Nitric Oxide: What a Vascular Surgeon Needs to Know

Daniel A. Popowich; Vinit Varu; Melina R. Kibbe


Vascular. 2007;15(6):324-335. 

In This Article

Local Delivery

The local delivery of drugs allows for the administration of the maximally effective dose of a drug without the unwanted systemic side effects. Because the target vessels are easily accessible during most vascular procedures, a local pharmacologic approach to administer a drug during the intervention can be easily performed.

Because of the mixed results seen in the use of systemic l-arginine supplementation, Kown and colleagues examined the effects of local l-arginine administration on neointimal hyperplasia after carotid artery bypass in rabbits.[39] Internal jugular veins were explanted and then treated ex vivo with the intraluminal administration of phosphate-buffered saline or l-arginine polymers with different lengths (R5, R7, or R9) and increasing concentrations (10 or 100 µmol/L) for 15 minutes. The grafts were then reimplanted as bypass conduits to the contralateral carotid artery. The animals were sacrificed 28 days later, and the arteries were examined for neointimal hyperplasia. The l-arginine polymer-treated group (tagged with biotin) demonstrated a dose- and length-dependent uptake into intimal and medial cells of treated vessels, and this correlated with an increase in NO levels and reduced neointimal hyperplasia.

Later studies sought to find out if the local administration of l-arginine would have the same effect in humans. Suzuki and colleagues performed a prospective, randomized, single-center clinical trial.[40] The study population consisted of patients with symptomatic ischemic heart disease who were undergoing coronary artery stent placement. After stent deployment, l-arginine (600 mg/6 mL) or saline (6 mL) was locally delivered via a catheter over 15 minutes. The patients were followed with serial angiography and intravascular ultrasonography to assess for neointimal thickness for up to 6 months. The authors found that in the l-arginine-treated groups, there was slightly less neointimal volume, but this was not statistically significant.

Because it was not known if the addition of l-arginine actually translated to increased NO production, several studies have focused on the addition of NO donors directly to the site of injury. An example of this method was performed by Harnek and colleagues.[41] They examined the ability of linsidomine to reduce neointimal hyperplasia after coronary angioplasty in pigs. After overdilation of a coronary artery by a balloon catheter, the authors locally infused linsidomine (7.2 ± 0.6 mL 10-4 M) for 10 minutes. The distal arterial segments served as controls. When evaluating for lumen loss radiographically, the linsidomine-treated groups were found to exhibit much less lumen loss compared with controls. Critics of this study point out that the evaluation of neointimal hyperplasia was performed radiographically, which could be subjectively biased. Furthermore, infusing the drug through a catheter for an extended period of time during the procedure to achieve an effect is not clinically feasible. Because of this, other studies have aimed to develop a clinically applicable approach to deliver NO locally to the site of injury.

A study by Kalinowski and colleagues examined the ability of three locally delivered drugs to the area of arterial injury to decrease neointimal hyperplasia compared with untreated controls.[38] The three drugs used were l-arginine, molsidomine, and r-hirudin, a direct thrombin inhibitor. The authors took cholesterol-fed New Zealand rabbits and balloon injured both common iliac arteries to induce stenosis. Four weeks later, one artery was dilated and received local application of one of the three drugs via a channeled balloon catheter, and the contralateral artery was injected with 0.9% saline. After 6 weeks, the vessels were harvested and studied. The authors found that there was a 59% reduction in neointimal hyperplasia in the l-arginine group, a 43% reduction in the molsidomine group, and a 20% reduction in the r-hirudin group. Whereas the l-arginine and molsidomine groups reached statistical significance, the r-hirudin group did not.

Although there are promising findings in these studies, the process by which the vessels were exposed to l-arginine or other drugs continues to be tedious and involved. Recent research in the field of creating NO-donating drugs has focused on their incorporation into polymers, films, and gels that would more reliably release NO and be easier to administer. Within this field of research, two of the most widely studied classes of NO-donating drugs are diazeniumdiolates and S-nitrosothiols.

Diazeniumdiolates are NO donors, which are formed by the reaction of secondary amine structures with two moles of NO under high pressure (see Table 1 ).[42,43,44] They generate bioactive NO in physiologic fluids (37°, pH 7.4) spontaneously (ie, without metabolism or redox activation), with reliable half-lives ranging from 2 seconds to many weeks depending on the ionic structure.[45] The major advantage of using diazeniumdiolates as NO donors is that they are stable as solids but can be triggered to release NO at controlled rates on hydrolysis or other chemical reactions. Given that most of these compounds generate two moles of NO on activation, the exact amount of NO delivered can be calculated by knowing the amount of diazeniumdiolate that was administered.[46] This quality has made them very useful research tools as reliable and stable sources of NO to probe into its protective or toxicologic roles. At present, diazeniumdiolates have not been used clinically, although they have been evaluated in experimental animal models of cardiovascular disease.

The second class of NO-donating drugs that are currently being widely studied is S-nitrosothiols (see Table 1 ). S-Nitrosothiols are thought to serve as a reservoir and transporter of NO within biologic systems.[47]S-Nitrosothiols are formed by the S-nitrosation of thiols or cysteine residues. Examples of endogenous S-nitrosothiols are S-nitrosoalbumin, S-nitrosoglutathione, and S-nitrosocysteine.[48] These S-nitrosothiols are present in the circulating blood and also found within cells.[49] They release NO by three known mechanisms:[50] copper ion-mediated decomposition,[51] direct reaction with ascorbate,[52] and homeolytic cleavage of the S-NO bond by light.[53] Like diazeniumdiolates, S-nitrosothiols are not clinically used at present, but they are being extensively studied for use in vascular pathology owing to the different mechanisms by which they release NO.

One mechanism to deliver NO locally to the area of vascular injury at the time of surgery is to use hydrogels. Hydrogels allow for uniform diffusion of drugs into the arterial wall while minimizing drug losses in the bloodstream and side branches.[54] One of the first studies to create NO-releasing hydrogels for the intended use as tissue coatings to provide local and sustained NO therapy following vascular injury was done by Fulton and colleagues.[55] The effect of a single dose of locally applied S-nitroso-N-acetylpenicillamine (SNAP) in a pluronic gel on the formation of neointimal hyperplasia was examined. The jugular veins of rabbits were harvested, and their outer surface was coated with this NO-containing gel before being reimplanted as carotid bypass grafts on the contralateral side. The animals were sacrificed at 28 days, and the authors found that the SNAP-treated groups exhibited a 36% decrease in mean intimal thickness compared with controls.

Further improvements in creating NO-releasing gels and making their delivery more simple were performed by Bohl and West.[56] They created hydrogels using different NO-releasing compounds (diazeniumdiolates and S-nitrosothiols). In vitro testing showed that all of their NO-hydrogel preparations were able to inhibit VSMC proliferation and exhibited significantly less platelet adhesion compared with controls. The authors speculated that the ease of creating these hydrogels and the ease of handling them would be well suited for their application to injured arteries in the operating room to prevent thrombosis and restenosis.

One of the first in vivo animal studies that used NO-releasing hydrogels was performed by Kaul and colleagues. A polymer of polylactic-polyglycolic acid matrix (Atrigel) containing the diazeniumdiolate 1-{N-[3-Aminopropyl]-N-[4-(3-aminopropylammoniobutyl)]}diazen-1-ium-1,2-diolate (SPER)/NO was applied to the outside of balloon-injured rat ileofemoral arteries.[37] Atrigel has the unique ability to exist as a liquid below body temperature but solidifies into a viscous mass when it comes into contact with tissue at body temperature. The polymer gel enables a depot drug delivery in which the NO donor is released over several days as it biodegrades, with complete resorption in about 14 days. There were no bleeding complications in any of the rats studied. The authors found that the NO donor reduced neointimal hyperplasia by about 75% compared with controls.

To examine if it was possible to use NO-releasing hydrogels via the endovascular route, Rolland and colleagues coated an angioplasty balloon with a hydrogel containing the NO donor molsidomine.[54] Angioplasties were then performed on the iliac arteries of pigs that were fed atherogenic diets. The arteries were examined at 3 hours, 24 hours, and 3 months after treatment. Although there were no observed effects at 3 hours, at 24 hours, the treated arteries exhibited better thromboresistance and vascular cell homeostasis. At 3 months, the treated groups exhibited fewer stenotic lesions.

Masters and colleagues developed a polyethylene glycol hydrogel that was covalently modified with the NO donor S-nitrosocysteine that could release NO for approximately 24 hours.[57] After performing in vitro studies to identify the optimal concentration of the NO donor to enhance endothelial cell growth while at the same time inhibiting VSMC proliferation, they examined its effect on reducing neointimal hyperplasia formation when applied directly to the external surface of the artery in a rat carotid artery injury model. The treatment groups were found to have 75% less neointimal formation at 14 days.

The results from these initial studies with NO-releasing hydrogels appear promising for their potential use in vascular surgical procedures. One drawback of their use is that they all have to be made at the time of the surgery, which could prove to be inconvenient. Another way to deliver NO to the arterial wall is by using biopolymers that can be stably stored and, when used, release various reliable amounts of NO over extended periods of time. Frost and colleagues noted that polymeric materials that can release or generate NO locally at fluxes that are equal to or greater than the normal endothelial cell layer for extended periods of time may provide the ultimate route to solve the issues of both thrombosis and stenosis that occur after biomedical implants and/or vascular procedures.[58]

Smith and colleagues were the first to use diazeniumdiolates as NO donors in polymers and then incorporate those polymers into vascular grafts.[59] They dipped 4 mm expanded polytetrafluoroethylene (ePTFE) grafts (GORE-TEX, W.L. Gore & Associates, Elkton, MD) into a freshly prepared solution of poly-(ethylenimine) (PEI) and a cross-linking agent such that the cross-linked PEI chains became intimately interwoven with the graft. The grafts were then exposed to gaseous NO to create diazeniumdiolate groups on the PEI. These grafts were shown to produce NO for several weeks in vitro, and they exhibited decreased platelet deposition and VSMC proliferation compared with untreated grafts. These grafts were then inserted as arteriovenous shunts into baboons for 1 hour, and analysis showed that there was significantly less platelet deposition on the NO-releasing vascular grafts compared with untreated grafts. Although this technique initially looked promising, the coating process changed the architecture of the graft, which the authors speculated might create long-term biocompatibility complications secondary to compliance mismatch.

In an attempt to preserve the graft architecture, Pulfer and colleagues incorporated polymeric diazeniumdiolate PEI/NO microspheres into the pores of an ePTFE graft (GORE-TEX).[60] These grafts retained the same physical properties as control grafts even after the addition of the microspheres, with the added benefit that they released NO in vitro for greater than 150 hours. Another method of incorporating NO donors into vascular grafts without altering their mechanical properties is to coat the inside lining with a NO-eluting polymer, which would not alter the architecture of the graft material. Zhang and colleagues created diazeniumdiolated silica nanoparticles and embedded them into hydrophobic matrixes, which were used to coat the inside of tubing used for extracorporeal venovenous circuits in a rabbit model.[61] NO release was substantial, and there was much less platelet consumption and activation when compared with controls when in contact with blood for 4 hours.

Fleser and colleagues coated 5 mm polyurethane grafts (Vectra, Thoratec Corporation, Pleasanton, CA) with multiple layers of a polyvinyl chloride film containing a diazeniumdiolate.[62] In vitro studies showed that NO release was linear up to 7 days and that NO release generated after 25 days was still above that which is produced by functional endothelial cells in vivo. Twelve grafts (three uncoated grafts, four sham-coated grafts, and five NO-releasing grafts) were implanted as arteriovenous grafts into sheep connecting the common carotid artery to the ipsilateral jugular vein. The results showed that all three uncoated grafts had occluded, two of the sham-coated grafts had occluded by 3 weeks, and only one of the NO-releasing grafts had occluded. Although there was a significant reduction in surface thrombus accumulation in the NO-releasing grafts compared with sham-coated and uncoated grafts, a statistically significant improvement in patency was not observed.

Another mechanism by which to use NO-releasing polymers is to incorporate them into currently used vascular prosthetics such as stents and bypass grafts. Yoon and colleagues incorporated sodium nitroprusside (SNP) into a polyurethane polymer and coated this onto metallic coiled stents.[63] Despite showing that there was a biologic effect of NO up to 14 days after implantation via increased cGMP levels, no reduction in neointimal hyperplasia in porcine carotid arteries was observed at 28 days.

Hou and colleagues examined the effects of a NO-eluting covered stent on neointimal formation in a porcine carotid balloon injury model.[64] The interior of a self-expanding ePTFE-covered aSpire stent was coated with silicone, which contained two different concentrations of SNP. When angiograms were performed at 28 days, a 24% reduction of vessel narrowing in the NO-treated groups was observed.

Jun and colleagues incorporated a diazeniumdiolate into a biocompatible polyurethane graft (PUBD-NO) to assess the ability of this material to promote graft endothelialization while preventing thrombus formation and neoinitimal hyperplasia.[65] The mechanical properties of this modified graft were comparable to unmodified polyurethane grafts and to native tissue. The grafts released a measurable amount of NO for up to 2 months. In vitro studies showed that the grafts were able to inhibit VSMC proliferation and stimulate endothelial cell growth while at the same time inhibit platelet adhesion. The authors speculated that the properties of this graft would be ideally suited for bypass grafting and exhibit less neointimal hyperplasia than conventional bypass grafts.[66] To date, there have been no clinical studies evaluating these types of grafts for the purpose of decreasing neointimal hyperplasia after vascular procedures.

Gene therapy represents another method by which to locally increase the level of NO at the site of vascular injury. Like other local therapies, one of the novel characteristics of gene therapy-based approaches is the ability to manipulate only the bypass graft or the target vessels with little or no systemic effects.[67] Unlike the local delivery of NO-releasing drugs, which would eventually be depleted and washed away, the local delivery of NOS genes provides the ability to release NO for longer periods of time at the site of vascular injury.

The first published study using nitric oxide synthase (NOS) gene transfer for the inhibition of neointimal hyperplasia was by von der Leyen and colleagues in 1995.[68] An endothelial NOS-expressing plasmid was used to increase NO production in injured rat carotid arteries via a modified liposome. It was found that not only was endothelium-dependent relaxation restored in the treatment group, but a 70% reduction in neointimal hyperplasia was also observed at 14 days (Figure 3). This study proved that it was possible to provide sustained levels of NO production following transfection of a NOS gene, and this NO had beneficial therapeutic effects on the vasculature after injury.[69]

Figure 3.

Effect of endothelial nitric oxide synthase infection on neointima formation in balloon injured rat carotid arteries. A, Uninjured; B, injured and untreated; C, injured and infected with the control vector; D, injured and infected with endothelial nitric oxide synthase. M = media; N = neointima (x25 original magnification, hematoxylin-eosin stains). Reproduced with permission from von der Leyen HE et al.68 Copyright (1995) National Academy of Sciences, U.S.A.

Subsequently, Shears and colleagues described the preclinical evaluation of NOS gene transfer using inducible nitric oxide synthase (iNOS).[70] iNOS was chosen because it had been shown to produce much larger quantities of NO in a calcium- and agonist-independent fashion.[71] An adenoviral vector carrying the human inducible nitric oxide synthase complementary deoxyribonucleic acid (DNA) (AdiNOS) was constructed. Transfection of balloon-injured rat carotid arteries with AdiNOS resulted in a > 95% reduction in neointimal hyperplasia 2 weeks postinjury. Next, AdiNOS was evaluated in a preclinical model, namely, the porcine iliac artery balloon injury model. AdiNOS-transfected arteries were found to have 51.8% less neointimal hyperplasia 3 weeks after injury. Kibbe and colleagues investigated the ability of this same AdiNOS construct to inhibit neointimal hyperplasia in the setting of a vein bypass graft.[72] Interposition jugular vein grafts transfected with AdiNOS anastomosed to the carotid arteries of pigs were found to have less vein graft neointimal hyperplasia at 21 days compared with controls.

Further demonstrating the ability to site-specifically overexpress the NOS gene, Pfeiffer and colleagues used a balloon catheter delivery system, the Infiltrator Drug Delivery Balloon System (Infiltrator, InterVentional Technologies Europe Ltd, Lisnenan, Letterkenny, Co Donegal, Ireland), to inject iNOS into the carotid arterial walls of foxhound dogs at the proximal and distal anastomosis of ePTFE bypass grafts.[73] At 6 months, less neointimal hyperplasia was found at both the proximal and distal anastomoses. This included the prosthetic wall, suture region, and arterial wall. At the proximal anastomosis, the reduction in neointimal hyperplasia was 43%, 52%, and 81%, and at the distal anastomosis, the reduction was 40%, 47%, and 52%, respectively, for the three defined locations. The authors pointed out that this study demonstrates the ability of a single local transfection to decrease neointimal hyperplasia even at 6 months.

Using another NOS gene, West and colleauges performed jugular-carotid interposition grafts in rabbits transfected with neuronal nitric oxide synthase (nNOS) via an adenoviral vector.[74] At 3 days, NOS activity was significantly increased, and this caused a substantial reduction in adhesion molecule expression and inflammatory cell infiltrate. At 28 days, although nNOS expression was no longer present, there was a reduction in neointimal hyperplasia by almost 50% and reduced vascular superoxide production.

Because of the promising in vitro and in vivo studies, several researchers attempted to start a randomized clinical trial in humans using gene therapy to reduce neointimal hyperplasia. In December 2000, the Recombinant DNA Advisory Committee at the National Institutes of Health voted unanimously to proceed with the first phase of clinical evaluation of iNOS lipoplex-mediated gene transfer, called REGENT-1: Restenosis Gene Therapy Trial.[69] The primary objective of this multicenter, prospective, single-blind, dose escalation study was to obtain safety and tolerability information of iNOS-lipoplex gene therapy for reducing restenosis following coronary angioplasty. As of 2002, 27 patients had been enrolled overseas and the process had been determined to be safe. To date, no results have been published as it appears that this trial lost its funding and closed. On April 5, 2002, a notification was issued that the trial had been closed without enrolling any individuals in the United States.[75]

Unfortunately, despite the promising findings shown with NOS therapy, the field of gene therapy has been mottled by two widely known complications. One case occurred as the result of administering a large viral load that led to the death of a patient.[67] In addition, in France, there were at least two cases of malignancy following retroviral gene therapy.[67]


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.