Cerebral perfusion and metabolism during profound hypothermia in children

J

THORAC CARDIOVASC SURG

1991;102:103-14

Cerebral perfusion and metabolism during profound hypothermia in children A study of middle cerebral artery ultrasonic variables and cerebral extraction of oxygen Flow velocity of the right middle cerebral artery was studied in eight children during cardiac operations performed with profound hypothermia. Cerebral oxygen consumption was estimated by relating the difference in oxygen content between arterial and venous blood (jugular bulb) to flow velocity. In another six children, also during profound hypothermic procedures, the diameter of the middle cerebral artery was studied with an electronic echo-tracking instrument connected to a real-time ultrasound scanner. Flow velocity and estimated oxygen consumption decreased during cooling in proportion to the temperature decrease (r = 0.67, p < 0.001, and r = 0.86, p < 0.001, respectively), whereas the diameter was unaffected by temperature. At a nasopharyngeal temperature of 16.9° ± 1.9° C flow velocity was reduced to 33.1 % ± 7.0% of the value obtained at 35° C after induction of anesthesia. Correspondingly, the oxygen consumption decreased to 20.1 % ± 6.4%. The increase in oxygen consumption per 10° C change in temperature was 3.6 (2.0 to 3.9) during surface cooling, 2.6 (1.9 to 2.7) during cardiopulmonary bypass cooling, and 2.7 (1.5 to 4.6) during rewarming. Flow velocity was not influenced by perfusion pressure during profound hypothermia within the range of 20 to 42 mm Hg (r = 0.14, .p = 0.52) but was related to pump flow (r = 0.73, p < 0.001). A pump flow down to 0.5 Ljminjm2 was found to be adequate during stable profound hypothermia, as judged from the maintained high jugular bulb venous oxygen saturation (70 % to 80 %). It is concluded that flow velocity is reduced at hypothermia in proportion to the reduced metabolic rate, although modified by other factors that influence cerebral blood flow.

J. van der Linden, MD, PhD,a R. Priddy, FFARACS,a R. Ekroth, MD, PhD,b C. Lincoln, FRCS,c W. Pugsley, FRCS,c M. Scallan, FFARCS,d and H. Tyden, MD, PhD,a

Uppsala, Sweden, and London, England

Operations with profound hypothermia in small children, with or without an arrest period, can be followed by cerebral dysfunction. The dysfunction is not always obviFrom the Departments of Anaesthesia" and Cardiothoracic Surgery," University Hospital, Uppsala, Sweden, and the Departments of Cardiothoracic Surgery" and Anaesthesia," Brompton Hospital, London, England. Supported by grants from the Gillberg Foundation, the Swedish 1987 Foundation for Stroke Research, the Swedish Heart and Lung Foundation, the Swedish Society of Medicine, and the University of Uppsala. Received for publication Jan. 15, 1990. Accepted for publication July 25, 1990. Address for reprints: C. Lincoln, Brompton Hospital, Fulham Rd., London SW3, England.

12/1/24598

ous on simple clinical examination but is detected by cerebral injury markers' and long-term neuropsychometric follow-up.l Its occurrence may suggest imperfections in the current cardiopulmonary bypass techniques. These may include inadequate perfusion and oxygenation of the brain. Little information is available regarding cerebral blood flow or metabolic requirements during profound hypothermic procedures in small children.e 4 Cerebral blood flow decreases with hypothermia, but the relationship to cerebral metabolic rate is unknown in this clinical setting. In the present study cerebral blood flow has been observed in small children undergoing profound hypothermic cardiac procedures by recording flowvelocity in the right middle cerebral artery continuously during all phases of the operations. This artery carries 80% of the 103

10 4

The Journal of Thoracic and Cardiovascular

van der Linden et al.

Surgery

Table I. Basic clinical variables of the patients included in the middle cerebral artery flow velocity (AJ-A8) and diameter (BJ-B6) studies

Patient

Al A2 A3 A4 AS A6 A7 A8 B1 B2 B3 B4 B5 B6

Age

Weight (kg)

BSA (m 2)

CPB (min)

1 yr 3 rno 6 rno 10rno 3 yr 6 rno 9rno 11 rno 1 day 6rno I yr 2 rno 5 rno 1 yr 2rno 1 yr 1 rno 1 yr

8.0 7.0 7.9 12.8 6.7 7.7 3.3 5.3 6.8 5.1 8.2 3.1 10.1 15.0

0.96 0.34 0.39 0.50 0.35 0.37 0.20 0.30 0.35 0.38 0.42 0.20 0.46 0.54

97 144 86 126 48 60 103 78 99 122 102 53 66 39

Diagnosis

Surface cooling

TGA AVe Fallot VSD VSD Fallot TGA VSD Fallot AVe VSD VSD VSD Fallot

No No Yes No Yes Yes Yes Yes No No No Yes No No

Nasopharyngeal temperature (lowest) C)

r

17.8 20.3 15.6 23.1 14.0 13.1 15.5 13.8 19.9 17.0 15.2 15.7 19.8 21.0

BSA, Body surface area; CPB, cardiopulmonary bypass; TGA, transposition ofthe great arteries; AYC, atrioventricular canal; YSD, ventricular septal defect.

flow to the hemisphere.' The potential of the noninvasive Doppler technique to estimate volume flowis based on the assumption that the diameter of intracranial arteries remains unchanged during an alteration in volume flow. To test the validity of this assumption, we studied the diameter of the middle cerebral artery with an electronic echo-tracking instrument connected to a real-time ultrasound scanner with a linear array 3.5 MHz transducer. With this technique diameter changes as small as 14 JLm can be detected." The cerebral metabolic rate was estimated by relating middle cerebral artery flow velocity to the difference in oxygen content between arterial and internal jugular bulb blood. The purpose of the study was to evaluate the relationship between cerebral blood flow and metabolic rate during all phases of profound hypothermic procedures, including those when pump flow was reduced to enhance surgical exposure. Patients and methods Patients. Eight consecutive children who underwent total correction of congenitalheart defects were includedin the Doppler studies, and another six consecutive children were included in the arterial diameter studies (Table I). The study protocol conformed to the rules of the Helsinki declaration and was approved by the Ethics Committee, Brompton Hospital. Informed consent was obtained from the parents of the patients participating in the study. Clinical management. Premedication consisted of chloral hydrate, 60 mg/kg body weight,givenorally, and atropine 0.02 mg/kg body weight given intramuscularly, I hour before the operation. Anesthesiawasinducedwith halothane, 1%to 3%(in one patient with propofol); nitrous oxide,70% to 50%;and fen-

tanyl, 5 to 15 ~g/kg body weight.Tracheal intubation was aided with pancuronium bromide, 0.2 rug/kg. After inductionof anesthesianitrousoxidewaswithdrawnand was not usedagain. Halothane in an air-oxygenmix was used to maintain anesthesia beforeand after cardiopulmonarybypassat 0.75%to 1.25% and 0.5%,respectively. During cardiopulmonary bypassisoflurane was used at 0% to 0.25%during core coolingand profound hypothermia and at 0.25%to 0.5%during core rewarming. Up to 3 ~g/kg body weight of fentanyl was given beforeclosureof the sternum. Three 20-guagecannulas wereplacedpercutaneouslyintothe internal jugular vein, two for routine clinical use and one for bloodsampling.The latter was inserted into the proximalinternaljugular veinwith its tip in the jugular bulb. A 22-gaugecannula was placed percutaneouslyinto the femoral artery for continuous monitoring of systemic arterial pressure and for blood gas analysis. Nasopharyngeal temperature (and arterial and venous temperatures during bypass) was recorded continuously. Hypothermia was attained either by surface cooling with ice to 26° C followed by further bypasscooling(six patients) or by bypass cooling only (eight patients). Acid-base management was in accordance with the alpha-stat pH principle? with blood gas levels measured with a Corning 178 pH/blood gas analyzer (Medfield, Mass.). After surgical exposure of the heart by a median sternotomy, cardiopulmonary bypasswas instituted by means of an ascendingaortic cannula and one or two cavalcannulas for venousdrainage. The priming solutionwas a mixture of whole blood and Hartmann's solution (compound sodium lactate) in a ratio calculated to achieve a hematocrit level of 25%. Masterflo infant and pediatric hollow fiber oxygenators (Dideco, Mirandola, Italy) were applied. During deep hypothermia extracorporeal perfusionwas maintained at 50% to 25% ofthe calculated flow (2.4 L X m2) for normothermia.After the lowflow period perfusionwas restored to 100%during rewarming.The arterial bloodtemperature did not exceed 39.5° C during rewarming,and the differencebetweennasopharyngeal and arterial temperature never exceeded 10° C. At 37° C in the

Volume 102

Profound hypothermia in children

Number 1 July 1991

10 5

100

75

MCAv X

50

25

r--0.97, p<0.0001

n-26 (5 patients)

0+---------+--------+---------+----------1 25 30 20 35 40

NASOPH.TEMP.Oegr-.C

Fig. 1. Middle cerebral artery flow velocity (MCAv) in percent of the anesthetized level as a function of nasopharyngeal temperature during surface cooling. Combined data from five patients (A3 and A5 to A8).

nasopharynx the child was weaned off extracorporeal circulation. Middle cerebral artery flow velocity measurements. A 2 MHz pulsed TCD-2-64B Doppler ultrasound velocimeter (EME, Uberlingen, Federal Republic of Germany) was used to obtain continuous signals from the middle cerebral artery. The ultrasound transducer was placed on the right side of the head just above the zygomatic arch, the thinnest part of the skull. The orientation and depth of the Doppler signals were adjusted to obtain the best possiblesignal strength. The position of the probe was fixed in two planes with a holder (EME, Uberlingen, Federal Republic of Germany). The holder was then meticulously fastened with plaster, and care was taken not to influence the position throughout the study, since a dislodgment of the probe influences the recording. The maximum flow velocity corresponds to the blood flow velocity in the center of the middle cerebral artery. Maximum flow velocity was followed continuously with minimum angle between probe and vessel. This allows almost no deviation from recording true maximum velocity, because the recorded velocity is proportional to the true maximum velocity multiplied by the cosine of the angle of incidence of the ultrasound beam. In this study an angle of incidence of 0 degrees was assumed. If the angle varies between 0 and I 5 degrees, the cosine remains greater than 0.965. Within this range any error in velocity measurements remains less than 3.5%. The instrument and exam-

ination procedure, including the justification of using maximum values instead of average values, has been discussed previously.f The time mean values from 8-second periods of maximum flow velocity were used for data analysis. Blood flow velocity is a function of vessel diameter and blood flow.? Because the diameter of the middle cerebral artery varies with age and between patients, the interindividual variation was normalized by expressing flow velocity as a percent of each individual's reference value. Thus comparisons between patients with different vessel diameters were made possible. In the present study the reference value was the value recorded after the induction of anesthesia. Admittedly, the value in the awake patient before the induction of anesthesia may seem more appropriate, but the former reference value simplified comparisons with metabolic data. For ethical reasons these could not be obtained during the awake state. Middle cerebral artery diameter measurements. The middle cerebral artery diameter was monitored with a Diamove echo-tracking unit (Teltec, Lund, Sweden) that includes a realtime ultrasound scanner with a linear array 3.5 MHz transducer. The system uses phase-locked-loop circuits to follow the arterial wall movements, as originally designed by Hokanson and colleagues. 10 Advantages of the phase locked echo-tracking principle are independence of the echo amplitude and high sensitivity for spatial movements. A phase-locked loop restores the position of an electronic gate relative to the moving echo, and the

10 6

The Journal of Thoracic and Cardiovascular

van der Linden et al.

Surgery

Table II. Changes in middle cerebral artery flow velocity and clinical variables during the study

Awake Anesthetized Before ice* 15 min of ice* 30 min of ice* After ice" 15 min after ice* Before CPB 5 min ofCPB 15 min ofCPB 30 min ofCPB 45 min ofCPB Before rewarming 15 min of rewarming 30 min of rewarming Beforeend of CPB After CPB 15 min after CPB 30 min after CPB After operation

MCAv

NT

MAP

CPP

(%)

(" C)

(mmHg)

(mmHg)

144 ± 100 ± 100 ± 57.7 ± 46.9 ± 39.8 ± 5\.2 ± 84.5 ± 663 ± 45.1 ± 33.1 ± 41.0 ± 39.9 ± 65.8± 132 ± 107 ± 98.2 ± 116 ± 110 ± 120 ±

34 0 0 5.7 5.8 4.8 \.8 19 17 13 7.0 8.7 9.5 14 35 24 6.6 14 16 34

35.3 ± 353 ± 32.4 ± 28.7 ± 26.7 ± 25.6 ± 26.4 ± 16.9 ± 16.2 ± 16.9 ± 17.9 ± 17.0 ± 303 ± 36.8 ± 36.5 ± 35.8 ± 34.5 ± 33.8 ± 33.8 ±

0.4 0.4 0.8 0.7 0.4 0.8 1.7 3.6 \.2 \.9 2.2 \.6 \.5 0.5 03 0.6 0.6 0.7 0.7

66.8 ± 68.0 ± 49.8 ± 48.4 ± 46.8 ± 4\.4 ± 44.6 ± 33.0 ± 27.4 ± 26.1 ± 33.2 ± 33.5 ± 49.8 ± 48.6 ± 523 ± 45.8 ± 51.1 ± 54.1 ± 54.4 ±

4.6 4.4 6.9 5.7 4.8 4.6 4.8 4.5 3.1 3.7 8.6 6.6 53 8.6 5.7 3.7 2.9 3.9 2.5

57.7 ± 57.3 ± 38.7 ± 39.5 ± 36.8 ± 33.4 ± 36.4 ± 30.6 ± 273 ± 26.3 ± 30.2 ± 3\.8 ± 48.1 ± 46.3 ± 47.9 ± 35.6 ± 41.3 ± 453 ± 453 ±

5.5 4.8 7.9 6.7 5.2 4.9 4.6 4.6 3.5 3.1 9.2 6.6 3.9 12 6.5 3.8 2.8 4.1 2.5

139 ± 0.98 ± 0.77 ± 0.92 ± 0.83 ± 2.30 ± 234 ± 2.28 ±

0.29 0.22 0.10 0.14 0.13 0.19 0.14 0.06

Mean ± SEM of eight patients (A I,AS); jugular venous blood samples for gas analyses were collected in six patients (A3-AS). MCAv. Flow velocity in the middle cerebral artery expressed in percent of the anesthetized values; NT, nasopharyngeal temperature; MAP, mean arterial pressure; CPP, cerebral perfusion pressure; CI, pump flow; Paco-, arterial carbon dioxide pressure; Asat, arterial oxygen saturation; Vjsat.jugular venous oxygen saturation; Paoj, arterial oxygen pressure; PVj02, jugular venous oxygen pressure; HCT, hematocrit level; D-CMR02, Doppler estimated cerebral oxygen consumption in percent of the anesthetized values; CPB, cardiopulmonary bypass. 'Five patients who were surface cooled to a nasopharyngeal temperature of Z5° C.

compensatory movement of the gate yields the movement of the echo. A dual echo-tracking loop includes the ability to track two separate echoes from opposite vessel walls simultaneously, the differential signals thereby representing the instantaneous change of vessel diameter. The smallest detectable movement is 14 ~m.6 The proximal segments of the middle cerebral arteries were imaged simultaneously and bilaterally close to the base of the skull through the anterior fontanelle. The probe was positioned to allow a frontal image of the vessels at an angle close to 90 degrees between the longitudinal axis of the vessels and the ultrasonic beams. Only paired parallel echoes, one moving caudally and the other moving cranially simultaneously with each pulse wave, were identified as opposite walls of the middle cerebral artery. The measurements were restricted to the artery that gave the best image. Time mean values from 8-second periods of diameter recordings were used for data analysis. Cerebral oxygen uptake. Oxygen content was derived from the following formula: Saturation X 1.34 X Hemoglobin

+ 0.003 X Po,

where oxygen content is in milliliters per 100 ml, saturation is a decimal, hemoglobin is in grams per 100 ml, and oxygen tension (Po-) is in millimeters of mercury. The arteriovenous oxygen content difference was determined for each patient and then multiplied with the middle cerebral artery blood flowvelocity to calculate the estimated cerebral oxygen uptake. Each patient's values are thus expressed in percent of the levels obtained after induction of anesthesia. The increase of metabolic rate per 10° C change in temperature (QIO) was calculated for each patient

separately, whereby the QIO represents the slope of the regression line. The data gathered before bypass cooling, during bypass cooling, and during rewarming were used in separate regression analyses yielding separate QIO values for the three periods. With a decrease in metabolic rate of 50% and a temperature reduction of 10° C, the QIO will be 2. Study protocol. The awake values of the middle cerebral artery flow velocity were recorded between 30 and 60 minutes after premedication, that is, when the child was calm enough to accept the examination. Flow velocity, mean arterial blood pressure, central venous pressure, pump flow,and temperatures were followed continuously. The measurements were started while the patient was still awake and continued throughout the procedure until 5 minutes after skin closure. Although these measurements were monitored continuously, the results are presented for practical reasons as a representative value of defined periods. Arterial and venous (jugular bulb) blood samples were obtained at fixed intervals: (1) 15 to 30 minutes after induction of anesthesia, when all lines had been inserted; (2) before surface cooling; (3) 15, 30, and 45 minutes after the start of surface cooling; (4) after removal of ice; (5) 15, 30, and 45 minutes later; (6) before cardiopulmonary bypass (7) after 5, 15,30, and 45 minutes of cardiopulmonary bypass; (8) before and after 15 and 30 minutes of rewarming; (9) before termination of cardiopulmonary bypass; (l0) 5, 15, 30, 45, and 60 minutes after cardiopulmonary bypass; and (11) within 5 minutes after completing the operation. Diameter recordings were performed in 8-second sequences one to three times every 10 minutes during the whole procedure.

Volume 102

Profound hypothermia in children 1 0 7

Number 1 July 1991

PaC02

Asat

Vjsat

Pa02

(mmHg)

(%)

(%)

(mm Hg)

32.8 ± 32.4 ± 28.6 ± 34.7 ± 32.5 ± 35.1 ± 38.1 ± 50.5 ± 40.7 ± 35.5 ± 34.1 ± 35.4 ± 27.6 ± 32.1 ± 30.0 ± 35.6 ± 38.8 ± 38.2 ± 40.3 ±

2.8 2.9 2.2 1.7 2.8 5.6 4.4 4.5 2.5 3.0 3.1 3.1 3.4 2.0 2.7 3.3 3.6 1.3 3.9

95.5 ± 96.3 ± 96.6 ± 93.6 ± 96.4 ± 98.0 ± 94.2 ± 99.6 ± 99.2 ± 99.0 ± 98.6 ± 99.3 ± 99.1 ± 99.4 ± 99.1 ± 91.2 ± 93.4 ± 93.6 ± 91.8 ±

1.2 1.2 1.4 5.2 3.4 1.1

2.8 0.2 0.2 0.1 0.8 0.2 0.6 0.2 0.5 5.0 4.0 3.9 4.3

67.3 ± 8.9 69.3 ± 8.8 55.8 ± 7.9 58.0 ± 4.9 56.8 ± 4.9 67.2 ± 6.5 73.7 ± 5.9 95.5 ± 1.8 79.7 ± 3.6 71.6 ± 6.7 72.3 ± 3.9 75.0 ± 2.2 55.8 ± 3.0 50.5 ± 13 47.0 ± 8.7 45.7 ± 6.9 63.0 ± 4.7 59.5 ± 5.5 67.3±11

154 ± 169 ± 204 ± 212 ± 358 ± 249 ± 222 ± 345 ± 230 ± 197 ± 169 ± 202 ± 219 ± 222 ± 216 ± 215 ± 214 ± 212 ± 180 ±

Statistical analysis. A commercially available, computerized statistical RS/ I package (BBN Corp., Cambridge, Mass.) was used for comparisons between samples. The results are expressed as mean ± standard error of the mean (SEM). The Wilk-Shapiro test was used to determine whether the data were normally distributed. Nonparametric tests!' were used when data were not normally distributed or when samples were small (n < 12). When the distribution was normal, standard parametric tests were applied. Differences were regarded as significant if the p value was less than 0.05. For the regression analyses to explain the variation in flow velocity, each patient was studied separately to obtain individual correlation coefficients. Spearman rank regression analysis was made for one independent variable. I I The means of partial correlation coefficients were tested versus 0; p < 0.05 was considered significant. Regression analyses of the combined data from all the patients were also performed. The role of each separate variable was tested for each patient with multiple stepwise regression analysis according to Draper and Smith,12 in which a linear regression analysis with more than one independent variable can be performed. The RS/l computer-assisted program was used to achieve the best fit of the cerebral oxygen consumption data during surface cooling to the van't Hoff equation. 13

Results Clinical course. All 14 children recovered well from the operation and were discharged from the hospital within 2 weeks from the operation. Most (13/14) had no seizures or signs of choreoathetosis. However, 1 child (patient A2, Table I) had a seizure on postoperative day 3 and signs of increased intracranial pressure during the first days after the operation.

PvP2

(mmHg)

62 62 98 92 115 82 66 30 26 15 27 16 33 17 31 65 53 53 59

39.8 ± 40.8 ± 31.0 ± 33.5 ± 32.6 ± 42.4 ± 49.0 ± 149 ± 56.6 ± 44.7 ± 43.3 ± 47.9 ± 29.1 ± 28.8 ± 26.2 ± 27.2 ± 36.6 ± 34.7 ± 35.4 ±

5.8 5.6 3.8 3.0 2.4 4.9 7.1 27 3.7 4.8 3.9 4.0 1.4 5.5 3.9 2.5 2.7 4.1 5.5

HCT (%)

37.6 ± 37.6 ± 32.0 ± 33.2 ± 32.7 ± 32.2 ± 37.3 ± 23.3 ± 23.3 ± 23.3 ± 23.8 ± 24.4 ± 24.3 ± 24.6 ± 24.6 ± 26.0 ± 30.1 ± 31.5 ± 32.6 ±

3.1 3.1 0.4 2.2 2.0 2.2 2.8 1.3 0.9 1.0 1.0 0.9 0.8 0.3 0.3 1.5

1.7 1.8 1.3

D-CM R 02 (%)

100.0 ± 97.2 ± 98.3 ± 70.4 ± 68.4 ± 53.1 ± 50.3 ± 7.6 ± 17.0 ± 20.1 ± 26.6 ± 23.5 ± 64.3 ± 150 ± 143 ± 135 ± 104 ± 113 ± 76.7 ±

0 02.8 30 19 20 14 12 3.7 2.7 6.4 5.5 5.0 16 53 36 40 21 36 21

Middle cerebral artery flow velocity. Before anesthesia was induced flowvelocity was 144% ± 34% of the subsequent anesthetized value (100%). With surface cooling to 26.7° ± 0.4° C it was further reduced to 39.8% ± 4.8%of the anesthetized (normothermic) value. Fifteen minutes after removal of the ice, flowvelocity had increased to 51.2% ± 1.8%. This increase occurred despite a continuing decrease in temperature. With the start of bypass cooling, flowvelocitydecreased again, and at 16.9° ± 1.9° C it was 33.1% ± 7.0%. With rewarming, flow velocity increased to 107% ± 24%. Flow velocity changes and clinical variables during the whole procedure are shown in Table II.

Correlation with temperature and other clinical variables. Flow velocity declined in proportion to decreasing temperature. This was particularly evident during surface cooling (Fig. 1, Table 11). The mean correlation coefficient for nasopharyngeal temperature versus flow velocity for the whole study period was 0.67 ± 0.06 (p < 0.01). In five of the patients regression analyses indicated that arterial carbon dioxide tension and/or cerebral perfusion pressure (mean arterial pressure minus central venous pressure) were positively related to flow velocityand improved the model, explaining the variation of flow velocity (Table III). During bypass the mean correlation coefficient for nasopharyngeal temperature and flow velocity was 0.59 ± 0.09,p < 0.01). A correlation was found between pump flow and flow velocity (0.80 ± 0.04, P < 0.01).

The Journal of

10 8

van der Linden et al.

Thoracic and Cardiovascular Surgery

Table III. Multiple regression analysis offactors influencing middle cerebral artery flow velocity for each separate patient and the function explaining individual variations in middle cerebral artery flow velocity (pump flow not included) Patient

n

r

p

Function

Al A2 A3 A4 AS A6 A7 A8

15 10 19 17 15 18 23 15

0.92 0.82 0.94 0.79 0.87 0.86 0.85 0.91

<0.001 <0.001 <0.0001 <0.001 <0.001 <0.0001 <0.0001 <0.001

2.6 X NT - 38 2.6 X CVP + 0.8 X Paco- - 0.7 X HCT + 18 2.4 X NT + 1.9 X Pacoz - 97 4.7 X NT - 82 1.1 X NT + 1.1 X Paco- + 0.7 X CPP - 76 1.3 X NT - 11 0.6 X MAP + 0.5 X Paco- - 20 1.7 X NT + 0.5 X Paco-- 37

NT. Nasopharyngeal temperature; Cv l', central venous pressure; PaC02. arterial carbon dioxide pressure; HCT, hematocrit level; Cf'P, cerebral perfusion pressure (mean arterial pressure minus central venous pressure); MAP, mean arterial pressure.

Table IV. Spearman rank correlations between middle cerebral artery flow velocity and other variables during the whole study, from the anesthetized state until after the operation Variable

Mean correlation Irs ± SEM)

Nasopharyngeal temperature Central venous pressure Mean arterial pressure Cerebral perfusion pressure Hematocrit level Arterial base excess Arterial carbon dioxide pressure Arterial saturation Arterial pH Arterial oxygen pressure Mean ± SEM of eight patients

(AJ-A~).

Table V. Spearman rank correlations between middle cerebral artery flow velocity and other variables during cardiopulmonary bypass Variable p

0.67 ± 0.06

<0.01

0.55 ± 0.09 0.55 ± 0.11 0.50.± 0.10

<0.01 <0.05 <0.05

0.30 ± 0.12 0.17 ± 0.16 0.14 ± 0.04

NS NS NS

0.07 ± 0.13 -0.04 ± 0.13 -0.11 ± 0.13

NS NS NS

Mean correlation Irs ± SEMi

Cardiac index Nasopharyngeal temperature Cerebral perfusion pressure Central venous pressure Hematocrit level Arterial base excess Arterial saturation Arterial pH Arterial carbon dioxide pressure Arterial oxygen pressure Mean ± SEM of eight patients (A

0.80 ± 0.04 0.59 ± 0.09

<0.01 <0.01

0.38 ± 0.11

<0.05

0.34 0.27 0.18 0.13 0.05 -0.03

± ± ± ± ± ±

0.12 0.20 0.18 0.16 0.18 0.17

0.01 ±0.15 J-A~).

p

<0.05 NS NS NS NS NS NS

NS, Not significant.

NS. Not significant.

However, lower pump flows were applied at lower temperatures, and to clarify the effect of pump flowper se, the relationship between pump flow and flow velocity was analyzed separately for the period of stable profound hypothermia (nasopharyngeal temperature <17° C). A significant correlation could still be detected (r = 0.73, P < 0.001, Fig. 2 B). During this period no correlation between mean arterial pressure (range 20 to 42 mm Hg) and flow velocity (r = 0.14, P = 0.52, Fig. 2, A) existed. Further details are given in Tables II through V. Middle cerebral artery diameter. Hypothermia did not influence the diameter of the middle cerebral artery. Thus no correlation between nasopharyngeal temperature and diameter existed (the mean Spearman rank correlation coefficient of all children was 0.06 ± 0.13, Table

VI). Furthermore, the mean diameter at a temperature more than 26° C was 2.15 ± 0.13 and 2.11 ± 0.12 mm at a temperature less than 26° C. A plot of a representative patient is shown in Fig. 3. Estimated cerebral oxygen consumption. Estimated cerebral oxygen uptake decreased during cooling. In contrast to middle cerebral artery flow velocity, there was no rebounding after the ice had been taken away. The discrepancy between middle cerebral artery flowvelocity and estimated cerebral oxygen consumption corresponded to a changing oxygen extraction during this period. Thus jugular venous oxygen saturation decreased during surface cooling and returned to normal levels after the ice was taken away. At a nasopharyngeal temperature of 25.6° ± 0.8° C (10° C lowerthan the first mea-

Volume 102

Profound hypothermia in children

Number 1 July 1991

1 09

80

80

•

o 60

DO

"'CAv 40

40

1lI

0

o

•

60

o

0

0 0

20

o

0 0

r::P

20

Q:]O

r-0.14, p-0.52

0.1-+----+--+----+--+----.4 40 45 30 35 20 25

r-0.73. p
o+---I---+----+--~-----+----+­

0.2

0.4

0.6

0.8

1.0

1.2

1.4

PUMP FLOW 1/m2/m1n

"'AP mmHg

Fig. 2. Left, Middle cerebral artery flow velocity (MCAv) in percent of the anesthetized level during profound hypothermia « 17° C) as a function of mean arterial pressure. Combined data of five patients (A3 and AS to A8). Right, Middle cerebral artery flow velocity (MCAv) in percent of the anesthetized level during profound hypothermia «17° C) as a function of pump flow rate. Combined data from five patients (A3 and AS to A8).

surements) the cerebral oxygen consumption was 53.1% ± 14%. During profound hypothermia (16.2° ± 0 1.20 C) the corresponding value was 17.0% ± 2.7%. With rewarming the estimated cerebral oxygen consumption returned to the normothermic level before bypass (Table II, Fig. 4). During the rewarming phase, jugular venous oxygen saturation decreased consistently and the extraction of oxygen hence increased. Fifteen minutes after bypass, when rewarming had been completed, normal jugular venous saturations were resumed. Further data are given in Table II. Correlation with temperature and other variables. Regression analyses indicated that temperature was the only measurement that could explain the variation in estimated cerebral oxygen consumption in each individual. A model, based on van't Hoff's law,':' provided the best fit to the combined data from all six patients (A3 to A8) from whom jugular venous blood samples were collected (r = 0.86, p < 0.001, Fig. 5).

Table VI. Spearman rank correlation coefficients and critical levels of significance (p < 0.05) of the individual diameters when tested versus nasopharyngeal temperature Patient

n

r,

p = 0.05

BI B2 B3 B4 B5 B6 Mean ± SEM

34 19 16 18 34 23 24 ± 3.3

-0.22 +OAI +0.03 +OA2 +0.04 -0.33 +0.06 ± 0.13

±0.34 ±OA6 ±0.51 ±OA8 ±0.34 ±OA2 ±OA4 ± 0.03

The QIO was calculated by regression analysis for each patient. During prebypass cooling the QIO median was 3.6 (2.0 to 3.9), during bypass cooling it was 2.6 (1.9 to 2.7), and during bypass rewarming it was 2.7 (1.5 to 4.6).

1 10

The Journal of Thoracic and Cardiovascular Surgery

van der Linden et al.

2.5

o

o 2

---------~--------rr--------------------o----------~---DO

ODD

0

0

o

o

0

DO

~-----g-------------------------I;J-----I;J---e [D

0

0 0

0-

8

1.5

mm 1

r--0.09,p-0.62

0.5

0+----+----+---+----+-----+---1----+-...,.-----+----+-----1 18 28 30 20 22 26 32 34 36 38 NASOPH.TEMP.Oegr-.C

Fig. 3. Nasopharyngeai temperature versus mean diameter of the middle cerebral artery (patient 85).

Discussion In the present study noninvasive methods were used to evaluate cerebral blood flow during profound hypothermic procedures in small children. The flow data were related to arteriojugular differences in oxygen content. The first major finding was constancy of the middle cerebral artery's mean diameter during the extreme temperature variations observed in this study. This observation supports the validity of flow velocity measurements to estimate changes in cerebral blood flow. The second major finding was the close match between the flow parameter and the calculated cerebral metabolic rate during profound hypothermia with periods of markedly reduced pump flow. Middle cerebral artery flow velocity versus volume flow. Flow velocity measurements of intracranial arteries may provide information of cerebral blood flow.Although absolute values are not obtained, the direction and magnitude of changes can be determined. The validity of this approach to estimate cerebral blood flow depends on whether the cross-sectional area of the vessel changes. Previous transcranial Doppler studies have assumed no or small changes of the middle cerebral artery cross-section-

al area, but actual measurements of the area 0t:the diameter have not been performed. In the present study we used an echo-tracking instrument that detects diameter changes as small as 14 }.Lm. 6 The results indicate that the middle cerebral artery, which carries approximately 80% of the hemispheric blood supply,' does not change its mean diameter during the conditions of the study. This supports the assumption that changes in flow velocity of the middle cerebral artery reflect changes in blood flow. Other workers have compared middle cerebral artery flow velocity with xenon-estimated cerebral blood flow during carbon dioxide-reactivity tests at normothermia 14 and with electromagnetically measured flow in the internal carotid artery during normothermia and moderate hypothermia. IS Bishop and colleagues 14 found a correlation coefficient of 0.85 between xenon washout measurements and changes in middle cerebral artery flow velocity, whereas Lindegaard and colleagues'> observed a correlation coefficient of 0.95 between changes in middle cerebral artery flow velocity and electromagnetically measured flow in the ipsilateral internal carotid artery. As far as we know, there are no published data on the validity of middle cerebral artery flow velocity measure-

Volume 102

Profound hypothermia in children

Number 1 July 1991

iOO

iOO

111

~

4

t

80

80

0 Ie

>

C

60

M R 0 2

c u 2:

X

40

60

40

20

20

O+---f-----1I---+---+--+----l iO 20 25 30 35 40 i5

0+---+--+---+--1----+----1 35 40 30 i5 20 25 10

NASOPH.TEMP.Degr.C

NASOPH.TEMP.Degr.C

Fig. 4. Middle cerebral artery flow velocity (MCAv) and Doppler-estimated. cerebral oxygen consumption (D-CMR02) measurements (both in percent of the anesthetized levels) during cooling (squares with solid line) and rewarming (triangles with dotted line) in one individual (patient A6).

ments during profound hypothermia. We have just completed a parallel study in which adult patients with acquired heart disease, undergoing profound hypothermic procedures (20 C), have been observed.l" Correlation coefficientsbetween 0.7 and 0.8 were found between middle cerebral artery flow velocity and internal jugular venous blood flow, measured with a thermodilution technique. Together, both direct measurements of arterial diameter and comparisons with different volume flow techniques support the validity of intracranial flow velocity measurements to estimate changes in arterial volume flow during conditions of profound hypothermic extracorporeal circulation. Hypothermia and middle cerebral artery flow velocity. The present results indicate that middle cerebral artery flow velocity in children undergoing profound hypothermic procedures follows temperature proportionally, although modified by marked alterations in arterial carbon dioxide tension, hematocrit value, and pump flow. This is in agreement with experimental work in animals'? 18 and humans," 19 0

When middle cerebral artery flow velocity deviated from the pattern of a close match between temperature and flow velocity, the oxygen extraction changed, conceivably to maintain cerebral oxygen consumption at the rate set by temperature. The seemingly inappropriate cerebral flow during surface cooling remains to be explained but may be due to a changed distribution of cardiac output, as reported by Mavroudis and colleagues.P It has been debated whether cerebral blood flow during cardiopulmonary bypass is autoregulated and matched to the metabolic rate. Accumulating data indicate that cerebral blood flowautoregulation is maintained during bypass." provided the pH management is based on the alpha-stat principle." Present data are in agreement with this concept. In contrast, with pH-stat management autoregulation is lost, and because of the generally higher arterial carbon dioxide tensions inherent in the pH-stat technique, a state of cerebral hyperperfusion is observed.21-23 This is illustrated by the higher middle cerebral artery flow velocities

1 12

The Journal of Thoracic and Cardiovascular Surgery

van der Linden et al.

o

o

50

o D-CMR02

o

o

100

»>

~

25

•.•••..••••.••••..••.••••

:II:

qg

"'!:J:l DO

o£.1:P

DO

o 0 ~)iB~ffilo o ?ti [j:J 0 o .>:~ DO 0 »>: @ rn B B o eroOB 0 B DO 0

o

0

.9.:»:-0

q 10

o

.':9

0

o

o

Ir-0.86,

p
O. 1+------+-----+-------f-----t-...,...----+-----~ 10

15

20

25

30

35

40

NASOPH.TEMP.Degr.C

Fig. 5. Doppler-estimated cerebral oxygen consumption (D-CMR02) in percent of the anesthetized levels as a function of nasopharyngeal temperature. Combined data from six patients (A3 to A8, in whom jugular venous blood samples were collected) and the function with the best fit.!" Log cerebral oxygen consumption = 0.042· nasopharyngeal temperature + 0.53.

reported by Lundar and colleagues' with pH-stat technique during profound hypothermia in children. It has been suggested that, although autoregulation is maintained during bypass with moderate hypothermia, it is lost during profound hypothermia." This is at variance with the data of Tanaka and colleagues'? and also with the present results. Thus we found no correlation between perfusion pressure and middle cerebral artery flow velocity when nasopharyngeal temperature was less than 17° C. The difference between the present data and those of Greeley and colleagues" may be explained by the wider perfusion pressure range in their study, with arterial pressures as low as 8 mm Hg, which possibly is below the limit of autoregulation. Differences in anesthetic regimen and pump flow rate also may account for the difference in perfusion pressure/flow response.

Hypothermia and cerebral oxygen consumption. The relationship between temperature and oxygen consumption has been studied extensively in animal experiments. Data from humans are sparse, but it appears that total body oxygen consumption in humans is reduced

much in the same way as indicated by animal studies. Thus an approximate 50% reduction of total body oxygen consumption is observed when temperature is reduced by 10° C. According to animal experiments, a similar relationship exists between temperature and cerebral metabolic rate.l '' )7 Clinical data relating the conditions during profound hypothermic procedures have not been available, as far as we know. The present data suggest that cerebral metabolism in children undergoing profound hypothermic procedures closely follows the pattern demonstrated by total body measurements in humans and by cerebral studies in animals. According to Kirklin and Barratt-Boyes,' 3 a function based on van't Hoff's law, which relates the logarithm of a chemical reaction rate directly to temperature, gives the best fit to available data on the relationship between temperature and metabolic rate. This function could explain 74% (r = 0.86, r2 = 0.74,p < 0.001) of the variation in estimated cerebral metabolic rate of the combined data of all the patients with a QIO of 2.6. The corresponding values for individual patients (2.0 to 3.9) are

Volume 102 Number 1 July 1991

also in agreement with the data of Harris" with total body QlO during surface cooling of children in the range of 1.9 to 4.2. Furthermore, in an elucidating review on the hypothermic brain QIO ranging between 2.13 and 4.6 was reported from different studies in animals and in humans.P According to Michenfelder.P these differences may be clarified largely by a canine study, wherein the effect of hypothermia on both functioning brain (active electroencephalogram) and nonfunctioning brain (isoe1ectric brain) was exarnined.i" In the absence of a functioning brain the QIO stayed between 2.1 and 2.5 (nasopharyngeal temperature down to 14° C). However, in the functioning brain the animals were reported to have a QIO of 5 or greater between 28° and 18° C. According to Michenfelderv" this striking increase is explained by the cessation of brain function (isoelectric electroencephalogram) that characteristically occurs between 18° and 21 ° C. Clinical implications. The present data have clinical implications. The measurements of oxygen extraction and jugular venous oxygen saturation indicate that a significant cerebral metabolic rate continues during profound hypothermia. The estimated cerebral metabolic rates at 15° and 20° C wereapproximately 14%and 23% of those at normothermia. The fact that jugular bulb blood maintained a high venous oxygen saturation when the pump flow was reduced to 0.5 L/min/m2 strongly suggests that this pump flow was adequate for oxygen delivery to the brain when the nasopharyngeal temperature was close to 15° C. This is in agreement with experimental findings that a flow of 15 to 30 nil/kg/min appears to be adequate at profound hypothermia.P: 27, 28 This study supports the view that low flow perfusion provides better cerebral protection than total circulatory arrest at profound hypothermia. This statement refers to the observation that oxygen delivery to the brain at profound hypothermia seemed adequate, even at low systemic blood flows. The equally large release of creatine kinase isoenzyme BB after arrest and low flowprocedures reported from both animal 18 and clinical studies, 1 however, is puzzling. This study focuses on cerebral perfusion and oxygen delivery to the brain. It is appreciated that other components of the extracorporeal management, not primarily related to oxygen delivery to the brain, may be equally important for the development of cerebral injury. Another possible mechanism for injury to the brain is related to a flow metabolism mismatch, not during profound hypothermia with low systemic flow but during cooling and rewarming with full but possibly still inadequate systemic flow. This topic has been discussed previously.l? Future studies will be directed at optimizing

Profound hypothermia in children

II3

the relationship between cerebral blood flow and metabolic demand during these periods, This may include a more precise control of arterial carbon dioxide tension and hematocrit, a moderate speed of rewarming, and the use of anesthetic agents that depress the cerebral metabolic rate, REFERENCES I. Rossi R, van der Linden J, Ekroth R, Scallan M, Thompson RJ, Lincoln C. No flowor lowflow? A study ofthe brain specific ischemic marker creatine kinase BB after deep hypothermic procedures. J THORAC CARDIOVASC SURG 1989;98:193-9. 2. Wells FC, Coghill S, Caplan HL, Lincoln C. Duration of circulatory arrest does influence the psychological development of children after cardiac operations in early life. J THORAC CARDIOVASC SURG 1983;86:823-31. 3. Lundar T, Lindberg H, Lindegaard KF, et al. Cerebral perfusion during major cardiac surgery in children. Pediatr Cardiol 1987;8:161-5. 4. Greeley WJ, Ungerleider RM, Smith L, Reves JG. The effects of deep hypothermic cardiopulmonary bypass and total circula tory arrest on cerebral blood flowin infants and children. J THORAC CARDIOVASC SURG 1989;97:737-45. 5. Toole JF. Cerebrovascular disorders. 3rd ed. New York: Raven Press, 1984:9. 6. Gustafsson D, Stale H, Bjorkman JA, Gennser G. Deviation of haemodynamic information from ultrasonic recordings of aortic diameter changes. Ultrasound Med BioI 1989;3:189-99. 7. Swan H. The importance of acid-base management for cardiac and cerebral preservation during open heart operations. Surg Gynecol Obstet 1984;158:391-414. 8. DeWitt LD, Wechsler LR. Transcranial Doppler. Stroke 1988;19:915-21. 9. Busija DW, Heistad DD, Marcus ML. Continuous measurements of cerebral blood flow in anesthetized cats and dogs. Am J Physiol 1981;241:H228-34. 10. Hokanson DE, Mozersky DJ, Summner DS, Strandness DE Jr. A phase locked echo tracking system for recording arterial diameter changes in vivo. J Appl Physiol 1972; 32:728-33. 11. Siegel S. Nonparametric statistics of behavioral sciences. Tokyo: McGraw-Hill, 1956. 12. Draper NR, Smith H. Applied regression analysis. New York: John Wiley, 1966. 13. Kirklin JW, Barratt-Boyes B. Cardiac surgery. New York: John Wiley, 1986:30-3. 14. Bishop CCR, Powel S, Rutt D, Browse NL. Transcranial Doppler measurements of middle cerebral artery blood flow velocity: a validation study. Stroke 1986;17:913-5. 15. Lindegaard KF, Lundar T, Wiberg J, Sjoberg D, Aaslid R, Nornes H. Variations in middle cerebral artery blood flow investigated with noninvasive transcranial blood velocity measurements. Stroke 1987;18:1025-30.

1 1 4 van der Linden et al.

16. vander Linden J, WesslenO, Ekroth R, Tyden H, von Ahn H. Transcranial Doppler-estimatedversusthermodilutionestimated measurement of cerebral blood flow during cardiac operations. J THORAC CARDIOVASC SURG [In press]. 17. Tanaka J, Shiki K, Asou T, Yasui H, Tokunaga K. Cerebral autoregulation during deep hypothermic nonpulsatile cardiopulmonarybypasswithselective cerebral perfusionin dogs. J THORAC CARDIOVASC SURG 1988;95:124-32. 18. Molina JE, Einzig S, Angeline RM, Bianco RW, Marks JA, Rasmussen TM. Brain damage in profound hypothermia: perfusion versus circulatory arrest. J THORAC CAR0I0VASC SURG 1984;87:596-604. 19. GovierAV, RevesJG, McKay RD, eta!. Factors and their influenceon regional cerebral blood flowduring nonpulsatile cardiopulmonary bypass. Ann Thorac Surg 1984; 38:592-600. 20. MavroudisC, BrownGL, Katzrnark SL, Howe WR, Gray LA. Flow distribution in infant pigs subjected to surface cooling, deep hypothermia, and circulatory arrest: deleterious effectsin pigswith left-to-rightshunts. J THORAC CAR0I0VASC SURG 1984;87:665-72. 21. Murkin JM, Farrar JK, Tweed WA, McKenzie FN, Guiradon G. Cerebral autoregulation and flow/metabolismcouplingduring cardiopulmonarybypass:the influence of paco-. Anesth Analg 1987;66:825-32. 22. Henriksen L, Hjelms E, Lindeburgh T. Brain hyperperfusion during cardiac operations. J THORAC CARDIOVASC SURG 1983;86:202-8.

The Journal of Thoracic and Cardiovascular Surgery

23. Soma Y, Hirotani T, Yozu R, et a!. A clinical study of cerebral circulation during extracorporeal circulation. J THORAC CAROIOVASC SURG 1989;97:187-93. 24. Harris EA. Metabolicaspectsof profoundhypothermia.In: Barratt-Boyes BG, Neutze JM, Harris E, eds. Heart disease in infancy. Baltimore: Williams & Wilkins, 1973;65. 25. Michenfelder JD. Anesthesia and the brain. New York: Churchill Livingstone, 1988:23-34. 26. Steen PA, Newberg LA, Milde JH, Michenfelder JD. Hypothermia and barbiturates: individual and combined effectson canine cerebral oxygenconsumption. Anesthesiology 1983;58:527-32. 27. Fox LS, BlackstoneEH, Kirklin JW, BishopSP, Bergdahl LA, BradleyEL. Relationshipof brain bloodflow and oxygen consumption to perfusionflow rate during profoundly hypothermiccardiopulmonarybypass.J THORAC CARDIaVASC SURG 1984;87:658-64. 28. Miyamoto K, Kawashima Y, Matsuda H, Okuda A, Maeda S, Hirose H. Optimal perfusionflow rate for the brain during deep hypothermic cardiopulmonary bypass at 20 C: an experimental study. J THORAC CARDIOVASC SURG 1986;92:1065-70. 29. van der Linden J, Ekroth R, Lincoln C, Pugsley W, Scallan M, Tyden H. Is cerebral blood flow/metabolic mismatch during rewarming a risk factor after profoundhypothermic procedures in small children? Eur J Cardiothorac Surg 1989;3:209-15.

Bound volumes availableto subscribers Boundvolumesof THEJOURNAL OFTHORACIC AND CARDIOVASCULAR SURGERY are availableto subscribers(only) for the 1991 issues from the Publisher,at a cost of$64.00 for domestic,$86.48 for Canadian, and $82.00for internationalsubscribersfor Vol. 101 (JanuaryJune) and Vol. 102 (July-December). Shipping charges are included. Each bound volume contains a subject and author index and all advertisingis removed.Copiesare shippedwithin60 days after publicationof the last issueof the volume. The bindingis durable buckram with the JOURNAL name, volumenumber, and year stamped in goldon the spine. Paymentmust accompany all orders. Contact MosbyYear Book,lnc., SubscriptionServices, 11830 Westline Industrial Drive,St. Louis, Missouri63146-3318, USA; phone (800) 325-4177, ext. 4351. Subscriptions must be in force to qualify. Boundvolumes are not available in place of a regular JOURNAL subscription.

0