The gastroepiploic artery (GEA) was introduced as a conduit for coronary artery bypass grafting in 1987 and has been applied as an additional conduit to internal thoracic arteries (ITAs) for myocardial revascularization with acceptable results.1–3 We have been using the GEA as the 3rd graft mainly for bypassing the right coronary artery system. In combination with both ITAs, it is possible to achieve complete revascularization with three in situ arterial conduits in most patients.1–9 In our institute, we began harvesting the GEA using the skeletonized technique in November 2001 and also began performing off-pump coronary artery bypass grafting (OPCAB) at the same time. We reported our experience with skeletonized GEA grafting in OPCAB, and the use of the skeletonized GEA graft in OPCAB surgery is considered both safe and effective with acceptable early clinical and angiographic outcomes.10
Recently, skeletonization of the ITA has attracted attention as an alternative harvesting technique that increases the length and caliber size of the graft compared with pedicled grafts,11–14 and it has also been applied for radial artery (RA) harvesting.15,16 This technique seems suitable for harvesting GEA grafts, increasing luminal caliber size, and reducing the possibility of flow competition. In previous studies of skeletonized ITA, there was no evidence of damage to the conduit when harvested with ultrasonic dissection (Harmonic Scalpel) or with bipolar cauterization.17–19 However, there have been only a few reports regarding skeletonization of GEA grafts.20–22
In the early days of coronary surgery, the endothelium was considered to be a passive barrier between blood and tissue, but this view changed with the observation that acetylcholine-induced arterial dilatation required the presence of endothelium.23 This was followed by the discovery of endothelium-derived relaxing factor, which was later shown to be nitric oxide (NO).24 The biologic functions of NO include actions as a vasodilator, neurotransmitter, cytotoxic agent, inhibitor of platelet aggregation, and roles in smooth muscle proliferation.25 NO released by vascular endothelial cells is a key controller molecule of vascular functions. Therefore, several indirect techniques have been used to evaluate the endothelial functions or NO bioavailability. However, there is a growing body of evidence suggesting limitations of the indirect methods for evaluation of NO bioavailability.
In this study, we compared the influence of GEA skeletonization using an ultrasonically activated device (USAD) with that using an electrosurgical unit from the viewpoint of real-time NO production by chemiluminescent assay. We also compared the inflammatory cytokine TNF-α in both groups by enzyme-linked immunosorbent assay. Furthermore, in excised specimens, we examined the expression of nitric oxide synthase (NOS) immunohistologically.
MATERIALS AND METHODS
The porcine stomach is anatomically and physiologically similar to that of humans. Fourteen healthy pigs weighing from 26.0 to 35.0 kg (average weight 29.8 ± 3.8 kg) were used in this study. All animals received humane care in compliance with the Principles of Laboratory Animal Care formulated by the National Society for Medical Research and the Guide for the Care and Use of Laboratory Animals prepared by the Institute of Laboratory Animal Resources and published by the National Institutes of Health (NIH Publication No. 86-23, revised 1985).
Pigs were divided randomly into the following two groups. In the first group (group USAD, n = 7), the GEA were harvested with an USAD (Harmonic Scalpel; Ethicon Endo-Surgery, CVG, Cincinnati, OH). In the second group (group E, n = 7), electrocautery (Force2 Electrosurgical Generator; Valleylab, Boulder, CO) was used to harvest the GEA.
Anesthesia and Surgical Procedures
Anesthesia was induced by intramuscular administration of ketamine hydrochloride (20 mg/kg). After tracheotomy, a cuffed endotracheal tube was inserted, and ventilation was performed with a volume-regulated ventilator (KMA-1300IIS; Acoma, Tokyo, Japan). Then, muscle relaxation was achieved by peripheral intravenous administration of pancuronium (0.1 mg/kg). Anesthesia was maintained with 1% halothane. A 22G catheter was inserted into the right carotid artery to achieve continuous arterial pressure and another catheter was also inserted into the right internal carotid vein. After administration of 1.5 mg/kg of heparin via the right internal carotid vein, a skin incision was made about 5-cm caudal from the xiphoid process. Following laparotomy, the stomach, GEA, and omentum were exposed and the omental artery at the middle portion of the GEA was ligated and divided. Using a harmonic scalpel (curved shears, level 2–3) or electrocautery, 1 cm from the GEA trunk, the omental and gastric branches including fatty tissue were separated. The GEA was harvested as a wide pedicle including the artery, satellite vein, and fat tissue before skeletonization. Then, the pedicled GEA was extracted from the intraabdominal space to facilitate further manipulation. The GEA was then skeletonized using the same apparatus. In group USAD, the tip of the harmonic scalpel was exchanged for a 5-mm dissecting hook, the top of which was applied gently to the GEA trunk and the whole surrounding tissues were divided from the GEA. The distal end of GEA was cut and clamped with bulldog forceps while performing skeletonization. Blood samples of 20 mL each were obtained at the following three times from the internal jugular vein and distal end of the GEA: (1) preskeletonization, (2) pedicle, and (3) postskeletonization. Whole blood samples were centrifuged at 3000 rpm for 15 minutes and separated plasma was stored at −80°C. The frozen plasma was defrosted and used for measurement of plasma levels of NO and TNF-α. After skeletonization, the GEA was excised at the root part, the specimens were fixed with 10% buffered formalin solution, and then made into paraffin blocks.
Plasma NO level was measured by chemiluminescence assay using a Sievers NO Analyzer (280NOA; Sievers, Boulder, CO). Ozone-based chemiluminescent assay has been shown to be highly sensitive for determination of nanomolar quantities of NO and related species in biologic fluids. NO generated in the living body reacts rapidly with molecular oxygen to form nitrite (NO2 −), and with oxyhemoglobin and superoxide anions (O2 −) to form nitrate (NO3 −). Therefore, the nitrate/nitrite concentration in plasma was used as an indicator of NO synthesis. The chemiluminescence assay efficiently reduces these NO metabolites to their elements in NO using a reducing agent, and by gas phase chemiluminescent reaction between NO and ozone, emission from electronically excited NO2 is in the red and near-infrared region of the spectrum, and can be detected by a thermoelectrically cooled, red-sensitive photomultiplier tube. This method has high sensitivity and reflects the amount of NO discharged into the lumen of the blood vessels in real-time. Plasma NO level was expressed in the form NOx (=NO2 + NO3) μmol/L.
Enzyme-Linked Immunosorbent Assay for TNF-α
Plasma TNF-α levels were determined by enzyme-linked immunosorbent assay according to the Griess method using a Swine TNF-α ELISA Kit (BioSource International, Camarillo, CA). Absorbance at 450 nm was measured with an ImmunoReader NJ-2000 (InterMed Japan, Tokyo, Japan). Antigen levels were determined from standard curves and protein levels were assayed using a BCA Protein Assay Kit (Pierce, Rockford, IL).
Immunohistochemical Staining of eNOS and iNOS
In biologic systems, NO is produced by the enzymatic oxidation of arginine. Three isoforms of the enzyme NOS have been identified in many cell types: endothelial NOS (eNOS), neural NOS (nNOS), and inducible NOS (iNOS). In this study, immunohistochemical staining for eNOS and iNOS was performed using paraffin sections 4-μm thick. After deparaffinization of the sections, epitope unmasking was performed by incubation in the presence of 0.5 mg/mL protease (p-4798; Sigma, St Louis, MO) for 15 minutes at room temperature. Intrinsic peroxidase activity was inhibited by addition of 3% hydrogen peroxide, and nonspecific binding was blocked with normal goat serum. Rabbit polyclonal antibodies against either human iNOS (N-20, SC-651; Santa Cruz Biotechnology, Santa Cruz, CA) or human eNOS (C-20, SC-654; Santa Cruz) were used as the primary antibodies, which were diluted 1:100 and incubated with the tissue sections for 24 hours at 4°C. The secondary antibody, an antimouse/antirabbit labeled polymer, prepared by combining acid polymers with peroxidase and goat antimouse immunoglobulin (Ig) and antirabbit Ig reduced to Fab’ fragments (Nichirei, Tokyo, Japan), was incubated with the tissue sections for 30 minutes at room temperature. Sections were then stained with 0.2 mg/mL of 3,3′-diaminobenzidine tetrahydrochloride for 10 minutes at room temperature. Between each step, the sections were washed with distilled water or 10 mmol/L sodium PBS (pH 7.2). The sections were then counterstained with hematoxylin.
All results are presented as the means ± standard deviation. Plasma NOx and TNF-α levels were analyzed by doubly multivariate repeated measures analysis of variance followed by Dunnett pairwise multiple comparison t test. P < 0.05 was considered statistically significant. Statistical analyses were performed using SPSS Advanced Models 10.0 (SPSS Japan, Tokyo, Japan).
The average time from preskeletonization to pedicle was 10.2 ± 2.3 minutes in group E and 11.7 ± 2.9 minutes in group USAD. The average time from preskeletonization to postskeletonization was 24.4 ± 5.6 minutes in group E and 28.9 ± 6.1 minutes in group USAD. Skeletonized GEA grafts were 2.8 ± 1.5 cm longer than pedicled grafts in group USAD and 2.6 ± 1.2 cm longer than pedicled grafts in group E.
Free Flow Amount of GEA
The time-related free flow amounts of GEA in both groups are shown in Figure 1. In group USAD, the free flow amount of GEA increased from 27.1 ± 5.1 mL/min at preskeletonization to 27.9 ± 4.5 mL/min at pedicle, and 31.2 ± 5.5 mL/min at postskeletonization. The free flow amount of GEA at postskeletonization was significantly greater than that at preskeletonization (P = 0.032). In contrast, in group E, free flow amount of GEA decreased from 28.8 ± 5.3 mL/min at preskeletonization to 28.1 ± 4.5 mL/min at pedicle, and 24.5 ± 6.1 mL/min at postskeletonization. The free flow amount of GEA at postskeletonization was significantly lower than that at preskeletonization (P = 0.021).
Plasma NOx Level in the Internal Jugular Vein
Time-related plasma NOx levels in the internal jugular vein are shown in Figure 2A. In the internal jugular vein, plasma NOx level did not change significantly after skeletonization of GEA in either group. The results of multivariate analysis indicated that the patterns of change of plasma NOx level in the internal jugular vein were not significantly different between the two groups (P = 0.302).
Plasma NOx Level in GEA
Time-related plasma NOx levels in GEA are shown in Figure 2B. In group USAD, the preskeletonization basal level of plasma NOx in GEA was 25.7 ± 10.9 μmol/L, which then increased to 26.9 ± 10.5 μmol/L (pedicle) and 32.2 ± 11.1 μmol/L (postskeletonization). In group E, the preskeletonization basal level of plasma NOx in GEA was 28.9 ± 10.4 μmol/L, which then changed to 27.5 ± 8.9 μmol/L (pedicle) and 21.8 ± 8.3 μmol/L (postskeletonization). The results of multivariate analysis indicated that the patterns of change of plasma NOx level in GEA were significantly different between the two groups (P = 0.024). Post hoc multiple comparison tests were then performed to determine whose means differed in each group. In group USAD, there was no significant difference in the plasma NOx level between preskeletonization and pedicle (P = 0.701), but a significant difference was observed between preskeletonization and postskeletonization (P = 0.037).
In group E, the plasma NOx level was significantly different between preskeletonization and postskeletonization (P = 0.026) and between pedicle and postskeletonization (P = 0.043).
Enzyme-Linked Immunosorbent Assay of TNF-α
Time-related plasma TNF-α levels in the internal jugular vein (A) and GEA (B) are shown in Figure 3. In the internal jugular vein, plasma TNF-α level did not change significantly after skeletonization of GEA in either group. In GEA, The results of multivariate analysis indicated that the patterns of change in plasma TNF-α level were significantly different between both groups (P = 0.033). Post hoc multiple comparison tests in group E demonstrated that the plasma TNF-α level increased significantly from preskeletonization to postskeletonization (14.1 ± 4.1 pg/mL versus 19.2 ± 4.6 pg/mL, P = 0.029).
Immunohistochemical Staining of eNOS and iNOS
GEA endothelial cells before the surgical procedure (preskeletonization) showed dense uniform luminal endothelial staining for endothelial NO synthase (eNOS).
When skeletonized GEA grafts were sectioned and stained with an antibody to eNOS, differences were seen between both groups (Fig. 4). In group USAD, the eNOS-positive GEA endothelial cells and smooth muscle cells were observed in all seven samples. In group E, however, eNOS-positive endothelial cells were observed only in three of seven samples. Furthermore, smooth muscle cells were partially degenerated in these three samples. There was a significant difference in the length of eNOS-stained lumen calculated as a percentage of the lumen (88 ± 5 arbitrary units in group USAD versus 58 ± 9 in group E, P < 0.01). There was a significant correlation between the percentage of the length of eNOS-stained lumen of skeletonized GEA and plasma NOx level in GEA at postskeletonization (P < 0.05, paired t test). Specifically, in group E, the three grafts harvested by electrocautery that showed poor eNOS staining revealed lower plasma NOx level at postskeletonization than the other four grafts.
The inducible NO synthase (iNOS)-positive endothelial cells were not observed in preskeletonized GEA. Skeletonized GEA also showed no evidence of iNOS immunoreactivity in either group.
The biologic functions of NO include actions as a vasodilator, neurotransmitter, cytotoxic agent, inhibitor of platelet aggregation, and roles in smooth muscle proliferation.25 NO released by vascular endothelial cells is a key controller molecule of vascular functions. There have been several previous reports regarding the relation between coronary artery bypass grafts and NO. Nishioka et al26 investigated in vivo NO release in patients after coronary bypass grafting. They concluded that in vivo ITA grafts showed more endothelium-derived NO release in response to acetylcholine than saphenous vein grafts after coronary bypass grafting. Using the IMA and RA segments taken from coronary artery bypass grafting patients, He et al27 compared NO release for IMA and RA. The basal and stimulated release of NO and EDHF-mediated hyperpolarization in the IMA were significantly greater than those in the RA, and it was suggested that the lower capacity of NO release may contribute to the susceptibility of RA to perioperative vasospasm and may have an impact on long-term graft patency.
The results of this study demonstrated that plasma NOx level in GEA increased in the case of ultrasonic skeletonization, and decreased in the case of skeletonization with electrocautery. Immunohistochemical staining of eNOS revealed a significant decrease in skeletonized GEA with electrocautery, although eNOS-positive endothelial and smooth muscle cells were observed fully in skeletonized GEA with an USAD. Plasma TNF-α level did not change significantly after skeletonization in group USAD, whereas plasma TNF-α level increased from preskeletonization to postskeletonization with electrocautery. It was suggested that an inflammatory response was induced in the GEA as a result of skeletonization with electrocautery.
This is the first report concerning the relation between ultrasonic skeletonization for GEA and real-time in vivo NO production, and it was suggested that GEA skeletonization with an ultrasonic scalpel led to the production of NO in the GEA endothelial cells. Although intra-arterial NOx level was higher in group USAD after skeletonization, NOx level in systemic veins was not different between the two groups. This suggests that increased NO had local rather than systemic effects. The question of which stimulus during graft harvesting was of greatest importance for endothelial NO production remains. Mechanical stimulation, thermal stimulation, change in flow amount, and pulsatile pressure change are all possible contributing factors. The most likely reason for increased NO production in ultrasonically skeletonized GEA is by the stimulation of shear stress. NO is an unstable radical with a half-life of 3 to 5 seconds. In the rat femoral arterial model, it was shown that increased blood pressure or flow (shear stress) stimulates the formation of NO and thereby dilates the femoral artery. Such increases in endothelial NO production by increase of shear stress are known to occur within a few minutes. In this study, endothelial cells were thought to be exposed to shear stress stimulation, which caused an increase in Ca2+ within the cells. This Ca2+ rise activated and enhanced the NO production in endothelial cells. Immunohistochemically, inducible NO synthase (iNOS)-positive endothelial cells were not observed in skeletonized GEA in either group. It is known that iNOS mRNA does not seem under normal conditions, but is induced by various cytokines and bacterial toxins in cells, such as macrophages and vascular smooth muscle cells. As this takes several hours, the increase in plasma NOx level in ultrasonically skeletonized GEA observed in this study may have been due to the increase of eNOS rather than that of iNOS.
Use of an ultrasonic scalpel allows atraumatic surgical dissection and maintenance of hemostasis and is therefore gentle on the tissues.28 The ultrasonically activated scalpel relies on the mechanical propagation of sound or pressure waves from an energy source, and these waves are then conducted to an active blade element. The ultrasonic scalpel is an ultrasonic surgical instrument for cutting and coagulation of tissue, operating at a frequency of 55.500 Hz, which yields four possible effects: cutting, cavitation, coaptation, and coagulation. These four possible effects can always be used both as single functions and in any chosen modified synergistic combination. In surgical practice, all four effects are applied simultaneously or consecutively. The various effects of ultrascision in the tissues are achieved at temperatures up to a maximum of 150°C. Coaptation leads to fragmentation of proteins, and coagulation and denaturing of protein compounds. Thus, there is no burning, carbonization, or smoke formation as with electrocautery, which can only be achieved at temperatures greatly in excess of 150°C and up to 400°C.29
GEA skeletonization with electrocautery seems to be affected in some points by extreme local heat produced by focused electrical density that causes vaporization and disruption of the tissue. We found loss of endothelium evident as reduced eNOS-stained areas after skeletonization with electrocautery. Plasma TNF-α levels increased significantly from preto postskeletonization when GEA were skeletonized with electrocautery. Vascular injury and increased stress to the artery may have resulted in the increase in plasma TNF-α level in this study. The GEA is a muscular artery similar to the RA, and has many fragile branches and a large satellite vein. Rough handling during harvesting will easily induce vasospasm and intimal fracture, which can frequently lead to conduit failure by virtue of thrombosis.30 Furthermore, in contrast to the internal mammary artery and the RA, the GEA is not fixed by the connective tissue of the chest wall or the forearm, and there seems to be a greater risk of injury during skeletonization of GEA compared with the internal mammary or radial arteries. The conventional pedicled GEA preparation with electrocautery frequently causes vasospasm of the vessel, as well as inaccurate evaluation of GEA size and vessel quality. Although the skeletonization of GEA was proposed to mitigate these disadvantages,20 the method with electrocautery and hemoclips requires meticulous work and sometimes causes bleeding and carries a higher risk of vessel injury. The use of bipolar cautery is associated with lower heat generation, but it is sometimes time-consuming. In this study, we demonstrated the potential advantages of GEA skeletonization with an ultrasonic scalpel, including visual inspection of the graft, increased effective length of the graft, higher free flow capacity, and increased endothelial NO production.
A limitation to this study is that the time frame of observation was short, and it is not clear whether the effects of ultrasonic skeletonization on GEA endothelial NO production may be temporary or persistent. In future, we will prepare further investigations including the determination of whether the endothelial NO production has any effect on the clinical and angiographic results in human clinical case.
In conclusion, the skeletonized GEA with an ultrasonic scalpel showed increased effective graft length, higher free flow capacity, and increased endothelial NO production. Because of the biologic functions of NO, such as vasodilatation and inhibition of platelet aggregation, ultrasonically skeletonized GEA seems a favorable graft without graft injury.
1.Pym J, Brown P, Pearson N, et al. Right gastroepiploic-to-coronary artery bypass. The first decade of use. Circulation.
2.Suma H, Wanibuchi Y, Terada Y. The right gastroepiploic artery: clinical and angiographic mid-term results in 200 patients. J Thorac Cardiovasc Surg.
3.Hirose H, Amano A, Takanashi S, et al. Coronary artery bypass grafting using the gastroepiploic artery in 1000 patients. Ann Thorac Surg.
4.Albertini A, Lochegnies A, El Khoury G, et al. Use of the right gastoepiploic artery as a coronary artery bypass graft in 307 patients. Cardiovasc Surg.
5.Bergsma TM, Grandjean JG, Voors AV, et al. Low recurrence of angina pectoris after coronary artery bypass surgery with bilateral internal thoracic and right gastroepipolic arteries. Circulation.
6.Grandjean JG, Boonstra PW, Heyer P, et al. Arterial revascularization with the right gastroepiploic artery and internal mammary arteries in 300 patients. J Thorac Cardiovasc Surg.
7.Isomura T, Sato T, Hisatomi H, et al. Intermediate clinical results of combined gastroepiploic and internal thoracic artery bypass. Ann Thorac Surg.
8.Nishida H, Endo M, Koyanagi H, et al. Coronary artery bypass grafting with the gastroepiploic artery and evaluation of flow with transcutaneous Doppler echocardiography. J Thorac Cardiovasc Surg.
9.Jegaden O, Eker A, Montagna P, et al. Techinical aspects and late functional results of the gastroepiploic bypass grafting (400 cases). Eur J Cardiothorac Surg.
10.kamiya H, Watanabe G, Takemura H, et al. Skeletonization of gastroepiploic artery graft in off-pump coronary artery bypass grafting: early clinical and angiographic assessment. Ann Thorac Surg.
11.Calafiore AM, Vitolla G, Iaco AL, et al. Bilateral internal mammary artery grafting: midterm results of pedicled versus skeletonized conduits. Ann Thorac Surg.
12.Higami T, Yamashita T, Nohara H, et al. Early results of coronary grafting using ultrasonically skeletonized internal thoracic arteries. Ann Thorac Surg.
13.Kim KB, Cho KR, Chang WI, et al. Bilateral skeletonized internal thoracic artery graftings in off-pump coronary artery bypass: early result of Y versus in situ grafts. Ann Thorac Surg.
14.Gauino M, Trani C, Glieca F, et al. Early vasoreactive profile of skeletonized versus pedicled internal thoracic artery grafts. J Thorac Cardiovasc Surg.
15.Taggart DP, Mathur MN, Ahmad I. Skeletonization of the radial artery: advantages over the pedicled technique. Ann Thorac Surg.
16.Amano A, Takahashi A, Hirose H. Skeletonized radial artery grafting: improved angiographic results. Ann Thorac Surg.
17.Higami T, Maruo A, Yamashita T, et al. Histological and physiologic evaluation of skeletonized internal thoracic artery harvesting with an ultrasonic scalpel. J Thorac Cardiovasc Surg.
18.Sasajima T, Wu MH, Shi Q, et al. Effect of skeletonizing dissection on the internal thoracic artery. Ann Thorac Surg.
19.Yoshida H, Wu MH, Kouchi Y, et al. Comparison of the effect of monopolar and bipolar cauterization on skeletonized, dissected internal thoracic arteries. J Thorac Cardiovasc Surg.
20.Gagliardotto P, Costa P, Lazreg M, et al. Skeletonized right gastroepiploic artery usesd for coronary artery bypass grafting. Ann Thorac Surg.
21.Asai T, Tabata S. Skeletonization of the right gastroepiploic artery using an ultrasonic scalpel. Ann Thorac Surg.
22.Ochi M, Bessho R, Saji Y, et al. Sequential grafting of the right gastroepiploic artery in coronary artery bypass surgery. Ann Thorac Surg.
23.Furchgott RF, Zawadski JV. The obligatory role of the endotherial cells and the relaxation of artery smooth muscle cells by acetylcholine. Nature.
24.Palmer RM, Ferrige AG, Moncada S. Nitric oxide release accounts for the biological activity of endothelium-derived relaxing factor. Nature.
25.Garg UC, Hassid V. Nitric oxide-generating vasodilators and 8-bromo cyclic guanosine monophosphate inhibit mitogenesis and proliferation of cultured rat vascular smooth muscle cells. J Clin Invest.
26.Nishioka H, Kitamura S, Kameda Y, et al. Difference in acetylcholine-induced nitric oxide release of arterial and venous grafts in patients after coronary bypass operations. J Thorac Cardiovasc Surg.
27.He GW, Liu ZG. Comparison of nitric oxide release and endothelium-derived hyperpolarizing factor-mediated hyperpolarization between human radial and internal mamaary arteries. Circulation.
2001;104(12 Suppl 1):I344–I349.
28.Amaral JF. Laparoscopic application of an ultrasonically activated scalpel. Gastrointest Endosc Clin North Am.
29.Amaral JF. Depth of thermal injury: ultrasonically activated scalpel vs electrosurgery. Surg Endosc.
30.Chardigny C, Jebara VA, Acar C, et al. Vasoreactivity of the radial artery. Comparison with the internal mammary and gastroepiploic arteries with implications for coronary artery surgery. Circulation.
1993;88(5 Part 2):II115–II127.
This is a well-performed experimental study examining the difference between right gastroepiploic artery skeletonization using an ultrasonically activated device compared with an electrosurgical unit, from the viewpoint of graft length, flow capacity, and nitric oxide production. The authors showed that ultrasonically skeletonized gastroepiploic arteries had increased graft length, higher free flow capacity, and increased endothelial nitric oxide production compared with those harvested with an electrosurgical unit. This is similar to results in the literature looking at radial artery harvest. This study is another piece of evidence suggesting that the ultrasonic scalpel allows for less traumatic surgical dissection and is gentler on tissues than the electrocautery unit. A principal limitation of this study is that the time frame of observation was short, and it is unclear whether the effect of ultrasonic skeletonization on nitric oxide production was temporary or persistent.