Correlates of Adverse Events During Saphenous Vein Graft Intervention With Distal Embolic Protection

Transcription

1 JACC: CARDIOVASCULAR INTERVENTIONS VOL. 1, NO. 2, BY THE AMERICAN COLLEGE OF CARDIOLOGY FOUNDATION ISSN /08/$34.00 PUBLISHED BY ELSEVIER INC. DOI: /j.jcin Correlates of Adverse Events During Saphenous Vein Graft Intervention With Distal Embolic Protection A PRIDE Substudy Ajay J. Kirtane, MD, SM, FACC,* Eugene R. Heyman, PHD, Christopher Metzger, MD, FACC, Jeffrey A. Breall, MD, PHD, FACC, Joseph P. Carrozza JR, MD, FACC New York, New York; Montgomery Village, Maryland; Kingsport, Tennessee; Indianapolis, Indiana; and Boston, Massachusetts Objectives We sought to identify correlates of 30-day adverse events among patients undergoing percutaneous coronary intervention (PCI) of saphenous vein grafts (SVG). Background Although the use of distal embolic protection devices during SVG intervention reduces major adverse cardiac events (MACE), many patients still experience MACE despite distal embolic protection, and the major predictors of MACE among these patients are not well-characterized. Methods Correlates of 30-day MACE and peri-procedural creatine kinase-myocardial band (CK-MB) elevation were assessed among 631 patients undergoing SVG intervention with distal embolic protection enrolled in the PRIDE (PRotection during saphenous vein graft Intervention to prevent Distal Embolization) study, a randomized comparison of the TriActiv System (Kensey-Nash Corp., Exton, Pennsylvania) with an active control group (Guardwire [Medtronic, Santa Clara, California] or Filterwire [Boston Scientific, Minneapolis, Minnesota]). Results Baseline covariates associated with MACE were longer lesion length, greater angiographically assessed estimated plaque volume, and higher SVG degeneration score. Graft age and angina class were not associated with adverse events. Angiographic lesion length was significantly correlated with more complex angiographic metrics such as estimated plaque volume and the SVG degeneration score. In multivariable analyses, angiographic lesion length was the strongest independent correlate of MACE (odds ratio [OR] 2.81 [95% confidence interval (CI) 1.82 to 4.34]/log-increase in lesion length, p 0.001) with a graded increase in MACE observed with increasing lesion lengths. Similarly, the strongest independent correlate of CK-MB elevation was lesion length (OR 2.54 [95% CI 1.59 to 4.04]/log-increase in lesion length, p 0.001). The associations between lesion length and both MACE and CK-MB elevation were consistent among the studied embolic protection devices (TriActiv, Guardwire, or Filterwire). Conclusions Angiographic lesion length was the strongest correlate of short-term adverse events among patients undergoing SVG intervention with distal embolic protection, with incremental effects noted at even relatively short lesion lengths. (J Am Coll Cardiol Intv 2008;1:186 91) 2008 by the American College of Cardiology Foundation From the *Center for Interventional Vascular Therapy, Columbia University Medical Center, New York, New York; Montgomery Village, Maryland; The Heart Center, Kingsport, Tennessee; Krannert Institute of Cardiology, Indiana University of Medicine, Indianapolis, Indiana; and the Section of Interventional Cardiology, Division of Cardiology, Department of Medicine, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, Massachusetts. This work was supported in part by a grant from Kensey-Nash, Inc. Manuscript received September 18, 2007; revised manuscript received December 13, 2007, accepted January 23, 2008.

2 JACC: CARDIOVASCULAR INTERVENTIONS, VOL. 1, NO. 2, The percutaneous treatment of diseased saphenous vein grafts (SVGs) represents one of the major current challenges of current percutaneous coronary intervention (PCI). Despite major advances in PCI technique and technology, patients with diseased SVGs still represent one of the highest-risk subgroups of patients undergoing PCI. Percutaneous treatment of SVGs has been associated with a greater incidence of slow flow and higher rates of periprocedural myocardial infarction, largely owing to distal embolization of the abundant and friable atherosclerotic debris in diseased SVGs. Although the embolic protection with either balloon occlusion devices or filter-based devices during SVG PCI reduces major adverse cardiac events (MACE), many patients undergoing SVG PCI still experience adverse events (1 5). Prior reports of the major correlates of MACE among patients undergoing contemporary SVG PCI have included a mixed cohort of patients treated both with and without embolic protection (3,6 10) or have been limited by small sample size (11). Therefore we sought to identify correlates of adverse events among a cohort of SVG PCI patients uniformly treated with distal embolic protection. Methods Data were compiled from 631 subjects undergoing SVG PCI in the PRIDE (PRotection during saphenous vein graft Intervention to prevent Distal Embolization) Study. The PRIDE was a prospective, randomized clinical trial of the TriActiv System (Kensey-Nash Corp., Exton, Pennsylvania) device with an active control (either Guardwire [Medtronic, Santa Clara, California] or Filterwire [Boston Scientific, Minneapolis, Minnesota]) in patients undergoing PCI of an SVG (12). Criteria for inclusion were age 21 years, angina or objective evidence of ischemia, lesion in an SVG ( 3.0 and 5.0 mm in diameter), Thrombolysis In Myocardial Infarction (TIMI) flow grade 0, and ability to provide informed consent. Patients were excluded if any of the following were present: pregnancy; lesion in a native artery or internal mammary graft; distal shoulder of the lesion within 2.0 cm of the distal anastamosis; left ventricular ejection fraction 25%; sequential grafts, unless the lesion was 2 cm proximal to the branch point; myocardial infarction (creatine kinase [CK] and creatine kinasemyocardial band [CK-MB] more than twice the upper limit of normal within 24 h); allergy to aspirin, clopidogrel, or ticlopidine; treatment of 2 SVGs; and comorbidities limiting life-expectancy to 6 months. The PRIDE study was approved by the institutional review board at each site; all patients provided written informed consent to participate. Randomization was stratified by intention to administer a glycoprotein IIb/IIIa receptor antagonist before intervention. Patients received aspirin before the procedure and either heparin or bivalirudin during the procedure. After the procedure, aspirin and either clopidogrel or ticlopidine were administered for a minimum of 1 month. Cardiac enzymes were routinely assessed every 8 h for 24 h. Patients were assessed clinically at 30 days. For this analysis, subjects included were those in Cohort 2 of the trial (randomized comparison of TriActiv System vs. Guardwire or Filterwire). Subjects in Cohort 1 (randomized comparison of TriActiv System vs. no embolic protection) were not included. All angiographic data were assessed offline by an independent core laboratory with quantitative coronary angiography methodology that has been described previously (3). The primary end point for this analysis was the occurrence of 30-day MACE (cardiac death, myocardial infarction [any post-procedure CK-MB level 3 the upper limit of normal], or target lesion revascularization) on a per-patient level. The MACE was adjudicated by an independent clinical events committee blinded to device assignment. Peri-procedural myocardial infarction (as defined earlier) was a secondary end point. Statistical analysis. Continuous variables are reported as means with SDs and categorical variables as rates or proportions. The t test was used to compare continuous variables, and the chi-square test was used for comparisons of categorical variables. Variables that were not normally distributed were appropriately transformed (e.g., lesion length was left-skewed and was therefore log-transformed with natural logarithms). Logistic regression multivariable analyses were performed, including candidate covariates associated with the outcome of interest (MACE, CK elevation) with p 0.05 in univariate analyses. Stepwise selection was used to develop final models. All analyses were performed by an independent study statistician (ERH) with SAS Version 8.2 (SAS Institute, Cary, North Carolina). Results Abbreviations and Acronyms CI confidence interval CK-MB creatine kinasemyocardial band MACE major adverse cardiac events OR odds ratio PCI percutaneous coronary intervention SVG saphenous vein graft The mean patient age was 69 years, with 82% male and 41% diabetic subjects (Table 1). The mean graft age was 12 years, and treated grafts had a mean reference diameter of 3.3 mm, a mean percent diameter stenosis of 65%, and a mean lesion length of 14 mm. Angiographic thrombus was present in 58% of lesions, and 11% of grafts had abnormal TIMI flow grade. Coronary stents were successfully delivered in 585 of 631 (92.7%) patients. After PCI, the mean residual diameter stenosis was 14%, with 11% of grafts having abnormal TIMI flow grade. Device success was observed in 99% of

3 188 JACC: CARDIOVASCULAR INTERVENTIONS, VOL. 1, NO. 2, 2008 Table 1. Baseline Characteristics Parameter MACE No MACE p Value Age, yrs (mean SD) Male gender 51 (76.1%) 465 (82.5%) 0.20 CCS class I 8 (12.1%) 75 (13.4%) 0.51 CCS class II 12 (18.2%) 146 (26.1%) CCS class III 26 (39.4%) 159 (28.4%) CCS class IV 20 (30.3%) 179 (32.0%) Prior MI 49 (75.4%) 352 (65.1%) 0.10 Diabetes mellitus 27 (40.3%) 235 (41.7%) 0.83 Dyslipidemia requiring medication 58 (86.6%) 486 (86.2%) 0.93 Hypertension 61 (91.0%) 472 (84.1%) 0.14 Cigarette smoking 18 (27.3%) 90 (16.4%) Family history of premature CAD 19 (36.5%) 180 (37.4%) 0.90 Positive functional test 22 (32.8%) 208 (37.3%) 0.48 GP IIb/IIIa planned 37 (55.2%) 297 (52.7%) 0.69 Before PCI Mean graft age (yrs) Graft to LAD 10 (15.2%) 146 (26.0%) Graft to LCX 26 (39.4%) 235 (41.8%) 0.71 Graft to RCA 31 (47.0%) 203 (36.1%) 0.08 Lesion length Reference vessel diameter Minimum lumen diameter Baseline % stenosis % with thrombus 44 (66.7%) 323 (57.7%) 0.16 Baseline TIMI flow grade SVG degeneration (4-point scale) Plaque volume After PCI Reference vessel diameter Minimum lumen diameter % stenosis % with thrombus 13 (19.7%) 54 (9.6%) TIMI flow grade CAD coronary artery disease; CCS Canadian Cardiovascular Society; GP IIb/IIIa glycoprotein IIb/IIIa receptor antagonist; LAD left anterior descending (artery); MACE major adverse cardiac events; MI myocardial infarction; PCI percutaneous coronary intervention; RCA right coronary artery; RCX right circumflex (artery); SVG saphenous vein graft; TIMI Thrombolysis In Myocardial Infarction. cases, with no significant differences between treatment groups in the overall cohort (p 0.68 for comparison with TriActiv System vs. active control). Adjudicated MACE was observed in 67 (10.6%) patients, with CK-MB elevation occurring in 56 of 549 (10.2%) patients. The univariate correlates of MACE and CK-MB elevation are shown in Table 2. The only baseline covariates associated with MACE were history of cigarette smoking, lesion length, estimated plaque volume, and SVG degeneration score. The baseline covariates associated with periprocedural CK-MB elevation were very similar and included female gender, history of cigarette smoking, lesion length, estimated plaque volume, and SVG degeneration score. Of note, neither graft age nor angina class at presentation was associated with adverse peri-procedural outcomes (Table 1). A strong and graded association between increasing lesion length and the incidence of MACE and CK-MB elevation was observed (Fig. 1). Lesion length was significantly correlated with estimated angiographically estimated plaque volume as well as the SVG degeneration score (r 0.83 for plaque volume; r 0.49 for SVG degeneration score). Because lesion length, SVG degeneration score, and angiographic plaque volume were highly correlated, only the covariate with the strongest association with the study end points (lesion length) was included in final multivariable models to avoid multicollinearity. In multivariable analyses, lesion length remained the strongest independent correlate of MACE (odds ratio [OR] 2.81 [95% confidence interval (CI) 1.82 to 4.34]/log-increase in lesion length, p 0.001). Similarly, the most significant independent correlate of peri-procedural CK-MB elevation

4 JACC: CARDIOVASCULAR INTERVENTIONS, VOL. 1, NO. 2, Table 2. Univariate Predictors of CK-MB Elevation and Adjudicated MACE CK-MB >3 ULN Adjudicated MACE Parameter OR (95% CI) p Value OR (95% CI) p Value Male gender 0.53 ( ) ( ) 0.2 Age (/yrs) 1.00 ( ) ( ) 0.52 CCS class 1.02 ( ) ( ) 0.51 Prior MI 1.35 ( ) ( ) 0.1 PVD 0.82 ( ) ( ) 0.46 Prior CVA/TIA 0.74 ( ) ( ) 0.8 Diabetes mellitus 0.72 ( ) ( ) 0.83 Dyslipidemia requiring medication 0.81 ( ) ( ) 0.93 Hypertension 2.34 ( ) ( ) 0.14 Cigarette smoking 2.09 ( ) ( ) Family history of premature CAD 1.17 ( ) ( ) 0.9 Positive functional test 0.82 ( ) ( ) 0.48 Age of vein graft 0.98 ( ) ( ) 0.51 Pre-PCI RVD 1.32 ( ) ( ) 0.27 Pre-PCI % stenosis 1.00 ( ) ( ) 0.75 Pre-PCI thrombus 1.15 ( ) ( Baseline TIMI flow grade ( ) ( ) 0.62 Log lesion length, mm 2.54 ( ) ( ) Thrombus grade 1.16 ( ) ( ) 0.16 SVG degeneration (4-point scale) 1.56 ( ) ( ) Baseline TIMI flow grade 1.35 ( ) ( ) 0.66 Baseline % stenosis 1.00 ( ) ( ) 0.75 Log plaque volume 1.97 ( ) ( ) CI confidence interval; CK-MB creatine kinase-myocardial band; CVA cardiovascular accident; OR odds ratio; RVD reference vessel diameter; TIA transient ischemic attack; ULN upper limit of normal; other abbreviations as in Table 1. 3 the upper limit of normal was lesion length (OR 2.54 [95% CI 1.59 to 4.04]/log-increase in lesion length, p 0.001). The association between lesion length and either MACE or CK-MB elevation was significant and consistent among the 3 different types of distal embolic protection device used in the trial and regardless of whether the use of glycoprotein IIb/IIIa inhibitors was planned (all p values for interaction not significant). Discussion In this analysis of 631 patients undergoing SVG PCI using distal embolic protection, angiographic lesion length was the only independent correlate of adverse clinical outcomes. Despite the use of embolic protection, the association between lesion length and adverse outcomes was graded and continuous, with 20% rates of peri-procedural MACE observed for lesions 20 mm in length. The use of embolic protection devices as an adjunct to SVG PCI has been demonstrated to reduce the occurrence of peri-procedural events by approximately 40% (1). Although this represents a significant relative and absolute reduction in adverse events for one of the highest-risk subsets of current PCI, the rate of adverse peri-procedural events remains high even with the use of embolic protection, and these peri-procedural events have been associated with significant morbidity and economic cost (13,14). Thus, identification of patients at the highest risk for periprocedural complications can lead to appropriate riskstratification of patients before SVG PCI. Patients with longer lesions in our study were at the highest risk of adverse events. Nonetheless, the observation of a 20% rate of MACE with lesions above 20 mm (a lesion length not traditionally considered extensive for conventional PCI) suggests that there remains significant room for improvement in the percutaneous treatment of SVG lesions, even if greater penetrance of these devices is achieved. A recent report from the American College of Cardiology-National Cardiovascular Data Registry describes 25% use of embolic protection devices for SVG PCI overall (15). Whether higher-risk patients such as those with longer lesions should therefore lower the threshold for the use of an embolic protection device remains to be determined, but longer lesion length was associated with a higher incidence of the use of an embolic protection device in that registry, sug-

5 190 JACC: CARDIOVASCULAR INTERVENTIONS, VOL. 1, NO. 2, 2008 % Incidence CK-MB >3X ULN Adjudicated MACE gesting that operators might be influenced by this lesion characteristic. The observed association between lesion length and outcomes in this study parallels that of other studies that have examined risk factors for adverse events in patients undergoing PCI with and without embolic protection (3,9). Longer lesions are thought to be representative of more severe degenerative disease within an SVG (and therefore a larger burden of debris with additionally greater potential to embolize). The observed correlation in this analysis between lesion length and angiographic plaque volume and the semi-quantitative SVG degeneration score validates this theory and is consistent with prior observations that estimated plaque volume and diffuse disease within SVGs is correlated with peri-procedural adverse events as observed almost 15 years ago (16). It is notable that in the present analysis the simple metric of lesion length was the strongest predictor of adverse events, particularly in comparison with more sophisticated angiographic measurements such as plaque volume. It is unclear why the relationship between lesion length and outcomes persisted in the current study population that was exclusively treated with embolic protection devices with high rates of technical success. Embolization might have occurred with wire or device crossing, and in this respect balloon occlusion devices might have a theoretical advantage over filter-based devices, which typically have larger crossing profiles. Although this study was not adequately powered to detect treatment effects between devices, we did not observe any interaction between lesion length and outcomes among the 3 tested devices. Nonetheless, the occurrence of adverse events even in patients with the shortest lesions in this study all of whom were treated <8 mm 8-11 mm mm mm >22 mm Lesion Length Figure 1. Association Between Lesion Length and Rates of Adverse Outcomes The figure depicts the relationship between increasing lesion length and the rates of adverse peri-procedural events (major adverse cardiac events [MACE] and creatine kinase-myocardial band [CK-MB] elevation). ULN upper limit of normal. with embolic protection devices stresses the high overall risk of this patient population, even when treated with adjunctive embolic protection. Interestingly, despite a mean graft age of 12 years and the presence of several high-risk angiographic features such as a high prevalence of thrombus, we did not observe an association between adverse outcomes and graft age, thrombus, or any other angiographic graft characteristics aside from lesion length. Several risk factors including thrombus and graft age have been described for SVG PCI without distal embolic protection (3,6,7,10). It might be that the use of embolic protection devices with high technical success rates in this study has the ability to mitigate the risk associated with some of the more traditional risk factors for peri-procedural events. Further confirmatory studies are necessary to validate these observations. Study limitations. This analysis is a nonrandomized retrospective analysis, and as such it is possible that both identified and unidentified confounders might have influenced the outcomes. The cohort of patients enrolled in PRIDE was a carefully selected clinical trial population that met specific clinical and angiographic inclusion criteria. Thus the results of this trial might not be fully generalizable to unselected patient populations. This study specifically excluded patients with significantly elevated CK-MB at baseline, patients with severely depressed ejection fractions, and patients with the highest-risk acute coronary syndromes, such as ST-segment elevation myocardial infarction. One could anticipate that rates of adverse events would be even higher among these patients, but whether the association we observed between lesion length and outcomes would persist in these patient populations is unknown. In addition, because troponin levels were not systematically or prospectively collected, the definitions of CK-MB elevation and MACE do not include patients with isolated troponin elevations. Finally, whether these results apply to embolic protection devices other than those studied in PRIDE is unknown. Reprint requests and correspondence: Dr. Joseph P. Carrozza Jr., Chief, Section of Interventional Cardiology, Beth Israel Deaconess Medical Center, 1 Deaconess Road, Baker 4, Boston, Massachusetts jcarrozz@bidmc.harvard.edu. REFERENCES 1. Baim DS, Wahr D, George B, et al. Randomized trial of a distal embolic protection device during percutaneous intervention of saphenous vein aorto-coronary bypass grafts. Circulation 2002;105: Stone GW, Rogers C, Hermiller J, et al. Randomized comparison of distal protection with a filter-based catheter and a balloon occlusion and aspiration system during percutaneous intervention of diseased saphenous vein aorto-coronary bypass grafts. Circulation 2003;108: Giugliano GR, Kuntz RE, Popma JJ, Cutlip DE, Baim DS. Determinants of 30-day adverse events after saphenous vein graft intervention

6 JACC: CARDIOVASCULAR INTERVENTIONS, VOL. 1, NO. 2, with and without a distal occlusion embolic protection device. Am J Cardiol 2005;95: Sievert H, Wahr DW, Schuler G, et al. Effectiveness and safety of the Proxis system in demonstrating retrograde coronary blood flow during proximal occlusion and in capturing embolic material. Am J Cardiol 2004;94: Mauri L, Rogers C, Baim DS. Devices for distal protection during percutaneous coronary revascularization. Circulation 2006;113: Keeley EC, Velez CA, O Neill WW, Safian RD. Long-term clinical outcome and predictors of major adverse cardiac events after percutaneous interventions on saphenous vein grafts. J Am Coll Cardiol 2001;38: Sdringola S, Assali AR, Ghani M, et al. Risk assessment of slow or no-reflow phenomenon in aortocoronary vein graft percutaneous intervention. Catheter Cardiovasc Interv 2001;54: Stone GW, Rogers C, Ramee S, et al. Distal filter protection during saphenous vein graft stenting: technical and clinical correlates of efficacy. J Am Coll Cardiol 2002;40: Halkin A, Masud AZ, Rogers C, et al. Six-month outcomes after percutaneous intervention for lesions in aortocoronary saphenous vein grafts using distal protection devices: results from the FIRE trial. Am Heart J 2006;151:915.e Kalyanasundaram A, Blankenship JC, Berger P, Herrmann H, Mc- Clure R, Moliterno D. Thrombus predicts ischemic complications during percutaneous coronary intervention in saphenous vein grafts: results from TARGET (do tirofiban and reopro give similar efficacy trial?). Catheter Cardiovasc Interv 2007;69:623 9; discussion van Gaal WJ, Choudhury RP, Porto I, et al. Prediction of distal embolization during percutaneous coronary intervention in saphenous vein grafts. Am J Cardiol 2007;99: Carrozza JP Jr., Mumma M, Breall JA, Fernandez A, Heyman E, Metzger C. Randomized evaluation of the TriActiv balloon-protection flush and extraction system for the treatment of saphenous vein graft disease. J Am Coll Cardiol 2005;46: Hong MK, Mehran R, Dangas G, et al. Creatine kinase-mb enzyme elevation following successful saphenous vein graft intervention is associated with late mortality. Circulation 1999;100: Cohen DJ, Murphy SA, Baim DS, et al. Cost-effectiveness of distal embolic protection for patients undergoing percutaneous intervention of saphenous vein bypass grafts: results from the SAFER trial. J Am Coll Cardiol 2004;44: Mehta SK, Frutkin AD, Milford-Beland S, et al. Utilization of distal embolic protection in saphenous vein graft interventions (an analysis of 19,546 patients in the American College of Cardiology-National Cardiovascular Data Registry). Am J Cardiol 2007;100: Liu MW, Douglas JS, Lembo NJ, King SB. Angiographic predictors of a rise in serum creatine kinase (distal embolization) after balloon angioplasty of saphenous vein coronary artery bypass grafts. Am J Cardiol 1993;72:514 7.