To the Editor:—
We congratulate Ditsworth et al. for the elegant biochemical demonstration of mechanisms triggering the apoptosis cascade to eventually cause neuronal death following deep hypothermic circulatory arrest (DHCA) in piglets. 1
The authors dismissed the significantly elevated caspase-8 of the cardiopulmonary bypass (CPB) group (see their figure 5) and declined making any comments regarding it, but how do they explain it? We believe it is the key to interpreting the article in the proper perspective. It is indeed documented that ischemic damage, even though reversible, had occurred without arrest, corroborating numerous reports of the overlooked deleterious effects of α-stat cooling. If circulatory arrest was not induced and α-stat hypothermia was innocuous, one would not anticipate activation of caspase-8 or −3, the case of control animals without CPB.
The authors did not find postrewarming (adenosine triphosphate [ATP]) differences between the various groups. However, the issue is the ATP during cooling just before arrest and after the arrest just before rewarming, not after rewarming, because the described caspase-8 and −3 are activated by ischemia, during which time ATP is depleted.
We postulate that the arrest (ischemic) period was the final blow to activate the pathway of caspase-3 and to trigger the irreversible apoptosis cascade whose grounds had been conditioned by the hypothermia-induced Bohr effect (tissue hypoxia caused by the increased affinity of the oxyhemoglobin), aggravated by the α-stat alkalosis during induction of hypothermia to the point of caspase-8 activation. In the context of ischemia, apoptosis and necrosis must be a continuum, the fate depending on the extent of spared ATP stores;2necrosis is found in the center of brain infarcts, and apoptosis in the surviving penumbra zone. Alkalosis within pH ranges of 7.0 to 8.0 depresses linearly creatine-kinase mediated phosphorylation (∼ P) reactions;3the extremes of pH below 7 or above 8 at which linearity is maintained is not known. During deep α-stat hypothermia, the actual pH often exceeds far above 8.0; if sustained long enough the decreased synthesis will lead to ∼ P depletion even without arrest, manifested as significant elevation of caspase-8 in the CPB-only group. We postulate that ∼ is further consumed during the arrest to levels below the threshold limits for activation of caspase-3 and leading to the apoptosis cascade. If arrest time is limited, sufficient ∼ still remains for the recovery of ATP synthesis mechanisms on rewarming and oxygenation, leading to the recovery of postrewarming [ATP] levels similar to controls, thus preventing acute cellular death by necrosis. However, because the apoptosis cascade has been already activated, those cells are destined to die 8 to 72 h post-DHCA without involving ATP at that time.
Following is a brief account of some of the widely documented deleterious effects during α-stat cooling that, despite being significant, were dismissed (as with the caspase-8 in the CPB group in the authors’s study) because of the overwhelming findings of circulatory arrest:
During cooling induction: (a) brain hypoxia, 4worse with α-stat hypothermia than with pH-stat hypothermia 5; (b) brain lactate production 6; and (c) brain glutamate and nitric oxide release. 7
During or after rewarming: (d) brain production of hypoxanthine and xanthine. 8
Functional outcome: (e) the time required for electroencephalographic recovery on rewarming correlated with time to electrocerebral silence, and even minimal hypocarbia increased time to electrocerebral silence;9(f) worse neurologic performance and worse brain histopathologic injury with longer pre-arrest CPB α-stat cooling duration. 10
Cerebral oxygen needs below 18°C could theoretically be met by the dissolved oxygen on which α-stat strategies rely, 11but dissolved oxygen cannot satisfy requirements at temperatures higher than 18°C, at which the role of oxyhemoglobin (whose dissociation is carbon dioxide–dependent or pH-dependent) is greater. Significant brain lactate production starts during cooling well before arrest, 6coinciding with brain hypoxia 4,5and excitotoxicity. 7Regardless of the temperature, the Ca2+extrusion pump is impaired by alkalosis. Alkalosis increases N -methyl d-aspartate currents sensitizing to glutamate, 12,13thus facilitating the intracellular Na+and Ca2+influx that cause cytotoxic edema, especially on reperfusion, which is minimized by acidosis. 14,15Mild acidosis reduces glutamate neurotoxicity by decreasing the activation of N -methyl d-aspartate receptors and, consequently, reducing Na+and Ca2+influx, thus minimizing oxygen-glucose deprivation excitotoxic and reperfusion neuronal injury. 12–16
The issue is preserving the metabolic machinery integrity and ∼ P levels during cooling induction. For millions of years, nature 17has exploited the protective hyperpolarizing effects of hypothermia due to increased Na+efflux 18and increased Cl−conductance 19of eucapnic ventilation–induced acidosis, which is equivalent to pH-stat perfusion management. Such a strategy maintains aerobic metabolism and integrity of energy requiring membrane pumps, regardless of age, 17while fully taking advantage of metabolic depression and decreased release of excitatory aminoacids 20induced by hypothermia.
Our contention that brain hypoxia (Bohr effect) develops with α-stat cooling severely enough to cause injury well before arrest induction has been corroborated by the authors’ elegant study. pH-stat management was demonstrated to be superior both functionally and histopathologically by the same group, 5,21and it is regrettable that pH-stat–managed (CPB and DHCA) piglets were not included in this study, for we believe that caspase-8 in a CPB group and apoptosis in a DHCA group would have been prevented or greatly minimized. Supplementary intravenous taurine further potentiated, equivalent to 1.2°C, the protection afforded by pH-stat hypothermia. 22