Stents Covered by Autologous Venous Grafts: Feasibility and Immediate and Long-Term Results

Christodoulos Stefanadis, MD, FACC, FESC, Konstantinos Toutouzas, MD, Eleftherios Tsiamis, MD, Charalambos Vlachopoulos, MD, Ioannis Kallikazaros, MD, Costas Stratos, MD, Manolis Vavuranakis, MD, FACC, Pavlos Toutouzas, MD, FACC, FESC, Department of Cardiology, Hippokration Hospital, University of Athens, Athens, Greece.

Am Heart J. 2000;139(3) 

In This Article


Fifty-eight AVGCS were implanted in 56 patients. The first 12 patients were selected on the basis of having lesions that were at low risk of causing major ischemic complications during AVGCS implantation and thus patients with totally occluded lesions or non-left anterior descending lesions were excluded. This was a prespecified number of patients to detect possible complications. Subsequent to this selection, the principal exclusion criteria were patients with small vessels (<2.5 mm), angiographically diffuse disease, tortuosity of the vessel, or lesions involving the origin of side branches. AVGCS were implanted in bail-out cases and in patients with ACS (Table I). The clinical application of the AVGCS was approved by the ethics committee of our institution. Written informed consent was obtained from every patient entering the catheterization laboratory for possible balloon angioplasty.

To compare the results of AVGCS and the control group, 114 patients were selected in which 138 conventional uncovered stents were implanted (Table I). These patients were matched with the AVGCS group for sex, age, diabetes mellitus, stenosis location, reference diameter, minimal luminal diameter (MLD) before the procedure, type of stents, and the ability to be seen for a follow-up angiographic and clinical evaluation. The patients of the control group were selected from a consecutive series of 529 patients who were treated between January 1995 and March 1998. The matching process was performed blindly, without knowledge of the clinical and angiographic outcome of the patients.

Venous graft harvest. Under local anesthesia, the right cephalic vein was isolated from the deltoideopectoral sulcus or from the anterobrachial region. Initially, the venous graft was harvested from the deltoideopectoral sulcus. However, because of the development of a small hematoma at the site of the incision in 2 patients, the vein was harvested from the anterobrachial region (41 patients, 43 lesions [74.13%]). The thickness of the venous wall was greater in veins from the anterobrachial region. A vein of 3 to 5 cm in length was adequate for the complete coverage of the stent. Thereafter, a 5F introducing sheath was inserted through the lumen of the vein and the venous graft was cleared from the boundary tissues. This procedure was performed with caution to avoid any harmful manipulation of the venous graft. In addition, venous segments with side branches were resected.

Stent coverage procedure. Two types of AVGCS were prepared. In type A AVGCS, the venous graft was trimmed to approximately twice the length of the stent and therefore it was inserted through the lumen of the stent. Subsequently, the 2 ends of the venous graft were folded back along the external surface of the stent and were joined by the application of 4 sutures (Prolene 7-0, Ethicon; Johnson and Johnson Co). Thus the stent was completely covered, both internally and externally, by the venous graft (Figure 1). In this type, a Palmaz or Palmaz-Schatz stent (Johnson & Johnson Interventional Systems, Warren, NJ) was used. Ten patients received type A AVGCS.

Type A AVGCS coverage. Venous graft was trimmed to approximately twice the length of stent; it was inserted through lumen of stent (A), and its 2 ends were folded back along external surface of stent (B). Ends were then joined with 4 stitches so that both internal and external surfaces of stent were covered by graft (C).

To address the problem of the increased profile of the endoprosthesis, the stent was covered only externally by the venous graft. In type B AVGCS, the endothelial surface of the venous graft faced toward the struts of the stent. Consequently, the venous graft was secured on the stent by application of 3 sutures (Prolene 7-0, Ethicon) at each site of the stent (Figure 2). However, the venous graft was not overstretched on the stent to avoid possible rupture of the venous graft. In type B AVGCS, the length of the trimmed venous graft was approximately 1 mm shorter than the length of the stent. Three stents were used during the course of the study: the Palmaz, the Palmaz-Schatz, and the Multilink (Advanced Cardiovascular Systems, Hampshire, UK). During the application of sutures for covering the premounted Multilink stent, care was taken to avoid rupture of the balloon by the needle of the suture. Each covered stent was prepared by 2 of 6 interventional cardiologists involved in the procedures.

Type B AVGCS coverage. Covering of stent by autologous vein graft is demonstrated in A. Venous graft was placed on external surface of stent in such a way that endothelium of tissue faced toward stent (B). Subsequently, venous graft was secured by applying 3 sutures at each end of stent (C).

In patients with ACS, the preparation progressed in parallel with coronary catheterization. In bailout cases, the venous graft was harvested after the procedure of balloon angioplasty. In the first 15 cases, a commercially available stent-delivery system (Johnson & Johnson, Interventional Systems) was used. The introducing sheath of the stent-delivery system was in contact with the AVGCS, and the distal part of the introducer sheath acted as a supportive device and prevented the AVGCS from sliding backward. After this initial experience, the AVGCS was mounted on the delivery balloon of a conventional Palmaz or Palmaz-Schatz stent and crimped down manually, except for the Multilink stent, which was premounted. For type A AVGCS, a 10F guiding catheter (Guidant, Coronary Atherectomy Guiding Catheter, DVI, Temecula, Calif) was used; for type B AVGCS, an 8F guiding catheter (Cordis Europe, NV, Roden, The Netherlands) was used. The target balloon-artery ratio was approximately 1.2. The MLD was calculated before and immediately after AVGCS placement by quantitative coronary angiography with the guiding catheter tip diameter as a reference (DCI-S, Automated Coronary Analysis, Philips, The Netherlands). The conventional regimen of anticoagulant therapy including aspirin, dextran, heparin, and acenocoumarol was used in the first 15 patients. However, after the initial phase of this study, all patients received ticlopidine and aspirin.

In 15 patients, intracoronary ultrasound images (Endosonics Corp, Pleasanton, Calif) were obtained to ensure optimal stent expansion. Optimal stent expansion was defined as full lesion coverage without inflow or outflow obstruction; adequate stent apposition and achievement of an intrastent cross-sectional area 80% of the average proximal and distal reference cross-sectional area were also required. Furthermore, the possibility of venous graft flapping was excluded.

The Palmaz, Palmaz-Schatz, and Multilink stents were used. The target balloon-artery ratio was approximately 1.2.

All patients underwent a treadmill stress test 3 and 6 months after the implantation. All patients were scheduled for a repeat angiography 6 to 12 months after the implantation. If symptoms recurred within 6 months, coronary angiography was performed earlier. Angiographic follow-up was performed at a mean of 9.7 ± 5.6 months after AVGCS implantation in 44 patients (46 lesions [79%]), and the MLD was calculated. In the control group, the follow-up angiography was completed after 8.1 ± 3.5 months in 99 patients (118 lesions [85%]). The measurements of MLD were performed by 3 experienced reviewers blinded to the patient's group. All patients were examined clinically or interviewed by telephone at a mean of 41.7 ± 11.4 months in the AVGCS group and 46.8 ± 13.1 months in the control group.

The following end points were evaluated: (1) procedural success: harvest of the autologous venous graft, covering of the stent, and deployment of the AVGCS to the desired diameter, (2) acute thrombosis: thrombosis of the stent within 24 hours of delivery, (3) subacute closure: stent thrombosis between 24 hours and 4 weeks after implantation, (4) clinical events: death, myocardial infarction, target vessel revascularization (TVR) rate, and (5) angiographic restenosis: diameter stenosis 50% of the reference diameter.

Continuous variables are expressed as mean ± SD. Rate of event-free survival was studied with the Kaplan-Meier analysis. In the AVGCS group, to assess the impact of multiple covariates (age, clinical status, sex, MLD, anatomic region of venous graft harvest) on clinical events, we used Cox regression analysis.