The critical velocity and 1500-m surface performances in Finswimming.

The purpose of this investigation was to determine whether the concepts of critical velocity (CV) and anaerobic swimming capacity (ASC) could be used by coaches as a reliable index in order to monitor 1500-m Surface (SF) performances in Finswimming. Thirteen Finswimmers (6 males and 7 females, 24+/-6 years), members of the Japanese national team, were instructed to swim three different swimming distances (400-, 800-, and 1500-m) at maximal effort in a 50m long course swimming pool. CV and the ASC were calculated using 400-m and 800-m swim times. Mean height and body mass were 170.2 cm and 69.7 kg in male and 160.5 and 61.0 kg in female. A highly positive correlation was found between the CV and the mean velocity of 1500-m SF (V1500) (r=0.91, P<0.01), but no correlation was found between the ASC and V1500. (r=0.46, P=0.11). However, a high correlation was found between the ASC and the residual error of V1500, calculated from the relationship between V1500 and the CV (r=0.89, P<0.01). These results suggest that the CV is a useful method for evaluating 1500-m SF performance and an aerobic performance expressed as the CV contributes to 1500-m SF performance.

Isoflurane and sevoflurane during reperfusion prevent recovery from ischaemia in mitochondrial KATP channel blocker pretreated hearts.

BACKGROUND AND OBJECTIVE: Inhalation anaesthetics given only during post-ischaemic reperfusion have some protective effect against reperfusion injury in the heart. Adenosine triphosphate-regulated mitochondrial potassium channels have been shown to be an important mediator of cardioprotection. Thus, we investigated whether 5-hydroxydecanoate, a putative mitochondrial potassium channel blocker, prevents the cardioprotective effect of volatile anaesthetics. METHODS: Forty rats were randomly allocated to four groups of equal size: control group, 5-hydroxydecanoate group, 5-hydroxydecanoate + sevoflurane group and 5-hydroxydecanoate + isoflurane group. Seven minutes after the start of perfusion, normal saline (control group) or 5-hydroxydecanoate (the other groups) was administered. Ten minutes after the start of perfusion, the heart was rendered globally ischaemic for 10 min. One minute before the end of the ischaemic period, 2.7% sevoflurane or 1.4% isoflurane were administered in the 5-hydroxydecanoate + sevoflurane or 5-hydroxydecanoate + isoflurane groups respectively. The heart was reperfused for 10 min. RESULTS: Adenosine triphosphate content at the end of reperfusion in the 5-hydroxydecanoate + sevoflurane group was significantly lower (P < 0.05) than those in the control and the 5-hydroxydecanoate + isoflurane groups (19.9 +/- 8.7, 28.1 +/- 3.4 and 30.4 +/- 2.3 micromol g(-1), respectively). In addition, the combination of inhalation anaesthetics and 5-hydroxydecanoate decreased the ratios of recovered hearts from ischaemia (5-hydroxydecanoate + sevoflurane group: 40%, 5-hydroxydecanoate + isoflurane group 50%). CONCLUSION: 5-hydroxydecanoate alone caused no significant changes in haemodynamics and myocardial metabolism. However, the combination of 5-hydroxydecanoate and volatile anaesthetics impaired the recovery from ischaemia. Although animal data cannot be extrapolated to human beings, we suggest that more attention be paid to patients on sulphonylurea drugs, which inhibit potassium channels, when they are anaesthetized with volatile anaesthetics.

Effects of nicorandil on myocardial function and metabolism in the post-ischaemic reperfused heart with or without inhalation anaesthetics.

BACKGROUND: Nicorandil, which is an ATP-sensitive K channel opener, has been reported to protect the ischaemic myocardium. However, its interaction with inhalation anaesthetics on the ischaemic myocardium has not been well elucidated. So, we have investigated whether isoflurane or sevoflurane modify the effects of nicorandil on cardiac function and metabolism in the rat heart-lung preparation. METHODS: Animals were allocated to 4 groups as follows: Control group, no drug; Nic group, nicorandil; Nic+Iso group, nicorandil and isoflurane; Nic+Sev group, nicorandil and sevoflurane. Seven minutes after the start of perfusion, nicorandil was administered and 10 min after the start of perfusion, the heart was rendered globally ischaemic for 10 min, and then the heart was reperfused for 10 min. RESULTS: LVdP/dt max in the Nic group was higher than those in the other groups. Right atrial pressure in the Nic+Iso and Nic+Sev groups was significantly higher than in the Control and Nic groups. Myocardial ATP in the Nic group was higher than in the other groups. DHBA levels in the perfusate in the Nic and Nic+Iso groups were lower than those in the Control and Nic+Sev groups, but those in the Nic+Sev group were higher than those in the other groups. CONCLUSIONS: Nicorandil improved post-ischaemic cardiac function and preserved high-energy phosphates. However, these beneficial effects of nicorandil were abolished by the combination with isoflurane or sevoflurane. In addition, sevoflurane increased hydroxyl radical formation in the post-ischaemic reperfused heart.

Effects of glibenclamide on hydroxyl radical formation in the postischaemic reperfused heart with or without inhalation anaesthetics.

BACKGROUND AND OBJECTIVE: We investigated whether glibenclamide (glyburide) affects myocardial metabolism and hydroxyl radical formation in the rat heart-lung preparation with or without inhalation anaesthetics. METHODS: Thirty-seven male Wistar rats were allocated to four groups: (a) control group (C), received vehicle only; (b) group G, received glibenclamide 10 microM L-1; (c) group I, received glibenclamide 10 microM L-1 and 1.4% isoflurane during perfusion; (d) group S, received glibenclamide 10 microM L-1 and 2.7% sevoflurane during perfusion. Glibenclamide was administered 7 min after the start of perfusion. Ten minutes later, the heart was rendered globally ischaemic for 10 min by reducing the preload and afterload to zero, and then the heart was reperfused for 10 min. The formation of hydroxyl radicals in perfusate blood and heart was measured with high performance liquid chromatography using salicylic acid. Hydroxyl radicals react with salicylic acid, yielding dihydroxybenzoic acids. RESULTS: The recovery time from ischaemia in group G was significantly longer than the other groups. However, there were no differences in myocardial metabolites and dihydroxybenzoic acids concentrations in the perfusate and heart among the four groups. CONCLUSIONS: Glibenclamide prolonged recovery time from ischaemia, but did not affect hydroxyl radical formation in the postischaemic reperfused heart. In addition, isoflurane and sevoflurane shortened this time. These facts suggest that mechanisms other than effects of volatile anaesthetics on hydroxyl radical formation are responsible for their protective effects in this model.

In vivo antitumour efficacy of MGI-114 (6-hydroxymethylacylfulvene, HMAF) in various human tumour xenograft models including several lung and gastric tumours.

MGI-114 (6-hydroxymethylacylfulvene, HMAF) is a semi-synthetic analogue of the cytotoxic sesquiterpenoid illudins. In the present study, the in vivo antitumour efficacy of MGI-114 was examined in a panel of human tumour xenograft models consisting mainly of human lung and gastric tumours, and compared with that of other antitumour drugs such as irinotecan, paclitaxel, cisplatin, doxorubicin, vindesine, etoposide and 5-fluorouracil (5-FU). When different administration schedules were compared, daily administration of MGI-114 was found to be more effective than intermittent administrations. In human tumour xenograft models of nasopharyngeal, breast and colon carcinoma and melanoma, MGI-114 exerted a strong antitumour activity with complete tumour regression being observed. Moreover, in four human lung and three gastric tumour xenograft models, MGI-114 showed a strong antitumour activity with complete tumour regression being observed in some of the models. The antitumour efficacy of MGI-114 was generally higher than or equivalent to that of other antitumour drugs such as irinotecan and paclitaxel. These results support the potential utility of MGI-114 in the treatment of a variety of human solid tumours.

Role of heparin and nitric oxide in the cardiac and regional hemodynamic properties of protamine in conscious chronically instrumented dogs.

BACKGROUND: Because protamine is administered to reverse heparin, a drug that might itself affect the pharmacologic properties of protamine, this study was designed to assess the properties of protamine alone and in the presence of heparin in conscious dogs. METHODS: Twelve dogs were instrumented to continuously record cardiac and regional hemodynamics. On separate occasions, a dose of protamine (0.5, 1, 3, 5, and 8 mg/kg) was randomly administered either alone or in the presence of heparin (ratio 100 IU/mg). Heparin (300 IU/kg) and protamine (3 mg/kg) were administered in the presence of N-methyl-L-arginine, a specific nitric oxide synthase inhibitor. Identical experiments were performed with protamine (8 mg/kg) in the absence of heparin on a separate occasion. RESULTS: Protamine alone produced limited cardiac and regional changes. In the presence of heparin, protamine produced hypotension at 3, 5, and 8 mg/kg, vasodilatation at 3 and 5 mg/kg, and a more pronounced dose-dependent increase in pulmonary pressure at 3, 5, and 8 mg/kg. Simultaneously, transient carotid vasodilatation at 3 and 5 mg/kg, coronary and hepatic vasodilatation at 3, 5, and 8 mg/kg, as well as a decrease in vertebral vascular resistance were recorded at 1, 3, and 8 mg/kg. Protamine produced an immediate increase followed by a secondary decrease in renal vascular resistance. Protamine-induced secondary pulmonary pressor effects were attenuated. In the presence of heparin, nitric oxide synthase blockade selectively attenuated protamine-induced immediate hypotension, systemic vasodilatation, and coronary, mesenteric, and hepatic vasodilations as well as the decrease in portal blood flow and accentuated the renal vasoconstriction. CONCLUSIONS: The presence of heparin accentuated the decrease in cardiac function induced by protamine as well as its effects on regional circulation. The data provide evidence that the nitric oxide pathway is involved in the systemic and selective regional heparin-protamine-mediated vasodilatation in conscious dogs.

Effects of propofol on ischemia and reperfusion in the isolated rat heart compared with thiamylal.

The aim of the present study was to investigate whether clinical doses of propofol and thiamylal affect oxygen free radical production and intracellular calcium concentration ([Ca2+]i) in the post-ischemic reperfused heart. Forty-eight rat hearts were perfused with a Langendorff system and loaded with Fura-2 / AM as a [Ca2+]i marker. The hearts were divided into 6 groups as follows (each group: n = 8); Group S (saline), Group TL (thiamylal 100 microM), Group TH (thiamylal 300 microM), Group I (Intralipid), Group PL (propofol 3 microM), and Group PH (propofol 10 microM). All hearts were initially perfused for 5 min as control aerobic perfusion. Afterwards, no-flow ischemia was induced for 15 min, followed by reperfusion for 20 min. The formation of hydroxyl radicals in the coronary effluent was measured with high performance liquid chromatography using salicylic acid. At the beginning of the ischemia and reperfusion periods, increases in systolic and diastolic [Ca2+]i were observed in all groups except Group TH. The high dose of thiamylal significantly suppressed this initial increase in cytosolic [Ca2+]i (Group S 1.30+/-0.15; Group TL 0.99+/-0.17; Group TH 0.70+/-0.09, at 1 min after reperfusion; systolic [Ca2+]i : p < 0.05). Total DHBAs in the coronary effluent of all groups increased significantly 1 min after reperfusion, however, there were no significant differences among the groups. Clinical doses of propofol had no significant effect on myocardial function and [Ca2+]i before and after ischemia, whereas thiamylal suppressed the increase in [Ca2+]i during ischemia and reperfusion. However, free radical formation during reperfusion was unaffected by thiamylal and propofol.

[The influence of N-nitro-L-arginine methyl ester (L-NAME) on hydroxyl free radical formation in the post-ischemic reperfused heart of anesthetized rats].

N-nitro-L-arginine methyl ester (L-NAME) has been reported to have protective action against hydroxyl free radicals. We have investigated whether L-NAME influences free radical formation in the post-ischemic reperfused heart of anesthetized rats. An isolated rat heart-lung preparation was used. Forty male Wistar rats were allocated into D (D-NAME 100 microMol.l-1), L (L-NAME 100 microMol.l-1), LH (L-NAME 100 microMol.l-1 and 1MAC halothane), LI (L-NAME 100 microMol.l-1 and 1MAC isoflurane), and LS (L-NAME 100 microMol.l-1 and 1MAC sevoflurane) groups. The heart was perfused initially at the cardiac output of 30 ml.min-1 and the atrial pressure of 70 mmHg. Drugs were administered into the reservor 7 min after the start of perfusion. Ten minutes after the start of perfusion, the heart was rendered globally ischemic for 10 min by reducing the preload and afterload to zero and then reperfused for 10 min. At the end of reperfusion, the heart was freeze-dried for 4 days. The perfusate blood was collected just before and after ischemia and at the end of reperfusion. The formation of hydroxyl radicals in the perfusate blood and heart was measured with high-performance liquid chromatography using salicylic acid. Hydroxyl radicals react with salicylic acid, yielding dihydroxybenzoic acid (DHBA). Before and after ischemia, there were no significant differences among the groups in cardiac output, systolic pressure, heart rate, and right atrial pressure. DHBAs in the heart of L, LH, LI, and LS groups were significantly lower than those of D group. However, there were no differences in the DHBA levels among 4 groups. The concentrations of DHBA in the perfusate blood after ischemia and reperfusion were significantly higher than those before ischemia in all groups. DHBAs in the perfusate blood after ischemia and reperfusion of L, LH, LI, and LS groups were significantly lower than those of D group. However, there were no differences in the DHBA levels among 4 groups administered L-NAME. This study indicates that L-NAME reduces hydroxyl free radical formation in the post-ischemic reperfused heart in anesthetized rats and volatile anesthetics do not influence the depressant effect of hydroxyl free radical formation by L-NAME.

A case of awareness with sevoflurane and epidural anesthesia in ovarian tumorectomy.

We experienced a case of awareness during ovarian tumorectomy in a patient who was anesthetized with sevoflurane and epidural anesthesia. A 74-year-old woman was scheduled for resection of an ovarian tumor. After epidural catheter insertion, anesthesia was induced with 60 mg of propofol and 6 mg of vecuronium, and anesthesia was maintained with epidural anesthesia (1% mepivacaine), 1 to 2% sevoflurane, and 66% nitrous oxide in oxygen. The operative course was uneventful and the total operation time was 2 hours and 50 minutes. Two days after the operation, we were surprised to learn that the patient complained of awareness during the surgery.

Measurement of cardiac function in conscious rats.

The goal of this study was to establish that 1. blood velocity profile in the rat aorta is parabolic, and 2. measure of left ventricular thickening fraction can be used in rats. Spontaneously hypertensive and normotensive Wistar Kyoto rats were instrumented with a 20-MHz pulsed Doppler flow probe around the thoracic aorta and a 20-MHz pulsed Doppler thickening probe on the left ventricle. Doppler frequency shifts were measured throughout the entire aorta diameter, and individual blood velocity profiles were constructed. It was demonstrated that blood velocity in the ascending aorta of rats is laminar; therefore, cardiac output can be measured using the pulsed Doppler method. In Wistar Kyoto rats, left ventricular thickening fraction was 24 +/- 1% and 25 +/- 1%, 2 and 3 weeks following surgery. In spontaneously hypertensive rats, left ventricular thickening fraction was 22 +/- 2%. Halothane depressed left ventricular thickening fraction, whereas isoproterenol increased left ventricular thickening fraction in conscious rats. Thus, pulsed Doppler technique is a valuable tool for evaluating cardiovascular function in conscious rats.

Effects of cibenzoline on cardiac function and metabolism in the rat heart--lung preparation.

BACKGROUND AND OBJECTIVE: Although there is concern that cibenzoline, an antidysrhythmic drug for the treatment of ventricular and supraventricular dysrhythmias, may be associated with dose-dependent inhibition of myocardial contractility there are few reports about the relationship between myocardial metabolism and cardiac function when it is used. The present study was designed to investigate the effects of cibenzoline on cardiac function and metabolism. The effects of cibenzoline on cardiac function and myocardial metabolism were assessed in the isolated rat heart-lung preparation. METHODS: Thirty-two male Wistar-ST rats were divided into four groups: control, and those to receive cibenzoline, either 300, 900 or 3000 ng mL(-1). The cibenzoline was administered into the perfusate 5 min after the start of perfusion. Heart rates in the 3000 ng mL(-1) group were significantly lower than those in the control group. Cardiac output in the 3000 ng mL(-1) group at 15 and 30 min was significantly lower than in the control group. In all groups, values for %LV dP/dt max (the ratio of values at each time to those at 5 min) at 20, 25, 30 min were significantly higher than at 5 min. Myocardial adenosine triphosphate concentration in the 3000 ng mL(-1) group was significantly lower than in controls. There was no difference between groups in the lactate/pyruvate ratio. CONCLUSION: The therapeutic range of cibenzoline has few effects on cardiac function and metabolism, although concentrations 10 times greater may cause a deterioration in myocardial metabolism.

Effects of epidural fentanyl and intravenous flurbiprofen for visceral pain during cesarean section under spinal anesthesia.

PURPOSE: Despite adequate levels of sensory blockade, patients sometimes complain of abdominal pain during cesarean section performed under spinal anesthesia. The aim of this study was to evaluate the effects of epidural fentanyl and intravenous flurbiprofen on visceral pain during cesarean section in patients having spinal anesthesia. METHODS: Thirty ASA physical status I and II patients undergoing elective cesarean section were studied. Spinal-epidural anesthesia was performed in all groups. Group A received no additional analgesics, group B received epidural fentanyl 100 mug, and group C received flurbiprofen 50 mg i.v. immediately after the delivery. Postdelivery, intraoperative visceral pain was evaluated by using the visual analog scale. Incidence and visual analog scale scores of visceral pain and incidence of intraoperative nausea and vomiting were obtained from each patient. RESULTS: Visual analog scale scores of pain were significantly lower in group B than in the other groups (P < 0.05). The incidence of nausea was comparable in all groups. The incidence of intraoperative vomiting was lower in group C than in the other groups (P < 0.05). CONCLUSION: Epidural fentanyl, but not intravenous flurbiprofen, decreases the incidence and severity of visceral pain during cesarean section.

Sevoflurane reduces dysrhythmias during reperfusion in the working rat heart.

PURPOSE: The effects of sevoflurane on myocardial reperfusion injury have not been well studied. The purpose of this study was to determine the effects of sevoflurane on myocardial function, arrhythmia, and metabolism during reperfusion in an isolated working rat heart model. METHODS: Thirty-two hearts were divided into four groups according to the timing of 2.5% sevoflurane administration: group I, control, no sevoflurane; group II, sevoflurane administered only before ischemia; group III, sevoflurane only during reperfusion; group IV, sevoflurane during the whole study period. Myocardial contractility, myocardial ATP, lactate, and glycogen levels were assessed in the reperfusion period following global heart ischemia of 15 min duration. The incidence and duration of ventricular fibrillation were also observed in the reperfusion period. RESULTS: There was no difference in cardiac output and left ventricular dP/ dt max among the four groups at 10, 15, and 20 min after reperfusion. There was no difference in myocardial ATP, lactate and glycogen contents between the groups. The incidences of ventricular fibrillation during reperfusion were 100%, 63%, 100%, and 25% (P < 0.05 vs control), and the durations of ventricular fibrillation during reperfusion were 375 +/- 269, 104 +/- 98 (P < 0.05 vs control), 303 +/- 189, and 93 +/- 245 (P < 0.05 vs control) in groups I, II, III, and IV, respectively (mean +/- SD). CONCLUSION: The administration of sevoflurane prior to reperfusion appears to provide myocardial protection, as assessed by reduced dysrhythmias during reperfusion.

[A multicenter study of cardiac events and anesthetic management of patients with ischemic heart diseases for noncardiac surgery--a follow-up report].

We have conducted a joint research project to investigate the incidence of ischemic heart disease in patients for noncardiac surgery and to define the risk of perioperative cardiac complications in these patients. From September to November 1997 we had 7288 patients scheduled to undergo noncardiac surgery in the 8 departments of anesthesiology. Of these patients, 228 (3.1%) patients had ischemic heart disease, and 30 of them (13.2%) developed perioperative cardiac events. Critical cardiac events, including perioperative myocardial ischemia and lethal arrhythmia, occurred in 7 of these patients. In our region of Japan, 3-4% of surgical patients tend to develop ischemic heart disease and 3.1% of them demonstrated severe cardiac complications perioperatively. Compared with United States we encounter fewer surgical patients with ischemic heart disease, but the risk of developing perioperative cardiac complications in such patients is almost the same for both countries.

[Multicenter study of cardiac events and anesthetic management of patients with ischemic heart diseases undergoing noncardiac surgery].

We designed a joint research project to investigate the incidence of ischemic heart diseases in patients undergoing noncardiac surgery and to define the risk of perioperative cardiac complications in these patients. Of the 8358 surgical patients in the 8 departments of anesthesiology between March 1997 and June 1997, 328 (3.9%) had ischemic heart diseases. Among the 328 patients, 54 (16.4%) developed perioperative cardiac events, including myocardial infarction (3 patients) and either lethal or potentially dangerous dysrhythmias (51 patients). Preoperative cardiac assessments were performed while the anesthetic techniques including intensive monitoring and perioperative prophylactic therapy were also employed. Patients with ischemic heart diseases received various types of preoperative evaluation to identify the degree of coronary artery disease and to assess the overall cardiac function. The patients were monitored using a multilead electrocardiogram, an arterial line, a central venous catheter, a pulmonary artery catheter, and by transesophageal echocardiography intraoperatively. Therapeutically, isosorbide, nitroglycerin, beta-blockers, calcium channel blockers, and/or nicorandil were administered to prevent perioperative ischemia. So far, no generally accepted management strategies have been established in patients with cardiovascular disorders based on large-scale outcome trials in Japan. Therefore, nationwide large multicenter trials are awaited with interest in order to establish helpful guidelines to improve the perioperative management and to reduce ischemia in cardiac patients undergoing noncardiac surgery.

[Premixing of 5% dextrose in Ringer's acetate solution with propofol reduces incidence and severity of pain on propofol injection].

The purpose of this study was to compare the effect of premixed 5% dextrose in Ringer's acetate solution and premixed lidocaine with propofpl on the reduction of pain during injection of propofol in adult patients. We conducted a prospective, randomized, double-blinded trial. Ninety-six patients were randomly allocated to one of three groups according to the agents added to 1% propofol 20 ml; Group C, normal saline 2 ml, Group L, 2% lidocaine 2 ml, and Group A, 5% dextrose in Ringer's acetate solution 2 ml. The pain on injection was rated as none, mild, moderate, or severe. Seventy percent of patients in the C group experienced pain, while 33% and 25% of patients experienced pain in the A group and the L group, respectively. Forty-two percent of patients in the C group complained moderate to severe pain but only one patient in both A group and L group. In conclusion, 5% dextrose in Ringer's acetate solution premixed with 200 mg propofol significantly reduces incidence and severity of pain associated with propofol injection and is easier to use than premixed lidocaine.

Hydroxyl radical formation during inhalation anesthesia in the reperfused working rat heart.

PURPOSE: To determine whether isoflurane, sevoflurane and halothane influenced hydroxyl radical production in the ischemic rat heart. METHODS: Twenty-four male Wistar rats were divided into four groups; control (C), isoflurane 1.4% (I), sevoflurane 2.5% (S) and halothane 1% (H). The hearts were perfused with modified Krebs-Henseleit bicarbonate buffer by a working heart model for 10 min. Then, whole heart ischemia was induced by severely restricting coronary perfusion for 15 min. Reperfusion of the hearts after this ischemic period lasted for 20 min. The coronary effluent was collected before and during ischemia and at 1, 5, 10, 20 min after reperfusion. At the end of reperfusion, hearts were removed and prepared for measurement. Hydroxyl radicals were identified by their reaction with salicylic acid to yield dihydroxybenzoic acids (DHBAs). RESULTS: Before and after ischemia, there were no differences in coronary flow and heart rate among the four groups, but cardiac output and LV dP/dt maximum in the anesthetic groups were lower than in the control group. Hydroxyl radical products in the heart were significantly lower in the I group than the other groups (e.g. C vs I, 278.1 +/- 24.3 vs 219.3 +/- 14.4 microM x g(-1), P < 0.05). The concentrations of DHBAs in the coronary effluent at some points in the I and H groups were less than in the C and S groups. CONCLUSION: These results indicate that isoflurane and halothane (to a lesser extent), reduce hydroxyl radical production in the ischemic heart, but sevoflurane does not.

Effects of melatonin and superoxide dismutase on free radical formation in the postischemic reperfused heart.

PURPOSE: Melatonin has been reported to protect against oxygen free radicals. We investigated whether melatonin or superoxide dismutase (SOD) would decrease hydroxyl radical concentration in the postischemic reperfused heart. METHODS: An isolated rat heart-lung preparation was used. Eighty-one male Wistar rats were allocated into control (no drug), S1 (SOD 400 U.ml(-1)), S2 (SOD 2000 U.ml(-1)), M1 (melatonin 0.1 microg.ml(-1)), M2 (melatonin 1.0 microg.ml(-1)), M3 (melatonin 10 microg.ml(-1)), SM (SOD 400 U.ml(-1) and melatonin 1.0 microg.ml(-1)) groups. The heart was perfused initially at the cardiac output of 30 ml.min(-1) and the mean arterial pressure of 70 mmHg. Drugs were administered into the reservoir 7 min after the start of perfusion. Ten minutes after the start of perfusion, the heart was rendered globally ischemic for 10 min by reducing the preload and afterload to zero and then reperfused for 10 min. At the end of reperfusion, the heart was freeze-dried for 6 days. The perfusate blood was collected just before and after ischemia and at the end of reperfusion. The formation of hydroxyl radicals in perfusate blood and heart was measured with high-performance liquid chromatography using salicylic acid. Hydroxyl radicals react with salicylic acid, yielding 2,3-, 2,4-, 2,5-, and 3,4-dihydroxybenzoic acid (DHBA). RESULTS: Before and after ischemia, there were no significant differences among the groups in cardiac output, systolic pressure, heart rate, and right atrial pressure. The concentrations of DHBAs in the perfusate blood and heart after ischemia and reperfusion in all groups were significantly higher than those before ischemia. DHBAs in the heart of all drug-administered groups were significantly lower than those in the control group. In the perfusate blood, DHBAs in the S2 group were significantly lower than those in the control group. CONCLUSIONS: SOD and melatonin decrease hydroxyl radical concentration in the postischemic reperfused heart.