PANCREATIC CELLULAR INJURY AFTER CARDIAC SURGERY WITH CARDIOPULMONARY BYPASS: FREQUENCY, TIME COURSE AND RISK FACTORS
Nys, Monique*†; Venneman, Ingrid*; Deby-Dupont, Ginette*†; Preiser, Jean-Charles*†; Vanbelle, Sophie‡; Albert, Adelin‡; Camus, Gérard†; Damas, Pierre*; Larbuisson, Robert*; Lamy, Maurice*†
*Departments of Anesthesia and Intensive Care Medicine, University Hospital of Liège; and †Center of Oxygen, Research and Development and ‡Department of Biostatistics, University of Liège, Liège, Belgium
Received 17 May 2006; first review completed 30 Jun 2006; accepted in final form 3 Oct 2006
Address reprint requests to Monique Nys, PhD, Department of Anesthesia and Intensive Care, CHU de Liège, B35, B4000, Sart Tilman, Liège, Belgium. E-mail: Monique.Nys@chu.ulg.ac.be.
Although often clinically silent, pancreatic cellular injury (PCI) is relatively frequent after cardiac surgery with cardiopulmonary bypass; and its etiology and time course are largely unknown. We defined PCI as the simultaneous presence of abnormal values of pancreatic isoamylase and immunoreactive trypsin (IRT). The frequency and time evolution of PCI were assessed in this condition using assays for specific exocrine pancreatic enzymes. Correlations with inflammatory markers were searched for preoperative risk factors. One hundred ninety-three patients submitted to cardiac surgery were enrolled prospectively. Blood IRT, amylase, pancreatic isoamylase, lipase, and markers of inflammation (α1-protease inhibitor, α2-macroglobulin, myeloperoxidase) were measured preoperatively and postoperatively until day 8. The postoperative increase in plasma levels of pancreatic enzymes and urinary IRT was biphasic in all patients: early after surgery and later (from day 4 to 8 after surgery). One hundred thirty-three patients (69%) experienced PCI, with mean IRT, isoamylase, and α1-protease inhibitor values higher for each sample than that in patients without PCI. By multiple regression analysis, we found preoperative values of plasma IRT ≥40 ng/mL, amylase ≥42 IU/mL, and pancreatic isoamylase ≥20 IU/L associated with a higher incidence of postsurgery PCI (P < 0.005). In the PCI patients, a significant correlation was found between the 4 pancreatic enzymes and urinary IRT, total calcium, myeloperoxidase, α1-protease inhibitor, and α2-macroglobulin. These data support a high prevalence of postoperative PCI after cardiac surgery with cardiopulmonary bypass, typically biphasic and clinically silent, especially when pancreatic enzymes were elevated preoperatively.
The presence of biological signs of pancreatic injury have been reported for several decades, occurring in patients after surgery including cardiac surgery with cardiopulmonary bypass (CPB) (1, 2) and in intensive care patients after clinical situations associated with splanchnic hypoperfusion and acute inflammation (3-9). Ischemia was involved to explain the pancreatic injury, as low cardiac output and indices of splanchnic ischemia including infarction of liver, spleen, or bowel were often found simultaneously (4, 10, 11). As reported in several animal models of hemorrhagic shock or arterial occlusion (12-15), the pancreas is highly susceptible to hypoperfusion resulting in hypoxia and inflammation (4-6, 16); and the phenomenon particularly affects the acini because of the structure of blood flow in pancreas (17) and the high metabolic activity of exocrine pancreas needing high oxygen supply.
The concept of inflammation after cardiac surgery with CPB is now widely accepted; and data have accumulated that demonstrated an important inflammatory reaction (18), neutrophil activation, and lymphocytes function dysregulation (19-24) that could play a key role in the pathogenesis of pancreatitis after cardiac surgery (25).
The diagnosis of acute pancreatitis (AP) requires the presence of clinical signs and hyperamylasemia and/or hyperlipasemia (1, 2, 9, 26, 27). More specific markers, such as the pancreatic isoamylase and immunoreactive trypsin (IRT), are probably more sensitive but less easily available. For instance, Fernandez-del Castillo et al. (28) detected evidence of the so-called pancreatic cellular injury (PCI) (as defined by hyperamylasemia and either increased serum lipase activity or an increase higher than reference range in the pancreatic isoamylase, or both) in 80 of 300 cardiac surgery patients. Only 23 of these 80 patients had associated abdominal signs or symptoms, and 3 developed clinically overt pancreatitis.
The present study was undertaken in patients submitted to heart surgery with CPB to detect biochemical signs of pancreatic injury. Over a 1-week follow-up, blood pancreatic enzyme levels (IRT, isoamylase, total amylase, and lipase) were systematically collected. A correlation between the severity of the PCI and markers of the acute phase reaction [α2-macroglobulin (α2-M) and α1-antiprotease inhibitor (α1-PI)] and of the polymorphonuclear neutrophil stimulation [myeloperoxidase (MPO), a specific marker of neutrophil activation] was searched (29, 30). This study did not intend to assess the clinical severity of pancreatic injury after heart surgery, but rather the time course and the presence of preoperative risk factors.
MATERIALS AND METHODS
Patient population and biochemical samples
With the approval of the Ethical Committee for Human Research of our institution, we studied 199 consecutive patients who underwent heart surgery with CPB. An informed consent was obtained from each patient before surgery. Demographic data and common risk factors for pancreatitis were recorded (Table 1).
Blood and urine samples were taken preoperatively (day 1); immediately after the end of surgery [day 0 (13 h)]; 8 h after the end of surgery [Day 0 (21 h)]; on the morning and the evening of the first postoperative day [day 1 (AM) and day 1 (PM)]; and on the second, third, fourth, and eighth postoperative days (days 2, 3, 4, and 8, respectively) for measurements of pancreatic enzymes (amylase, lipase, IRT, and pancreatic isoamylase), total and free calcium, proteins, MPO, α1-PI, and α2-M and for urinary IRT.
A standardized anesthetic technique was used (19). Anesthesia was induced and maintained with sulfentanil, diazepam, plus vecuronium or pancuronium for muscle relaxation. Patients were intubated and ventilated with a FiO2 = 1.0. Before CPB, heparin was given (3 to 5 mg/kg) to achieve a target activated clotting time longer than 400 s. Cardiopulmonary bypass was established through a standard median sternotomy, aortic root cannulation, and single or bicaval atrial cannulation for venous return. Haemaccel solution and mannitol were used for the priming. The flow rate through the bypass was calculated for each patient according to his body surface area.
Systemic temperature was kept between 28°C and 32°C. The MAP was continuously maintained between 90 and 100 mmHg before CPB and between 20 and 30 mmHg lower during CPB by using vasoactive drugs. Heparin neutralization was obtained with protamine at the end of the operation. Postoperatively, the patients received mechanical ventilation and were given morphine and diazepam as necessary until the next morning to keep cardiac, respiratory, renal, and metabolic functions stable.
The following variables were also recorded: anesthesia, type of surgery [coronary-artery by pass grafting (CABG) or valve replacement surgery (VRS)], details of CPB [duration, aortic cross-clamping time, and bypass temperature (normothermia or hypothermia), vasopressor/inotropic support, aprotinin administration, intra-aortic balloon counterpulsation, urine output, and/or filtration volume during surgery] (Table 2).
Biological markers analysis
Serum amylase was measured by a colorimetric enzymatic technique as described by Lorentz (31). Normal reference values ranged from 20 to 90 IU/L.
The pattern of serum isoamylases was determined by polyacrylamide gel electrophoresis as described by Warshaw and Lee (32). This technique enables the differentiation of the 5 pancreatic isoamylases from salivary or other tissue isoamylases. Normal reference values for pancreatic isoamylase ranged from 0 to 35 IU/L.
The enzymatic activity of lipase was measured by a nephelometric method according to the Ziegenhorn technique (33) (Beckman Instruments Inc, Fullerton, Calif). Normal reference values ranged from 0 to 200 IU/L.
Immunoreactive trypsin was measured in plasma or serum by on-site-developed radioimmunoassay technique as described previously (7, 34). This technique allows the determination of trypsinogen, free trypsin, and α1-PI-bound trypsin (but not α2-M-bound trypsin). Mean normal values of blood and urinary IRT are 32 ± 12 ng/mL and ≤1 ng/mL, respectively. Plasma IRT value of 44 ng/mL (mean value + SD) was considered as the upper normal limit.
Total calcium was measured by a colorimetric method in the form of an O-cresolphthalein complex (Roche Diagnostics kit, Mannheim, Germany). The presence of 8-hydroxyquinoline in the reactive milieu excluded the magnesium interference. Normal reference values ranged from 2.15 to 2.55 mmol/L. Free calcium was calculated as 50% of total calcium as described by Thomas (35).
The total protein concentration was measured in blood samples by the biuret method. Normal reference values ranged from 66 to 83 g/L.
This neutrophil enzyme was measured in plasma by on-site-developed double antibody radioimmunoassay as described previously (36). Myeloperoxidase value in normal individuals ranged in the interval 166 ± 100 ng/mL (mean value ± SD).
α1-Protease inhibitor and α2-macroglobulin-
The α1-PI and α2-M were detected in sera by immunonephelometric methods (Dade Behring Kit, Brussels, Belgium). Normal reference values ranged from 0.98 to 2.23 g/L for α1-PI and from 1.59 to 2.75 g/L for α2-M.
Biochemical definition of PCI
We used the following definition: PCI was defined as plasma IRT and pancreatic isoamylase levels higher than 70 ng/mL (mean ± 3 SD) and 35 IU/L, respectively, in 3 or more consecutive samples, excluding the preoperative (day 1) sample, regardless of the levels of amylase and lipase.
Radiological examination of the pancreas
In 15 patients with early hyperamylasemia (plasma amylase ≥90 IU/L during the first 36 h after surgery) and from whom an informed consent was obtained, a contrast-enhanced abdominal computed tomographic (CT) scan was performed. Blood amylase was chosen because this parameter could be obtained from the emergency laboratory.
Results from quantitative variables were generally as means and SD whenever appropriate; proportions were calculated for categorical data. PCI patients were compared at each time point with patients without PCI (no PCI) by the Student t test for quantitative variables and by the chi-square test for categorical findings. Multiple regression was used to assess the relationship between a continuous dependent variable and a set of independent covariables, whereas logistic regression was applied when the dependent variable was binary. Logistic regression was used to determine the intersection point (cutoff) or point of equal likelihood between the distribution of any covariate (i.e., IRT) in PCI and no PCI patients (37) when the 2 distributions of the 2 variables did not entirely overlap. Afterward, any value above the cutoff point was more likely in PCI patients; and any value below was more likely to be observed in no PCI patients. Results were considered to be significant at the 5% critical level (P < 0.05). All calculations were performed with SAS version 8.2 for Windows (SAS Inc, Cary, NC).
Among the 199 patients enrolled in the study, 6 died either in the operating room or early in the postoperative period at day 0 (13 h); and the data recorded in these patients were not included in the analysis. When using total amylase and lipase activities conventionally used to define AP (1, 2, 9, 26, 27), 85 patients (44%) experienced this complication. By contrast, according to our definition of PCI (plasma IRT >70 ng/mL and pancreatic isoamylase >35 IU/L), 133 patients (69%) were classified in the PCI group and 60 patients in the no PCI group. None of the 193 patients died during the time course of the study, but 6 (3%) patients (2 from the PCI group) died after the 8-day study period.
Demographic and surgical characteristics of the patients
The main demographic and surgical characteristics of the 2 groups of patients (PCI and no PCI) are displayed in Tables 1 and 2. No statistical difference was observed between the 2 groups. As compared with VRS, CABG was more common in men (86%) than in women (47%), was performed in 14% of patients operated in normothermic conditions (vs. 2% of VRS, P < 0.0001), and required similar CPB (85 ± 34 min vs. 89 ± 33 min for VRS) but a shorter clamping duration (mean value, 40 ± 17 min vs. 58 ± 21 min for VRS) (P < 0.0001).
Preoperative biological values
The preoperative values (day 1) of the biological variables are summarized in Tables 3 and 4. Remarkably, hematocrit; the concentrations of plasma IRT, isoamylase, total amylase, lipase, α1-PI, total calcium, and protein; and urinary IRT were already higher in the PCI than in the no PCI group. Among the PCI patients, 116 (87%), 90 (68%), and 85 (64%) patients had preoperative hyperlipasemia, hyperamylasemia, or both, respectively.
Postoperative biological values
Time course of blood amylase, isoamylase, lipase, IRT, calcium, and protein
Amylase. Among the 193 patients, 104 (54%) had at least once a concentration of total amylase higher than 90 IU/L. Overall, the total amylase values ranged from 5 to 1224 IU/L.
Pancreatic isoamylase. For pancreatic isoamylase, a concentration higher than 35 IU/L was found at least once in 153 (79.3%) of the patients. Overall, the isoamylase values ranged from 2 to 696 IU/L.
Lipase. Ninety-six (50%) of the 193 patients had at least once a concentration of lipase higher than 200 IU/L. The highest individual value was 4629 IU/L and was observed at day 2.
Immunoreactive trypsin. The IRT values measured in all patients ranged from 1 to 558 ng/mL. A concentration of IRT higher than 70 ng/mL was found at least once in 140 (72.5%) patients. Fifty patients exhibited plasma concentrations of IRT over 70 ng/mL preoperatively, and 13 of these 50 patients presented abnormal concentrations of both IRT and isoamylase and were thus already PCI patients.
The time course of the 4 plasma markers of pancreatic injury, namely, total amylase, isoamylase, lipase, and IRT, were similar in patients with PCI and no PCI, with a first increase between 8 and 15 h after surgery and a second later on (from day 3); but the magnitude of the changes were dramatically different between groups (Fig. 1, A-D). In the no PCI group, the plasma levels of the enzymes slightly increased but remained in the reference range. In the PCI group, VRS patients tended to have higher mean values of pancreatic enzymes in plasma compared with CABG patients. As shown by the regression analysis, the time course of the plasma levels of the 4 pancreatic markers was correlated in both groups. In addition, the preoperative values were unlikely to influence the postoperative course of pancreatic enzymes, as the percentage of patients having experienced a ≥2 times increase in preoperative values of the pancreatic enzymes was much higher in the PCI than in the no PCI group (Table 5).
Urinary IRT. Although much lower, the urinary IRT profile was similar to the plasma profile (Fig. 2), namely, a sharp rise and fall between day 0 (13 h) and day 1 (AM), followed by a progressive and long-lasting increase starting on the second or third day. A statistically significant difference in the mean values was observed at each time point between PCI and no PCI patients.
Blood calcium. The mean values of total and free calcium increased from day 1 to day 0 (13 h). This increase was the consequence of the peroperative calcium administration (700 mg/L) within the Haemaccel solution used to prime CPB (Fig. 3). The values remained higher than the day 1 value afterward (P < 0.0001). The time course of total calcium differed according to the presence or absence of PCI. In the PCI group, calcium continued to increase up to day 0 (21 h), and slowly decreased afterward. In contrast, calcium decreased from day 0 (13 h) in the no PCI group and was lower than in the PCI group for each time point.
Plasma proteins. Despite a progressive increase in both groups, the time course of plasma proteins differed between the PCI and no PCI patients. Plasma protein concentration was already higher in the PCI than in the no PCI group preoperatively, and increased postoperatively more markedly in the PCI than in the no PCI group (Table 4).
Time course of plasma inflammatory markers
Myeloperoxidase. The time course of plasma MPO mean concentrations was similar for the 2 groups of patients, with a first peak at day 0 (13 h) and a lower surge starting at day 1 (PM) to day 8 (Table 4). Interestingly, MPO was higher in the no PCI than in the PCI patients.
α1-Protease inhibitor. The time course of α1-PI was similar in the 2 groups of patients, although the plasma concentrations were systematically higher in the PCI than in the no PCI group, including preoperatively. There was a significant decrease after CPB compared with preoperative value and a statistically significant increase in the postoperative days. All the mean values were significantly increased from day 0 (21 h) to day 8 compared with day 1 for PCI and no PCI patients (Table 4).
α2-Macroglobulin. As opposed to α1-PI, α2-M decreased after surgery; and the mean values remained low up to day 1 (am) and slowly increased in the postoperative days without a complete return to the preoperative value (Table 4). No statistically significant difference was observed between PCI and no PCI patients.
Predictive values of preoperative levels
In an attempt to determine the risk of developing a postoperative PCI from preoperative analyte levels, logistic regression was applied to determine critical cutoff points. It was found that any preoperative plasma value (day 1) of IRT ≥40 ng/mL (P < 0.0001 and percentage concordance 80%), total amylase ≥42 IU/L (P < 0.001 and percentage concordance 69%), and pancreatic isoamylase ≥20 IU/L (P < 0.005 and percentage concordance 66%) was associated with postsurgery PCI (Fig. 4, A-C). None of the other variables assessed were predictive of PCI.
Correlation between biological parameters
From the separate analyses of the biological parameters, it seemed that the time courses of the 4 pancreatic enzymes (IRT, isoamylase, lipase, and total amylase) were highly interrelated. By regression analysis, the association between these 4 variables and potentially independent variables, including urinary IRT, MPO, α1-PI, α2-M, or total calcium, was assessed at each time point and in each group. It was found that the preoperative (day 1) values of the 4 enzymes were correlated with the urinary IRT concentration in both groups (P < 0.0001 for PCI, P < 0.05 for no PCI), total calcium in PCI group only (P < 0.05), MPO in no PCI group (P < 0.05), α2-M in PCI group (P < 0.0001), and creatinine in PCI group (P < 0.05). Postoperatively, the values of the 4 pancreatic enzymes were significantly correlated with urinary IRT and total calcium throughout the observation period in the PCI group, with MPO in the PCI group from day 2, with α1-PI in the PCI group from day 1 (AM), and with α2-M in the PCI group, up to day 0 (21 h) but no longer with creatinine from day 4.
Tomodensitometric examination (15 patients)
Early (18-36 h after surgery and total amylase >90 IU/L) contrast-enhanced abdominal CT scan images of the pancreas were obtained from 11 PCI patients and from 4 no PCI patients. A typical aspect of pancreatitis was found in 5 PCI patients, with swelling in 2 patients and a periglandular exudate in 3. The CT scan of the 6 other PCI patients and of the 4 no PCI patients did not reveal any sign of pancreatic injury. In only 1 patient with no evidence of pancreatitis on early CT scan, clinical and biochemical features of pancreatitis were found on the second postoperative day.
In this study, we found systematic increases in the serum and urine levels of pancreatic enzymes after cardiac surgery with CPB with values higher than the reference ranges in 69% of the patients. The single-center design of this study can, of course, limit the generalizability of these findings. However, using our stringent definition, we found a higher number of PCI patients than by using as markers of pancreas injury total hyperamylasemia and/or hyperlipasemia. It thus seemed that the use of isoamylase and trypsin, both exclusively released from the pancreas, may increase the sensitivity for the detection of clinically silent pancreatic injury as compared with total amylase and lipase, which were used in previous studies (1, 3, 26, 28, 38-41). These 2 variables are not pancreas-specific because amylase can also be released by salivary glands and the activity of the pancreatic lipase cannot be distinguished from the activity of the liver, intestinal, subendothelial, and lipoprotein lipase (40-43). We used 2 variables, IRT and isoamylase, that cannot be released from any other tissue than the pancreas; and a simultaneous elevation of these 2 variables above the upper limit of normal in at least 3 consecutive samples was required to define PCI. For IRT, we used a stringent criterion, with the upper limit of normal being set at the mean value + 3 SD.
The time course study demonstrated that the release of the pancreatic enzymes was typically biphasic, similar for IRT (in blood and urine), isoamylase, total amylase, and lipase, with a first peak value directly after surgery and a second progressive increase starting 4 to 8 days after surgery. Interestingly, in the group of no PCI patients, we observed a similar biphasic profile for the 4 enzymes, but with mean values remaining within the reference range.
Without surprise, we observed that PCI was associated with male gender. Renal insufficiency could represent another theoretical explanation for the increase in IRT, but is unlikely because no significant difference was observed in the renal function evaluated by plasma creatinine and urine output between groups. Another argument against the involvement of renal failure in the increase in IRT is that the profile of IRT in urine was similar to that of plasma, but with much lower values. Inasmuch as kidneys only eliminate trypsinogen (44), this suggest that the IRT in plasma was mainly trypsin complexes with antiproteases, which are eliminated by macrophages (45).
In contrast to Fernandez-del Castillo et al. (28), we did not find any relation between the amount of calcium given and the incidence of PCI. This lack of relation is not surprising in view of the systematic administration of a fixed calcium dose of 700 mg/L (or 1400 mg added to the priming solution) in our institution, contrasting with the use of a higher and variable dose peroperatively (1683 ± 879 mg for patients with injury and 1366 ± 663 mg for patients without injury) and postoperatively (0 to 6000 mg) by Fernandez-del Castillo et al. (28). According to these findings, these authors suggested that an elevated calcium dose could play a role in the onset of the PCI. The data presented here cannot assess the effect of supplemental calcium on the incidence of PCI, although we found a slight but significant difference (P = 0.032) in the preoperative blood Ca2+ concentration between groups, with a higher level of calcium being associated with the presence of PCI. In animal models (46, 47), increased intra-acinar Ca2+ concentration activates trypsinogen with acini damage, provoking or at least worsening acute pancreatitis. Hypercalcemia also exacerbates hypotension-induced pancreatic injury (48, 49) and neutrophil activation in pancreatic microcirculation (50).
An intriguing finding was the biphasic profile of the pancreatic enzyme release. The first peak could be attributed to ischemia or a peroperative inflammation reaction, and the second to inflammation. Indeed, a peroperative inflammation reaction with cellular acidosis and stimulation of neutrophils can be due to the surgical procedure (51, 52) and to the contact of blood with the material of the CPB circuit, increasing the adherence of phagocytes to the vessel walls. The CPB is accompanied by tissue hypoxia in splanchnic organs including pancreas (53), splanchnic hypoxia that is likely from a circulatory origin, even in the absence of significant decrease in hepatosplanchnic blood flow (54, 55), but could also originate from mechanical compression of the splanchnic area. In clinical practice, the presence of pancreatic ischemia is difficult to assess (56). Ischemia has been demonstrated to be linked to an increase in intracellular and intramitochondrial calcium concentrations (57). If ischemia of the splanchnic area contributed to the development of pancreatic injury, this ischemia has increased intra-acinar Ca2+ concentration, with a possible intracellular trypsinogen activation and neutrophil stimulation into the pancreatic circulation.
Regarding the late increase, stimulation of pancreas or the reabsorption of digestive enzymes from the intestine after food intake could also explain, at least partially, the second increase in the pancreatic enzymes. An additional argument for an inflammatory component for the second increase in pancreatic enzymes lies on the increase in α1-PI. The multivariate analysis suggests a direct link between the evolution of inflammatory parameters (MPO, α1-PI, α2-M) and pancreatic enzymes starting from day 2 in the PCI group. The importance of the early release of MPO, despite important hemodilution, reflects a sudden neutrophil activation due to CPB and the use of heparin for surgery (19, 20). Myocardial neutrophil sequestration and activation have been reported during CPB (21). In the late phase, systemic MPO release was smaller, perhaps reflecting a decrease in the activation of the neutrophils in the blood flow; but we cannot exclude that local neutrophil trapping and activation were present in the pancreas and in other splanchnic organs. The trapping of leucocytes in pancreas is an early phenomenon during pancreatitis (58, 59). The MPO release was higher in no PCI than in PCI patients, but these patients also had a higher preoperative MPO value.
Another interesting and surprising point of our study was the demonstration by multivariate analysis of a positive predictive value for preoperative IRT, total amylase, and pancreatic isoamylase values. Hence, the systematic measurement of the preoperative value of IRT, the most sensitive predictor of postoperative PCI, could be an indication to follow the pancreatic function in these patients. Actually, the high IRT values that we measured could be related to a minimal preoperative pancreatic injury, as IRT reflects the leakage of unactivated proenzymes from injured acinar cells (60). The PCI observed after CPB could thus have been facilitated by a "primed" pancreas. The priming of pancreas could result from preoperative ischemia, although this was not assessed.
In conclusion, we confirmed that PCI was common in CPB patients; and with severe and highly specific criteria, we found a higher frequency of PCI than previously reported for AP. In agreement with previous studies, the PCI was linked to the serum concentrations of calcium and was not associated to clinical signs of pancreatitis. Therefore, the clinical relevance of the present findings is uncertain. We also demonstrated that the release of pancreatic enzymes was biphasic: a direct postoperative release that could be linked to peroperative ischemia and a late phase that could be attributed to postsurgery inflammation reaction. Interestingly, we determined that IRT, isoamylase, and total amylase preoperative values were predictive of a postoperative PCI, a phenomenon that could be taken into consideration for the clinical follow-up of the CPB patients.
These findings could have an important implication for the management of cardiac surgery patients at risk of AP/PCI. It has been largely demonstrated that trypsin and other serine proteases released by suffering pancreas or reabsorbed from ischemic gut have numerous effects and play an important role in initiating or perpetuating an inflammation response (61-63). Further studies would test the hypothesis that normothermia and/or decreased calcium concentrations can decrease the frequency of postoperative pancreatic injury.
1. Perryman RG, Hoerr S: Observations on postoperative pancreatitis and postoperative elevation of the serum amylase. Am J Surg 88:417-423, 1954.
2. Manhaffey JH, Howard JM: The incidence of postoperative pancreatitis: study of 131 surgical patients utilizing serum amylase concentration. Arch Surg 70:348-352, 1955.
3. Tilney NL, Bailey GL, Morgan AP: Sequential system failure after rupture of abdominal aortic aneurysms. Ann Surg 178:117-122, 1973.
4. Warshaw AL, O'Hara PJ: Susceptibility of the pancreas to ischemic injury in shock. Ann Surg 188:197-201, 1978.
5. Jones MS, Dawson H: Studies of the ischemic pancreas in shock. Adv Shock Res 1:197-207, 1979.
6. Gmaz-Nikulin E, Nikulin A, Plamenac P, Hegewald G, Gaon D: Pancreatic lesions in shock and their significance. J Pathol 135:223-236, 1981.
7. Deby-Dupont G, Haas M, Pincemail J, Braun M, Lamy M, Deby C, Franchimont P: Immunoreactive trypsin in adult respiratory distress syndrome. Intensive Care Med 10:7-12, 1984.
8. Nicod L, Leuenberger C, Seydoux C, Rey F, Van Melle G, Perret C: Evidence for pancreatic injury in adult respiratory distress syndrome. Am Rev Respir Dis 131:696-699, 1985.
9. Weaver DW, Busuito MJ, Bouwman DL, Wilson RF: Interpretation of serum-amylase levels in critically ill patients. Crit Care Med 13:532-533, 1985.
10. Feiner H: Pancreatitis after cardiac surgery. Am J Surg 684:131-135, 1976.
11. Haas GS, Warshaw AL, Daggett WM, Aretz HT: Acute pancreatitis after cardiopulmonary bypass. Am J Surg 149:508-515, 1985.
12. Forsyth RP, Hoffbrand BI, Melmon KL: Redistribution of cardiac output during hemorrhage in the unanesthetized monkey. Circ Res 27:311-320, 1970.
13. Göber I: Die Auswirkung des hypovolämischen Shocks auf das Pankreas. Wien Klin Wochenschr 87:2-11, 1975.
14. Ferguson JL, Merrill GF, Miller HI, Spitzer JJ: Regional blood flow redistribution during early burn shock in the guinea pig. Circ Shock 4:317-326, 1977.
15. Florholmen J, Riepl R, Almdahl SM, Myklebust R, Burhol PG, Giercksky KE, Malm D: Impact of experimental endogenous gram-negative peritonitis on the pancreas of the rat as evaluated by cationic trypsin-like immunoreactivity in peritoneal fluid and serum and by electron microscopy of pancreatic tissue. Scand J Gastroenterol 22:313-320, 1987.
16. Hegewald G, Nikulin A, Gmaz-Nikulin E, Plamenac P, Barenwald G: Ultrastructural changes of the human pancreas in acute shock. Pathol Res Pract 179:610-615, 1985.
17. Endrich B, Klar E, Hammersen F: The microcirculation of the pancreas: state of the art. In: Messmer K, Hammersen F (eds.): Gastrointestinal Microcirculation. Progress in Applied Microcirculation, Vol 17. Basel, Switzerland: Karger AG, pp 144-174, 1990.
18. Gessler P, Pretre R, Hohl V, Rousson V, Fischer J, Dahinden C: CXC-chemokine stimulation of neutrophils correlates with plasma levels of myeloperoxidase and lactoferrin and contributes to clinical outcome after pediatric cardiac surgery. Shock 22:513-520, 2004.
19. Faymonville ME, Pincemail J, Duchateau J, Paulus JM, Adam A, Deby-Dupont G, Deby C, Albert A, Larbuisson R, Limet R, et al.: Myeloperoxidase and elastase as markers of leukocyte activation during cardiopulmonary bypass in humans. J Thorac Cardiovasc Surg 102:309-317, 1991.
20. Hashimoto K, Miyamoto H, Suzuki K, Horikoshi S, Matsui M, Arai T, Kurosawa H: Evidence of organ damage after cardiopulmonary bypass. J Thorac Cardiovasc Surg 104:666-673, 1992.
21. Farah B, Vuillemenot A, Lecompte T, Bara L, Pasquier C, Jebara V, Carpentier A, Fabiani JN: Myocardial neutrophil sequestration and activation related to the reperfusion of human heart during coronary artery surgery. Cardiovasc Res 28:1226-1230, 1994.
22. Kong C-W, Huang C-H, Hsu T-G, Tsai KK, Hsu C-F, huang M-C, Chen L-C: Leukocyte mitochondrial alterations after cardiac surgery involving cardiopulmonary bypass: clinical correlations. Shock 21:315-319, 2004.
23. Halter J, Steinberg J, Fink G, Lutz C, PCIone A, Maybury R, Fedors N, DiRocco J, Lee HM, Nieman G: Evidence of systemic cytokine release in patients undergoing cardiopulmonary bypass. J Extra Corpor Technol 37:272-277, 2005.
24. Raja SG, Dreyfus GD: Modulation of systemic inflammatory response after cardiac surgery. Asian Cardiovasc Thorac Ann 13:382-395, 2005.
25. Granger J, Remick D: Acute pancreatitis: models, markers, and mediators. Shock 24(suppl 1):45-51, 2005.
26. Morrissey R, Berk JE, Fridhandler L, Pelot D: The nature and significance of hyperamylasemia following operation. Ann Surg 178:67-71, 1974.
27. Paajanen H, Nuutinen P, Harmoinen A, Pöyhönen M, PitKänen O, Nordback I, Grönroos J, Nevalainen TJ: Hyperamylasemia after cardiopulmonary bypass: pancreatic cellular injury or impaired renal excretion of amylase? Surgery 123:504-510, 1998.
28. Fernandez-Del Castillo C, Harringer W, Warshaw AL, Vlahakes GJ, Koski G, Zaslavsky AM, Rattner DW: Risk factors for pancreatic cellular injury after cardiopulmonary bypass. N Engl J Med 325:382-387, 1991.
29. Caimi G, Hoffmann E, Montana M, Incalcaterre E, Dispensa F, D'Amico T, Canino B, Lo Presti R: Plasma markers of platelet and polymorphonuclear leukocyte activation in young adults with myocardial infarction. Clin Hemorheol Microcirc 32:67-74, 2005.
30. Gash O, Nys M, Deby-Dupont G, Chapelle J-P, Lamy M, Piérard LA, Legrand V. Acute neutrophil activation in direct stenting: comparison of stable and unstable angina patients. Int J Cardiol. 112:59-65, 2005.
31. Lorentz K: α-Amylase assay: current state and future development. J Clin Chem Clin Biochem 17:499-504, 1979.
32. Warshaw AL, Lee KH: Characteristic alterations of serum isoenzymes of amylase in diseases of liver, pancreas, salivary gland, lung, and genitalia. J Surg Res 22:362-369, 1977.
33. Ziegenhorn J, Neumann U, Knitsch KW, Zwez W: Determination of serum lipase. Clin Chem 25:1067, 1979.
34. Malvano R, Marchisio M, Massaglia A, Giacosa PA, Zannino M, Andriulli A, Burlina A: Radioimmunoassay of trypsin-like substance in human serum. Scand J Gastroenterol 15:2-10, 1980.
35. Thomas L: Calcium (Ca)., In: Thomas L (ed.): Clinical Laboratory Diagnostics: Use and Assessment of Clinical Laboratory Results, 3rd ed. Frankfurt, Germany: TH-Books, pp 213-237, 1998.
36. Pincemail J, Deby-Dupont G, Deby C, Thirion A, Torpier G, Faymonville ME, Tomassini M, Lamy M, Franchimont P: Fast double antibody radioimmunoassay of human granulocyte myeloperoxidase and its application to human plasma. J Immunol Methods 137:181-191, 1991.
37. Albert A, Harris EK: Multivariate Interpretation of Clinical Laboratory Data. Statistics: Textbooks and Monographs, Vol. 75. New York: Marcel Dekker, 1987. 328 pp.
38. Panebianco AC, Scott SM, Dart CH, Takaro T, Echegaray HM: Acute pancreatitis following extracorporeal circulation. Ann Thorac Surg 9:562-567, 1970.
39. Svensson LG, Decker GD, Kinsley RB: A prospective study of hyperamylasemia and pancreatitis after cardiopulmonary bypass. Ann Thorac Surg 39:409-411, 1985.
40. Kazmierczak SC, Van Lente F: Incidence and source of hyperamylasemia after cardiac surgery. Clin Chem 34:916-919, 1988.
41. Rattner DW, Gu ZY, Vlahakes GJ, Warshaw A: Hyperamylasemia after cardiac surgery. Ann Surg 209:279-283, 1989.
42. Yadav D, Agarwal N, Pitchumoni CS: A critical evaluation of laboratory tests in acute pancreatitis. Am J Gastroenterol 97:1309-1318, 2002.
43. Kylänpää-Bäck M-L, Kemppainen E, Puolakkainen P: Trysin-based laboratory methods and carboxypeptidase activation peptide in acute pancreatitis. JOP 3:34-48, 2002.
44. Borgström A, Ohlsson K: Studies on the turnover of endogenous cathodal trypsinogen in man. Eur J Clin Invest 8:379-382, 1978.
45. Debanne MT, Beel R, Dolovich J: Uptake of proteinase-α2-macroglobulin complexes by macrophages. Biochem Biophys Acta 411:295-304, 1975.
46. Frick TW, Fernandez-del Castillo C, Bimmler D, Warshaw AL: Elevated calcium and activation of trypsinogen in rat pancreatic acini. Gut 41:339-343, 1997.
47. Raraty M, Ward J, Erdemli G, Vaillant C, Neoptolemos JP, Sutton R, Petersen OH: Calcium-dependent enzyme activation and vacuole formation in the aPCIal granular region of pancreatic acinar cells. Proc Natl Acad Sci U S A 97:13126-13131, 2000.
48. Hlouschek V, Finkes T, Turi S, Weber IA, Singh J, Domschke W, Schnekenburger J, Kruger B, Lerch MM: Early changes in pancreatic acinar cell calcium signaling after pancreatic duct obstruction. J Biol Chem 278:9361-9369, 2003.
49. Mithofer K, Warshaw AL, Frick TW, Lewandroski KB, Koski G, Rattner DW, Fernandez-Del Castillo C: Calcium administration augments pancreatic injury and ectoPCI trypsinogen activation after temporary systemic hypotension in rats. Anesthesiology 83:1266-1273, 1995.
50. Zhou ZG, Chen YQ, Liu XB, Hu WM, Tian BL, Chen HQ: Changes of cytosolic [Ca2+] i in neutrophils in pancreatic microcirculation of rats with caerulein-induced acute pancreatitis under fluid shear stress. World J Gastroenterol 10:3185-3187, 2004.
51. Biglioli P, Cannata A, Alamanni F, Naliato M, Porqueddu M, Zanobini M, Tremoli E, Parolari A: Biological effects of off-pump vs. on-pump coronary artery surgery: focus on inflammation, hemostasis and oxidative stress. Eur J Cardiothorac Surg 24:260-269, 2003.
52. Chello M, Mastroroberto P, Quirino A, Cuda G, Perticone F, Cirillo F, Covino E: Inhibition of neutrophil apoptosis after coronary bypass operation with cardiopulmonary bypass. Ann Thorac Surg 73:123-130, 2002.
53. Kumle B, Boldt J, Suttner SW, Piper SN, Lehmann A, Blome M: Influence of prolonged cardiopulmonary bypass times on splanchnic perfusion and markers of splanchnic organ function. Ann Thorac Surg 75:1558-1564, 2003.
54. Thorén A, Jakob SM, Pradl R, Elam M, Ricksten SE, Takala J: Jejunal and gastric mucosal perfusion versus splanchnic blood flow and metabolism: an observational study on postcardiac surgical patients. Crit Care Med 28:3649-3654, 2000.
55. Andrasi TB, Bielik H, Blazovics A, Zima E, Vago H, Szabo G, Juhasz-Nagy A: Mesenteric vascular dysfunction after cardiopulmonary bypass with cardiac arrest is aggravated by coexistent heart failure. Shock 23:324-329, 2005.
56. Sakorafas GH, Tsiotos GG, Satt MG: Ischemia/Reperfusion-induced pancreatitis. Dig Surg 17:3-14, 2000.
57. Di Lisa F, Canton M, Menabo R, Dodoni G, Bernardi P: Mitochondria and reperfusion injury. The role of permeability transition. Basic Res Cardiol 98:235-241, 2003.
58. De Coninck B, Meert P, Hanique G, Jadoul M, Dugernier T, el Gariani A, Marion E, Ferrant A, Reynaert M: Scintigraphy with Indium-labelled leukocytes in acute pancreatitis. Acta Gastroenterol Belg 54:176-173, 1991.
59. Hartwig W, Carter EA, Jimenez RE, Werner J, Fischman AJ, Fernandez-Del Castillo C, Warshaw AL: Chemotactic peptide uptake in acute pancreatitis: correlation with tissue accumulation of leucocytes. J Appl Physiol 87:743-749, 1999.
60. Appelros S, Petersson U, Toh S, Johnson C, Borgstrom A: Activation peptide of carboxypeptidase B and anionic trypsinogen as early predictors of the severity of acute pancreatitis. Br J Surg 88:216-221, 2001.
61. Deitch EA, Shi HP, Lu Q, Feketeova E, Xu DZ: Serine proteases are involved in the pathogenesis of trauma-hemorrhagic shock-induced gut and lung injury. Shock 19:452-456, 2003.
62. Waldo SW, Rosario HS, Penn AH, Schmid-Schonbein GW: Pancreatic digestive enzymes are potent generators of mediators for leukocyte activation and mortality. Shock 20:138-143, 2003.
63. Acosta JA, Hoyt DB, Schmid-Schonbein GW, Hugli TE, Anjaria DJ, Frankel DA, Coimbra R: Intraluminal pancreatic serine protease activity, mucosal permeability, and shock: a review. Shock 26:3-9, 2006.
Pancreas; cardiac surgery; bypass; myeloperoxidase; trypsin; amylase
©2007The Shock Society
Highlight selected keywords in the article text.