Electrocardiographic Changes During Surface-Induced Deep Hypothermia

Electrocardiographic Changes During Surface-Induced Deep Hypothermia

Electrocardiographic Changes During Surface-Induced Deep Hypothermia The Influence of Ether, Halothane, Carbon Dioxide, and Perfusion Rewarming Murra...

1MB Sizes 0 Downloads 29 Views

Electrocardiographic Changes During Surface-Induced Deep Hypothermia The Influence of Ether, Halothane, Carbon Dioxide, and Perfusion Rewarming

Murray P. Sands, Shigekazu Sato, M.D., Hitoshi Mohri, M.D., Warren G. Guntheroth, M.D., and K. Alvin Merendino, M.D. ABSTRACT The influence of halothane, ether, carbon dioxide, and perfusion rewarming on the electrocardiogram was studied in 37 ,dogs subjected to surface-induced deep hypothermia. Significant anesthetic-related differences in P-R,QRS, Q-T, and R-R intervals during cooling were not apparent; however, reduced arterial pressure, ventricular fibrillation,and a greater tendency for bradycardia requiring supportive measures were noted at low temperatures with halothane anesthesia. The use of 95% 0 2 1 5 % COZ significantlyreduced the QTc at low temperatures. Other phenomena, including the occurrence and significance of J waves, are discussed. The relationship of the electrocardiogram to clinical and pathological results was evaluated and indicates that (1) properly managed resuscitation (manual massage and defibrillation) is not a serious hazard, (2) ether in 100% oxygen is the agent of choice for surfaceinduced deep hypothermia with prolonged circulatory arrest, and (3)halothane may be used in a procedure combining surface cooling and perfusion rewarming if given in a mixture of oxygen and carbon dioxide.

T

here are basically two successful approaches to hypothermia for cardiac operations in infants. One technique, as described by Mohri and associates [5, 71, relies on the use of surface cooling and rewarming under deep ether anesthesia given in 100% oxygen. Subsequently Barratt-Boyes and co-workers [I] modified the surface cooling method by following it with a short period of perfusion cooling and perfusion rewarming under halothane anesthesia given in oxygen and carbon dioxide. The advantages of the earlier technique include complete absence of ventricular fibrillation or other dysrhythmias during cooling, avoidance of the difficulties of extracorporeal perfusion in small infants, and attainment of safe circulatory arrest for periods up to one hour. The latter method utilizes a nonexplosive anesthetic and obviates the need for resuscitative procedures such as cardiac massage. The nonexplosive nature of halothane has prompted us to explore its use experimentally as the anesthetic agent in pure surface-induced hypothermia without perfusion. Unfortunately, halothane has proved inferior to ether for From the Department of Surgery, Divison of Cardiothoracic Surgery, and the Department of Pediatrics, University Hospital, University of Washington School of Medicine, Seattle, Wash. This study was aided by U.S. Public Health Service Grant No. 13517, Unit 7, and NICH No. HD 02272, American Heart Association, and the Northeastern Chapter of the Washington State Heart Association Grants. Accepted for publication Nov. 25, 1974. Address reprint requests to Dr. Mohri, Department of Surgery RF-25, University Hospital, Seattle, Wash. 98195.

386

T H E ANNALS OF THORACIC SURGERY

ECG Changes during Hypothermia surface cooling to deep levels, particularly if surface rewarming is to follow clinically useful periods of circulatory arrest. The untoward effects of halothane with this type of procedure are discussed later; many, however, relate to cardiac function and include ventricular fibrillation prior to elective arrest, severe bradycardia at levels of deep hypothermia, and low cardiac output following resuscitation. The current study was undertaken to determine if there were electrocardiographic differences that might help explain the poor results with surface cooling and rewarming under halothane anesthesia as compared with the good results achieved with ether anesthesia. In addition, the effects of carbon dioxide, perfusion rewarming, and duration of circulatory arrest on the electrocardiogram and clinically judged results were evaluated.

Materials and Methods A method of surface-induced hypothermia developed in this laboratory has been detailed elsewhere [5, 71. The essentials of the method include deep ether anesthesia given in 100% 0 2 , maintenance of ventilation at levels adequate for normothermia so as to gradually induce respiratory alkalosis (pH = '7.6 to 7.8) during cooling to 18" to 20°C, and administration of 10%low-molecular-weight dextran (10 ml per kilogram of body weight) between 35" and 25°C. A total of 37 adult mongrel dogs of both sexes, weighing 14 to 22 kg, were used in this study. Dogs were divided into seven groups, and in each of these only one variation of the method was employed, such as the anesthetic agent, the gas mixture, or the use of core (perfusion) rewarming. Details of the method for each group are shown in Table 1. Anesthesia was induced with 30 mg per kilogram of thiamylal sodium given intravenously. Animals were then intubated and anesthesia was maintained with the appropriate agent, ether or halothane, for the particular study being done. Ventilation was maintained with a Palmer vclT-me-constant respirator. Ether in 100%O2was administered through a closed circuit with soda-limecanisters. Ether in 95% 0 2 was provided from a copper kettle vaporizer in a nonrebreathing semiclosed circuit. This same circuit with a Fluotec vaporizer was used in the TABLE 1 . DETAILS OF TECHNIQUE

No. of Dogs

Anesthetic Used

Breathing Mixture

5

Ether

100% 0

6

Ether Ether Ether Halothane Halothane Halothane

100% 0, 95% 0215% COz 95% OJ5% C o t 100% 02. 95% 02/5% Coz

4 5 5 6 6

2

100% 0,

Circulatory Arrest Time (min)

0 30 60 30 30

30 30

VOl.. 19, NO. 4.

Rewarming Method Surface (warm water immersion) Surface Surface Surface Surface Core (bypass) Core (bypass)

APKIl..

1975

3x7

SANDS ET AL. groups receiving halothane. Arterial and venous catheters for pressure monitoring, blood sampling, and infusion purposes were installed by femoral cutdown. Rectal and esophageal temperatures were monitored with Yellow Springs Telethermometer probes. All dogs were cooled by immersion in ice water. Active cooling was stopped when body temperature reached 22°C; thereafter the temperature continued to drift downward to 17" to 18°C. Except for dogs in the ether-nonarrest group, which were rewarmed after a 30-minute stabilization period, all dogs underwent thoracotomy through the right fourth intercostal space immediately after the cessation of active cooling. Circulatory arrest was established by occluding the cavae and cross-clamping the aorta and pulmonary artery. Cardioplegia was induced by injecting Young's solution* into the aortic root proximal to the crossclamp. During the occlusion period, dogs to undergo core rewarming were prepared by cannulation with a Bardic catheter to the right atrium and an arterial catheter to the ascending aorta. The bypass circuit included an Olson roller pump and a Bentley Temptrol pediatric-sized oxygenator with an additional heat exchanger. The pump prime consisted of fresh acid-citrate-dextrose (ACD)blood to which heparin had been added; also added was 600 mg calcium chloride and 4.5 mEq sodium bicarbonate per unit. In the groups undergoing core rewarming, resuscitation and rewarming were achieved by starting perfusion and gradually establishing a flow of 100 ml per kilogram after occlusive tapes and clamps were released. Water temperatures in the heat exchangers were gradually increased to a maximum of 40°C. Defibrillation,when necessary, was done as soon as it seemed possible. At the completion of rewarming the animals were decannulated and the chest closed. Circulation was reestablished in the groups undergoing surface rewarming by ending caval and arterial occlusion and applying gentle cardiac massage. Calcium chloride (300 mg) was given intravenously to counteract the potassium effect of Young's solution. Occasionally,one or two 50-volt AC electroshocks were required to defibrillate the heart. Usually an additional 100 or 200 mg of calcium chloride and mechanical or electrical pacing for a few minutes were all that were required to maintain adequate cardiac output except in the one group of dogs which underwent an arrest of 60 minutes; in these dogs 50 pg doses of epinephrine were used in the early postresuscitation stage. As soon as cardiac massage was started, surface warming was initiated by floating the animals in warm water (42°C)with a plastic sheet interposed. Rewarming was terminated when the rectal temperature reached 35°C. The six standard limb ECG leads were monitored utilizing a Hewlett-Packard 8811A amplifier and model 7788A recorder. This same machine with appropriate amplifiers was used to monitor and record the arterial and venous pressures. ECG leads were recorded at 25 mm per second paper speed immediately *Young'ssolution: 0.81 gm potassium citrate,2.46 gm magnesium sulfate, 0.001 gni neostigmine methylsulfate, water in sufficient quantity to make 100 ml, and pH adjusted to 7.4 with sodiuin bicarbonate.

3 88

T H E ANNALS OF THORACIC SURGERY

ECG Changes during Hypothermia after induction of anesthesia; prior to cooling; during cooling at 35", 30", 25", and 20°C; and during rewarming at 20", 25", 30", and 35°C. Lead I1 was constantly monitored for unusual rhythms or patterns. P-R, QRS, Q-T, and R-R intervals from each ECG were measured, and a careful inspection was made for changes in wave form. Statistical analyses were done with a CDC-6500 computer and included correlation and paired t-test matrices for determination, means 1 standard deviation, Q-T/R-R ratios, QTc (Q-T and unpaired Student's t-test matrices, comparing each measurement to its counterpart in the dogs undergoing 30-minute arrest using ether in 100% 0 2 .

m),

*

Results All dogs were long-term survivors. The noticeable hemodynamic differences were confined to lower arterial blood pressures during cooling in groups receiving halothane; there was also one episode of ventricular fibrillationjust prior to elective arrest in a dog receiving halothane in 95% oz/5% COz. There was no postoperative neurological disturbance in the form of "highstepping gait" [5] in any of the dogs receiving ether anesthesia except a mild to moderate involvement of the forelimbs in the dogs undergoing 60-minute arrest. All 5 dogs receiving halothane in 100%oxygen which were rewarmed by perfusion developed high-stepping gait, as did 1 of the 5 receiving halothane in 5% COz that were surface-rewarmed. None of the dogs receiving halothane in 5% COz and perfusion rewarming developed postoperative neurological disturbance. The general effects on the ECG observed in all groups included a progressive decline in heart rate and prolongation of the P-R, QRS, and Q-T intervals as body temperature was reduced. Rewarming resulted in a reversal of this sequence. P waves widened with cooling and occasionally became notched at temperatures below 25°C. Cooling did not cause significant QRS vector changes in the presence of a conducted S-A rhythm. Nonspecific S-T and T changes, such as S-T depression or elevation and flattened T waves, were universally present below 30°C. Diphasic T waves and T direction reversals were also seen and, as the temperature dropped to around 20"C, it was not unusual for previously isoelectric or diphasic T waves to become large and broad.

Effects of Alterations in Anesthetic, Breathing Mixture, and Duration of Circulatoly Arrest RHYTHM

With 1 exception, sinus rhythm was retained during cooling regardless of anesthetic or breathing mixture until the body temperature had decreased below 20°C. The heart of 1 dog receiving halothane in 95% 0 2 / 5 %COz fibrillated at the end of the cooling period, possibly as the result of manipulation during operative preparations for elective arrest. During surface rewarming, ether-anesthetized animals receiving 100% 0 2 with circulatory arrest periods of up to 30 minutes all returned to sinus rhythm by 20°C. When the circulatory arrest period was extended to 60 minutes, 25% ( 1 of 4) VOI.. 19. NO. 4, APRIL, 1975

389

SANDS ET AL. returned to sinus rhythm by 20°C and all were in sinus rhythm by 30°C.Forty percent (2of 5) of the ether-anesthetized dogs receiving 95% O3/5% CO, and subjected to 30 minutes of circulatory arrest were in sinus rhythm by 20°C during rewarming, and the remaining 60% (3of 5)reverted by the time the temperature reached 25°C. Of the dogs given halothane in 100%On that were rewarmed by perfusion and subjected to 30 minutes of circulatory arrest, 50% (3of 6)had returned to sinus rhythm at 20°C during rewarming, with the rest reverting by 30°C.Dogs given halothane in 95% 0 ~ / 5 % C02 that were rewarmed by perfusion and also subjected to 30 minutes of circulatory arrest showed a 67% return (4of6) to sinus rhythm by 20°C during rewarming, with the remainder reverting by 30°C.This information is summarized in Table 2. P-RINTERVAL

The P-R intervals in all dogs were similar during cooling to 25°C.At 20°C ether-anesthetized dogs tended to have shorter P-R intervals, but the difference was not statistically significant. During rewarming there were no P-R interval differences that could be attributed to the anesthetic agent (ether or halothane) when a sinus rhythm was present. Similarly, the use of 5% C02with either ether or halothane did not seem to cause significant P-R alterations during warming. However, among the three groups of dogs cooled with ether in 100% O2 with either no arrest, 30 minutes arrest, or 60 minutes arrest, P-R intervals during rewarming at 20",25",and 30°C were significantly shorter Cp < 0.05, < 0.01, and < 0.02,respectively) in the nonarrested dogs. Furthermore, when the P-R intervals during rewarming of dogs undergoing cardiac arrest were compared with the precooling control value, all were significantly greater @I < 0.05),whereas in the dogs not undergoing arrest, rewarming P-R intervals taken above 20°C were not significantly greater than the precooling control Cp > 0.05). QRS COMPLEX

During coolingQRS trends were similar in all groups (Fig. 1).However, when compared to the dogs given ether, the halothane-anesthetized animals showed TABLE 2. NUMBER OF ANIMALS IN CONDUCTED SINUS RHYTHM A T DESIGNATED TEMPERATURE DURING REWARMING

Method

Rectal Temperature 20°C 25°C

30°C

Surface rewarming Ether in 100% 0 2 , no arrest Ether in 100% 02,30 inin arrest Ether in 100% 0 2 , 60 inin arrest Ether in 5 % C 0 2 ,30 inin arrest Halothane in 575 C02, 30 inin arrest

5of5 6of6 1 of4 2of5 1 of5

... 3 of 4 5 of 5 5of5

4 of 4

Core rewarming Halothane in 100% 0 2 , 30 inin arrest Halothane in 57%C02, 30 min arrest

3of6 4 of 6

5of6 ...

6of6 6of6

390

THE ANNALS OF THORACIC SURGERY

...

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

ECG Changes during Hypothmia significantly shorter QRS durations at normothermia, 35", and 20°C (P values < 0.05, < 0.02,< 0.005, respectively). Although QRS complexes became wider, there was little change in pattern during cooling to 22" to 20°C.Below this temperature thoracotomy and cardiac manipulation resulted in changes typical of direct mechanical ventricular stimulation. With the single exception of the animal receiving halothane in 95% 02/5% COz whose heart fibrillated, all hearts maintained adequate output and most immediately returned to a unifocal sinus or AV nodal rhythm even at 18°C (the usual temperature for induced cardioplegia) when interfering stimuli ceased. The configuration of the QRS complex generally reflected the stage of rewarming regardless of anesthetic or breathing mixture. Immediately following resuscitation and for a period of 15 to 30 seconds, the QRS usually originated from a low ventricular focus. After this short initial period, a unifocal rhythm with normally oriented complexes supervened. No significant differences in QRS duration were detected in any group undergoing circulatory arrest during rewarming. In the dogs receiving ether in 100%O2but not undergoing arrest, the QRS duration was shorter only at 20°C during rewarming (P < 0.02). Q-T INTERVAL

Throughout cooling and rewarming there were no statistically significant differences between Q-T interval changes in any of the groups receiving ether anesthesia. FIG. 1. Mean values (in msec) of P-R, QRS, Q-T, and R-R interval data from the four major groups in this study.

35

30

25

20

20

25

30

35

O C

VOI..

19, NO. 4. APRII.. 1975

391

SANDS ET AL. Dogs given halothane in 100% 0, had significantly greater Q-Tintervals than those receiving ether prior to cooling, at 30" and 25°C during cooling, and at 25°C during (perfusion) rewarming (p < 0.01, < 0.05, < 0.05, and < 0.02). When halothane was provided in 95% OJ5% C02, significant differences occurred prior to cooling (p < 0.005) and at 25°C during (surface or perfusion) rewarming (1, < 0.02). R-R INTERVAL

The only statistically significant difference in R-R intervals between any of the groups receiving ether anesthesia occurred at 25°C during rewarming in the group being given 95% o d 5 % COz (p < 0.05), at which temperature the R-R interval was greater. Comparing dogs receiving halothane in 100%O2and rewarmed by perfusion to those given ether in 100% O2 revealed a significantly greater R-R interval (p < 0.001) immediately following induction of anesthesia but prior to cooling. Although the numerical differences between the means for R-R intervals between these two groups was fairly well preserved during cooling to 20°C and rewarming to 30"C, the differences were not statistically significant due to considerably increased standard deviations in the halothane group. In dogs receiving halothane in 95% 02/5%C 0 2 there were considerably longer R-R intervals as compared to dogs given ether in 100% 0 2 . The differences were significant during cooling to 25°C (p < 0.02). This trend continued at 25°C and below but the differences were not significant. When undergoing surface rewarming these animals showed significantly longer R-R intervals at 20" and 25°C (1, < 0.005), while perfusion rewarming resulted in prolonged R-R intervals at 25" and 30°C (p < 0.05 and < 0.01). These interval data are summarized in Figure 1. QTC ( Q - T K R )

Figure 2 shows the mean QTc for each group. All groups had significantly increased mean QTc (p< 0.05) by 30°C during cooling. In animals anesthetized with ether in 100%02 there was a gradual increase in QTc of from 0.36 2 0.044 prior to cooling to 0.69 2 0.06 1 at 20°C during cooling. During rewarming after 30 minutes of circulatory arrest the QTc declined fairly abruptly from 0.66 2 0.044 at 20°C to 0.48 2 0.066 at 25°C and then continued to decline slowly to a value of 0.41 2 0.045 at 35°C. In contrast to results in this group, the group given ether in 95% Od5% C 0 2had lower QTc throughout cooling and rewarming, the differences being significant (p < 0.05) at 25" and 20°C during cooling. The dogs given halothane in 100% O2 had QTC somewhat larger than those given ether in 100% 0 2 until 20°C during cooling, although the differences were not significant. At 25°C during rewarming, dogs given halothane in 100%0 2 had significantly greater QTc as compared to those receiving ether, but this difference gradually diminished and was negligible by 35°C. As in those gettingether, the use of 95% 0 2 1 5 % C02 resulted in smaller QTc values throughout cooling and rewarming under halothane anesthesia. These differences were significant at 25°C during cooling @ < 0.05) but not at other points (Fig. 2). 392

THE ANNALS OF THORACIC SURGERY

ECG Changes during Hypothemnia

-

D.--

Ether/100% 02 Elher/95% q / 5 % C q

6-d

Halothanc/95%02/5%C~

0-0

Halothone /loo% 02

FIG. 2 . Mean QTc ( Q - T m )values. The w e of carbon dioxide consistmtly results in smaller QTc values during surface hypothemia. J WAVES

Characteristic J deflections or early positive S-T junctional deflections are represented in Figure 3. These deflections appeared in 50% of the dogs given ether in 100 0 2 but only at temperatures of 20°C or less, and all were of relatively low amplitude. The incidence was also 50% in dogs given ether in 95% 02/5% COZ,and again all occurred at 20°C or less. Two of 6 dogs given halothane in 100%0 2 developed J waves, but the onset was much earlier (at 35" and 25°C) and the amplitude of deflection greater. Only 3 of the 11 dogs cooled with halothane in 95% 0215% C02 developed J waves at 20" and 22°C. J activity was not seen during rewarming in any dogs. U WAVES

Two dogs receiving ether in 100%O2developed transient U waves following the induction of anesthesia and prior to cooling. Another dog receiving ether in 95% 0915% COi also showed U waves following anesthesia induction. N o dog

FIG. 3. J waves. These early S-T junctional deflections (arrows) may vary in amplitude and duration. Characteristically, their onset is in the terminal portion of the major QRS deflection. Both examples are of Lead 11 taken at 20°C during cooling under standard conditions. (1 mu = 1 cm deflection at 25 mmlsec paper speed.)

VOL. 19. NO. 4, APRIL, 1975

393

SANDS E T AL. receiving ether had U waves at temperatures below 35°C during cooling. During rewarming, 2 receiving ether in 100%0, had transient U activity at 25" and 35"C, respectively. Two of the 6 dogs in the group given halothane in 100% 0, developed U waves followinganesthesia induction. One of these also showed a brief period of U activity at 25°C during cooling. No animal given halothane in 95% 0215% CO, developed U waves. Although U waves occurred in some dogs that also developed J deflections, the two were never seen simultaneously.

Comment Surface cooling to 20°C followed by 30 minutes of circulatory arrest and then surface rewarming produces markedly different results with our technique if halothane is used rather than ether anesthesia [6].Halothane anesthesia appears to increase the danger of ventricular fibrillation at levels of deep hypothermia and arterial blood pressures are somewhat reduced. Resuscitation is not difficult after 30 minutes of circulatory arrest, but vasopressors or cardiotonic drugs are necessary to maintain cardiac output and the initial rewarming phase is prolonged. Postoperatively all dogs developed a neurological disturbance in the form of "high-stepping gait" [6]. None of these problems have been encountered with ether under similar conditions. In contrast to these clinically observed differences, there are no clearly important differences in P-R, QRS, Q-T, R-R, or QTc intervals when halothane as opposed to ether anesthesia was used with our technique. This may be due to two factors: (1) many important physiological events elicit little or no ECG response, and (2) the importance of qualitative ECG changes may not be revealed by quantitative analysis of the various ECG intervals. Trends were observed, however, that may help explain the clinical differences. Relatively greater lengthening of P-R and R-R intervals in halothaneanesthetized dogs at below 25°C suggests that bradycardia or asystole may occur earlier with this agent. This observation is further supported by another study currently in progress utilizing halothane in 100% 0, with surface cooling and rewarming but without thoracotomy or circulatory arrest: in this series all 12 dogs developed severe bradycardia at around 20°C and required pacing and cardiotonic drugs; in 2 dogs externally applied electroshock was necessary for hearts that fibrillated in response to emergency closed-chest mechanical pacing. This is again in contrast to similarly studied ether-anesthetized dogs that can be reliably cooled to 18"to 20°C and routinely stabilized at that temperature for periods in excess of 30 minutes. Reports from this laboratory have shown considerable differences in the PCOZ of dogs in the present [ 101and other [7] series when ether or halothane was given in 95% o&% COz rather than 100% 0 2 at temperatures below 30°C. Although the importance of this observation is not known, it is apparent that the addition of 5% CO, will appreciably shorten the QTc under hypothermic conditions. When

394

T H E ANNALS OF THORACIC SURGERY

ECG Changes during Hypothermia used with ether anesthesia, 5% COZ causes no other ECG change except to prolong the return to a normal sinus rhythm during rewarming, but morbidity and mortality increase over that achieved with 100% O2 if circulatory arrest is extended to 60 minutes [7]. When used with halothane anesthesia, 5% COZ additionally prolongs R-R intervals at levels of deep hypothermia; however, the initial rewarming phase is shortened and the incidence of postoperative neurological disorders decreases from 100% to 25% following 30 minutes of circulatory arrest. Interestingly, the 1 dog in this series receiving halothane in 95% 0 2 1 5 % C 0 2 that fibrillated prior to elective arrest survived the subsequent 30-minute arrest period, was rewarmed, and had no detectable postoperative disturbance. Since it was known that halothane in 100%O2 resulted in difficulties during surface rewarming, perfusion rewarming was tried in the current series. Even so, all animals developed high-stepping gait postoperatively following 30 minutes of arrest. On the other hand, with halothane in 95% 0 2 / 5 % COz and perfusion rewarming, no animal developed postoperative disturbance following 30 minutes of arrest. These results could not be correlated with any ECG difference other than a quicker return to a normal sinus rhythm after arrest in perfusionrewarmed dogs receiving 95% 0 2 / 5 % COz. J waves have been reported with induced and accidental hypothermia. Some early reports stated that their appearance was a warning of impending ventricular fibrillation [2, 91 although this is not always the case [4]. Recently it has been demonstrated that J-type phenomena may result from a variety of clinical and experimental “stress response” situations, all of which are related to increased sympathetic activity and endogenous catecholamine release [3]. The onset of J waves at highly variable temperatures with halothane anesthesia seen in this study supports the concept that such phenomena represent generalized sympathetic adrenergic activity with stress rather than a specific response to cold. Under deep ether anesthesia, J waves were seen only at temperatures of 20°C or less, which corresponds to the point at which circulating norepinephrine levels, although still low, begin to rise [81. Since anesthetic agents and arterial Pcoz and their interactions all have been shown to influence sympathetic adrenergic activity, it seems unwarranted to attribute specificity to J and related S-T changes at this time. Delayed return to a normal sinus rhythm and prolongation of all the measured intervals during the initial rewarming phase as circulatory arrest was increased from 0 to 60 minutes in the dogs given ether in 100% O2can probably be related to anoxic changes. The effects of properly managed resuscitation manual massage and defibrillation - appear to be surprisingly benign, as shown by all dogs in the 30-minute arrest group returning to a normal sinus rhythm before 20°C during rewarming. Furthermore, insofar as the QTc reflects the integrity of repolarization metabolism, circulatory arrest for periods up to 60 minutes causes only minimal metabolic derangement with ether anesthesia. We conclude that the electrocardiogram is most useful as a clinical monitor during hypothermia and that studies of other factors are probably more important in increasing present knowledge of the physiology of hypothermia. In the

VOL.. 19, NO. 4, APRIL, 1975

395

SANDS ET AL.

latter realm the interaction of cold and various anesthetic agents is of great importance, and enough is now known to provide some rationale for the selection of an anesthetic agent. Thus, as the results of this and cited studies demonstrate, ether given in 100% O2is the agent of choice in infants for surface hypothermia to deep levels with circulatory arrest for periods of up to one hour. Halothane is unsuitable for surface hypothermia at deep levels but may be used in a procedure combining surface cooling and perfusion rewarming if given in a mixture of oxygen and carbon dioxide.

References 1. Barratt-Boyes, B. G., Simpson, M., and Neutze, J. M. Intracardiac surgery in neonates and infants using deep hypothermia with surface cooling and limited cardiopulmonary bypass. Circulation 43 (Suppl I): 1, 1971. 2. Boba, A. An abnormal electrocardiographic pattern and its relation to ventricular fibrillation: Observations during clinical and experimental hypothermia. Am Heart J 57:255, 1959. 3. DeSweit,J. Changes simulating hypothermia in the electrocardiogram in subarachnoid hemorrhage. J Electrocardiol 5: 193, 1972. 4. Emslie-Smith, D., Sladden, G. E., and Stirling, G. R. The significance of changes in the electrocardiogram in hypothermia. Br Heart J 21:343, 1959. 5. Mohri, H., Dillard, D. H., Crawford, E. W., Martin, W. E., and Merendino, K. A. Method of surface-induced deep hypothermia for open heart surgery in infancy. J Thorac Cardiovasc Surg 58:262, 1969. 6. Mohri, H., Dillard, D. H., and Merendino, K. A. Hypothermia: Halothane anesthesia and the safe period of total circulatory arrest. Surgery 72:345, 1972. 7. Mohri, H., Hessel, E. A., 11, Nelson, R. J., Matano, I., Anderson, H. N., Dillard, D. H., and Merendino, K. A. Use of Rheomacrodex and hyperventilation in prolonged circulatory arrest under deep hypothermia induced by surface cooling. A m J Surg 112:241, 1966. 8. Mohri, H., Pitts, C. L., Sands, M. P., Manhas, D. R., Dillard, D. H., and Merendino, K. A. Effects of surface-induced hypothermia and circulatory occlusion on plasma catecholamines. Surgery 72:596, 1972. 9. Osborn, J. J. Experimental hypothermia: Respiratory and blood pH changes in * relation to cardiac function. Am J Physiol 175:389, 1953. '10. Sato, S., Vanini, V., Mohri, H., and Merendino, K. A. A comparative study of the effects of carbon dioxide and perfusion rewarming on limited circulatory occlusion during surface hypothermia, under halothane and ether anesthesia. A n n Surg 180:192, 1974.

396

T H E ANNALS OF THORACIC SURGERY