- Email: [email protected]
Hemoconcentration induced by surface hypothermia in infants
J
THORAC CARDIOVASC SURG
80:236-241, 1980
Hemoconcentration induced by surface hypothermia in infants The effects of surface hypothermia (25 0 C) on arterial hematocrit value (by microcentrifuge) and plasma protein concentration (by refractometry) were studied in infants undergoing surface cooling for cardiac operations. To analyze in detail the mechanisms leading to the observed changes in patients, we performed parallel studies on normal dogs and permanently splenectomized dogs. In these dogs, measurements were also made of plasma volume (by 12"I-albumin) and red cell volume (by o'Cr-erythrocytes). Arterial hematocrit value increased progressively during surface cooling in infants. Assuming that red cell volume remained constant and that the ratio of whole body red cell percentage to arterial hematocrit value increased during surface cooling in infants as in splenectomized dogs, we estimated percent changes in plasma volume in infants from arterial hematocrit data. The computed plasma volume decreased progressively as the body temperature was decreased. Since plasma protein concentration remained constant, the loss of plasma volume suggested a sequestration of whole plasma in portions of the circulatory bed and/or an extravasation of whole plasma into the interstitial space,
Richard Y. Z. Chen, M.D.,* Anthony E. Wicks, M.D.,** and Shu Chien, M.D., Ph.D.,*** New York, N. Y.
Surface hypothermia in conjunction with limited cardiopulmonary bypass has been employed frequently for early surgical correction of congenital heart defects in young infants. 1-4 It has been shown in animal experiments that surface cooling induces hemoconcentration.":" However, this phenomenon has not been documented in human infants. In this investigation, we have studied the effect of surface cooling on arterial hematocrit and plasma protein concentration in infants undergoing surface hypothermia for cardiac operations. To analyze in detail the mechanisms leading to the From the Departments of Anesthesiology and Physiology, College of Physicians and Surgeons, Columbia University, New York, N. Y.
Supported in part by U.S. Public Health Service Grants HL 12738 and HL 16851. Presented in part at 1978 Annual Meeting of the American Society of Anesthesiologists. Received for publication Jan. 23, 1980. Accepted for publication Feb. 8, 1980. Address for reprints: Dr. Richard Y. Z. Chen, 622 West 168th SI., New York, N. Y. 10032. *Assistant Professor of Anesthesiology. **Assistant Professor of Anesthesiology. Current address: Department of Anesthesiology, University of Miami, School of Medicine, Miami, Fla. ***Professor of Physiology.
236
observed changes in patients, we" 9 performed parallel experiments on normal dogs and chronically splenectomized dogs. In these dogs, plasma volume and red cell volume were also measured. Correlation of experimental results with clinical findings has allowed an estimation of the effect of surface cooling on changes in plasma volume in these infants.
Methods Ten patients, aged 2 1/ 2 to 21 months with body weight between 2.5 to 7.5 kg, were studied. Informed consent regarding the nature and risks of the study, which had been approved by the Institutional Review Board, was obtained from the parent or guardian of each patient. These patients were given nothing by mouth for 6 to 8 hours prior to the induction of anesthesia without premedication. Anesthesia was induced with ketamine (10 mg/kg intramuscularly [1M]). An intravenous catheter was introduced percutaneously or by cutdown. Succinylcholine (I mg/kg intravenously [IV]) was given to facilitate nasal endotracheal intubation. A thermistor attached to an esophageal stethoscope was inserted to the heart level for temperature monitoringv'" An arterial line in a radial or femoral artery and a central venous line via a femoral vein were used for pressure monitoring and blood sampling. Anesthesia was maintained with morphine (I mg/kg IV)
0022-5223/80/080236+06$00.60/0 © 1980 The C. V. Mosby Co.
Volume 80
Hemoconcentration induced by surface cooling
Number 2
237
August, 1980
Table I. Control and post-cooling values in infants undergoing surface hypothermia (mean ± SEM)
Control Hypothermia
Esophageal temp. (OC)
Arterial hematocrit (%)
PPC (gm/IOO ml)
MCV (JL3)
36.2 ± 0.3 25.2 ± 0.3*
41.6 ± 4.2 44.8 ± 4.2*
5.65 ± 0.75 5.62 ± 0.18
90.3 ± 6.4 90.6 ± 7.1
Legend: PPC. Plasma prole in concentration. MCV. Mean corpuscular volume . • Student's t lest on difference between hypothermia and control: p < 0.01.
in divided doses. Pancuronium bromide, 0.1 mg/kg/hr IV was given to prevent shivering and to facilitate ventilation. The procedures usually took 45 minutes before active surface cooling started. At all times the patients were manually ventilated with 100% oxygen. Since there was no effort to maintain normal body temperature, the esophageal temperature decreased by approximately 1°C during the preparatory period. Surface hypothermia was achieved by covering the body surface with small plastic bags filled with crushed ice. The IV line was kept closed during the active cooling period except for flushing with 2 ml of saline solution after each medication. Immediately prior to the application of ice bags, an arterial blood sample was taken for the determination of hematocrit and plasma protein concentration as control. Successive blood samples were taken at every 1°C change in body temperature until cardiopulmonary bypass commenced. The duration of active surface cooling ranged from 30 to 60 minutes, and the body temperature was lowered to approximately 25° C. Total dead space of the arterial catheter was 0.5 ml. The first I ml of blood withdrawn from the arterial catheter was discarded and the following 0.5 ml of blood was taken as the arterial sample. In preliminary studies, this maneuver was found to be adequate to avoid sampling errors and to yield a consistent hematocrit value. II Hematocrit reading of each arterial blood sample was determined in duplicate after 5 minutes of centrifugation at 15,000 x g in a microcentrifuge. The reading was then multiplied with a plasma trapping factor of 0.99 12 to give a corrected arterial hematocrit value. In four infants, the red cell count of each arterial blood sample was determined with a Coulter counter (Model B, Coulter Electronics, Hialeah, Fla.). The mean corpuscular volume was calculated from the hematocrit value and the cell count. Plasma protein concentration in each arterial blood sample was determined in duplicate by a refractometric method. 1:1 The blood volume is equal to the sum of cell volume (CV) and plasma volume (PV). The overall cell percentage (H o ) in the entire circulation is calculated as H" = 100 x CV/(PV + CV)
(I)
The F(,l'lls factor (F,.), which reflects the uneven distribution of cells and plasma in the circulatory bed ;'" is calculated as F,. = H"IH"
(2)
where H, is arterial hematocrit value. Rearranging Eq I gives PV = CV x (lOa - H")/H,,
(3)
The combination of Eq 2 and Eq 3 yields PV = CV x
(100 I) F,. x H"
(4)
It is unlikely for cell volume to change suddenly in the absence of transfusion or bleeding, and this constancy of cell volume in hypothermia was indeed found in dog experiments." Therefore, the relative changes in plasma volume during surface cooling can be estimated from Eq 5, which was derived from Eq 4 with the use of F('t'ils factor and arterial hematocrit values at control and hypothermic temperatures: PV«) - PV",) = PV(,.)
(5)
where the suffix (c) denotes the control values before surface cooling and (t) denotes the subsequently determined values following the application of ice bags. ICV(L) is the cumulative amount of cell volume removed because of sampling loss. Since the spleen in the infant does not have a reservoir function like that in the dog, the F(,l'lls factor, which reflects red cell distribution in the circulation.v' is comparable in infants to that in the splenectomized dogs. Rosenthal and associates" reported that the value of F('(') in patients with congenital heart disease was 0.898. Since the value of F(. in splenectomized dogs increased from 0.851 at 37° C to 0.901 at 25° C body temperature," the values of Fell) in infants during surface cooling were obtained by assuming that the Feells factor increases with hypothermia in a manner similar to that found in the splenectomized
The Journal of
238
Chen, Wicks, Chien
Thoracic and Cardiovascular Surgery
uQ..
14
Q..
Z ~
12
~
10
~
8
g
.'"
~
w
0 Z
oj."..
ell:
,,,,#~....e-
-c
::I:
.~
U
~'
....
c.ft."
Z
w
~
....
4
ex w Q.
I
Dog s with Intacl Spleen «(1- 7 ) Splenectom iud Dogs ( n . 7 ) Humon ( n - 10 )
+4 +2 0
D'
6
Z w U
EW.l
U
w C)
:I: U
+6
W Q..
-2
meon t SEM
-2
-4
-6
- 8
- 10
- 12
CHANGE IN ESOPHAGEAL TEMPERATURE (OC) Fig. 2. Percent changes in plasma protein concentration
2
(PPC) for infants in relation to the changes in esophageal temperature. The values are mean ± SEM. The shaded areas
oOC-..L-
_~2----1L..
_- ..1..-. .L. .
_.. .L6-- -1L..
_-8~...L--:-&:--'-~----I
CHANGE IN ESOPHAGEAL lENIlPEltAlUIE Fig. 1. Percent changes in arterial hematocrit value for infants in relation to the changes in esophageal temperature. The values are mean ± SEM. Closed circles denote those values statistically different from the control values by Student's t test (p < 0.05). The shaded areas show those values obtained from dog experiments" for comparison. dog. With the assumption of a linear relation between Few and body temperature, Fell) at each temperature was calculated for use in Eq 5. The statistical significance of the difference between paired data was evaluated with Student's t test.
Results Values of esophageal temperature, arterial hematocrit, mean corpuscular volume, and plasma protein concentration determined prior to the application of ice bags and after surface cooling prior to the commencement of cardiopulmonary bypass in infants are shown in Table I. There is considerable variation of arterial hematocrit value and plasma protein concentration in infants with congenital heart disease according to the nature and severity of the disease process. In each individual, the values of arterial hematocrit and plasma protein concentration determined prior to the application of ice bags were used as their own control. Values obtained in the subsequent blood samples during active surface cooling were normalized with the control value. Percent change over the control value was plotted against the change in esophageal tempera-
show those values obtained from dog experiments comparison.
(8)
for
ture. Fig. 1 shows the changes in arterial hematocrit value as a result of active surface cooling. In infants, the arterial hematocrit value increased progressively during surface cooling. When the body temperature was lowered 11 C, the arterial hematocrit value increased by approximately 8% above the control value. The results obtained from infants were in general agreement with those from the splenectomized dogs. The plasma protein concentration did not change significantly in infants or in the dogs during surface cooling (Fig. 2). The mean corpuscular volume also remained constant in hypothermia (Table I). Fig. 3 shows the estimated percent changes of plasma volume in infants during hypothermia computed from Eq 5. As the body temperature was lowered by surface cooling, there was a progressive reduction of plasma volume. The dotted line with squares depicts the results without correction for sampling loss (i.e., ICV ,u = 0 in Eq 5). When the body temperature was decreased by 110 C, the plasma volume was reduced by 20%. When correction was made for this sampling loss, the actual reduction of plasma volume with II ° C decrease in body temperature was found to be 23.3%. The changes in plasma volume were statistically significant from the control values (p < 0.05) even when the body temperature was decreased only 2° C by surface cooling. 0
Discussion In the present study, we have demonstrated that the arterial hematocrit value increases during surface cool-
Volume 80 Number 2 August, 1980
ing in infants. For obvious reasons , the blood volume was not measured with the isotope dilution technique in these infants. We have correlated the clinical findings with those results obtained from parallel experiments performed on dogs . The experimental protocol for dogs was similar to the clinical setting for surface hypothermia . The changes of arterial hematocrit value in infants during surface cooling agreed in general with those found in splenectomized dogs. (see Fig. I). The increase in arterial hematocrit value during hypothermia was greater in normal than in splenectomized dogs . This was attributed to the contraction of canine spleen in response to cold, releasing cell-rich blood into the circulation. H It has been shown that canine spleen behaves as a principal blood reservoir and is capable of actively releasing cell-rich blood upon sympathetic nerve stimulation.J'"'!" whereas the spleen in infants does not have a reservoir function. On the assumption of a constant red cell volume, the plasma volume of infants was found to decrease progressively with hypothermia . Since there was neither blood transfusion nor significant blood loss during surface cooling and since alterations in red cell size can be ruled out by the constancy of mean corpuscular volume, the assumption of a constant red cell volume in these infants is warranted . The assumption of a constant red cell volume in infants during surface cooling is further supported by direct measurement of red cell volume in dog experirnents ." It should be emphasized, however, that this kind of assessment can be regarded only as an estimation of the changes in plasma volume in infants . More accurate results will have to be obtained by direct measurement of plasma volume, which unfortunately is not feasible in infants at the present time. As early as 1921, Smith and associates 19 suggested that there is an uneven distribution of cells in the overall circulation . With advances in the techniques for measuring cell volume and plasma volume, it was established that there is •'extra plasma" present in the minute vessels, so that the overall cell percentage is lower than the large vessel cell percentage. 20- 23 In infants and splenectomized dogs , the total blood volume can be viewed as being contained in two compartments H: the large vessels such as the arteries, veins, and heart chambers versus the minute vessels such as the arterioles, venules, and capillaries (Fig. 4). The large vessels have a higher cell percentage, and the minute vessels have "extra plasma" and a lower cell percentage . Therefore, the overall cell percentage in the entire circulation, i.e ., the cell volume as a percentage of the sum of cell volume and plasma volume (Eq
Hemoconcentration induced by surface cooling
239
o ~
w
~
::::>
-5
'6
\
.......
~
\l
r\ ~
z o
«
",I
I\',{ I
w
z
With Correction for Sampling lou
,
- 10
Q..
I:
-. -
~\\ \1
o
~
If)
:5
--.--Without Correction for Sampling lou
I\ '\1
-15
IT\ ',r..,1
u
~
Z w U 00: ~
Human
t-,r-, '\1
-20
1
- 25
Mean t S.E.M. (n=10)
L..-_...L.-_........_
o
........._
!\
I
.........._ _"------'
-2 -4 -6 -8 -10 12 CHANGE IN ESOPHAGEAL TEMPERATURE ("C)
Fig. 3. Estimatedpercentchanges in plasma volumefor infants in relation to changes in esophageal temperature. The values are mean ± SEM. Squares denote values calculated without correction for sampling loss (see text). Circles denote values calculated with correction for sampling loss (Eq 5). Closed squares and circles denote those values statistically different from the control values by Student's t test (p < 0.05).
100 %
Lo rg e Vesse l.
M inut e Ve sse l.
~
0::
U
0 ~ 50 % ~ w I
0
He
l ~~ I~ ~~~ ~~li BLO OD
--
-Ho
He l11li11:11il1
VO LUME
Fig. 4. Schematic drawings of hematocrit distribution in the circulation in normothermia for human beings and splenectomized dogs. Shaded area indicates cell volume. Open area indicates plasma volume. H,,, Overall cell percentage of entire circulation. H II • Large vessel cell percentage. H ,.. Minute vessel cell percentage.
The Journal of
240 Chen, Wicks, Chien
I), is smaller than the cell percentage determined from large vessels. The ratio of overall percentage to the large vessel cell percentage (hematocrit corrected for plasma trapping) is termed F""lls factor (Eq 2).14 For the plasma volume to be estimated from large vessel cell percentage, the F"ells factor must be taken into consideration. A relative decrease of the minute vessel volume and/or a relative increase in minute cell percentage may cause an increase in the F"l'lls factor in the face of an increase in the arterial hematocrit value. In hypothermia, there is an intense vasoconstriction" that may shut down a large portion of capillary bed and thereby lead to an increase in F"l'lls factor, as proposed by Chaplin and associates.v' The constancy of plasma protein concentration in hypothermia (see Fig. 2) indicates that the reduction in plasma volume is due to a loss of whole plasma from the effective circulating blood. Since the thoracic duct lymph flow was found to decrease at the same time," the loss of plasma volume represents a sequestration of whole plasma in certain portions of the circulatory bed and/or an extravasation of plasma into the interstitial space. Pang and colleagues" recently have shown an increase in bradykinin during surface hypothermia in infants. This could lead to extravasation of plasma and interstitial edema. Svanes' group'" reported that the capillary filtration coefficient remained constant in the omemtum of rabbits following active surface cooling. Further studies are needed to determine the capillary filtration coefficent of other vascular regions in hypothermia. Cold diuresis" and inhibition of active transport of sodium from cells at low temperature'" may contribute to hemoconcentration in hypothermia. These mechanisms should result in an increase in plasma protein concentration, but this value was found to remain constant in the present study. It is worth noting that with an infusion of lowmolecular-weight dextran at a rate of 10 mIlkg body weight, Mohri and associates'" found that the hematocrit values remained relatively constant in both infants and dogs during surface cooling from 37° to 25° C. The volume of administered dextran is equivalent to approximately 23% of plasma volume, an amount which was found to have been lost from the circulating blood volume following surface cooling to 25° C in the present study. Their findings can thus be explained by the replacement of the lost plasma in surface hypothermia. In conclusion, the results of the present study demonstrate that surface cooling in infants undergoing cardiac operations induces an increase in arterial hematocrit value and a decrease in plasma volume similar to the changes found in splenectomized dogs. At 25° C
Thoracic and Cardiovascular Surgery
body temperature, the reduction of plasma volume was approximately 23% of the precooling value. The hemoconcentration in hypothermia has been found to contribute significantly to the increase in blood viscosity and the rise in peripheral flow resistance in dogs." This may lead to an increase in cardiac workload. Therefore, the hemoconcentration induced by surface cooling may be detrimental to those infants with compromised cardiac functions. We are grateful to Mr. Daniel Batista and Mr. Juan Rodriquez for their excellent technical assistance.
REFERENCES Rittenhouse EA, Mohri H, Dillard DH, Merendino KA: Deep hypothermia in cardiovascular surgery. Ann Thorac Surg 17:63-98, 1974 2 Barratt-Boyes BG, Neutze JM: Primary repair of tetralogy of Fallot in infants using profound hypothermia with circulatory arrest and limited cardiopulmonary bypass. Ann Surg 178:406-411, 1973 3 Ingersoll I, Lell W, Allarde R, Corsser G: The role of profound hypothermia in infants undergoing surgical correction of complicated heart defects. Anesth Analg (Cleve) 54:660-668, 1975 4 Subramanian S, Wagner H: Correction of transpositionof the great arteries in infants under surface-induced deep hypothermia. Ann Thorac Surg 16:391-401, 1973 5 D'Amato HE, Hegnauer H: Blood volume in hypothermic dog. Am J Physiol 173:703-705, 1963 6 Thauer R: The circulation in hypothermia of nonhibernating animals and men, Handbook of Physiology. Circulation, Sect 2, Vol III, Chap 54, Washington, D. C., 1965, American Physiological Society, pp 1899-1920 7 Popovic C, Popovic P: Hypothermia in Biology and in Medicine. New York, 1974, Grune & Stratton, Inc., pp 79-167 8 Chen RYZ, Chien S: Plasma volume, red cell volume and thoracic duct lymph flow in hypothermia. Am J Physiol 233:H605-H612, 1977 9 Chen RYZ, Chien S: Hemodynamic functions and blood viscosity in surface hypothermia. Am J Physiol 235: HI36-H143, 1978 10 Cooper KE, Kenyon JR: A comparison of temperatures measured in the rectum, esophagus and on the surface of the aorta during hypothermia in man. Br J Surg 44:616619, 1957 II Bourke DL: Error in intraoperative hematocrit determination. Anesthesiology 45:357-359, 1976 12 Chien S, Dellenback RJ, Usami S, Gregersen MI: Plasma trapping in hematocrit determination. Differences among animal species. Proc Soc Exp Bioi Med 119:1155-1158, 1965 13 Neuhausen BS, Rioch DM: The refractometric determination of serum proteins. J Bioi Chern 55:353-356, 1923
Volume 80 Number 2 August, 1980
14 Gregersen MI, Rawson RA: Blood volume. Physiol Rev 39:307-342, 1959 15 Rosenthal A, Button LN, Nathan DG, Miettinen OS, Nadas AS: Blood volume changes in cyanotic congenital heart disease. Am J Cardiol 27:162-167, 1971 16 Brooksby GA, Donald DE: Dynamic changes in splanchnic blood flow and blood volume in dogs during activation of sympathetic nerves. Circ Res 29:227-238, 1971 17 Donald DE, Aarhus LL: Active and passive release of blood from canine spleen and small intestine. Am J Physiol 227:1166-1172, 1974 18 Carneiro 11, Donald DE: Blood reservoir function of dog spleen, liver and intestine. Am J Physiol 232:H62-H67, 1977 19 Smith HP, Arnold HR, Whipple GH: Blood volume studies. VII. Comparative values of Welcker carbon monoxide and dye methods for blood volume determinations. Accurate estimation of absolute blood volume. Am J Physiol 56:336-360, 1921 20 Gibson JG II, Seligman AM, Peacock WC, Aub JC, Fine J, Evans RO: Distribution of red cells and plasma in large and minute vessels of the normal dog determined by radioactive isotope of iron and iodine. J Clin Invest 25:848-857, 1946 21 Chaplin H Jr, Mollison PL, Vetler H: The body/venous hematocrit ratio. Its constancy over a wide hematocrit range. J Clin Invest 32:1309-1316, 1953
Hemoconcentration induced by surface cooling
24 I
22 Huggins RA, Smith EL, Deavers S: Distribution of red cells and plasma in dogs. Am J Physiol 191: 163-166, 1957 23 Chien S, Gregersen MI: Determination of body fluid volume, Physical Techniques in Biological Research, Vol 4, WS Nastuk, ed., New York, 1962, Academic Press Inc., pp 1-105 24 Kunn LA, Turner JK: Alterations in pulmonary and peripheral vascular resistance in immersion hypothermia. Circ Res 7:366-374, 1959 25 Pang LM, Stalcup SA, Lipset CF: Bradykinin generation during hypothermic cardiopulmonary bypass. Am Rev Respir Dis 115:362, 1977 26 Svanes K, Zweifach BW, Intaglietta M: Effect of hypothermia on transcapillary fluid exchange. Am J Physiol 218:981-989, 1970 27 Segar WE, Riley PA, Barila TG: Urinary composition during hypothermia. Am J Physiol 185:528-532, 1956 28 Skow JC: Enzymatic basis for active transport across cell membrane. Physiol Rev 45:596-617, 1965 29 Mohri H, Hessel EA, Nelson RJ, Matano I, Anderson HN, Dillard DH, Merendino KA: Use of Rheomacrodex and hyperventilation in prolonged circulatory arrest under deep hypothermia induced by surface cooling. Method for open heart surgery in infants. Am J Surg 112:241-250, 1966
THORAC CARDIOVASC SURG
80:236-241, 1980
Hemoconcentration induced by surface hypothermia in infants The effects of surface hypothermia (25 0 C) on arterial hematocrit value (by microcentrifuge) and plasma protein concentration (by refractometry) were studied in infants undergoing surface cooling for cardiac operations. To analyze in detail the mechanisms leading to the observed changes in patients, we performed parallel studies on normal dogs and permanently splenectomized dogs. In these dogs, measurements were also made of plasma volume (by 12"I-albumin) and red cell volume (by o'Cr-erythrocytes). Arterial hematocrit value increased progressively during surface cooling in infants. Assuming that red cell volume remained constant and that the ratio of whole body red cell percentage to arterial hematocrit value increased during surface cooling in infants as in splenectomized dogs, we estimated percent changes in plasma volume in infants from arterial hematocrit data. The computed plasma volume decreased progressively as the body temperature was decreased. Since plasma protein concentration remained constant, the loss of plasma volume suggested a sequestration of whole plasma in portions of the circulatory bed and/or an extravasation of whole plasma into the interstitial space,
Richard Y. Z. Chen, M.D.,* Anthony E. Wicks, M.D.,** and Shu Chien, M.D., Ph.D.,*** New York, N. Y.
Surface hypothermia in conjunction with limited cardiopulmonary bypass has been employed frequently for early surgical correction of congenital heart defects in young infants. 1-4 It has been shown in animal experiments that surface cooling induces hemoconcentration.":" However, this phenomenon has not been documented in human infants. In this investigation, we have studied the effect of surface cooling on arterial hematocrit and plasma protein concentration in infants undergoing surface hypothermia for cardiac operations. To analyze in detail the mechanisms leading to the From the Departments of Anesthesiology and Physiology, College of Physicians and Surgeons, Columbia University, New York, N. Y.
Supported in part by U.S. Public Health Service Grants HL 12738 and HL 16851. Presented in part at 1978 Annual Meeting of the American Society of Anesthesiologists. Received for publication Jan. 23, 1980. Accepted for publication Feb. 8, 1980. Address for reprints: Dr. Richard Y. Z. Chen, 622 West 168th SI., New York, N. Y. 10032. *Assistant Professor of Anesthesiology. **Assistant Professor of Anesthesiology. Current address: Department of Anesthesiology, University of Miami, School of Medicine, Miami, Fla. ***Professor of Physiology.
236
observed changes in patients, we" 9 performed parallel experiments on normal dogs and chronically splenectomized dogs. In these dogs, plasma volume and red cell volume were also measured. Correlation of experimental results with clinical findings has allowed an estimation of the effect of surface cooling on changes in plasma volume in these infants.
Methods Ten patients, aged 2 1/ 2 to 21 months with body weight between 2.5 to 7.5 kg, were studied. Informed consent regarding the nature and risks of the study, which had been approved by the Institutional Review Board, was obtained from the parent or guardian of each patient. These patients were given nothing by mouth for 6 to 8 hours prior to the induction of anesthesia without premedication. Anesthesia was induced with ketamine (10 mg/kg intramuscularly [1M]). An intravenous catheter was introduced percutaneously or by cutdown. Succinylcholine (I mg/kg intravenously [IV]) was given to facilitate nasal endotracheal intubation. A thermistor attached to an esophageal stethoscope was inserted to the heart level for temperature monitoringv'" An arterial line in a radial or femoral artery and a central venous line via a femoral vein were used for pressure monitoring and blood sampling. Anesthesia was maintained with morphine (I mg/kg IV)
0022-5223/80/080236+06$00.60/0 © 1980 The C. V. Mosby Co.
Volume 80
Hemoconcentration induced by surface cooling
Number 2
237
August, 1980
Table I. Control and post-cooling values in infants undergoing surface hypothermia (mean ± SEM)
Control Hypothermia
Esophageal temp. (OC)
Arterial hematocrit (%)
PPC (gm/IOO ml)
MCV (JL3)
36.2 ± 0.3 25.2 ± 0.3*
41.6 ± 4.2 44.8 ± 4.2*
5.65 ± 0.75 5.62 ± 0.18
90.3 ± 6.4 90.6 ± 7.1
Legend: PPC. Plasma prole in concentration. MCV. Mean corpuscular volume . • Student's t lest on difference between hypothermia and control: p < 0.01.
in divided doses. Pancuronium bromide, 0.1 mg/kg/hr IV was given to prevent shivering and to facilitate ventilation. The procedures usually took 45 minutes before active surface cooling started. At all times the patients were manually ventilated with 100% oxygen. Since there was no effort to maintain normal body temperature, the esophageal temperature decreased by approximately 1°C during the preparatory period. Surface hypothermia was achieved by covering the body surface with small plastic bags filled with crushed ice. The IV line was kept closed during the active cooling period except for flushing with 2 ml of saline solution after each medication. Immediately prior to the application of ice bags, an arterial blood sample was taken for the determination of hematocrit and plasma protein concentration as control. Successive blood samples were taken at every 1°C change in body temperature until cardiopulmonary bypass commenced. The duration of active surface cooling ranged from 30 to 60 minutes, and the body temperature was lowered to approximately 25° C. Total dead space of the arterial catheter was 0.5 ml. The first I ml of blood withdrawn from the arterial catheter was discarded and the following 0.5 ml of blood was taken as the arterial sample. In preliminary studies, this maneuver was found to be adequate to avoid sampling errors and to yield a consistent hematocrit value. II Hematocrit reading of each arterial blood sample was determined in duplicate after 5 minutes of centrifugation at 15,000 x g in a microcentrifuge. The reading was then multiplied with a plasma trapping factor of 0.99 12 to give a corrected arterial hematocrit value. In four infants, the red cell count of each arterial blood sample was determined with a Coulter counter (Model B, Coulter Electronics, Hialeah, Fla.). The mean corpuscular volume was calculated from the hematocrit value and the cell count. Plasma protein concentration in each arterial blood sample was determined in duplicate by a refractometric method. 1:1 The blood volume is equal to the sum of cell volume (CV) and plasma volume (PV). The overall cell percentage (H o ) in the entire circulation is calculated as H" = 100 x CV/(PV + CV)
(I)
The F(,l'lls factor (F,.), which reflects the uneven distribution of cells and plasma in the circulatory bed ;'" is calculated as F,. = H"IH"
(2)
where H, is arterial hematocrit value. Rearranging Eq I gives PV = CV x (lOa - H")/H,,
(3)
The combination of Eq 2 and Eq 3 yields PV = CV x
(100 I) F,. x H"
(4)
It is unlikely for cell volume to change suddenly in the absence of transfusion or bleeding, and this constancy of cell volume in hypothermia was indeed found in dog experiments." Therefore, the relative changes in plasma volume during surface cooling can be estimated from Eq 5, which was derived from Eq 4 with the use of F('t'ils factor and arterial hematocrit values at control and hypothermic temperatures: PV«) - PV",) = PV(,.)
(5)
where the suffix (c) denotes the control values before surface cooling and (t) denotes the subsequently determined values following the application of ice bags. ICV(L) is the cumulative amount of cell volume removed because of sampling loss. Since the spleen in the infant does not have a reservoir function like that in the dog, the F(,l'lls factor, which reflects red cell distribution in the circulation.v' is comparable in infants to that in the splenectomized dogs. Rosenthal and associates" reported that the value of F('(') in patients with congenital heart disease was 0.898. Since the value of F(. in splenectomized dogs increased from 0.851 at 37° C to 0.901 at 25° C body temperature," the values of Fell) in infants during surface cooling were obtained by assuming that the Feells factor increases with hypothermia in a manner similar to that found in the splenectomized
The Journal of
238
Chen, Wicks, Chien
Thoracic and Cardiovascular Surgery
uQ..
14
Q..
Z ~
12
~
10
~
8
g
.'"
~
w
0 Z
oj."..
ell:
,,,,#~....e-
-c
::I:
.~
U
~'
....
c.ft."
Z
w
~
....
4
ex w Q.
I
Dog s with Intacl Spleen «(1- 7 ) Splenectom iud Dogs ( n . 7 ) Humon ( n - 10 )
+4 +2 0
D'
6
Z w U
EW.l
U
w C)
:I: U
+6
W Q..
-2
meon t SEM
-2
-4
-6
- 8
- 10
- 12
CHANGE IN ESOPHAGEAL TEMPERATURE (OC) Fig. 2. Percent changes in plasma protein concentration
2
(PPC) for infants in relation to the changes in esophageal temperature. The values are mean ± SEM. The shaded areas
oOC-..L-
_~2----1L..
_- ..1..-. .L. .
_.. .L6-- -1L..
_-8~...L--:-&:--'-~----I
CHANGE IN ESOPHAGEAL lENIlPEltAlUIE Fig. 1. Percent changes in arterial hematocrit value for infants in relation to the changes in esophageal temperature. The values are mean ± SEM. Closed circles denote those values statistically different from the control values by Student's t test (p < 0.05). The shaded areas show those values obtained from dog experiments" for comparison. dog. With the assumption of a linear relation between Few and body temperature, Fell) at each temperature was calculated for use in Eq 5. The statistical significance of the difference between paired data was evaluated with Student's t test.
Results Values of esophageal temperature, arterial hematocrit, mean corpuscular volume, and plasma protein concentration determined prior to the application of ice bags and after surface cooling prior to the commencement of cardiopulmonary bypass in infants are shown in Table I. There is considerable variation of arterial hematocrit value and plasma protein concentration in infants with congenital heart disease according to the nature and severity of the disease process. In each individual, the values of arterial hematocrit and plasma protein concentration determined prior to the application of ice bags were used as their own control. Values obtained in the subsequent blood samples during active surface cooling were normalized with the control value. Percent change over the control value was plotted against the change in esophageal tempera-
show those values obtained from dog experiments comparison.
(8)
for
ture. Fig. 1 shows the changes in arterial hematocrit value as a result of active surface cooling. In infants, the arterial hematocrit value increased progressively during surface cooling. When the body temperature was lowered 11 C, the arterial hematocrit value increased by approximately 8% above the control value. The results obtained from infants were in general agreement with those from the splenectomized dogs. The plasma protein concentration did not change significantly in infants or in the dogs during surface cooling (Fig. 2). The mean corpuscular volume also remained constant in hypothermia (Table I). Fig. 3 shows the estimated percent changes of plasma volume in infants during hypothermia computed from Eq 5. As the body temperature was lowered by surface cooling, there was a progressive reduction of plasma volume. The dotted line with squares depicts the results without correction for sampling loss (i.e., ICV ,u = 0 in Eq 5). When the body temperature was decreased by 110 C, the plasma volume was reduced by 20%. When correction was made for this sampling loss, the actual reduction of plasma volume with II ° C decrease in body temperature was found to be 23.3%. The changes in plasma volume were statistically significant from the control values (p < 0.05) even when the body temperature was decreased only 2° C by surface cooling. 0
Discussion In the present study, we have demonstrated that the arterial hematocrit value increases during surface cool-
Volume 80 Number 2 August, 1980
ing in infants. For obvious reasons , the blood volume was not measured with the isotope dilution technique in these infants. We have correlated the clinical findings with those results obtained from parallel experiments performed on dogs . The experimental protocol for dogs was similar to the clinical setting for surface hypothermia . The changes of arterial hematocrit value in infants during surface cooling agreed in general with those found in splenectomized dogs. (see Fig. I). The increase in arterial hematocrit value during hypothermia was greater in normal than in splenectomized dogs . This was attributed to the contraction of canine spleen in response to cold, releasing cell-rich blood into the circulation. H It has been shown that canine spleen behaves as a principal blood reservoir and is capable of actively releasing cell-rich blood upon sympathetic nerve stimulation.J'"'!" whereas the spleen in infants does not have a reservoir function. On the assumption of a constant red cell volume, the plasma volume of infants was found to decrease progressively with hypothermia . Since there was neither blood transfusion nor significant blood loss during surface cooling and since alterations in red cell size can be ruled out by the constancy of mean corpuscular volume, the assumption of a constant red cell volume in these infants is warranted . The assumption of a constant red cell volume in infants during surface cooling is further supported by direct measurement of red cell volume in dog experirnents ." It should be emphasized, however, that this kind of assessment can be regarded only as an estimation of the changes in plasma volume in infants . More accurate results will have to be obtained by direct measurement of plasma volume, which unfortunately is not feasible in infants at the present time. As early as 1921, Smith and associates 19 suggested that there is an uneven distribution of cells in the overall circulation . With advances in the techniques for measuring cell volume and plasma volume, it was established that there is •'extra plasma" present in the minute vessels, so that the overall cell percentage is lower than the large vessel cell percentage. 20- 23 In infants and splenectomized dogs , the total blood volume can be viewed as being contained in two compartments H: the large vessels such as the arteries, veins, and heart chambers versus the minute vessels such as the arterioles, venules, and capillaries (Fig. 4). The large vessels have a higher cell percentage, and the minute vessels have "extra plasma" and a lower cell percentage . Therefore, the overall cell percentage in the entire circulation, i.e ., the cell volume as a percentage of the sum of cell volume and plasma volume (Eq
Hemoconcentration induced by surface cooling
239
o ~
w
~
::::>
-5
'6
\
.......
~
\l
r\ ~
z o
«
",I
I\',{ I
w
z
With Correction for Sampling lou
,
- 10
Q..
I:
-. -
~\\ \1
o
~
If)
:5
--.--Without Correction for Sampling lou
I\ '\1
-15
IT\ ',r..,1
u
~
Z w U 00: ~
Human
t-,r-, '\1
-20
1
- 25
Mean t S.E.M. (n=10)
L..-_...L.-_........_
o
........._
!\
I
.........._ _"------'
-2 -4 -6 -8 -10 12 CHANGE IN ESOPHAGEAL TEMPERATURE ("C)
Fig. 3. Estimatedpercentchanges in plasma volumefor infants in relation to changes in esophageal temperature. The values are mean ± SEM. Squares denote values calculated without correction for sampling loss (see text). Circles denote values calculated with correction for sampling loss (Eq 5). Closed squares and circles denote those values statistically different from the control values by Student's t test (p < 0.05).
100 %
Lo rg e Vesse l.
M inut e Ve sse l.
~
0::
U
0 ~ 50 % ~ w I
0
He
l ~~ I~ ~~~ ~~li BLO OD
--
-Ho
He l11li11:11il1
VO LUME
Fig. 4. Schematic drawings of hematocrit distribution in the circulation in normothermia for human beings and splenectomized dogs. Shaded area indicates cell volume. Open area indicates plasma volume. H,,, Overall cell percentage of entire circulation. H II • Large vessel cell percentage. H ,.. Minute vessel cell percentage.
The Journal of
240 Chen, Wicks, Chien
I), is smaller than the cell percentage determined from large vessels. The ratio of overall percentage to the large vessel cell percentage (hematocrit corrected for plasma trapping) is termed F""lls factor (Eq 2).14 For the plasma volume to be estimated from large vessel cell percentage, the F"ells factor must be taken into consideration. A relative decrease of the minute vessel volume and/or a relative increase in minute cell percentage may cause an increase in the F"l'lls factor in the face of an increase in the arterial hematocrit value. In hypothermia, there is an intense vasoconstriction" that may shut down a large portion of capillary bed and thereby lead to an increase in F"l'lls factor, as proposed by Chaplin and associates.v' The constancy of plasma protein concentration in hypothermia (see Fig. 2) indicates that the reduction in plasma volume is due to a loss of whole plasma from the effective circulating blood. Since the thoracic duct lymph flow was found to decrease at the same time," the loss of plasma volume represents a sequestration of whole plasma in certain portions of the circulatory bed and/or an extravasation of plasma into the interstitial space. Pang and colleagues" recently have shown an increase in bradykinin during surface hypothermia in infants. This could lead to extravasation of plasma and interstitial edema. Svanes' group'" reported that the capillary filtration coefficient remained constant in the omemtum of rabbits following active surface cooling. Further studies are needed to determine the capillary filtration coefficent of other vascular regions in hypothermia. Cold diuresis" and inhibition of active transport of sodium from cells at low temperature'" may contribute to hemoconcentration in hypothermia. These mechanisms should result in an increase in plasma protein concentration, but this value was found to remain constant in the present study. It is worth noting that with an infusion of lowmolecular-weight dextran at a rate of 10 mIlkg body weight, Mohri and associates'" found that the hematocrit values remained relatively constant in both infants and dogs during surface cooling from 37° to 25° C. The volume of administered dextran is equivalent to approximately 23% of plasma volume, an amount which was found to have been lost from the circulating blood volume following surface cooling to 25° C in the present study. Their findings can thus be explained by the replacement of the lost plasma in surface hypothermia. In conclusion, the results of the present study demonstrate that surface cooling in infants undergoing cardiac operations induces an increase in arterial hematocrit value and a decrease in plasma volume similar to the changes found in splenectomized dogs. At 25° C
Thoracic and Cardiovascular Surgery
body temperature, the reduction of plasma volume was approximately 23% of the precooling value. The hemoconcentration in hypothermia has been found to contribute significantly to the increase in blood viscosity and the rise in peripheral flow resistance in dogs." This may lead to an increase in cardiac workload. Therefore, the hemoconcentration induced by surface cooling may be detrimental to those infants with compromised cardiac functions. We are grateful to Mr. Daniel Batista and Mr. Juan Rodriquez for their excellent technical assistance.
REFERENCES Rittenhouse EA, Mohri H, Dillard DH, Merendino KA: Deep hypothermia in cardiovascular surgery. Ann Thorac Surg 17:63-98, 1974 2 Barratt-Boyes BG, Neutze JM: Primary repair of tetralogy of Fallot in infants using profound hypothermia with circulatory arrest and limited cardiopulmonary bypass. Ann Surg 178:406-411, 1973 3 Ingersoll I, Lell W, Allarde R, Corsser G: The role of profound hypothermia in infants undergoing surgical correction of complicated heart defects. Anesth Analg (Cleve) 54:660-668, 1975 4 Subramanian S, Wagner H: Correction of transpositionof the great arteries in infants under surface-induced deep hypothermia. Ann Thorac Surg 16:391-401, 1973 5 D'Amato HE, Hegnauer H: Blood volume in hypothermic dog. Am J Physiol 173:703-705, 1963 6 Thauer R: The circulation in hypothermia of nonhibernating animals and men, Handbook of Physiology. Circulation, Sect 2, Vol III, Chap 54, Washington, D. C., 1965, American Physiological Society, pp 1899-1920 7 Popovic C, Popovic P: Hypothermia in Biology and in Medicine. New York, 1974, Grune & Stratton, Inc., pp 79-167 8 Chen RYZ, Chien S: Plasma volume, red cell volume and thoracic duct lymph flow in hypothermia. Am J Physiol 233:H605-H612, 1977 9 Chen RYZ, Chien S: Hemodynamic functions and blood viscosity in surface hypothermia. Am J Physiol 235: HI36-H143, 1978 10 Cooper KE, Kenyon JR: A comparison of temperatures measured in the rectum, esophagus and on the surface of the aorta during hypothermia in man. Br J Surg 44:616619, 1957 II Bourke DL: Error in intraoperative hematocrit determination. Anesthesiology 45:357-359, 1976 12 Chien S, Dellenback RJ, Usami S, Gregersen MI: Plasma trapping in hematocrit determination. Differences among animal species. Proc Soc Exp Bioi Med 119:1155-1158, 1965 13 Neuhausen BS, Rioch DM: The refractometric determination of serum proteins. J Bioi Chern 55:353-356, 1923
Volume 80 Number 2 August, 1980
14 Gregersen MI, Rawson RA: Blood volume. Physiol Rev 39:307-342, 1959 15 Rosenthal A, Button LN, Nathan DG, Miettinen OS, Nadas AS: Blood volume changes in cyanotic congenital heart disease. Am J Cardiol 27:162-167, 1971 16 Brooksby GA, Donald DE: Dynamic changes in splanchnic blood flow and blood volume in dogs during activation of sympathetic nerves. Circ Res 29:227-238, 1971 17 Donald DE, Aarhus LL: Active and passive release of blood from canine spleen and small intestine. Am J Physiol 227:1166-1172, 1974 18 Carneiro 11, Donald DE: Blood reservoir function of dog spleen, liver and intestine. Am J Physiol 232:H62-H67, 1977 19 Smith HP, Arnold HR, Whipple GH: Blood volume studies. VII. Comparative values of Welcker carbon monoxide and dye methods for blood volume determinations. Accurate estimation of absolute blood volume. Am J Physiol 56:336-360, 1921 20 Gibson JG II, Seligman AM, Peacock WC, Aub JC, Fine J, Evans RO: Distribution of red cells and plasma in large and minute vessels of the normal dog determined by radioactive isotope of iron and iodine. J Clin Invest 25:848-857, 1946 21 Chaplin H Jr, Mollison PL, Vetler H: The body/venous hematocrit ratio. Its constancy over a wide hematocrit range. J Clin Invest 32:1309-1316, 1953
Hemoconcentration induced by surface cooling
24 I
22 Huggins RA, Smith EL, Deavers S: Distribution of red cells and plasma in dogs. Am J Physiol 191: 163-166, 1957 23 Chien S, Gregersen MI: Determination of body fluid volume, Physical Techniques in Biological Research, Vol 4, WS Nastuk, ed., New York, 1962, Academic Press Inc., pp 1-105 24 Kunn LA, Turner JK: Alterations in pulmonary and peripheral vascular resistance in immersion hypothermia. Circ Res 7:366-374, 1959 25 Pang LM, Stalcup SA, Lipset CF: Bradykinin generation during hypothermic cardiopulmonary bypass. Am Rev Respir Dis 115:362, 1977 26 Svanes K, Zweifach BW, Intaglietta M: Effect of hypothermia on transcapillary fluid exchange. Am J Physiol 218:981-989, 1970 27 Segar WE, Riley PA, Barila TG: Urinary composition during hypothermia. Am J Physiol 185:528-532, 1956 28 Skow JC: Enzymatic basis for active transport across cell membrane. Physiol Rev 45:596-617, 1965 29 Mohri H, Hessel EA, Nelson RJ, Matano I, Anderson HN, Dillard DH, Merendino KA: Use of Rheomacrodex and hyperventilation in prolonged circulatory arrest under deep hypothermia induced by surface cooling. Method for open heart surgery in infants. Am J Surg 112:241-250, 1966