Research in
Res Exp Med (1992) 192:269-279
Experimental Medicine 9 Springer-Verlag 1992
Hemodynamic effects following intraperitoneal infusion of pancreatic ascites fluid B.Vollmar, H.Waldner, M. Vierl, T. Kerner, P. Lehnert, and L. Schweiberer Institute of Surgical Research and the Department of Surgery, KlinikumInnenstadt, Ludwig-Maximilians-Universitfit,Marchioninistrasse15, W-8000 M0nchen 70, Federal Republic of Germany Received December 6, 1991 / accepted March 17, 1992
Summary. Severe necrotizing pancreatitis is accompanied by release of hemorrhagic ascites fluid (HAF), which is thought to be related to the occurrence and frequency of cardiocirculatory and pulmonary failure as a consequence of acute pancreatitis. The purpose of this study was to evaluate the role of HAF due to these systemic complications. Experiments were performed in 25 pigs (mean b.wt. 22 + 1 kg) under general anesthesia and mechanical ventilation. The animals received 50 ml/kg b.wt. i.p. of either physiologic saline solution (control CO, n = 9) or hemorrhagic ascites fluid (HAF, n = 16). HAF was obtained from 16 pigs with pancreatitis induced by intraductal infusion of bile salt. Eight animals in the HAF group were pretreated with indomethacin (10mg/kg i.v. INDO/HAF). All animals were followed up for 6 h. Mean arterial pressure, cardiac output, and stroke volume fell significantly in the H A F ( - 2 5 % , - 2 7 % , -27%) and in the INDO/HAF groups ( - 2 4 % , - 2 0 % , -17%) as compared with controls ( - 6 % , - 6 % , - 6 % ) . Also, left ventricular end-diastolic pressure (LVEDP) decreased by 52% and 48% in both HAF recipient groups, whereas LVEDP was unchanged in the control group. Myocardial contractility (Vm,x) remained unaltered in all experimental groups. No significant differences in gas exchange and lung dry/wet weight ratio were observed. Lipase and PGI2 of the unpretreated HAF group rised to 203%and 198% in arterial blood at 6 h compared with unaltered levels in the control group. No increase of prostanoid concentrations was detected in the indomethacin-pretreated group, whereas lipase increase by a comparable extent as in the HAF group. We conclude that the early consequences of HAF are mainly characterized by systemic hypotension due to hypovolemia. Key words: Pancreatic ascites fluid - Hemodynamics - Hypovolemia Peritoneal lavage Correspondence to: B. Vollmar
270
Introduction H e m o r r h a g i c pancreatitis represents the most severe stage of acute pancreatitis. Cardiocirculatory and p u l m o n a r y dysfunction occurring within the first 2 weeks after the onset of severe acute pancreatitis are t h o u g h t to be due to the effects of liberated vasoactive and toxic substances [5, 6, 8]. In a previous study, we were able to show that large a m o u n t s of e n z y m e s and vasoactive mediators released f r o m the pancreas accumulate in h e m o r r h a g i c ascites fluid ( H A F ) [30]. These noxious substances are t h o u g h t to be absorbed into the general circulation, consequently affecting r e m o t e organs, such as lung and heart. Studies in patients [3-5, 10, 24] and animals [25, 28, 29] suggest a link b e t w e e n peritoneal liberation of these c o m p o u n d s and f r e q u e n c y of organ dysfunction in the early course of the disease. Several investigations indicate that peritoneal lavage of the accumulating ascites might reverse circulatory shock associated with h e m o r rhagic pancreatitis and r e d u c e d mortality [23, 25, 27]. We therefore hypothesized that pancreatitis-associated ascites fluid contributes to the systemic complications seen during pancreatitis. We investigated this hypothesis by examining i.p.-infused ascites fluid in a porcine model.
Materials and methods All experiments were done following approval bv the local animal research committee.
Induction of hemorrhagic pancreatitis and production of HAF For production of hemorrhagic ascites fluid, 16 pigs (German hybrid strain, mean b.wt. 30+2kg) were used. Following pretreatment with ketamine (10mg/kg), flunitrazepam (0.1 mg/kg), and atropine (0.5 mg), the animals underwent mechanical ventilation. For hemodynamic monitoring and for infusion of balanced electrolyte solution and drugs, catheters were placed in the abdominal aorta, vena cava inferior, and pulmonary artery. Under general anesthesia with the i.v. opioid piritramide (10mg/kg per h) and halothane in the inspired gas (0.3 vol. To), a laparotomy was performed. The pancreatic duct was cannulated extraduodenally. A retrograde infusion with pressure not exceeding 30 cm H~O into the pancreatic duct contained 1 ml/kg 5% sodium taurocholate solution. The HAF was collected at hourly intervals for 6h, pooled, and stored at -70~
In vivo porcine model To set up the model, 25 pigs (German hybrid strain, mean b.wt. 22 +_ 1 kg) were pretreated by i.m. injection of ketamine (10mg/kg), flunitrazepam (1 mg/kg), and atropine (0.5 rag). After tracheotomy and intubation, the animals were mechanically ventilated with a mixture of O~ and N;O at a rate of 18 breaths/min. The end-expiratory CO2 concentration was continuously monitored and the tidal volume of the respirator was set to maintain end-expiratory CO2 at 4.5vol %. Anesthesia was maintained by i.v. infusion of piritramide (10mg/kg per h) and halothane in the inspired gas (0.3 vol. To). Monitoring of hemodynamics and blood sampling for biochemical and blood gas analyses required two large-bore catheters advanced through the femoral vessels into the abdominal aorta for measurement of mean arterial pressure and for blood sampling and into the vena cava inferior for infusion of drugs and of isotonic electrolyte solution. Two thermistor-equipped, flow-directed catheters (5F and 7F), advanced through the jugular veins into the pulmonary artery, were used for determination of pulmonary artery pressure, pulmonary capillary wedge pressure, and cardiac output (5F) and for taking blood samples
271 (7F). Cardiac output was determined in triplicate by injection of 5 ml ice-cold saline (CO-Computer, Mansfield, MSS, USA). A catheter-tip manometer (Millar Instruments, Houston, Tex., USA) was placed in the left ventricle through the right common carotid artery. This high-fidelity pressure transducer was used to determine left ventricular enddiastolic pressure and to calculate Vmax as a parameter of myocardial contractility. The catheter was calibrated immediately before insertion according to the manufacturer's instructions. The abdomen was then entered through an upper midline incision. A large-bore polyethylene catheter was placed in the bursa omentalis for infusion of H A F . A urinary catheter, advanced by direct puncture of the bladder, allowed hourly urine collection. The abdomen was then closed and a period of 60 min was allowed to elapse for hemodynamic stabilization. Isotonic electrolyte solution was administered at 10 ml/kg per h during the time required for preparation and the subsequent 6-h observation period. Mean arterial pressure (MAP), pulmonary artery pressure (PAP), pulmonary capillary wedge pressure (PCWP), heart rate (HR), cardiac output (CO), and left ventricular enddiastolic pressure (LVEDP) were measured. The following parameters were calculated using standard formula: stroke volume (SV), pulmonary vascular resistance (PVR), systemic vascular resistance (SVR), and alveolo-arterial oxygen difference (AaDO2). Arterial and mixed venous blood-gas tensions were analyzed using an ABL 300 (Radiometer, Copenhagen, Denmark).
Determination of myocardial contractility: The maximal contractile element velocity of shortening (Vmax) was determined as the measure of left ventricular contractility. Left ventricular pressure, derived from the catheter-tip manometer, was digitized in intervals of 4 ms and fed into a PDP 11/34 computer. Data sampling was performed during all recording times for at least 1 rain. The drift-corrected raw values were then analyzed offline using specifically designed software. The theoretical background for using V .... as a parameter for left ventricular contractility and the method of its determination have been previously described elsewhere [9].
Measurement of extravascular lung water: Gravimetric determination of the extravascular lung water (EVLW) was performed at the end of each experiment using the method of Pearce et al. [20] modified by Gray et al. [13]. Gravimetric E V L W was equal to the weight of the blood-free wet lung minus the blood-free dry weight. Following halothane overdose and KC1 infusion, the lungs were immediately removed through a thoracotomy. The lungs were weighed and then placed in a blender with a known amount of water. The lungs were completely homogenized. Samples of the homogenate were dried in an oven at 110~ together with a sample of blood withdrawn prior to KC1 infusion. Samples of the homogenized lungs were also placed in a highspeed centrifuge for 30 min. Lung homogenate supernatants and the blood sample were assayed for hemoglobin content and specific gravity. Biochemical analyses: Parameters measured were activity of lipase by the method of Weber [31] and concentrations of arachidonic acid metabolites prostacyclin (PGI2) and thromboxane As (TXA2). Radioimmunoassays (Pasteur Diagnostics, Paris, France) were used for all. Because PGI2 and TXA2 are chemically unstable, they were measured by their stable degradation products, 6-keto-PGFl~ and thromboxane B2, respectively. The sensitivities of our assays are 20pg/ml for TXB2 and 50pg/ml for 6-keto-PGFl~. Cross-reactivities of the antibodies as indicated by the manufacturer were < 10% for the tested metabolites. In the range of 200-2000 pg TXB2 or 6-keto-PGF~ per ml incubation mixture, the coefficient of variation for quadruplicates was 0.5-5%. For interassay precision, the coefficient of variation calculated from six consecutive assays was < 10%.
Experimental groups and experimental protocol Animals were randomly assigned to three experimental groups. Group 1 served as the control group and consisted of nine pigs. These animals received 50 ml/kg b.wt. of 0.9% NaC1 warmed to 37~ i.p. over i h. The animals in group II, which consisted of eight pigs, received 50ml/kg b.wt. of H A F warmed to 37~ i.p. over 1 h (HAF). The eight animals in group III were pretreated with indomethacin (10mg/kg) to block cyclooxygenase completely as soon as the cen-
272 tral venous catheter had been placed (INDO/HAF). The experimental protocol was otherwise the same as in group II. After the end of surgical preparation, closure of the abdomen, and stabilization period, the settings for anesthesia, mechanical ventilation, and volume substitution were not changed for the whole observation time. Hemodynamic recordings and withdrawal of arterial blood samples were performed before (time 0, baseline), immediately, and at hourly intervals for 6 h after HAF infusion.
Statistical analyses The results are expressed as means + SEM. Comparisons between groups were performed by the Kruskal-Wallis analysis followed by the Wilcoxon U-test. Results within groups were analyzed by Friedman rank analysis of variance followed bv Wilcoxon and Wilcox multiple comparisons. For inner group comparison, only data obtained at time 0 and at 6 h were considered. A probability level exceeding 95% was considered statistically significant.
Results Within 15 min after intraductal infusion of bile-salt solution in the donor animals, the pancreas became edematous and areas of hemorrhage and necroses developed within 30 min. The volume of ascites fluid collected over 6 h was in the range of 850-1400ml. Ascites fluid had a mean osmolality of 300 + 5 mosmol/1, m e a n lipase levels of 1074 +_ 129 U/l, mean values of 6-keto-PGF1,, and TXB2 of 8822 + 1104 pg/ml and 1099 + 101 pg/ml, respectively. Results for systemic and pulmonary hemodynamics, myocardial contractility, and blood gases are summarized in Tables 1 and 2. Serum lipase activities and plasma concentrations of 6-keto-PGFl~ and TXB2 are illustrated in Fig. 1.
Hernodynarnic variables and myocardial contractility In the control group, h e m o d y n a m i c parameters did not exhibit significant differences during the whole observation time. M A P , CO, and SV decreased slightly, by about 6%. H R remained unchanged at a mean value of 110 + 5 beats/min. L V E D P did not change with an average value of 6 + 0.5 m m H g (Table 1). Infusion of H A F (group II) caused a continuous and significant fall of m e a n arterial pressure from 83 + 5 m m H g at baseline to 77 + 5 m m Hg after 3 h and to 63 + 4 m m H g after 6h. This decrease of M A P in the H A F group was accompanied by a significant drop of cardiac output by 17.5% after 3 h and 35% after 6h. Owing to nearly unchanged H R s during the observation period, SV decreased significantly from 31 + 4 m l at baseline to 23 _+ 4 m l at the end of the experiment. L V E D P fell within 6h by 67%, from 6 + 0 . 4 to 2 + 0 . 6 m m H g (Table 1). Pretreatment with indomethacin (group III) did not prevent the systemic h e m o d y n a m i c effects observed in the H A F group. The fall in M A P following H A F infusion was comparable to that observed in the untreated animals. C O and SV decreased significantly by 12% within 6 h. H R ranged between 103 +_ 6 and 118 + 14 beats/min. L V E D P fell significantly by 45% (Table 1). M e a n pulmonary artery pressure remained nearly unchanged in the three experimental groups. Pulmonary vascular resistance increased slightly over time in the groups receiving H A F and in the control group, with no statistically significant differences between or within groups (Table 2).
273
+1 +1 +1 +1 +1 +1
+1 +1 +1 +1 +1 +1 +1 +1 +1
+1 +1 +1 0
+1 +1 +1
+1 +1 +1
+1 +1 +1
+1 +1 +]
+1 +1 +1
+1 +1 +1 +1 +1 +1
+1 +1 +1
+1 +t +1
+t +1 +1
~q
0
0
~
0
0
~
+1 +1 +1 +1 +1 +1
,,-:.C
~q
d B. eq
-FI +1 +1 +1
+1 +1 +1
+1 +1 +1
+I +1 +1 +l +1 +1 +1 +1 +1
~
eq
~ ~.,-.~
.~= -u +1 +1 +l
+1 +1 +1
+1 +1 +1
+1 +1 +1 +1 +1 +l
+1 +1 +1
+1 +1 +1
+1 +1 +1 +1 +1 +1 +1 +1 +1
+1 +I +1
+1 +1 +1
m ~
~ 2 e
~
~
d d d
d d d
+1 +1 +1
+1 +1 +1
+1 +1 +1
+1 +1 +1 +l +1 +1 +1 +1 +1
Cq
9
9 Intraperitoneal H A F led to reductions in arterial blood pressure, cardiac CO, and SV. 9 Myocardial contractility remained unchanged. The reductions in CO and SV, accompanied by a significant decrease in L V E D P , indicate reduced ventricular filling. The effects on cardiovascular function are therefore dominated by the effects due to hypovolemia. 9 There were no marked changes in pulmonary hemodynamics. Pulmonary gas e x c h a n g e and E V L W as markers of lung edema were not attenuated within the observation period of 6 h. 9 Inhibition of the cycloxygenase pathway failed to prevent the circulatory effects seen after i.p. H A F infusion. Hypovolemia and shock have long been known to be life-threatening complications in the early stage of severe necrotizing pancreatitis. The early stage is characterized by release of mediators and vasoactive substances into ascites fluid
277 and blood with consecutive cardiocirculatory and other organ complications as a result of inflammation and the necrotizing process. The vasoactive substances may incite ascites formation, cardiovascular depression, and decrease of plasma volume. In the present study, i.p.-infused ascites fluid led to a decrease in MAP CO, and SV. Coincidence of unchanged myocardial contractility and significantly reduced LVEDP confirmed a decreased cardiac filling due to hypovolemia. Accordingly, a study by Altimari et al. [1] demonstrates that the effects on cardiovascular system during pancreatitis are related to reduced cardiac filling rather than to decreased contractility. Using systolic time intervals for measurement of cardiac function, R~m6 et al. [22] could not find any evidence for alteration of the contactile state during pancreatitis in dogs. In a comparable experimental setting, R/~m6 et al. [21] showed that after induction of pancreatitis constant and significant decreases in SV, CO, LVEDV, and end-diastolic pressure developed, whereas parameters of left ventricular contractility were not affected to the same extent. In accordance with the present study. R~im6 et al. [21] suggested that circulatory failure was due primarily to a prompt reduction in preload and not caused by loss of contractility. Goldfarb et al. [11] even reported slightly improved ventricular contractility assessed by the systolic pressurediameter ratio at 4 h following experimental pancreatitis. In the present study, influences of inadequate volume status can be excluded, because of the strict regimen of volume substitution in all animals. Thus, hypovolemia with consecutively altered loading conditions rather than decreased intrinsic myocardial contractility is responsible for the hemodynamic effects following i.p. infusion of HAF. Patients with severe pancreatitis have a high-cardiac-output, low-resistance picture resembling the hemodynamic pattern seen in sepsis [6]. Changes in cardiac function are not due to morphological alterations, but probably associated with circulating compounds, which have a profound impact on the cardiovascular system, or are caused by other mechanisms, e.g. metabolic abnormalities, resulting in loss of vascular reactivity and opening of arteriovenous shunts [4, 6, 8]. Simultaneously with the hyperdynamic circulatory status, which is related to the necrotizing course of acute pancreatitis, patients regularly show signs of hypovolemia (i.e., lowered pressures in the right and left atria) in the initial phase of acute pancreatitis [4]. H A F given i.v. to various animals, including dogs, pigs, and rats [2, 19, 29] has been shown to cause transient hypotension. At present, one study is available where H A F was given i.p. to assess the toxicity of H A F by survival rates in rats [15]. In accordance with our results, the authors reported a fall in arterial pressure following i.p. infusion of HAF. Kiviniemi et al. [14] showed that a fall in arterial blood pressure during acute pancreatitis in dogs was associated with a rise of plasma 6-keto-PGFl~. Pretreatment with the prostaglandin synthesis inhibitor ibuprofen prevented the fall in blood pressure. The authors further reported unchanged serum thromboxane B2 levels. The conclusion reached in this study was that prostacyclin at least mediates the initial hypotension in hemorrhagic pancreatitis in dogs, whereas thromboxane production has a negligible role in the development of hemodynamic changes. In the present study, thromboxane levels remained nearly unchanged, but there was no correlation between the rise of 6-keto-PGFl~ and the fall in blood pressure in the present study. Concomitantly, blockade of PGI2 by indomethacin treatment did not alter the hemodynamic effects caused by i.p. HAF. Thus, our conclusion is that
278 prostanoids do not causally contribute to mediation of h e m o d v n a m i c effects following i.p. infusion of ascites fluid. P u l m o n a r y h e m o d y n a m i c s and gas exchange r e m a i n e d nearly u n c h a n g e d . Also, dry/wet weight ratios did not differ b e t w e e n the control and the two experimental groups. Dry/wet weight ratio was used as a measure of lung e d e m a because it is t h o u g h t to be accurate and sensitive as a m e a s u r e m e n t of e d e m a [26]. T h e r e are several studies reporting m a r k e d changes of p u l m o n a r y h e m o dynamics and permeability during acute pancreatitis. G o u l b o u r n e et al. [12] r e p o r t e d that wet lung weight was significantly increased after 18 h pancreatitis in rats. Burnweit and H o r t o n [7] s h o w e d a significant increase of p u l m o n a r y artery pressure and an increase of lung water, but without i m p a i r m e n t of pulmonary gas exchange during pancreatitis in dogs. Lungerella et al. [18] observed a significant difference in dry/wet weight ratio as early as l h after pancreatitis. This discrepancy f r o m our results might be due to differences in species and in experimental setup, i.e. the failure of induction of pancreatitis in the recipient of ascites and therefore in the severity of the condition in the m o d e l e m p l o y e d . In s u m m a r y , we f o u n d that h e m o r r h a g i c ascites fluid mediates the hypotensive state seen during the early course of pancreatitis. This strongly underlines the pivotal role of ascites fluid with regard to the onset and the clinical course of pancreatitis. F r o m this point of view, r e m o v i n g ascites fluid by therapeutic peritoneal lavage seems a logical and promising concept.
References 1. Altimari AF, Prinz RA, Leutz DW, Sandberg L, Kober PM, Raymond RM (1986) Myocardial depression during acute pancreatitis: Fact or fiction? Surgery 100:724-731 2. Amundson E, Ofstad E, Hagen PO (1968) Experimental acute pancreatitis in dogs: I. Hypotensive effect induced by pancreatic exudate. Scand J GastroenteroI 3 : 659-664 3. Basran GS, Ramasubramanian R, Verma R (1987) Intrathoracic complications of acute pancreatitis. Br J Dis Chest 81:326-331 4. Beger HG, Bittner R, Bfichler M, Hess W, Schmitz JE (1986) Hemodynamic data pattern in patients with acute pancreatitis. Gastroenterology 90 : 74-79 5. Beger HG, Bfichler M, Bittner R, Block S. Nevalainen T, Roscher R (1988) Necrosectomy and postoperative local lavage in necrotizing pancreatitis. Br J Surg 75 : 207-212 6. Bradley EL, Hall JR, Lutz J, Hamner L, Lattouf O (1983) Hemodynamic consequences of severe pancreatitis. Ann Surg 198 : 130-133 7. Burnweit CA, Horton JW (1988) Extravascular lung water as an indicator of pulmonary dysfunction in acute hemorrhagic pancreatitis. Ann Surg 207 : 33-38 8. Cobo LC, Abraham E, Bland RD (1984) Sequential hemodynamic and oxygen transport abnormalities in patients with acute pancreatitis. Surgery 95 : 324-330 9. Conzen PF, Hobbhahn J, Goetz AE, Habazettl H, Granetzny T, Peter K, Brendel W (1989) Myocardial contractility, blood flow, and oxygen consumption in healthy dogs during anesthesia with isoflurane of enflurane. J Cardiothor Anesth 3 : 70-77 10. Geokas MC, Rinderknecht H, Brodrich JW, Largman C (1978) Studies on the ascites fluid of acute pancreatitis in man. Am J Dig Dis 23 : 182-188 11. Goldfarb RD, Tambolini W, Nightingale L, Lefkowitz M, Kish P, Loegering D J, Weber PB (1985) Canine left ventricular function during experimental pancreatitis. J Surg Res 38 : 125-133 12. Goulbourne JA, Watson H, Davies GC (1987) lll-In-platelet and 125-1-flbrinogen deposition in the lungs in experimental acute pancreatitis. J Surg Res 43 : 521-526 13. Gray BA, Beckett RC, Allison RC, McCaffree DR, Smith RM, Sivak ED, Carlile JR (1984) Effect of edema and hemodynamic changes on extravascular thermal volume of the lung. J Appl Physioi Respir Environ Exercise Physiol 56 : 878-886
279 14. Kiviniemi H, Ram6 J, Stahlberg M, Laitinen S, Jalowaara P, Viinikka L, Kairaluoma M (1987) Prostacyclin and thromboxane in acute hemorrhagic pancreatitis in dogs. J Surg Res 42 : 232-236 15. Lankisch PG, Koop H, Winckler K, Schmidt H (1979) Continuous peritoneal dialysis as treatment of acute experimental pancreatitis in the rat: II. Analysis of its beneficial effect. Dig Dis Sci 24:117-122 16. Lee CT, Fein AM, Lippmann M, Holtzman H, Kimbel P, Weinbaum G (1981) Elastolytic activity in pulmonary lavage fluid from patients with adult respiratory distress syndrome. N Engl J Med 304:192-196 17. Lee WK, Frasca M, Lee C, Haider B, Regan T (1981) Depression of myocardial function during acute pancreatitis. Circ Shock 8 : 369-374 18. Lungarella G, Gardi C, Margherita de Santi M, Luzi P (1985) Pulmonary vascular injury in pancreatitis: evidence for a major role played by pancreatic elastase. Exp Mol Pathol 42 : 44-59 19. Ofstad E, Amundsen E, Hagen PO (1969) Experimental acute pancreatitis in dogs: II. Histamine release induced by pancreatic exudate. Scand J Gastroenterol 4 : 75-79 20. Pearce ML, Yamashita J, Beazell J (1965) Measurement of pulmonary edema. Circ Res 16 : 482-491 21. R~im60J, R~im6 P, Karkunen M, Kiviniemi H, Kettunen R (1992) Impaired venous return causes circulatory failure in experimental pancreatic shock in dogs. Intensive Care Med 15:111-115 22. R~im6 P, Karhunen M, R~im60J (1989) Systolic time intervals in canine experimental acute hemorrhagic pancreatitis. J Surg Res 46 : 212-215 23. Ranson JHC, Spencer FC (1978) The role of peritoneal lavage in severe acute pancreatitis. Ann Surg 187 : 565-575 24. Ranson JHC, Turner JW, Roses DF, Rifkind KM, Spencer FC (1974) Respiratory complications in acute pancreatitis. Ann Surg 179 : 557-566 25. Rosato EF, Mullis WF, Rosato FE (1973) Peritoneal lavage therapy in hemorrhagic pancreatitis. Surgery 74 : 106-115 26. Staub N (1974) Pulmonary edema. Physiol Rev 54 : 678-807 27. Stone HH, Fabian TC (1980) Peritoneal dialysis in the treatment of acute alcoholic pancreatitis. Surgery 74:106-115 28. Takada Y, Appert HE, Howard JM (1976) Vascular permeability induced by pancreatic exudate formed during acute pancreatitis. Surg Gynecol Obstet 143 : 779-783 29. Traverso LW, Pullos TG, Frey CF (1983) Hemodynamic characterization of porcine hemorrhagic pancreatitis ascites fluid. J Surg Res 34 : 254-262 30. Vollmar B, Waldner H, Schmand J, Conzen P, Goetz AE, Habazettl H, Schweiberer L, Brendel W (1989) Release of arachidonic acid metabolites during acute pancreafitis in pigs. Scand J Gastroenterol 24:1253-1264 31. Weber H (1965) Mikromethode zur Bestimmung der Pankreaslipase im Serum. Dtsch Med Wochenschr 26:1170-1176
Res Exp Med (1992) 192:269-279
Experimental Medicine 9 Springer-Verlag 1992
Hemodynamic effects following intraperitoneal infusion of pancreatic ascites fluid B.Vollmar, H.Waldner, M. Vierl, T. Kerner, P. Lehnert, and L. Schweiberer Institute of Surgical Research and the Department of Surgery, KlinikumInnenstadt, Ludwig-Maximilians-Universitfit,Marchioninistrasse15, W-8000 M0nchen 70, Federal Republic of Germany Received December 6, 1991 / accepted March 17, 1992
Summary. Severe necrotizing pancreatitis is accompanied by release of hemorrhagic ascites fluid (HAF), which is thought to be related to the occurrence and frequency of cardiocirculatory and pulmonary failure as a consequence of acute pancreatitis. The purpose of this study was to evaluate the role of HAF due to these systemic complications. Experiments were performed in 25 pigs (mean b.wt. 22 + 1 kg) under general anesthesia and mechanical ventilation. The animals received 50 ml/kg b.wt. i.p. of either physiologic saline solution (control CO, n = 9) or hemorrhagic ascites fluid (HAF, n = 16). HAF was obtained from 16 pigs with pancreatitis induced by intraductal infusion of bile salt. Eight animals in the HAF group were pretreated with indomethacin (10mg/kg i.v. INDO/HAF). All animals were followed up for 6 h. Mean arterial pressure, cardiac output, and stroke volume fell significantly in the H A F ( - 2 5 % , - 2 7 % , -27%) and in the INDO/HAF groups ( - 2 4 % , - 2 0 % , -17%) as compared with controls ( - 6 % , - 6 % , - 6 % ) . Also, left ventricular end-diastolic pressure (LVEDP) decreased by 52% and 48% in both HAF recipient groups, whereas LVEDP was unchanged in the control group. Myocardial contractility (Vm,x) remained unaltered in all experimental groups. No significant differences in gas exchange and lung dry/wet weight ratio were observed. Lipase and PGI2 of the unpretreated HAF group rised to 203%and 198% in arterial blood at 6 h compared with unaltered levels in the control group. No increase of prostanoid concentrations was detected in the indomethacin-pretreated group, whereas lipase increase by a comparable extent as in the HAF group. We conclude that the early consequences of HAF are mainly characterized by systemic hypotension due to hypovolemia. Key words: Pancreatic ascites fluid - Hemodynamics - Hypovolemia Peritoneal lavage Correspondence to: B. Vollmar
270
Introduction H e m o r r h a g i c pancreatitis represents the most severe stage of acute pancreatitis. Cardiocirculatory and p u l m o n a r y dysfunction occurring within the first 2 weeks after the onset of severe acute pancreatitis are t h o u g h t to be due to the effects of liberated vasoactive and toxic substances [5, 6, 8]. In a previous study, we were able to show that large a m o u n t s of e n z y m e s and vasoactive mediators released f r o m the pancreas accumulate in h e m o r r h a g i c ascites fluid ( H A F ) [30]. These noxious substances are t h o u g h t to be absorbed into the general circulation, consequently affecting r e m o t e organs, such as lung and heart. Studies in patients [3-5, 10, 24] and animals [25, 28, 29] suggest a link b e t w e e n peritoneal liberation of these c o m p o u n d s and f r e q u e n c y of organ dysfunction in the early course of the disease. Several investigations indicate that peritoneal lavage of the accumulating ascites might reverse circulatory shock associated with h e m o r rhagic pancreatitis and r e d u c e d mortality [23, 25, 27]. We therefore hypothesized that pancreatitis-associated ascites fluid contributes to the systemic complications seen during pancreatitis. We investigated this hypothesis by examining i.p.-infused ascites fluid in a porcine model.
Materials and methods All experiments were done following approval bv the local animal research committee.
Induction of hemorrhagic pancreatitis and production of HAF For production of hemorrhagic ascites fluid, 16 pigs (German hybrid strain, mean b.wt. 30+2kg) were used. Following pretreatment with ketamine (10mg/kg), flunitrazepam (0.1 mg/kg), and atropine (0.5 mg), the animals underwent mechanical ventilation. For hemodynamic monitoring and for infusion of balanced electrolyte solution and drugs, catheters were placed in the abdominal aorta, vena cava inferior, and pulmonary artery. Under general anesthesia with the i.v. opioid piritramide (10mg/kg per h) and halothane in the inspired gas (0.3 vol. To), a laparotomy was performed. The pancreatic duct was cannulated extraduodenally. A retrograde infusion with pressure not exceeding 30 cm H~O into the pancreatic duct contained 1 ml/kg 5% sodium taurocholate solution. The HAF was collected at hourly intervals for 6h, pooled, and stored at -70~
In vivo porcine model To set up the model, 25 pigs (German hybrid strain, mean b.wt. 22 +_ 1 kg) were pretreated by i.m. injection of ketamine (10mg/kg), flunitrazepam (1 mg/kg), and atropine (0.5 rag). After tracheotomy and intubation, the animals were mechanically ventilated with a mixture of O~ and N;O at a rate of 18 breaths/min. The end-expiratory CO2 concentration was continuously monitored and the tidal volume of the respirator was set to maintain end-expiratory CO2 at 4.5vol %. Anesthesia was maintained by i.v. infusion of piritramide (10mg/kg per h) and halothane in the inspired gas (0.3 vol. To). Monitoring of hemodynamics and blood sampling for biochemical and blood gas analyses required two large-bore catheters advanced through the femoral vessels into the abdominal aorta for measurement of mean arterial pressure and for blood sampling and into the vena cava inferior for infusion of drugs and of isotonic electrolyte solution. Two thermistor-equipped, flow-directed catheters (5F and 7F), advanced through the jugular veins into the pulmonary artery, were used for determination of pulmonary artery pressure, pulmonary capillary wedge pressure, and cardiac output (5F) and for taking blood samples
271 (7F). Cardiac output was determined in triplicate by injection of 5 ml ice-cold saline (CO-Computer, Mansfield, MSS, USA). A catheter-tip manometer (Millar Instruments, Houston, Tex., USA) was placed in the left ventricle through the right common carotid artery. This high-fidelity pressure transducer was used to determine left ventricular enddiastolic pressure and to calculate Vmax as a parameter of myocardial contractility. The catheter was calibrated immediately before insertion according to the manufacturer's instructions. The abdomen was then entered through an upper midline incision. A large-bore polyethylene catheter was placed in the bursa omentalis for infusion of H A F . A urinary catheter, advanced by direct puncture of the bladder, allowed hourly urine collection. The abdomen was then closed and a period of 60 min was allowed to elapse for hemodynamic stabilization. Isotonic electrolyte solution was administered at 10 ml/kg per h during the time required for preparation and the subsequent 6-h observation period. Mean arterial pressure (MAP), pulmonary artery pressure (PAP), pulmonary capillary wedge pressure (PCWP), heart rate (HR), cardiac output (CO), and left ventricular enddiastolic pressure (LVEDP) were measured. The following parameters were calculated using standard formula: stroke volume (SV), pulmonary vascular resistance (PVR), systemic vascular resistance (SVR), and alveolo-arterial oxygen difference (AaDO2). Arterial and mixed venous blood-gas tensions were analyzed using an ABL 300 (Radiometer, Copenhagen, Denmark).
Determination of myocardial contractility: The maximal contractile element velocity of shortening (Vmax) was determined as the measure of left ventricular contractility. Left ventricular pressure, derived from the catheter-tip manometer, was digitized in intervals of 4 ms and fed into a PDP 11/34 computer. Data sampling was performed during all recording times for at least 1 rain. The drift-corrected raw values were then analyzed offline using specifically designed software. The theoretical background for using V .... as a parameter for left ventricular contractility and the method of its determination have been previously described elsewhere [9].
Measurement of extravascular lung water: Gravimetric determination of the extravascular lung water (EVLW) was performed at the end of each experiment using the method of Pearce et al. [20] modified by Gray et al. [13]. Gravimetric E V L W was equal to the weight of the blood-free wet lung minus the blood-free dry weight. Following halothane overdose and KC1 infusion, the lungs were immediately removed through a thoracotomy. The lungs were weighed and then placed in a blender with a known amount of water. The lungs were completely homogenized. Samples of the homogenate were dried in an oven at 110~ together with a sample of blood withdrawn prior to KC1 infusion. Samples of the homogenized lungs were also placed in a highspeed centrifuge for 30 min. Lung homogenate supernatants and the blood sample were assayed for hemoglobin content and specific gravity. Biochemical analyses: Parameters measured were activity of lipase by the method of Weber [31] and concentrations of arachidonic acid metabolites prostacyclin (PGI2) and thromboxane As (TXA2). Radioimmunoassays (Pasteur Diagnostics, Paris, France) were used for all. Because PGI2 and TXA2 are chemically unstable, they were measured by their stable degradation products, 6-keto-PGFl~ and thromboxane B2, respectively. The sensitivities of our assays are 20pg/ml for TXB2 and 50pg/ml for 6-keto-PGFl~. Cross-reactivities of the antibodies as indicated by the manufacturer were < 10% for the tested metabolites. In the range of 200-2000 pg TXB2 or 6-keto-PGF~ per ml incubation mixture, the coefficient of variation for quadruplicates was 0.5-5%. For interassay precision, the coefficient of variation calculated from six consecutive assays was < 10%.
Experimental groups and experimental protocol Animals were randomly assigned to three experimental groups. Group 1 served as the control group and consisted of nine pigs. These animals received 50 ml/kg b.wt. of 0.9% NaC1 warmed to 37~ i.p. over i h. The animals in group II, which consisted of eight pigs, received 50ml/kg b.wt. of H A F warmed to 37~ i.p. over 1 h (HAF). The eight animals in group III were pretreated with indomethacin (10mg/kg) to block cyclooxygenase completely as soon as the cen-
272 tral venous catheter had been placed (INDO/HAF). The experimental protocol was otherwise the same as in group II. After the end of surgical preparation, closure of the abdomen, and stabilization period, the settings for anesthesia, mechanical ventilation, and volume substitution were not changed for the whole observation time. Hemodynamic recordings and withdrawal of arterial blood samples were performed before (time 0, baseline), immediately, and at hourly intervals for 6 h after HAF infusion.
Statistical analyses The results are expressed as means + SEM. Comparisons between groups were performed by the Kruskal-Wallis analysis followed by the Wilcoxon U-test. Results within groups were analyzed by Friedman rank analysis of variance followed bv Wilcoxon and Wilcox multiple comparisons. For inner group comparison, only data obtained at time 0 and at 6 h were considered. A probability level exceeding 95% was considered statistically significant.
Results Within 15 min after intraductal infusion of bile-salt solution in the donor animals, the pancreas became edematous and areas of hemorrhage and necroses developed within 30 min. The volume of ascites fluid collected over 6 h was in the range of 850-1400ml. Ascites fluid had a mean osmolality of 300 + 5 mosmol/1, m e a n lipase levels of 1074 +_ 129 U/l, mean values of 6-keto-PGF1,, and TXB2 of 8822 + 1104 pg/ml and 1099 + 101 pg/ml, respectively. Results for systemic and pulmonary hemodynamics, myocardial contractility, and blood gases are summarized in Tables 1 and 2. Serum lipase activities and plasma concentrations of 6-keto-PGFl~ and TXB2 are illustrated in Fig. 1.
Hernodynarnic variables and myocardial contractility In the control group, h e m o d y n a m i c parameters did not exhibit significant differences during the whole observation time. M A P , CO, and SV decreased slightly, by about 6%. H R remained unchanged at a mean value of 110 + 5 beats/min. L V E D P did not change with an average value of 6 + 0.5 m m H g (Table 1). Infusion of H A F (group II) caused a continuous and significant fall of m e a n arterial pressure from 83 + 5 m m H g at baseline to 77 + 5 m m Hg after 3 h and to 63 + 4 m m H g after 6h. This decrease of M A P in the H A F group was accompanied by a significant drop of cardiac output by 17.5% after 3 h and 35% after 6h. Owing to nearly unchanged H R s during the observation period, SV decreased significantly from 31 + 4 m l at baseline to 23 _+ 4 m l at the end of the experiment. L V E D P fell within 6h by 67%, from 6 + 0 . 4 to 2 + 0 . 6 m m H g (Table 1). Pretreatment with indomethacin (group III) did not prevent the systemic h e m o d y n a m i c effects observed in the H A F group. The fall in M A P following H A F infusion was comparable to that observed in the untreated animals. C O and SV decreased significantly by 12% within 6 h. H R ranged between 103 +_ 6 and 118 + 14 beats/min. L V E D P fell significantly by 45% (Table 1). M e a n pulmonary artery pressure remained nearly unchanged in the three experimental groups. Pulmonary vascular resistance increased slightly over time in the groups receiving H A F and in the control group, with no statistically significant differences between or within groups (Table 2).
273
+1 +1 +1 +1 +1 +1
+1 +1 +1 +1 +1 +1 +1 +1 +1
+1 +1 +1 0
+1 +1 +1
+1 +1 +1
+1 +1 +1
+1 +1 +]
+1 +1 +1
+1 +1 +1 +1 +1 +1
+1 +1 +1
+1 +t +1
+t +1 +1
~q
0
0
~
0
0
~
+1 +1 +1 +1 +1 +1
,,-:.C
~q
d B. eq
-FI +1 +1 +1
+1 +1 +1
+1 +1 +1
+I +1 +1 +l +1 +1 +1 +1 +1
~
eq
~ ~.,-.~
.~= -u +1 +1 +l
+1 +1 +1
+1 +1 +1
+1 +1 +1 +1 +1 +l
+1 +1 +1
+1 +1 +1
+1 +1 +1 +1 +1 +1 +1 +1 +1
+1 +I +1
+1 +1 +1
m ~
~ 2 e
~
~
d d d
d d d
+1 +1 +1
+1 +1 +1
+1 +1 +1
+1 +1 +1 +l +1 +1 +1 +1 +1
Cq
9
9 Intraperitoneal H A F led to reductions in arterial blood pressure, cardiac CO, and SV. 9 Myocardial contractility remained unchanged. The reductions in CO and SV, accompanied by a significant decrease in L V E D P , indicate reduced ventricular filling. The effects on cardiovascular function are therefore dominated by the effects due to hypovolemia. 9 There were no marked changes in pulmonary hemodynamics. Pulmonary gas e x c h a n g e and E V L W as markers of lung edema were not attenuated within the observation period of 6 h. 9 Inhibition of the cycloxygenase pathway failed to prevent the circulatory effects seen after i.p. H A F infusion. Hypovolemia and shock have long been known to be life-threatening complications in the early stage of severe necrotizing pancreatitis. The early stage is characterized by release of mediators and vasoactive substances into ascites fluid
277 and blood with consecutive cardiocirculatory and other organ complications as a result of inflammation and the necrotizing process. The vasoactive substances may incite ascites formation, cardiovascular depression, and decrease of plasma volume. In the present study, i.p.-infused ascites fluid led to a decrease in MAP CO, and SV. Coincidence of unchanged myocardial contractility and significantly reduced LVEDP confirmed a decreased cardiac filling due to hypovolemia. Accordingly, a study by Altimari et al. [1] demonstrates that the effects on cardiovascular system during pancreatitis are related to reduced cardiac filling rather than to decreased contractility. Using systolic time intervals for measurement of cardiac function, R~m6 et al. [22] could not find any evidence for alteration of the contactile state during pancreatitis in dogs. In a comparable experimental setting, R/~m6 et al. [21] showed that after induction of pancreatitis constant and significant decreases in SV, CO, LVEDV, and end-diastolic pressure developed, whereas parameters of left ventricular contractility were not affected to the same extent. In accordance with the present study. R~im6 et al. [21] suggested that circulatory failure was due primarily to a prompt reduction in preload and not caused by loss of contractility. Goldfarb et al. [11] even reported slightly improved ventricular contractility assessed by the systolic pressurediameter ratio at 4 h following experimental pancreatitis. In the present study, influences of inadequate volume status can be excluded, because of the strict regimen of volume substitution in all animals. Thus, hypovolemia with consecutively altered loading conditions rather than decreased intrinsic myocardial contractility is responsible for the hemodynamic effects following i.p. infusion of HAF. Patients with severe pancreatitis have a high-cardiac-output, low-resistance picture resembling the hemodynamic pattern seen in sepsis [6]. Changes in cardiac function are not due to morphological alterations, but probably associated with circulating compounds, which have a profound impact on the cardiovascular system, or are caused by other mechanisms, e.g. metabolic abnormalities, resulting in loss of vascular reactivity and opening of arteriovenous shunts [4, 6, 8]. Simultaneously with the hyperdynamic circulatory status, which is related to the necrotizing course of acute pancreatitis, patients regularly show signs of hypovolemia (i.e., lowered pressures in the right and left atria) in the initial phase of acute pancreatitis [4]. H A F given i.v. to various animals, including dogs, pigs, and rats [2, 19, 29] has been shown to cause transient hypotension. At present, one study is available where H A F was given i.p. to assess the toxicity of H A F by survival rates in rats [15]. In accordance with our results, the authors reported a fall in arterial pressure following i.p. infusion of HAF. Kiviniemi et al. [14] showed that a fall in arterial blood pressure during acute pancreatitis in dogs was associated with a rise of plasma 6-keto-PGFl~. Pretreatment with the prostaglandin synthesis inhibitor ibuprofen prevented the fall in blood pressure. The authors further reported unchanged serum thromboxane B2 levels. The conclusion reached in this study was that prostacyclin at least mediates the initial hypotension in hemorrhagic pancreatitis in dogs, whereas thromboxane production has a negligible role in the development of hemodynamic changes. In the present study, thromboxane levels remained nearly unchanged, but there was no correlation between the rise of 6-keto-PGFl~ and the fall in blood pressure in the present study. Concomitantly, blockade of PGI2 by indomethacin treatment did not alter the hemodynamic effects caused by i.p. HAF. Thus, our conclusion is that
278 prostanoids do not causally contribute to mediation of h e m o d v n a m i c effects following i.p. infusion of ascites fluid. P u l m o n a r y h e m o d y n a m i c s and gas exchange r e m a i n e d nearly u n c h a n g e d . Also, dry/wet weight ratios did not differ b e t w e e n the control and the two experimental groups. Dry/wet weight ratio was used as a measure of lung e d e m a because it is t h o u g h t to be accurate and sensitive as a m e a s u r e m e n t of e d e m a [26]. T h e r e are several studies reporting m a r k e d changes of p u l m o n a r y h e m o dynamics and permeability during acute pancreatitis. G o u l b o u r n e et al. [12] r e p o r t e d that wet lung weight was significantly increased after 18 h pancreatitis in rats. Burnweit and H o r t o n [7] s h o w e d a significant increase of p u l m o n a r y artery pressure and an increase of lung water, but without i m p a i r m e n t of pulmonary gas exchange during pancreatitis in dogs. Lungerella et al. [18] observed a significant difference in dry/wet weight ratio as early as l h after pancreatitis. This discrepancy f r o m our results might be due to differences in species and in experimental setup, i.e. the failure of induction of pancreatitis in the recipient of ascites and therefore in the severity of the condition in the m o d e l e m p l o y e d . In s u m m a r y , we f o u n d that h e m o r r h a g i c ascites fluid mediates the hypotensive state seen during the early course of pancreatitis. This strongly underlines the pivotal role of ascites fluid with regard to the onset and the clinical course of pancreatitis. F r o m this point of view, r e m o v i n g ascites fluid by therapeutic peritoneal lavage seems a logical and promising concept.
References 1. Altimari AF, Prinz RA, Leutz DW, Sandberg L, Kober PM, Raymond RM (1986) Myocardial depression during acute pancreatitis: Fact or fiction? Surgery 100:724-731 2. Amundson E, Ofstad E, Hagen PO (1968) Experimental acute pancreatitis in dogs: I. Hypotensive effect induced by pancreatic exudate. Scand J GastroenteroI 3 : 659-664 3. Basran GS, Ramasubramanian R, Verma R (1987) Intrathoracic complications of acute pancreatitis. Br J Dis Chest 81:326-331 4. Beger HG, Bittner R, Bfichler M, Hess W, Schmitz JE (1986) Hemodynamic data pattern in patients with acute pancreatitis. Gastroenterology 90 : 74-79 5. Beger HG, Bfichler M, Bittner R, Block S. Nevalainen T, Roscher R (1988) Necrosectomy and postoperative local lavage in necrotizing pancreatitis. Br J Surg 75 : 207-212 6. Bradley EL, Hall JR, Lutz J, Hamner L, Lattouf O (1983) Hemodynamic consequences of severe pancreatitis. Ann Surg 198 : 130-133 7. Burnweit CA, Horton JW (1988) Extravascular lung water as an indicator of pulmonary dysfunction in acute hemorrhagic pancreatitis. Ann Surg 207 : 33-38 8. Cobo LC, Abraham E, Bland RD (1984) Sequential hemodynamic and oxygen transport abnormalities in patients with acute pancreatitis. Surgery 95 : 324-330 9. Conzen PF, Hobbhahn J, Goetz AE, Habazettl H, Granetzny T, Peter K, Brendel W (1989) Myocardial contractility, blood flow, and oxygen consumption in healthy dogs during anesthesia with isoflurane of enflurane. J Cardiothor Anesth 3 : 70-77 10. Geokas MC, Rinderknecht H, Brodrich JW, Largman C (1978) Studies on the ascites fluid of acute pancreatitis in man. Am J Dig Dis 23 : 182-188 11. Goldfarb RD, Tambolini W, Nightingale L, Lefkowitz M, Kish P, Loegering D J, Weber PB (1985) Canine left ventricular function during experimental pancreatitis. J Surg Res 38 : 125-133 12. Goulbourne JA, Watson H, Davies GC (1987) lll-In-platelet and 125-1-flbrinogen deposition in the lungs in experimental acute pancreatitis. J Surg Res 43 : 521-526 13. Gray BA, Beckett RC, Allison RC, McCaffree DR, Smith RM, Sivak ED, Carlile JR (1984) Effect of edema and hemodynamic changes on extravascular thermal volume of the lung. J Appl Physioi Respir Environ Exercise Physiol 56 : 878-886
279 14. Kiviniemi H, Ram6 J, Stahlberg M, Laitinen S, Jalowaara P, Viinikka L, Kairaluoma M (1987) Prostacyclin and thromboxane in acute hemorrhagic pancreatitis in dogs. J Surg Res 42 : 232-236 15. Lankisch PG, Koop H, Winckler K, Schmidt H (1979) Continuous peritoneal dialysis as treatment of acute experimental pancreatitis in the rat: II. Analysis of its beneficial effect. Dig Dis Sci 24:117-122 16. Lee CT, Fein AM, Lippmann M, Holtzman H, Kimbel P, Weinbaum G (1981) Elastolytic activity in pulmonary lavage fluid from patients with adult respiratory distress syndrome. N Engl J Med 304:192-196 17. Lee WK, Frasca M, Lee C, Haider B, Regan T (1981) Depression of myocardial function during acute pancreatitis. Circ Shock 8 : 369-374 18. Lungarella G, Gardi C, Margherita de Santi M, Luzi P (1985) Pulmonary vascular injury in pancreatitis: evidence for a major role played by pancreatic elastase. Exp Mol Pathol 42 : 44-59 19. Ofstad E, Amundsen E, Hagen PO (1969) Experimental acute pancreatitis in dogs: II. Histamine release induced by pancreatic exudate. Scand J Gastroenterol 4 : 75-79 20. Pearce ML, Yamashita J, Beazell J (1965) Measurement of pulmonary edema. Circ Res 16 : 482-491 21. R~im60J, R~im6 P, Karkunen M, Kiviniemi H, Kettunen R (1992) Impaired venous return causes circulatory failure in experimental pancreatic shock in dogs. Intensive Care Med 15:111-115 22. R~im6 P, Karhunen M, R~im60J (1989) Systolic time intervals in canine experimental acute hemorrhagic pancreatitis. J Surg Res 46 : 212-215 23. Ranson JHC, Spencer FC (1978) The role of peritoneal lavage in severe acute pancreatitis. Ann Surg 187 : 565-575 24. Ranson JHC, Turner JW, Roses DF, Rifkind KM, Spencer FC (1974) Respiratory complications in acute pancreatitis. Ann Surg 179 : 557-566 25. Rosato EF, Mullis WF, Rosato FE (1973) Peritoneal lavage therapy in hemorrhagic pancreatitis. Surgery 74 : 106-115 26. Staub N (1974) Pulmonary edema. Physiol Rev 54 : 678-807 27. Stone HH, Fabian TC (1980) Peritoneal dialysis in the treatment of acute alcoholic pancreatitis. Surgery 74:106-115 28. Takada Y, Appert HE, Howard JM (1976) Vascular permeability induced by pancreatic exudate formed during acute pancreatitis. Surg Gynecol Obstet 143 : 779-783 29. Traverso LW, Pullos TG, Frey CF (1983) Hemodynamic characterization of porcine hemorrhagic pancreatitis ascites fluid. J Surg Res 34 : 254-262 30. Vollmar B, Waldner H, Schmand J, Conzen P, Goetz AE, Habazettl H, Schweiberer L, Brendel W (1989) Release of arachidonic acid metabolites during acute pancreafitis in pigs. Scand J Gastroenterol 24:1253-1264 31. Weber H (1965) Mikromethode zur Bestimmung der Pankreaslipase im Serum. Dtsch Med Wochenschr 26:1170-1176
