Patches for Carotid Artery Endarterectomy: Current Materials and Prospects

Abstract

Patch angioplasty is commonly performed after carotid endarterectomy. Randomized prospective trials and meta-analyses have documented improved rates of perioperative and long-term stroke prevention as well as reduced rates of restenosis for patches compared to primary closure of the arteriotomy. Although use of vein patches is considered to be the “gold standard” for patch closure, newer generations of synthetic and biological materials rival outcomes associated with vein patches. Future bioengineered patches are likely to optimize patch performance, both by achieving minimal stroke risk and long-term rates of restenosis, as well as by minimizing the risk of unusual complications of prosthetic patches such as infection and pseudoaneurysm formation; in addition, lessons from bioengineered patches will likely enable construction of bioengineered and tissue-engineered bypass grafts.

Since surgical repair of symptomatic carotid artery stenosis was first reported by Eastcott and colleagues in 1954,¹ carotid artery endarterectomy (CEA) has remained the standard management strategy for significant carotid stenosis in both symptomatic and asymptomatic patients. Approximately 100,000 CEA are performed annually in the United States.^2–4 The standard surgical approach for CEA involves a longitudinal arteriotomy from the common to the internal carotid artery, allowing plaque removal. Unfortunately, closure of the longitudinal arteriotomy also allows for the possibility of narrowing the artery, either immediately or in delayed fashion, thus mimicking the stenosis for which the surgery was originally performed. Closure of the arteriotomy with a patch minimizes the effect of neointimal hyperplasia and scarring, maintaining arterial lumen diameter after the procedure. Eversion endarterectomy is an alternative surgical technique that allows plaque removal without longitudinal arteriotomy and potentially avoids placement of a patch; however, this technique is less frequently practiced. In addition, patches are commonly used to close arteriotomies in other vascular beds, such as after common femoral endarterectomy or profundaplasty.

To avoid restenosis after CEA, carotid patching was routinely used by Imparato as early as 1965,⁵ and many articles have since supported the use of a patch.^6–11 Another early advocate of vein patching after CEA was Dr. Thor Sundt, who was also a pioneer in his laboratory investigations examining the healing of carotid patches in a dog model.¹² Recent reviews continue to highly recommend patch angioplasty after CEA to avoid restenosis, compared to primary arterial closure.¹³ Patching is now thought to be part of optimal care of the patient undergoing traditional CEA.^14,15 In this review, we describe traditional and novel patch materials, and review the characteristics and clinical results of these patches.

Why patch at all?

Important factors for all successful surgical procedures include simplicity, ease, safety, short duration, and cost-effectiveness. As such, primary closure for CEA can be seen as a good choice, with patch closure being somewhat more complicated and of longer duration. Therefore, it is a valid question to ask whether or not patching after CEA is a reasonable activity at all.

In 2004, Bond and colleagues reviewed the outcome of 7 randomized trials that compared primary closure with patch angioplasty after CEA.^10,11 These 7 randomized trials, involving 1281 procedures, showed that patch angioplasty is associated with reduced 30-day risk of ipsilateral stroke (1.6% vs. 4.8%, p=.001), reduced risk of stroke or death (2.5% vs. 6.1%, p=.007), reduced rates of return to surgery (1.1% vs. 3.1%, p=.01), and reduced rates of arterial occlusion (0.5% vs. 3.6%, p=.0001), compared to primary closure. In addition, carotid patching was associated with reduced long-term rates of ipsilateral stroke (1.6% vs. 4.8%, p=.001), reduced risk of stroke or death (14.6% vs. 24.1%, p=.004), and reduced rates of recurrent stenosis (4.8% vs. 18.6%, p<.0001) compared to primary closure. This meta-analysis provides strong evidence that carotid patching provides both perioperative as well as long-term benefits for patient care, and is consistent with standard use of patching during CEA. The benefit that is probably the most generally agreed upon is the reduced rate of restenosis in the long term. Several important series are reviewed in Table 1; although these series are heterogeneous, and reflect different patch materials and times of followup, this Table nevertheless shows that placement of a carotid patch is associated with fewer strokes and less restenosis compared with primary arterial closure.

Table 1.

Improved outcome after patch closure compared to primary closure after CEA. Time and degree of restenosis were defined by the authors of the referenced study.

		n		stroke		restenosis
Study	ref	primary	patch	primary	patch	primary	patch
Hertzer 1987	70	483	434	3.1%	0.7%	31%	9%
Ranaboldo 1993	71	104	109	5.8%	1.8%	16%	6%
AbuRahma 1996	72	135	264	5.2%	1.5%	12%	3%
Katras 2001	73	97	107	2.8%	1.0%	9%	6%
Ali 2005	74	117	119	7.7%	1.7%	25%	7%
Rockman 2005	75	233	1377	5.6%	2.2%
Verhoeven 2005	76	83	236	6.0%	2.5%	11%	7%
Hertzer 2006	77	783	1479	2.8%	1.4%	29%	15%

Other accepted indications for patch angioplasty after CEA traditionally include a very small internal carotid artery (< 4 mm), an extended, complex, or irregular arteriotomy, and concomitant repair of a distal internal carotid artery that contains a kink or coil. As such, patching may allow optimization of blood flow, vessel geometry, and biomechanics, although the influence of these physical parameters on long-term patient outcome is not well described.

Types of patches

The ideal requirements for any patching material include: 1) long term stability and durability, 2) low risk of restenosis, 3) compliance near that of the host artery, 4) comfortable handling characteristics, 5) easy harvest or ready to use, 6) anti-coagulant function, and 7) resistance to infection and late degeneration (Table 2). As described below, there are a variety of materials in common use for arteriotomy closure during CEA, each with advantages and disadvantages.

Table 2.

Ideal requirements for a carotid patch material.

-	long term stability and durability
-	low risk of restenosis
-	compliance near that of the host artery
-	comfortable handling characteristics
-	easy harvest or ready to use
-	anti-coagulant function
-	resistance to infection and late degeneration

Prosthetic patches

The most commonly used prosthetic patching materials are expanded polytetrafluoroethylene (PTFE) and Dacron. PTFE is a fluoride resin composed of only carbon and fluoride; expanded PTFE (ePTFE) has a porous structure with 20–30 µm fibril distance. ePTFE is also commonly used as vascular grafts, and it has properties that include resistance to thrombosis and the ability to support re-endothelialization. More recently, ePTFE patches have an elastomeric coating such as polyurethane applied to its outside surface to minimize suture hole bleeding.¹⁶ Dacron is a polyester fiber, a condensation polymer of ethylene glycol and terephthalic acid. Dacron shows high tensile strength and resistance to stretching, and woven or knitted sheets of Dacron are commonly used in vascular surgery, including use as vascular grafts.

An early Italian trial first showed the importance of prosthetic patch angioplasty in preventing restenosis after carotid endarterectomy.¹⁷ Prosthetic patches have a significant advantage since they are ready-to-go, i.e. available, by just opening the package. In addition, outcomes of recent generations of various prosthetic materials show no differences when compared to autologous vein patches.^18–20 For example, Naylor et al. reviewed the Leicester experience with 269 patients randomized to vein or thin-walled Dacron (Hemashield Finesse) patch closure. After 3 years, cumulative freedom from death or ipsilateral stroke was 93.0% in the Dacron-patched group compared to 95.5% in the vein-patched group (p=0.42). Interestingly, cumulative freedom from recurrent stenosis (>70%) was 92.9% in the Dacron-patched group compared to 98.4% in the vein-patched group (p=0.03).²⁰ Similar results were reported from the Cleveland Clinic, with synthetic and vein patches having similar low rates of late stroke (2.1% vs. 2.0%) and slightly higher but not statistically significant incidence of restenosis (>60%) in synthetic patches compared to vein patches (6.3% vs. 4.8%, p=0.99).¹⁹ A recent Cochrane review suggested an odds ratio for risk of restenosis of 1.34 for PTFE patches compared to vein, but with a very wide 95% CI (0.71–2.51).²¹

Although PTFE patches were originally very commonly used for CEA, collagen-impregnated Dacron patches became more commonly used upon recognition of their advantage in hemostatic function, i.e. reduced bleeding from the suture holes.¹³ Carney et al reported that PTFE patches showed significantly longer time to attain hemostasis after release of the clamps at the end of the operation, as well increased incidence of blood loss > 300 ml, and greater use of oxidized cellulose to stop the bleeding, compared to vein or Dacron patches.²² In addition, AbuRahma reported significantly longer hemostasis times in PTFE-patched patients compared to Dacron-patched patients (14.4 vs. 3.4 minutes, p<0.001).²³ In addition, this group also reported that the long term results of older collagen-impregnated Dacron patches compared unfavorably with ePTFE, with higher rates of postoperative stroke (7% vs. 0%, p=0.02) and carotid restenosis (12% vs, 2%, p=0.013). Recent newer developments include the development of sealing ePTFE patches (ACUSEAL; W. L. Gore & Associates, Flagstaff, AZ) and less thrombogenic Dacron patches (Hemashield Finesse), with each generation attempting to improve upon previous versions; for example, hemostasis times are typically shorter (3–4 minutes) and comparable between materials. AbuRahma and colleagues recently reported no significant differences in perioperative stroke (2% vs. 2%, p=1.0) and short-term restenosis risk (0% vs. 4%, p=0.12) between these patches.²⁴ These results were recently confirmed in longer term followup, with cumulative stroke-free rates of 98% (ePTFE) and 97% (Dacron) at 3 years (p=0.7), whereas cumulative freedom from restenosis (>70%) was 89% (ePTFE) and 79% (Dacron) at 3 years (p=0.04).²⁵

In a recent Cochrane review, ePTFE and Dacron patches showed similar rates of arterial restenosis and occlusion compared to vein patches (OR 1.01, 95% CI 0.61–1.66).²¹ Several large series have suggested increased rates of restenosis in Dacron patches (Table 3). It is likely that as additional materials are made available for clinical use the results of these two prosthetic materials will converge. We believe that one advance in reduction of bleeding time traditionally associated with ePTFE patches is the use of superiorly swagged needles, with the needle diameter not significantly larger than the suture diameter, creating less empty space for bleeding around the suture. Thus time to hemostasis may be less patch-dependent with the use of newer sutures and needles.

Table 3.

Outcome after patch closure – effects of different patch materials. Time and degree of restenosis were defined by the authors of the referenced study.

		n			stroke			restenosis
Study	ref	vein	Dacron	ePTFE	vein	Dacron	ePTFE	vein	Dacron	ePTFE
AbuRahma 1996	72	130	-	134	0.8%	-	2.2%	2.9%	-	2.2%
Archie 2000	78	903	359	27				0.6%	6.4%	3.7%
Jacobowitz 2001	32	159	90	-	2.0%	2.2%	-	2.2%	8.5%	-
Grego 2003	79	80	-	80	1.3%	-	6.4%	9.3%	-	13.3%
Naylor 2004	20	134	133	-	4.5%	7%	-	1.6%	7.1%	-
AbuRahma 2008	25	-	100	100	-	3%	2%	-	21%	11%

Other significant long term sequelae associated with prosthetic patches include pseudoaneurysm formation and development of infection.^26,27 Both complications are unusual but treatable. On the other hand, these complications certainly suggest that use of prosthetic patch materials may require life-long surveillance in susceptible populations, and thus are clearly not perfect materials. In particular, the rates of infection are unfavorably higher compared to other materials, and remain as a point of improvement for future developments.

An early report by Branch and Davis linked infection and pseudoaneurysm formation after CEA; this review of 57 cases estimated an incidence of 0.30% of pseudoaneurysm after CEA, and estimated that CEA performed with primary closure had half this rate of postoperative infection.²⁸ One of the largest series of prosthetic patch infections was described by Dr. Cooley’s group at the Texas Heart Institute.²⁹ This report of 13 cases of patch infections associated with pseudoaneurysms discussed several points from this group’s extensive experience; these points remain critical for optimal patient management: ¹) Infection complicates patches, both prosthetic and vein, more frequently than primary arterial closures. ²) Braided sutures such as silk can trap bacteria in their interstices, providing a nidus of infection in earlier series; the use of monofilament sutures, especially polypropylene, has eliminated this risk. ³) Untreated patch infection can lead to patch blowout and massive hemorrhage, as well as sepsis, abscess formation, or stroke; these presentations must be treated urgently. ⁴) Staphylococcus species were the most commonly cultured organism (29/30 cases), followed by gram-negative rods (7/30 cases). ⁵) Surgical repair involves extensive dissection and usually needs general anesthesia. ⁶) Use of a shunt is desirable but potentially hazardous, due to the friable tissues, and usually impractical. ⁷) Debridement of the arterial wall usually precludes safe primary closure; 75% of the cases treated by reclosure of the arteriotomy required early reoperation or died. ⁸) The majority of repairs were treated by resection and saphenous vein replacement; treatment with autogenous tissue is thought to be mandatory in the presence of gross infection. ⁹) Minor infections and pseudoaneurysms not associated with infection may be treated with partial aneurysmectomy and patch repair, although even saphenous vein patches can be re-infected. ¹⁰) 10% of patients so treated developed postoperative strokes or died, most commonly in patients needing carotid artery ligation (50%), and least frequently in patients in whom repeat patch placement was possible (12%). ¹¹) Recurrent patch infection was common in patients receiving Dacron patches. ¹²) Donor vein site infection was also possible. ¹³) Cranial nerve injury was less common in these procedures compared to primary CEA. This series also estimated that infected patches and pseudoaneurysms occur in 0.18% of CEA.²⁹

A more recent review of prosthetic patch infections has estimated that patch infection occurs in approximately 0.37% of all cases, ranging from 0.26% to 0.71% in several reported series.²⁷ Repair of these prosthetic patch infections is associated with increased morbidity compared to the primary CEA procedure; although the reported postoperative mortality rate was 2.6%, the postoperative stroke rate was 2.6%, the rate of cranial nerve injury was 12.8%, and the rate of recurrent infection was 7.7%, all of which were greater than rates associated for elective repair.²⁷ General recommendations from this group included reconstruction with either a vein patch or an interposition graft, depending on the quality of the remaining artery after removal of all prosthetic and infection. The use of a muscle flap to cover the site has been reported, ³⁰ although the few overall number of case reports makes it unable to determine whether this adjunctive technique is popular or not.

Venous patches

Patching with autologous venous tissue remains the most commonly used option for arterial patching during CEA, and continues to show superb results in the literature (Table 3). This patch continues to enjoy popularity with surgeons as it is commonly used, has excellent handling, and is resistant to thrombosis and restenosis due to its endothelial lining on the luminal surface.³¹ There have been many reports comparing results of CEA after use of autologous or synthetic patches.^{6,11,13,19,21,32} Most analyses have shown that there are no significant differences in early outcomes when comparing venous with prosthetic patches, with very low risk of any events. O'Hara reported the results from the Cleveland Clinic that randomized 207 cases to vein or synthetic patch closure; the stroke rate in the vein patch group was 3.0% compared to 2.1% in the synthetic patch group (p=0.99). Recurrent stenosis (>60%) was present in 4.8% of the vein patch group compared to 6.3% of the synthetic patch group (p=0.99).¹⁹ Similarly, Jacobowitz reported the New York University experience.³² In 159 vein patches compared to 90 Dacron patches, the rate of perioperative stroke was similar (vein 1.3% vs. Dacron 1.1%, p=NS); the rate of late stroke was also similar (vein 2.0% vs. Dacron 2.2%, p=NS). Restenosis (50–79%) was present in 2.2% of vein patches and 8.5% of Dacron patches (p=NS).

It is of interest that autologous vein was the first material to be used for CEA patching. Imparato was an early proponent for vein patches to be used routinely to prevent restenosis.³³ Originally surgeons used the proximal saphenous vein in the thigh, but patients and surgeons objected to the additional incision in the thigh for an operation that should be confined to the neck. This preference to avoid proximal saphenous vein harvest led to use of the distal saphenous vein at the ankle or the cervical veins harvested within the CEA incision. There was concern that these veins were weaker and could potentially lead to a catastrophic blowout.^34–38 Some authors have shown that veins > 3.5 mm are generally safe to use with reduced risk of rupture.^39,40 A clever development deployed everted cervical vein (external jugular or facial vein), thereby creating a double-walled vein patch with increased tensile strength comparable to the saphenous vein. Since cervical veins are harvested within the CEA operative field, the additional leg incision is obviated.⁴¹ The double layered everted cervical vein patch has demonstrated durable outcomes compared to the traditional saphenous vein patch.⁴²

Another option is the cryopreserved homograft saphenous vein patch. Plestis et al reported a series of 1006 consecutive cases of CEA repaired with saphenous vein segments that were harvested from coronary artery bypass procedures, and cryopreserved in 10% DMSO at −120°C.⁴³ Results were excellent, with 1.2% perioperative strokes, and 96% 10-year cumulative freedom from ipsilateral stroke. Recurrence of severe (>75%) stenosis was 2%, and freedom from >20% restenosis was 84% at 10 years. These results suggest that modified vein may be a durable substitute for autologous vein.

There are few reports of infection after vein patch placement. In an early series, Thompson reported no cases of vein patch infection in 1,140 CEA, although there were 7 cases of Dacron patch pseudoaneurysms (0.6%) in the series.⁴⁴ A notable case report of a vein patch infection reported repair with resection, segmental replacement with a vein graft, topical irrigation for 2 weeks, and systemic antibiotics for 3 weeks.⁴⁵ Yamamoto et al. reported their experience with 2888 CEA closed with vein patches at the Mayo Clinic in 23 years; there were only 3 cases of infection, all of which involved Dacron or Teflon mesh reinforcement of the site without involvement of the vein patch itself.⁴⁶ These cases were treated with removal of the synthetic material without disturbance of the vein patch. This group also reported 5 patch ruptures of uninfected vein patches, 3 of which led to death or severe disability, and 4 cases of late (1–9 year) aneurysmal expansion. Interestingly, the group concluded that use of a synthetic material was preferable to a vein patch.⁴⁶

Biomaterial patches

Bovine pericardium has served for many years as a popular option as a biomaterial patch for CEA.^47,48 Kim et al reported their preliminary experience with this patch for CEA, with little differences in outcome compared to vein patches (no early strokes; >50% restenosis 3.3% vs. 1.6%, p=NS).⁴⁹ Hines et al have recently confirmed these results, with excellent handling and early results; <50% restenosis occurred in 25% of cases by 2 years, but 16% of cases had 50–79% restenosis, and no cases were detected with >80% restenosis.⁴⁸ Bovine pericardium offers the benefits of off-shelf availability, durability, and biocompatibility, as well as the ability to ultrasound through the patch immediately after placement. In addition, the satisfactory use of bovine pericardial patches in infected fields has been reported. However, bovine pericardium has had reduced popularity following reports in the lay press of bovine spongiform encephalopathy (BSE) in certain cattle herds, although BSE has never been reported after placement of a carotid patch.

When compared to outcomes following use of polyester patches, bovine pericardial patches show comparable results but may have a lower incidence of recurrent stenosis; one study reported 4% restenosis in bovine patches compared to 7.6% restenosis in polyester patches, although the mean length of followup in the groups (bovine, 12 months; polyester 24.5 months) was not comparable.⁵⁰ Just as the newer prosthetic patches and standard vein patches show no significant differences regarding durability and outcomes, it is likely that bovine pericardial patches are also equivalent. Bovine pericardial patches have shown significantly decreased intraoperative suture line bleeding compared to prosthetic patches.⁵¹ Although there are no reports comparing bovine pericardium with other conduits regarding rates of postoperative infection, bovine pericardium has been used in other infected cardiovascular fields.^52,53 The low risk of infection after autologous vein and biomaterial patches may be an important factor for the future development of tissue engineered vascular patches, although this assertion also awaits confirmation in large series.

One other biomaterial that has been reported after use in an animal study is a combination patch, with one side of the patch constructed from glutaraldehyde-fixed bovine peritoneum/fascia, and the other side constructed from polyester.⁵⁴ This interesting patch was tested in femoral arteries of dogs and found to have, at 6 months, no degeneration and complete reendothelialization; the mechanical strength was superior to that of bovine pericardial patches. Other biomaterials for potential use include amnion,⁵⁵ decellularized bovine inferior vena cava,⁵⁶ and decellularized human pericardium.⁵⁷ Decellularized venous patches have similar burst and suture-holding strength as native veins.⁵⁸ As the tissue engineering field matures, additional biomaterials for use a carotid patch can be expected, such as patches with textured surfaces that promote cell migration and tissue healing.⁵⁹

Future directions

As excellent results are currently being obtained with available patch materials, directions for future development may lie in the prevention of unusual complications such as infection and pseudoaneurysm. However, additional benefits may become evident as the field develops.

An interesting option for carotid patching was reported by Jenkins et al, in which they used the superior thyroid artery.⁶⁰ Use of this autologous vessel has the advantages of excellent material strength, reduction of surgical cost and possibly infection risk, excellent compliance match to the host artery, and availability within the operative field. However, the superior thyroid artery has limitations including reduced patch size, focal arteriosclerosis and limited followup data. Ultimately the use of any artery is limited by its potential for creating distal ischemia in the original locus. Use of an endarterectomized occluded femoral artery has been reported, eliminated the potential for distal ischemia; ⁶¹ nevertheless this option has not become popular. However, the lessons derived in using autogenous artery suggest that strength and compliance matching are critical determinants for successful patch materials.

Tissue engineering, the combination of scaffolds and cells to develop neo-tissues, has recently become a popular field, as there is potential to create neo-tissues with both off-shelf availability as well as device alterations that can be “personalized” to individual patient requirements. Novel synthetic conduits formed by tissue engineering will very likely replace structural arterial tissue after surgery. In the late 1980’s, the concept of tissue engineering grafts, with polymer scaffolds absorbed by newly replaced tissue, was reported by Langer and Vacanti. Based on this concept, the first report of a tissue engineered conduit applied in the human vascular system was the use of a tissue engineered venous graft to repair congenital defects in the pulmonary artery.⁶³ Shin’oka and colleagues used autologous bone marrow cells to seed a scaffold copolymer of L-lactide and ε-caprolactone reinforced with a polyglycolic acid sleeve and reported satisfactory midterm clinical results in congenital pulmonary artery system repair.⁶⁴ This tissue engineered conduit has only been used in the venous system, but has potential as an arterial patch once its long-term strength and pathophysiological changes to the arterial circulation are better understood.

Specific biological cellular approaches for tissue engineering arterial grafts have started to be applied in clinical practice, with reports of dermal fibroblasts rolled into sheets and then used as tissue engineered vascular grafts.^65–67 This method has focused on use of tissue engineered vascular grafts as arteriovenous grafts for hemodialysis access, and has shown fair prospects as a conduit. The advantage of this conduit, rolled sheets of dermal fibroblasts, for application as a carotid patch, is the ability to be tolerated within the human arterial environment, including arterial pressure. In addition, this construct is based on autologous cells and thus is not likely to lead to rejection. Another recent interesting approach to tissue engineering an arterial conduit has used cross-linked elastic salmon collagen as a vascular graft.⁶⁸ Although this particular conduit is not yet mature, the development of this graft shows that the concepts of using only biological material, that is gradually biodegradable, and with similar wall compliance as native vessels, may ultimately lead to the development of even better improved tissue-engineered devices.

Tissue engineered grafts and patches still have obstacles to surmount. One of the major requirements of a tissue engineered vascular patch is long-term stability. Another important requirement is to ensure a stable supply of cells within the neotissue. If the source of the cells for a tissue-engineered patch include cells from middle-aged or elderly patients, then these cells may not grow well, or very slowly if at all. If these cells are taken from other humans or large animals, then infection and rejection need to be eliminated. The application of stem cells or novel types of stem cells, such as embryonic stem (ES) or induced pluripotent stem (iPS) cells, for tissue engineered conduits has not yet been worked out; problems with ethical issues, accurate induction of specific cell types, carcinogenic gene transduction, and other problems still exist. For the potential of “off-shelf” use, hybrid biomaterials such as tissue engineered conduits still have to surmount numerous obstacles.

Conclusion

Although minimally invasive carotid artery angioplasty and stenting preoccupies much discussion in the treatment of carotid artery stenosis, the majority of patients with carotid disease are still treated by open surgery.^4,13,69 As such, development of carotid patches, as well as patches for use in arteriotomy closure elsewhere, continues to remain a critical adjunct for vascular surgery. Reports comparing most synthetic and vein patches show acceptable satisfactory results after CEA, although problems such as infection, rupture, handling, stroke, cost, and others, remain. Although our preference is for biological patches such as bovine pericardium, we believe that the clinical results obtained with most available synthetic patches are currently similar enough to prevent clear recommendation of any particular one. Recent novel conduits are not yet mature enough for clinical use, although the future of biomaterials and tissue engineered patches, especially for use as vascular bypass grafts, is likely to be bright.