Measurement of cerebral blood flow during cardiopulmonary bypass with near-infrared spectroscopy

OBJECTIVES: We sought to assess the incidence and clinical relevance of examination data to recurrent ischemia within an international randomized trial. BACKGROUND: Ischemic symptoms commonly recur after thrombolysis for acute myocardial infarction. METHODS: Patients (n = 40,848) were prospectively evaluated for recurrent angina and transient electrocardiographic (ECG) or hemodynamic changes. Five groups were developed: Group 1, patients with no signs or symptoms of recurrent ischemia; Group 2, patients with angina only; Group 3, patients with angina and ST segment changes; Group 4, patients with angina and hemodynamic abnormalities; and Group 5, patients with angina, ST segment changes and hemodynamic abnormalities. Baseline clinical and outcome variables were compared among the five groups. RESULTS: Group 1 comprised 32,717 patients, and Groups 2 to 5 comprised 20% of patients (4,488 in Group 2; 3,021 in Group 3; 337 in Group 4; and 285 in Group 5). Patients with recurrent ischemia were more often female, had more cardiovascular risk factors and less often received intravenous heparin. Significantly more extensive and more severe coronary disease, antianginal treatment, angioplasty and coronary bypass surgery were observed as a function of ischemic severity. The 30-day reinfarction rate was 1.6% in Group 1, 6.5% in Group 2, 21.7% in Group 3, 13.1% in Group 4 and 36.5% in Group 5 (p < 0.0001); in contrast, the 30-day mortality rate was significantly lower (p < 0.0001) in Groups 1, 2 and 3 (6.6%, 5.4% and 7.7%, respectively) than in Groups 4 and 5 (21.8% and 29.1%). CONCLUSIONS: Postinfarction angina greatly increases the risk of reinfarction, especially when accompanied by transient ECG changes. However, mortality is markedly increased only in the presence of concomitant hemodynamic abnormalities.     

Gut mucosal ischemia during normothermic cardiopulmonary bypass results from blood flow redistribution and increased oxygen demand

Impaired gut mucosal perfusion has been reported during cardiopulmonary bypass. To better define the adequacy of gut blood flow and oxygenation during cardiopulmonary bypass, we measured overall gut blood flow and ileal mucosal flow and their relationship to mucosal pH, mesenteric oxygen delivery and oxygen consumption in immature pigs (n = 8). Normothermic, noncross-clamped, right atrium-to-aorta cardiopulmonary bypass was maintained at 100 ml/kg per minute for 120 minutes. Animals were instrumented with an ultrasonic Doppler flow probe on the superior mesenteric artery, a mucosal laser Doppler flow probe in the ileum, and pH tonometers in the stomach, ileum, and rectum. Radioactive microspheres were injected before and at 5, 60, and 120 minutes of cardiopulmonary bypass for tissue blood flow measurements. Overall gut blood flow significantly increased during cardiopulmonary bypass as evidenced by increases in superior mesenteric arterial flow to 134.1% +/- 8.0%, 137.1% +/- 7.5%, 130.3% +/- 11.2%, and 130.2% +/- 12.7% of baseline values at 30, 60, 90, and 120 minutes of bypass, respectively. Conversely, ileal mucosal blood flow significantly decreased to 53.6% +/- 6.4%, 49.5% +/- 6.8%, 58.9% +/- 11.6%, and 47.8% +/- 10.0% of baseline values, respectively. Blood flow measured with microspheres was significantly increased to proximal portions of the gut, duodenum and jejunum, during cardiopulmonary bypass, whereas blood flow to distal portions, ileum and colon, was unchanged. Gut mucosal pH decreased progressively during cardiopulmonary bypass and paralleled the decrease in ileal mucosal blood flow. Mesenteric oxygen delivery decreased significantly from 67.0 +/- 10.0 ml/min per square meter at baseline to 42.4 +/- 4.6, 44.9 +/- 3.5, 46.0 +/- 3.6, and 42.9 +/- 3.9 ml/min per square meter at 30, 60, 90, and 120 minutes of bypass. Despite the decrease in mesenteric oxygen delivery, mesenteric oxygen consumption increased progressively from 10.8 +/- 1.4 ml/min per square meter at baseline to 13.4 +/- 1.2, 15.9 +/- 1.2, 16.7 +/- 1.4, and 16.6 +/- 1.54 ml/min per square meter, respectively. We conclude that gut mucosal ischemia during normothermic cardiopulmonary bypass results from a combination of redistribution of blood flow away from mucosa and an increased oxygen demand.     

The effects of hemodilution on cerebral hemodynamics and oxygen consumption during nonpulsatile hypothermic cardiopulmonary bypass

The effects of hemodilution on cerebral hemodynamics and oxygen consumption during nonpulsatile hypothermic cardiopulmonary bypass (CPB) were investigated in 16 patients undergoing elective open heart surgery under CPB. They were divided into 2 groups; high (25%) (H) and low hematocrit (15%) (L) groups according to the hematocrit values during CPB. Simultaneous measurements of blood flow velocity in the middle cerebral artery (VMCA) and oxygen tension in the jugular venous bulb (PjvbO2) were performed at prebypass, bypass, and post bypass periods. There were no significant differences in nasopharyngeal temperature (NPT), cerebral perfusion pressure, pump flow rate, and PaCO2 between the two groups at any period. VMCA changed in correlation with NPT in both groups, and VMCA in the L group was faster than that in the H group at any period. PjvbO2 increased during aortic cross clamp, and decreased on rewarming in both groups. However, there were no significant differences in VMCA and PjvbO2 between the two groups at any period. These findings suggest that increased cerebral blood flow may compensate the decreased oxygen content due to the hemodilution during CPB. Therefore, the profound hemodilution up to 15% of hematocrit is considered to be a safe technique during CPB.     

Pulsatile versus nonpulsatile flow. No difference in cerebral blood flow or metabolism during normothermic cardiopulmonary bypass in rabbits

BACKGROUND: Although pulsatile and nonpulsatile cardiopulmonary bypass (CPB) do not differentially affect cerebral blood flow (CBF) or metabolism during hypothermia, studies suggest pulsatile CPB may result in greater CBF than nonpulsatile CPB under normothermic conditions. Consequently, nonpulsatile flow may contribute to poorer neurologic outcome observed in some studies of normothermic CPB. This study compared CBF and cerebral metabolic rate for oxygen (CMRO2) between pulsatile and nonpulsatile CPB at 37 degrees C. METHODS: In experiment A, 16 anesthetized New Zealand white rabbits were randomized to one of two pulsatile CPB groups based on pump systolic ejection period (100 and 140 ms, respectively). Each animal was perfused at 37 degrees C for 30 min at each of two pulse rates (150 and 250 pulse/min, respectively). This scheme created four different arterial pressure waveforms. At the end of each perfusion period, arterial pressure waveform, arterial and cerebral venous oxygen content, CBF (microspheres), and CMRO2 (Fick) were measured. In experiment B, 22 rabbits were randomized to pulsatile (100-ms ejection period, 250 pulse/min) or nonpulsatile CPB at 37 degrees C. At 30 and 60 min of CPB, physiologic measurements were made as before. RESULTS: In experiment A, CBF and CMRO2 were independent of ejection period and pulse rate. Thus, all four waveforms were physiologically equivalent. In experiment B, CBF did not differ between pulsatile and nonpulsatile CPB (72 +/- 6 vs. 77 +/- 9 ml.100 g-1.min-1, respectively (median +/- quartile deviation)). CMRO2 did not differ between pulsatile and nonpulsatile CPB (4.7 +/- 0.5 vs. 4.1 +/- 0.6 ml O2.100 g-1.min-1, respectively) and decreased slightly (0.4 +/- 0.4 ml O2.100 g-1.min-1) between measurements. CONCLUSIONS: During CPB in rabbits at 37 degrees C, neither CBF nor CMRO2 is affected by arterial pulsation. The absence of pulsation per se is not responsible for the small decreases in CMRO2 observed during CPB.     

Autologous blood transfusion during cardiopulmonary bypass

Immediately prior to cardiopulmonary bypass, two units of blood were removed from each of 25 patients undergoing open-heart surgery, and the autologous blood was reinfused after cessation of support with the pump oxygenator. Pertinent data on blood balance and hematologic measurements were compared to a matched group of control subjects. There were no significant differences in the amount of operative or postoperative bleeding, the requirements for homologous blood and blood products, or the amount of protamine needed for neutralization of heparin.     

Comparison of pH-stat and alpha-stat cardiopulmonary bypass on cerebral oxygenation and blood flow in relation to hypothermic circulatory arrest in piglets

Melagatran, a new, competitive and rapid inhibitor of thrombin with a molecular mass of 429 Da is described. Melagatran is well tolerated when administered in very high doses, and the oral bioavailability in the dog is relatively high. The aim of the study was to determine, in the preclinical setting, the degree of selectivity against the fibrinolytic system required for entering the clinical development phase. Melagatran was compared with two structurally similar thrombin inhibitors, inogatran and H 317/86. The potent inhibition of thrombin by melagatran was demonstrated by a low inhibition constant (Ki) for thrombin (0.002 micromol/l) and prolongation of clotting time to twice the control value in coagulation assays at low concentrations (0.010, 0.59 and 2.2 micromol/l for thrombin time, activated partial thromboplastin time and prothrombin time, respectively). Furthermore, thrombin-induced platelet aggregation was inhibited at the same concentration (IC50-value 0.002 micromol/l) as the Ki-value for thrombin. In two assays of global fibrinolysis, inhibition was observed at a concentration of 1.1 micromol/l in a euglobulin plasma fraction model, while no inhibition was observed at a concentration of < or = 10 micromol/l in a plasma model. In an in vivo model of endogenous fibrinolysis in the rat, inhibition of fibrinolysis was observed at > or = 1.0 micromol/l. In all assays, except the Ki-ratio determinations, the compounds could be graded with regard to selectivity against the fibrinolytic system: inogatran > melagatran > H 317/86. For melagatran, inhibition of fibrinolysis was not observed at concentrations below the upper limit of the proposed therapeutic plasma concentration interval (< 0.5 micromol/l). Thus, melagatran seems to have a sufficient selectivity against the fibrinolytic system, while H 317/86 was considered to be insufficient for clinical development.     

Cerebral blood flow velocity in relation to cerebral blood flow, cerebral metabolic rate for oxygen, and electroencephalogram analysis during isoflurane anesthesia in dogs

The purpose of this study was to correlate changes in cerebral blood flow velocity (Vmean) with cerebral blood flow (CBF) during isoflurane anesthesia in dogs. The relation between cerebral oxygen consumption (CMRO2) and electroencephalogram (EEG) analysis also was investigated. Blood flow velocity was measured in the middle cerebral artery using a pulsed transcranial Doppler (TCD). CBF was measured with radioactive microspheres. EEG was measured over both hemispheres and median EEG frequency (median frequency) was calculated after fast Fourier transformation. Baseline anesthesia was maintained with 50% nitrous oxide in oxygen and 50 micrograms.kg-1 x h-1 fentanyl. Animals of Group I (control, n = 6) were not given isoflurane. Data were recorded at baseline, and at 30, 60, and 90 min. There was no significant change in any variable over time. In Group II (n = 7), data were recorded at baseline and at 1%, 2%, and 3% end-tidal isoflurane. Mean arterial pressure was maintained at baseline levels by phenylephrine infusion. CBF increased from 70.8 +/- 10.6 mL.100g-1 x min-1 at baseline to 146.1 +/- 36.9 mL.100 g-1 x min-1 with 3% isoflurane (P < 0.01). Vmean increased from 38.3 +/- 6.7 cm/s to 65.6 +/- 9.7 cm/s (P < 0.01). The correlation between relative changes in CBF and Vmean was r = 0.94 (P < 0.01). With 1% isoflurane the EEG shifted to slow-wave, high-voltage activity, and median frequency decreased from 5.9 +/- 0.7 Hz to 1.4 +/- 0.4 Hz (P < 0.05). Median frequency was not decreased further during 2% and 3% isoflurane anesthesia.(ABSTRACT TRUNCATED AT 250 WORDS)     

Changes of alpha hANP concentration in arterial and coronary sinus blood during and after cardiopulmonary bypass

The Digit Memory Test (DMT) (Hiscock & Hiscock, 1989), a forced-choice test for detecting malingering, was administered to 27 patients with unequivocal cerebral dysfunction, 5 patients with postconcussional syndrome, 6 suspected malingerers and 10 normal controls. Results indicate that, even in patients with severe, but static cerebral dysfunction and unequivocal memory disorder, DMT performance is between 95% to 100% correct. By contrast, the 6 patients in whom malingering was seriously considered performed at a level much below the other three groups (74% correct) but not significantly below chance. The DMT may be helpful in evaluating patients suspected of malingering even when they do not score significantly below chance.     

Emboli from an extraluminal blood flow hollow fiber oxygenator with and without an arterial filter during cardiopulmonary bypass in a pig model

The effect of an arterial filter on visceral emboli was quantified with autologous indium-111 labeled platelets (INPLT) during cardiopulmonary bypass (CPB) in Yorkshire pigs. Biodistribution of INPLT was determined in 12 control pigs (30-35 kg, unoperated control [n = 6] and sham operated control [n = 6]). CPB was carried out with (n = 6) and without (n = 6) an arterial filter in 12 pigs at a flow rate of 2.5-3.5 L/min. Platelets labeled with In-111 tropolone (650-780 microCi) were injected intravenously 24 hr before CPB. All pigs were systemically heparinized (activated coagulation time > 400 sec); CPB was instituted with a roller pump, an extraluminal blood flow oxygenator (Bentley Univox, 1.8 m2), and an arterial filter (0.25 m2) and continued for 3 hr. Platelet kinetics, pooling, and counts were monitored by a Geiger probe and a Coulter counter. The thrombi in the oxygenator and arterial filter and emboli in viscera and brain were imaged with a gamma camera and measured with an ion chamber and gamma counter. Percentage of INPLT (mean +/- SD) in organs, tissues, and components of the circuit in four groups of pigs was calculated. Flow cytometry with antibodies to CD61 (GPIIIa) and CD62P (GMP-140: control) of porcine platelets was carried out with blood samples taken before, during, and after CPB for estimation of circulating platelet aggregates and platelet microparticles. Pulmonary, renal, cardiac, and cerebral emboli in pigs undergoing CPB with and without a filter were similar (p < 0.1). The amount of filter adherent thrombi was small (0.04 +/- 0.01%); oxygenator adherent thrombus in both groups was similar (p < 0.1). Emboli were found in the cerebral medulla, hippocampus, and posterior cerebral cortex in both groups. During CPB, the arterial filter functioned minimally as a trap for platelet thrombi detached from the oxygenator and circulating emboli. Flow cytometry of blood demonstrated the shift of equilibria from single platelets to platelet aggregates and microparticles during CPB and their gradual reversal to single platelets after CPB; the loosely adherent emboli disaggregated and further shifted these equilibria to single platelets and smaller aggregates, probably through the action of endogenous nitric oxide and prostacyclin. The emboli were trapped in organs and tissues and microparticles were sequestered by the reticuloendothelial system.     

Regional blood flow during pulsatile cardiopulmonary bypass and after circulatory arrest in an infant model

BACKGROUND: Pulsatile perfusion systems have been proposed as a means of improving end-organ perfusion during and after cardiopulmonary bypass. Few attempts have been made to study this issue in an infant model. METHODS: Neonatal piglets were subjected to nonpulsatile (n = 6) or pulsatile (n = 7) cardiopulmonary bypass and 60 minutes of circulatory arrest. Cerebral, renal, and myocardial blood flow measurements were obtained at baseline, on bypass before and after circulatory arrest, and after bypass. RESULTS: Cerebral blood flow did not differ between groups at any time and was diminished equally in both groups after circulatory arrest. Renal blood flow was diminished in both groups during bypass but was significantly better in the pulsatile group than in the nonpulsatile group prior to, but not after, circulatory arrest. Myocardial blood flow was maintained at or above baseline in the pulsatile group throughout the study, but in the nonpulsatile group, it was significantly lower than baseline during CPB prior to circulatory arrest and lower compared with baseline and with the pulsatile group 60 minutes after CPB. CONCLUSIONS: Pulsatile bypass does not improve recovery of cerebral blood flow after circulatory arrest, may improve renal perfusion during bypass but does not improve its recovery after ischemia, and may have beneficial effects on myocardial blood flow during bypass and after ischemia compared with nonpulsatile bypass in this infant model.     

Cerebral cortex regional blood flow and tissue oxygen tension during the trigeminal depressor response in rabbits

The paradox of pancytopenia despite cellular bone marrows (BM) was investigated in 120 patients with myelodysplastic syndromes (MDS). Detailed cell cycle kinetics were examined following in vivo infusions of iodo--and/or bromodeoxyuridine (IUdR/BrdU), while the incidence of apoptosis was measured by in situ end labeling (ISEL) of fragmented DNA. Results showed that MDS are highly proliferative disorders with an equally high incidence of apoptotic intramedullary cell death accounting for the paradox of cellularity/cytopenia. By double-labeling BM biopsy sections for ISEL/BrdU we found the peculiar situation of "signal antonymy" where S-phase cells were frequently apoptotic, a phenomenon so far only seen in MDS biopsies. The cause-effect relationship of this excessive proliferation/apoptosis is discussed at length.     

Hypothermic modulation of cerebral ischemic injury during cardiopulmonary bypass in pigs

BACKGROUND: The aim of this study was to determine whether progressive levels of hypothermia (37, 34, 31, or 28 degrees C) during cardiopulmonary bypass (CPB) in pigs reduce the physiologic and metabolic consequences of global cerebral ischemia. METHODS: Sagittal sinus and cortical microdialysis catheters were inserted into anesthetized pigs. Animals were placed on CPB and randomly assigned to 37 degrees C (n = 10), 34 degrees C (n = 10), 31 degrees C (n = 11), or 28 degrees C (n = 10) management. Next 20 min of global cerebral ischemia was produced by temporarily ligating the innominate and left subclavian arteries, followed by reperfusion, rewarming, and termination of CPB. Cerebral oxygen metabolism (CMRO2) was calculated by cerebral blood flow (radioactive microspheres) and arteriovenous oxygen content gradient. Cortical excitatory amino acids (EAA) by microdialysis were measured using high-performance liquid chromatography. Electroencephalographic (EEG) signals were graded by observers blinded to the protocol. After CPB, cerebrospinal fluid was sampled to test for S-100 protein and the cerebral cortex was biopsied. RESULTS: Cerebral oxygen metabolism increased after rewarming from 28 degrees C, 31 degrees C, and 34 degrees C CPB but not in the 37 degrees animals; CMRO2 remained lower with 37 degrees C (1.8 +/- 0.2 ml x min[-1] x 100 g[-1]) than with 28 degrees C (3.1 +/- 0.1 ml x min[-1] x 100 g[-1]; P < 0.05). The EEG scores after CPB were depressed in all groups and remained significantly lower in the 37 degrees C animals. With 28 degrees C and 31 degrees C CPB, EAA concentrations did not change. In contrast, glutamate increased by sixfold during ischemia at 37 degrees C and remained significantly greater during reperfusion in the 34 degrees C and 37 degrees C groups. Cortical biopsy specimens showed no intergroup differences in energy metabolites except two to three times greater brain lactate in the 37 degrees C animals. S-100 protein in cerebrospinal fluid was greater in the 37 degrees C (6 +/- 0.9 microg/l) and 34 degrees C (3.5 +/- 0.5 microg/l) groups than the 31 degrees C (1.9 +/- 0.1 microg/l) and 28 degrees C (1.7 +/- 0.2 microg/l) animals. CONCLUSIONS: Hypothermia to 28 degrees C and 31 degrees C provides significant cerebral recovery from 20 min of global ischemia during CPB in terms of EAA release, EEG and cerebral metabolic recovery, and S-100 protein release without greater advantage from cooling to 28 degrees C compared with 31 degrees C. In contrast, ischemia during 34 degrees C and particularly 37 degrees C CPB showed greater EAA release and evidence of neurologic morbidity. Cooling to 31 degrees C was necessary to improve acute recovery during global cerebral ischemia on CPB.     

Renal blood flow velocity in acute renal failure following cardiopulmonary bypass surgery

Two lysosomal storage diseases, aspartylglucosaminuria and mannosidosis, are associated with highly elevated serum dolichol concentrations. To elucidate possible mechanisms leading to elevated serum dolichols, we studied the effects of Triton WR 1339 (known to increase serum cholesterol) and orotic acid (known to decrease serum cholesterol) on blood and biliary dolichol and beta-hexosaminidase levels in rats. In Triton WR 1339-treated rats, serum dolichol was markedly increased compared with saline-treated controls 1 (400 +/- 70 ng/mL, n = 7 v 85 +/- 11 ng/mL, n = 8, P < .001), 4 (789 +/- 70 ng/mL, n = 10 v 110 +/- 10 ng/mL, n = 7, P < .0001), and 8 (549 +/- 43 ng/mL, n = 8 v 87 +/- 8 ng/mL, n = 7, P < .001) days after administration of the drug. By contrast, serum dolichol was decreased (64 +/- 5 ng/mL, n = 8 v 119 +/- 7 ng/mL, n = 8, P < .0001) after a 7-day orotic acid feeding compared with controls. Serum beta-hexosaminidase was unaffected by both treatments. Orotic acid also increased biliary dolichol (280 +/- 47 ng/100 g body weight [BW]/h, n = 7 v 83 +/- 15 ng/100 g BW/h, n = 7, P < .01) and beta-hexosaminidase (21 +/- 3 mU/100 g BW/h, n = 7 v 8.3 +/- 2 mU/100 g BW/h, n = 9, P < .01) excretion compared with controls. Thus, both Triton WR 1339 and orotic acid have an effect on dolichol metabolism, and it is conceivable--based on our results--that serum dolichol concentrations are regulated, at least in part, by a mechanism similar to that for serum cholesterol levels.     

Arterial-venous carbon dioxide tension difference after hypothermic cardiopulmonary bypass

Arterial-venous carbon dioxide tension difference (Pv-aCO2) is known to become high after severe hemorrhage shock and resuscitation. We hypothesized that Pv-aCO2 might be high after cardiac surgery because of the oxygen debt occurred during hypothermic cardiopulmonary bypass (CPB). Blood pressure, cardiac index, hemoglobin, the arterial and mixed venous blood gases were repeatedly measured every 6 hours for 24 hours following cardiac surgery in 60 adult patients who underwent hypothermic CPB. Immediately after the surgery, Pv-aCO2 was extremely high, then gradually decreased to within normal ranges 12 hours later (8.0 +/- 2.9 mmHg vs 5.9 +/- 3.1 mmHg. p < 0.01). Factors which significantly correlated to Pv-aCO2 were cardiac index, oxygen delivery, minimum rectal temperature and duration of CPB. Oxygen debt during hypothermic CPB might cause significantly high Pv-aDO2. At least 12 hours were necessary to recover from anaerobic status to physiological condition.     

The effect of furosemide on renal blood flow and renal tissue oxygen tension in dogs

The effect of furosemide (2 mg/kg i.v.) on blood flowin the renal artery and the tissue oxygen tension of the renal cortex and medulla was investigated in six dogs. The flow was measured with an electromagnetic flowmeter and the tissue oxygen tension with IBC tissue oxygen electrodes. The flow in the renal artery increased significantly (p less than 0.05) 5-15 minutes after furosemide administration. 40 minutes after the injection the flow had returned to its initial level. The tissue oxygen tension of both the cortex and the medulla showed significant elevation following furosemide administration. The maximum increase of tissue oxygen tension was recorded 10-20 minutes after the injection. The oxygen tension values exceeded the initial level by 22% in the cortex and by 48% (p less than 0.001) in the medulla. Simultaneously, with the changes in oxygen tension, urine output increase considerably and urine osmolality declined. Application of these renal effects of furosemide in clinical work, particularly in cardiopulmonary bypass surgery, is discussed.     

Endothelial dysfunction in cerebral microcirculation during hypothermic cardiopulmonary bypass in newborn lambs

OBJECTIVES: Inflammatory stimuli or mechanical stresses associated with hypothermic cardiopulmonary bypass could potentially impair cerebrovascular function, resulting in inadequate cerebral perfusion. We hypothesize that hypothermic cardiopulmonary bypass is associated with endothelial or vascular smooth muscle dysfunction and associated cerebral hypoperfusion. Therefore we studied the cerebrovascular response to endothelium-dependent vasodilator, acetylcholine, endothelium-independent nitric oxide donor, sodium nitroprusside, and vasoactive amine, serotonin, in newborn lambs undergoing hypothermic cardiopulmonary bypass (nasopharygeal temperature = 18 degrees C). METHODS: Studies were performed on 13 newborn lambs equipped with a closed cranial window, allowing for direct visualization of surface pial arterioles. Six animals were studied while undergoing hypothermic cardiopulmonary bypass, whereas seven served as nonbypass, warm (37 degrees C) controls. Pial arteriolar caliber (range = 111 to 316 microm diameter) was monitored using video microscopy. RESULTS: Topical application of acetylcholine caused a dose-dependent increase in arteriolar diameter in the control group that was absent in animals undergoing hypothermic cardiopulmonary bypass. Hypothermic cardiopulmonary bypass did not alter the vasodilation in response to sodium nitroprusside. Furthermore, the contractile response to serotonin was fully expressed during hypothermic cardiopulmonary bypass. CONCLUSIONS: The specific loss of acetylcholine-induced vasodilation suggests endothelial cell dysfunction rather than impaired ability of vascular smooth muscle to respond to nitric oxide. It is speculated that loss of endothelium-dependent regulatory factors in the cerebral microcirculation during hypothermic cardiopulmonary bypass may enhance vasoconstriction, and impaired cerebrovascular function may be a basis for associated neurologic injury during or after hypothermic cardiopulmonary bypass.