Studies of the effects of hypothermia on regional myocardial blood flow and metabolism during cardiopulmonary bypass

Myocardial high-energy phosphate replenishment during ischemic arrest: aerobic versus anaerobic metabolism

An in vivo, isolated pig heart preparation was used to study the effect of L-glutamate added to crystalloid and blood potassium cardioplegia on the myocardial high-energy phosphate compounds, adenosine triphosphate (ATP) and creatine phosphate (CP). Studies were performed during a three-hour arrest interval and during 60 minutes of reperfusion. Levels of ATP remained at or above control levels during arrest in animals receiving either unmodified blood or glutamate-enriched crystalloid cardioplegia. While glutamate significantly improved the ability of the crystalloid solution to preserve ATP during arrest, when added to blood, it contributed to a depressed ATP after a three-hour arrest. Creatine phosphate declined during arrest in all animals, but those receiving unenriched blood cardioplegia consistently had the highest levels (p less than 0.05). Addition of glutamate to crystalloid cardioplegia provided a significantly (p less than 0.05) higher level of CP at the end of three hours of arrest, which was still lower than that noted with unenriched blood. Comparable to its effect on the ATP level, when glutamate was added to blood cardioplegia, a decrease (p less than 0.05) in CP was noted after three hours of arrest. Attempts to enhance high-energy phosphate production by supplementing blood cardioplegia with L-glutamate are ineffective, while increased high-energy phosphate production results when glutamate is added to crystalloid cardioplegia. This implies that L-glutamate functions where anaerobic and not aerobic metabolism is the major component of preservation. With reperfusion, the only group of animals displaying depressed levels of ATP and CP was that receiving glutamate-enriched blood cardioplegia.

Pressurized rapid cardioplegia versus administration of exogenous substrate and topical hypothermia

Five hundred fifty-nine patients undergoing aortocoronary operation were analyzed retrospectively according to the type of myocardial protection implemented during the period of ischemia. In Group 1 (253 patients), a rapid method of hypothermic cardioplegia alone was utilized. In Group 2 (306 patients), slower infusion of the same solution with topical hypothermia was implemented. Cardiac isoenzymes (CPK-MB, LDH1, LDH2, serum glutamic oxaloacetic transaminase [SGOT]) and myocardial infarct index (MII) were measured postoperatively for 48 hours. Immediately after operation, a significant difference was found between Groups 1 and 2 in the CPK-MB isoenzyme mean value levels--12.1 versus 18.6 IU, p less than 0.01--and MII mean values--5.2 versus 8.1, p less than 0.01. CPK-MB variances between subgroups receiving two, three, and four grafts were significantly different in favor of Group 1. Differences were also found in LDH1, LDH2, total lactic dehydrogenase (LDH), and SGOT: Group 2 levels were significantly higher than in Group 1. There were ten intraoperative infarctions in Group 2 and none in Group 1. In 45% of the patients in Group 2, inotropic agents were necessary in the postoperative period versus none in Group 1. Spontaneous cardiac rhythm following ischemia occurred in 89.7% of the patients in group 1 versus 29% in Group 2. A method of pressurized high-flow rapid cardioplegia with intermittent reperfusions alone, seems to provide adequate protection of the myocardium during ischemia over a slower low-flow method of infusion combined with topical hypothermia.

Myocardial preservation during anoxic arrest. Experimental model ventricular fibrillation

An experimental model with anaesthetized healthy mongrel dogs on extracorporeal circulation is described. Anaesthesia and cardiopulmonary bypass are the same as used in clinical practice. Various methods of myocardial preservation were investigated and their protective effect was judged by cardiac performance after termination of 60 min of anoxic arrest. In this study, the first part of an experimental series, electrically-induced fibrillation during 60 min of normothermic and local hypothermic anoxic arrest was investigated. In group I, the hearts were fibrillated immediately after cross-clamping. In group II, which served as controls, the hearts were allowed to fibrillate spontaneously after aortic cross-clamping. All the hearts in group I went into an ischaemic contracture, whereas those in group II showed a 50% recovery, but with a strongly reduced cardiac performance after termination of anoxic arrest and cardiopulmonary bypass. Measurements of myocardial surface pH demonstrated a rapidly developed acidosis during the period of anoxic arrest. The most impressive finding by light microscopy was pronounced myocardial oedema. External cooling by 4 degrees C glucose 5.5% continuously flushed into the pericardial sac in combination with electrically-induced fibrillation proved to be ineffective as a protective method. None of the eight dogs in this group survived. External cooling combined with intraventricular injection of 4 degrees C glucose 5.5% seemed to protect the hearts against ischaemic damage, insofar that all six hearts in this group were able to take over the circulation after declamping. The working capacity was, however, impaired and a relatively long period of mechanical support and stimulation with inotropic drugs was necessary.

A paracorporeal rat heart model for ischemic and reperfusion studies

Paracorporeal rat hearts were supplied with blood derived from the abdominal aorta of a supporting rat. The circulatory stability and working capacity of the supporting animal was analyzed in the experimental situation in terms of PO2, SO2, PCO2, HCO3, pH and electrolytes, all of which were within the normal range before and during a 60 min period of paracorporeal perfusion. For evaluation of ischemic damage in this model studies were made on three groups of excised hearts. They were subjected to 10, 15 or 20 min of complete global ischemia at 37 degrees C (ambient temperature) and reperfused for 30 min, including ECG, observations of contractility and an analysis of creatine kinase efflux in the coronary effluent. The results showed good reproducibility and the data were in accordance with reports from similar studies on Langendorff preparations. The model, which is easily set up, inexpensive and based upon pulsatile blood perfusion, should be more physiologic than the conventional Langendorff preparation.

Myocardial flow distribution. II : Empty beating heart, ventricular fibrillation and cardiac arrest

Myocardial oxygen consumption (MVO2) and coronary blood flow (CBF) distribution were studied in 21 isolated, metabolically supported dog hearts. Measurements of MVO2 and CBF distribution were carried out in three different experimental conditions : empty beating heart (EBH), ventricular fibrillation (VF) and high potassium-induced cardiac arrest (CA). MVO2 was approximately the same in EBH and VF (4.09 +/- 0.77 and 4.28 +/- 0.68 ml O2 min-1 100 g-1 respectively), and significantly lower in the group with CA (2.40 +/- 0.18 ml O2 min-1 100 g-1, P less than 0.05). Total CBF showed no significant differences among the three groups (84 +/- 7 ml/min in EBH; 78 +/- 7 ml/min in VF and 83 +/- 7 ml/min in CA). Subendocardial CBF per unit of tissue mass was significantly lower in hearts with VF (0.43 +/- 0.01 ml/min-1 g-1, P less than 0.05) when tested against the other two groups of experiments (0.69 +/- 0.03 ml min-1 g-1 in EBH and 0.65 +/- +/- 0.04 ml min-1 g-1 in CA). This was also reflected in the endo/epi ratio, that was significantly lower in VF (1.41 +/- 0.07, P less than 0.05) with respect to the other two groups (2 +/- 0.09 in EBH and 2.21 +/- 0.07 in CA). From data presented here we can conclude that cardioplegia, even in absence of hypothermia, is a method that will assure myocardial protection providing : (1) a lower subendocardial MVO2; (2) a higher subendocardial CBF, which helps for a prompt recovery during reperfusion.

Myocardial protection during aortic valve replacement. Cardiac metabolism and enzyme release following continuous blood cardioplegia

Cardiac metabolism following hypothermic potassium cardioplegia with blood as cardioplegia vehicle was studied in two groups of patients undergoing aortic valve replacement. In 15 patients, blood was given as single dose infusion (single dose group) and in 18 patients the same initial bolus was followed by a continuous perfusion (25-30 ml/min) with modified blood from the heart-lung machine (continuous blood group). Simultaneous samples were drawn from arterial and coronary sinus blood before and during the first 60 min after cardioplegia. In the continuous blood group, samples were also drawn during the period of cardioplegic perfusion. The samples were analyzed for PO 2, O2-saturation and content, PCO2, pH, lactate, pyruvate, glucose, potassium, myoglobin, creatine kinase (CK), its isoenzyme MB, and aspartate aminotransferase (ASAT). In addition myoglobin and enzymes were followed in peripheral venous blood for 24 hours. Myocardial biopsies were taken from the left ventricle at the beginning and end of cardioplegia and analyzed for adenosine triphosphate (ATP), creatine (C) and creatinephosphate (CP). The pattern of metabolic changes after cardioplegia was similar in both groups with decreased myocardial oxygen extraction, marked lactate and potassium release, increased glucose uptake and significant enzyme and myoglobin release. However, the degree of changes was significantly smaller in the continuous blood group. The myocardial biopsies also showed significantly less ATP and CP decrease in the continuous blood group, suggesting, together with the other metabolic results, that the myocardial protection afforded by continuous blood cardioplegia was superior to that of the single dose group. Furthermore, continuous perfusion permitted easy control of myocardial temperature during the period of aortic cross-clamping.

Improved oxygen delivery to the myocardium during hypothermia by perfusion with 2,3 DPG-enriched red blood cells

The oxygen affinity of red cells increases stepwise with temperature reductions below 37 degrees C. In vitro studies demonstrated that biochemically modified red cells with increased 2,3 diphosphoglycerate (2,3 DPG) (150% and 250% of normal) exhibited significantly less oxygen affinity at 24 degrees C than did unmodified cells. At 15 degrees C, significant attenuation of affinity was observed with 250%, but not 150%, of normal 2,3 DPG cells. Measurements made of isolated fibrillating dog hearts during perfusion at 24 degrees C alternately with unmodified (80% of normal 2,3 DPG) and modified (300% of normal 2,3 DPG) red cells demonstrated significantly greater oxygen consumption, higher coronary sinus partial pressures of oxygen and carbon dioxide, higher in vitro P50 values, and lower arterial and coronary sinus lactate levels during perfusion with modified as compared with unmodified cells. This evidence, indicating improved oxygen delivery to hypothermic dog hearts by red cells with 300% of normal 2,3 DPG activity, suggests that high 2,3 DPG cells might protect myocardial tissue in patients undergoing hypothermic cardiac operation.

Clinical experience with cold blood as the vehicle for hypothermic potassium cardioplegia

Intermittent cold ischemic arrest was compared with hypothermic potassium cardioplegia using cold blood as the vehicle in two consecutive series of patients having isolated coronary bypass grafting. Between January 1, 1977, and June 30, 1977, 196 patients were operated on using cold ischemic arrest. The incidence of perioperative infarction was 14.3%, and mean total myocardial ischemia time was 42 +/- 1.2 minutes. From July 1, 1977, to June 30, 1978, there were 428 operations done using cold blood with potassium. The incidence of perioperative infarction was 5.6% (p less than 0.005), and the mean total myocardial ischemic time was 80 +/- 2.1 minutes. In the five years prior to this study, the incidence of perioperative infarction was constant at 13% while operative mortality was declining from 5 to 1% and the need for postoperative myocardial support was declining also. Use of cold blood potassium cardioplegia compared with cold ischemic arrest for myocardial protection during coronary artery operations has significantly reduced the incidence of perioperative infarction while doubling cross-clamp time.

The importance of monitoring intramyocardial temperature during hypothermic myocardial protection

In 50 patients undergoing cardiac operation, hypothermic cardioplegic solution was infused into the root of the aorta immediately after aortic cross-clamping. Cardiac standstill was achieved within 1 to 3 minutes. However, monitoring of intramyocardial temperature with a needle thermistor revealed that such core cooling is unpredictable (the intramyocardial temperature achieved ranged from 7 degrees to 33 degrees C), unstable (this temperature can rise at more than 0.5 degrees C per minute), and uneven (a difference of up to 17 degrees C was observed between the intramyocardial temperature of the anterior and posterior left ventricular sites). The area supplied by the stenotic coronary artery was least protected. Monitoring of intramyocardial temperature enables one to know when supplementary cooling is indicated. We conclude that widespread differences in this temperature during cardiac operation make monitoring advisable for optimal myocardial protection.

Ventricular fibrillation during cardiopulmonary bypass: long-term effects on myocardial morphology and function

Mongrel dogs were subjected to hypothermic (28 degrees to 30 degrees C) cardiopulmonary bypass with hemodilution by 50%. In two groups of 8 dogs each, ventricular fibrillation was induced for 60 and 90 minutes, respectively, while the dogs were on bypass. A group of 6 dogs with the heart beating but nonworking served as control. Seven weeks after operation, hemodynamic measurements were made in the survivors (6 in each group) and the heart was fixed by perfusion with glutaraldehyde. Multiple transmural samples were taken from both ventricles. Light microscopy revealed solitary left ventricular scars (0.5 to 3 mm wide) in 2 hearts each from Groups 2 and 3. None of the hearts exhibited diffuse subendocardial fibrosis indicative of healed ischemic injury. All animals were hemodynamically normal. We conclude that in the nonhypertrophied heart, ventricular fibrillation up to 90 minutes with continuous bypass-sustained coronary perfusion (perfusion pressure at or above 70 mm Hg) offers protection from permanent myocardial injury.

Cardia lymph in electrical ventricular fibrillation: an experimental study

The flow velocity of cardiac lymph during electrical ventricular fibrillation under normothermic cardiopulmonary bypass was studied experimentally in dogs. The time needed for the cardiac lymph node to become stained after injection of dye into the apex myocardium of the left ventricle was measured as an indicator in determining flow velocity of cardiac lymph. The flow velocity was markedly decelerated immediately after the commencement of electrical ventricular fibrillation. It was accelerated, however, after 2 hours of continuous electrical ventricular fibrillation. The difference between the two values was significant (p less than 0.01). Absent contractility of the heart influenced the deceleration of flow velocity of cardiac lymph immediately after the commencement of electrical ventricular fibrillation. Acceleration after 2 hours involved stasis of cardiac lymph as a result of absent contractility and increment of lymph production due to the nonphysiological condition of the myocardium.

Intraoperative changes in coronary resistance during aortic valve replacement

Coronary vascular resistance was investigated in 10 patients undergoing aortic valve replacement using continuous constant-pressure coronary perfusion at 32 degrees C. After coronary flow was initiated, resistance was low but increased steadily until it reached a certain resting level. The plateau was attained faster after a short period of anoxia than after a longer period. The initial postischemic resistance was dependent on the duration preceding anoxia, being of the same magnitude after short and moderate periods of anoxia but significantly higher after a long period. This resistance difference between the groups lasted for the whole perfusion. The total coronary resistance and flow reached a plateau in 30 minutes, while resistance increased threefold but flow decreased to half of the initial postanoxia flow. Our results indicate the importance of initiating coronary perfusion soon after aortic cross-clamping to avoid increase in the initial vascular resistance and subsequent inadequate myocardial flow.

Left ventricular subendocardial necrosis

The hearts of as many as 90% of patients who die after open-heart operations have left ventricular subendocardial necrosis. This form of myocardial infarction depresses myocardial performance postoperatively and may result in late myocardial fibrosis. It occurs without anatomical obstruction of the coronary arteries and is caused by a discrepancy between subendocardial oxygen supply and demand during the perioperative period. This review of subendocardial necrosis summarizes the author's current understanding of: (1) why the subendocardium is especially vulnerable to this injury; (2) how to predict which patients are most susceptible to it; (3) how interventions before, during, and after extracorporeal circulation can either contribute to it, minimize its severity, or prevent it; and (4) where future study of this problem should be directed.

Effects of conventional hypothermic ischemic arrest and pharmacological arrest on myocardial supply/demand balance during aortic cross-clamping

Aortic cross-clamping may produce ischemic damage due to a discrepancy between supply and demand. Supply is determined by noncoronary collateral flow and substrate stores, and demand by electromechanical activity, wall tension, and temperature. The effects of 60 minutes of conventional hypothermic ischemic arrest were compared to those of pharmacological arrest. Noncoronary collateral blood supply was comparable in both groups during cross-clamping. With ischemic arest, mechanical activity and endocardial electrical activity persisted and wall tension fell progressively. With pharmacological arrest, electromechanical activity stopped in less than 1 minute but returned (with increased wall tension) nearly 1 hour. Thirty minutes following reperfusion, coronary flow was redistributed away from subendocardium after ischemic arrest and toward endocardium after pharmacological arrest. Myocardial performance was depressed severely after conventional arrest and only mildy after pharmacological arrest. We conclude that aortic cross-clamping is safer with pharmacological arrest than with ischemic arrest. The cardioplegic solution modifies the supply/demand balance favorably, but it is washed out by noncoronary collateral blood supply.