Lobar torsion after pulmonary resection: presentation and outcome.

OBJECTIVE: We reviewed our experience on postoperative lobar torsion. METHODS: Between January 1972 and January 1998, 7887 patients underwent pulmonary resection at our institution. Seven (0.089%; 4 women and 3 men; median age, 68 years) patients required surgical reintervention for lobar torsion. RESULTS: The indications for pulmonary resection were non-small cell carcinoma in 5 patients, lymphoma in 1 patient, and metastatic prostate carcinoma in 1 patient. The right upper lobe was resected in 3 patients, the left lower lobe in 2 patients, and the right middle and right lower lobe in 1 patient each. Postoperative radiographs demonstrated pulmonary infiltrates and volume loss in 5 patients and complete opacification in 2 patients. The median white blood cell count was 10.6 x 10(9) cells/L (range, 9.3-14.9 x 10(9) cells/L), and the median peak temperature was 38.4 degrees C (range, 37.8 degrees C-40.2 degrees C) during the first 48 hours postoperatively. The diagnosis of lobar torsion was made a median of 10 days (range, 2-14 days) after the initial operation; 4 patients underwent completion pneumonectomy, and 3 had lobectomy. Median hospitalization was 24 days and ranged from 10 to 56 days. There were no postoperative deaths. Complications after reoperation included respiratory failure in 2 patients, atrial arrhythmia in 2 patients, and empyema, urinary tract infection, and a transient ischemic attack in 1 patient each. CONCLUSIONS: Lobar torsion represents a difficult diagnostic dilemma in the early postoperative period after pulmonary resection. A high index of suspicion is necessary to avoid a delay in treatment. Late diagnosis results in further pulmonary resection and prolonged hospitalization in the majority of cases.

Gene therapy for vein graft disease.

Applying gene therapeutics to vein graft disease requires foundational knowledge of the underlying pathophysiology. This review details a brief description of vein graft disease, examines published and unpublished data on gene transfer to veins, and reviews the genes, which have significantly altered vascular biology.

Surgical treatment of valvular heart disease in patients with acromegaly.

BACKGROUND AND AIM OF THE STUDY: Acromegaly is associated with heart disease in one-third of patients, and diastolic dysfunction may precede global systolic dysfunction. Patients with acromegalic heart disease may have valvular disease, but the role of surgery in such patients has not been established. The purpose of this study was to document the outcome of surgery in a series of these patients from one institution. METHODS: Among 951 patients with the diagnosis of acromegaly seen at our institution since 1972, 10 (eight men, two women) have undergone operation for valvular heart disease. Average patient age was 62.2 +/- 11.5 years; average body weight was 84 +/- 13 kg; average height was 178 +/- 12 cm. The mean duration of acromegaly was 15.2 +/- 12.7 years. At the time of heart surgery, seven patients had active disease, defined by elevation of growth hormone levels, while three had inactive disease. Treatment of pituitary adenomas before valvular surgery included surgical resection in three patients and external-beam radiation treatment in four. The preoperative ejection fraction was 42 +/- 19% (range: 20% to 66%). Valve lesions included aortic stenosis in four patients, aortic regurgitation in four, and mitral regurgitation in three (one patient had double valve disease). RESULTS: Valve replacement was performed in all patients with aortic disease (two bioprostheses, six mechanical), and three patients with mitral regurgitation had repair. Concomitant procedures performed in seven patients included coronary bypass (two), left ventricular aneurysmectomy (two), and ligation of the left atrial appendage, septal myectomy and defibrillator insertion (one each). Early complications included endocarditis, low cardiac output and arrhythmia in one patient each. There were no perioperative deaths. One patient underwent reoperation ten years later for a perivalvular leak. CONCLUSION: Valvular surgery can be performed safely in acromegalic patients, even those with active endocrinopathy.

Acutely sodded expanded polytetrafluoroethylene grafts produce only prostacyclin: a qualitative difference from saphenous veins.

Select subsets of patients require prosthetic graft material for revascularization. Although arterial prosthetic grafts of large caliber perform acceptably, grafts of <6 mm exhibit a high attrition rate. Microvessel endothelial sodding, a method resulting in the lining of prosthetic grafts with autologous endothelium, improves graft patency; however, aggressive antiplatelet therapy is still required, because terminating an antiplatelet regimen accelerates graft attrition. The present investigation was designed to address the acute production of vasoactive substances in microvessel endothelial cell sodded expanded polytetrafluoroethylene (ePTFE) grafts in an attempt to delineate a possible mechanism behind the continued requirement for antiplatelet therapy. Equal lengths of acutely sodded ePTFE grafts (canine falciform ligament source) and saphenous veins (SV) (canine source) were evaluated by superfusion bioassay. Basal secretion from ePTFE grafts relaxed the biodetector ring 1 +/- 3%, whereas SV relaxed the ring 10 +/- 3% (p < 0.05, ePTFE vs. SV). Relaxation with acetylcholine stimulation was 49 +/- 7% in grafts and 50 +/- 10% in veins (p = NS). Calcium ionophore stimulation produced relaxation of 37 +/- 9% from ePTFE grafts and 100 +/- 23% from SV (p < 0.05). Indomethacin added to perfusate reduced relaxations from sodded ePTFE grafts to 20.2 +/- 9.2% with acetylcholine stimulation and 12.5 +/- 4.3% with calcium ionophore (p < 0.05 vs. control); addition of N(G)-monomethyl-L-arginine (L-NMMA) had no effect on the release of vasoactive substances from ePTFE grafts. In contrast, relaxations of effluent from SV stimulated by acetylcholine and calcium ionophore were significantly attenuated with indomethacin and L-NMMA (p < 0.05 vs. control). Scanning electron microscopy demonstrated confluent endothelium in SV and a nonconfluent endothelial cell layer in grafts. Acutely sodded ePTFE grafts produce vasoactive substances that quantitatively and qualitatively differ from those produced by canine SV. The ePTFE grafts produce mainly prostanoids, whereas SV produce both nitric oxide and prostanoids. The endothelial cell isolation procedure and absence of immediate graft luminal confluence may contribute to the observed differences.

Different effects of protamine on canine coronary microvessel and conductance arteries: evidence of hyperpolarizing factor release.

BACKGROUND: Protamine administration may lead to systemic hypotension, perhaps because of vasodilatation produced by endothelial nitric oxide. This study compared release of vasoactive substances from canine coronary microvessels with that from paired conductance arteries. METHODS: Microvessels were mounted in a videoscopic no-flow system, and circumflex arteries were studied in organ chambers; both were induced to contract by endothelin-1. RESULTS: Protamine (10 to 160 micrograms/mL) produced concentration-dependent relaxation in both microvessel and conductance arteries (46% +/- 14% maximal relaxations in microvessel and 82% +/- 15% in conductance arteries, n = 10 each). Removal of the endothelium abolished this relaxation (P < .05, n = 6). Indomethacin (10(-5) mol/L) did not alter the relaxation in either group (51% +/- 10% in microvessel and 103% +/- 7% in conductance arteries, n = 6 each). NG-monomethyl-L-arginine (L-NMMA, 10(-4) mol/L) attenuated relaxation in conductance arteries (38% +/- 12%, P = .04, n = 6) but had no effect on microvessel arteries (58% +/- 10%, n = 6). Tetraethylammonium chloride (10(-3) mol/L), an inhibitor of voltage-dependent potassium channels, had no effect on conductance arteries (103% +/- 9%, n = 6) but abolished relaxation in microvessels (-25% +/- 11%, P = .03, n = 6). CONCLUSIONS: Protamine sulfate causes endothelium-dependent relaxation in microvessel and conductance arteries in the heart by different mechanisms--that is, by nitric oxide release in conductance arteries and by endothelium-derived hyperpolarizing factor (EDHF) release in microvessels. This is the first description of the release of EDHF in response to protamine administration.

Recombinant gene transfer of endothelial nitric oxide synthase augments coronary artery relaxations during hypoxia.

BACKGROUND: Coronary arteries respond to hypoxia with transient relaxations, which increases coronary blood flow, in part, by release of nitric oxide. We hypothesized that increased expression of nitric oxide synthase might further augment blood vessel relaxation during hypoxia. The present study examined the effect of adenovirus-mediated transfer of bovine endothelial nitric oxide synthase (eNOS) on hypoxia-induced transient relaxations in canine coronary arteries. METHODS AND RESULTS: Paired segments of coronary arteries were exposed to vehicle (phosphate-buffered saline with albumin) or an adenovirus encoding either E coli beta-galactosidase (Ad.CMVLacZ, viral control; 10(10) pfu/mL) or eNOS (Ad.CMVeNOS; 10(10) pfu/mL) for 2 hours at 37 degrees C. Immunohistochemistry with a monoclonal antibody specific for eNOS documented both endothelial and adventitial expression in Ad.CMVeNOS arteries, whereas vehicle and viral controls demonstrated only constitutive expression. Levels of cGMP were increased 5-fold in Ad.CMVeNOS arteries compared with controls. In arteries exposed to Ad.CMVeNOS, maximum contraction to prostaglandin F(2alpha) was reduced compared with viral controls, and this effect was eliminated by pretreatment with a competitive inhibitor of eNOS (N(G)-monomethyl-L-arginine, 10(-3) mol/L). Hypoxia-induced transient relaxation (95% N(2)-5% CO(2)) in Ad.CMVeNOS arteries (45.2+/-8.8%, n=6) was augmented compared with vehicle (26.3+/-6.0%) or viral (27.2+/-7.1%) controls. CONCLUSIONS: Adenovirus-mediated gene transfer of nitric oxide synthase reduces receptor-dependent contractions and augments hypoxia-induced relaxations in canine coronary arteries; this method of augmentation of NO production might be advantageous for reduction of coronary artery vasospasm.

The role of gene therapy for intimal hyperplasia of bypass grafts.

BACKGROUND: Proliferation of the intima is an early lesion of saphenous vein graft disease. Early patency rates of radial artery grafts are acceptable, but little is known about their risk of intimal hyperplasia. METHODS AND RESULTS: To develop a model of intimal hyperplasia, we incubated human saphenous veins, internal mammary arteries, and radial arteries (n=6, 8, and 10, respectively) in an organ culture with Roswell Park Memorial Institute 1640 (30% serum) for 0, 4, 7, 10, and 14 days. Quantitative histological studies were performed, and the average intimal-to-medial (I/M) ratio was calculated for each incubation interval. After 10 and 14 days of culture, the I/M ratio increased in the saphenous veins (P=0. 03, P=0.04 versus 0 day, respectively). No significant increase occurred in the I/M ratio in either the internal mammary or radial arteries. Next, the ability of adenoviral gene transfers to inhibit intimal hyperplasia in the saphenous veins was evaluated. Adenoviral-mediated gene transfer of nitric oxide synthase significantly reduced the I/M ratio at 14 days compared with vehicle (P=0.001) and virus (P=0.004) controls. CONCLUSIONS: The human saphenous vein has a greater propensity for intimal hyperplasia than arterial grafts; the human radial artery behaves similarly to the internal mammary artery. In the future, gene therapy may augment nitric oxide synthase, limiting vein graft disease.

A model of xenograft hyperacute rejection attenuates endothelial nitric oxide production: a mechanism for graft vasospasm?

BACKGROUND: The deposition of complement components within grafts, complement consumption, and prolongation of graft function by complement inactivation imply a pivotal role for complement in xenograft hyperacute rejection. The current investigations examined the endothelial production of vasoactive substances in pulmonary arteries during simulated hyperacute rejection. METHODS AND RESULTS: Canine pulmonary arteries were suspended in organ chambers and exposed to either autologous canine serum for 90 minutes or heterologous porcine serum for 30, 60, or 90 minutes. Following serum exposure, the vessels were allowed a one-hour equilibration in buffered crystalloid solution. Dose-response curves were obtained with acetylcholine, sodium nitroprusside, and calcium ionophore A23187 following contraction with phenylephrine (10(-6) M) in the presence of indomethacin (10(-5) M). Receptor-dependent, endothelial-dependent relaxations to acetylcholine (10(-9)-10(-4) M) were impaired with 30-, 60-, or 90-minute porcine serum exposure when compared to vessels exposed to autologous canine serum (n = 10, 7, 9, respectively; p < .05; 2-way ANOVA). Receptor-independent, endothelial-dependent relaxations to calcium inophore (10(-9)-10(-6) M) were significantly impaired at 60- and 90-minute porcine exposures only (n = 7, 8; p < .05). Endothelial-independent relaxations to sodium nitroprusside (10(-9)-10(-4) M) were not impaired with either canine or porcine serum exposure. Oxyhemoglobin (10(-6) M) abolished acetylcholine-mediated relaxations, indicating that nitric oxide was the predominant mediator. CONCLUSIONS: Simulated hyperacute xenograft rejection impairs endothelium-dependent relaxation of canine pulmonary arteries. Both basal and stimulated production of nitric oxide is impaired by heterologous serum exposure and, subsequently, complement activation. Reduced production of nitric oxide may explain, in part, the vasospasm and thrombosis of xenografts during hyperacute rejection.

Endothelial regulation of vascular contraction in radial and internal mammary arteries.

BACKGROUND: The extent to which the endothelium regulates radial artery (RA) contractions is unknown. The goals of this study were to characterize endothelium-dependent relaxations in the RA, compare these responses with those in the internal mammary artery (IMA), and, subsequently, manipulate nitric oxide production in the RA with adenovirus-mediated gene transfer. METHODS: Segments of RA and IMA from 43 patients were studied initially in organ chambers. Endothelial function was evaluated and gene transfer, was examined. RESULTS: After precontraction to 80% maximum tension with prostaglandin F2alpha, acetylcholine produced lesser relaxations in the RA (21.5%+/-5.8%) than in the IMA (66.7%+/-10.6%); human thrombin and adenosine 5'-diphosphate yielded similar results. Reduced relaxations in the RA (16.8%+/-4.2%) compared with those in the IMA (71.6%+/-11.9%) were noted with calcium ionophore. Superfusion bioassay demonstrated a similar baseline release in both arteries but a reduced stimulated production of vasoactive substances in the RA, results confirmed by cyclic guanosine monophosphate level determination. The RA produced less 6-keto-prostaglandin F1alpha than the IMA. Light microscopy demonstrated an intact endothelium in both arteries. Adenovirus-mediated gene transfer of nitric oxide synthase augmented relaxations of the RA to acetylcholine. CONCLUSIONS: Reduced production of endothelium-derived relaxing factors suggests diminished endothelial regulation of vascular smooth muscle in the RA compared with the IMA. This finding may explain, in part, the predisposition to vasoconstriction in RA grafts.

The Allen test.

There has been renewed interest in radial artery bypass conduits for coronary artery revascularization, and surgeons should be familiar with methods of assessing circulation of the hand to determine which patients are candidates for this procedure. The Allen test originally was described in 1929, and this report reviews its initial description and contribution to our surgical heritage.

Ex vivo gene transfer of endothelial nitric oxide synthase to atherosclerotic rabbit aortic rings improves relaxations to acetylcholine.

Cholesterol feeding results in impaired endothelium dependent vasorelaxation. The role of nitric oxide in this process is unclear. The aim of this study was to evaluate the role of nitric oxide in cholesterol-induced vasomotor dysfunction by examining the effect of overexpression of eNOS in the hypercholesterolemic rabbit aorta on vascular reactivity. Vascular rings from the thoracic aorta of hypercholesterolemic rabbits were exposed ex vivo either to an adenoviral vector encoding endothelial nitric oxide synthase (AdeNOS) or Escherichia coli beta Galactosidase (AdbetaGal). Transgene expression was examined by histochemistry for beta galactosidase, immunohistochemistry for eNOS and cyclic GMP measurements and vasomotor studies were performed. Transgene expression was found to localize to the endothelium and adventitia. cGMP levels were significantly greater in AdeNOS compared to AdbetaGal transduced rings. Acetylcholine mediated relaxation was significantly impaired in cholesterol fed rabbits and was markedly improved by overexpression of eNOS. These results suggest that reduced NO bioavailability observed in cholesterol-induced vascular dysfunction can be partially overcome by eNOS gene transfer.

New approaches to prevention and treatment of radial artery graft vasospasm.

BACKGROUND: There has been renewed interest in radial artery (RA) conduits for coronary artery bypass because of the relative resistance of arterial grafts to atherosclerosis compared with autogenous vein grafts. Although improved drug therapy for arterial spasm is now available, vasospasm still occurs in at least 5% to 10% of RA grafts. We systematically evaluated the effectiveness of calcium channel blockers and organic nitrates for inhibition or reversal of RA contraction in vitro. Additionally, we investigated the efficacy of novel gene therapy with endothelial nitric oxide synthase (eNOS) to inhibit RA contractions. METHODS AND RESULTS: Segments of RA from 28 patients undergoing coronary artery bypass grafting were mounted in organ chambers. In control experiments, KCl (5 to 50 mmol/L) produced dose-dependent increases in tension (maximum tension, 14.3 +/- 3.0 g, n = 7). Addition of diltiazem or verapamil had no significant effect on KCl contraction (128 +/- 36% and 88 +/- 24% control, respectively); however, nifedipine markedly inhibited KCl contraction (27 +/- 4% control, P = 0.005). Norepinephrine (NE, 10(-9) to 10(-4) M) produced dose-dependent increases in tension (maximum tension, 15.7 +/- 2.7 g in control rings, n = 8). Diltiazem and verapamil pretreatment had no significant effect on NE contraction (103 +/- 14% and 90 +/- 14% control, respectively); nifedipine significantly inhibited NE contraction (70 +/- 11% control, P = 0.02). Isosorbide dinitrate and nitroglycerin markedly inhibited KCl contractions (47 +/- 9% and 30 +/- 8% of controls, n = 6) and NE contractions (42 +/- 10% and 31 +/- 9% of controls, n = 6). Nifedipine, isosorbide, and nitroglycerin were further evaluated for the ability to reverse an established contraction (KCl 40 mmol/L); nitroglycerin was most effective in reversing RA contraction. In separate experiments, RA underwent adenoviral-mediated gene transfer with vehicle, Escherichia coli beta-galactosidase, or eNOS (eNOS, 10(10) PFU/mL x 1 hour). Transgene expression was confirmed by beta-galactosidase activity and eNOS immunohistochemistry after 40 hours of ex vivo incubation. Immunohistochemistry demonstrated recombinant NOS in adenovirus encoding bovine eNOS (Ad.CMVeNOS) RA only. Ad.CMVeNOS arteries contracted only 46.6 +/- 13.7% of controls to KCl (n = 5) and 48.2 +/- 11.4% of controls to prostaglandin F2 alpha a (10(-9) to 10(-6) M, n = 5). CONCLUSIONS: Diltiazem, which is used empirically to prevent RA vasospasm, had little effect on human RA contractions (receptor-independent and receptor-dependent). Organic nitrates inhibited and reversed RA contractions. Adenoviral transfer of NOS suggests that future clinical application of gene therapy may play an important role in prevention of RA vasospasm.

Minimally invasive saphenous vein harvesting: endothelial integrity and early clinical results.

BACKGROUND: Endoscopic harvesting of vein grafts may reduce wound complications but the effect on the venous endothelium is unknown. Endothelium-derived vasoactive substances may be important in vein graft disease prevention. We investigated the impact of endoscopic harvesting on the release of these factors and proceeded to evaluate the clinical applicability. METHODS AND RESULTS: Porcine veins were harvested in either an endoscopic or open fashion. Superfusion bioassay from endoscopic veins had a similar basal secretion as control veins (6.5% +/- 1.5% versus 3.2% +/- 2.2%, respectively; n = 5, p = 0.39). Calcium ionophore A23187 stimulation was similar in both groups (24.6% +/- 5.1% versus 27.3% +/- 9.6%; n = 5, p = 0.68). Light and electron microscopy documented a normal endothelial monolayer in both groups with no endothelial cell or connective tissue loss. Encouraged by these results, 38 patients have subsequently undergone this procedure at our institution. Total operative time for harvesting 35 to 45 cm of saphenous vein was 62.3 +/- 5.3 minutes (range, 35 to 120 minutes). The procurement time in the most recent five patients was 41.6 +/- 3.3 minutes. Patients had little incisional pain, but did have mild ecchymosis. CONCLUSIONS: Endothelial release of vasoactive substances after endoscopic harvesting is similar to that after the traditional, extended incision technique, and microscopy confirmed similar histology. These laboratory findings support the satisfactory early clinical experience with endoscopic harvesting of saphenous veins.

Selective convective brain cooling during normothermic cardiopulmonary bypass in dogs.

OBJECTIVE: Although normothermic cardiopulmonary bypass results in improved cardiac outcome, patients do not benefit from hypothermia-mediated brain protection and thus may be at high risk for ischemic brain injury. The present study evaluated the efficacy of selective forced-air cerebral cooling. METHODS: Sixteen dogs were anesthetized with either intravenous pentobarbital or inhaled halothane (n = 8 for each group). Temperatures were monitored in the esophagus (i.e., core), parietal epidural space, and brain parenchyma. Normothermic atrial-femoral cardiopulmonary bypass and forced-air pericranial cooling (to approximately 13 degrees C) were maintained for 150 minutes. Data between groups were compared by means of repeated-measures analysis of variance and two-sample t test. Within each group, brain-to-core temperature gradients were compared to zero by means of the one-sample t test. RESULTS: In pentobarbital-anesthetized dogs, after 30 minutes of cerebral cooling, temperatures in the parietal epidural space and 1 cm and 2 cm beneath the dura were 3.3 degrees +/- 1.4 degrees C (mean +/- standard deviation), 2.6 degrees +/- 1.3 degrees C, and 1.1 degrees +/- 0.6 degrees C cooler than the core temperature, respectively. At the conclusion of the study (i.e., 150 minutes), these temperatures were 4.5 degrees +/- 1.8 degrees C, 3.9 degrees +/- 1.6 degrees C, and 2.0 degrees +/- 0.9 degrees C cooler than the core temperature, respectively. Similar changes were observed in halothane-anesthetized dogs. CONCLUSIONS: Regardless of the background anesthetic, the magnitude of selective cerebral cooling observed in our study was larger than the 1 degree to 2 degrees C changes previously reported to modulate ischemic brain injury.

Anesthetic technique influences brain temperature during cardiopulmonary bypass in dogs.

BACKGROUND: Because different anesthetics have different effects on cerebral blood flow and cerebral metabolism, we hypothesized that they also may have different effects on brain temperature during hypothermic cardiopulmonary bypass (CPB) and subsequent rewarming. METHODS: Sixteen dogs were anesthetized either with inhaled halothane, 1.0 minimum alveolar concentration (ie, an anesthetic that should increase cerebral blood flow and minimally affect cerebral metabolism; n = 8), or with intravenous high-dose pentobarbital (ie, an anesthetic that should reduce cerebral blood flow and cerebral metabolism by approximately one half; n = 8). Normocapnia (alpha-stat technique) and a blood pressure near 90 mm Hg were maintained. Thermistors were placed in the esophagus (ie, the body core), in the parietal epidural space, and in the parietal brain parenchyma at depths of 1 and 2 cm. Initially, all temperatures were controlled at 38.0 degrees +/- 0.2 degrees C (mean +/- standard deviation). Thereafter, atrial-femoral artery CPB was initiated, and after 15 minutes at 38 degrees C, the core temperature was decreased to 28 degrees C over approximately 21 minutes. After 30 minutes at 28 degrees C, the core temperature was returned to 38 degrees C over approximately 21 minutes and was maintained at 38 degrees C for the next 30 minutes. RESULTS: In halothane-anesthetized dogs, the mean brain-to-core temperature gradient always was 1.0 degrees C or less for all brain sites during all phases of CPB. In contrast, in pentobarbital-anesthetized dogs, the mean brain temperature during active cooling typically exceeded the core temperature by 1.7 degrees to 2.2 degrees C. This brain-to-core temperature gradient persisted into the period of stable hypothermia. During the rewarming phase of CPB, the mean brain temperature was 2.9 degrees to 3.4 degrees C cooler than the core temperature. This trend of relative cerebral hypothermia persisted well into the period in which the core temperature was 38 degrees C. CONCLUSIONS: Deep barbiturate anesthesia resulted in a brain-to-core temperature gradient during CPB that was of a magnitude greater than the 1 degrees C previously reported to modulate ischemic neurologic injury. We speculate that the timely administration of barbiturates (eg, during the latter stages of CPB) may be useful as part of a cerebroprotective regimen in humans undergoing CPB, in part because the barbiturates influence brain temperature.

Selective convective brain cooling during hypothermic cardiopulmonary bypass in dogs.

BACKGROUND: Neurologic complications, primarily resulting from ischemic insults, represent the leading cause of morbidity and disability, and the second most common source of death, after cardiac operations. Previous studies have reported that increases (as occur during the rewarming phase of cardiopulmonary bypass [CPB]) or decreases in brain temperature of a mere 0.5 degrees to 2 degrees C can significantly worsen or improve, respectively, postischemic neurologic outcome. The purpose of the present study was to evaluate a novel approach of selectively cooling the brain during hypothermic CPB and subsequent rewarming. METHODS: Sixteen dogs were anesthetized with either intravenous pentobarbital or inhaled halothane (n = 8 per group). Normocapnia (alpha stat technique) and a blood pressure near 75 mm Hg were maintained. Temperatures were monitored by placing thermistors in the esophagus (i.e., core), parietal epidural space, and brain parenchyma at depths of 1 and 2 cm beneath the dura. During CPB, core temperature was actively cycled from 38 degrees C to 28 degrees C, and then returned to 38 degrees C. Forced air pericranial cooling (air temperature of approximately 13 degrees C) was initiated simultaneous with the onset of CPB, and maintained throughout the bypass period. Brain-to-core temperature gradients were calculated by subtracting the core temperature from regional brain temperatures. RESULTS: In halothane-anesthetized dogs, brain temperatures at all monitoring sites were significantly less than core during all phases of CPB, with one exception (2 cm during systemic cooling). Brain cooling was most prominent during and after systemic rewarming. For example, during systemic rewarming, average temperatures in the parietal epidural space, and 1 and 2 cm beneath the dura, were 3.3 degrees +/- 1.3 degrees C (mean +/- standard deviation), 3.2+/-1.4 degrees C, and 1.6 degrees +/-1.0 degrees C, cooler than the core, respectively. Similar trends, but of a greater magnitude, were noted in pentobarbital-anesthetized dogs. For example, during systemic rewarming, corresponding brain temperatures were 6.5 degrees +/-1.7 degrees C, 6.3 degrees +/-1.6 degrees C, and 4.2+/-1.3 degrees C cooler than the core, respectively. CONCLUSIONS: The magnitude of selective brain cooling observed in both study groups typically exceeded the 0.5 degrees to 2.0 degrees C change previously reported to modulate ischemic injury, and was most prominent during the latter phases of CPB. When compared with previous research from our laboratory, application of cold forced air to the cranial surface resulted in brain temperatures that were cooler than those observed during hypothermic CPB without pericranial cooling. On the basis of the assumption that similar beneficial brain temperature changes can be induced in humans, we speculate that selective convective brain cooling may enable clinicians to improve neurologic outcome after hypothermic CPB.

Complement mediates attenuation of endothelium-dependent relaxations in canine coronary arteries after porcine serum exposure: a mechanism for vascular thrombosis in xenograft hyperacute rejection.

BACKGROUND: Evidence for complement activation in xenograft hyperacute rejection includes prolongation of graft survival after complement inactivation as well as component deposition and consumption during hyperacute rejection. The current investigations examined the endothelial production of vasoactive substances during heterologous serum exposure. METHODS: Segments of canine coronary artery were exposed to either autologous canine serum for 90 minutes or heterologous porcine serum for 30, 60, or 90 minutes. After replacement of the serum with buffered saline, segments were contracted with phenylephrine (10(-6) mol/L) in the presence of indomethacin (10(-5) mol/L). RESULTS: Compared with responses of vessels exposed to autologous canine serum, receptor-dependent relaxation to acetylcholine was impaired in arteries after 60 or 90 minutes of exposure to porcine serum. Receptor-independent relaxation to calcium ionophore A23187 was not significantly impaired at any length of porcine serum exposure. Endothelial-independent relaxation to sodium nitroprusside was not impaired with either canine or porcine serum exposure. Oxyhemoglobin (10(-6) mol/L) abolished acetylcholine-mediated relaxation, indicating that nitric oxide was the predominant mediator of the impaired pathway. Basal release of nitric oxide after a 60-minute porcine serum exposure was reduced by half compared with coronary arteries exposed to autologous canine serum. Serum pretreated by either heat inactivation of complement or immunoadsorption with anti-C3 antibodies failed to depress endothelial-dependent relaxation on 60 minutes of exposure to canine coronary arteries. Scanning electron microscopy revealed an intact endothelial layer in coronary arteries exposed to either porcine or canine serum for 60 minutes. CONCLUSIONS: Hyperacute xenograft rejection impairs receptor-dependent relaxation of canine coronary arteries at 60 and 90 minutes. These data strongly suggest that impairment of endothelial production of nitric oxide during acute xenograft rejection is mediated by complement activation.

Recombinant endothelial nitric oxide synthase-transduced human saphenous veins: gene therapy to augment nitric oxide production in bypass conduits.

BACKGROUND: Nitric oxide is a potent vasodilator that also inhibits platelet aggregation and smooth muscle cell proliferation, properties that may prevent early and late occlusion of saphenous vein coronary bypass conduits. We determined whether human saphenous veins can be transduced with adenovirus vector-encoding bovine endothelial nitric oxide synthase (Ad.CMVeNOS), resulting in functional expression of recombinant nitric oxide synthase. METHODS AND RESULTS: Harvested segments of human saphenous vein were exposed for 1 hour at 37 degrees C to replication-deficient Ad.CMVeNOS (5 x 10(9) PFU/mL) or control adenovirus-encoding Escherichia coli beta-galactosidase (Ad.CMVLacZ; 5 x 10(9) PFU/mL). The vein segments were analyzed for recombinant endothelial nitric oxide synthase expression and activity 48 hours later. Histochemical staining for recombinant beta-galactosidase activity was localized to the luminal endothelium and adventitia of vein segments transduced with Ad.CMVLacZ. Similarly, immunohistochemical staining with a monoclonal antibody for nitric oxide synthase localized recombinant gene expression to endothelial and adventitial cells in Ad.CMVeNOS veins; only endogenous nitric oxide synthase was identified in the endothelium of Ad.CMVLacZ veins. Nitrite generation after stimulation with calcium ionophore increased in Ad.CMVeNOS veins (1420.0+/-298.2 nM/mg versus 130.3+/-19.9 nM/mg; n=3; P<.05). Isometric tension recording demonstrated augmented maximal relaxation to calcium ionophore (32+/-4.5% versus 17.4+/-7.4%; n=6; P<.05) after precontraction with norepinephrine. Bioassay superfusion demonstrated a twofold augmentation of the biodetector ring relaxation during calcium ionophore stimulation of Ad.CMVeNOS veins. CONCLUSIONS: Adenovirus-mediated gene transfer to human saphenous veins resulted in functional transgene expression with increased nitric oxide release. These or similar molecular techniques to increase nitric oxide production may reduce the risk of early thrombosis in saphenous vein grafts.

Hypothermic storage with University of Wisconsin solution preserves endothelial and vascular smooth-muscle function.

BACKGROUND: The effect of University of Wisconsin (UW) solution on coronary endothelial function is debated. Although earlier studies demonstrated preservation of stimulated release of endothelial vasoactive factors, the effect of UW solution on basal nitric oxide release has not been investigated, and this was the purpose of the present study. METHODS AND RESULTS: Isolated canine coronary arteries were evaluated immediately after being harvested (group 1) and after 24-hour hypothermic (4 degrees C) storage in either buffered physiological saline solution (group 2) or UW solution (group 3). Smooth-muscle vascular response to sodium nitroprusside (10(-9) to 10(-6) mol/L), KCl (5 to 50 mmol/L), and prostaglandin F2alpha (10(-9) to 10(-6) mol/L) was comparable among the groups, as was the endothelial response to acetylcholine (10(-9) to 10(-6) mol/L) and calcium ionophore A23187 (10(-9) to 10(-6) mol/L). Vessels in group 3 exhibited better preservation of basal release of nitric oxide, as measured by contraction to either N(G)-monomethyl L-arginine (10(-4) mol/L) or methylene blue (10(-5) mol/L), in comparison with control arteries exposed to saline solution (group 2). There was no difference in the basal production of nitric oxide between group 1 and group 3 vessels. CONCLUSIONS: The present study demonstrated preservation of vascular smooth muscle and endothelium-dependent relaxations after prolonged hypothermic storage in UW solution. Furthermore, storage in this solution preserved basal release of nitric oxide.