Catheterization and Cardiovascular Interventions 83:1035–1042 (2014)
Long-Term Outcome Following Percutaneous Coronary Intervention With Drug-Eluting Stents Compared With Bare-Metal Stents in Saphenous Vein Graft Lesions: From Western Denmark Heart Registry Mikkel Hougaard,1* MD, Per Thayssen,1 MD, DMSci, Anne Kaltoft,2 MD, PhD, Hans-Henrik Tilsted,3 MD, Michael Maeng,2 MD, PhD, Jens Flensted Lassen,2 MD, PhD, Leif Thuesen,2 MD, DMSci, and Lisette Okkels Jensen,1 MD, DMSci, PhD Objectives: We used the Western Denmark Heart Registry to assess one-year and long-term all-cause mortality and stent failure following Percutaneous Coronary Intervention (PCI) with drug-eluting stents (DES) or bare-metal stents (BMS). Background: The use of DES compared with BMS during PCI has reduced the risk of restenosis in native coronary artery lesions. In saphenous vein grafts (SVG) the outcome after DES compared with BMS is insufficiently described. Methods: From January 1, 2002 to December 31, 2010 all patients with PCI of SVG lesions were identified among 3.0 million inhabitants. Stent failure was defined as clinically driven target lesion revascularization, graft occlusion without intervention, or stent thrombosis. Results: The study cohort consisted of 529 patients with 755 SVG lesions (348 DES patients with 510 lesions and 181 BMS patients with 245 lesions). Mean age did not differ between patients with DES-treated lesions compared to patients with BMS-treated lesions (67.5 6 9.1 years vs. 67.6 6 9.3 years; P 5 0.85). The median follow-up time was 3.0 years (25th–75th percentile: 1.4–5.1 years). One-year (n 5 27 (8.2%) vs. n 5 12 (6.7%), log rank P 5 0.60) and 3-year cumulative mortality (n 5 31 (18.8%) vs. n 5 59 (21.8%), log rank P 5 0.64) did not differ significantly between DES- and BMS-treated patients. One-year cumulative stent failure was seen in 39 (6.6%) DES-treated lesions vs. 24 (10.8%) BMStreated lesions (P 5 0.088), and 3-year cumulative stent failure in 48 (15.4%) vs. 34 (18.8%) lesions (P 5 0.25), respectively. Conclusion: In SVG lesions, DES showed no long-term benefit compared to BMS in rates of all-cause mortality or stent failure. VC 2013 Wiley Periodicals, Inc. Key words: ANGIO, coronary angiography; CAD, coronary artery disease; CATH, diagnostic cardiac catheterization; CABG, cronary artery bypass grafting; GRFT, bypass grafts coronary; PCI; percutaneous coronary intervention
INTRODUCTION
Drug eluting stents (DES) are superior to bare metal stents (BMS) in preventing restenosis after treatment of
1
Department of Cardiology, Odense University Hospital, Odense, Denmark 2 Department of Cardiology, Aarhus University Hospital, Skejby, Aarhus, Denmark 3 Department of Cardiology, Alborg University Hospital, Aalborg, Denmark Conflict of interest: Nothing to report. C 2013 Wiley Periodicals, Inc. V
native coronary vessels with percutaneous coronary intervention (PCI) [1–3]. The use of BMS compared to angioplasty alone in treatment of stenosis in saphenous
*Correspondence to: Mikkel Hougaard, Department of Cardiology, Odense University Hospital, Sdr. Boulevard 29, 5000 Odense C, Denmark. E-mail:
[email protected] Received 2 May 2013; Revision accepted 23 October 2013 DOI: 10.1002/ccd.25279 Published online 31 October 2013 in Wiley Online Library (wileyonlinelibrary.com)
1036
Hougaard et al.
vein grafts (SVG) has been shown to reduce the risk of major cardiac events (MACE) and the need of reintervention [4]. The benefits of DES in relation to PCI of SVG lesions has, however, not been clearly established. The safety and efficacy of DES versus BMS in SVG remains uncertain due to contradictory reports [5,6]. Atherosclerotic disease in SVG has several peculiarities that make it difficult to extrapolate outcomes of the use of DES as compared to BMS, from outcomes observed in native coronary arteries [7–10]. We used the Western Denmark Heart Registry (WDHR) to compare mortality and stent failure in patients with SVG lesions treated with PCI and DES or BMS implantation.
PATIENTS AND METHODS Setting and Design The study was conducted using WDHR, which covers the region’s entire population of 3.0 million inhabitants (55% of the Danish population). A detailed description of the databases has been reported previously [11]. Patients and Procedures
In WDHR we identified all patients with PCI of SVG lesions performed at the three participating centers in the period of January 1, 2002 to December 31, 2010. PCI was performed according to the standard practices of the participating centers and stent type, use of glycoprotein IIb/IIIa receptor inhibitor or distal protection devices were left at the operator’s discretion. Post PCI medical treatment consisted of lifelong acetylsalicylic acid (75–150 mg daily) and clopidogrel with a loading dose of 300 mg followed by 75 mg daily. Since November 2002, the recommended duration of clopidogrel treatment after stent implantation has been 12 months regardless of stent type. Endpoints
The endpoints were time to all-cause mortality or stent failure.
Stent thrombosis was characterized using the Academic Research Consortium (ARC) definition [13]. Definite stent thrombosis had to be confirmed angiographically and accompanied by one or more of the following signs within 48 hours: New onset of ischemic symptoms at rest, new electrocardiographic changes suggestive of acute ischemia, or typical rise and fall in cardiac biomarkers. In-stent restenosis was defined as a stenosis of more than 50% (visual assessed) within the stent or within a 5 mm bordering segment proximal or distal to the stent in patients with symptoms. Angiographic follow-up was not scheduled, and patients were re-admitted due to symptoms and TLR was therefore clinically driven due to in-stent restenosis or stent thrombosis. Graft occlusion without TLR was identified by angiograms with occluded SVG where revascularization was not attempted or succeeded. Statistical Methods
We counted endpoint events that occurred during the follow-up period and compared rates for the two groups. Follow-up began on the date of the PCI procedure and continued until date of target lesion revascularization, observation of an occluded vein graft (without further treatment), March 28, 2011 or after three years of follow-up (to ensure at least 10% of the study population at risk), whichever came first. The statistical analysis was carried out using SPSS 20.0. Categorical data were presented as counts and percentages and compared by the Pearson Chi-Square test or the Fisher exact test. For continuous variables with a normal distribution, the mean 6 1 SD was reported. For variables not normally distributed, the median and interquartile ranges were reported. We constructed Kaplan-Meier curves for DES- and BMStreated patients. Cox proportional hazards regression analysis was used to estimate the hazard ratio (HR) for mortality or stent failure. We controlled for sex, age, GPIIb/IIIa receptor inhibitor use and diabetes mellitus in mortality regression analyses and further we controlled for reference vessel size, final TIMI flow and lesion complexity in lesion related regression analysis.
Endpoint Definitions
Data on mortality were obtained from the Danish National Registry of Patients, which has kept electronic records on the gender, date of birth, changes in address, date of emigration, and changes in vital status of the entire Danish population since 1968 [12]. Stent failure was defined as a clinical driven TLR, graft occlusion without intervention or stent thrombosis.
RESULTS Descriptive Data A total of 643 consecutive patients were treated with PCI for a SVG lesion at one of the three university hospitals (Odense University Hospital, Aarhus University Hospital and Aalborg University Hospital) between January 1, 2002 and December 31, 2010. Mortality
Catheterization and Cardiovascular Interventions DOI 10.1002/ccd. Published on behalf of The Society for Cardiovascular Angiography and Interventions (SCAI).
Long-Term Outcome Following Percutaneous Coronary Intervention
1037
TABLE I. Patient Characteristics
Number of patients – no. Male gender – no. (%) Age – years (SD) Graft age – years (SD) Family history – no. (%) Smoking – no. (%) Body Mass Index – kg/m2 (SD) Diabetes mellitus – no. (%) Hypertension – no. (%) Previous Percutaneous coronary intervention – no. (%) Previous myocardial infarction – no. (%) Lipid lowering therapy – no. (%) Glycoprotein IIb/IIIa receptor blocker – no (%) Indication % STEMI NSTEMI/UAP SAP
BMS
Valid cases
DES
Valid cases
181 151 (83.4) 67.6 (9.3) 11.4 (5.6) 87 (52.1) 47 (29.2) 27.1 (4.2) 32 (19.6) 99 (55.9) 59 (34.1) 111 (63.4) 143 (81.3) 57 (37.3)
181 181 181 170 167 161 130 163 177 173 175 176 153 177 24 57 96
348 285 (81.9) 67.5 (9.1) 11.0 (6.6) 181 (55.8) 78 (27.6) 27.9 (4.2) 77 (25.4) 224 (67.3) 110 (33.1) 185 (56.4) 296 (88.9) 88 (25.4)
348 348 348 322 324 283 298 303 333 333 328 333 347 341 34 122 185
13.6 32.2 54.2
data was not available for 4 patients, who were foreign citizens. Patients treated with balloon angioplasty without stent implantation (n ¼ 85) or with unknown stent type (n ¼ 25), were excluded. Thus, the final study population consisted of 529 patients who had 755 SVG lesions treated (348 DES patients with 412 lesions and 181 BMS patients with 212 lesions). The mean age was 67.5 6 9.2 years and age did not differ between patients with DES-treated lesions or patients with BMS-treated lesions (67.5 6 9.1 years vs. 67.6 6 9.3 years; P ¼ 0.85). Most of the baseline characteristics were well balanced between the two groups, although the prevalence of lipid lowering therapy and hypertension was higher among the DES-treated, whereas the use of a glycoprotein IIb/IIIa receptor blocker was more frequent among the BMS-treated (Table I). There was no significant difference between the DES- and the BMS-treated patients when comparing gender, family history, smoking, body mass index, diabetes, previous PCI, or myocardial infarction. The mean graft age among the DES- and BMS treated was similar in the two groups [11.0 6 6.6 years and 11.4 6 5.6 years respectively (P ¼ 0.53)]. Procedural Data
Procedural data are listed in Table II. The DES- and BMS-treated groups differed in the distribution of the grafted vessels. More patients in the BMS group had a SVG supplying the right coronary artery and more patients in the DES group had SVG supplying the circumflex system. The prevalence of SVG supplying left anterior descending artery was low due to the high use of the left mammarian artery for this vessel. Both preintervention and final TIMI flow grades were higher in
10.0 35.8 54.3
P Value 0.66 0.85 0.53 0.43 0.71 0.10 0.16 0.01 0.83 0.13 0.02 0.01 0.41
the DES group, and DES were more commonly used in complex (type B2/C) lesions. Reference segment and minimum luminal diameter were larger among the BMS-treated patients, and lesion and stent lengths were similar in the two groups. A higher contrast volume was used in the BMS-group. Mortality and Clinical Events The median follow-up time was 3.0 years (25th– 75th percentile: 1.4–5.1 years). The overall 1-year mortality was 7.4% (n ¼ 39) and the 3-year mortality was 17.0% (n ¼ 90). Among patients treated with DES and BMS, 1-year cumulative mortality was 8.2% (n ¼ 27) and 6.7% (n ¼ 12), respectively (log rank P ¼ 0.60) and 3-year cumulative mortality was n ¼ 31 (18.8%) and n ¼ 59 (21.8%), respectively (log rank P ¼ 0.64) (Fig. 1). After adjustment for sex, age, GPIIb/IIIa receptor inhibitor use and diabetes we found no significant difference in mortality with DES compared with BMS after 1 year (HR ¼ 1.85, 95% CI 0.804.25) or after 3 years (HR ¼ 1.25, 95% CI 0.762.07). One-year cumulative stent failure was seen in 39 (6.6%) DES-treated lesions vs. 24 (10.8%) BMS-treated lesions (P ¼ 0.088), and three-year cumulative stent failure in 48 (15.4%) vs. 34 (18.8%) lesions (P ¼ 0.25), respectively (Fig. 2). After adjustment for sex, age, GPIIb/IIIa receptor inhibitor use, diabetes, reference vessel size, final TIMI flow and lesion complexity, our data showed no significant stent failure benefit associated with DES after 1 year (HR ¼ 0.73, 95% CI 0.381.97) or 3 years (HR ¼ 0.75, 95% CI 0.411.35). Cumulative rates of stent thrombosis, TLR and graft occlusion without revascularization after 1 and 3 years were similar between DES and BMS treated lesions (Table III).
Catheterization and Cardiovascular Interventions DOI 10.1002/ccd. Published on behalf of The Society for Cardiovascular Angiography and Interventions (SCAI).
1038 TABLE II.
Hougaard et al. Procedure Characteristics
Number of lesions – no. Vessel grafted – no. (%) Left anterior descending artery RD-1 RD-2 Left circumflex artery RM-1 RM-2 Right coronary artery Lesion length – mm (SD) Reference segment – mm Minimum lumen diameter – mm Relative preinterventional stenosis (%) Relative postinterventional stenosis (%) Stent length – mm (SD) Stent number – no. (%) 1 2 3þ Number of vessels treated (%) 1 2 3 Number of lesion’s treated (%) 1 2 3þ Max balloon pressure – atm Max balloon diameter – mm Final TIMI flow – no. (%) Grade 0 Grade 1 Grade 2 Grade 3 Preintervention TIMI flow – no. (%) Grade 0 Grade 1 Grade 2 Grade 3 Lesion type (%) A B1 B2 C Procedure time – minutes (SD) Flouro time – minutes (SD) Contrast – ml (SD)
BMS
Valid cases
DES
Valid cases
P Value
212
212 209
412
412 412
0.009
30 (14.4) 16 (7.7) 0 (0.0) 14 (6.7) 45 (21.5) 8 (3.8) 96 (45.9) 16.3 (13.5) 4.0 (0.7) 0.5 (0.6) 88.2 (11.4) 0.7 (5.3) 19.9 (10.6)
240 240 233 245 245 244
46 (11.2) 49 (11.9) 2 (0.5) 50 (12.1) 90 (21.8) 33 (8.0) 142 (34.5) 16.5 (13.3) 3.5 (0.6) 0.4 (0.4) 89.0 (10.9) 0.8 (6.6) 20.7 (14.0)
506 510 509 510 510 506
84.1 13.5 2.4
206 33 6
83.3 11.2 5.5
425 57 28
82.4 16.4 1.2
201 40 3
76.5 20.0 3.5
390 102 18
67.6 25.8 6.6 16.9 (4.0) 4.1 (0.7)
165 63 16 245 245
64.7 25.1 10.2 18.3 (4.4) 3.7 (0.6)
330 128 52 510 510
0.0 0.8 7.3 91.8
0 2 18 255
0.2 1.0 1.6 97.3
1 5 8 496
14.3 5.3 22.4 58.0
35 13 55 142
12.2 6.9 14.3 66.7
62 35 73 340
14.8 21.4 6.6 57.1 31.7 (19.7) 12.7 (9.4) 152.3 (99.8)
27 39 12 104 178 181 178
13.6 8.8 12.7 64.9 30.6 (19.4) 12.1 (8.7) 130.3 (83.7)
66 43 62 316 340 348 344
0.84