|
|
||||||||
Ann Thorac Surg 1995;59:1321-1325
© 1995 The Society of Thoracic Surgeons
Division of Cardiothoracic Anaesthesia and Intensive Care, Huddinge University Hospital, Stockholm, Sweden
Abstract
Temperature induces depression of cerebral perfusion and cerebral oxygen metabolism in particular, and this seems to explain why a reduced pump flow above a critical level is well tolerated during hypothermic cardiopulmonary bypass with apparent full metabolic recovery afterward. It only partly explains why a longer period of hypothermic circulatory arrest leads to a protracted recovery of cerebral perfusion and cerebral metabolism. This review suggests there is evidence that energy metabolism can easily be compromised during and after rewarming after hypothermic cardiopulmonary bypass with low flow and with circulatory arrest. Although data indicate that cerebral metabolism and cerebral energy state are better after low flow than after circulatory arrest, the risk of energy crises appears imminent with both techniques.
The introduction of hypothermia for whole-body protection and cerebral protection in particular enabled successful surgical correction of cardiac anomalies in children some 30 years ago. Initially, surface cooling, core cooling to deep hypothermia (<20°C), or a combination of these was used to depress brain metabolism enough to allow apparently safe periods of total circulatory arrest. The repair then was accomplished during optimal surgical conditions with an atonic heart in a bloodless field. The duration of cardiopulmonary bypass (CPB) was kept as short as possible, as CPB itself at that time was associated with great morbidity and mortality.
More recently, the indications for the profound hypothermic circulatory arrest technique have expanded to include surgical interventions for cerebral and complex aortic aneurysms in adults, reoperations for cardiac valve replacements, and operations for tumors within the venae cavae. Also, improved detection techniques for severe arteriosclerotic disease of the ascending aorta have promoted circulatory arrest in the field of coronary artery bypass surgery.
Brain injury still occurs despite the protection of hypothermia, and with improvements in CPB techniques, the concept of total circulatory arrest, instead of continuous pump flow whenever possible, has been questioned. Low flow is being used as an alternative to deep hypothermic circulatory arrest when a limited blood return to the operative field is acceptable. It has the advantage of maintaining flow to vital organs, especially the brain. Several new approaches to improve cerebral protection during circulatory arrest have been introduced, including intermittent circulatory arrest with reperfusion in between, selective antegrade and retrograde perfusion, cerebroplegia, and elective use of cerebroprotective drugs.
Hypothesis
The main hypothesis is that hypothermia depresses cerebral metabolism enough to allow a safe period of total circulatory arrest or maintained low pump flow after which there is full functional recovery. It also implies that structural derangements without detectable functional derangement do not occur, which might be of importance later in life. However, most of our methods do not allow this distinction, and in humans, we most often rely on indirect signs to measure brain damage.
Determinants of Cerebral Hemodynamics
A vast number of factors may influence cerebral hemodynamics and cerebral metabolism during and after deep hypothermic procedures: (A) subjects: species, age (different sensitivity to hypoxia), vascular disease (diabetes, hypertension, arteriosclerosis), vascular anomalies, preoperative state (marginal blood pressure, cyanosis), preoperative cerebral state (edema, earlier lesion); (B) CPB: pH management, timing of low flow or arrest, arrest time, flow rates, pulsatile flow, blood pressure, intracranial pressure, central venous pressure, surface activation of cascade systems, introduction of air or emboli, arterial filters, cannulation sites, selective antegrade or retrograde cerebral perfusion; (C) temperature: brain, perfusate, cooling or warming rates, core or surface cooling, topical cooling of skull, selective antegrade or retrograde cerebral perfusion; (D) blood: hematocrit, viscosity, arterial carbon dioxide level, arterial saturation, oncotic pressure, clotting, glucose levels, leukocyte, platelet, complement activation, blood protease cascades, cytokines; (E) drugs: anesthetics, anesthesia depth, cerebroplegia, cerebroprotective agents; (F) heart: cardiac and main vessel anatomy, cardiac function.
Temperature
Cerebral blood flow (CBF) in healthy humans is directly related to temperature, cerebral metabolism, and arterial carbon dioxide tension. The influence of temperature on cerebral perfusion is dependent on carbon dioxide management during CPB. Alpha-stat management (temperature-uncorrected blood gases) maintains a better CBF to metabolism ratio during hypothermia than pH-stat management but has the potential disadvantage of cerebral hypoperfusion because of low arterial carbon dioxide tension [1]. Alternatively, pH-stat management (temperature-corrected blood gases) with strongly uncoupled CBF and metabolism has the theoretical advantage of increasing cerebral perfusion and enhancing brain cooling, but possible drawbacks include raising intracranial pressure and increasing brain edema. Cerebral perfusion is reduced during hypothermic CPB with alpha-stat management, and there appears to be a linear relationship between temperature and CBF or CBF velocity [24]. This increase in cerebral vascular resistance during hypothermia seems to be unaffected by reductions in pump flow, which is in contrast to findings for systemic vascular resistance [5].
Hypothermia is of paramount importance for brain protection because cerebral metabolism in humans correlates exponentially with temperature [6, 7]. The relationship can be expressed according to van't Hoff's equation, where the 10logarithm of the metabolic rate is related linearly to temperature [6, 7]. Thus, cerebral metabolism is reduced approximately 5% to 7% per each degree centigrade, which results in a prolonged tolerance for ischemia. Cerebral metabolism is attenuated more than cerebral perfusion during cooling, even with alpha-stat management, because of the exponential relationship between temperature and cerebral metabolism. This results in increased CBF to cerebral metabolism ratios during hypothermia with apparent CBF in excess of demand, thus causing the arterialjugular venous oxygen content to narrow [6, 8].
Brain temperature is usually not measured. It is dependent on local heat production, regional CBF, and temperature of perfusing blood. Nasopharyngeal temperature (NPT) reflects only blood supplying one part of the brain, and substantial temperature differences can occur between arterial temperature and NPT during active cooling and rewarming [9]. These temperature differences diminish with time on hypothermic CPB. Thus, NPT may underestimate brain temperature and cerebral metabolism during cooling (vice versa during rewarming). This may lead to the introduction of deep hypothermic circulatory arrest at a relatively high cerebral metabolic rate with unexpected brain damage, including brain swelling, decreased cerebral perfusion, and decreased cerebral metabolism after circulatory arrest.
A clever approach to assure complete brain cooling before circulatory arrest is to continue cooling until jugular venous blood is almost fully saturated, thus indicating depressed metabolism of the entire brain [5, 6, 8]. Other proposed strategies to lower brain temperatures include pH-stat management during cooling [10], cooling until encephalographic silence [11], long fixed cooling times irrespective of NPT [12] and addition of head surface cooling before and during circulatory arrest [13]. In one study [10], piglets were randomized to either pH-stat or alpha-stat management before and after 1 hour of circulatory arrest at an NPT of 15°C. With the pH-stat strategy, a higher cerebral perfusion was observed in all regions of the brain, including the basal ganglia, during cooling. Also, pH-stat management resulted in an earlier recovery of adenosine triphosphate (ATP) levels and intracellular pH and a lower cerebral water content during rewarming.
Critical Cerebral Perfusion and Metabolism During Hypothermia
During hypothermia at 20°C (NPT), autoregulation is preserved in monkeys, and CBF is maintained despite reductions in pump flow to 25% [14]. In children, autoregulation seems to cease at an NPT of 20°C or less [3, 15], thus making cerebral perfusion dependent on pump flow [7]. Both CBF and cerebral metabolic rate for oxygen remained unaffected by a 30% to 45% reduction in pump flow at 26°C and at 18° to 22°C [5]. A further 45% to 70% reduction in pump flow decreased both CBF and cerebral metabolic rate for oxygen and was accompanied by an increased oxygen extraction at temperatures higher than but not less than 20°C. This suggests that there is a luxuriant brain perfusion at temperatures lower than 20°C and that further reduction in pump flow can be tolerated until the point is reached where the cerebral metabolic rate for oxygen becomes dependent on oxygen supply (CBF). On the basis of the exponential correlation between temperature and cerebral metabolic rate for oxygen in children during nonpulsatile CPB, minimal flow rates at different temperatures can be predicted [5]. Indeed, with differences between species taken into consideration, this prediction seems consistent with findings reported in several studies [14, 1619]. For example, the predicted minimal flow would be 8% in sheep at 15°C, and a flow at 10% of full flow was observed to maintain normal levels of ATP, creatine phosphate, and brain pH [18]. With a further reduction to 5% (5 mL kg-1 min-1), ATP levels decreased and brain pH began to fall but to a lesser extent than that occurring with circulatory arrest. This is consistent with an animal study [20] where cortical ATP levels fell linearly and lactate levels rose significantly with reductions in pump flow at 21°C.
Thus, despite the dramatic reductions in cerebral metabolism and requirements with profound hypothermia, there is still an ongoing, continuous need of substrates to the brain. Circulatory arrest will lead to a continuous temperature-dependent depletion of energy stores and an increase in intracellular acidosis that is due to anaerobic glycolysis and production of lactate [21]. This cerebral ischemic insult is aggravated by hyperglycemia [21], and serum glucose levels during reperfusion after circulatory arrest have been observed to correlate with the release of the cerebral injury marker CK-BB in children [22]. The increased release of lactate from the brain in animals seems to be linearly related to the duration of arrest [23]. Studies in rats [24] showed that moderate cerebral ischemia with moderate cerebral tissue lactate levels produces only selective neuronal damage in vulnerable areas. In contrast, higher levels of lactate damaged endothelium, neurons, and astrocytes and resulted in infarction and progressive edema. In children, the duration of increase in anterior fontanel pressure, an accurate estimate of intracranial pressure, is directly related to the NPT at the end of arrest and to the duration of arrest [25]. Serum levels of CK-BB in children after circulatory arrest at 15°C were noted to be related to the duration of arrest [22]. A moderate release was observed at less than 40 minutes, whereas the release was markedly greater after 40 minutes. Also, the impact of temperature during arrest is well illustrated in an animal study [26], where fewer neurodeficits and less histologic evidence of injury were found in dogs sustaining 2 hours of circulatory arrest at 6°C than at 18°C.
These findings all argue for continuous flow and against circulatory arrest, especially at high temperatures. Also, the data argue for very low temperatures when longer periods of arrest are anticipated. However, determinations of critical flow rates at different temperatures cannot explain why neuronal damage as judged by neuronal Golgi morphology was worse after a marginal flow of 10 mL kg-1 min-1 at 18°C than after 1 hour of circulatory arrest [27]! Animals maintained at high pump flows had significantly fewer Golgi abnormalities, but approximately 20% in this control group showed changes consistent with a ``severe'' designation [27]. Nor can it explain why children undergoing a heart operation at 15°C with continuous low flow (25%), well above the theoretical critical pump flow rate, were found to have the same rise in CK-BB as after a 40-minute period of circulatory arrest [28].
Cerebral Perfusion and Metabolism During and After Rewarming
Cerebral perfusion and cerebral oxygen metabolism follow temperature and usually return to or are much higher than normal during rewarming after hypothermic CPB provided there has been adequate cerebral perfusion and no period of circulatory arrest or marginal flow [1, 68, 29]. After a period of ischemia, such as after circulatory arrest, cerebral perfusion is depressed in proportion to both the length of ischemia and temperature [2, 3032]. The transcranial Doppler technique enables continuous measurements of cerebral perfusion, including changes during the cardiac cycle. Absent diastolic blood flow velocities were observed after profound hypothermic circulatory arrest procedures in children [32]. This indicates a lower than expected cerebral perfusion and suggests increased cerebral vascular resistance because of increased intracranial pressure. It is consistent with the findings of increased anterior fontanel pressures after circulatory arrest [25]. The diastolic pattern returned to normal after 4 to 6 hours. The observation of a normal diastolic blood flow velocity in 2 children who had a period of cold reperfusion despite hypothermic circulatory arrest suggested a favorable effect of cold reperfusion [32]. Animal studies [3335] have indicated increased cerebral vascular resistance and decreased CBF lasting at least 4 hours after hypothermic CPB with circulatory arrest at 13°C. In contrast, cerebral vascular resistance was only marginally elevated after CPB with low flow (25%) at 13°C [3335].
During rewarming after profound hypothermia with maintained pump flow, cerebral oxygen rate increases with or more than cerebral perfusion. In both children and adults, this may result in an increased arteriovenous oxygen extraction and jugular venous oxygen saturation values of 40% or less [8, 29]. The situation where oxygen delivery becomes rate limiting for aerobic oxidation may be approached, but it probably does not usually pass critical levels, as in children there is no general net release of lactate during this period [36]. However, the period during and after rewarming is delicate with respect to perfusion pressure, hematocrit, and anesthesia. Systemic vascular resistance, and accordingly blood pressure, falls with increased temperature and return to full flow, which often necessitates volume infusions in an already hemodiluted patient. If blood products are avoided, additional hemodilution occurs and because of the fall in blood pressure, the anesthetist may be prone to avoid additional anesthetics. All these factors may aggravate the already delicate balance between cerebral oxygen delivery and oxygen consumption during and after rewarming.
A low cardiac output after CPB might also worsen this balance. Indeed, this might explain why the cerebral release of CK-BB correlates with low levels of hemoglobin after hypothermia in children [22]. Hemoconcentration by hemofiltration during rewarming probably promotes oxygen delivery to the brain if hematocrit is very low and also releases excessive water and toxic metabolites from the body and probably the brain [37]. Hemofiltration during rewarming in children who had a limited period of hypothermic circulatory arrest or continuous low flow resulted in an equal recovery of cerebral perfusion during and after rewarming in both groups. However, the cerebral metabolic status in the circulatory arrest group was worse, as indicated by a protracted release of lactate from the brain [36]. The increased rate of oxygen demand during rewarming might be a response to a preceding period of ischemia, insufficient anesthesia, or too rapid rewarming, as suggested by the correlation found between speed of rewarming and desaturation of jugular blood in children [8]. A logical approach would be to set a lower rate of rewarming and a lower target temperature.
An energy deficit is present subsequent to circulatory arrest at the start of rewarming [18, 38]. However, cerebral oxygen metabolism and cerebral oxygen extraction remain depressed along with cerebral perfusion during and early after rewarming, thus suggesting disordered cerebral metabolism and oxygen utilization [6, 36]. The duration until recovery of cerebral oxygen metabolism correlates with the length of circulatory arrest in animals [39, 40]. The depression of cerebral oxygen consumption may be explained by persistent intracellular acidosis [17, 38], known to depress many enzyme systems including the Krebs cycle [41]. Thus, optimal aerobic glucose metabolism may be impeded and anaerobic glycolysis, furthered. Anaerobic glycolysis may elucidate the protracted release of lactate from the brain seen in children up to 6 hours after CPB with hypothermic circulatory arrest [36]. This may be consistent with the observation of an up to 8 hours' increase in cerebral vascular resistance and depressed cerebral perfusion together with recovery of cerebral oxygen and glucose metabolism within a few hours after CPB with hypothermic circulatory arrest [33, 35]. Thus, during the first hours after CPB with hypothermic circulatory arrest, a situation with a marginal oxygen delivery is observed, and a recovered cerebral metabolic rate is maintained by an abnormally high oxygen and glucose extraction. This situation, which is more severe after circulatory arrest than after low flow, creates a vulnerable period postoperatively in which suboptimal hemodynamic and anesthesia conditions, including decreased cardiac output and blood pressure, inadequate sedation, low arterial oxygen saturation, and inappropriate ventilation, could result in a critical and possibly irreversible cerebral function impairment.
New Techniques
Different techniques of promoting cerebral perfusion after hypothermic perfusion and after hypothermic circulatory arrest in particular have been investigated. Animal studies [42] have confirmed that pulsatile perfusion promotes cerebral perfusion and metabolic recovery considerably after 60 minutes or more of hypothermic circulatory arrest. Pulsatile perfusion during hypothermia has been observed to improve cerebral metabolic recovery both after hypothermic circulatory arrest and after low flow in animals [17], and pulsatile perfusion prevented the net release of lactate from the brain during a low flow of 25 mL kg-1 min-1 in dogs at 20°C [43]. A clinical trial during profound hypothermic procedures is warranted.
Other methods to decrease the fall in cerebral energy state during hypothermic circulatory arrest are currently being investigated. Retrograde cerebral blood perfusion with venous pressures lower than 25 mm Hg in dogs at 18°C resulted in lower cerebral temperatures, increased cerebral oxygen tensions and ATP levels, and decreased cerebral carbon dioxide tensions and lactate levels in comparison with circulatory arrest [44]. However, in this animal model, retrograde cerebral perfusion did not prevent anaerobic brain metabolism, thus indicating suboptimal perfusion. Antegrade perfusion seems to protect the brain better than retrograde perfusion, as better postoperative neurologic scores were observed in sheep that received antegrade crystalloid cerebroplegia or topical head cooling during a 2-hour period of circulatory arrest at 15°C than in sheep with retrograde crystalloid cerebroplegia or systemic hypothermia alone [45].
Short intermittent antegrade cold cerebral blood reperfusion every 15 to 20 minutes during deep hypothermic circulatory arrest in animals promotes full recovery of cerebral metabolism and prevents cerebral lactate release during reperfusion [22, 39], and two separate 30-minute periods of hypothermic circulatory arrest separated by a short period of warm antegrade reperfusion gave earlier recovery of cerebral metabolism than a single 60-minute period of arrest [46]. With intermittent antegrade crystalloid cerebroplegia delivered at 4°C, significantly lower brain temperatures were observed in puppies during 2 hours of circulatory arrest at an NPT of 15°C than in puppies with conventional arrest [47]. Interestingly, short-term recovery of CBF, cerebral metabolism, ATP levels, and intracellular pH was augmented by the use of cerebroplegia. However, this improved recovery may not be explained solely by the effect of temperature on cerebral metabolism. Intermittent antegrade or retrograde crystalloid cerebroplegia or selective cold blood perfusion during circulatory arrest probably washes out cerebrotoxic metabolites produced during ischemia. Also, retrograde cerebral perfusion definitely reduces the incidence of severe stroke during operation on the aortic arch of adults by washing out air and debris before the return to normal antegrade perfusion.
Conclusion
The temperature-induced depression of cerebral perfusion and cerebral oxygen metabolism in particular seems to explain why a reduced pump flow above a critical level is well tolerated during hypothermic CPB with apparent full metabolic recovery afterward. It only partly explains why a longer period of hypothermic circulatory arrest leads to a protracted recovery of cerebral perfusion and cerebral metabolism. This review suggests there is evidence that energy metabolism can easily be compromised during and after rewarming after hypothermic CPB with low flow and with circulatory arrest. Although data indicate that cerebral metabolism and cerebral energy state are better after low flow than after circulatory arrest, the risk of energy crises appears imminent with both techniques.
Acknowledgments
The present study was supported by grants from the Karolinska Institute.
Footnotes
Presented at the Conference on CNS Dysfunction After Cardiac Surgery: Defining the Problem, Fort Lauderdale, FL, Dec 1011, 1994.
Address reprint requests to Dr van der Linden, Division of Cardiothoracic Anaesthesia and Intensive Care, Huddinge University Hospital, S-141 86 Huddinge, Stockholm, Sweden.
References
This article has been cited by other articles:
![]() |
H. Kaukuntla, D. Harrington, I. Bilkoo, T. Clutton-Brock, T. Jones, and R. S. Bonser Temperature monitoring during cardiopulmonary bypass--do we undercool or overheat the brain? Eur. J. Cardiothorac. Surg., September 1, 2004; 26(3): 580 - 585. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. Hagl, N. Khaladj, D. J. Weisz, N. Zhang, L. J. Guo, C. A. Bodian, D. Spielvogel, and R. B. Griepp Impact of high intracranial pressure on neurophysiological recovery and behavior in a chronic porcine model of hypothermic circulatory arrest Eur. J. Cardiothorac. Surg., October 1, 2002; 22(4): 510 - 516. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. P. Ehrlich, J. McCullough, D. Wolfe, N. Zhang, H. Shiang, D. Weisz, C. Bodian, and R. B. Griepp Cerebral effects of cold reperfusion after hypothermic circulatory arrest J. Thorac. Cardiovasc. Surg., May 1, 2001; 121(5): 923 - 931. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. I. Buss, R. F. McLean, B. I. Wong, S. E. Fremes, C. D. Naylor, E. M. Harrington, W. G. Snow, and M. Gawel Cardiopulmonary Bypass, Rewarming, and Central Nervous System Dysfunction Ann. Thorac. Surg., May 1, 1996; 61(5): 1423 - 1427. [Abstract] [Full Text] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |