- Email: [email protected]
Corticosteroid and immune responses to cardiac surgery
ELSEVIER
Corticosteroid and immune responses to cardiac surgery Alexander A. Tinnikov,* Marina V. Legan,t Natalya A. Sheveluk,t Galina A. Cvetovskaya,t Sergei E. Naumenko,t and Sergei G. Sidelnikovt *Institute o f Cytology and Genetics and f lnstitute o f Pathology o f Circulation, Novosibirsk, Russia
Serum corticosteroid-binding globulin (CBG) and cortisol levels as well as subsets of circulating immunocompetent cells (1CCs) were measured during cardiac surgery. Closed heart surgeD' (closed mitral commissurotomy) resulted in an elevation of cortisol levels (up to 32 +_5 I,tg/dL by the end of the surgery) with no changes in CBG and ICC levels observed. Open heart surgery (open reconstruction of the mitral valve) in surface-induced hypothermia (without extracorporeal bypass) caused a dramatic drop in CBG activity (from 250 +_ 17 I.tM before the beginning of anesthesia to 198 +_15 pM by the end of cooling (just before cardiac arrest) and 158 + 13 IxM after 30 min of reperfusion), whereas cortisol levels were only slightly elevated by the end of cooling, and a significant increase (up to 17 + 2 I~g/dL) was observed only at the end of the surgery (60 rain of reperfusion and warming). Similar to CBG, a significant decline in circulating 1CC contents" occurred in response to cooling and circulator)., arrest. (Steroids 61:411-415, 1996)
Keywords:stress; corticosteroid-binding globulin; immunocompetent cells; hypothermia; cardiac arrest
Introduction Open heart surgery is a most severe challenge for the body. Study of the responses of the homeostatic mechanisms to such surgery may shed new light on their nature. In a previous study we found substantial differences in the corticosteroid response to two different types of cardiovascular surgery in children: with (open heart surgery) and without (closed heart surgery) cardiac arrest. ~ During open heart surgery in surface-induced hypothermia a dramatic drop in the activity of serum corticosteroid-binding globulin (CBG) was observed, which was most pronounced after resuscitation. At the same time, closed heart surgery in normothermia without cardiovascular arrest had no effect on CBG activity, in accordance with numerous data suggesting that CBG level, in contrast to corticosteroids, is a rather stable parameter. 23 The only w e l l - k n o w n acute stressor affecting C B G seems to be inflammation. 4 The close relation between immune and pituitary-adrenal systems is a generally accepted fact. 5'6 A rapid decline in C B G activity during inflammation fits naturally into the system of these relations. Being a serine protease inhibitor, 7 C B G was reported to undergo
Address reprint requests to Dr. Alexander A. Tinnikov, Institute of Cytology and Genetics, Novosibirsk 90, 630090, Russian Federation. Fax: 0073832-356-558. E-mail: [email protected]. Received June 26, 1995; accepted March 15, 1996. Steroids 61:411-415, 1996 © 1996 by Elsevier Science Inc. 655 Avenue of the Americas. New York, NY 10010
enzymatic cleavage in sites of inflammation, resulting in the release of free corticosteroids known as strong antiinflammatory agents. 8 Thus, because of its molecular nature, CBG may be involved in the corticosteroid regulation of the immune system. In the present study, corticosteroid and immune responses to heart surgery in adults were observed. As in the study on children, l we investigated responses to two types of the surgery: without cardiac arrest (closed heart surgery) at normal body temperature, and with circulatory arrest (open heart surgery) in surface-induced hypothermia (without extracorporeal bypass). The immune response was assessed by changes in the circulating immunocompetent cell (ICC) contents. 9
Experimental Subjects Seventeen subjects undergoing corrective operations for acquired mitral valve disease were under observation. Twelve subjects (eight men, four women, mean age 42 years) underwent open reconstruction of the mitral valve (open heart surgery, under conditions of hypothermia). In five patients, closed mitral commissurotomy was performed (four men, one woman, mean age 40 years) (closed heart surgery). Anesthesia in the hypothermic group was performed as follows. After premedication with morphine (0.14-0.16 mg/kg), dimedrol (0.14-0.16 mg/kg), and diazepam (0.14-0.16 mg/kg), anesthesia was induced with sodium thiopental (7-9 mg/kg), and tracheal 0039-128X/96/$15.00 PII S0039-128X(96)00060-0
Papers intubation was performed with the aid of succinylcholine (2.5-3.3 mg/kg). Artificial ventilation with normal tidal volume for normothermia was carried out before and after circulatory arrest. Anesthesia was maintained with diethyl ether-O 2 mixture (III~) supplemented with morphine (up to 1 mg/kg) and droperidol (up to 0.1 mg/kg). Muscle relaxation was accomplished with pipecurii bromidum (Arduan, Hungary) (0.1-0.15 mg/kg). Subjects were cooled with crushed ice, which was applied to the subject's head and body until the esophageal temperature was 30°C. Thereafter, ice was removed from the subject's body, and a left-sided thoracotomy was performed. During thoracotomy there was an additional fall in temperature of about 3-4°C, so that before cardiac arrest the esophageal temperature reached 27.2 _+0.6°C. Corrective procedures were performed during the period of circulatory arrest. After surgical repair, calcium chloride was administrated intravenously, manual cardiac massage was carried out, and the heart was warmed with water irrigation. When needed, electrical defibrillation was performed. After rewarming of the subject to 33-34°C with an electric blanket and irrigation of the pleural cavity with warm water, the chest was closed and the subject was transported to the intensive care unit. In three complicated cases the resuscitation was accompanied by intravenous administration of prednisolone (90-150 mg). In subjects undergoing closed mitral commissurotomy, anesthesia was maintained with ketamin (2.5 mg/kg/h), with supplemental doses of morphine (up to 0.5 mg/kg). Operative bleeding never exceeded 500 ml, and the volume of plasma or blood transfusions was in the range of 250-500 mL.
Blood In this study we have used serum specimens obtained from the blood taken for current medical control. The work has been approved by the Ethics Committee of the Institute for Pathology of Circulation. During open heart surgery, blood was taken from the subclavian vein (1) before the beginning of anesthesia; (2) at a body temperature of 30°C; (3) at a body temperature of 27.2°C, just before occlusion; (4) just after the end of the occlusion and restoration of heart action; (5) after 30 min of reperfusion, at a body temperature of 28°C; (6) after 60 min of reperfusion, at a body temperature of 30°C; (7) 24 h after the surgery; and (8) 3 days after the surgery. During closed heart surgery, blood was taken from the subclavian vein (1) before the beginning of anesthesia, (2) before skin incision, (3) before closed mitral commissurotomy, (4) at the end of the surgery, and (5) 24 h after surgery. Blood was centrifuged at 2000 x g for 20 min, and sera were stored at -20°C until assay. During the surgeries, hemoglobin levels were monitored automatically with STAT Profile-5 (Nova, USA).
Measurement of corticosteroid-binding globulin CBG was measured by a radioligand method using [3H]cortisol (one-point assay of maximum corticosteroid-binding capacity) as described previously.1 Briefly, examined sera specimens were previously treated with dextran-coated charcoal to remove endogenous steroids. The final dilution of the serum was 1:100. Diluted steroid-free serum (0.1 mL) was incubated with 60 nM [3H]cortisol without (total binding) or with (nonspecific binding) the addition of 20 ~M nonradioactive cortisol for 10 min at 37°C and for 1 h in an ice bath. Nonbound ligand was removed by charcoaldextran suspension (4°C). The time taken to add the suspension to all tubes did not exceed 10 seconds. After 30 seconds without mixing, tubes were centrifuged in a microfuge for 20-25 seconds. The radioactivity of the supernatant was counted at 50% efficiency.
Statistics All results are given as mean _+ SEM. Results were analyzed by repeated-measures ANOVA, and values of P < 0.05 were considered to be significant.
Results Closed-heart surgery Serum cortisol levels rose progressively in the course of the closed heart surgery at normal body temperature, with a peak at the end of the operation (Figure 1). A day after the surgery, cortisol levels were within normal basal ranges. CBG activity did not change either during or after surgery. Circulating immunocompetent cell subsets also were not affected substantially (data not shown).
Open-heart surgery. During open heart surgery, the serum cortisol level increased slightly during the initial stage as the subject's body was cooled; however, it was almost the same as the level observed at the beginning of anesthesia (Figure 2). A sig25O
200
4O
~,D 150
30
~
35
.=
Immunocompetent cell subpopulations The percentages of total T-lymphocytes (CD4 + CD8) and B-cells (CD22) of blood were determined from monoclonal antibodies (from Sorbent, St. Petersburg, Russia) and flow cytometry as described elsewhere. 9 Measurements were made in total lymphoid cell population obtained after Ficoll-hypaque (Sigma Chemical Co., St. Louis, MO, USA) separation.
~,
100
"~ ~
£
25
N
20
"~ o
..~
8 5o o 0
I l
I 2
I 3
I 4
t
1D
rO
Steps o f Operation
Cortisol assay Cortisol in serum was measured by the method of competitive protein binding, l° using human transcortin, cortisol (Sigma Chemical Co.), and [1,2,6,7-3H]cortisol (specific activity 90 Ci/ mmol; Isotop, St. Petersburg, Russia). Intra- and interassay coefficients of variation were less than 5% and 10%, respectively.
412
Steroids,
1996, vol. 61, July
Figure 1 Serum CBG (O) and cortisol (O) levels during closed heart surgery (closed mitral commissurotomy). Steps of operation: 1, before the beginning of anesthesia; 2, before skin incision; 3, before correction of the lesion; 4, the end of the surgery; 1D, 24 h after surgery. Each point represents the mean _+SEM of five subjects. *P< 0.05; * * * P < 0.001; compared with the starting point.
Corticosteroid and immune responses to surgery: Tinnikov et al. 50
~
250 35
L~T
~-
"~ 20(I
40
3o o~
¢J
o
25
/
. ~ 150
2o .~ w 15 ~
mo
.. 50 1= o
~
30
~ P~.
2o
10 10
~
0 ]
2
3
4
5
6
ID
3D
I 1
I 2
I 3
Serum CBG (e) and cortisol (©) levels during open heart surgery (open reconstruction of the mitral valve, in surface hypothermia). Steps of operation: 1, before the beginning of anesthesia; 2, body temperature 30°C; 3, body temperature 27.2°C, just before occlusion; 4, after resuscitation; 5, 30 rain of reperfusion, body temperature 28°C; 6, 60 min of reperfusion, body temperature 30°C; 1D, 24 h after surgery; 3D, 3 days after surgery. Each point represents the mean + SEM of 9-12 subjects. * P < 0.05; * * P < 0.01; * * * P < 0.001; compared with the starting point.
nificant increase in cortisol level was observed only at the end of the operation (60 rain after reperfusion) as well as 1 day after surgery. In three cases, patients were given a prednisolone injection (90-150 mg) after resuscitation (no postprednisolone data were included in the results presented in figures). Very high corticosteroid levels measured by the radioligand method were observed in these cases: 100, 110, and 150 btg/dL after 30 min of reperfusion and 60, 45, and 110 txg/dL after 60 min of reperfusion, respectively. CBG activity declined progressively during cooling and surgical procedures, with a significant sharp decline observed at the end of cooling (body temperature 27,2°C), reaching a minimum after resuscitation. One and three days after surgery, serum CBG remained at a lowered level. The course of circulating immunocompetent cell subsets during and after the surgery is shown in Figure 3. Since patterns for CD4 cells (T-helpers) and CD8 cells (Tsuppressors) were very similar, they were combined as total T-cells (CD4 + CD8). Compared with the starting point, the percentage of T-cells and B-cells declined considerably during the surgery. A significant drop in circulating T-cells occurred by the end of cooling. The percentage of T-cells returned to normal ranges 3 days after the operation. A significantly lowered percentage of B-cells was observed immediately after occlusion, after 30 min of reperfusion, and 3 days after surgery. Hemoglobin levels in blood during open heart surgery did not change during first steps of the operation and declined significantly after resuscitation (Figure 4).
Discussion The present study demonstrated substantial differences in corticosteroid responses to closed and open heart surgery, in accordance with previous observations of children with congenital heart defects.~ During closed heart surgery, an el-
I 5
I ¢i
I 1D
I 3D
Stepsof Operation
Steps of Operation
Figure 2
I 4
Figure 3
Circulating T-cell (CD4 + CD8) subset (e) and B-cell (CD22) subset (&) during open heart surgery. The steps of operation are the same as in Figure 2. *P< 0.05 compared with the starting point.
evation in cortisol levels occurred, with a peak (about 30 ixg/dL) at the end of the operation. CBG levels did not change in this case. Such a picture is typical of the pituitaryadrenal response to stress and agrees with the generally accepted concept of corticosteroid regulation: the key role is played by corticosteroids, namely by the free fraction, and the role of CBG is to provide the circulatory transport of bound (biologically inactive) hormones. A considerable and rather rapid decline in CBG activity during open heart surgery in adults as well as in children j suggests a more active role for this serum protein in the stress responses of the body. Furthermore, a course of serum cortisol by no means corresponded with the severe conditions of open heart surgery. This finding contrasts with the picture observed in children when serum cortisol was elevated to about 80 ~g/dL after resuscitation.~ It seems that available data on serum cortisol levels after cardiac arrest are contradictory. Murray t| observed extremely high cortisol levels (about 100 I~g/dL) immediately after cardiac arrest. On the other hand, there are some data providing evi140 I
- I ~ ~ 1 = _
~ _ ___~~
*
120
-3[
100 80
O =
60
40
20
0
1,
2,
3,
4,
5,
6,
Stepsof Operation Figure 4 Hemoglobin levels in blood during open heart surgery. The steps of the operation (1-6) are the same as in Figure 2. *P < 0.05 compared with the starting point.
S t e r o i d s , 1996, v o l . 61, J u l y
413
Papers dence tbr a rapid decline in serum cortisol after heart arrest, perhaps accounted for by an impairment of cortisol release from the adrenal cortex caused by global hypoxia. 12 Undoubtedly, cooling is a major contributing factor in the reactions of the body observed in this study. Swan et al. 13 reported that the pituitary-adrenal response to cardiovascular surgery under conditions of hypothermia was observed only during the beginning of hypothermia; hypothermia then suppressed further activation of the pituitary-adrenal system in spite of surgical injuries. It was also reported that secretion of corticosteroids in response to surgical stress was moderately elevated at body temperatures above 30°C, but was lowered at temperatures below 28°C. 14 Bernhard et al. 15 observed no increase in corticosteroid levels during surgery in humans and dogs in surface hypothermia. In light of the data cited, the lack of pronounced response of serum cortisol to open heart surgery does not seem too paradoxical. The striking difference between children I and adults in the response of cortisol to open heart surgery under conditions of hypothermia remains to be explained. Despite slight changes in serum cortisol, CBG response to open heart surgery in adults was the same as in children. This suggests that CBG may play a certain role, perhaps independent of corticosteroids, in responses of the body to such stress. In the case of inflammation, a drop in CBG activity was supposed to be a way of elevating the free fraction (biologically significant) of corticosteroids, the total level of which is actually very high during inflammation~'S; it is unclear, at the same time, whether the drop in CBG activity is of the same significance, considering the lack of a pronounced corticosteroid response. As estimated by the time course of hemoglobin, some degree of blood dilution occurred after resuscitation. No signs of dilution were seen before occlusion (cardiac arrest) when CBG levels were already significantly lowered. Blood dilution after resuscitation may contribute to the further drop in CBG activities observed after occlusion. However, on the whole, the maximum estimated blood dilution (a 12% decline in hemoglobin concentration) is much less than the maximum drop in corticosteroid-binding capacity (a decline of 37% from the initial level). Therefore, leaving aside the cause of the drop in CBG activities before occlusion, the drop in CBG activity during the surgery cannot be fully explained by blood dilution. Based on data available, some suppositions may be made about possible mechanisms of the drop in serum CBG activity. First, the drop may result from the uptake of circulating CBG by target cells. 2 Redistribution of CBG between blood and interstitial fluid may also account for physiological fluctuations in serum CBG levels. ~6 CBG is known to share a high degree of homology with serine protease inhibitors, 7 a superfamily of proteins that also includes proteins controlling blood pressure and coagulation. 17 It can be speculated that CBG may be mobilized and employed as a serine protease inhibitor to maintain homeostatic parameters during dramatic changes in the cardiovascular system that are related to open heart surgery. It should be noted that a significant lowering in CBG activity was observed as early as the end of cooling, before circulatory arrest. In this connection it should be noted that, in experiments with dogs, it
414
Steroids, 1996, vol. 61, July
was shown that a lowering of body temperature to 20°C was accompanied by lowering cardiac output by 80% and arterial pressure by 40%. ~8 Substantial functional shifts in the cardiovascular system may take place in patients undergoing surface-induced hypothermia. Stress-induced immunosuppression is usually associated with the action of corticosteroids. In the present study a "standard" stress reaction (closed heart surgery) had no effect on the parameters of the immune system under study. During open heart surgery, a considerable decrease in the percentage of circulating ICC subsets occurred against a background of only slight or even no elevation of serum cortisol. It is unclear whether suppression or activation is reflected by the observed course of circulating ICC when percentages of B-cells and T-cells changed in a similar manner during the operation. It seems that there is no general agreement on the immune response to surgical trauma. There are reports of suppression of humoral and cellular immunity ~9 as well as studies providing evidence for the lack of suppression 2° or even stimulation 2~ of immune response. It was recently demonstrated that ischemia and reperfusion caused injuries of the vascular bed, a key factor being the adherence and activation of leukocytes in the field of the endothelium. 22"23 The same phenomenon may be responsible for lowered circulating ICCs during open heart surgery accompanied by partial (due to hypothermic effects on cardiac function) and global ischemia followed by reperfusion. In this connection, the similarity between the course of CBG activity and that of ICC subsets is worth noting. A rather close relation of CBG to the immune system is demonstrated by lowered CBG activity during acute inflammation4; a probable cause of this response is cleavage of CBG by elastase (a serine protease) of activated leukocytes, s Activated leukocytes that adhere during ischemia-reperfusion also release proteases, which are a key factor responsible lbr injuries of the endothelium. 22'23 Thus, the decline in CBG activity during open heart surgery may be of the same nature as the similar decline during inflammation. It is of some interest that the application of exogenous serine protease inhibitors resulted in some protective effect during ischemia-reperfusion. 23 CBG may be an emergency endogenous inhibitor that counteracts proteases released by leukocytes in response to partial and global ischemia and subsequent reperfusion, thus playing a protective role in response of the body to open heart surgery. It should be noted that the role of CBG may be independent of corticosteroids.
Acknowledgments We thank Mrs. Valentina Maslova for her excellent technical assistance. The study was supported by grant 93-0406499 from the Russian Foundation for Fundamental Sciences.
References 1.
2.
Tinnikov AA, Legan MV, Pavlova IP, Litasova EE, Ivanova LN (1993). Serum corticosteroid-binding globulin levels in children undergoing heart surgery. Steroids 58:536-539. Rosner W (1990). The functions of corticosteroid-binding globulin
Corticosteroid and i m m u n e responses to surgery: Tinnikov et al.
3.
4.
5. 6.
7.
8.
9.
10.
11. 12.
and sex-hormone-binding globulin: recent advances. Endocr Rev 11:80-91. Tinnikov AA (1993). Corticosteroid-binding globulin levels in the rat serum under conditions of starvation and restriction of motions. Horm Metab Res 25:88-89. Vrancks R, Savu L, Maya M, Nunez EA (1987). ~x-Fetoprotein and transcortin behave as acute phase reactants in the maternal and fetal compartments of the inflammatory pregnant mouse. Endocrinology 120:1782-1789. Besedovsky H, Sorkin E (1977). Network of immune-neuroendocrine interactions. Clin Exp lmmunol 27:1-12. Blalock JE (1989). A molecular basis for bidirectional communication between the immune and neuroendocrine systems. Physiol Rev 69:1-32. Hammond GL, Smith CL, Goping IS, Underhill DA, Harley M J, Reventos J, Musto NA, Gunsalus GL, Bardin CW (1987). Primary structure of human corticosteroid binding globulin, deduced from hepatic and pulmonary cDNAs, exhibits homology with serine protease inhibitors. Proc Natl Acad &'i USA 84:5153-5157. Hammond GL, Smith CL, Underhill CM, Nguyen VTT (1990). Interaction between corticosteroid binding globulin and activated leukocytes in vitro. Biochem Biophys Res Commun 172:172-177. Fleisher TA, Marti GE, Hagengruber C (1988). Two-color flow cytometric analysis of monocyte depleted human lymphocyte subsets. Cytometrv 9:309-315. Murphy BE (1967). Some studies of protein binding of steroids and their application to the routine micro and ultramicro measurement o1' various steroids in body fluids by competitive protein-binding radioassay. J Clin Endocrinol Metab 27:973-990. Murray D (1967). Cortisol binding to plasma proteins in man in health, stress and at death. J Endocrinol 39:571-591. Lindner KH, Strohmenger HU, Ensinger H, Hetzel WD, Ahnefeld
13. 14.
15.
16.
17. 18.
19.
20. 2t.
22.
23.
FW, Georgieff M (1992). Stress hormone response during and after cardiopulmonary resuscitation. Anesthesiology 77:662-668. Swan H, Jenkins D, Helmreich ML (1957). The adrenal cortical response to surgery. Surge O" 42:202-217. Van Brunt EE, Ganong WF (1963). The effects of preanesthetic medication, anesthesia, and hypothermia on the endocrine response to injury. Anesthesiology 24:500-514. Bernhard WF, McMurrey JD, Ganong WF, Lenninan R (1956). The effect of hypothermia on the peripheral serum levels of free 17hydroxycorticosteroids in the dog and man. Ann Surg 143:210-215. Hsu BRS, Kuhn RW (1988). The role of the adrenal in generating the diurnal variation in circulating levels of corticostemid-binding globulin in the rat. Endocrinology 122:421-426. Carrel R, Travis J (1985). %-Antitripsin and the serpins: variation and concentration. Trends Biochem Sci 10:20-24. Riittenhouse EA, Mohri H, Dillard DH, Merendino KA (1974). Deep hypothermia in cardiovascular surgery. Ann Thorac Surg 17: 63-98. Cooper AJ, Irvine JM, Turnbull AR (1974). Depression of immunological responses due to surgery: a model in the guinea pig. bnmunology 27:393-399. Markley K, Smallman E, Evans G (1967). Antibody production in mice after thermal and tourniquet trauma. Surgery 61:896-9(13. Kinnaen P, Mahieu A, Van Geertruyden N (1978). Stimulation of antibody synthesis induced by surgical trauma in rats. Clin Exp hnmunol 32:243-252. Hemandes LA, Grisham MB, Twohig B, Arfors KE, Harlan JM, Granger DN (1987). Role of neutrophils in ischemia-reperfusioninduced microvascular injury. Am J Physiol 253:H699-H703. Zimmerman BJ, Granger DN (1990). Perfusion-induced leukocyte infiltration: role of elastase. Am J Phvsiol 259:H390 H394
Steroids, 1996, vol. 61, July
415
Corticosteroid and immune responses to cardiac surgery Alexander A. Tinnikov,* Marina V. Legan,t Natalya A. Sheveluk,t Galina A. Cvetovskaya,t Sergei E. Naumenko,t and Sergei G. Sidelnikovt *Institute o f Cytology and Genetics and f lnstitute o f Pathology o f Circulation, Novosibirsk, Russia
Serum corticosteroid-binding globulin (CBG) and cortisol levels as well as subsets of circulating immunocompetent cells (1CCs) were measured during cardiac surgery. Closed heart surgeD' (closed mitral commissurotomy) resulted in an elevation of cortisol levels (up to 32 +_5 I,tg/dL by the end of the surgery) with no changes in CBG and ICC levels observed. Open heart surgery (open reconstruction of the mitral valve) in surface-induced hypothermia (without extracorporeal bypass) caused a dramatic drop in CBG activity (from 250 +_ 17 I.tM before the beginning of anesthesia to 198 +_15 pM by the end of cooling (just before cardiac arrest) and 158 + 13 IxM after 30 min of reperfusion), whereas cortisol levels were only slightly elevated by the end of cooling, and a significant increase (up to 17 + 2 I~g/dL) was observed only at the end of the surgery (60 rain of reperfusion and warming). Similar to CBG, a significant decline in circulating 1CC contents" occurred in response to cooling and circulator)., arrest. (Steroids 61:411-415, 1996)
Keywords:stress; corticosteroid-binding globulin; immunocompetent cells; hypothermia; cardiac arrest
Introduction Open heart surgery is a most severe challenge for the body. Study of the responses of the homeostatic mechanisms to such surgery may shed new light on their nature. In a previous study we found substantial differences in the corticosteroid response to two different types of cardiovascular surgery in children: with (open heart surgery) and without (closed heart surgery) cardiac arrest. ~ During open heart surgery in surface-induced hypothermia a dramatic drop in the activity of serum corticosteroid-binding globulin (CBG) was observed, which was most pronounced after resuscitation. At the same time, closed heart surgery in normothermia without cardiovascular arrest had no effect on CBG activity, in accordance with numerous data suggesting that CBG level, in contrast to corticosteroids, is a rather stable parameter. 23 The only w e l l - k n o w n acute stressor affecting C B G seems to be inflammation. 4 The close relation between immune and pituitary-adrenal systems is a generally accepted fact. 5'6 A rapid decline in C B G activity during inflammation fits naturally into the system of these relations. Being a serine protease inhibitor, 7 C B G was reported to undergo
Address reprint requests to Dr. Alexander A. Tinnikov, Institute of Cytology and Genetics, Novosibirsk 90, 630090, Russian Federation. Fax: 0073832-356-558. E-mail: [email protected]. Received June 26, 1995; accepted March 15, 1996. Steroids 61:411-415, 1996 © 1996 by Elsevier Science Inc. 655 Avenue of the Americas. New York, NY 10010
enzymatic cleavage in sites of inflammation, resulting in the release of free corticosteroids known as strong antiinflammatory agents. 8 Thus, because of its molecular nature, CBG may be involved in the corticosteroid regulation of the immune system. In the present study, corticosteroid and immune responses to heart surgery in adults were observed. As in the study on children, l we investigated responses to two types of the surgery: without cardiac arrest (closed heart surgery) at normal body temperature, and with circulatory arrest (open heart surgery) in surface-induced hypothermia (without extracorporeal bypass). The immune response was assessed by changes in the circulating immunocompetent cell (ICC) contents. 9
Experimental Subjects Seventeen subjects undergoing corrective operations for acquired mitral valve disease were under observation. Twelve subjects (eight men, four women, mean age 42 years) underwent open reconstruction of the mitral valve (open heart surgery, under conditions of hypothermia). In five patients, closed mitral commissurotomy was performed (four men, one woman, mean age 40 years) (closed heart surgery). Anesthesia in the hypothermic group was performed as follows. After premedication with morphine (0.14-0.16 mg/kg), dimedrol (0.14-0.16 mg/kg), and diazepam (0.14-0.16 mg/kg), anesthesia was induced with sodium thiopental (7-9 mg/kg), and tracheal 0039-128X/96/$15.00 PII S0039-128X(96)00060-0
Papers intubation was performed with the aid of succinylcholine (2.5-3.3 mg/kg). Artificial ventilation with normal tidal volume for normothermia was carried out before and after circulatory arrest. Anesthesia was maintained with diethyl ether-O 2 mixture (III~) supplemented with morphine (up to 1 mg/kg) and droperidol (up to 0.1 mg/kg). Muscle relaxation was accomplished with pipecurii bromidum (Arduan, Hungary) (0.1-0.15 mg/kg). Subjects were cooled with crushed ice, which was applied to the subject's head and body until the esophageal temperature was 30°C. Thereafter, ice was removed from the subject's body, and a left-sided thoracotomy was performed. During thoracotomy there was an additional fall in temperature of about 3-4°C, so that before cardiac arrest the esophageal temperature reached 27.2 _+0.6°C. Corrective procedures were performed during the period of circulatory arrest. After surgical repair, calcium chloride was administrated intravenously, manual cardiac massage was carried out, and the heart was warmed with water irrigation. When needed, electrical defibrillation was performed. After rewarming of the subject to 33-34°C with an electric blanket and irrigation of the pleural cavity with warm water, the chest was closed and the subject was transported to the intensive care unit. In three complicated cases the resuscitation was accompanied by intravenous administration of prednisolone (90-150 mg). In subjects undergoing closed mitral commissurotomy, anesthesia was maintained with ketamin (2.5 mg/kg/h), with supplemental doses of morphine (up to 0.5 mg/kg). Operative bleeding never exceeded 500 ml, and the volume of plasma or blood transfusions was in the range of 250-500 mL.
Blood In this study we have used serum specimens obtained from the blood taken for current medical control. The work has been approved by the Ethics Committee of the Institute for Pathology of Circulation. During open heart surgery, blood was taken from the subclavian vein (1) before the beginning of anesthesia; (2) at a body temperature of 30°C; (3) at a body temperature of 27.2°C, just before occlusion; (4) just after the end of the occlusion and restoration of heart action; (5) after 30 min of reperfusion, at a body temperature of 28°C; (6) after 60 min of reperfusion, at a body temperature of 30°C; (7) 24 h after the surgery; and (8) 3 days after the surgery. During closed heart surgery, blood was taken from the subclavian vein (1) before the beginning of anesthesia, (2) before skin incision, (3) before closed mitral commissurotomy, (4) at the end of the surgery, and (5) 24 h after surgery. Blood was centrifuged at 2000 x g for 20 min, and sera were stored at -20°C until assay. During the surgeries, hemoglobin levels were monitored automatically with STAT Profile-5 (Nova, USA).
Measurement of corticosteroid-binding globulin CBG was measured by a radioligand method using [3H]cortisol (one-point assay of maximum corticosteroid-binding capacity) as described previously.1 Briefly, examined sera specimens were previously treated with dextran-coated charcoal to remove endogenous steroids. The final dilution of the serum was 1:100. Diluted steroid-free serum (0.1 mL) was incubated with 60 nM [3H]cortisol without (total binding) or with (nonspecific binding) the addition of 20 ~M nonradioactive cortisol for 10 min at 37°C and for 1 h in an ice bath. Nonbound ligand was removed by charcoaldextran suspension (4°C). The time taken to add the suspension to all tubes did not exceed 10 seconds. After 30 seconds without mixing, tubes were centrifuged in a microfuge for 20-25 seconds. The radioactivity of the supernatant was counted at 50% efficiency.
Statistics All results are given as mean _+ SEM. Results were analyzed by repeated-measures ANOVA, and values of P < 0.05 were considered to be significant.
Results Closed-heart surgery Serum cortisol levels rose progressively in the course of the closed heart surgery at normal body temperature, with a peak at the end of the operation (Figure 1). A day after the surgery, cortisol levels were within normal basal ranges. CBG activity did not change either during or after surgery. Circulating immunocompetent cell subsets also were not affected substantially (data not shown).
Open-heart surgery. During open heart surgery, the serum cortisol level increased slightly during the initial stage as the subject's body was cooled; however, it was almost the same as the level observed at the beginning of anesthesia (Figure 2). A sig25O
200
4O
~,D 150
30
~
35
.=
Immunocompetent cell subpopulations The percentages of total T-lymphocytes (CD4 + CD8) and B-cells (CD22) of blood were determined from monoclonal antibodies (from Sorbent, St. Petersburg, Russia) and flow cytometry as described elsewhere. 9 Measurements were made in total lymphoid cell population obtained after Ficoll-hypaque (Sigma Chemical Co., St. Louis, MO, USA) separation.
~,
100
"~ ~
£
25
N
20
"~ o
..~
8 5o o 0
I l
I 2
I 3
I 4
t
1D
rO
Steps o f Operation
Cortisol assay Cortisol in serum was measured by the method of competitive protein binding, l° using human transcortin, cortisol (Sigma Chemical Co.), and [1,2,6,7-3H]cortisol (specific activity 90 Ci/ mmol; Isotop, St. Petersburg, Russia). Intra- and interassay coefficients of variation were less than 5% and 10%, respectively.
412
Steroids,
1996, vol. 61, July
Figure 1 Serum CBG (O) and cortisol (O) levels during closed heart surgery (closed mitral commissurotomy). Steps of operation: 1, before the beginning of anesthesia; 2, before skin incision; 3, before correction of the lesion; 4, the end of the surgery; 1D, 24 h after surgery. Each point represents the mean _+SEM of five subjects. *P< 0.05; * * * P < 0.001; compared with the starting point.
Corticosteroid and immune responses to surgery: Tinnikov et al. 50
~
250 35
L~T
~-
"~ 20(I
40
3o o~
¢J
o
25
/
. ~ 150
2o .~ w 15 ~
mo
.. 50 1= o
~
30
~ P~.
2o
10 10
~
0 ]
2
3
4
5
6
ID
3D
I 1
I 2
I 3
Serum CBG (e) and cortisol (©) levels during open heart surgery (open reconstruction of the mitral valve, in surface hypothermia). Steps of operation: 1, before the beginning of anesthesia; 2, body temperature 30°C; 3, body temperature 27.2°C, just before occlusion; 4, after resuscitation; 5, 30 rain of reperfusion, body temperature 28°C; 6, 60 min of reperfusion, body temperature 30°C; 1D, 24 h after surgery; 3D, 3 days after surgery. Each point represents the mean + SEM of 9-12 subjects. * P < 0.05; * * P < 0.01; * * * P < 0.001; compared with the starting point.
nificant increase in cortisol level was observed only at the end of the operation (60 rain after reperfusion) as well as 1 day after surgery. In three cases, patients were given a prednisolone injection (90-150 mg) after resuscitation (no postprednisolone data were included in the results presented in figures). Very high corticosteroid levels measured by the radioligand method were observed in these cases: 100, 110, and 150 btg/dL after 30 min of reperfusion and 60, 45, and 110 txg/dL after 60 min of reperfusion, respectively. CBG activity declined progressively during cooling and surgical procedures, with a significant sharp decline observed at the end of cooling (body temperature 27,2°C), reaching a minimum after resuscitation. One and three days after surgery, serum CBG remained at a lowered level. The course of circulating immunocompetent cell subsets during and after the surgery is shown in Figure 3. Since patterns for CD4 cells (T-helpers) and CD8 cells (Tsuppressors) were very similar, they were combined as total T-cells (CD4 + CD8). Compared with the starting point, the percentage of T-cells and B-cells declined considerably during the surgery. A significant drop in circulating T-cells occurred by the end of cooling. The percentage of T-cells returned to normal ranges 3 days after the operation. A significantly lowered percentage of B-cells was observed immediately after occlusion, after 30 min of reperfusion, and 3 days after surgery. Hemoglobin levels in blood during open heart surgery did not change during first steps of the operation and declined significantly after resuscitation (Figure 4).
Discussion The present study demonstrated substantial differences in corticosteroid responses to closed and open heart surgery, in accordance with previous observations of children with congenital heart defects.~ During closed heart surgery, an el-
I 5
I ¢i
I 1D
I 3D
Stepsof Operation
Steps of Operation
Figure 2
I 4
Figure 3
Circulating T-cell (CD4 + CD8) subset (e) and B-cell (CD22) subset (&) during open heart surgery. The steps of operation are the same as in Figure 2. *P< 0.05 compared with the starting point.
evation in cortisol levels occurred, with a peak (about 30 ixg/dL) at the end of the operation. CBG levels did not change in this case. Such a picture is typical of the pituitaryadrenal response to stress and agrees with the generally accepted concept of corticosteroid regulation: the key role is played by corticosteroids, namely by the free fraction, and the role of CBG is to provide the circulatory transport of bound (biologically inactive) hormones. A considerable and rather rapid decline in CBG activity during open heart surgery in adults as well as in children j suggests a more active role for this serum protein in the stress responses of the body. Furthermore, a course of serum cortisol by no means corresponded with the severe conditions of open heart surgery. This finding contrasts with the picture observed in children when serum cortisol was elevated to about 80 ~g/dL after resuscitation.~ It seems that available data on serum cortisol levels after cardiac arrest are contradictory. Murray t| observed extremely high cortisol levels (about 100 I~g/dL) immediately after cardiac arrest. On the other hand, there are some data providing evi140 I
- I ~ ~ 1 = _
~ _ ___~~
*
120
-3[
100 80
O =
60
40
20
0
1,
2,
3,
4,
5,
6,
Stepsof Operation Figure 4 Hemoglobin levels in blood during open heart surgery. The steps of the operation (1-6) are the same as in Figure 2. *P < 0.05 compared with the starting point.
S t e r o i d s , 1996, v o l . 61, J u l y
413
Papers dence tbr a rapid decline in serum cortisol after heart arrest, perhaps accounted for by an impairment of cortisol release from the adrenal cortex caused by global hypoxia. 12 Undoubtedly, cooling is a major contributing factor in the reactions of the body observed in this study. Swan et al. 13 reported that the pituitary-adrenal response to cardiovascular surgery under conditions of hypothermia was observed only during the beginning of hypothermia; hypothermia then suppressed further activation of the pituitary-adrenal system in spite of surgical injuries. It was also reported that secretion of corticosteroids in response to surgical stress was moderately elevated at body temperatures above 30°C, but was lowered at temperatures below 28°C. 14 Bernhard et al. 15 observed no increase in corticosteroid levels during surgery in humans and dogs in surface hypothermia. In light of the data cited, the lack of pronounced response of serum cortisol to open heart surgery does not seem too paradoxical. The striking difference between children I and adults in the response of cortisol to open heart surgery under conditions of hypothermia remains to be explained. Despite slight changes in serum cortisol, CBG response to open heart surgery in adults was the same as in children. This suggests that CBG may play a certain role, perhaps independent of corticosteroids, in responses of the body to such stress. In the case of inflammation, a drop in CBG activity was supposed to be a way of elevating the free fraction (biologically significant) of corticosteroids, the total level of which is actually very high during inflammation~'S; it is unclear, at the same time, whether the drop in CBG activity is of the same significance, considering the lack of a pronounced corticosteroid response. As estimated by the time course of hemoglobin, some degree of blood dilution occurred after resuscitation. No signs of dilution were seen before occlusion (cardiac arrest) when CBG levels were already significantly lowered. Blood dilution after resuscitation may contribute to the further drop in CBG activities observed after occlusion. However, on the whole, the maximum estimated blood dilution (a 12% decline in hemoglobin concentration) is much less than the maximum drop in corticosteroid-binding capacity (a decline of 37% from the initial level). Therefore, leaving aside the cause of the drop in CBG activities before occlusion, the drop in CBG activity during the surgery cannot be fully explained by blood dilution. Based on data available, some suppositions may be made about possible mechanisms of the drop in serum CBG activity. First, the drop may result from the uptake of circulating CBG by target cells. 2 Redistribution of CBG between blood and interstitial fluid may also account for physiological fluctuations in serum CBG levels. ~6 CBG is known to share a high degree of homology with serine protease inhibitors, 7 a superfamily of proteins that also includes proteins controlling blood pressure and coagulation. 17 It can be speculated that CBG may be mobilized and employed as a serine protease inhibitor to maintain homeostatic parameters during dramatic changes in the cardiovascular system that are related to open heart surgery. It should be noted that a significant lowering in CBG activity was observed as early as the end of cooling, before circulatory arrest. In this connection it should be noted that, in experiments with dogs, it
414
Steroids, 1996, vol. 61, July
was shown that a lowering of body temperature to 20°C was accompanied by lowering cardiac output by 80% and arterial pressure by 40%. ~8 Substantial functional shifts in the cardiovascular system may take place in patients undergoing surface-induced hypothermia. Stress-induced immunosuppression is usually associated with the action of corticosteroids. In the present study a "standard" stress reaction (closed heart surgery) had no effect on the parameters of the immune system under study. During open heart surgery, a considerable decrease in the percentage of circulating ICC subsets occurred against a background of only slight or even no elevation of serum cortisol. It is unclear whether suppression or activation is reflected by the observed course of circulating ICC when percentages of B-cells and T-cells changed in a similar manner during the operation. It seems that there is no general agreement on the immune response to surgical trauma. There are reports of suppression of humoral and cellular immunity ~9 as well as studies providing evidence for the lack of suppression 2° or even stimulation 2~ of immune response. It was recently demonstrated that ischemia and reperfusion caused injuries of the vascular bed, a key factor being the adherence and activation of leukocytes in the field of the endothelium. 22"23 The same phenomenon may be responsible for lowered circulating ICCs during open heart surgery accompanied by partial (due to hypothermic effects on cardiac function) and global ischemia followed by reperfusion. In this connection, the similarity between the course of CBG activity and that of ICC subsets is worth noting. A rather close relation of CBG to the immune system is demonstrated by lowered CBG activity during acute inflammation4; a probable cause of this response is cleavage of CBG by elastase (a serine protease) of activated leukocytes, s Activated leukocytes that adhere during ischemia-reperfusion also release proteases, which are a key factor responsible lbr injuries of the endothelium. 22'23 Thus, the decline in CBG activity during open heart surgery may be of the same nature as the similar decline during inflammation. It is of some interest that the application of exogenous serine protease inhibitors resulted in some protective effect during ischemia-reperfusion. 23 CBG may be an emergency endogenous inhibitor that counteracts proteases released by leukocytes in response to partial and global ischemia and subsequent reperfusion, thus playing a protective role in response of the body to open heart surgery. It should be noted that the role of CBG may be independent of corticosteroids.
Acknowledgments We thank Mrs. Valentina Maslova for her excellent technical assistance. The study was supported by grant 93-0406499 from the Russian Foundation for Fundamental Sciences.
References 1.
2.
Tinnikov AA, Legan MV, Pavlova IP, Litasova EE, Ivanova LN (1993). Serum corticosteroid-binding globulin levels in children undergoing heart surgery. Steroids 58:536-539. Rosner W (1990). The functions of corticosteroid-binding globulin
Corticosteroid and i m m u n e responses to surgery: Tinnikov et al.
3.
4.
5. 6.
7.
8.
9.
10.
11. 12.
and sex-hormone-binding globulin: recent advances. Endocr Rev 11:80-91. Tinnikov AA (1993). Corticosteroid-binding globulin levels in the rat serum under conditions of starvation and restriction of motions. Horm Metab Res 25:88-89. Vrancks R, Savu L, Maya M, Nunez EA (1987). ~x-Fetoprotein and transcortin behave as acute phase reactants in the maternal and fetal compartments of the inflammatory pregnant mouse. Endocrinology 120:1782-1789. Besedovsky H, Sorkin E (1977). Network of immune-neuroendocrine interactions. Clin Exp lmmunol 27:1-12. Blalock JE (1989). A molecular basis for bidirectional communication between the immune and neuroendocrine systems. Physiol Rev 69:1-32. Hammond GL, Smith CL, Goping IS, Underhill DA, Harley M J, Reventos J, Musto NA, Gunsalus GL, Bardin CW (1987). Primary structure of human corticosteroid binding globulin, deduced from hepatic and pulmonary cDNAs, exhibits homology with serine protease inhibitors. Proc Natl Acad &'i USA 84:5153-5157. Hammond GL, Smith CL, Underhill CM, Nguyen VTT (1990). Interaction between corticosteroid binding globulin and activated leukocytes in vitro. Biochem Biophys Res Commun 172:172-177. Fleisher TA, Marti GE, Hagengruber C (1988). Two-color flow cytometric analysis of monocyte depleted human lymphocyte subsets. Cytometrv 9:309-315. Murphy BE (1967). Some studies of protein binding of steroids and their application to the routine micro and ultramicro measurement o1' various steroids in body fluids by competitive protein-binding radioassay. J Clin Endocrinol Metab 27:973-990. Murray D (1967). Cortisol binding to plasma proteins in man in health, stress and at death. J Endocrinol 39:571-591. Lindner KH, Strohmenger HU, Ensinger H, Hetzel WD, Ahnefeld
13. 14.
15.
16.
17. 18.
19.
20. 2t.
22.
23.
FW, Georgieff M (1992). Stress hormone response during and after cardiopulmonary resuscitation. Anesthesiology 77:662-668. Swan H, Jenkins D, Helmreich ML (1957). The adrenal cortical response to surgery. Surge O" 42:202-217. Van Brunt EE, Ganong WF (1963). The effects of preanesthetic medication, anesthesia, and hypothermia on the endocrine response to injury. Anesthesiology 24:500-514. Bernhard WF, McMurrey JD, Ganong WF, Lenninan R (1956). The effect of hypothermia on the peripheral serum levels of free 17hydroxycorticosteroids in the dog and man. Ann Surg 143:210-215. Hsu BRS, Kuhn RW (1988). The role of the adrenal in generating the diurnal variation in circulating levels of corticostemid-binding globulin in the rat. Endocrinology 122:421-426. Carrel R, Travis J (1985). %-Antitripsin and the serpins: variation and concentration. Trends Biochem Sci 10:20-24. Riittenhouse EA, Mohri H, Dillard DH, Merendino KA (1974). Deep hypothermia in cardiovascular surgery. Ann Thorac Surg 17: 63-98. Cooper AJ, Irvine JM, Turnbull AR (1974). Depression of immunological responses due to surgery: a model in the guinea pig. bnmunology 27:393-399. Markley K, Smallman E, Evans G (1967). Antibody production in mice after thermal and tourniquet trauma. Surgery 61:896-9(13. Kinnaen P, Mahieu A, Van Geertruyden N (1978). Stimulation of antibody synthesis induced by surgical trauma in rats. Clin Exp hnmunol 32:243-252. Hemandes LA, Grisham MB, Twohig B, Arfors KE, Harlan JM, Granger DN (1987). Role of neutrophils in ischemia-reperfusioninduced microvascular injury. Am J Physiol 253:H699-H703. Zimmerman BJ, Granger DN (1990). Perfusion-induced leukocyte infiltration: role of elastase. Am J Phvsiol 259:H390 H394
Steroids, 1996, vol. 61, July
415