Early vascular responses to everolimus-eluting cobalt–chromium stent in the culprit lesions of st-elevation myocardial infarction: results from a multicenter prospective optical coherence tomography study (MECHANISM-AMI 2-week follow-up study)

The MECHANISM-AMI-2W (Multicenter Comparison of Early and Late Vascular Responses to Everolimus-eluting cobalt–CHromium Stent and platelet AggregatioN studIeS for TreatMent of Acute Myocardial Infarction 2 weeks FD-OCT follow-up) study was planned for specific purposes (ClinicalTrials.gov ID: NCT02014753, UMINID: UMIN000012616). Prior to MECHANISM-AMI-2W, a pilot study (MECHANISM-Pilot) was conducted to determine the optimal FD-OCT parameters [11]. The MECHANISM-AMI-2W study enrolled patients with STEMI who underwent FD-OCT immediately after primary PCI and were eligible for follow-up FD-OCT assessment at 2 weeks after implantation. Patients were recruited between April 2014 and May 2015 from 13 Japanese institutions. STEMI was defined in accordance with the Third Universal Definition of Myocardial Infarction [12]. Exclusion criteria were: cardiogenic shock, culprit lesion of left main coronary artery, reference vessel diameter < 2.0 or ≥ 4.5 mm, chronic kidney disease as indicated by serum creatinine > 2.0 mg/dl, maintenance hemodialysis, comorbid cancer with expected survival < 2 years, scheduled surgery within 3 months, history of adverse reactions to aspirin or clopidogrel, warfarin intake before STEMI onset, age < 20 years, pregnancy, and STEMI at prior stented segment.

Culprit lesions were treated with either one or two CoCr-EES (Xience Prime/Xpedition, Abbott vascular, Santa Clara, CA, USA), following thrombus aspiration therapy if needed. Concurrent use of distal protection devices was dependent on the operator’s discretion. Immediately after the procedure, FD-OCT image acquisition was performed throughout the stented segment, with a margin segment ≥ 5 mm from the culprit lesion. A loading dose of aspirin (81–200 mg) and either clopidogrel (300 mg) or prasugrel (20 mg) (the standard Japanese loading doses) was administered before PCI. Dual antiplatelet therapy with aspirin and thienopyridine was recommended for 12 months after PCI. This study complies of the Declaration of Helsinki. The study protocol was approved by the Ethical Committee of each participating institution. All patients provided written informed consent before inclusion.

Coronary imaging was performed immediately after PCI, and after 2 weeks (14 ± 4 days) during staged PCI for residual stenotic lesions or follow-up angiography for pre-discharge evaluation. Intracoronary injection of nitroglycerin (isosorbide dinitrate or nitroglycerin) was administered. Angiograms were obtained from a standard series of multiple projections. All angiographic images were stored on a CD/DVD-ROM for offline analysis. FD-OCT images were acquired with an OCT system (ILUMIEN or ILUMIEN OPTIS) and an FD-OCT catheter (Dragonfly/Dragonfly JP/Dragonfly OPTIS) (St. Jude Medical, St. Paul, MN, USA). All FD-OCT images of the target lesion site (coronary segment from ≥ 5 mm distal and 5 mm proximal to the target lesion) were recorded with an automatic pullback system. Although the use of a contrast agent as a flushing liquid was recommended, lactated Ringer’s solution or low-molecular weight dextran L were also acceptable. FD-OCT images were digitally stored and submitted for offline analyses.

Quantitative coronary angiography was performed at an independent core laboratory (Cardiocore Japan, Tokyo, Japan) with a standard software package (QAngio XA; Medis medical imaging systems, Leiden, The Netherlands). The stented segment and 5-mm peri-stent segments were analyzed using a standard technique. FD-OCT image analysis was performed by an independent core laboratory (Kobe Cardiovascular Core Laboratory, Kobe University Graduate School of Medicine, Kobe, Japan) using ILIMIEN proprietary software (St. Jude Medical, St. Paul, MN, USA).

Quantitative FD-OCT analyses were performed with every 1-mm interval for the assessment of lumen, stent, and intra-stent tissue area (stent area–lumen area). Intra-stent tissue thickness was measured from the lumen border to the center of the strut blooming. Struts with intra-stent tissue thickness equal to 0 μm were defined as uncovered struts. A strut that was partially covered with tissue was classified as an uncovered strut. A maximum distance of > 108 μm between the center reflection of the strut and adjacent vessel surface was defined as malapposed struts.

Qualitative FD-OCT analyses was performed for every single frame to assess (1) thrombus, a mass ≥ 250 μm in height with an irregular surface, protrusion into the lumen, and significant attenuation behind the mass; (2) prolapse, a signal-rich protruding tissue ≥ 250 μm in height, located between the struts with a smooth surface and no attenuation; (3) intra-stent tissue, combinations of entire intra-stent tissue between lumen and stent, and (4) dissection, evidence of flaps ≥ 200 μm in height within the stent segment and stent margin, according to the published methods [13, 14]. Maximal area, maximal height or depth, and longitudinal length were measured for each mass or segment individually (Fig. 1). Total length was defined as the sum of the longitudinal length of each composition. Each % longitudinal length to stent was calculated as: the number of cross-sections with thrombus/prolapse/malapposition/dissection × 100/the total number of cross-sections within the stented segment.

The primary endpoint was % uncovered strut. Secondary endpoints included the representative described FD-OCT parameters of thrombus, prolapse, strut coverage/malapposition, and dissection, also corresponding to them.

The baseline, lesion and procedural characteristics of the study population are shown in Table 1. A total of 52 patients (age 63.7 ± 11.7 years, male 85.0%) were enrolled with high frequencies of diabetes mellitus (42.3%) and smoking (50.0%). Following thrombus aspiration in 80.8% of patients, 1.15 ± 0.36 stents were used. Two patients (3.8%) received chronic antiplatelet therapy with aspirin prior to the onset of STEMI. All patients started to receive thienopyridine from this event and continued dual antiplatelet therapy, primarily with clopidogrel (98.1%), for 2 weeks. Neither major adverse events nor target vessel revascularizations occurred within 2 weeks.

Table 2 demonstrates the findings of quantitative and qualitative coronary angiography. Thrombolysis in Myocardial Infarction (TIMI) grade 3 flow was achieved in 88.5% of patients. The angiographic reference diameter was 2.88 ± 0.53 mm at post-procedure. There were no statistical differences in angiographic parameters between post-procedure and 2 weeks follow-up except in-lesion reference diameter.

Figure 2 depicts the study flow and lesions eligible for analysis. Among the 52 patients enrolled, complete FD-OCT assessment was performed in 45 lesions at post-procedure and 49 lesions at 2-week follow-up. The primary endpoint of % uncovered struts was 20.7 ± 13.7% (of 49 patients). The other general quantitative FD-OCT variables based on these entire patients (n = 45 at post-procedure and n = 49 at 2-weeks follow-up) are summarized in Table 3.

Table 4 displays the results of serial quantitative FD-OCT analysis. A total of 44 lesions were available for serial analyses. A representative case example of serial FD-OCT cross-sections is shown in Fig. 3. There were significant improvements in uncovered struts and malapposed struts from post-procedure to 2-week follow-up, with decreases from 63 ± 20 to 21 ± 14% (p < 0.0001) and 7.3 ± 9.0 to 4.7 ± 5.9% (p = 0.005), respectively. Their individual serial changes are plotted in Fig. 4.

The number of thrombus masses ≥ 250 μm was significantly decreased. The majority of quantitative parameters of intra-stent thrombus, including longitudinal length, height, and area, were significantly reduced (% longitudinal length to stent: 28.8 ± 22.7 vs. 18.1 ± 20.2%, p = 0.0001; maximum area 0.93 ± 0.87 vs. 0.65 ± 0.63 mm², p = 0.034). A similar trend was observed for prolapse. As a consequence, the minimum and average lumen areas were larger at 2 weeks (4.96 ± 1.56 vs. 5.27 ± 1.56 mm², p = 0.040 and 6.49 ± 1.82 vs. 6.71 ± 1.89 mm², p = 0.048, respectively). Individual serial changes in minimal lumen area, maximum thrombus area, and percent longitudinal length of thrombus from post-procedure to 2-week follow-up are plotted in Fig. 5. The number of dissection flaps > 200 μm was also significantly reduced.

Despite the thrombus-rich environment, this study demonstrated a significant reduction of in-stent thrombus within 2 weeks; this corresponded to a substantial period in which the risk of thrombotic complications was increased after coronary stenting. The present study indicates that the thrombus size can be expected to decrease by an average of approximately 50% (with 60% of the length and 80% of the area of the thrombus preserved inside stent) at 2 weeks after CoCr-EES implantation.

The high antithrombotic potential of the fluorinated polymer of CoCr-EES has been indicated by several experiments [3, 18], which partially accounts for the striking reduction of in-stent thrombus in this study, as well as the low rate of stent thrombosis in other clinical studies [4, 5, 19]. Furthermore, in terms of drugs, the other important component of a drug-eluting stent, a biolimus-eluting stent with a bare metal (non-polymer) luminal surface was found to reduce thrombotic events in comparison to bare metal stents [20], indicating that exposure of the vascular wall to biolimus has an independent inhibiting effect on thrombus formation. Thus, both the polymer and drug components of drug-eluting stents have independent antithrombotic effects, and optimal combinations may be advantageous for preventing early stent thrombosis.

The present results support the current expert consensus on the treatment of STEMI. This study described the approximate temporal changes after stent implantation during emergent procedures, which should be extremely helpful for physicians. Furthermore, the potential of FD-OCT-guided techniques were revealed. Because high-resolution FD-OCT enables the precise visualization of stent expansion, strut apposition, strut embedment, and the residual thrombus burden during the procedure [21, 22], it may be associated with accelerated vascular healing reactions within the initial 2 weeks. However, to confirm this hypothesis, it will be necessary to perform further randomized clinical studies to compare the vascular responses after stent implantation with or without FD-OCT guidance.

Furthermore, the findings may contribute to the ongoing debate about the optimal strategies of dual antiplatelet therapy, which have been the greatest concerns in the clinical arena. The vast majority of patients in this study received clopidogrel, which was mostly identical to the EXAMINATION study [4]; this prompts speculation as to what happened in that monumental study. Currently, other regimens that include prasugrel or ticagrelor are more frequently used in the actual clinical setting. This factor is likely to have a different impact on the clinical outcome [23, 24]. Since the publication of key study findings [25], prolonged dual antiplatelet therapy remains the recommended strategy. Nevertheless, the drug regimen should be repeatedly re-evaluated and altered based on the elevated bleeding risk. Currently, several prospective, multicenter, randomized controlled clinical trials, such as the GLOBAL LEADERS trial [26] and the STOPDAPT 2 study, in which dual antiplatelet therapy is continued for 1 month, are ongoing. The results of the present study, which focus on the safety mechanisms of drug-eluting stents, should be incorporated into comprehensive clinical trials for future constructive discussions on shortening the duration of dual antiplatelet therapy.

The population of the present study was relatively small. However, considering the extensive effort required for FD-OCT analysis, the sample size was at the most attainable goal. In this respect, it was similar to the goals of previous investigations. Another limitation is the lack of a control group for statistical comparisons. Although direct comparison with patients who received bare metal stents would be ideal, designing a head-to-head comparison between bare metal stent (BMS) and drug-eluting stents is ethically difficult due to strong clinical evidence [4, 5, 20]. Furthermore, similar vascular responses might be observed if other DESs or BMSs were applied in this study, as indicated in our previous pilot study, which included a very small number of patients (n = 6 for each) [11]. However, considering the significant reduction in the incidence of clinical thrombotic events within the initial 2 weeks in the EXAMINATION study, the early vascular responses of patients who receive a CoCr-EES might be found to differ from those of patients who receive a BMS if an adequate number of patients could be enrolled. A third restriction is a potential selection bias due to the small number of patients who were eligible for FD-OCT evaluation during emergent procedures: the enrolment of cases with severe hemodynamic conditions tended to be avoided. Moreover, the lesion morphology and site must be suitable for FD-OCT evaluation. This study mainly included patients with multi-vessel disease because FD-OCT assessments were scheduled during staged PCI. A fourth limitation is that it might be difficult to accurately distinguish thrombus from tissue prolapse because white (platelet) thrombus can also be detected as a homogenous high signal without attenuation. In the present study, thrombus was defined as a mass of ≥ 250 μm in height with an irregular surface, protrusion into the lumen, and significant attenuation behind the mass. Accordingly, some white thrombi might have been misclassified as protrusions in the present study. However, the classification methods for thrombi and protrusions that we applied had previously been established by Soeda et al. [21], whose findings are considered to be important for predicting long-term clinical outcomes. Furthermore, the combination of such intra-stent tissues was significantly reduced from after intervention to the 2-week follow-up examination, which is consistent with the temporal changes in thrombi that were defined in this study. Finally, although several definitions of OCT parameters have been reported [10, 21], we decided to perform our analysis using a pre-specified definition based on the pilot study. The results of clinical studies are affected to varying degrees by differences in definitions or the sampling rates of cross-sections [27]; however, the serial data set that was used in the present study was considered to substantially minimize the influences on vascular behaviors.