In the setting of acute myocardial infarction (AMI), spontaneous and percutaneous coronary intervention (PCI)-related embolisation results in a decreased efficacy of mechanical reperfusion and myocardial salvage. Direct stenting without predilation may decrease embolisation and the incidence of the no-reflow phenomenon.1,2 More specific approaches to the problem of microvessel embolisation during PCI include thrombectomy by different techniques and the use of anti-embolic protection devices. One randomised trial has reported rheolytic thrombectomy to be effective in decreasing embolisation and adverse clinical events in patients who underwent PCI on venous grafts or native coronary vessel with massive thrombosis,3 while the results of two concluded placebo-controlled randomised studies were conflicting.4,5
The Florence-AngioJet Randomised Trial
The Florence-AngioJet randomised trial was the first study that assessed the efficacy of rheolytic thrombectomy before direct infarct artery stent implantation in patients who underwent PCI for AMI.4 This study was based on a sample of 100 patients with a first AMI and the end-points of the study were early ST-segment elevation resolution, the corrected thrombolysis in myocardial infarction (TIMI) frame count and the infarct size as assessed by technetium-99m (Tc-99m) sestamibi scintigraphy at one month. All end-points were reached. Patients randomised to thrombectomy before direct stenting had a higher incidence of early ST-segment elevation resolution (90% vs 72%, p = 0.022), lower corrected TIMI frame counts (18.2 ± 7.7 vs 22.5 ± 11.0, p = 0.032) and smaller infarcts (13.0 ± 11.6% vs 21.2 ± 18.0%, p = 0.010) compared with patients randomised to direct stenting alone. By multivariate analysis, the only variables related to the early ST-segment resolution were randomisation to thrombectomy (OR 3.56, 95% CI 1.11 to 11.42, p = 0.032) and diabetes mellitus (OR 0.24, 95% CI 0.07 to 0.86, p = 0.029). At one month, no patient died or had reinfarction or the need for target vessel revascularisation (TVR), and the six-month clinical outcomes were similar in the two arms: the mortality rate was 2% in both groups and no reinfarction occurred, while TVR rate was 14% in the stent alone arm and 23% in the rheolytic thrombectomy arm (p = 0.270).
Some characteristics of the study design and other figures of the study deserve a specific comment. The study tried to avoid the confounding effects of some procedural variables that may affect the effectiveness of reperfusion such as predilation or postdilation after stenting, different types of stents, use or non-use of glycoprotein IIb/IIIa inhibitors. Direct stenting was attempted in all patients and was successfully performed in 84% of patients, while in 16 patients, predilation was performed after direct stenting attempt failure. After stenting, no patient needed further balloon dilation, and only two types of bare tubular stents were used (one with a closed-cell design, and the other with an open-cell design; the latter was used in target lesions involving a major branch). Nearly all patients received the same anti-thrombotic treatment (all but two patients had abciximab (ReoPro®, Eli Lilly, Indianapolis, Indiana) treatment), while the cross-over to thrombectomy occurred in only four patients of the direct stenting alone arm. The AngioJet® device (Possis Medical Inc., Minneapolis, Minnesota) directly crossed the target lesion in nearly all cases, and only two patients needed predilation with a small balloon before thrombectomy. Early ST-segment resolution and infarct size were assessed in two predefined and narrow temporal windows (30 minutes from the end of the procedure for ST-segment resolution and one month for scintigraphy). The study was powered to detect a difference in early ST-segment resolution and no differences in clinical outcome could be revealed because of the small number of patients and the criteria used for the enrollment (patients with a history of previous MI were excluded).
The AngioJet in Acute Myocardial Infarction Trial
The AngioJet in Acute Myocardial Infarction (AiMI) trial is a multi-centre, randomised trial that compared rheolytic thrombectomy before angioplasty and stenting of the infarct artery and is based on a sample of 480 patients.5 The primary end-point of the study was final infarct size as assessed by sestamibi scintigraphy. Secondary end-points were TIMI flow grade, tissue perfusion myocardial blush according to the TIMI criteria, >70% ST-segment elevation resolution, and major adverse cardiac events (MACEs) at one month. The study did not show any benefit of rheolytic thrombectomy on the effectiveness of reperfusion.
The final infarct size was 12.5% + 12.1% in the thrombectomy arm and 9.8% + 10.9% in the control arm (p = 0.03), while the final TIMI grade 3 flow was more frequent in the control arm (97% vs 91.8%, p <0.02). No significant differences between arms were revealed in the ST-segment resolution or myocardial perfusion blush score. MACE rate was higher in the thrombectomy arm (6.7% vs 1.7%, p = 0.01) and the difference was driven by the mortality rate (4.6% vs 0.8%, p = 0.02).
The unexpected and harmful results of this study raised several concerns about the study design, patient enrolment and thrombectomy technique. First, the incidence of patients with an occluded infarct artery at baseline angiography was low in both groups and a baseline TIMI grade 3 flow was more frequent in the control arm (26% vs 19%, p <0.05). This figure may explain, at least in part, the low incidence of baseline moderate to large thrombus (21.9% in the thrombectomy arm and 19.6% in the control arm) and the smaller infarct size of the control arm compared with the thrombectomy arm. Again, the high incidence of already spontaneously reperfused patients before PCI may explain the small infarct sizes in both arms. At the same time, a median value of infarct size of only 12.5% of the thrombectomy arm makes difficult to understand the unexpected high mortality rate at one month (4.6%). Another concern is the thrombectomy technique: the investigators used a distal to proximal approach without the device activation while crossing the occlusion, and this procedural characteristic could have favoured embolisation.
The conflicting results of the two randomised trials suggested the opportunity to perform another study with a different design that overcomes the confounding effect of several procedural variables (predilation, postdilation or direct stenting at the discretion of the operator, thrombectomy performed with a retrograde technique).
The JETSTENT Trial
The AngioJET Thrombectomy and STENTing for Treatment of Acute Myocardial Infarction (JETSTENT) trial is a multi-centre, international randomised trial that will compare rheolytic thrombectomy before direct infarct artery stenting with direct stenting alone in patients with acute MI (principal investigators: the author and Antonio Colombo).
There are some differences in study design compared with the Florence trial design. In order to enrol a real-world AMI population, patients with a history of previous MI will be enrolled. No restriction based on clinical status on presentation or high-risk coronary anatomy will be used. Thus, patients with cardiogenic shock will be included. Inclusion angiographic criteria are the evidence of a TIMI grade 3-5 thrombus and a reference vessel diameter 2.5mm. Thrombectomy will be performed using an antegrade technique (the first rheolytic thrombectomy pass is made using a proximal-to-distal approach with activated device). The two co-primary end-points of the study are ST-segment resolution at 30-45 minutes post-PCI, assessed by 12-lead electrocardiogram (ECG) and infarct size as measured by Tc-99m sestamibi imaging at 30 days. Key secondary clinical end-points are: six-month and 12-month mortality and hospital admission for congestive heart failure.
The primary hypothesis of the study is that rheolytic thrombectomy before infarct artery stenting provides a better reperfusion as assessed by ST-segment elevation resolution and scintigraphic infarct size. The subsequent increase in myocardial salvage could result in a decreased incidence of death and congestive heart failure due to left ventricle (LV) remodelling at a mid-term follow-up.
The sample size of the JETSTENT trial will be 500 patients and is calculated on the statistical hypothesis of a 30% reduction in infarct size in patients randomised to thrombectomy (power 90%). The study started on January 2006, and the end of enrolment is planned for December 2006.