VEST IV is a unique study, presenting for the first time the long-term follow-up of a randomized controlled trial evaluating the impact of external stenting of SVGs used in CABG surgery. Our key finding is that the benefits of external stents on improving Fitzgibbon perfect patency and reducing IH that was demonstrated at 1 year are maintained at 4.5 years. Similarly, with the exception of an additional vein graft occlusion in the nonstented group, there was little further deterioration at longer-term follow up. A second key finding, as discussed subsequently, is that the distance of the VEST stent from the SVG lumen was an important determinant of the degree of IH. The MACCE rate of 18% in this study was comparable to that reported by the SYNTAX trial group for their 3-vessel disease subgroup at 4 years and at 5 years post CABG at 19 and 24% respectively [11, 12].
This study strengthens and corroborates previous pre-clinical and clinical reports suggesting a protective biomechanical effect of a braided cobalt chrome external stent on vein graft remodeling [7, 8, 13, 14]. A particular strength of our trial was the paired study design with each patient acting as their own control, thereby eliminating many of the potential factors that could affect SVG disease progression. The stented and nonstented graft groups were well balanced with respect to baseline anatomical and physiological parameters that might contribute to the development of IH, including the diameter of the native coronary artery and the severity of the proximal coronary artery stenosis. This was also evidenced by the similarity of measured graft flows in both the stented and nonstented SVGs, intraoperatively by TTFM and at 1-year by angiography [7]. Furthermore, while previous IVUS studies on SVGs were limited only to the proximal part of the grafts at one time point [15], our study provides, for the first time, new insight into the diffuse nature of the disease by recording intimal development along the entire SVG length at both 1 and 4.5 years.
The early phase of vein graft remodeling is dominated by luminal enlargement followed by a later phase of vein graft thickening and IH. Luminal enlargement is generated mainly by the exposure to the high shear stress of the arterial circulation [3]. During the first months post implantation, as a reaction to the elevated wall tension, there is a proliferative thickening of the venous wall and changes in wall composition [3]. Hozumi et al. have shown that between 1 to 12 months post implantation, intimal area was significantly increased from 0.90 mm2 to 5.26 mm2 (p < 0.001) [15]. Our findings concur with Hozumi et al., suggesting that most SVG remodeling occurs over the first year after grafting and, consequently, that early intervention is crucial in order to effectively lessen the pathological changes which serve as the foundation for occlusive atheromatous lesions 5–10 years following surgery [15]. The mitigation of IH and lumen deformation achieved by external stenting was most pronounced in the first year after implantation and was maintained over the subsequent 3.5 years. This observation is also in line with experimental data on biodegradable external stents which demonstrated that the majority of the inhibitory effect on SVG remodeling was achieved 6 months post implantation [16].
In both arteries and veins, lumen irregularities generate low and oscillatory shear stress that is directly correlated with accelerated vascular disease [13, 17]. A high proportion of SVGs, mainly from the upper leg, demonstrate caliber irregularities even during harvesting while > 25% of SVG show severe segmental ectasia with > 50% dilatation at 1 year [7]. In their landmark publication, Fitzgibbon and colleagues described the progression of SVG disease over the course of 15 years [9]. At 5, 10 and 15 years, only 52, 23 and 19% of the patent grafts respectively demonstrated “perfect” patency with no lumen irregularities. The diffuse progressive nature of vein graft disease was further supported by contemporary studies that demonstrate that only 12–25% of SVG are perfectly patent 10 years after surgery [18]. We found that reduced SVG lumen irregularity is associated with decreased intimal area and thickness, most probably due to the improved hemodynamics and reductions in oscillatory shear stress provided by external stents [13, 17]. External stenting prevented SVG deformation early after implantation and mitigated development of further lumen irregularities at 4.5 years.
Mechanical external support of SVGs has been a focus of intense research, with their use intended to reduce well-documented pathophysiological changes that occur in the SVG following implantation. These devices have developed substantially over recent years, and there is now a large body of data in both animal models and human patients related to their biomechanical effects [17, 19, 20]. External stenting targets some of the key factors initiating the pathological cascade in SVG post implantation such as high circumferential wall stress and disturbed flow patterns due to luminal irregularities [13, 14, 17]. The rationale and the physiological basis for mechanical external stents is further supported by the reported benefits of the “no touch” technique, in which the saphenous vein is harvested with a pedicle of surrounding tissue that serves as an external support with both mechanical and biological roles [21]. Randomized prospective studies showed that no touch technique is associated with superior patency rates compared with conventionally prepared vein grafts both in the short-term (18 months) and long-term (8.5 years) at angiographic follow-up [22, 23].
Early clinical experience with external stents showed conflicting data with early patency rates ranging from 28 to 92% at 6–12 months [24, 25]. High SVG failure rates were attributed to the design of the first generation of external stents that required incorporation of the device into the anastomoses, constriction of the graft and fixation of the stent to the SVG with fibrin glue [25]. Several studies have shown that fibrin glue on the external surface of vein grafts lead to aneurysmal degeneration and excessive IH which may jeopardize vein graft patency [25,26,27]. Second generation technologies which did not require the use of fibrin glue or incorporation to the anastomoses site showed comparable patency rates to non-stented SVG at 1 year [7].
The appropriate size of external stents for vein grafts has been a focus of controversy and another possible confounding factor for success of external stenting. Zilla and colleagues concluded that constrictive external stents, which reduce diameter mismatch with the target artery, are more effective in mitigating IH compared to non-restrictive stents [28]. In contrast, Izzat and colleagues have shown that loose-fitting external stents are more effective in suppressing intimal proliferation [29]. The rationale behind using oversized external supports was to provide sufficient space to promote adventitial neovascularization, which was potentially interrupted by constrictive stents [20]. The distance of the external stent from the lumen is equivalent to the SVG wall thickness. The closer this measure is to the vein baseline wall thickness, the better the conformity of the VEST to the vein. We postulated that a closer conformity would improve VEST performance in mitigating IH, and indeed we found that the correlation between the distance of the VEST from the lumen and the IH thickness was indeed significant in our study. As shown in Fig. 6, the closer the value is to the baseline literature reported vein wall thickness, the lesser the proliferation of IH [30, 31]. This finding suggests that mildly constricting the vein has a beneficial effect on intimal proliferation, most probably due to more effective reduction in wall tension, SVG dilatation, and lumen irregularities. It is expected that future studies, in which appropriate model selection will ensure a mildly constrictive stent, a more effective reduction in IH will be achieved.
Limitations
Together with the small sample size, a further limitation of our study is that the study cohort was based on a first in human trial and exhibited learning curve effects of both the technology and the implantation technique (which were largely resolved in VEST II [8]). The technical failures and inappropriate model selection, with suboptimal dimensional match between the SVG and the stent, likely have adversely affected the stent’s ability to even more effectively mitigate IH. VEST III, a large randomized trial that is currently underway (NCT02511834), will address both the population size and learning curve issues.