Methods

Data on all patients undergoing PCI are routinely collected in our center. To identify patients with ST we performed a systematic search in our electronic database of all repeated PCI procedures performed in 11,346 patients who had undergone a first stent placement from January 2002 through May 2010 at our center. After filtering out elective repeated PCI procedures, all remaining coronary angiograms were systematically reviewed for evidence of ST by an experienced cardiologist (F.Z.). Subsequently ST was independently confirmed by 2 experienced cardiologists (B.S. or A.H.) based on coronary angiograms and hospital records. Also, subsequent angiograms and hospital records were reviewed for occurrence of target vessel revascularization and recurrence of ST. Baseline characteristics and follow-up were collected from a prospectively recorded database and missing data were completed by hospital record review and telephone interviews. Follow-up on all-cause 30-day and 1-year mortality was collected using municipal civil registries. These registries provide completeness of follow-up for vital status in >99% of patients admitted to our department. All patients with definite ST were included in this analysis.

All patients underwent PCI and were pretreated with aspirin (500 mg), clopidogrel (600 mg), and heparin (5,000 IU). Periprocedural glycoprotein IIb/IIIa inhibitors were used unless contraindicated. Thrombus aspiration was performed at the operator’s discretion. For all patients the first procedural step was the passing of a steerable guidewire through the target lesion. For patients in the “thrombus aspiration” group, this step was followed by the advancing of a 6Fr Export Aspiration Catheter (Medtronic, Minneapolis, Minnesota) or a 6Fr Diver Clot Extraction Catheter (Invatec, Bethlehem, Pennsylvania) into the target coronary segment during continuous aspiration. For patients in the “conventional PCI” group, this step was followed by balloon dilatation to establish anterograde flow. When judged necessary by the operator, additional balloon dilatation and/or stent placement was performed in the 2 groups. Patients were categorized into the thrombus aspiration group when thrombus aspiration was attempted regardless of its success. Because all patients had ST, aspirin was usually prescribed indefinitely and clopidogrel for ≥1 year.

Thrombus aspiration was deemed successful when thrombotic material was retrieved. In 6 patients histopathologic analysis of filtered material obtained by thrombus aspiration was performed. Material was placed in formalin and fixed for 24 hours. Thereafter, filtered material was pelleted by centrifugation in liquid agar at 65°C in an Eppendorf tube. After the agar pellet was solidified at 4°C, it was embedded in paraffin using an automated tissue processor. Paraffin sections were cut at 4 μm and stained with hematoxylin and eosin for microscopic examination (100×). Immunostaining was performed to optimize visualization of endothelial cells, smooth muscle cells, and macrophage foam cells. Identified material was analyzed for the presence of 3 components: platelets, erythrocytes, and atheromatous plaque (defined as any fragment of the vessel wall, cholesterol crystals, inflammatory cells, or collagen tissue). Size was classified into 5 groups: residue (very small filter casts of loosely cohesive platelets) and well-formed thrombi <0.5, 0.5 to 1.0, 1.0 to 2.0, and >2 mm.

ST was defined as an acute coronary syndrome with angiographic evidence of thrombus or occlusion, thus meeting Academic Research Consortium (ARC) criteria for “definite” ST. Furthermore, patients were classified by timing of ST by ARC classification: early ST (onset of ST 0 day to 30 days after initial stent placement), late ST (31 to 360 days), and very late ST (>360 days). Hypertension and hypercholesterolemia were defined as a documented history of this condition warranting medical therapy. Ischemic time was defined as time from symptom onset to initial intracoronary therapy by thrombus aspiration or balloon inflation of the infarct-related coronary artery. Thrombolysis In Myocardial Infarction (TIMI) flow and myocardial blush grade were recorded as previously described. Lesion calcium was identified as radio-opacities within the vascular wall of the target vessel on angiogram. When multiple stents were placed, total stent length was defined as the sum of individual stent lengths placed in the target coronary artery and minimal stent diameter was defined as the smallest diameter of any stent placed in the target coronary artery. Postprocedure distal embolization was defined as an angiographically visible distal filling defect with abrupt cutoff in the vessel located distally of the ST-related lesion. Distal embolization could be assessed when epicardial coronary flow was present through the infarct-related lesion and distal vessels were visualized. TIMI thrombus grade at the target lesion site was defined according to Gibson et al. Myocardial infarction was defined as recurrent symptoms with new ST-segment elevation and increase of levels of cardiac markers to ≥2 times the upper limit of the normal range. Recurrence of ST was defined as recurrence of definite ST in the target vessel according to ARC criteria. Target vessel revascularization was defined as ischemia-driven revascularization of the ST-related vessel performed using PCI or coronary artery bypass grafting. Occurrence of major adverse cardiac events (MACEs) was defined as death, myocardial infarction, recurrence of ST, or target vessel revascularization.

Continuous variables were summarized as mean ± SD or median and interquartile range. Discrete variables were presented as fractions and percentages. To calculate p values we used Student’s t test for normally distributed continuous variables, Mann–Whitney U test for nonparametric continuous variables, and Pearson chi-square test for categorical variables. For ordinal variables the p value for trend was obtained. Occurrence of MACEs and the separate components was estimated and plotted using the Kaplan–Meier method and tested with log-rank test.

Multivariable logistic regression models with bootstrap model selection were fitted to assess independent predictors of postprocedure TIMI grade 3 flow and postprocedure myocardial blush grade 2 or 3. This method has been used previously in the context of an ST population to avoid an overfit model. Among all baseline and procedural characteristics listed in Tables 1 and 2 , potential predictors of postprocedure TIMI grade 3 flow and myocardial blush grade 2 or 3 with a p value <0.15 in univariable analysis were selected for multivariable analysis. For these variables bootstrap selection with 200 models was performed. Predictors selected in >140 of the models (70%) were included in the final model. Goodness of fit was assessed with Hosmer–Lemeshow test. A potentially confounding effect of improvements other than thrombus aspiration throughout the study period was investigated in our multivariable models but was found to be absent. For all analyses a p value <0.05 (2-tailed) was considered statistically significant. Statistical analyses were performed with SPSS 16.0.2 (SPSS, Inc., Chicago, Illinois) and STATA 11.0 (STATA Corp., College Station, Texas).