VEST external stent for coronary artery bypass grafts

Taggart et al. 2015 (VEST-I)

Study size and design

Prospective randomised study in 30 people having on-pump multivessel SVG including a left internal mammary artery graft to the left anterior descending coronary artery and SVGs to right and circumflex territories.

Intervention and
comparator(s)

Each patient had the intervention (VEST-stented SVG) and control (non-stented SVG); 1 VEST device was randomly assigned intraoperatively to a single SVG, with 1 or more SVG remaining non-stented.

Randomisation of which SVG had VEST (either circumflex or right coronary artery) was done intraoperatively by sealed envelope after the distal anastomosis.

Key outcomes

The SVG mean intimal hyperplasia area, assessed by IVUS at 12 months in 43 SVGs, was significantly reduced with VEST (4.37±1.40 mm²) versus non-stented SVG (5.12±1.35 mm², p=0.04). Overall SVG failure rates did not differ significantly between the 2 groups (30% stented vs 28.2% non-stented SVG; p=0.55). The stented versus non-stented SVG failure rates were significantly lower in the circumflex territory (17.6% vs 27.5%; p=0.02) and significantly higher in the right territory (46.2% vs 13.4%; p=0.01).
Lumen regularity, using the Fitzgibbon classification, showed a higher proportion of stented SVGs were in class I (62% vs 39%; p=0.08) and with a lower incidence of SVG ectasia (6.7% vs 28.2%; p=0.05).
It was hypothesised that the high failure rate of VEST-stented SVG to the right territory was a result of using metal clips to ligate SVG side branches and fix VEST to the anastomoses, which was not part of the company's instructions. In the left territory, the patency rates were higher than the control. This was the rationale for initiating the VEST-II study.

Strengths and limitations

VEST-I was a first-in-man randomised study comparing vein grafts in the same patients with and without VEST.

It had only 1-year follow-up measuring intimal hyperplasia, which is a surrogate outcome for vein graft disease. It included a small number of patients and there was no separate control group (all patients had both VEST and control), so no differences in clinical outcomes could be reported. It was also a company-funded study; the first author is a company shareholder and has received honoraria from the company.

Taggart et al. 2017 (VEST-II)

Study size and design

Prospective single-arm study in 30 people having on-pump multivessel SVG, with at least 1 vein graft bypass indicated for the right coronary artery. Single centre in UK.

Researchers investigated whether avoidance of both fixation of the external stent to the anastomoses and the use of metal clips to ligate SVG side branches would improve the early patency of externally stented SVG to the right coronary artery. This was because of high failure rate in the right territory observed in VEST-I (see above). Because the right and left territories are different in anatomical structure and haemodynamics and cannot be compared (unless properly randomised), the results were compared with published historical VEST-I data on SVG to the right territory.

Intervention and comparator(s)

In each patient, an SVG to the right territory had VEST, and all other venous grafts to the left side were unsupported. SVG patency was assessed by CT angiography at 3 to 6 months (graft failure defined as occlusion or >50% stenosis). No control arm was used for this study: instead, results were compared with the historical VEST-I data of SVG failure rate to the right territory.

Key outcomes

29 patients (96.6%) completed follow-up and CT angiography data were available for a total of 43 SVGs, (29 stented and 14 non-stented SVGs) and 47 arterial grafts. Patency of stented SVGs was 86.2% (25/29 on CTA). Avoidance of both metallic clips to ligate side branches and of fixation of VEST stents to 42 of the anastomoses improved patency of stented SVG to the right coronary territory.

The patency rates were comparable to those generally described in literature on early patency rates of SVGs to the right coronary territory (in the range of 71% to 86% at 6 to 12 months). This study showed that when specific recommendations for use are followed, the VEST external stent for SVGs to the right territory is safe.

Strengths and limitations

This was a non-randomised study with no control group, using historical control data. It was done at a single centre and had a short follow-up. It was also another company-funded study; the first author is a company shareholder and has received honoraria from the company.

Meirson et al. 2015

Study size and design

Post-hoc computational fluid dynamics analysis of data from 29 patients in the VEST-I trial. Diffuse flow patterns were assessed using mean values of various hemodynamic parameters, including time-averaged wall shear stress and OSI. Focal flow disturbances were characterised using percentile analysis of each parameter.

Intervention and comparator(s)

See VEST-I.

Key outcomes

In both diffuse and focal flow-pattern analyses, OSI was significantly lower in the stented vs non-stented SVG group (p=0.009 and p<0.003 respectively). No statistically significant differences were observed in time-averaged wall shear stress values. High OSI values were correlated with the development of intimal hyperplasia (p=0.01).

Strengths and limitations

This study has the same limitations as VEST-I, as well as being a post-hoc study (the original study did not specify these outcomes in the original protocol).

Webb et al. 2015

Study size and design

This VEST-I sub-study assessed SVGs with and without OCT in the 24 patients in VEST-I that had coronary angiography with OCT imaging using a non-occlusive technique.

Intervention and comparator(s)

See VEST-I (1-year follow-up).

Key outcomes

Mean cross-sectional area was greater in unstented vs stented grafts (8.4±3 vs 7.6±2.7 mm; p=0.005). The lumen of the stented grafts was more homogeneous (difference between maximum and minimum lumen diameter was significantly smaller in stented compared with unstented grafts, 0.28±0.19 vs 0.33±0.23 mm respectively; p=0.006), and more circular (mean eccentricity index 0.08±0.06 vs 0.10±0.06, stented vs unstented respectively; p=0.019). Adherent thrombus was identified in 3 grafts (all unstented).

Strengths and limitations

This study has the same limitations as VEST-I. The main VEST results did not report the OCT results because many of the conduits were too large to image the entire circumferential depth of the vessel wall. OCT findings may also highlight the early changes occurring in SVGs after implantation of aorto-coronary bypass conduits, changes that may accelerate vein graft failure.

Abbreviations: IVUS, intravascular ultrasonography; OCT, optical coherence tomography; OSI, oscillatory shear index; SVG, saphenous vein graft.