Hypothermia in the Neonate and Infant Kelly Calvert, BSN, CCP Chief of Perfusion Services West Virginia University Hospital (Prepared for the Course: Basis and Techniques of Neonatal Perfusion - Brazilian Page of Perfusion Line)

I. RATIONALE

Although the rationale for hypothermia in cardiac surgery may seem somewhat obvious there may be a need to review these issues as they apply specifically to the neonatal patient. The theory behind each issue certainly applies to the adult patient as well albeit somewhat less exaggerated.

A. Cerebral Protection

Certainly there is concern for all organs to gain protection through induced hypothermia, but hypothermia is specifically a must for cerebral protection. The goal to provide organ protection is by a result of a reduction in cellular metabolism and to also preserve high energy phosphate stores and lower ATP consumption. Hypoxic injury begins when these high energy stores are depleted, and although other organs are at risk, the brain is the most sensitive and is the limiting factor in both how we achieve it as well as the outcome.

B. Surgical Exposure

Hypothermia allows for a reduction in total blood flow to provide exposure at the surgical field. This is emphasized as congenital repairs are primarily intracardiac repairs and direct visualization is paramount. In the presence of significant collateral flow the difference may be low-flow bypass(LFB) or deep hypothermic circulatory arrest(DHCA). This is realized when one considers a "pump sucker" is used to clear a field of blood in a larger patient, whereas a cotton swab is as effective in clearing the site in a neonatal heart.

C. Organ Protection from Cardiopulmonary Bypass

This speaks to the deleterious effects of cardiopulmonary bypass(CPB) itself. Hypothermia allows for flow reduction and possibly ultimately in less complement activation and negative influences of CPB. On the other hand infants seem to respond better to pulsatile flow and hypothermia may provide a protective mechanism by which to tolerate prolonged laminar non-pulsatile flow.

D. Myocardial Protection

The role of hypothermia in myocardial protection is two-fold. Through systemic cooling in a global fashion the myocardium will be protected from collateral rewarming. Cold cardioplegic solutions provide additional direct and topical hypothermic protection. Hypothermia is an adjunctive protectant in the role of the myocardium, whereas it is a mandatory component in cerebral protection.

E. Safety Margin

As in adult CPB procedures, hypothermia provides an element of safety in the event of catastrophic failure of products, equipment malfunction or embolic event.

---

II. EFFECTS OF HYPOTHERMIA

A. Cellular

After establishing the surface rationale for use of induced hypothermia it is of importance to view the effects of reduced temperature at the cellular level. Through the basis of understanding these effects a judgement can then be made as to the outcomes at a clinical level.

We need to first look at expectation. What temperatures should we cool to and for how long? For the answer we need to review the Temperature Coefficient, or Q10. Q10 is the relationship between temperature and metabolic activity, ultimately the reduction in metabolic reactions resultant from hypothermia. Q10 states that for every 10⁰ C change in temperature there is a multiple by which a reaction rate will change. Thus a Q10 of 2 means that a reaction rate will double with an increase of 10⁰ C.

There is a different Q10 value for most physiologic changes in the body. Most reactions have a Q10 of greater than 2 or 3. Processes of diffusion and ionization, on the contrary, have a low Q10 value, closer to 1. The Q10 is not necessarily stable at all temperature ranges. For example from 25⁰C to 35⁰C the Q10 is 2.2. From 15-25⁰C the Q10 is 1.9 and from 5-15⁰C the Q10 is 1.6. Q10 values also vary for different organs and between cells within an organ, and even for different biologic enzymatic processes with individual cells. The above can be summarized by stating that any change in temperature results in a heterogenous response among various cellular processes. It is thus important to monitor several temperatures to assure global cooling. Our current practice will be discussed further in this text.

So how is this relative to what we practice? If one looks at the Q10 of the pediatric brain, we can make some assumptions. The Q10 for the pediatric brain metabolic activity is 3.6. This means that for a reduction of 10⁰C the cerebral metabolism, or CMRO2, is decreased by a multiple of 3.6. We can use this to predict safe limits of protection with hypothermia. If the brain can tolerate an ischemic period of 3-5" at 37⁰C, then at ten degrees less, or 27⁰C we can estimate that a tolerance of ischemia could be up to 18 minutes. At 17⁰C, a common temperature for DHCA, up to 61 minutes of ischemic arrest could be tolerated. These values relative to safe periods of circulatory arrest for a given temperature are certainly what we read in the literature and base our standards of practice.

These estimated safety margins also are reinforced when we look at ionic transport. Passive diffusion has a Q10 value less than ionic transport. At normothermia 50-75% of the high energy phosphate expenditures are used in maintaining transmembrane ionic gradients. The depletion of ATP stores slows as temperature decreases. The levels are usually maintained following 30-45 minutes of DHCA. Beyond 60 minutes there is recovery but it takes much longer.

Another protective mechanism of hypothermia can be noted. In an ischemic event there is a loss of K+, Ca++ and Na+ eletromechanical gradients. An influx of Ca++ can occur resulting in a reperfusion injury. Hypothermia protects from injury by restricting the membrane permeability and preventing Ca++ entry into the cell (specifically the brain), thus limiting the reperfusion damage. Neuronal response is also affected as hypothermia slows the release of neuroransmitters, providing a protective mechanism against ischemia. Protein kinase, an important enzyme involved in regulation of neurotransmitter release and function, has severely depressed activity during an ischemic event at normothermia. Reductions in protein kinase activity of 39% have been noted in brain tissue ischemia at normothermia. When temperature was reduced to just 32⁰C the protein kinase activity was unchanged.

The effects of hypothermia on pH can also be used to determine options for blood gas strategies. There is an inverse relationship to temperature and pH. As temperature decreases there is an alkaline shift on pH. For every 1⁰C change in temperature there is a 0.015 change in pH. A transcellular gradient of 0.6-0.8 pH units is needed to eliminate acid metabolites and CO2. The biochemical basis for alpha-stat and pH-stat strategies is based on these gradients The dissociation constant of H2O is highly dependent on temperature. H2O represents the greatest primary source of intracellular H+ and the shift in pH at lower temperatures is an attempt to maintain electrochemical neutrality of the intracellular environment. As the solubility of CO2 increases with a decrease in temperature, there is a perception of respiratory alkalosis. Alpha-stat management maintains the gradient to allow H+ to leave the cell therefore the elevated pH is apppropriate for the given temperature change to maintain the same transcellular gradient. PH-stat strategy allows H+ to accumulate as temperature decreases. With a low intracellular pH electrochemical neutrality is lost and enzymatic functions are impaired. In terms of cellular function it would seem that alpha-stat management of blood gases is the preferred method. This can not be an absolute assumption that it is then the obvious choice for neonatal and pediatric perfusion protocols. One must also weigh the effects of cerebral blood flow and clinical outcomes. These considerations will be discussed below.

B. Cerebral

As temperature decreases, hypothermic vasoconstriction occurs, systemic vascular resistance increases and the percent of flow proportional to the cerebral circulation increases. These normal vascular responses occur at moderate hypothermia. Cerebral blood flow (CBF) and cerebral metabolism (CMRO2) both decrease with cooling. CBF decreases in a linear fashion whereas CMRO2 decreases in an exponential fashion. CBF is dependent on brain metabolism. If metabolic needs are high, the cerebral vascular resistance decreases and CBF increases. This relationship is known as flow/metabolism coupling. At moderate hypothermia both flow/metabolism coupling and pressure flow autoregulation remains intact. Thus, despite a range of pressures CBF is not altered as cerebral vasculature responds by dilating or constricting accordingly. At deep hypothermic temperatures this normal autoregulation is affected. The normal vascular responses are last and CBF is relatively dependent on blood gas strategy. Alpha-stat management maintains this autoregulation, but there is a loss of flow/metabolism coupling. With pH-stat strategy, although there is a loss of enzymatic function but cerebral vasodilation is significant.

This CBF: CMRO2 (flow: metabolism) relationship is about 20:1 at normothermic temperatures with alpha-stat blood gas management. It may be even higher in neonates, as the metabolic demand for neuronal growth is much greater than in older children or adults. As cooling begins this ratio increases to 30:1 at moderate hypothermia ranges. However under pH-stat management this relationship is 60:1, providing gross luxuriant flow. This luxuriant flow becomes advantageous relative to global cooling of brain tissue as well as minimizing proportional steal of flow from collaterals. This global cooling may have more benefit to outcome than the advantages noted from alpha-stat regulation. A modified alpha-stat strategy may seem in order.

When looking at clinical outcomes I again refer to Boston Children's efforts in following DHCA. Through their retrospective studies on choreoathetosis and developmental outcomes, they identified a direct correlation with blood gas strategy, especially in the presence of aortopulmonary collaterals. Through 1996 the patients managed by pH-stat strategy exhibited an overall lower mortality, improved oxygenation post-CPB, improved coagulation post-CPB and specifically the TGV patients experienced shorter ICU and hospital stays. There was a greater recovery of high energy phosphates during rewarming and less postop seizure activity. Certainly if alpha-stat strategies are used one must employ a longer cooling phase (20 min or greater) to assure global cooling. These potential benefits of pH-stat management are not applicable to the adult patient, however, secondary to atherosclerotic changes there are steal flow areas of the brain and only alpha-stat management should be used.

C. Renal

Hypothermia, in conjunction with nonpulsatile perfusion, decreases the mean arterial pressure in pediatric patients. This leads to an increase release of renin, angiotensin, ADH and other catecholamines, thereby causing further renal vasoconstriction and a decrease in renal blood flow. As pulsatile flow alone returns there is a significant decrease in renin release. Thus hypothermia alone is not a substantial factor in influencing renal blood flow. This overall effect does not seem to be significant postop. Postop renal failure is primarily related to low cardiac output states or preop renal dysfunction.

The greatest influence in the neonate is organ immaturity. There is generally a decrease in glomerular filtration rate in neonates and therefore a prolonged CPB time may result in greater fluid retention than adults. This, compounded with greater hemodilution ratios in the newborn may lead to detriment.

D. Pulmonary

Hypothermia in conjunction with nonpulsatile perfusion may be more contributory to lung dysfunction than exposure to the bypass circuit itself. A significant leukocyte activation and degranulation occurs and capillary membrane injury results in extravasation and leak. This leads to a decrease in lung compliance and decrease functional residual capacity. This is of concern especially in the cyanotic infant with pulmonary hypertension. Leukocyte depletion of all blood components is a must in an attempt to thwart the untoward pulmonary effects on this subset of patients.

E. Myocardial

Intracardiac repairs obviously require some period of ischemic time. Hypothermia aids in global myocardial preservation. Although with hypothermia there is a greater overall reduction in metabolic activity the greatest impact on myocardial oxygen consumption (mVO2) is contractility. The electromechanical activity of contraction accounts for 85% of myocardial energy demands. Thus the greatest impact on decreasing those demands is a direct result of the slowing of heart rate in response to hypothermia.

Spontaneous fibrillation of the pediatric heart does not necessarily occur even at extremely cold temperatures. Thus this energy demand is occurring during the entire cooling phase prior to aortic cross-clamp placement. During this rapid cooling phase, there is a hypercontractile response of the myocardium as Ca++ influxes intracellularly as a positive inotropic effort. During this hypercontractile state there is an increase in wall tension (Tw) and significant ATP consumption as a flaccid diastolic arrest is not realized. Cautious use of exogenous calcium and avoidance of overdistension is recommended.

This exposes a bit of a controversy regarding neonatal tolerance to ischemia. In the normal, immature heart there are increased glycogen stores and available anaerobic capacity to combat ischemia. There are larger high energy stores in immature myocytes for recovery post ischemia. The cyanotic sick infant, who has often suffered from congestive heart failure of some degree, has a myocardial substrate reserve that is less than the normal newborn heart. This is due to impaired substrate delivery and marginal myocardial energy reserves secondary to expenditure. Thus the normal neonatal heart is more resistant to ischemia and reperfusion injury but this is not the case with the sick neonate with a stressed myocardium. Considerations must be made relative to the use of blood cardioplegia to enhance substrate delivery as well as the timing of aortic cross clamp placement.

---

III. RECOMMENDATIONS

In summary there are a few basic recommendations for practice, based on our protocols at West Virginia University Hospital. Cooling methods must employ surface cooling in the form of hypothermia blankets as well as ice packs around the head. Core cooling should be enhanced with vasodilation to achieve true global cooling. Temperatures monitored should cool at relatively even gradients. We monitor tympanic, rectal, nasopharyngeal and skin toe temperatures. The toe temperature reflects peripheral perfusion and is often a good indicator for patient management. Cooling prior to DHCA should be a minimum of 20-30 minutes with a rate of 1⁰C per minute. We employ pH-stat management on all cases <28⁰C.