Importance of acid-base strategy in reducing myocardial and whole body oxygen consumption during perfusion hypothermia

J

THORAC CARDIOVASC SURG

1990;100:699-707

Importance of acid-base strategy in reducing myocardial and whole body oxygen consumption during perfusion hypothermia During induced hypothermia with cardiopulmonary bypass, acid-base management usually follows one of two strategies: the so-called ectothermic or alpha-stat strategy, in which the pH of the arterial blood increases 0.015 pH units for every degree Celsius decrease in body temperature, or the pH-stat strategy, in which pH rema~ 7.4 at aU temperatures. It has been assumed that oxygen consumption decreases approximately equally during hypothermia with either strategy, although there are biochemical reasons to hypothesize that oxygen consumption would be better maintained with the alpha-stat strategy. We also hypothesized that venous oxygen tension would be lower with the more alkaline alpha-stat strategy than with the pH-stat acid-base strategy, because of the Bohr effect. We tested these hypotheses by placing 10 anesthetized immature domestic pigs on cardiopulmonary bypass. We measured whole body oxygen consumption and myocardial oxygen consumption. Control measurements were made at 37° C. Then the animals were cooled to 27° C and the measurements were repeated. The alpha-stat strategy (pH 7.554 ± 0.020 at 27° C) was used in five animals and five animals received pH-stat management (pH 7.409 ± 0.012 at 27° C). Whole body and myocardial oxygen consumption rate decreased in both groups, but more so in the alpha-stat animals than in the pH-stat animals. The unexpectedly high oxygen consumption in the pH-stat animals also resulted in a lower than expected venous oxygen tension. Thus the effect of hypothermia in reducing oxygen consumption was less pronounced with pH-stat acid-base management.

David C. Willford, PhD, William Y. Moores, MD, Shangyi Ji, MD, Zhung Tung Chen, MD, Arturo Palencia, BS, and Pat 0. Daily, MD, La Jolla, Calif.

Human blood samples at constant gas content, as well as the blood of many ectothermic vertebrates, increase in pH as temperature decreases. 1•3 In many cases this increase amounts to about 0.015 pH units per degree Celcius. Because this pH strategy appears to maintain a constant ratio of charged to uncharged alpha-imidazole groups of histidine, it is often referred to as alpha-stat acid-base regulation. 4• 5 In hibernating mammals, on the From the Department of Medicine, Division of Physiology, University of California, San Diego, La Jolla, Calif. This research was supported by National Institutes of Health Grant HL-17731 and a Medical Research Grant from the Veterans Administration. Dr. Willford also received support as a Parker B. Francis Fellow in Pulmonary Research. Read at the Fifteenth Annual Meeting of The Western Thoracic Surgical Association, Monterey, Calif., June 22-25, 1989. Address for reprints: David C. Willford, PhD, Department of Medicine M-023, Division of Physiology, University of California, San Diego, La Jolla, CA 92093.

12/6/23658

other hand, pH tends to remain constant at 7.4 as body temperature decreases, 3 and this strategy is often referred to as pH-stat acid-base regulation. It should be noted that the acid-base state of some ectotherms is not regulated by variance of the pH with a slope of -O.Q15;o C. Many ectotherms do follow this line, some appear to be pH-stat, and some utilize intermediate strategies. 2 Because alpha-stat acid-base regulation is observed in many ectotherrnic vertebrates, it has been proposed as a management strategy for patients during hypothermia. 3• 6-8 The rationale for the use of this strategy is based on the observation that myocardial performance and the function of many enzymes is maintained better with alpha-stat than with pH-stat regulation4• 9· 11 and that this could also result in a higher metabolic rate with alpha-stat regulation. 3• 7 However, because of the increased affinity of hemoglobin for oxygen (decreased P 5o*) during hypo*Pso =Oxygen tension (millimeters of mercury) at which hemoglobin is half saturated with oxygen.

699

The Journal of Thoracic and Cardiovascular Surgery

7 0 0 Willford et a/.

Reservoir Oxygenator Heat Exchanger Fig. 1. Diagram of swine heart model. SVC. Superior vena cava; LA, left atrium; RA, right atrium; LV. left ventricle; RV. right ventricle; IVC, inferior vena cava.

thermia with alpha-stat regulation, peripheral oxygen delivery could be compromised, and at least one study 12 suggests improvement with a more acidotic strategy. Wood 13 has suggested that one of the adaptations that allows ectotherms to tolerate large changes in body temperature is reduced thermal sensitivity of the oxygen-hemoglobin equilibrium curve when compared with that of endothermic mammals such as man. The purpose of our study was to test two hypotheses: First, that metabolic rate would be maintained-ata higher rate with alpha-stat acid-base regulation because of better enzyme function. The second hypothesis deals with venous oxygen tension (Po 2). Our earlier theoretical work had suggested that, because of the Bohr effect on the oxygen-hemoglobin equilibrium curve, the venous Poz (and presumably tissue Poz) would be lower with alphastat acid-base regulation. 14 On the other hand, a higher oxygen consumption rate (hypothesis 1) would also tend to lower venous Poz. Thus we also wanted to test the hypothesis that venous Po 2 would be lower with alphastat acid-base regulation. Cardiopulmonary bypass

(CPB) was used to allow rapid, efficient cooling of the animals, as well as to control the perfusion flow rate and thus provide a means for ensuring adequate oxygen delivery to the tissues during hypothermia. Because the oxygen consumption rate, as well as the mixed venous Poz, can be dependent on the blood flow rate, 15 the ability to clamp the blood flow with CPB represented a powerful advantage of this experimental model.

Materials and methods Ten immature domestic pigs weighing approximately 30 kg each were used in this study. The animals were divided into two groups, one of which was designated for alpha-stat acid-base management during hypothermia and the other pH-stat acidbase management. Animal preparation. All animals were initially given ketamine (25 mg/kg) and atropine (2 to 4 mg) intramuscularly. Halothane was then given initially by mask and later (after a tracheostomy) through a cuffed endotracheal tube. Before initiation of extracorporeaJ perfusion, morphine sulfate ( 10 mgj kg) was given intravenously and the halothane was withdrawn. Supplemental30 mg injections of morphine were given every 30 minutes during normothermia and every hour during hypother-

Volume 100 Number 5 November 1990

Oxygen consumption during hypothermia 7 0 I

a-STAT

l:

c. 0

> ·:;

.e

---------t---. f ,,',' t ___.

0'

10'

20'

f ~~. . . ~..

......... . . .

....I I

pH-STAT

60'

30'

27oC

Fig. 2. Mean and standard error of the mean in vivo arterial pH at each of the sampling periods. Desired pH for the normothermic control in both groups and for the pH-stat hypothermic animals was 7.40 ± 0.05. For the alphastat animals, the desired pH at 27° C was 7.55 ± 0.05.

·Table I. Blood gas and chemistry data from 10 alpha-stat and 10 pH-stat animals (mean ± standard error of the mean) Temperature (C) Alpha-stat control Alpha-stat pH -Stat control pH-Stat

36.4 27.2 37.3 26.8

± 0.3 ± 0.1 ± 0.2

± 0.1

Hemoglobin (gmjdl) 8.3 7.5 6.4 6.1

± 0.5 ± 0.3 ± 0.2

± 0.1

Arterial Po2 (mmHg) 191.1 235.6 227.0 151.2

± ± ± ±

42.3 26.5 39.1 27.2

Venous Po2 (mmHg) 24.6 24.8 26.3 23.0

± ± ± ±

2.3 2.5 3.7 2.1

Arterial pH

Norepinephrine (pgjml)

± ± ± ±

99.7* 85.7* 72.3* 342.7*

7.390 7.554 7.404 7.409

0.029 0.020 0.017 0.008

*Norepinephrine concentrations were measured in four animals, two alpha-stat and two pH-stat animals.

mia. The animals were paralyzed with and subsequently maintained with succinylcholine to block shivering. (This was important because oxygen consumption resulting from shivering thermogenesis would have compromised our ability to evaluate differences in the oxygen consumption rate caused by different strategies of acid-base management.) Before CPB the animals' lungs were ventilated with an Ohio anesthesia ventilator (Ohio Medical Products, Madison, Wis.). The desired acid-base strategy (either alpha-stat or pH-stat) was achieved by adjusting ventilation to control arterial carbon dioxide tension (Pco2). After CPB was begun, arterial Pco2 was controlled by altering the fractional concentration of oxygen and gas flow rate to the membrane oxygenator (William Harvey HF-4000, C.R. Bard, Inc., Santa Ana, Calif.). Once the desired arterial Pco2 was obtained, the ventilator or gas flow settings were maintained and metabolic acidosis was treated with sodium bicarbonate. Blood gases were monitored with an IL-813 blood gas analyzer. Hemoglobin concentration and percent hemoglobin saturation with oxygen were determined with an IL-282 Co-Oximeter. The blood gas analyzer and the Co-Oximeter are manufactured by

Instrumentation Laboratories, Inc., Lexington, Massachusetts. Blood gases were corrected to the in vivo temperature by means of our measured blood gas correction factors for pigs. 16• 17 We attempted to maintain arterial Po2 between 150 and 250 mm Hg. The animals were connected to an extracorporeal perfusion circuit similar to that used by our group previously. 15• 18 • 19 The perfusion circuit is illustrated in Fig. 1. Extracorporeal circulation was used to allow us to study blood gases and oxygen consumption during hypothermia while maintaining total blood flow at a constant level. As discussed previously, this was important because both oxygen consumption and venous Po2 can be dependent on the total blood flow. The perfusion circuit allowed for drainage of the cavae and perfusion of the aortic root. Pressure monitoring catheters (model p23la, Statham Instruments, Inc., Hato Rey, Puerto Rico) were placed in the aortic root. Blood gas sampling catheters were connected to the aortic perfusion line and mixed venous return line and were placed in the coronary sinus. A calibrated Sarns pump (Sarns, Inc., Ann Arbor, Mich.) was used with the disposable mem-

The Journal of Thoracic and Cardiovascular Surgery

7 0 2 Willford et a/.

100

0

90

0....

-

80

0

0

-

70

c

C1l

60

Q.

50

·> r:::

....C1l

', pH-STAT

'\,

------- I l' ---- -----r·--------r

---------t---J.. f 1

(..)

a-STAT

40 30

Control

0'

37°C

27"C

10'

20'

30'

60'

Fig. 3. Mean and standard error of the mean of the whole body oxygen consumption rates (V0 2) for both alphastat and pH-stat animals at each of the sampling periods expressed as a percent of the normothermic control. With a 10° c decrease in temperature, vo2 in the alpha-stat animals decreased to about 50% of control. In the pH-stat animals the vo2 during hypothermia remained at more than 70% of the normothermic control.

Table II. Oxygen consumption (V02) and total blood flow (mean ± standard error of the mean) Temperature (C)

vo2 (ml/minfkg)

% Control V02

36.4 ± 0.3 27.2 ± 0.1

6.19 ± 0.49 3.26 ± 0.41

100.0 52.5 ± 4.0

37.3 ± 0.1 26.8 ± 0.1 p Value*

4.42 ± 0.44 3.04 ± 0.21

100.0 71.6 ± 6.6 p < 0.05

NS

Q/0 vo2

(Q1) datafromfive alpha-stat and five pH-stat animals MV02 (ml/min/100 gm)

Alpha-stat 6.44 ± 0.57 2.08 ± 0.16 3.21 ± 0.31 pH-stat 5.26 ± 1.23 1.47 ± 0.17 3.78 ± 1.17 NS p < 0.05

% Control MV02

Q/0 MV02

100.0 48.0 ± 5.7

2.40 ± 0.41

100.0 66.0 ± 7.0 p < 0.09

1.57 ± 0.16 p < 0.09

Qr . (ml/min/kg) 75.80 ± 8.88 62.75 ± 2.63 68.06 ± 9.48 59.59 ± 5.54

NS

Alpha-stat hypothermia different from pH-stat hypothermia.

brane oxygenators mentioned earlier. Temperature was controlled by routing extracorporeal blood flow throu.gh a heat exchanger. Temperature probes (Yellow Springs Instrument Co., Yellow Springs, Ohio) were placed in the esophagus and in the heat exchanger. The extracorporeal perfusion circuit was primed with fresh, heparinized blood from a donor pig, and the animals were perfused at normothermia (37° C). A mean aortic root pressure of 80 to I 00 mm Hg was maintained during normothermic CPB with a total blood flow rate of approximately 70 mlfminfkg. During hypothermia, the blood flow rate for both experimental groups was adjusted to 60 ml/minjkg. The actual values for total blood flow rate during hypothermia were 62.75 ± 2.63 mijminjkg for the alpha-stat animals and 59.59 ± 5.54 mij minjkg for the pH-stat animals (see Table II). Pressures were recorded on a Gould physiologic recorder (Spectramed Inc., Critical Care Division, Oxnard, Calif.).

Approximately 2 hours from anesthesia induction until the time of the first measurements was required to finish this surgical preparation. Measurements. A I 0-minute equilibration period was allowed before physiologic measurements were made. We were concerned that some of the differences in oxygen consumption we measured might be due to time-dependent deterioration of the surgical preparation. Therefore we maintained the desired pH for a 1-hour period during hypothermia. Control measurements were made at 3r C followed by measurements at 0, I 0, 20, 30, and 60 minutes at 27° C. Total blood flow was determined by counting the revolutions per minute on the previously calibrated pump and multiplying revolutions per minute by milliliters per revolution. Myocardial blood flow was determined by timing the coronary sinus and right ventricular drainage into a graduated cylinder. Blood oxygen contents were determined with the IL-282 Co-Oximeter.

Volume 100 Number 5

Oxygen consumption during hypothermia 7 0 3

November 1990

Oxygen consumption could then be determined for both the .whole body and the heart by the Fick principle:

7.55

V02 = (>t(Ca02- Cv02) Where Y02 is the oxygen consumption, reported as milliliters per minute per kilogram for the body and milliliters per minute per 100 gm for the heart; Q1 is the blood flow rate in deciliters per minute; and Ca02 and Cv02 are the arterial and venous oxygen contents, respectively, in milliliters of oxygen per deciliter. Because there was a large variation between animals, we also report the whole body and myocardial oxygen consumption rates as percent control. We also described the effect of temperature on the oxygen consumption rates in terms of the Q 10, or the change in rate per !Oo C change in temperature. The Q10 can be calculated with the following equation:

~I -:I: >

7.4

100 .!!!

90

~

80 70

-c E .... N

Ql

O u0

·> Qj c.

E

~ 60 50 40

Results To test differences in metabolic rate and venous Po2 resulting from differences in pH strategy, we attempted to maintain the pH within 0.05 pH units of the target pH. Thus the target pH during normothermia for both groups and for the pH-stat hypothermic group was 7.40 ± 0.05. For the alpha-stat hypothermic group, the target pH at 27° C was 7.55 ± 0.05. The values actually obtained are presented in Table I, and they are consistent with the values we desired. The stability of our preparation with time is illustrated in Figs. 2 and 3. As can be seen in Fig. 2, we were able to maintain the desired arterial pH values for both experimental groups for the 1-hour period during hypothermia. Fig. 3 illustrates the effect of temperature on oxygen consumption rate expressed as percent control. Note that there appears to be no time-dependent effect on the oxygen consumption rate. This lack of a time-dependent effect is verified by performing a linear regression on the 27° C data for both hypothermic groups. The resultant slopes are not statistically different from 0. Thus there was no change in pH or oxygen consumption rate with time over the 1-hour hypothermic test period. Because there was no time-dependent difference, we were able to pool all the data collected during hypothermia and compare differences using an unpaired t test. Originally we 14 had predicted that if oxygen consumption and cardiac output decreased proportionally during

7.45

.E

QIO= (R,jR2)10/(T2-TI) where R 1 and R 2 are the oxygen consumption (or any other) rates at temperatures T 1and T 2, respectively. A 010 of 2 means that the rate will. decrease to one half of its original value with a 10° C decrease in temperature. Similarly, a Q 10 of 3 means that the rate will decrease to one third of its original value, and a Q10 of I means that there is no change in rate with change in temperature. Many biologic systems have a Qw of approximately 2. In later experiments, the concentration of norepinephrine was measured by the radioenzymatic assay ofPeuler and Johnson. 20

7.5

Q.

ns

C

90

·~Ql 80

gGI.c uN

.

-

100

::E ....CP 0E a.. .... 0

z

70 60 50 40 27

30

35

37

Temperature ("C) Fig. 4. Pooled measurements from the !-hour hypothermic period are compared with the normothermic control measurements. Top, Our acid-base management strategies for the two hypothermic groups. Middle, Difference in whole body Y02 between the alpha-stat and pH-stat animals. Bottom, Difference in myocardial oxygen consumption (MV02).

hypothermia, then Ca02 - Cv02 would remain constant and venous Po 2 would decrease because of the temperature-induced leftward shift of the oxygen dissociation curve. Furthermore, we had predicted that because the leftward shift of the curve would be more pronounced with alpha-stat than with pH-stat conditions, the mixed venous Po 2 would belower in the alpha-stat animals than in the pH-stat animals. Surprisingly, as can be seen in Table I, venous Po2 was not different between the two hypothermic groups. In fact, the mean value was lower

704

The Journal of Thoracic and Cardiovascular Surgery

Willford eta/.

-'E

400

.......

Cl

.e Cll

c: ..c:

pH-STAT

300

·;:: Q.

Cll

c:

·a. Cll

...0

z

200 100

•

a-STAT

27

37 Temperature

rc)

Fig. 5. Measured norepinephrine concentrations from the last four experimental animals. Two of the animals were from the alpha-stat group and two from the pH-stat group. Samples were drawn during the normothermic control and at the end of l hour at 27° C. Note the dramatic increase in norepinephrine in the pH-stat animals.

(but not significantly) inthe pH-stat animals than in the alpha-stat animals. Table II and Figs. 3 and 4 summarize oxygen consumption measurements. A 10° C decrease in temperature resulted in an oxygen consumption rate of approximately 50% of the normothermic value in the alpha-stat animals. In the pH-stat animals in 10° C decrease in temperature resulted in an oxygen consumption of approximately 75% of the normothermic value. In other words, although the Q10 for whole body oxygen consumption was 2.08 for the alpha-stat animals, it was 1.57 for the pH-stat animals. The beating, nonworking heart studied in our experiment showed a similar pattern. The myocardial oxygen consumption rate Qw was 2.40 in the alpha-stat animals but only 1.57 in the pH-stat animals. As it became apparent that the oxygen consumption rates were unexpectedly high in the pH-stat animals, we began searching for a mechanism to account for this observation. In the last four animals (two pH-stat and two alpha-stat), we measured the concentration of norepinephrine. As can be seen in Fig. 5, although the concentration of norepinephrine remained below 100 pg/ml in the alpha-stat animals, it increased dramatically (average 474%) in the two pH-stat animals. Discussion

We were intrigued by the finding of an unexpectedly high oxygen consumption of the pH-stat animals. The high oxygen consumption rate observed in the pH-stat

animals was exactly counter to our original hypothesis about the metabolic effects of alpha-stat and pH-stat strategies. Although we did not measure myocardial function, our hypothesis concerning the metabolic effects of the two strategies was based in part on previous studies6• 7 that did report the effects of different acid-base management strategies on myocardial function during hypothermia. These studies have reported that myocardial contractility in hypothermic dogs is maintained higher in alpha-stat regulation than with pH-stat regulation. However, these previous studies are difficult to interpret because the "alpha-stat" pH used in both cases was higher than the predicted alpha-stat value. For instance, McConnell and associates7 compared pH 7.4 with pH 7. 7, which was more alkaline than the predicted alpha-stat value of 7 .55. Becker and co-workers6 actually compared alpha-stat with a more alkaline pH of approximately 7.8. 10 In a study that actually compared alphastat with pH-stat acid-base regulation, Swain, White, and Peters9 reported better cardiac electrical stability with alpha-stat regulation. Whereas the oxygen consumption rate decreased during hypothermia with alpha-stat regulation, as expected, oxygen consumption decreased only half as much during pH-stat regulation. Although shivering often is responsible for elevating oxygen consumption during accidental hypothermia, shivering was blocked with succinylcholine in our animals. We believe that the increased oxygen consumption in the pH-stat animals is consistent with

Volume 100

Oxygen consumption during hypothermia 7 0 5

Number 5 November 1990

activation of the sympathetic nervous system with this ·strategy. Catecholamines have been implicated in the initiation of nonshivering thermogenesis, which increases oxygen consumption. 21 Maintaining the pH constant at 7.4 may actually represent an acidosis during hypothermia if the in vivo control system is designed to defend the charge ratio on alpha-imidazole groups. Evidence that 22 this might be the case was recently obtained by Nattie, acidosis to response who was able to block the ventilatory by painting the medullary chemoreceptors of cats with diethyl pyrocarbonate, an agent that binds the imidazole group. Evidence for the hypothesis that the sympathetic nervous system is being activated by a "perceived" acidosis with pH-stat management is also consistent with the observation that inhalation of 5% to 6% carbon dioxide increases heart rate and cardiac output in both man and dog. 23 · 24 Simon and Riedel 25 have investigated the effects of severe hypercapnia (10% carbon dioxide) in rabbits and have found it to increase sympathetic nervous system activity. Interestingly, the effect was preserved even after denervation of the peripheral chemoreceptors. Because the sympathetic response to hypercapnia was preserved under these . conditions, they postulate that other chemoreceptive structures, possibly the medullary chemoreceptors, were able to mediate the response to increased arterial carbon dioxide. We considered the apparent activation of the sympathetic nervous system with pH-stat acid-base management such an important finding in our study that we have recently attempted to verify it in another series of animal experiments. We used 17 halothane-anesthetized pigs cooled to 29 o C in a non-CPB closed chest model. This preliminary study also suggests that maintaining the pH at 7.4 during hypothermia activates the sympathetic nervous system, leading to an increase in norepinephrine 26 concentration and the oxygen consumption rate. The effect of the increased oxygen consumption observed in our pH-stat animals was to depress the mixed venous Po 2. We .had previously determined that the venous Po2 was very sensitive to changes in oxygen consumption. 14 Presumably the venous Po 2 reflects the average tissue Po 2,27 and there is a critically low venous Po2 below which aerobic metabolism in the tissues cannot be maintained. 28 Cain and Bradley29 speculated that, because of the decreased oxygen consumption rate in hypothermia, the critical Po 2 would also be lower. Subsequently, we 30 observed this to be the case in anesthetized pigs. We found that in normothermic pigs the critical mixed venous Po 2 was 22 mm Hg. Below this value the oxygen consumption rate decreased linearly and blood lactate accumulated. During hypothermia we found the critical venous Po 2to be about 15 mm Hg at 29° C. Thus

we found that the critical venous Po2 was decreased during hypothermia. Our data are consistent with Cain's hypothesis that critical venous Po2 should be lower at the lower oxygen consumption rate that occurs during hypothermia.29 It is also consistent with our previous theoretical predictions in which we 14 explained Cain's hypothesis in terms of Fick's law of diffusion. That is, since the venous blood is representative of end capillary blood, we believe the decrease in critical venous Po2 occurs because a lower diffusion driving pressure for oxygen is required as the oxygen consumption rate decreases, as it does during hypothermia. Critieal Po2 and its relationship to oxygen consumption is an important concept because clinically induced hypothermia is used primarily to protect the tissues from hypoxic injury. In the current study we have found a significantly higher oxygen consumption rate with pH-stat acid-base management than with alpha-stat acid-base management during moderate hypothermia at 27° C. Thus there may be less protection of tissues from hypoxic damage during hypothermia with pH-stat acid-base management. We gratefully acknowledge the valuable contributions of Cynthia J. Willford and Muriel Spooner.

1.

2.

3. 4.

REFEREN CES Rosenthal TB. The effect of temperature on the pH of blood and plasma in vitro. J Bioi Chern 1948;173:25-30. Shelton G, Jones DR, Milsom WK. Control of breathing in ectothermic vertebrates. In: Handbook of Physiology. Section 3: The respiratory system. Vol II. Bethesda, Maryland: American Physiological Society, 1986:857-909. White FN. A comparative physiological approach to hypothermia. J THORAC CARDIOVASC SURG 1981;82:821-31. Reeves RB. An imidazole alphastat hypothesis for vertebrate acid-base regulation: tissue carbon dioxide content and body temperature in bullfrogs. Resp Physiol 1972; 14:219-36.

5.

Rahn H, Reeves RB, Howell BJ. Hydrogen ion regulation, temperature and evolution. Am Rev Resp Dis 1975;

112:165-72. 6. Becker H, Vinten-Johansen J, Buckberg GD, et al. Myocardial damage caused by keeping pH 7.40 during deep

systemic hypothermia. J THORAC CARDIOVASC SURG 1981 ;82:81 0-20.

7. McConnell DH, White FN, Nelson RL, et al. Importance of alkalosis in maintenance of "ideal" blood pH during hypothermia. Surg Forum 1975;26:263-5. 8. White FN, Somero G. Acid-base regulation and phospholipid adaptations to temperature: time courses and physiologic significance of modifying the milieu for protein function. Physiol Rev 1982;62:40-90. 9. Swain JA, White FN, Peters RM. The effect of pH on the

70 6

The Journal of Thoracic and Cardiovascular Surgery

Willford et a/.

hypothermic ventricular fibrillation threshold. J THORAC CARDIOVASC SURG 1984;87:445-51. 10. Swain JA. Regulation of pH during hypothermia [Letter]. J THORAC CARDIOVASC SURG 1983;85:147-8. II. Buckberg GD. Regulation of pH during hypothermia [Reply]. J THORAC CARDIOVASC SURG 1983;85:148-9. 12. Osborn JJ, Gerbode F, Johnston JB, Ross JK, Ogata T, Kerth WJ. Blood chemical changes in perfusion hypothermia for cardiac surgery. J THORAC CARDIOVASC SURG 1961;42:462-76. 13. Wood SC. Adaptation of red cell function to hypoxia and temperature in ectothermic vertebrates. Am Zoo! 1980; 20:163-72. 14. Willford DC, Hill EP, Moores WY. Theoretical analysis of oxygen transport during hypothermia. J Clin Monitor 1986;2:30-43. 15. Hill EP, Willford DC, Moores WY, Bellamy R, Heydorn WH. Oxygen transport and oxygen consumption vs. cardiac output at different haematocrits. Perfusion 1987;2:3950. 16. Willford DC, Hill EP. Modest effect of temperature on the porcine oxygen dissociation curve. Resp Physiol 1986; 64:113-23. 17. Willford DC, Hill EP. Temperature corrections for blood gas values. In: Tembleson MD, ed. Swine in biomedical research. Vol2. New York: Plenum Press, 1986:1429-37. 18. Moores WY, Hannon JP, Crum JD, Willford DC, Rodkey WG, Geasling JW. Coronary flow distribution and dynamics during continuous and pulsatile extracorporeal circulation in the pig. Ann Thorac Surg 1977;24:582-90. 19. Moores WY, De VenutoF, Heydorn WH, eta!. Extending the limits of hemodilution on cardiopulmonary bypass using stroma-free hemoglobin solution. J THORAC. CARDIOV ASC SURG 1981 ;81: 155-62. 20. Peuler JD, Johnson GA. Simultaneous single isotoperadioenzymatic assay of plasma norepinephrine, epinephrine, and dopamine. Life Sci 1977;21 :625-36. 21. Horwitz BA. Metabolic aspects of thermogenesis: neuronal and hormonal control. Fed Proc 1979;38:2147-9. 22. Nattie EE. Diethyl pyrocarbonate (an imidazole binding substance) inhibits rostral VLM C02 sensitivity. J Appl Physiol1986;61:843-50. 23. Korner PI. Integrative neural cardiovascular control. Physiol Rev 1971;51:312-67. · 24. McGregor M, Donevan RE, Anderson NM. Influence of carbon dioxide and hyperventilation on cardiac output in man. J Appl Physiol1962;17:933-7. 25. SimonE, Riedel W. Diversity of regional sympathetic outflow in integrative cardiovascular control: patterns and mechanisms. Brain Res 1975;87:323-33. 26. Willford DC, Hill EP, Schaffartzik W, Bain R, Ziegler MG. Elevated norepinephrine level and oxygen consumption rate (V02) with pH-stat acid-base management during hypothermia [Abstract]. Physiologist 1989;32:203. 27. Tenney SM. A theoretical analysis of the relationship between venous blood and mean tissue oxygen pressures. Respir Physiol 1974;20:283-96.

28. Cain SM. Oxygen delivery and uptake in dogs during anemic and hypoxic hypoxia. J Appl Physiol1977;42:228-34. 29. Cain SM, Bradley WE. Critical 0 2 transport values at lower body temperatures in rats. J Appl Physiol 1983; 55:1713-7. 30. Willford DC, Hill EP, White FC, Moores WY. Decreased critical mixed venous oxygen tension and critical oxygen transport during induced hypothermia in pigs. J Clin Monit 1986;2:155-68.

Discussion Dr. Hillel Laks (Los Angeles, Calif.). Dr. Willford and his group in San Diego have made many contributions to the understanding of acid-base management over the years, and this is certainly a very important and interesting clinical problem. They have shown that the pH management during hypothermia is important and that there are various factors at work which can either improve or reduce oxygen consumption. My first question concerns methods. Dr. Willford, you observed that the baseline oxygen consumption for the pH-stat group was actually lower at the control normothermic level than for the alpha-stat group. Although there was a percentage decrease with hypothermia that was statistically significant, the absolute value of oxygen consumption during hypothermia in the alpha-stat and pH-stat groups was actually the same or very close, around 3. Could you comment on this question of the statistical interpretation? Dr. Willford. In a large number of animals in which we have measured oxygen consumption rate during normothermia, we have seen oxygen values ranging from about 3 ml/min/kg body weight to 7 or 8 mljminjkg body weight. The average in a large number of animals is usually around 5 ml/min/kg. As it turned out, the five pH-stat animals had a slightly lower oxygen consumption rate. They tended to be on the lower end of the spectrum, and the five alpha-stat animals had a slightly higher value in absolute terms. That was one of the reasons we chose to express these data in terms of percent control, using each animal as its own control. Dr. Laks. From a statistical point of view, one way that this has been handled in some of our studies was to pool the control values and then compare each group with that pooled group of controls, looking for statistical significance. Were you able to do that in your study? Dr. Willford. We certainly could, but we chose to use each animal as its own control. We may reevaluate the data with your comments in mind. Dr. Laks. Another elegant study that you did earlier compared the oxyhemoglobin dissociation curves of the pig, the human, and the dog at various temperatures. You showed that the pig oxyhemoglobin dissociation curve is less affected by temperature than that of the human or the dog. Would this make it a less viable experimental model than perhaps the dog? Dr. Willford. Yes, that is a disadvantage of using the pig. What Dr. Laks is saying is that the effect of temperature on the pig oxygen dissociation curve is only about 70% of that on the human dissociation curve, and that is a slight disadvantage. All else being equal, the mixed venous Poz does not decrease as much during hypothermia in the pig as it does in humans. However, the pig offers some other advantages in terms of its physiologic responses in our study. Dr. Laks. I was also interested in the norepinephrine levels

Volume 100 Number 5 November 1990

in your study. They were elevated in the pH-stat group, which implies that the body somehow reacts to hypothermia with an increasing norepinephrine release. Kevin Turley and his group have studied deep hypothermia in puppies and have shown a profound increase of catecholamine levels, particularly with surface cooling. It is interesting that in the alpha-stat group this was not the case whereas you found a profound increase in the four animals in the pH-stat group. Was there any other sign of norepinephrine release, such as a change in the peripheral resistance in heart rate or contractility, to support this explanation? Dr. Willford. Yes. In this study there was a suggestion of increased contractility but no changes in peripheral resistance. In another study of 23 pigs, which I am currently conducting, we have seen a statistically higher rate of rise of left ventricular pressure in the pH-stat animals. We have yet another study ongoing at this time, in which we are cooling the animals to a much lower temperature. This allows a better separation in the pH groups. Again, the alpha-stat animals show no increase in norepinephrine levels but the pH-stat animals, at 22° C, have norepinephrine concentrations exceeding 1000 pgjml, over 10 times those observed in the alpha-stat animals at the same temperature. Dr. Laks. We did a study in the isolated neonatal perfused heart in which we controlled pH in the acid-stat, pH-stat, alpha-stat, and alkaline-stat methods. With I hour of perfusion

Oxygen consumption during hypothermia 7 0 7

at 20° C these different pH levels had no effect on myocardial function after rewarming. Do you have any information about increased tolerance of the neonatal heart to these different pH controls? Dr. Willford. Were they isolated from the central nervous system? Dr. Laks. These were isolated hearts, and therefore there was no effect of epinephrine release because the blood was cooled as it reached the isolated heart and then rewarmed before it returned to the support pig. Dr. Willford. I have no specific knowledge of the neonatal heart. However, I think that Andre Malan in Strasbourg has suggested that the heart is fairly capable of defending its intracellular pH. Thus the sort of metabolic depressant effect one would expect from acidity may not occur so much in the heart as elsewhere in the animal. I should mention that Malan's studies of intracellular pH were conducted in hibernating mammals. Dr. Laks. This is a very important subject, first broached in Paul Ebert's classic paper in 1962, in which he controlled pH and showed that there was no effect on general body oxygen consumption with acidic pH, but that myocardial function seemed to be impaired by that perfusion. There have been many other controversial studies since then, and your study certainly adds to our body of knowledge.