yield local concentrations high enough to not only inhibit further thrombus formation but actively disperse thrombi that have already formed. Although studies have compared IC and IV administrations of the initial bolus, these were limited in size and yielded conflicting or inconclusive results. Because many patients undergoing percutaneous coronary intervention (PCI) will receive GPIs, it is important to identify the optimal strategy to maximize the benefit from GPIs. We therefore conducted a systematic review and meta-analysis of randomized controlled trials (RCTs) to compare the clinical outcomes of IC and IV administrations of GPIs.
Methods
We systematically searched the Cochrane Library, EMBASE, and MEDLINE databases from inception through to January 2011 without language restriction to identify all RCTs comparing the effects of IV administration of GPIs to those of IC administration. The Medical Subject Headings search string for this literature search was “intracoronary[All Fields] and intravenous[All Fields] and ((‘abciximab’[Substance] or ‘abciximab’[All Fields]) or (‘eptifibatide’[Substance] or ‘eptifibatide’[All Fields]) or (‘tirofiban’[Substance] or ‘tirofiban’[All Fields])).” We limited our literature search to studies conducted in humans and published in peer-reviewed journals. Reviews and reference lists of retrieved articles were manually searched for potentially relevant publications not previously identified in the database search. The retrieved studies were examined to eliminate potential duplicates or overlapping data.
We included a study in our systematic review if (1) it was a trial in which the subjects were randomly assigned to IC or IV administration in a parallel-group design, (2) it reported clinical outcomes, and (3) its duration of follow-up was ≥1 month. We excluded conference abstracts because their results may not be final. We included studies whether they were open label or blinded but performed a quality assessment.
Two reviewers (S.F. and A.S.) independently extracted data from each trial. Results were compared and any disagreements were resolved by consensus. Data extracted for each RCT included first author, year of publication, study design, length of follow-up, number of participants and their characteristics, drug, and drug protocol. Pooled outcomes were mortality, target vessel revascularization, Thrombolysis In Myocardial Infarction grade flow after PCI, and bleeding events. Selected outcomes of interest that were not appropriate for pooling were reported and qualitatively compared.
The systematic review and meta-analysis were performed according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement for reporting systematic reviews and meta-analyses of RCTs. Each study was evaluated using the Jadad scale. One point was assigned for an affirmative answer to each of the following 5 questions: (1) Was the study described as randomized? (2) Was the method of randomization described and appropriate? (3) Was the study described as double blinded? (4) Was the method of double blinding described and adequate? (5) Was there a description of withdrawals/dropouts?
We used fixed-effect or DerSimonian–Laird random-effect models, which accounts for within-study and between-study variabilities, to estimate pooled risk ratios (RRs) with their 95% confidence intervals (CIs). In these types of models, if there are no events in the 2 arms, then any measurement of effect summarized as an RR cannot be defined. Hence, RCTs with no events in either arm are discarded from pooled analyses. Forest plots were created for each outcome. Statistical heterogeneity was assessed using the Q statistic (with p <0.1 considered significant) and I 2 statistics were calculated to estimate the proportion of variance owing to between-study heterogeneity. Where there was no heterogeneity, fixed-effect models were preferentially reported. All analyses were conducted using STATA 9.0 (STATA Corp., College Station, Texas).
Results
Our literature search identified 147 studies excluding duplicates ( Figure 1 ) . After reviewing titles and abstracts to exclude irrelevant studies, case reports, editorial comments, and reviews, 25 studies were retrieved for further consideration. We conducted a manual search of references of retrieved articles but no additional studies were found. We then excluded observational studies, retrospective subgroup analyses, and articles that did not report outcomes of randomized head-to-head trials of IC versus IV administration. One RCT reported 1- and 6-month outcomes in separate publications. Ultimately, 10 RCTs reported in 11 articles were included in our systematic review.
Table 1 presents characteristics of the included RCTs. These trials involved 1,590 patients, with 813 and 777 receiving IC and IV administration, respectively. The number of subjects in each RCT ranged from 41 to 534, and length of follow-up ranged from 1 month to 21 months. One RCT examined eptifibatide, 2 examined tirofiban, and the rest examined abciximab. Participants were predominantly men (studies ranging from 55% to 82% men), with mean or median age ranging from 59 to 76 years. Seven of the 10 RCTs included only patients with ST-elevation myocardial infarction; in the other 3 studies the percentage of patients presenting with ST-elevation myocardial infarction ranged from 26% to 63%.
| Study | GPI | Age (years)/Men (%) | Population | Access Site | Drug Protocol § | |
|---|---|---|---|---|---|---|
| IC | IV | |||||
| Short term (1 month–3 months) ⁎ | ||||||
| Deibele et al, 2010 | eptifibatide | 59/69% | ACS (26% STEMI) | femoral | IC bolus, then 18-hour IV infusion | IV bolus, then 18-hour IV infusion |
| Gu et al, 2010 | abciximab | 64/74% | STEMI | — | IC bolus proximal to lesion, then no 12-hour IV infusion | IV bolus during PCI, then no 12-hour IV infusion |
| Iversen et al, 2011 | abciximab | 62/81% | STEMI | — | IC bolus, then 12-hour IV infusion | IV bolus, then 12-hour IV infusion |
| Dominguez-Rodriguez et al, 2009 | abciximab | 68/76% | STEMI | femoral | IC bolus after thrombectomy, then 12-hour IV infusion | IV bolus after thrombus aspiration, then 12-hour IV infusion |
| Wu et al, 2008 | tirofiban | 76/55% | ACS (63% STEMI) | — | IC bolus (over 3 minutes), then 36-hour IV infusion | IV bolus (over 3 minutes), then 36-hour IV infusion |
| Thiele et al, 2008 ‡ | abciximab | 65/61% | STEMI | — | IC bolus before balloon dilation, then 12-hour IV infusion | IV bolus, then 12-hour IV infusion |
| Yang et al, 2007 | tirofiban | 59/79% | STEMI | — | IC bolus before first balloon inflation, then 36-hour IV infusion | IV bolus before angiography, then 36-hour IV infusion |
| Bellandi et al, 2004 | abciximab | 62/78% | STEMI | femoral | IC bolus below occlusion, then 12-hour IV infusion | IV bolus, then 12-hour IV infusion |
| Mid/long term (≥6 months) † | ||||||
| Eitel et al, 2010 ‡ | abciximab | 65/61% | STEMI | — | IC bolus, then 12-hour IV infusion | IV bolus, then 12-hour IV infusion |
| Bertrand et al, 2010 | abciximab | 59/80% | STEMI | transradial | IC bolus, then 12-hour IV infusion | IV bolus, then 12-hour IV infusion |
| Galache Osuna et al, 2006 | abciximab | 61/82% | ACS (41% STEMI) | radial or femoral | IC bolus, then 12-hour IV infusion | IV bolus, then 12-hour IV infusion |
⁎ Yang et al reported 1- to 3-month follow-up; all others reported 1-month follow-up.
† Galache Osuna et al reported 6- to 21-month follow-up; all others reported 6-month follow-up.
‡ Same randomized controlled trial with 2 terms of follow-up reported separately.
§ Unless otherwise noted, the abciximab intracoronary and intravenous bolus doses are 0.25 mg/kg body weight and the intravenous infusion dose is 0.125 μg/kg/min; the eptifibatide intracoronary and intravenous bolus doses are 180 μg/kg body weight and the intravenous infusion dose is 2 μg/kg/min or 1 μg/kg/min in patients with renal insufficiency; the tirofiban intracoronary and intravenous bolus doses are 10 μg/kg body weight and the intravenous infusion dose is 0.15 μg/kg/min.
Within each study the IC and IV groups received the same GPI at the same dosage and duration. The initial bolus was delivered by IC or IV administration depending on the group. The maintenance infusion was delivered by IV regardless of group. No difference was observed in the safety or tolerability of IC and IV administration.
Three of the 10 RCTs, reported in 4 publications, achieved a score of ≥3 points for quality. Methodologic shortcomings included undescribed method of randomization (5 studies), uncertain or absent double blinding (9 studies), and undescribed withdrawals/dropouts (6 studies). Although we initially intended to perform a subgroup analysis of high-quality studies, we did not because we found only 2 high-quality studies in each of the short-term and mid-/long-term follow-up periods. However, we observed that the results of high-quality trials were not distinguishable from those of the other trials.
Thrombolysis In Myocardial Infarction grade flow outcomes after PCI were pooled from 8 studies with short-term follow-up ( Table 2 ). Most of these studies specified that Thrombolysis In Myocardial Infarction grade flow was assessed by ≥1 observer blinded to a patient’s treatment status. Compared to subjects randomized to IV administration those randomized to IC administration were more likely to have complete perfusion (Thrombolysis In Myocardial Infarction grade 3 flow) after PCI (n = 1,348, RR 1.08, 95% CI 1.02 to 1.15; Figure 2 ) . A random-effects model was used because there was evidence of mild heterogeneity (I 2 = 20.1%, p = 0.270) between studies.
| Study | Number of Patients | TIMI 3 Flow After PCI | TVR | Death | Any Bleeding | |||||
|---|---|---|---|---|---|---|---|---|---|---|
| IC | IV | IC | IV | IC | IV | IC | IV | IC | IV | |
| Short term (1 month–3 months) ⁎ | ||||||||||
| Deibele et al, 2010 | 22 | 19 | 21 (95%) | 17 (89%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) | 0 (0%) |
| Gu et al, 2010 | 271 | 263 | 241 (89%) | 226 (86%) | 9 (3%) | 10 (4%) | 5 (2%) | 7 (3%) | 31 (11%) | 27 (10%) |
| Iversen et al, 2011 | 185 | 170 | 149 (81%) | 124 (73%) | 7 (4%) | 16 (9%) | 2 (1%) | 9 (5%) | 21 (11%) | 28 (15%) |
| Dominguez-Rodriguez et al, 2009 | 25 | 25 | 22 (88%) | 17 (68%) | — | — | 0 (0%) § | 0 (0%) § | 2 (8%) | 3 (12%) |
| Wu et al, 2008 | 58 | 57 | 51 (88%) | 41 (72%) | 0 (0%) | 1 (2%) | 1 (2%) | 2 (3%) | 7 (12%) ¶ | 9 (16%) ¶ |
| Thiele et al, 2008 ‡ | 77 | 77 | 65 (84%) | 66 (86%) | 0 (0%) | 2 (3%) | 2 (3%) | 3 (4%) | 4 (5%) | 5 (6%) |
| Yang et al, 2007 | 28 | 26 | 26 (93%) | 19 (73%) | — | — | 0 (0%) | 2 (7%) | 7 (25%) | 3 (11%) |
| Bellandi et al, 2004 | 22 | 23 | 22 (100%) | 20 (87%) | — | — | 1 (5%) | 1 (5%) | — | — |
| Mid/long term (≥6 months) † | ||||||||||
| Eitel et al, 2010 ‡ | 77 | 77 | 65 (84%) | 66 (86%) | 2 (3%) | 4 (5%) | 4 (5%) | 4 (5%) | 4 (5%) | 5 (6%) |
| Bertrand et al, 2010 | 53 | 52 | 50 (94%) | 47 (90%) | — | — | — | — | — | — |
| Galache Osuna et al, 2006 | 72 | 65 | 72 (100%) | 64 (98%) | — | — | 2 (3%) | 3 (4%) | — | — |
⁎ Yang et al reported 1- to 3-month follow-up; all others reported 1-month follow-up.
† Galache Osuna et al reported 6- to 21-month follow-up; all others reported 6-month follow-up.
‡ Same randomized controlled trial with 2 terms of follow-up reported separately.
§ In-hospital mortality outcomes (30-day mortality outcomes not reported).
¶ Fourteen-day bleeding event outcomes (30-day bleeding outcomes not reported).
