Conduits for Vascular Reconstruction in the Pediatric Patient 

Autogenous arterial and venous grafts, such as great saphenous and internal iliac veins and internal mammary, radial, and hypogastric arteries, remain the criterion standard conduits for vascular reconstruction.

Vascular reconstruction in pediatric patients requires an approach substantially different from that used in the adult population. Many factors must be considered when dealing with this patient population. Such factors include the small caliber of their vessels, the possibility of spasm, the risk of infection, the propensity for children to rapidly form collateral circulation, the inevitability of growth, and the strong tendency for stenosis and growth arrest to occur. See the images depicted below.

Carrel is credited as the pioneer of vascular surgery and of the use of cardiovascular tissue allografts; he is recognized for his work on vascular anastomoses.^[1] Initial results with his techniques were disappointing because of the high rate of degeneration and vascular failure due to poor preservation methods, immunologic rejection, and inflammatory reactions that produced fibrosis, calcification, and aneurysms. However, using his descriptions, major advances have been accomplished.

One of the risks of tissue transplantation is the potential for disease transmission from the donor to the recipient. Viral, bacterial, and fungal infections have been transmitted by means of tissue allografts. Tissue donors should be tested for transmittable infectious diseases. Synthetic materials are also prone to a high rate of failure in small-to-medium–sized vessels because of thrombosis and stenosis at the site of anastomosis. In addition, these grafts are associated with an increased rate of infection that leads to repeated operations. Experimental studies have demonstrated that vascular allografts are more resistant to infection than synthetic grafts.

The use of autologous grafts has provided the best results for vascular reconstruction. The internal and external saphenous veins or the basilic and cephalic veins have been used with success, as have the neck, radial, and pelvic veins. Longer conduits and smaller veins (< 3.5 mm) are technical risk factors for graft failure. In children, preserving the saphenous vein is advisable. Their small diameter is insufficient for large vessels reconstruction. Currently, 40% of pediatric patients are overweight and may later experience a high incidence of conditions such as cardiac disease, hypertension, and diabetes; thus, preserving these veins in order to have them available for coronary bypass once these patients become adults is recommended. No severe late complications (eg, venous stasis, ulcers, edema) are reported after vein harvesting from the lower extremities in children.

The authors have successfully used the contralateral internal iliac vein as a graft to replace a segment of external iliac artery that was resected en block with a pelvic tumor. The authors have also used contralateral saphenous veins to reconstruct arteries and veins during limb salvage procedures in which the vessels have to be removed en block with the tumor in order to achieve clear margins of resection. In such instances, bypass before tumor removal may be beneficial to decrease ischemic damage and reperfusion injury because the vascular repair frequently must be delayed during the orthopedic reconstruction. Also, the ipsilateral greater saphenous vein can be divided distally and flipped over to be anastomosed to the external iliac vein while leaving its normal drainage to the femoral vein intact. Cryopreserved cadaveric iliac arteries and veins have also been successfully used as conduits for vascular reconstruction.

Research is being conducted using the umbilical artery or using venous homografts for small-diameter vascular reconstruction in newborn patients. In addition, minimally invasive harvesting techniques have been described. When the host veins are not useful or available for reconstruction, a synthetic or allogenic conduit is necessary.

Prosthetic graft research began in 1952, when Voorhees and colleagues developed the first prosthetic graft using woven nylon. In the late 1950s, nylon, Orion, Ivalon, polyethylene terephthalate (Dacron), polytetrafluoroethylene (PTFE), and other polymers were tested as prosthetic grafts. In 1957, Creech and colleagues concluded that Dacron and PTFE were physically stable, whereas Nylon, Orion, and Ivalon were not. Since then, Dacron and PTFE have been adopted as the standard materials most commonly used in the manufacturing of synthetic prosthetic grafts. PTFE grafts exhibit little inflammatory and thrombogenic reactivity, and PTFE continues to be the material of choice for small diameter vascular grafts.

PTFE is recognized by the body as an implanted foreign material. The blood-graft and tissue-graft interfaces create interactions that ultimately determine the patency of the implanted prosthetic graft. These interactions are complex but largely predictable. The typical reaction cascade starts at the blood-biomaterial interface with the absorption of plasma proteins, followed by deposition and activation of platelets, infiltration of neutrophils and monocytes, and migration and proliferation of endothelial and smooth muscle cells, followed by deposition of extracellular matrix.

The implantation of a biomaterial into an immunocompetent host typically elicits an inflammatory response analogous to wound healing and involves several types of inflammatory cells. These cells synthesize and secrete cytokines that affect the cells in surrounding tissues, promoting cellular and capillary ingrowth. The ideal healing process of prosthetic grafts involves rapid endothelialization of inner surfaces and spatially and temporally limited subendothelial smooth muscle cells ingrowth, followed by phenotypic and functional differentiation of cell components and controlled remodeling of the extracellular matrix.

Endothelial cells play an important role in the maintenance of graft patency. The presence of endothelium on the flow surface prevents platelet deposition and subsequent graft thrombosis; the presence of luminal endothelial cells also inhibits smooth muscle cells migration to the subintimal space, theoretically inhibiting neointimal hyperplasia. Optimizing these processes in order to obtain desirable results is a major concern of current research.

Expanded polytetrafluoroethylene (e-PTFE) is the most commonly used synthetic conduit because of its mechanical properties (see images below). It withstands pulsatile arterial pressures given its good suture-retention strength, with compliance close to that of native artery. The material has increased porosity on the abluminal surface that allows for tissue ingrowth and has low porosity on the luminal surface to prevent graft leakage. In animals, the patency rate was 75%, even in grafts smaller than 4.5 mm.^[2] Furthermore, e-PTFE allows neovascularization response with transmural capillary ingrowth and has good resistance to infection.

The ideal vascular graft has yet to be developed, and all materials have advantages and disadvantages. Desired characteristics of a vascular bypass grafts include mechanical stability, biocompatibility, nonthrombogenicity, infection resistance, availability, and cost-effectiveness. In children, prolonged life expectancy must also be considered during fabrication of the ideal graft. Synthetic conduits are prone to infection, and, in small pediatric patients, grafts less than 6 mm in diameter have a high thrombosis rate, with poor long-term patency at 5 years (40-50%). Porcine small-intestinal submucosa has been used to construct vascular conduits. It is an acellular collagen matrix that is remodeled into host artery, with good resistance to infection.

Bovine pericardium has been used to fabricate conduits for vascular reconstruction. It has a uniform thickness, is easy to handle and suture, has a low thrombogenicity, and grows adequately without causing unwarranted dilation.

Sophisticated visualization systems and miniaturized multiarticulated instruments for robotic manipulation in small spaces have proved their usefulness by enabling procedures that were technically too demanding to perform in the past. These systems permit three-dimensional visualization, allow for precise dissection, and are safe and effective. However, some authors still believe that these systems and instruments are too expensive and cumbersome. Additional research is needed to assess their true usefulness in this type of surgery.

The vast majority of the data regarding vascular reconstructive surgery in pediatric patients comes from experience with right ventricular outflow obstruction, midaortic syndrome, renal-artery occlusive disease and hypertension, iatrogenic vascular trauma, and access for hemodialysis. This article addresses the options and rationale behind choosing vascular reconstruction conduits in the pediatric population, as related to the most common circumstances requiring intervention. This article is limited to a discussion of conduits used for reconstructing the renal artery and aorta, trauma, and dialysis access in pediatric patients.

Every year, 3-5 per 1 million children develop chronic renal failure in the United States. Approximately 70% of these patients require temporary hemodialysis, and 23% of these patients require long-term hemodialysis. Dialysis is intended to be temporary; the ultimate goal is renal transplantation. However, this approach is not always rapidly feasible. Hence, many of the techniques developed for hemodialysis in adult patients have been adapted to the pediatric population. Because of the small diameters and low flow rates in the arteries of children, success in maintaining long-term access for hemodialysis can be challenging.^[3]

Indwelling central venous dialysis catheters are currently used extensively; however, problems such as kinking, infection, thrombosis, and vascular stenosis have plagued central venous dialysis catheter use in children and adults. Children encounter the same issues as adults when arteriovenous access is being considered. The surgeon must consider the location, protection, maintenance, vascular immaturity, vascular size, and patient factors (eg, the potential inability of children to protect or care for their access sites).

Creation of an arteriovenous fistula is the method of choice in children who weigh more than 30 kg; however, a vein of adequate caliber must be used. Fistulae are usually placed in radiocephalic or brachiocephalic positions. Interrupted sutures and magnification loupes are suggested during construction.

In 1980, Applebaum et al reported satisfactory use of e-PTFE bridge fistulae in children.^[4] Such expanded polytetrafluoroethylene (e-PTFE) grafts have been used in children who weigh as little as 3.8 kg with long-term patency rates of 88%. In 1994, Lumsden et al reported a mean patency of 11 months with e-PTFE bridge grafts; however, grafts placed in children who weighed less than 30 kg demonstrated poor function.^[5] In addition, they found that patency time was lower for femoral e-PTFE grafts when compared with upper-extremity grafts. This result was attributed to rapidly developing outflow stenosis, thrombosis, and graft infection.

Dialysis access in children remains a persistent problem. Each conduit has limitations, which are usually related to durability. Arteriovenous fistulae are still considered the primary choice when feasible. Otherwise, an e-PTFE bridge graft in the upper extremity is a good alternative. Polytetrafluoroethylene (PTFE) grafts placed in the groin may also be considered. However, because of consistently poor patency rates, early use of PTFE grafts is discouraged in this region. When traditional methods are exhausted, nonconventional conduits (eg, free saphenous vein) have been used.

Fibromuscular dysplasia (FMD) and middle aortic syndrome (MAS) are the most common causes of surgically remediable hypertension in young patients, and a considerable amount of research has focused on surgical intervention in this group. Takayasu arteritis (see images below) or temporal arteritis, neurofibromatosis, retroperitoneal fibrosis, mucopolysaccharidosis, and Williams syndrome are other etiologic factors. Developmental anomaly in the fusion and maturation of the paired embryonic dorsal aortas are mentioned as congenital causes for MAS. Patients with aortic obstruction, as observed with coarctation or MAS, typically present with severe hypertension, weak or absent femoral pulses, and abdominal bruit. MAS results from diffuse narrowing of the distal thoracic and abdominal aorta, commonly involving the visceral and renal arteries.

Surgical treatment of patients with MAS requires an individualized approach. The proper procedure and conduit must be selected with future growth and durability in mind. The timing for reconstruction of the aorta, renal and visceral arteries in children, must be carefully considered in each patient. Timing for surgery depends on the severity of the hypertension and patient's age. If possible, postponing surgery until the patient has completed growth is considered best. However, in patients with severe hypertension with high risk for end-organ damage, surgical intervention should not be delayed.

Regardless of the patient's age, surgical relief of the aortic constriction can be accomplished with patch angioplasty, bypass, or both. Patch aortoplasty with polyethylene terephthalate (Dacron) is recommended for young patients who are still growing. This allows for thoracoabdominal bypass at an older age when growth is completed. Dacron patches have a high success rate. Aortic cryopreserved homografts have also been used with success, but further research is needed to establish their usefulness and durability.

Despite the growing amount of literature supporting early one-stage operations with arterial autografts, Panayiotopoulos et al (1996) suggested that children with MAS who are younger than 6 years should be medically treated with every effort to avoid surgical intervention.^[6] If the patient cannot be medically controlled and surgery must be performed earlier, splenorenal or hepatorenal bypass (if the celiac axis is normal) is adequate. Otherwise, they advocate thoracoabdominal-to-infrarenal aortic bypass with reconstruction of the renal artery, with or without renal autotransplantation. Panayiotopoulos et al suggested these recommendations with the accepted understanding that further surgery may be necessary as the patient outgrows the graft.

Renal-artery occlusive disease is another curable cause of pediatric hypertension. Renovascular problems account for as many as 20% of cases of severe hypertension observed in tertiary-care pediatric centers. Renal revascularization is the preferred method for the management of this disease. Debate continues regarding the appropriate time for surgery because the potential for growth and second operations versus the progression of disease to the point of permanent end-organ damage must be weighed. When children and young adults with renal occlusive disease are treated, therapeutic options include medical care, balloon-dilation angioplasty, and surgical reconstruction or bypass.

Percutaneous transluminal renal angioplasty (PTRA) has been described. It is technically feasible and usually provides worthwhile clinical improvement; however, experience in the pediatric population is still limited. Several options are available for aortorenal arterial bypass. Suggested bypass conduits include an autogenous hypogastric artery, saphenous vein, or prosthetic grafts. Patch angioplasty, splenorenal bypass, and renal autotransplantation have also been used.

In general, prosthetic grafts for renal artery reconstruction have been associated with poor long-term outcomes. High intimal proliferative response leading to stenosis has been reported. In addition, prosthetic grafts do not grow, and lead to recurrence of hypertension later in life. Prosthetic conduits also pose a lifelong risk of infection. As such, prosthetic conduits are not recommended as the first choice for reconstruction of the renal artery in young patients.

The most extensive experience with surgical management of pediatric renovascular hypertension comes from the University of Michigan. In 1995, Stanley et al published their experience with 57 pediatric patients aged 10 months to 18 years.^[7] They strongly disagreed with balloon dilation to manage renovascular occlusive disease with or without MAS and favored a one-stage surgical approach.

Stanley et al evaluated several conduits for the aortorenal reconstructive procedures. Saphenous vein and hypogastric artery autografts resulted in the highest success rates. However, they noted significant dilatation of the vein grafts in the pediatric population, particularly those receiving aortorenal bypass grafts. They reported a 20% incidence of aneurysm formation when the vein was used for renal-artery bypass or as replacement grafts in children. Hypogastric arterial conduits are relatively resistant to dilation and have demonstrated improved durability. For ostial lesions, they favor direct reimplantation of the renal artery into the aorta or a branch artery.

Messina et al (1986) also reported that arterial autografts are preferred in young patients who are still growing and have small vessels, and advocate prosthetic grafts for patients who have reached maturity.^[8]

In 1978, Novick et al found that splenorenal bypasses had poor durability; they frequently became kinked and showed a high thrombotic rate.^[9] In addition, disease could progress in the celiac artery, leading to compromised flow and requiring further intervention. In situ aortic or iliac-inflow renal autotransplantation are other options to manage renovascular hypertension. In situ aortic reimplantation is useful in patients with short orificial stenoses. Autotransplantation into the iliac fossa may be useful after complex ex vivo reconstruction of renal-branch lesions. Renal autotransplantation is also recommended in patients with MAS who are too small for definitive surgery but whose renal function is in serious jeopardy.

Over the years, surgery for pediatric renovascular hypertension has changed as experience revealed shortcomings with saphenous vein grafts, splenorenal bypass, and prosthetic conduits in young patients. The hypogastric artery has the greatest durability and has become the preferred conduit in pediatric renal artery bypass. Renal autotransplantation has also been successful in selected patients. Regardless of the conduit, comprehensive patient follow-up is essential.

Various approaches have been suggested for recurrent aortic narrowing after a proximal aortic coarctation or interrupted aortic arch is repaired. Options include percutaneous balloon dilatation with or without stents, resection of the narrowed segment with direct anastomosis, patch aortoplasty with polyethylene terephthalate (Dacron), interposition tube grafts, or extra-anatomic grafts between the ascending and descending aorta.

Kanter et al (2000) studied extra-anatomic bypass in depth.^[10] They stressed that surgery is not the first option, and bypass should not be a first choice. They examined 19 pediatric patients aged two months to 18 years. Fifteen had coarctation with hypoplastic arch, 3 had an interrupted arch, and 1 had diffuse aortic hypoplasia. Kanter et al preferred resection of the affected area with primary anastomosis or patch aortoplasty. However, they advocate extra-anatomic bypass in selected patients. Patients who are unlikely to have success with traditional techniques, such as those with underdeveloped collateral vessels, and patients who need concomitant cardiac repair may require bypass. The preferred technique for extra-anatomic bypass involves sternolaparotomy with extra-anatomic bypass from the ascending to the descending aorta by using a Dacron conduit.

Other extra-anatomic bypass procedures performed using a multitude of incisions and repeated left thoracotomy have been described with success. In some children, regardless of the incision, extra-anatomic Dacron bypass grafting to manage recurrent aortic obstruction is a feasible option in cases of complex aortic reconstruction.

Aortic aneurysmal disease is exceedingly rare in children.^[11] Thoracic and thoracoabdominal aneurysms have been encountered and treated with some success. In 1996, Hashimoto et al successfully reconstructed a thoracoabdominal aneurysm with a surplus polyethylene terephthalate (Dacron) graft.^[12] Surplus grafts are designed to eventually straighten with the growth of the child. In 1998, Clark et al repaired an abdominal aortic aneurysm (AAA) and an injury to the left common femoral artery in an 11-year-old child with aneurysmorrhaphy and reversed saphenous vein interposition.^[13] During 25-year follow-up, normal caliber was demonstrated in the distal aorta and only slight dilatation was observed in the left femoral vein graft.

Vascular injuries in infants and children are rare and usually iatrogenic and require a substantially different approach from that used to manage similar injuries in older children and adults. Many injuries, particularly injuries in infants, must be approached with the understanding that surgical intervention may not be fully successful because of the technical difficulties related to small vessels and to vascular spasm. Also, peripheral vascular injuries that may result in limb loss in adults rarely do so in infants or children. Children have a particular propensity to rapidly develop collateral vessels, which often allows for limb preservation without surgical intervention. However, the development of an adequate collateral circulation for limb preservation is not necessarily sufficient to ensure normal patterns of limb growth. This quandary has created controversy surrounding the indications for surgical intervention.

In 1983, Flanigan et al reviewed iatrogenic pediatric vascular injuries.^[14] They reported a 23% incidence of leg length discrepancy after nonoperative treatment, compared with a 9% incidence when the injuries were aggressively treated. On the contrary, Klein et al (1982) supported nonoperative management for iatrogenic vascular injuries in small children, noting poor surgical results.^[15] They advocated close observation with a possible delay in surgical reconstruction until later in life. In 1981, Smith and Green asserted that good surgical outcomes occurred least often in children younger than 2 years.^[16]

The timing of operation has also been a topic for debate. Whitehouse et al (1976) evaluated 4 children who underwent late operations and reported no improvement in leg-length discrepancy.^[17] Klein et al (1982) reported three late operations with equalization of leg lengths after surgery.^[15] In 2001, Cardneau et al presented a series of 14 children with saphenous-vein bypasses who underwent lower-extremity revascularization at a mean time of 5.7 years after the initial ischemic insult.^[18] Five children had limb-length discrepancy, which markedly improved after late revascularization. Although the timing of surgery remains controversial, most agree that surgical reconstruction is indicated in all but the youngest patients.

Meagher et al (1979) conducted a 20-year retrospective evaluation of vascular trauma in 53 infants and children with blunt or penetrating vascular injuries.^[19] The most common sites of injury were the brachial and superficial femoral arteries and the inferior vena cava. The authors concluded that, when surgical management is indicated, primary anastomosis with interrupted suture should be attempted. When this is not feasible and a graft is needed, they recommend using the internal iliac artery in small patients and the saphenous vein in larger children.

To accommodate for growth, the vessel used should be spatulated to create an anastomotic lumen of 3 times the diameter of the vessel in question, and the anastomosis should be performed with interrupted suture by using the principles of meticulous microvascular arterial reconstruction. Polyethylene terephthalate (Dacron) or expanded polytetrafluoroethylene (e-PTFE) conduits should be reserved to repair larger arteries in children whose arteries are similar in size to those of adults.

In 2006, Lazarides et al reported 23 children aged 13 years or younger who had arterial extremity trauma.^[20] They concluded that school-aged children (>6 y) can safely undergo surgical repair but that neonates, infants, and preschool children, are best treated nonoperatively if they have an ischemic but nonthreatened extremity. As long as a distal Doppler signal was present, limb loss was rare. Patients treated nonoperatively received systemic heparin, and limb-length discrepancy was noted in only one patient. They reported an 87% limb-salvage rate with this approach.

Aortic injury usually requires resection of the injured or pseudoaneurysmal segment and reconstruction with a synthetic graft. The size of the child obviously presents a dilemma because the conduit does not grow with the child's aorta.

No single therapeutic strategy can be applied to all young patients with vascular injury. In a child of any age with a severely ischemic extremity, surgery remains the first choice. However, the surgeon must recognize that good outcomes may not be forthcoming in small patients, and in patients who require complex reconstructions, unless the surgeon is experienced in using advanced microvascular surgical techniques.

Arterial reconstructions in the pediatric population pose notable challenges to cardiac, vascular, oncologic, and pediatric surgeons. Small vessels, spasm, and the potential for growth over time should be considered. Moreover, the surgeon must be familiar with microvascular techniques, which require tremendous discipline and skill.

The choice of conduit for vascular reconstruction in the pediatric patient depends on the patient's specific circumstance. No single therapeutic strategy can be applied to all children. The patient's age, developmental status, and indications for reconstruction dictate the choice of conduit. Further clarification of the alternatives in young patients is still required. Research is underway on tissue-engineered conduits to decrease thrombogenicity and to investigate prosthetic material lined with endothelial cells or heparin. In the future, stem cells derived from bone marrow or embryonic stem cells might be used to regenerate vascular structures, probably over a synthetic absorbable conduit bed. The reconstruction of complex vascular structures by using computer-aided design may be used to fabricate custom-made tissue-engineered replacement conduits in the future.

Shinoka et al from Japan has been successful in placing tissue-engineered grafts in humans, using autologous bone marrow to seed a dissolvable prosthetic graft; however, these grafts have been used only for venous replacement in children. The technique of vascular tissue engineering, in combination with stem cell research, may hold the key for the creation of a practical and successful small diameter prosthetic graft.

Endoscopic saphenous vein harvest techniques facilitate their use in lower extremity limb salvage procedures. Vascular surgeons should become familiar with these techniques to minimize vein harvest wound complications and extend the options for limb salvage conduits, including use of both the ipsilateral and contralateral saphenous vein and the short saphenous vein. Meticulous hemostasis within the tunnel after endoscopic conduit harvest and avoidance of postoperative anticoagulation should help to prevent postoperative hematoma formation and early graft occlusion.

Widespread acceptance of minimally invasive intraluminal bypass surgery in pediatric patients will increase as flexible, small-diameter grafts, and low-profile insertion systems become available. In adults, these designs have already markedly reduced perioperative morbidity, shortened hospital stays, and hastened recovery.

In robotic surgery, future advances in reducing the size of instruments, improving image guiding systems, and incorporating tactile feedback, may expand the application of this technology to treat patients with cardiovascular disease.