- Email: [email protected]
The effect of modified ultrafiltration on the amount of circulating endotoxins in children undergoing cardiopulmonary bypass
The Effect of Modified Ultrafiltration on the Amount of Circulating Endotoxins in Children Undergoing Cardiopulmonary Bypass Stig Yndgaard, MD, Lars W. Andersen, MD, Claus Andersen, MD, Go¨sta Petterson, MD, and Leif Baek, MD Objective: To determine whether the use of modified ultrafiltration during pediatric cardiopulmonary bypass (CPB) diminishes the load of circulating endotoxins. Design: Single-arm prospective observational study. Setting: A university hospital operating room and intensive care unit. Participants: Twenty children undergoing CPB for correction of various congenital heart diseases. Interventions: The amount of endotoxins in plasma was measured during CPB and before and after modified ultrafiltration. The ultrafiltrate was assayed for the presence of endotoxins. Postoperatively, the children were followed with relevant infectious parameters and cultures. Measurements and Main Results: The amount of endotoxins increased significantly during the CPB procedure (from a
median of 1.3 ng [range, 0 to 13.7 ng] to 24.2 ng [range, 2.1 to 75.9 ng]). After termination of CPB, modified ultrafiltration was shown to lower the amount of circulating endotoxins in blood (from a median of 24.2 ng [range, 2.1 to 75.4 ng] to 9.0 [range, 0.1 to 40.6 ng]). The major bulk of this reduction in endotoxin load was retrieved in the ultrafiltrate (median of 11.9 ng [range, 0 to 12.1 ng]). Conclusion: This study strongly suggests that modified ultrafiltration decreases the amount of circulating endotoxins in children undergoing cardiac surgery. Copyright r 2000 by W.B. Saunders Company
C
intubation, an arterial catheter was inserted in the femoral artery together with a central venous catheter in the internal jugular vein. Hypothermic, nonpulsatile CPB using membrane oxygenators was performed in all children. Pump flow at 2.4 L/min/m2 was maintained at all temperatures. Unacceptable perfusion pressures (ie, mean perfusion pressure ⬍30 mmHg) were treated with bolus doses of phenylepinephrine. Antegrade blood cardioplegia was used in all children. CPB was weaned with no inotropes or small-to-moderate doses (2 to 6 µg/kg/min) of dopamine. Nitroglycerin was used in all children in the rewarming phase to obtain a uniform rewarming of the body. CPB was terminated at a peripheral temperature of greater than 31°C and a rectal temperature of 36°C. After termination of CPB, MUF9,10 was initiated. The aortic cannula was left in place, and blood was siphoned from the aorta, pumped through a dialysis filter, and returned to the right atrium. An Amicon dialyzing filter (Sorin Biomedical, Glostrup, Denmark), with cutoff at 17,000 d, was used for children with body weights less than 6 kg. A Gambro dialyzing filter (Jydekrogen 8, Vallensbaek, Denmark), with cutoff at 52,000 d, was used in children weighing more than 6 kg. Volume was infused from the reservoir as necessary to maintain hemodynamic stability. The MUF was ceased at a hematocrit of 35% to 40%, which was obtained within 10 minutes in all patients. Optimal hemodynamics during the MUF procedure were achieved by transfusion of blood from the heart-lung machine to the right atrium. After termination of MUF, transfusion was taken over by the anesthesiologist. CPB and MUF produce large volume shifts, and the measured endotoxin concentrations must be related to the actual blood volume of the child. The actual blood volume at the different time points was estimated on weight, priming volume, and amount of ultrafiltrate. Consequently, actual endotoxin concentrations were recalculated at the different time points to a whole body endotoxin load. The endotoxin load was calculated as the concentration load (ng/mL) ⫻ the calculated actual blood volume (mL).
ARDIOPULMONARY BYPASS (CPB) is associated with endotoxemia in adults and children undergoing cardiac surgery.1-3 The authors have previously shown that endotoxemia coincides with splanchnic hypoperfusion, together with reperfusion injuries caused by the CPB procedure.4,5 Under normal conditions, endotoxins are contained within the intestinal lumen by an intact mucosal barrier. Hypoperfusion and reperfusion injury, however, may produce a leaky intestinal mucosa with subsequent translocation of endotoxins to the portal and systemic circulation. Endotoxins have numerous biologic activities and aggravate the systemic inflammatory response created by the nonbiologic surfaces of the extracorporeal circuit.6-8 Circulating endotoxins clinically may cause hypotension, fever, hypermetabolism, tissue damage, and coagulopathies. Various degrees of endotoxemia during and after cardiac surgery may be an important pathogenetic factor leading to increased morbidity and mortality in some children. It is important to find a way by which the endotoxin load can be reduced to improve outcome after cardiac surgery. Previously, it has been shown that the use of modified ultrafiltration (MUF) after termination of CPB in children is associated with significant improvements in cardiac performance.9,10 A decreased endotoxin load, together with decreased myocardial edema, may explain the improved cardiac performance observed in pediatric cardiac surgical patients after MUF. This study investigated whether MUF removed circulating endotoxins from the circulation in children undergoing cardiac surgery. METHODS The authors investigated 20 consecutive children with different congenital heart diseases (Table 1) scheduled for cardiac surgery. Ethics committee approval and written informed parental consent were obtained. Patients older than 3 months of age were premedicated with rectal diazepam and morphine/scopolamine 45 minutes before operation. Patients younger than 3 months were not premedicated. Anesthesia was induced with midazolam, 100 µg/kg; fentanyl, 10 µg/kg; and pancuronium, 200 µg/kg. Nasal intubation was performed in all cases. Anesthesia was maintained with fentanyl bolus doses to a total dose of 30 to 40 µg/kg, supplemented with 0.4% isoflurane in oxygen and air. Routine antibiotics (cefuroxime, 30 mg/kg, 3 times a day for 2 days) were administered from induction of anesthesia. All patients received methylprednisolone, 30 mg/kg, just after induction of anesthesia. After
KEY WORDS: children, cardiopulmonary bypass, endotoxins, modified ultrafiltration
From the Department of Cardiothoracic Anesthesia and Cardiac Surgery, The Heart Center, University of Copenhagen, Rigshospitalet, Copenhagen; and Department of Clinical Microbiology, University of Copenhagen, Herlev Hospital, Herlev, Denmark. Address reprint requests to Stig Yndgaard, MD, Department of Cardio Thoracic Anesthesia (4142), The Heart Center, Rigshospitalet, 2100 Copenhagen, Denmark. Copyright r 2000 by W.B. Saunders Company 1053-0770/00/1404-0008$10.00/0 doi:10.1053/jcan.2000.7944
Journal of Cardiothoracic and Vascular Anesthesia, Vol 14, No 4 (August), 2000: pp 399-401
399
400
YNDGAARD ET AL
Table 1. Patient Characteristics
Diagnosis
Operation
Abnormal coronary artery Aortic stenosis AV-canal AV-canal AV-canal CCTGA CCTGA DORV Fallot Fallot Fallot Fallot Sing vent Sing vent Sing vent Sing vent TGA TGA VSD VSD
Reconstruction Shunt AV-patch AV-patch AV-patch Mitral prosthesis Mitral ⫹ tric prosthesis TCPC Total correction Total correction Total correction Total correction Glenn Septum resection Coarctation-resection TCPC Switch Switch Patch Patch
CPB Clamping Time Weight Time (min) (min) (kg)
5.2 3.67 4.52 5.4 4.1 8.8 15.4 18 7.6 10.8 10.3 6.8 9.6 16.5 3.69 15.8 3.9 3.9 15 6.8
193 146 126 203 93 58 136 96 138 162 157 135 117 54 114 179 200 67 71 74
33 42 75 109 59 43 90 52 79 113 88 95 0 26 55 60 116 45 41 38
Abbreviations: CPB, cardiopulmonary bypass; AV, atrioventricular; TGA, transposition of the great arteries; VSD, ventriculoseptal defect. CCTGA, congenit corrected transposition of the great arteries; DORV, double outlet right ventricle; tric, tricuspid; TCPC, total cava pulmonary connection.
Blood samples for endotoxin measurements were drawn at the following time points: after induction of anesthesia (P1), 1 minute after the start of CPB (P2), at target temperature (P3), at the start of rewarming (P4), after aortic cross-clamp release (P5), and at the end of CPB before MUF. MUF was performed after all procedures, once the patients were safely weaned from bypass, and the amount of endotoxins in blood after MUF was measured as well (P6). The amount of endotoxins in the ultrafiltrate (P7) was measured. Postoperatively, endotoxin levels, C-reactive protein levels, leukocyte counts, and temperature were followed daily until discharge from the intensive care unit (ICU). The endotoxin concentrations in blood and ultrafiltrate were measured with the Limulus amebocyte lysate test combined with a rocket immunoelectrophoretic assay. The sensitivity of the test is 2 pg/mL.11,12 All results are given as median values and ranges. Friedmann’s test was followed by Wilcoxon nonparametric test for paired data. Regression analysis was used to test the correlation between endotoxin load and bypass and aortic cross-clamp time together with temperature rises postoperatively; p values ⬍0.05 were considered statistically significant.
during the bypass period and reached peak values at the end of the bypass period (24.2 ng [range, 2.1 to 75.9 ng]). The MUF procedure significantly decreased the endotoxin load from 24.2 ng (range, 2.1 to 75.4 ng) to 9.0 ng (range, 0.1 to 40.6 ng). The major part of this reduction in endotoxin load (11.9 ng [range, 0 to 12.1 ng]) was recovered in the ultrafiltrate. This group of patients is too small to allow prediction of any correlation between the type of congenital heart disease, repair, and endotoxin load. No correlation was found between endotoxin load and CPB and aortic cross-clamp time. DISCUSSION
During pediatric cardiac surgery, a consistently elevated endotoxin load was found. MUF performed after weaning from CPB reduced this load significantly, and the bulk of the load could be retrieved in the ultrafiltrate. CPB in children often results in increased capillary permeability and accumulation of excess total body water. In some children, this leads to multiple organ dysfunction—the so-called postperfusion syndrome. This syndrome may be caused partly by endotoxins and their activation of different potent inflammatory mediators. Other important contributing pathogenetic factors to the postperfusion syndrome are maturity, length of CPB, degrees of hypothermia, and contact activation. Besides contact activation, which is the generally accepted major cause of the CPB-induced systemic inflammatory response, the presence of circulating endotoxins is another major additional cause. The authors and other research groups worldwide in previous studies have shown that endotoxemia is a consistent phenomenon in adults and children undergoing cardiac surgery.4,5,13,14 The source of endotoxemia is primarily the gut.15 The CPB procedure causes suboptimal oxygen delivery to the gut partly because of the countercurrent exchange system in the villi together with reperfusion injuries of the mucosa. This suboptimal oxygen delivery causes increased
RESULTS
The average stay in ICU was 2 days (range, 1 to 17 days). Three children died within 30 days from unexpected cardiac arrest. None of the children developed postoperative signs of infections, according to clinical status, cultures, levels of C-reactive protein, and leukocyte counts. All children had varying temperature rises postoperatively, which could not be correlated to levels of circulating (not bound to effector cells) endotoxins. Endotoxin data were given as median load (range). The endotoxin loads at the different time points are shown in Fig 1. After induction of anesthesia, the endotoxin load was 1.3 ng (range, 0 to 13.7 ng). The endotoxin load increased significantly
Fig 1. Changes in total endotoxin amount before, during, and after cardiopulmonary bypass (CPB) and modified ultrafiltration shown as median and quartiles. ", ng endotoxin found in ultrafiltrate; 䊉, ng endotoxin in blood and quartiles. *p F 0.01, **p F 0.001.
MODIFIED ULTRAFILTRATION AND ENDOTOXEMIA
401
mucosal permeability, leading to translocation of endotoxins from the lumen of the gut to portal and systemic circulations. Environmental endotoxins from the priming fluids is another important source, according to investigations done by this group.1,3,4 The clearance capacity of the reticuloendothelial system, primarily the liver, is the limiting factor for spillover of endotoxins into the systemic circulation. The reticuloendothelial system may be variably compromised as a result of the CPB procedure. Endotoxins are potent inflammatory mediators of the immune system, which are unsubstituted, and weigh about 5000 d. They are often present in the circulation in aggregates with protein molecules. In cardiac surgery, anti-inflammatory therapy, such as corticosteroids, aprotinin, anticytokine antibodies, and circuit modifications, has shown beneficial effects when instituted well before the trauma (the CPB procedure). MUF is instituted long after the trauma has begun, however, which is one of the drawbacks of MUF from an immunologic point of view, because the release and the maximal deleterious effects of endotoxins, cytokines, complement factors, and other mediators may have taken place already. MUF has been shown to produce significant beneficial changes in hemodynamics at the crucial time of weaning the patient from bypass, however. The amount of pharmacologic inotropic support has decreased significantly in all groups of patients since the introduction of this procedure, according to the experience of this center and others. The physiologic benefits of MUF may be due to reductions in total body water and to an increase in hematocrit with a concomitant decrease in transfused blood products.10 The authors’ theory is that endotoxins cause some of the major problems observed after CPB in children (ie, increased
capillary leak, myocardial depression, and an aggravated systemic inflammatory response). Suffredini et al16 have shown in healthy volunteers that endotoxins in a dose of 4 ng/kg depress myocardial function directly, not related to changes in enddiastolic volume and vascular resistance. Several studies have explored the role of MUF in modifying the inflammatory response caused by the exposure of the blood to the nonendothelialized surfaces of the pump. El Habbal et al17 showed a decrease in interleukin-8 after MUF. Other authors have shown decreased amounts in tumor necrosis factor-␣, interleukin-6, interleukin-10, complement C3a, and complement C5a.18,19 The present study shows that MUF reduces the endotoxin load. The procedure does not eliminate endotoxins already bound to cells. Looking at the improved cardiac performance after pediatric CPB using MUF, however, it is plausible that the reduced endotoxin load may be beneficial for the child despite its ‘‘late and incomplete removal.’’ At this institution, MUF was introduced 3 years ago, and since then significantly fewer hemodynamic problems have been experienced after CPB in all categories of children, including those with the most complicated anatomic lesions and surgical corrections. Because of the good experience with MUF at this center and others, it was not acceptable to include a control group in this study. In conclusion, this study showed that MUF reduced the endotoxin load in children undergoing CPB. This reduction in endotoxins may modify the systemic inflammatory response and may decrease morbidity and mortality after pediatric cardiac surgery.
REFERENCES 1. Andersen LW, Larsen B, Baek L: Transient endotoxemia during pediatric surgery. J Cardiothorac Anesth 131:S6, 1990 2. Casey WF, Hauser GJ, Hannallah RS, et al: Circulating endotoxemia and tumor necrosis factor during pediatric cardiac surgery. Crit Care Med 20:1090-1096, 1992 3. Andersen LW, Baek L, Degn H, et al: Presence of circulating endotoxins during cardiac surgery. J Thorac Cardiovasc Surg 93:115119, 1987 4. McNicol L, Andersen LW, Liu G, et al: Markers of splanchnic perfusion and intestinal translocation of endotoxins during cardiopulmonary bypass: Effects of dopamine and milrinone. J Cardiothorac Vasc Anesth 13:292-298, 1999 5. Andersen LW, Landow L, Baek L, et al: Association between gastric intramucosal pH and splanchnic endotoxin, antibody to endotoxin, and tumor necrosis factor alpha concentrations in patients undergoing cardiopulmonary bypass. Crit Care Med 21:210-217, 1993 6. Ghosh S, Latimer RD, Gray BM, et al: Endotoxin-induced organ injury. Crit Care Med 21:S19-S24, 1993 7. Morrison DC, Ulevitch RJ: The effects of bacterial endotoxins on host mediation systems: A review. Am J Pathol 93:526-617, 1978 8. Andersen LW, Baek L: Transient endotoxemia during cardiopulmonary bypass. J Perfusion 1:36-41, 1992 9. Elliot MJ: Ultrafiltration and modified ultrafiltration in pediatric open heart operations. Ann Thorac Surg 56:1518-1522, 1993 10. Naik SK, Blaji S, Elliott MJ: Modified ultrafiltration improves haemodynamics after cardiopulmonary bypass in children. J Am Coll Cardiol 19:37, 1993 11. Levin J, Bang FB: The role of endotoxins in the extracellular
coagulation of Limulus blood. Bull John Hopkins Hosp 115:265-274, 1964 12. Baek L: New, sensitive rocket immunoelectrophoretic assay for measurement of the reaction between endotoxin and Limulus amoebocyte lysate. J Clin Microbiol 17:1013-1020, 1983 13. Ohri SK, Bjarnason I, Pathi V, et al: Cardiopulmonary bypass impairs small intestinal transport and increases gut permeability. Ann Thorac Surg 55:1080-1086, 1993 14. Ohri SK, Somasundaram S, Koak Y, et al: The effect of intestinal hypoperfusion on intestinal absorption and permeability during cardiopulmonary bypass. Gastroenterology 106:318-323, 1994 15. Martinez-Pellus AE, Merino P, Bru M, et al: Can selective digestive decontamination avoid the endotoxemia and cytokine activation promoted by cardiopulmonary bypass? Crit Care Med 21:16841691, 1993 16. Suffredini AF, Fromm RE, Parker MM, et al: The cardiovascular response of normal humans to the administration of endotoxin. N Engl J Med 3:280-287, 1989 17. El Habbal MH, Smith L, Strobel S, et al: Modified ultrafiltration after cardiopulmonary bypass for repair of ventricular septal defect reduces serum IL-8. Circulation 88:0505A, 1995 18. Montenegro LM, Greeley WJ: The use of modified ultrafiltration during pediatric cardiac surgery is a benefit. J Cardiothorac Vasc Anesth 12:480-482, 1998 19. Andreasson S, Gothberg S, Bergren H, et al: Hemofiltration modifies complement activation after extracorporeal circulation in infants. Ann Thorac Surg 56:1515-1517, 1993
median of 1.3 ng [range, 0 to 13.7 ng] to 24.2 ng [range, 2.1 to 75.9 ng]). After termination of CPB, modified ultrafiltration was shown to lower the amount of circulating endotoxins in blood (from a median of 24.2 ng [range, 2.1 to 75.4 ng] to 9.0 [range, 0.1 to 40.6 ng]). The major bulk of this reduction in endotoxin load was retrieved in the ultrafiltrate (median of 11.9 ng [range, 0 to 12.1 ng]). Conclusion: This study strongly suggests that modified ultrafiltration decreases the amount of circulating endotoxins in children undergoing cardiac surgery. Copyright r 2000 by W.B. Saunders Company
C
intubation, an arterial catheter was inserted in the femoral artery together with a central venous catheter in the internal jugular vein. Hypothermic, nonpulsatile CPB using membrane oxygenators was performed in all children. Pump flow at 2.4 L/min/m2 was maintained at all temperatures. Unacceptable perfusion pressures (ie, mean perfusion pressure ⬍30 mmHg) were treated with bolus doses of phenylepinephrine. Antegrade blood cardioplegia was used in all children. CPB was weaned with no inotropes or small-to-moderate doses (2 to 6 µg/kg/min) of dopamine. Nitroglycerin was used in all children in the rewarming phase to obtain a uniform rewarming of the body. CPB was terminated at a peripheral temperature of greater than 31°C and a rectal temperature of 36°C. After termination of CPB, MUF9,10 was initiated. The aortic cannula was left in place, and blood was siphoned from the aorta, pumped through a dialysis filter, and returned to the right atrium. An Amicon dialyzing filter (Sorin Biomedical, Glostrup, Denmark), with cutoff at 17,000 d, was used for children with body weights less than 6 kg. A Gambro dialyzing filter (Jydekrogen 8, Vallensbaek, Denmark), with cutoff at 52,000 d, was used in children weighing more than 6 kg. Volume was infused from the reservoir as necessary to maintain hemodynamic stability. The MUF was ceased at a hematocrit of 35% to 40%, which was obtained within 10 minutes in all patients. Optimal hemodynamics during the MUF procedure were achieved by transfusion of blood from the heart-lung machine to the right atrium. After termination of MUF, transfusion was taken over by the anesthesiologist. CPB and MUF produce large volume shifts, and the measured endotoxin concentrations must be related to the actual blood volume of the child. The actual blood volume at the different time points was estimated on weight, priming volume, and amount of ultrafiltrate. Consequently, actual endotoxin concentrations were recalculated at the different time points to a whole body endotoxin load. The endotoxin load was calculated as the concentration load (ng/mL) ⫻ the calculated actual blood volume (mL).
ARDIOPULMONARY BYPASS (CPB) is associated with endotoxemia in adults and children undergoing cardiac surgery.1-3 The authors have previously shown that endotoxemia coincides with splanchnic hypoperfusion, together with reperfusion injuries caused by the CPB procedure.4,5 Under normal conditions, endotoxins are contained within the intestinal lumen by an intact mucosal barrier. Hypoperfusion and reperfusion injury, however, may produce a leaky intestinal mucosa with subsequent translocation of endotoxins to the portal and systemic circulation. Endotoxins have numerous biologic activities and aggravate the systemic inflammatory response created by the nonbiologic surfaces of the extracorporeal circuit.6-8 Circulating endotoxins clinically may cause hypotension, fever, hypermetabolism, tissue damage, and coagulopathies. Various degrees of endotoxemia during and after cardiac surgery may be an important pathogenetic factor leading to increased morbidity and mortality in some children. It is important to find a way by which the endotoxin load can be reduced to improve outcome after cardiac surgery. Previously, it has been shown that the use of modified ultrafiltration (MUF) after termination of CPB in children is associated with significant improvements in cardiac performance.9,10 A decreased endotoxin load, together with decreased myocardial edema, may explain the improved cardiac performance observed in pediatric cardiac surgical patients after MUF. This study investigated whether MUF removed circulating endotoxins from the circulation in children undergoing cardiac surgery. METHODS The authors investigated 20 consecutive children with different congenital heart diseases (Table 1) scheduled for cardiac surgery. Ethics committee approval and written informed parental consent were obtained. Patients older than 3 months of age were premedicated with rectal diazepam and morphine/scopolamine 45 minutes before operation. Patients younger than 3 months were not premedicated. Anesthesia was induced with midazolam, 100 µg/kg; fentanyl, 10 µg/kg; and pancuronium, 200 µg/kg. Nasal intubation was performed in all cases. Anesthesia was maintained with fentanyl bolus doses to a total dose of 30 to 40 µg/kg, supplemented with 0.4% isoflurane in oxygen and air. Routine antibiotics (cefuroxime, 30 mg/kg, 3 times a day for 2 days) were administered from induction of anesthesia. All patients received methylprednisolone, 30 mg/kg, just after induction of anesthesia. After
KEY WORDS: children, cardiopulmonary bypass, endotoxins, modified ultrafiltration
From the Department of Cardiothoracic Anesthesia and Cardiac Surgery, The Heart Center, University of Copenhagen, Rigshospitalet, Copenhagen; and Department of Clinical Microbiology, University of Copenhagen, Herlev Hospital, Herlev, Denmark. Address reprint requests to Stig Yndgaard, MD, Department of Cardio Thoracic Anesthesia (4142), The Heart Center, Rigshospitalet, 2100 Copenhagen, Denmark. Copyright r 2000 by W.B. Saunders Company 1053-0770/00/1404-0008$10.00/0 doi:10.1053/jcan.2000.7944
Journal of Cardiothoracic and Vascular Anesthesia, Vol 14, No 4 (August), 2000: pp 399-401
399
400
YNDGAARD ET AL
Table 1. Patient Characteristics
Diagnosis
Operation
Abnormal coronary artery Aortic stenosis AV-canal AV-canal AV-canal CCTGA CCTGA DORV Fallot Fallot Fallot Fallot Sing vent Sing vent Sing vent Sing vent TGA TGA VSD VSD
Reconstruction Shunt AV-patch AV-patch AV-patch Mitral prosthesis Mitral ⫹ tric prosthesis TCPC Total correction Total correction Total correction Total correction Glenn Septum resection Coarctation-resection TCPC Switch Switch Patch Patch
CPB Clamping Time Weight Time (min) (min) (kg)
5.2 3.67 4.52 5.4 4.1 8.8 15.4 18 7.6 10.8 10.3 6.8 9.6 16.5 3.69 15.8 3.9 3.9 15 6.8
193 146 126 203 93 58 136 96 138 162 157 135 117 54 114 179 200 67 71 74
33 42 75 109 59 43 90 52 79 113 88 95 0 26 55 60 116 45 41 38
Abbreviations: CPB, cardiopulmonary bypass; AV, atrioventricular; TGA, transposition of the great arteries; VSD, ventriculoseptal defect. CCTGA, congenit corrected transposition of the great arteries; DORV, double outlet right ventricle; tric, tricuspid; TCPC, total cava pulmonary connection.
Blood samples for endotoxin measurements were drawn at the following time points: after induction of anesthesia (P1), 1 minute after the start of CPB (P2), at target temperature (P3), at the start of rewarming (P4), after aortic cross-clamp release (P5), and at the end of CPB before MUF. MUF was performed after all procedures, once the patients were safely weaned from bypass, and the amount of endotoxins in blood after MUF was measured as well (P6). The amount of endotoxins in the ultrafiltrate (P7) was measured. Postoperatively, endotoxin levels, C-reactive protein levels, leukocyte counts, and temperature were followed daily until discharge from the intensive care unit (ICU). The endotoxin concentrations in blood and ultrafiltrate were measured with the Limulus amebocyte lysate test combined with a rocket immunoelectrophoretic assay. The sensitivity of the test is 2 pg/mL.11,12 All results are given as median values and ranges. Friedmann’s test was followed by Wilcoxon nonparametric test for paired data. Regression analysis was used to test the correlation between endotoxin load and bypass and aortic cross-clamp time together with temperature rises postoperatively; p values ⬍0.05 were considered statistically significant.
during the bypass period and reached peak values at the end of the bypass period (24.2 ng [range, 2.1 to 75.9 ng]). The MUF procedure significantly decreased the endotoxin load from 24.2 ng (range, 2.1 to 75.4 ng) to 9.0 ng (range, 0.1 to 40.6 ng). The major part of this reduction in endotoxin load (11.9 ng [range, 0 to 12.1 ng]) was recovered in the ultrafiltrate. This group of patients is too small to allow prediction of any correlation between the type of congenital heart disease, repair, and endotoxin load. No correlation was found between endotoxin load and CPB and aortic cross-clamp time. DISCUSSION
During pediatric cardiac surgery, a consistently elevated endotoxin load was found. MUF performed after weaning from CPB reduced this load significantly, and the bulk of the load could be retrieved in the ultrafiltrate. CPB in children often results in increased capillary permeability and accumulation of excess total body water. In some children, this leads to multiple organ dysfunction—the so-called postperfusion syndrome. This syndrome may be caused partly by endotoxins and their activation of different potent inflammatory mediators. Other important contributing pathogenetic factors to the postperfusion syndrome are maturity, length of CPB, degrees of hypothermia, and contact activation. Besides contact activation, which is the generally accepted major cause of the CPB-induced systemic inflammatory response, the presence of circulating endotoxins is another major additional cause. The authors and other research groups worldwide in previous studies have shown that endotoxemia is a consistent phenomenon in adults and children undergoing cardiac surgery.4,5,13,14 The source of endotoxemia is primarily the gut.15 The CPB procedure causes suboptimal oxygen delivery to the gut partly because of the countercurrent exchange system in the villi together with reperfusion injuries of the mucosa. This suboptimal oxygen delivery causes increased
RESULTS
The average stay in ICU was 2 days (range, 1 to 17 days). Three children died within 30 days from unexpected cardiac arrest. None of the children developed postoperative signs of infections, according to clinical status, cultures, levels of C-reactive protein, and leukocyte counts. All children had varying temperature rises postoperatively, which could not be correlated to levels of circulating (not bound to effector cells) endotoxins. Endotoxin data were given as median load (range). The endotoxin loads at the different time points are shown in Fig 1. After induction of anesthesia, the endotoxin load was 1.3 ng (range, 0 to 13.7 ng). The endotoxin load increased significantly
Fig 1. Changes in total endotoxin amount before, during, and after cardiopulmonary bypass (CPB) and modified ultrafiltration shown as median and quartiles. ", ng endotoxin found in ultrafiltrate; 䊉, ng endotoxin in blood and quartiles. *p F 0.01, **p F 0.001.
MODIFIED ULTRAFILTRATION AND ENDOTOXEMIA
401
mucosal permeability, leading to translocation of endotoxins from the lumen of the gut to portal and systemic circulations. Environmental endotoxins from the priming fluids is another important source, according to investigations done by this group.1,3,4 The clearance capacity of the reticuloendothelial system, primarily the liver, is the limiting factor for spillover of endotoxins into the systemic circulation. The reticuloendothelial system may be variably compromised as a result of the CPB procedure. Endotoxins are potent inflammatory mediators of the immune system, which are unsubstituted, and weigh about 5000 d. They are often present in the circulation in aggregates with protein molecules. In cardiac surgery, anti-inflammatory therapy, such as corticosteroids, aprotinin, anticytokine antibodies, and circuit modifications, has shown beneficial effects when instituted well before the trauma (the CPB procedure). MUF is instituted long after the trauma has begun, however, which is one of the drawbacks of MUF from an immunologic point of view, because the release and the maximal deleterious effects of endotoxins, cytokines, complement factors, and other mediators may have taken place already. MUF has been shown to produce significant beneficial changes in hemodynamics at the crucial time of weaning the patient from bypass, however. The amount of pharmacologic inotropic support has decreased significantly in all groups of patients since the introduction of this procedure, according to the experience of this center and others. The physiologic benefits of MUF may be due to reductions in total body water and to an increase in hematocrit with a concomitant decrease in transfused blood products.10 The authors’ theory is that endotoxins cause some of the major problems observed after CPB in children (ie, increased
capillary leak, myocardial depression, and an aggravated systemic inflammatory response). Suffredini et al16 have shown in healthy volunteers that endotoxins in a dose of 4 ng/kg depress myocardial function directly, not related to changes in enddiastolic volume and vascular resistance. Several studies have explored the role of MUF in modifying the inflammatory response caused by the exposure of the blood to the nonendothelialized surfaces of the pump. El Habbal et al17 showed a decrease in interleukin-8 after MUF. Other authors have shown decreased amounts in tumor necrosis factor-␣, interleukin-6, interleukin-10, complement C3a, and complement C5a.18,19 The present study shows that MUF reduces the endotoxin load. The procedure does not eliminate endotoxins already bound to cells. Looking at the improved cardiac performance after pediatric CPB using MUF, however, it is plausible that the reduced endotoxin load may be beneficial for the child despite its ‘‘late and incomplete removal.’’ At this institution, MUF was introduced 3 years ago, and since then significantly fewer hemodynamic problems have been experienced after CPB in all categories of children, including those with the most complicated anatomic lesions and surgical corrections. Because of the good experience with MUF at this center and others, it was not acceptable to include a control group in this study. In conclusion, this study showed that MUF reduced the endotoxin load in children undergoing CPB. This reduction in endotoxins may modify the systemic inflammatory response and may decrease morbidity and mortality after pediatric cardiac surgery.
REFERENCES 1. Andersen LW, Larsen B, Baek L: Transient endotoxemia during pediatric surgery. J Cardiothorac Anesth 131:S6, 1990 2. Casey WF, Hauser GJ, Hannallah RS, et al: Circulating endotoxemia and tumor necrosis factor during pediatric cardiac surgery. Crit Care Med 20:1090-1096, 1992 3. Andersen LW, Baek L, Degn H, et al: Presence of circulating endotoxins during cardiac surgery. J Thorac Cardiovasc Surg 93:115119, 1987 4. McNicol L, Andersen LW, Liu G, et al: Markers of splanchnic perfusion and intestinal translocation of endotoxins during cardiopulmonary bypass: Effects of dopamine and milrinone. J Cardiothorac Vasc Anesth 13:292-298, 1999 5. Andersen LW, Landow L, Baek L, et al: Association between gastric intramucosal pH and splanchnic endotoxin, antibody to endotoxin, and tumor necrosis factor alpha concentrations in patients undergoing cardiopulmonary bypass. Crit Care Med 21:210-217, 1993 6. Ghosh S, Latimer RD, Gray BM, et al: Endotoxin-induced organ injury. Crit Care Med 21:S19-S24, 1993 7. Morrison DC, Ulevitch RJ: The effects of bacterial endotoxins on host mediation systems: A review. Am J Pathol 93:526-617, 1978 8. Andersen LW, Baek L: Transient endotoxemia during cardiopulmonary bypass. J Perfusion 1:36-41, 1992 9. Elliot MJ: Ultrafiltration and modified ultrafiltration in pediatric open heart operations. Ann Thorac Surg 56:1518-1522, 1993 10. Naik SK, Blaji S, Elliott MJ: Modified ultrafiltration improves haemodynamics after cardiopulmonary bypass in children. J Am Coll Cardiol 19:37, 1993 11. Levin J, Bang FB: The role of endotoxins in the extracellular
coagulation of Limulus blood. Bull John Hopkins Hosp 115:265-274, 1964 12. Baek L: New, sensitive rocket immunoelectrophoretic assay for measurement of the reaction between endotoxin and Limulus amoebocyte lysate. J Clin Microbiol 17:1013-1020, 1983 13. Ohri SK, Bjarnason I, Pathi V, et al: Cardiopulmonary bypass impairs small intestinal transport and increases gut permeability. Ann Thorac Surg 55:1080-1086, 1993 14. Ohri SK, Somasundaram S, Koak Y, et al: The effect of intestinal hypoperfusion on intestinal absorption and permeability during cardiopulmonary bypass. Gastroenterology 106:318-323, 1994 15. Martinez-Pellus AE, Merino P, Bru M, et al: Can selective digestive decontamination avoid the endotoxemia and cytokine activation promoted by cardiopulmonary bypass? Crit Care Med 21:16841691, 1993 16. Suffredini AF, Fromm RE, Parker MM, et al: The cardiovascular response of normal humans to the administration of endotoxin. N Engl J Med 3:280-287, 1989 17. El Habbal MH, Smith L, Strobel S, et al: Modified ultrafiltration after cardiopulmonary bypass for repair of ventricular septal defect reduces serum IL-8. Circulation 88:0505A, 1995 18. Montenegro LM, Greeley WJ: The use of modified ultrafiltration during pediatric cardiac surgery is a benefit. J Cardiothorac Vasc Anesth 12:480-482, 1998 19. Andreasson S, Gothberg S, Bergren H, et al: Hemofiltration modifies complement activation after extracorporeal circulation in infants. Ann Thorac Surg 56:1515-1517, 1993