Author + information
- Received June 28, 1999
- Revision received December 3, 1999
- Accepted January 20, 2000
- Published online May 1, 2000.
- Javed M Ahmed, MD, MRCPa,
- Mun K Hong, MD, FACCa,* (, )
- Roxana Mehran, MD, FACCa,
- Gary S Mintz, MD, FACCa,
- Alexandra J Lansky, MDa,
- Augusto D Pichard, MD, FACCa,
- Lowell F Satler, MD, FACCa,
- Kenneth M Kent, MD, PhD, FACCa,
- Hongsheng Wu, PhDa,
- Gregg W Stone, MD, FACCa and
- Martin B Leon, MD, FACCa
- ↵*Reprint requests and correspondence: Dr. Mun K. Hong, Division of Cardiology, Cornell University-New York Presbyterian Hospital, Starr Pavilion 4, 520 E. 70th St, New York, New York 10021
We compared in-hospital and one-year clinical outcomes in patients undergoing debulking followed by stent implantation versus stenting alone for saphenous vein graft (SVG) aortoostial lesions.
Stent implantation in SVG aortoostial lesions may improve procedural and late clinical outcomes. However, the impact of debulking before stenting in this complex lesion subset is unknown.
We studied 320 consecutive patients (340 SVG aortoostial lesions) treated with Palmaz-Schatz stents. Debulking with excimer laser or atherectomy was performed in 133 patients (139 lesions) before stenting (group I), while 187 patients (201 lesions) underwent stent implantation without debulking (group II). Procedural success and late clinical outcomes were compared between the groups.
Overall procedural success (97.6%) was similar between the groups. Procedural complications were also similar (2.2% for group I and 2.6% for group II). At one-year follow-up, target lesion revascularization (TLR) was 19.4% for group I and 18.2% for group II (p = 0.47). There was no difference in cumulative death or Q wave myocardial infarction between the groups. Overall cardiac event-free survival was similar (69% for group I and 68% for group II). By Cox regression analysis, the independent predictors of late cardiac events were final lumen cross-sectional area (CSA) by intravascular ultrasound (IVUS) (p = 0.001) and restenotic lesions (p = 0.01). Similarly, final IVUS lumen CSA (p = 0.0001) and restenotic lesions (p = 0.006) were found to predict TLR at one year.
These results suggest that, in most patients with SVG aortoostial lesions, debulking before stent implantation may not be necessary.
Treatment of saphenous vein graft (SVG) aortoostial lesions by conventional balloon angioplasty or new devices (laser or atherectomy) has been associated with suboptimal results, frequent periprocedural complications and higher rates of restenosis (1–3). These disappointing results can be attributed, in part, to the excessive residual stenosis and unyielding plaque. Stents have been found to improve short- and long-term outcomes compared with percutaneous transluminal coronary angioplasty (PTCA) or atherectomy when used for aortoostial lesions (4–6). However, even with stents the risk of restenosis remains high. It has been suggested that debulking the atherosclerotic plaque or altering the lesion compliance with atheroablative devices before stent implantation may improve procedural and late clinical outcomes in this complex lesion subset.
Thus, to determine the acute and late clinical outcomes associated with stent implantation with or without prior debulking for SVG aortoostial lesions, we evaluated procedural success, major in-hospital complications and one-year clinical outcomes in a large consecutive series of patients.
The patient cohort included a consecutive series of 320 patients (340 SVG aortoostial lesions) in the Cardiology Research Foundation Angioplasty Database treated with Palmaz-Schatz stent implantation between September 1995 and March 1997. All patients gave written informed consent approved by the Institutional Review Board of the Washington Hospital Center before the procedure. These patients represent the majority (>95%) of patients with SVG aortoostial lesions treated during the study period, as the strategy was to implant stents in all SVG aortoostial lesions, and only patients with distal embolization and reduced flow did not receive stents. Patients were divided into two groups according to the type of intervention used. All indications for stent use (planned elective use to improve acute procedural safety and reduce late clinical events, provisional use to treat suboptimal primary device results or urgent use to treat abrupt or threatened closure) are included in this study. Debulking with excimer laser (n = 90), directional atherectomy (n = 39) and transluminal extraction atherectomy (n = 10) was performed in 139 lesions before stenting (group I), and 201 lesions were treated with stents alone (group II).
Baseline demographic characteristics and in-hospital events were recorded by independent hospital chart review. Clinical follow-up data were obtained by office visits or serial telephone interviews by research nurses. All records relevant to late clinical events (death, Q wave myocardial infarction [MI] and any revascularization) and target lesion revascularization were obtained and evaluated by an independent adjudication committee. In addition to target lesion revascularization, repeat revascularization is also reported per patient (as any repeat revascularization) and includes all target lesion and target vessel revascularizations.
Procedural success was defined as a final diameter stenosis <50% in the absence of major in-hospital complications (death, Q wave MI and coronary artery bypass surgery [CABG]). A diagnosis of Q wave MI was made when there was documentation of pathological Q waves (>0.04 s) on an electrocardiogram (ECG) in conjunction with elevation of creatinine phosphokinase greater than twice the upper limit of normal. Non-Q wave MI was defined as creatinine kinase-MB (CK-MB) enzyme elevation ≥5 times the upper limit of normal in the absence of new pathological Q waves. Degenerated SVG were those with ectasia or lumen irregularity comprising ≥50% of the SVG shaft length. Ectasia was a lumen >20% larger than the user-defined reference segment. For the purpose of this study, ostial lesions were defined as within 3 mm of the proximal anastomosis.
Stent types and deployment technique
The types of stents used were either coronary (n = 114, 28.7%) or “biliary” Palmaz-Schatz (n = 282, 71.3%) (Johnson and Johnson Interventional Systems, Warren, New Jersey). Stent implantation procedures were performed using techniques described previously (7,8). Coronary stents were used for vessels ≤4 mm and a larger biliary version was selected for vessels >4 mm in diameter. The lesion was crossed with a 0.014 in floppy or extrasupport guide wire, and either prior balloon angioplasty or atheroablation was performed before stent implantation. Adjunct high-pressure PTCA was performed after initial stent deployment (14 to 16 atms) in all cases. Intravascular ultrasound (IVUS) imaging was performed in 270 (79.4%) lesions pretreatment, poststent implantation and postadjunct PTCA. Optimal stent implantation was carefully monitored using an interactive technique with prespecified IVUS end points and additional high-pressure inflations if needed. The IVUS was used to optimize stent apposition, expansion (minimal stent area ≥80% of the distal reference lumen area) and lesion coverage. In addition, IVUS enabled the detection of outflow obstruction and residual dissection at stent margins.
All patients were pretreated with aspirin and continued their standard antianginal therapy. A bolus of 10,000 U of intravenous heparin was given after insertion of femoral arterial sheath. After the initial administration, a repeat bolus of 5,000 U was given to maintain the activated clotting time (ACT) >300 s, if necessary. Heparin was stopped immediately after the procedure, and the sheath was removed when the ACT was <150 s. All patients received ticlopidine at a dose of 250 mg twice daily for four weeks. Abciximab (Centocor, Malvern, Pennsylvania) was used in <5% of patients. Emboli containment devices were not used in any patients.
Device selection was based on vessel size, preprocedural lesion morphology (lesion eccentricity, ulceration, degeneration, irregularity, calcification and thrombus) and the absence of clinical contraindications at the discretion of the operator. Directional atherectomy (Devices for Vascular Intervention, Redwood City, California) was chosen for eccentric lesions located in the nondegenerated SVG ≥3 mm in diameter (n = 39, 28%). The procedure was performed according to the previously described technique (9). Most commonly a 7F atherectomy device was used. In the event of residual stenosis >30%, localized dissection or intimal flap, a larger device was used to treat the residual stenosis.
The Transluminal Extraction Atherectomy (TEC; InterVentional Technologies, San Diego, California) was used for ostial lesions (n = 10, 7.1%) in degenerated SVG or thrombus containing lesions. It was performed by the methods described previously (10).
The excimer laser angioplasty (Spectranetics/Advanced Interventional System, Colorado Springs, Colorado) was selected for smaller or degenerated SVG (n = 90, 64.7%). The ELCA procedure was performed as described elsewhere (11). In brief, the range of the laser fiber catheter used was 1.7 mm to 2.0 mm. Energy densities ranged from 25 to 65 mJ/mm2 (mean 57.9 ± 5.4 mJ/mm2), and the number of pulses ranged from 30 to 1,880 (mean, 339 ± 352). A single laser pass technique was used in most of the cases.
Adjunct balloon angioplasty was performed in all atheroablation cases before stent implantation to ensure adequate stent expansion.
All cineangiograms were analyzed with the use of a computer-assisted, automated edge-detection algorithm (ARTREK, Quantitative Cardiac System, Ann Arbor, Michigan) by a core angiographic laboratory that was blinded to the clinical outcome. With the outer diameter of the contrast filled catheter as the calibration standard, reference and minimal lumen diameters (MLDs) were determined before and after stent implantation, and post final PTCA from multiple projections and the results from the “worst” view were recorded. Based upon these measurements, percent diameter stenoses were determined. Standard morphologic criteria were used for the identification of lesion length (“shoulder-to-shoulder”), eccentricity, irregularity, fluoroscopic calcification and ulceration.
IVUS imaging protocol
Intravascular ultrasound studies were performed with the Boston Scientific Corporation/Cardiovascular Imaging System. This system incorporated a single-element 30-MHz transducer and an angled mirror mounted on the tip of a flexible shaft that was rotated at 1,800 rpm within a 3.2F short monorail polyethylene imaging sheath to form planar cross-sectional images in real time. All IVUS studies were performed after administration of 0.2 mg of intragraft nitroglycerin. The ultrasound catheter was advanced ≈10 mm beyond the target lesion, and a slow imaging run (using automated transducer pullback at 0.5 mm/s) was performed from beyond the target lesion to the aortoostial junction. A number of cross-sectional measurements were made, and validation of these measurements by IVUS has been reported previously (12,13). By using computerized planimetry (TapeMeasure, Indec Systems, Mountain View, California), lesion site and reference segment external elastic membrane (EEM) cross-sectional area (CSA) and lumen CSA were measured. Because media thickness cannot be measured accurately, plaque plus media (P + M) CSA (EEM CSA minus lumen CSA) was used as a measure of atherosclerotic plaque. Plaque burden (cross-sectional narrowing) was calculated as the P + M CSA divided by the EEM CSA. When the atherosclerotic plaque encompassed the catheter, the lumen was assumed to be the size of the imaging catheter. The reference segment selected was the most visually normal cross section within 10 mm distal to the lesion.
Statistical analysis was performed using StatView 4.5 or SAS (14) (both SAS Institute Inc, Cary, North Carolina). Categorical data are presented as percent frequency and compared between groups by chi square statistics. Continuous variables are presented as mean ± one standard deviation, and comparison between the groups was performed using unpaired t test. Cox (15) proportional hazard regression analysis was used to determine independent correlates of target lesion revascularization (TLR) and late cardiac events. Survival curves were constructed by Kaplan Meier method and displayed using the SAS LIFETEST procedure. The Wilcoxon log rank test was used for survival comparison between groups. A p value ≤0.05 was considered statistically significant.
The baseline characteristics of all treated patients are listed in Table 1. The groups were well matched except for a higher incidence of previous MI in the stent alone group and a trend for more diabetic patients in the debulking and stent group. The treated grafts were relatively old with an average graft age of 96.3 ± 58.6 months.
Angiographic analysis and procedural characteristics
Lesion characteristics and quantitative angiographic measurements are presented in Tables 2 and 3. ⇓⇓Eccentric lesions were more frequent in group I as compared with group II (63.6% vs. 46.8%, p = 0.02). The prevalence of degenerated SVG, ulceration, thrombus containing and restenotic lesions was similar for the two groups, most likely related to the small sample size. Likewise, there was no difference in the average pretreatment and posttreatment angiographic reference diameter. Interestingly, despite a smaller MLD and greater baseline diameter stenosis in group I, the final MLD was greater and diameter stenosis was lower.
A total of 396 stents were deployed in 320 patients with 340 lesions. The mean number of stents per lesion was 1.17 ± 0.45. In 289 (85.0%) lesions, a single stent was used, and 51 (15.0%) lesions were treated with two stents. Average balloon size was 4.36 ± 0.79, and mean final inflation pressure was 15.1 ± 4.2. The average balloon-to-artery ratio used for final stent expansion was 1.30 ± 0.25.
Quantitative IVUS findings are presented in Table 4. Preintervention, lesions treated with plaque debulking and stenting had smaller MLD and lumen CSA. Conversely, these lesions had larger P + M CSA and greater plaque burden (Table 4). ⇓Postintervention, these lesions had lesser plaque burden as compared with lesions treated with stents alone.
Overall procedural success was high and similar for the two groups (Table 5). Similarly combined major in-hospital complications (death, Q wave MI and emergent CABG) did not differ significantly between the groups (2.2% vs. 2.6%, p = 0.24). Likewise, the incidence of subacute stent thrombosis, vascular complications and the need for repeat in-hospital target lesion angioplasty was similar for both groups. The presence of any CK-MB enzyme elevation postintervention was found in 34.1% vs. 31.0% (p = 0.55), and non-Q wave MI (CK-MB ≥5 times normal occurred in 18.2% vs. 18.5% (p = 0.94) of the patients. Intravascular ultrasound guidance was used more frequently in patients who underwent atheroablation before stenting as compared with patients who were treated with stents alone (87.8% vs. 73.6%, p = 0.002). The periprocedural use of abciximab was similar for both groups (2.9% for debulking and stenting vs. 4.7% for stents alone, p = 0.41). A representative procedural result before and after laser treatment and stent implantation to treat an ostial SVG lesion is shown in Figure 1.
Late clinical outcomes
Clinical follow-up up to one year was available in 130 (97.7%) patients in group I and 181 (96.7%) patients in group II (Table 5). There was no difference in late mortality or Q wave MI between the two groups. Overall TLR at one year was 19.4% for patients treated with debulking and stenting and 18.2% for patients treated with stenting alone (p = 0.47). The rate of any repeat revascularization was also similar for both groups (23.7% vs. 19.9, p = 0.39). Actuarial event free survival curves for any cardiac events (death, Q wave MI, angioplasty or CABG) and for TLR at one-year follow-up are shown in Figure 2. Event-free survival was similar in both groups for both end points (p = 0.48 for any event and p = 0.12 for TLR).
Cox regression analysis was performed to determine the independent correlates of any adverse cardiac events and TLR at one-year follow-up. The following variables were entered into the multivariable model: age, gender, graft age, diabetes, hypertension, previous history of MI, types of stents, prestent atheroablation, MLD before and after intervention, preintervention and final reference lumen diameter, plaque burden and final lumen CSA by IVUS. The independent predictors of late cardiac events were: final IVUS lumen CSA (relative risk: 0.61 CI: 0.47–0.78, p = 0.001) and restenotic lesions (relative risk: 1.67, CI: 1.11–2.49, p = 0.01). Similarly, final IVUS lumen CSA (relative risk: 0.61, CI: 0.62–0.89, p = 0.0001) and restenotic lesions (relative risk: 1.94, CI: 1.20–3.14, p = 0.006) were found to predict TLR at one year.
This study shows that patients with SVG aortoostial lesions treated with stent implantation with or without prior atheroablation have: 1) similar in-hospital procedural results and major complications, and 2) similar one-year cardiac events, including TLR and any repeat revascularization. In this large patient cohort, we also identified independent predictors of TLR and late cardiac events. Lumen CSA by IVUS and restenotic lesions were found to predict both TLR and adverse cardiac events at one year. Interestingly, in none of these analyses did prestent atheroablation predict TLR or adverse cardiac end points.
Treatment of SVG aortoostial lesions
Overall procedural success rate of balloon angioplasty in SVG varies between 84% and 92% (1,2). Procedural success is lower in aortoostial location compared with nonostial locations (1,2). Conventional balloon angioplasty is frequently unsuccessful in this location due to greater elastic recoil. Several investigators have demonstrated that stent use in SVG aortoostial lesions is associated with high procedural success and low complication rates (16–18). Furthermore, in a retrospective comparison of stenting and balloon angioplasty for SVG aortoostial lesions, Brener et al. (4) have shown that patients in the stent group had a lower incidence of composite end points of death, MI and repeat revascularization (23% vs. 45%, p < 0.001) at one-year follow-up (12). Likewise, Rocha-Singh et al. (6) reported a 93% procedural success rate and 27% restenosis rate with Palmaz-Schatz stent implantation in ostial SVG lesions. In another study involving stent implantation in SVG aortoostial in patients with unstable angina, Rechavia et al. (5) have shown high immediate success and low 30-day and long-term event rates.
However, no study has addressed the issue of whether atheroablation before stenting in this lesion subset could improve the immediate and late clinical outcomes compared with stenting alone. In this study comparing two strategies for the treatment of SVG aortoostial lesions, we found no difference in the immediate and late outcomes after stenting with or without prestent atheroablation. Device selection was based on vessel size, anatomy, preprocedural lesion morphology (lesion eccentricity, ulceration, irregularity, calcification and thrombus) and degeneration of the graft. De novo lesions involving the ostium of the larger (>3 mm) nondegenerated graft were treated with directional atherectomy followed by stent implantation or balloon angioplasty followed by stent insertion. Lesions located in the ostium of the smaller vein grafts were treated with excimer laser angioplasty and stents or stents alone. Due to the relatively small sample size in each group, there was no statistical difference in the baseline lesion characteristic.
Overall major in-hospital complications (death, Q wave MI and CABG) were infrequent in both treatment groups despite lesion complexity. However, there was a relatively high frequency of periprocedural non-Q wave MI after both types of treatment (18.2% in group I and 18.5% in group II, p = 0.60). This high prevalence of non-Q wave MI perhaps represents a very complex subset of lesions in diffusely diseased SVG grafts. Several reports have indicated an association between periprocedural CK-MB elevations and late adverse cardiac events (19,20), and this could also account for high late mortality in our study. Only a minority (<5%) of patients was treated with glycoprotein IIb/IIIa inhibitor, and distal protection devices were not used in this study.
Stent synergy (prestent atheroablation) approach
The rationale of this strategy was to test the hypothesis of whether atheroablation before stenting could facilitate subsequent optimal stent expansion and whether this optimal stent expansion translates into improved long-term outcome. Although with this approach we were able to achieve a larger final MLD (3.43 ± 0.76 mm vs. 3.21 ± 0.70 mm; p = 0.05) and less residual plaque burden (57.6 ± 10.69% vs. 64.9 ± 11.05%; p = 0.01), no impact on the procedural and late clinical outcome was demonstrated. The most likely explanation may be that the atheroablation is not sufficient to have a long-term impact. Compared with the reference diameter of 3.5 mm, the atheroablative devices create relatively small lumen (1.7 mm to <3 mm). This is illustrated by the relatively large residual plaque burden (approximately 60%).
This study demonstrates that stent implantation in SVG aortoostial lesions is associated with high procedural success rate, lower complications and acceptable one-year follow-up clinical events. As there is no difference in procedural outcomes, late clinical events and target lesion revascularization between patients undergoing debulking before stenting or stenting alone, debulking before stent implantation may not be necessary.
There are several limitations of our study. First, only lesions treated with Palmaz-Schatz stents were included in the analysis, and it is unknown whether different stent designs may have influenced the clinical outcome. Second, this study was a retrospective analysis of the clinical, angiographic and IVUS data from a large group of consecutively studied patients and, therefore, the results and conclusions are subject to limitations inherent in all such reports. Third, IVUS was performed in a majority (79.4%) of the cases, and the results of this study may not apply to those cases where IVUS is not used. Fourth, the clinical follow-up in the matched cohort was truncated at one year, and long term follow-up may have produced different results because vein graft failure continues to develop with time. Finally, the use of different devices and IVUS guidance was based on operator decision and not on random assignment; therefore, the strength of the conclusions made with respect to the effect of these variables on late outcome and TLR is limited. Despite these limitations, this study provides clinically relevant information about the impact of different revascularization strategies in patients with SVG aortoostial lesions.
☆ This study was supported, in part, by Cardiovascular Research Foundation, Washington, DC.
- activated clotting time
- coronary artery bypass surgery
- creatinine kinase-MB
- cross-sectional area
- external elastic membrane
- intravascular ultrasound
- myocardial infarction
- minimal lumen diameter
- P + M
- plaque plus media
- percutaneous transluminal coronary angioplasty
- saphenous vein graft
- target lesion revascularization
- Received June 28, 1999.
- Revision received December 3, 1999.
- Accepted January 20, 2000.
- American College of Cardiology
- de Feyter P.J.,
- van Sylen R.J.,
- de Jaegere P.P.T.,
- Topol E.J.,
- Serruys P.W.
- Webb J.G.,
- Myler R.K.,
- Shaw R.E.,
- et al.
- Abdel-Meguid A.E.,
- Whitlow P.L.,
- Simpfendorfer C.,
- et al.
- Rechavia E.,
- Litvack F.,
- Macko G.,
- Eigler N.L.
- Safian R.D.,
- Gelbfish J.S.,
- Erny R.E.,
- Schnittet S.J.,
- Schmidt D.A.,
- Baim D.S.
- Bittle J.A.,
- Sanborn T.A.
- Tobis J.M.,
- Mallery J.,
- Mahon D.,
- et al.
- Potkins B.N.,
- Bartorelli A.L.,
- Gessert J.M.,
- et al.
- Ray A.A.
- Cox D.R.
- Wong S.C.,
- Baim D.S.,
- Schatz R.A.,
- et al.
- Abdelmeguid A.E.,
- Topol E.J.,
- Whitlow P.L.,
- Sapp S.K.,
- Ellis S.G.
- Abdelmeguid A.E.,
- Topol E.J.