Calcium administration augments pancreatic injury and ectopic trypsinogen activation after temporary systemic hypotension in rats

BACKGROUND

Calcium infusion and hypotension have been described as the most important risk factors for pancreatic injury after cardiopulmonary bypass.

METHODS

Rats were randomly allocated to three experimental groups undergoing either sham operation and saline infusion (Control, n = 30), hemorrhagic reduction of mean arterial pressure to 30 mmHg for 30 min alone (hypotension, n = 51), or hypovolemic hypotension followed by bolus infusion of CaCl2 (200 mg.kg-1; hypercalcemia, n = 85). Serum ionized calcium, amylase activity, trypsinogen activation peptide in pancreatic tissue homogenates, pancreatic wet/dry weight ratio, histologic changes, and mortality were assessed for 24 h.

RESULTS

Control rats showed no significant changes of any parameter throughout the experiments. In contrast, hypotension significantly increased serum amylase (P < 0.001), tissue trypsinogen activation peptide (P < 0.01), wet/dry weight ratio (P < 0.001), and histologic scores for edema (P < 0.001) and pancreatic necrosis (P < 0.05). Subsequent CaCl2 administration transiently increased [Ca2+] (P < 0.001) with the concentration rapidly returning to baseline within 3 h. That infusion of CaCl2 further increased amylase (P < 0.05), tissue trypsinogen activation peptide (P < 0.05), wet/dry weight ratio (P < 0.001), and histologic evidence of pancreatic edema (P < 0.05) and acinar necrosis (P < 0.05) when compared with hypotension alone. Whereas all Control animals survived the experiments, 22% (P < 0.05) and 47% (P < 0.05 vs. hypotension) of animals died in the hypotension and hypercalcemia groups, respectively.

CONCLUSIONS

Temporary hypotension alone causes ectopic trypsinogen activation and lethal acute pancreatitis. Super-imposed hypercalcemia significantly aggravates hypotension-induced pancreatic injury and mortality in rats.

Peritoneal exfoliative cytology and Ki-ras mutational analysis in patients with pancreatic adenocarcinoma

This is the first description of the detection of pancreatic adenocarcinoma peritoneal metastasis by established radiolabeled polymerase chain reaction (PCR) based Ki-ras mutational analysis. The present study evaluates both routine cytology and Ki-ras mutational analysis in the detection of peritoneal micrometastases in 24 subjects with pancreatic adenocarcinoma compared to seven control cases of chronic pancreatitis and seven control cases of cholecystitis. Locoregional extension, vascular invasion, and distal metastases were confirmed in 21/24 (88%) of the subjects with pancreatic adenocarcinoma by compute tomography, angiography, endosonography, or laparoscopy. The most common site of histologically confirmed extrapancreatic involvement was the vasculature (29%), followed by the liver (25%), duodenum (17%), peritoneum (17%), and lymph nodes (12%). Peritoneal lavage cytology was positive in 3/24 (12%) cases of pancreatic carcinoma while Ki-ras codon 12 mutational analysis was positive in 2/24 (8%). Two histologically confirmed cases of peritoneal metastases were not detected by either methodology, while peritoneal lavage cytology detected malignant cells in one case with histologically confirmed lymph node metastasis.

Cystic tumors of the pancreas

Cystic neoplasms of the pancreas are relatively uncommon lesions that should not be confused with pancreatic pseudocysts. Guidelines for this differential diagnosis, the characteristics of the more common varieties, and the authors' experience with 130 cystic tumors are described.

Effect of microcirculatory perfusion on distribution of trypsinogen activation peptides in acute experimental pancreatitis

Extraintestinal trypsinogen activation peptides (TAP) have been shown to correlate with severity of acute pancreatitis in humans as well as in various animal models. Ischemia superimposed on experimental pancreatitis, however, increases acinar cell injury without increasing TAP in plasma. We speculated that TAP generated in the pancreas might not reach the circulation in necrotizing pancreatitis due to decreased pancreatic perfusion. To test the hypothesis that generation of TAP in plasma is related to pancreatic perfusion and that plasma TAP may therefore underestimate acinar cell injury in necrotizing disease, we correlated TAP in pancreatic tissue and body fluids with capillary pancreatic blood flow in necrotizing and edematous pancreatitis. The ratio between necrosis and TAP in tissue was similar in both models; the ratio between TAP in plasma and tissue, however, was significantly lower in necrotizing pancreatitis, indicating that a certain amount of TAP generated in the pancreas did not reach the circulation. Decreased pancreatic perfusion found in necrotizing pancreatitis was consistent with this finding. Our data suggest that TAP in tissue is most reliable to indicate severity of acute pancreatitis, whereas plasma TAP may underestimate pancreatic injury in necrotizing disease due to decreased pancreatic perfusion.

Quantitative assay of trypsinogen by measurement of cleaved activation peptide after activation with enterokinase

Measurement of trypsinogen by quantitative immunosorbent assay of the highly specific and inert trypsinogen activation peptide following complete activation of trypsinogen by enterokinase offers a simple and sensitive method that provides reliable results over a wide range of trypsinogen concentrations. This method offers a significant advantage in being applicable to complex biological tissues.

Postburn pancreatitis

OBJECTIVE

The authors examined the prevalence and complications of pancreatitis in severely burned patients. Factors predictive for the development of pancreatitis after burns are considered.

SUMMARY BACKGROUND DATA

Pancreatitis has been documented at necropsy after burns; however, it is not clinically recognized as a common complication of burn injury. Recent improvements in survival rates could yield previously unrecognized complications, such as pancreatitis, particularly in those patients who previously would have not survived. The hypothesis is that pancreatitis is a frequent complication after major burn injury and causes significant morbidity for patients with large burns.

METHODS

This retrospective review of adult patients with large burns examines postburn pancreatitis using stepwise logistic regression analysis.

RESULTS

Forty-nine of 121 (40%) patients developed hyperamylasemia or hyperlipasemia well after the admission period (23 +/- 3 days), and all enzyme abnormalities were temporally associated with emerging infections. Most of these patients (40/49, 82%) had symptoms of pancreatitis. Three patients (6%) had pancreatic pseudocysts or abscesses. Inhalation injury (p = 0.0001), associated trauma (p = 0.0311), and escharotomy (p = 0.0415) were risk factors for pancreatitis. Using Fischer's exact test, patients with pancreatitis had increased mortality and length of stay. Patients with high enzyme elevations and > or = 50% body surface area burned were at severe risk of pancreatic pseudocyst or abscess development (43%; 90% confidence interval of 23-77%).

CONCLUSIONS

Pancreatitis is a frequent complication after large burn injuries. Patients at high risk for pancreatitis complications should receive surveillance examinations during their acute hospitalization.

Pathogenesis and prevention of early pancreatic infection in experimental acute necrotizing pancreatitis

OBJECTIVE

The authors test antibiotic strategies aimed at either mitigating bacterial translocation from the gut or delivering antibiotics specifically concentrated by the pancreas for prevention of early secondary infection after acute necrotizing pancreatitis.

BACKGROUND

Infection currently is the principal cause of death after severe pancreatitis. The authors have shown that the risk of bacterial infection correlates directly with the degree of tissue injury in a rodent model of pancreatitis. Bacteria most likely arrive by translocation from the colon.

METHODS

Severe acute necrotizing pancreatitis was induced in rats by a combination of low-dose controlled intraductal infusion of glycodeoxycholic acid superimposed on intravenous cerulein hyperstimulation. At 6 hours, animals were randomly allocated to five treatment groups: controls, selective gut decontamination (oral antibiotics and cefotaxime), oral antibiotics alone, cefotaxime alone, or imipenem. At 96 hours, surviving animals were killed for quantitative bacterial study of the cecum, pancreas, and kidney.

RESULTS

The 96-hour mortality (35%) was unaffected by any treatment regimen. Cecal gram-negative bacteria were significantly reduced only by the oral antibiotics. Pancreatic infection was significantly reduced by full-gut decontamination and by imipenem, but not by oral antibiotics or by cefotaxime alone. Renal infection was reduced by both intravenous antibiotics.

CONCLUSIONS

Early pancreatic infection after acute necrotizing pancreatitis can be reduced with a full-gut decontamination regimen or with an antibiotic concentrated by the pancreas (imipenem) but not by unconcentrated antibiotics of similar spectrum (cefotaxime) or by oral antibiotics alone. These findings suggest that 1) both direct bacterial translocation from the gut and hematogenous seeding interplay in pancreatic infection while hematogenous seeding is dominant at extrapancreatic sites and 2) imipenem may be useful in clinical pancreatitis.

Further experience with laparoscopy and peritoneal cytology in the staging of pancreatic cancer

A total of 114 patients with pancreatic cancer and no evidence of metastatic disease by computed tomography underwent laparoscopy. Intra-abdominal spread was present in 27 patients (24 per cent). Metastases were 2.4 times more common in tumours of the distal pancreas (11 of 25, 44 per cent) than in pancreatic head tumours (16 of 89, 18 per cent) (P < 0.05). None of the 27 patients with metastases underwent further surgery. Of the 87 patients without metastatic spread, 42 were found to have vascular invasion by angiography and were offered radiation therapy. Thirty of 40 patients explored surgically were resected; two of the remaining ten had peritoneal spread that had been missed at laparoscopy (false-negative rate of 7 per cent). Cytological examination of peritoneal washings was performed in 94 patients at the time of laparoscopy; cytology was positive in 16 (17 per cent). There was a significant correlation between positive cytology and presence of visible liver or peritoneal metastases (ten of 22 with metastases versus six of 72 without, P < 0.001). Positive cytology was present in six patients (8 per cent) without visible metastases, but none had resectable disease.

Acute hypercalcemia causes acute pancreatitis and ectopic trypsinogen activation in the rat

BACKGROUND & AIMS

Clinical and experimental observations have associated acute and chronic hypercalcemia with pancreatitis. The aim of this study was to determine whether acute hypercalcemia can induce acute pancreatitis and, if so, whether the pathogenesis involves premature protease activation.

METHODS

Rats given bolus infusions of CaCl2 (200 mg/kg; n = 76) were compared with saline-treated controls (n = 40). Serum [Ca2+], serum amylase activity, trypsinogen activation peptide (TAP) concentration in serum and pancreatic tissue, pancreatic wet/dry weight ratio, and histology were assessed for 24 hours. For dose-response analysis, CaCl2 was injected at a dose of 50-200 mg/kg, and the aforementioned indices were assayed for 1 hour (n = 5 each).

RESULTS

There were no significant changes in the controls. Calcium infusion increased serum [Ca2+] 3-fold after 5 minutes (P < 0.001). Within 1 hour, serum amylase (2.5-fold) and tissue TAP (3-fold) levels increased along with macroscopic and microscopic edema formation and leukocytic infiltration. The extent of the changes at 1 hour correlated with the calcium dose. Amylase and tissue TAP concentrations remained elevated until 24 hours when serum TAP concentration had increased (P < 0.001) and focal acinar necrosis became evident.

CONCLUSIONS

Acute experimental hypercalcemia induces dose-dependent morphological alterations characteristic of acute pancreatitis, acute hyperamylasemia, and early ectopic trypsinogen activation. This supports the pathophysiological relevance of excess calcium and offers a possible pathogenetic mechanism for its association with clinical pancreatitis.

Cyst fluid NB/70K concentration and leukocyte esterase: two new markers for differentiating pancreatic serous tumors from pseudocysts

Cystic lesions of the pancreas include inflammatory pseudocysts, serous cystadenomas, and mucinous tumors, some of which are malignant. Preoperative clinical and radiological parameters are unreliable and may result in incorrect diagnosis and inappropriate treatment. Cyst fluid analysis for cytology, viscosity, carcino-embryonic antigen, CA 72-4, and CA 15-3 will distinguish mucinous from nonmucinous lesions and usually help in determining malignancy. Currently, there is no reliable method to differentiate inflammatory pseudocysts from serous cystadenomas. This distinction is important because the treatment of these two lesions is different; pseudocysts are either observed or drained, whereas serous tumors are usually resected. The tumor marker NB/70K was measured in aspirated cyst fluid from 13 inflammatory pseudocysts and 11 serous cystadenomas by a commercial immunoassay. Leukocyte esterase was measured using Chemstrip SG urine test strips and amylase and lipase on a routine chemistry analyzer. The cyst fluid NB/70K concentration was significantly higher in pseudocysts (mean, 555 U/ml; range, 42-1,926 U/ml) than in serous cystadenomas (mean, 12 U/ml; range 0-130 U/ml) and this difference was significant (p < 0.0002). Leukocyte esterase was detected in 7 of 11 pseudocysts but was absent in 10 of 10 serous tumors (p = 0.002). Amylase and lipase values were generally higher in pseudocysts but these markers were unreliable due to marked outliers. Cyst fluid NB/70K and leukocyte esterase are promising markers to help differentiate pseudocysts from serous tumors on percutaneous aspirates. When combined with previously reported cyst fluid parameters (amylase, lipase, cytology, and amylase isoenzymes), these two cystic lesions can be reliably distinguished.

Standards for pancreatic resection in the 1990s

OBJECTIVE

To describe the current indications and operative outcomes of pancreatic resection.

DESIGN

Retrospective case series.

SETTING

Referral practice in a university hospital.

PATIENTS

Two hundred thirty-one consecutive patients undergoing pancreatoduodenectomy (PD), distal pancreatectomy (DP), or total pancreatectomy (TP) over a 44-month period. Their ages ranged from 16 to 85 years, with a mean of 54 years; 20% of the patients were 70 years old or older.

MAIN OUTCOME MEASURES

Mortality, complications, and length of hospital stay.

RESULTS

Operative mortality was 0.4% (one death following DP); there were no deaths in 142 PDs or in 18 TPs. The most common complication following PD was delayed gastric emptying. Pancreatic fistula occurred in 6.3% of PD and in 9.8% of DP patients. Overall, 58% of PD, 80% of DP, and 78% of TP patients had no complications. The mean +/- SD length of hospital stay was 15 +/- 7, 10 +/- 5, and 15 +/- 6 days for PD, DP, and TP, respectively. Reoperation for any cause was necessary in only 1.2% (3/231). The most frequent indication for PD was pancreatic cancer (36%) followed by chronic pancreatitis (26%); for DP it was chronic pancreatitis (28%) and cystic neoplasms (27%); and for TP, chronic pancreatitis (55%). Newer indications for pancreatic resection included mucinous ductal ectasia and intraductal papillary tumors (eight cases, 4%) and metastatic tumors (eight cases, 4%).

CONCLUSIONS

Current indications for pancreatic resection have expanded. These procedures are associated with a low risk for death and postoperative complications when performed in a high-volume setting.

Increased intrapancreatic trypsinogen activation in ischemia-induced experimental pancreatitis

OBJECTIVE

The potential of pancreatic ischemia to cause acute pancreatitis as indicated by morphologic changes and ectopic trypsinogen activation was investigated.

BACKGROUND

Experimental evidence has shown that pancreatic ischemia is important in the evolution of severe pancreatitis, but whether ischemia can initiate pancreatitis has been disputed.

METHODS

Pancreatic ischemia was induced in rats by hemorrhagic hypotension (30 mm Hg for 30 min; n = 64). Changes of pancreatic microcirculatory perfusion were studied using diffuse reflectance spectroscopy. Serum amylase, trypsinogen activation peptide (TAP) in serum and pancreatic tissue, wet/dry weight ratio, and histology were determined over 24 hours and compared with sham-operated control subjects (n = 35).

RESULTS

In control animals, serum amylase (47.9 +/- 2.1 units/L), serum (7.9 +/- 0.7 nmol/L) and tissue TAP (63.0 +/- 5.4 nmol/L x g), wet/dry weight ratio (2.8 +/- 0.1), and histology remained unchanged. Temporary hypotension markedly decreased pancreatic perfusion with incomplete recovery after reperfusion. Pancreatic isoamylase activity increased within 1 hour (110 +/- 5 units/L, p < 0.05) and further to 151 +/- 18 units/L at 24 hours. Tissue TAP was elevated at 1 hour (134 +/- 16 nmol/L x g, p < 0.05) and increased to 341 +/- 43 nmol/L x g (p < 0.001) after 24 hours, whereas serum TAP remained unchanged (8.3 +/- 0.5 nmol/L). Morphologic alterations included elevated wet/dry weight ratio (4.1 +/- 0.3, p < 0.01) and increased histologic scores for edema (p < 0.05) and acinar necrosis (p < 0.05) at 24 hours. Trypsinogen activation preceded the development of pancreatic necrosis.

CONCLUSIONS

In addition to its potentiating role, severe pancreatic ischemia can play a pathogenetic role in the initiation of acute pancreatitis.

Alcohol selectively impairs oxygenation of the pancreas

BACKGROUND

Alcohol predisposes to acute pancreatitis by an unknown mechanism and is known to reduce regional pancreatic blood flow.

OBJECTIVE

To investigate whether increased blood alcohol causes functional impairment of pancreatic microperfusion as indicated by reduced tissue oxygenation.

DESIGN

Prospective, randomized, controlled study.

SETTING

University hospital laboratory.

SUBJECTS

Forty adult female Wistar rats.

INTERVENTIONS

Intravenous infusion of ethanol (2 g/kg) or saline over 60 minutes. Tissue hemoglobin oxygenation saturation and hemoglobin content were measured using reflectance spectroscopy in the pancreas, stomach, and kidney at baseline and at 10-minute intervals for 1 hour.

RESULTS

Blood ethanol levels (mean +/- SEM) peaked at 1810 +/- 94 mg/L. Compared with saline controls, pancreatic hemoglobin oxygen saturation in ethanol-treated rats had significantly decreased by 40 minutes and remained depressed, while pancreatic hemoglobin content was unchanged. The magnitude of the depression was at least as great as that previously observed in acute experimental pancreatitis of moderate severity. Neither hemoglobin oxygen saturation nor hemoglobin content was affected in the stomach or kidney by ethanol.

CONCLUSION

A raised blood ethanol level was associated with acutely decreased hemoglobin oxygen saturation in the pancreas but not in the stomach or kidney. This observation of provoked hypoxia provides a possible mechanism by which alcohol contributes to pancreatic injury.

Intravenous contrast medium aggravates the impairment of pancreatic microcirculation in necrotizing pancreatitis in the rat

BACKGROUND

Previous reports demonstrated that radiographic contrast medium, as used in contrast-enhanced computed tomography, increases acinar necrosis and mortality in experimental pancreatitis. The authors studied the possibility that these changes may be related to an additional impairment of pancreatic microcirculation.

METHODS

Fifty Wistar rats had acute pancreatitis induced by intraductal glycodeoxycholic acid (10 mmol/L for 10 min) and intravenous cerulein (5 micrograms/kg/hr for 6 hrs). After rehydration (16 mL/kg), pancreatic capillary perfusion was quantified by means of intravital microscopy at baseline before intravenous infusion of contrast medium (n = 25) or saline (n = 25), and 30 and 60 minutes thereafter. In addition to total capillary flow, capillaries were categorized as high- or low-flow (> or < 1.6 nL/min).

RESULTS

Pancreatic capillary flow did not change in either high- or low-flow capillaries after saline infusion. However, contrast medium infusion induced a significant decrease of total capillary flow (p < 0.001). Analysis according to the relative flow rate revealed that this was primarily because of a significant additional reduction of perfusion in low-flow capillaries (p < 0.0001). Furthermore, complete capillary stasis was observed in 15.9 +/- 3.4% after contrast medium as compared with 3.2 +/- 1.2% after saline infusion (p < 0.006).

CONCLUSION

Radiographic contrast medium aggravates the impairment of pancreatic microcirculation in experimental necrotizing pancreatitis.

Pancreatic surgery. A paradigm for progress in the age of the bottom line

The past few decades have seen great change in the capabilities of medical care. The next decade will emphasize great change in its delivery, driven mainly by the necessity of containing costs presently running at more than 13% of the gross national product. The current perception is that two of the principal causes of the excessive cost of medical care in the United States are the price of new technology and the fact that much of the care rendered is by specialists. In fact, most surgical care can be rendered by general surgeons, and the coming changes will revitalize the demand for and self-esteem of general surgeons. Managed care will recognize this by using general surgeons and keeping patients "down on the farm," a line drawn from the World War I era song entitled "How can you keep them down on the farm after they've seen Paree?" But some things are still unique to the medical equivalent of Paris, perhaps including more complex forms of treatment, the acquisition of knowledge, and teaching. The questions are: what should we decentralize and how do we discriminate what should remain decentralized in the community for economy, and what might be concentrated to good advantage in the centers? I would like to offer the pancreas as a paradigm in thinking about these issues.

Standards for pancreatic resection in the 1990s

OBJECTIVE

To describe the current indications and operative outcomes of pancreatic resection.

DESIGN

Retrospective case series.

SETTING

Referral practice in a university hospital.

PATIENTS

Two hundred thirty-one consecutive patients undergoing pancreatoduodenectomy (PD), distal pancreatectomy (DP), or total pancreatectomy (TP) over a 44-month period. Their ages ranged from 16 to 85 years, with a mean of 54 years; 20% of the patients were 70 years old or older.

MAIN OUTCOME MEASURES

Mortality, complications, and length of hospital stay.

RESULTS

Operative mortality was 0.4% (one death following DP); there were no deaths in 142 PDs or in 18 TPs. The most common complication following PD was delayed gastric emptying. Pancreatic fistula occurred in 6.3% of PD and in 9.8% of DP patients. Overall, 58% of PD, 80% of DP, and 78% of TP patients had no complications. The mean +/- SD length of hospital stay was 15 +/- 7, 10 +/- 5, and 15 +/- 6 days for PD, DP, and TP, respectively. Reoperation for any cause was necessary in only 1.2% (3/231). The most frequent indication for PD was pancreatic cancer (36%) followed by chronic pancreatitis (26%); for DP it was chronic pancreatitis (28%) and cystic neoplasms (27%); and for TP, chronic pancreatitis (55%). Newer indications for pancreatic resection included mucinous ductal ectasia and intraductal papillary tumors (eight cases, 4%) and metastatic tumors (eight cases, 4%).

CONCLUSIONS

Current indications for pancreatic resection have expanded. These procedures are associated with a low risk for death and postoperative complications when performed in a high-volume setting.

Late histopathologic changes and healing in an improved rodent model of acute necrotizing pancreatitis

Studies of experimental pancreatitis have generally focussed on early time points rather than the stages of healing and resolution or scarring. We recently characterized a new pancreatitis model of moderate severity by combining intraductal infusion of very low concentrations of glycodeoxycholic acid with intravenous caerulein. This study evaluates late histopathologic changes and gross complications in this pancreatitis model compared to the traditionally used high-dose bile salt model in rats. After 14 days, histopathologic features of caerulein pancreatitis were not different from saline controls. High-dose intraductal bile salt infusion resulted in widespread chronic inflammation, acinar dilation and atrophy, marked reactive stromal proliferation, and ductular budding with periductal fibrosis. In contrast, animals receiving low-dose intraductal bile salt infusion combined with intravenous caerulein demonstrated a moderate degree of chronic inflammation and acinar atrophy along with an intermediate degree of periductal fibrosis and stromal reaction. We conclude that due to its moderate degree of injury, this model may be useful for the study of tissue injury and repair following acute pancreatitis.

Isovolemic hemodilution with dextran prevents contrast medium induced impairment of pancreatic microcirculation in necrotizing pancreatitis of the rat

BACKGROUND

Previous studies demonstrated that intravenous contrast medium (CM), as used in contrast enhanced computed tomography, aggravates the impairment of pancreatic microcirculation (PM) characteristic of severe pancreatitis and increases necrosis and mortality in necrotizing pancreatitis (NP) in rats. This study evaluates the use of isovolemic hemodilution, which can enhance the microcirculation in severe pancreatitis, for preventing CM-induced injury.

METHODS

NP was induced in 30 dextran-tolerant Wistar rats by intraductal glycodeoxycholic acid and intravenous cerulein for 6 hours. PM was quantified by intravital microscopy using fluorescein isothiocyanate labeled erythrocytes. Based on previous results, areas with low blood flow (< 1.6 nL/min/cap) were identified and baseline recordings of capillary blood flow taken. A reduction of hematocrit to 75% of baseline was achieved by replacement of 5 mL/kg of blood with 25 mL/kg Ringer's lactate (RL) or by exchange of 8 mL/kg of blood for the same amount of dextran 70.6%. Thereafter, the nonionic CM iopamidol (Solutrast, Byk Gulden, Konstanz, Germany) was injected during 1 minute and PM measurements repeated after 30 and 60 minutes.

RESULTS

Despite hemodilution with RL, pancreatic capillary perfusion was significantly decreased to 87% of baseline (0.83 +/- 0.04 mL/min/cap; n = 216) 60 minutes after CM infusion (P < 0.05). In contrast, capillary blood flow was significantly increased to 161% (1.56 +/- 0.05 nL/min/cap; n = 278) in the group treated with dextran. Moreover, the percentage of capillaries developing complete stasis was significantly lower in the dextran group (2.3 +/- 1.2%) compared to animals diluted with RL (22.3 +/- 4.8%) (P < 0.002).

CONCLUSION

Isovolemic hemodilution with dextran prevents the additional impairment of pancreatic microcirculation induced by CM in NP.

Relation of epidermal growth factor receptor and estrogen receptor-independent pS2 protein to the malignant transformation of mucinous cystic neoplasms of the pancreas

OBJECTIVE

To evaluate the role of epidermal growth factor receptor (EGF-R) and pS2 protein in the evolution of malignancy in mucinous cystic tumors of the pancreas.

BACKGROUND

Mucinous cystic tumors of the pancreas include histologically benign but premalignant mucinous cystic neoplasms and mucinous cystadenocarcinoma. The molecular events leading to transformation from a benign to a malignant mucinous tumor are not known. Overexpression of EGF-R and detection of an estrogen-induced protein (pS2) has been demonstrated in ductal adenocarcinomas of the pancreas, but these factors have not been evaluated in mucinous cystic tumors.

DESIGN

Twenty-six mucinous tumors were examined for EGF-R, pS2 protein, and estrogen and progesterone receptors.

RESULTS

Eight (61.2%) of 13 malignant tumors exhibited increased expression of EGF-R, whereas EGF-R was not detected in any of the 13 benign tumors (P = .002). The pS2 protein was detected in nine of 11 malignant and 11 of 11 benign tumors (P = .480). Estrogen and progesterone receptors were not detected in the epithelium of either tumor type. The median survival time of the patients with EGF-R-negative tumors was 29.0 months compared with 14.5 months for those with EGF-R-positive tumors, but this difference did not reach significance owing to the small population size.

CONCLUSIONS

Overexpression of EGF-R in mucinous cystic tumors, as in ductal adenocarcinomas, may be an important feature associated with malignancy and may have prognostic significance. Failure to detect EGF-R in histologically benign epithelium suggests that the upregulation of EGF-R may be important in the evolution of aggressive behavior. The expression of pS2 protein appears to be independent of estrogen and may play a role in the proliferative activity of mucinous tumors. However, pS2 expression is not a feature associated exclusively with malignancy.

Hypercalcemia causes acute pancreatitis by pancreatic secretory block, intracellular zymogen accumulation, and acinar cell injury

BACKGROUND

Because hypercalcemia is a known etiologic factor for human acute pancreatitis, studies of the pancreatic pathophysiology and pathomorphology of experimental hypercalcemia have potential clinical significance.

MATERIALS AND METHODS

Rats received central venous infusion of either 0.6 mmol/kg per hour CaCl2 or 0.9% NaCl infusion for 12 hours. Pancreatic tissue samples were obtained and prepared for electron microscopy. Tissue homogenates were examined for DNA, lactate dehydrogenase (LDH), protein, amylase, and calcium contents. Basal or stimulated (cerulein 0.25 microL/kg per hour) pancreatic secretions were analyzed for volume, protein, and amylase output, as well as protein composition on sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE).

RESULTS

The tissue calcium content and the ratio of LDH to DNA was unchanged after calcium infusion, but the ratios of total protein to DNA and of amylase to DNA were significantly larger. Basal output of pancreatic juice volume, protein, and amylase were significantly lower. SDS-PAGE of pancreatic juice revealed weakening of a 70,000-d band and appearance of lower molecular weight bands in two samples. Ultrastructural examination demonstrated accumulation of zymogen granules in the acinar cell, large autophagic vacuoles containing remnants of condensing vacuoles.

CONCLUSIONS

These findings suggest that hypercalcemia induces pancreatic injury via a secretory block, accumulation of secretory proteins, and possibly activation of proteases.

Failure of pentoxifylline to ameliorate severe acute pancreatitis in the rat: results of a prospective, randomized, controlled study

OBJECTIVE

To investigate the benefit of pentoxifylline in severe experimental pancreatitis.

DESIGN

Prospective, randomized, controlled study.

SETTING

Experimental animal laboratory in a University hospital.

SUBJECTS

Forty-two adult male Sprague-Dawley rats.

INTERVENTIONS

Acute pancreatitis was induced by supramaximal stimulation with cerulein plus a pressure and volume controlled 10 min intraductal infusion of 10-mM glycodeoxycholic acid. Thirty minutes after pancreatitis was induced, animals were randomized to receive pentoxifylline (60 mg/kg over 2.5 hrs), or saline. All animals received fluid resuscitation with lactated Ringer's solution (8 mL/kg/hr), and surviving animals were killed at 24 hrs.

MEASUREMENTS AND MAIN RESULTS

There was a progressively significant decrease in mean arterial pressure after pancreatitis was induced, with no difference between pentoxifylline-treated rats and controls. Hematocrit increased significantly in both groups at 6 hrs, and returned to baseline values at 24 hrs. Ascites volume and levels of trypsinogen activation peptide in plasma and ascites were similar in both groups. Twenty-four hour mortality was 47% for the pentoxifylline group and 52% for the control group. Histologic scores for necrosis, edema, inflammation, and hemorrhage showed no significant differences between the two groups.

CONCLUSION

Treatment with pentoxifylline failed to improve outcome in a model of severe acute pancreatitis in the rat.

Pancreatic microcirculatory changes in experimental pancreatitis of graded severity in the rat

BACKGROUND

A technique with two complementary methods, intravital microscopy (IVM) and diffuse reflectance spectroscopy (DRS), was developed to analyze pancreatic tissue perfusion.

METHODS

After initial in vivo and in vitro validation of the techniques, we studied pancreatic microcirculation in models of mild, moderate, and severe pancreatitis. Anesthetized Sprague-Dawley rats were randomly allocated to the three models or to serve as controls. Stable systemic hemodynamic parameters were maintained with normal saline solution infusion. Exocrine capillary perfusion was assessed by IVM; hemoglobin oxygenation and hemoglobin content were measured by DRS.

RESULTS

Capillary perfusion in mild pancreatitis initially increased significantly at 30 minutes to 155% +/- 38% of baseline values but returned to baseline within 3 hours. Hemoglobin content and oxygen saturation remained stable. In moderate and severe pancreatitis capillary perfusion significantly decreased versus the control group to 12% +/- 6% and 6% (range, 0% to 14%) of baseline values, respectively, at 6 hours. Oxygen saturation decreased significantly in moderate pancreatitis from 48.5% +/- 2.3% to 41.6% +/- 3.5% (p < 0.05) and in severe pancreatitis from 47.2% +/- 1.5% to 38.9% +/- 0.5% (p < 0.05), whereas hemoglobin content did not change.

CONCLUSIONS

We conclude that (1) IVM and DRS provide both unique and complementary data on tissue perfusion of the pancreas, (2) that moderate and severe experimental pancreatitis are accompanied by progressive tissue ischemia, and (3) that significant stasis (decreased perfusion) and decreased oxygen saturation occur whereas generalized vasoconstriction (decreased hemoglobin levels) was not found. In contrast, mild experimental pancreatitis was accompanied by initial hyperperfusion and normal oxygen delivery was maintained.

Exocrine hyperstimulation but not pancreatic duct obstruction increases the susceptibility to alcohol-related pancreatic injury

OBJECTIVE

To evaluate the factors thought to be involved in the pathogenesis of acute pancreatitis associated with alcohol.

BACKGROUND

The mechanism of alcohol-induced pancreatitis is believed to involve synergistic effects of various pathogenetic factors. The present study was designed to evaluate the possible contribution of pancreatic duct obstruction, physiologic exocrine stimulation, or secretory hyperstimulation to alcohol-induced pancreatic injury.

METHODS

Wistar rats were allocated randomly to a control group (group 1), or to a group with pancreatic duct obstruction (group 2), physiologic exocrine stimulation (group 3), ductal obstruction and exocrine stimulation (group 4), or exocrine hyperstimulation with the cholecystokinin analogue cerulein (group 5). Three hours after this pretreatment, animals in each experimental group were randomly divided into two subgroups for intragastric administration of either water (groups 1A through 5A) or beer (groups 1B through 5B). Test solutions were instilled over 9 hours (total amount of alcohol administered, 4.8 g/kg). Twenty-four hours after beginning the test infusion, animals were killed for histologic evaluation of pancreatic edema and determination of an acinar cell necrosis score. Serum amylase levels were determined at 3, 9, and 24 hours.

RESULTS

No increase in amylase levels or significant morphologic changes were found in control animals (group 1A) or in animals subjected to physiologic exocrine stimulation (group 2A). Pancreatic duct obstruction, with or without physiologic exocrine hyperstimulation (groups 3A and 4A), and exocrine hyperstimulation (group 5A) induced pancreatitis of similar severity with minor acinar cell damage. Alcohol superimposed on exocrine hyperstimulation (group 5B) increased acinar cell injury (group 5A, 0.4 +/- 0.1 points vs 5B, 1.0 +/- 0.2 points; P < .05) and serum amylase levels at 24 hours (group 5a, 41 +/- 6 U/L vs group 5B, 72 +/- 11 U/L; P < .05), whereas no differences between subgroups A and B (water vs beer) were found in groups 1 through 4.

CONCLUSION

Our findings suggest that the pathogenesis of acute alcoholic pancreatitis may require a state of exocrine hyperstimulation, perhaps via cholecystokinin, but do not support a role for constriction or obstruction of Oddi's sphincter.

Impairment of pancreatic microcirculation correlates with the severity of acute experimental pancreatitis

BACKGROUND

Altered microcirculatory perfusion may be an important factor in the pathogenesis of necrotizing pancreatitis. Diffuse reflectance spectroscopy (DRS) can measure dynamic alterations in the microcirculation over a larger area than intravital microscopy.

STUDY DESIGN

We used DRS to characterize changes in the microcirculation during the evolution of pancreatitis of varying severity. Male Sprague Dawley rats (330 to 400 g) were randomly allocated to four groups: control (n = 18), mild pancreatitis (n = 18), moderate pancreatitis (n = 18), or severe pancreatitis (n = 34). Within each group, rats were studied 0.5, 3.0, or 6.0 hours after induction of pancreatitis to determine total hemoglobin content (IHb) and hemoglobin oxygenation (ISO2).

RESULTS

Total hemoglobin content in the pancreas remained constant in all groups. Hemoglobin oxygenation increased significantly in rats in the control group and in rats with mild pancreatitis for the duration of the experiment, but not in rats with moderate or severe pancreatitis. Rats with severe pancreatitis had a significant decrease in ISO2 six hours after the induction of pancreatitis compared with baseline values (49.18 +/- 1.55 versus 52.01 +/- 0.19, p < 0.01) as well as rats in the control group that were studied after six hours (49.18 +/- 1.55 versus 55.92 +/- 1.07, p < 0.02). Furthermore, there was marked variability in IHb and ISO2 at different locations within the same pancreas in rats with severe pancreatitis, which was not observed in the control group or in the rats with mild or moderate pancreatitis.

CONCLUSIONS

Impaired microcirculatory perfusion characterizes severe, but not mild, pancreatitis. The predominant early change is stasis rather than vasoconstriction. As the changes become more severe, necrosis occurs in a heterogeneous distribution.

Interstitial protease activation is the central event in progression to necrotizing pancreatitis

BACKGROUND

Although intracellular protease activation is thought to be an early event in acute pancreatitis, factors determining progression from edematous to necrotizing pancreatitis are largely unknown. With enterokinase as a probe and an immunoassay quantifying free trypsinogen activation peptides (TAP), we sought evidence for the presence of interstitial trypsinogen in edematous pancreatitis and documented the effects of its ectopic activation.

METHODS

Edematous pancreatitis in the rat was induced by supramaximal stimulation with cerulein (5 micrograms/kg/hr) and coupled with enterokinase infused into the pancreatic duct at 30 mm Hg. Blue dextran infusion at this pressure corroborated interstitial delivery. Rats with no stimulation, maximal physiologic stimulation (0.25 microgram/kg/hr of cerulein), or intraductal saline infusion served as controls. TAP levels measured by enzyme-linked immunosorbent assay, 6-hour survival, and histopathology were used as end points.

RESULTS

Intraductal enterokinase infusion alone or in combination with maximal physiologic stimulation generated only slight increases in TAP level and no or minimal pancreatic injury. In contrast, enterokinase superimposed on edematous pancreatitis (supramaximal cerulein stimulation) produced fulminant pancreatitis and rapid death of all animals within 6 hours. Pancreatic histopathology showed severe intrapancreatic hemorrhage, acinar inflammation, and necrosis. TAP levels were significantly higher in plasma (p = 0.02), urine (p = 0.05), and ascites (p < 0.001) when compared with all other groups.

CONCLUSIONS

In edematous pancreatitis a large pool of trypsinogen accumulates in the interstitial space. Activation of these proenzymes leads to catastrophic consequences and may underlie progression from mild to necrotizing pancreatitis.

Time course of bacterial infection of the pancreas and its relation to disease severity in a rodent model of acute necrotizing pancreatitis

BACKGROUND

Bacterial infection of pancreatic necrosis is thought to be a major determinant of outcome in acute necrotizing pancreatitis. The determinants and possibilities for prophylaxis are unknown and difficult to study in humans.

OBJECTIVE

The time course of bacterial infection of the pancreas in a rodent model of acute necrotizing pancreatitis was characterized. The authors ascertained if there is a correlation with the degree of necrosis.

METHODS

Acute pancreatitis (AP) of graded severity was induced under sterile conditions by an intravenous infusion of cerulein (5 micrograms/kg/hr) for 6 hours (mild AP), or a combination of intravenous cerulein with an intraductal infusion of 10-mM glycodeoxycholic acid (0.2 mL for 2 min for moderate AP, 0.5 mL for 10 min for severe AP). Sham-operated animals (intravenous and intraductal NaCl 0.9%) served as controls. Ninety-six hours after induction, animals were killed for quantitative bacterial examination and histologic scoring of necrosis. In addition, groups of animals with severe AP were investigated at 12, 24, 48, 96, and 144 hours.

RESULTS

No significant pancreatic necrosis was found in control animals (0.3 +/- 0.1) or animals with mild AP (0.6 +/- 0.1) killed at 96 hours. Necrosis scores were 1.1 +/- 0.2 for animals with moderate AP and 1.9 +/- 0.2 for animals with severe AP. Control animals did not develop significant bacterial infection of the pancreas (> or = 10(3) CFU/g). At 96 hours, the prevalence of infection was 37.5% in animals with mild AP and 50% in animals with moderate AP. In animals with severe AP, infection of the pancreas increased from 33% in the first 24 hours to 75% between 48 and 96 hours (p < 0.05). The bacterial counts and the number of different species increased with time and was maximal (> 10(11) CFU/g) at 96 hours.

CONCLUSION

Bacterial infection of the pancreas in rodent AP increases during the first several days, and its likelihood correlates with the severity of the disease. This model, which closely mimics the features of human acute pancreatitis, provides a unique opportunity to study the pathogenesis of infected necrosis and test therapeutic strategies.

Intravenous contrast medium impairs oxygenation of the pancreas in acute necrotizing pancreatitis in the rat

OBJECTIVE

Contrast-enhanced computed tomography is widely used to evaluate severe acute necrotizing pancreatitis (ANP) by demonstrating areas of malperfusion, which might indicate irreversible necrosis. Because of our prior finding that the intravenous contrast medium (CM) accentuates the severity of ANP by promoting further necrosis and higher mortality, we sought to investigate the mechanism by which this injury is mediated.

DESIGN

Mild acute pancreatitis was induced in Sprague-Dawley rats with intravenous caerulein hyperstimulation; and severe ANP, with intravenous caerulein plus intraductal glycodeoxycholic acid. Control animals and rats with pancreatitis were randomized to be given intravenous CM or saline.

MAIN OUTCOME MEASURE

Diffuse reflectance spectroscopy was used to measure the index of hemoglobin content and oxygen saturation in pancreatic tissues in vivo.

RESULTS

Oxygen saturation of hemoglobin was increased in animals with mild acute pancreatitis (AP) (mean [+/- SEM], 58.7% +/- 1.2% vs 55.2% +/- 1.5% in control animals; P < .05) and was decreased in animals with ANP (51.2% +/- 1.2% vs 55.2% +/- 1.5%; P < .05). Fifteen minutes after the infusion of CM, oxygen saturation of hemoglobin significantly decreased further in animals with ANP (51.4% +/- 1.8% before infusion of CM vs 46.1% +/- 1.7% at 15 minutes; P < .05) and remained significantly below the comparable group receiving intravenous saline for the entire 60-minute test. This decrement was not observed in animals with ANP given saline or in animals with mild AP or in control animals after infusion of saline or CM. The index of hemoglobin content remained unchanged throughout the experiment in all groups.

CONCLUSIONS

The prolonged reduction of oxygen saturation of hemoglobin in the pancreas following the administration of intravenous CM in rats with severe ANP indicates that CM impairs the pancreatic microcirculation in necrotizing forms of AP. This may explain our previous finding that CM increases pancreatic injury and mortality in rodents with ANP, and it underlines our concern that the use of contrast-enhanced computed tomography early in human AP may promote the evolution of pancreatic necrosis.

Evidence for extraluminal trypsinogen activation in three different models of acute pancreatitis

BACKGROUND

Inappropriate extraluminal activation of trypsin is assumed to play a part in the pathogenesis of acute pancreatitis (AP), but proof has been elusive because active trypsin is transient and difficult to measure. We have previously shown increased levels of trypsinogen activation peptides (TAP), a direct measure of trypsin activation, to correlate with severity of AP, tissue necrosis, and survival in a rodent model induced by cerulein hyperstimulation and bile salt infusion. The present study seeks to show that increased trypsinogen activation also characterizes three other models of experimental AP in rodents to give credence to the generality of the phenomenon and to its potential relevance to human AP.

METHODS

Experimental AP was induced in mice by a choline-deficient diet supplemented with ethionine and in rats by creation of a closed duodenal loop or by ligation of the biliopancreatic duct plus physiologic stimulation. TAP were quantified by an immunoassay in tissue and plasma at various time points after onset of AP.

RESULTS

In the group with choline-deficient diet supplemented with ethionine a significant increase in tissue and plasma TAP was found at 48 and 72 hours, respectively. In the group with closed duodenal loop significant TAP elevations were found in plasma as early as 6 hours and in the group with ligation of the biliopancreatic duct plus physiologic stimulation at 24 hours.

CONCLUSIONS

These experiments provide further evidence that extraluminal protease activation is a pathophysiologic event common to the evolution of various models of experimental acute pancreatitis and therefore increase the likelihood that this phenomenon is important in the human disease as well.

Cyst fluid cytologic analysis in the differential diagnosis of pancreatic cystic lesions

Pancreatic cystic lesions include inflammatory pseudocysts, benign serous cystadenomas, and mucinous neoplasms, some of which are malignant. Cytologic analysis of cyst fluid has been proposed to diagnose pancreatic cysts before definitive therapy. The authors report an analysis of 31 pancreatic cyst aspirates: 9 pseudocysts, 5 serous cystadenomas, 8 mucinous cystic neoplasms, 4 mucinous cystadenocarcinomas, 2 papillary cystadenocarcinomas, 1 mucinous ductal adenocarcinoma with cystic degeneration, and 2 cystic islet cell tumors. All pseudocysts were correctly classified as probable inflammatory lesions, because of the presence of abundant acute inflammation and histiocytes and the absence of glandular epithelium. Three of five serous cystadenomas were correctly classified, based on the presence of small cuboidal cells in clusters with microvesicular cytoplasm containing glycogen. Eleven of 12 mucinous tumors contained round cells with large cytoplasmic mucin vacuoles or columnar cells containing cytoplasmic mucin. Malignancy was diagnosed in 5 of 7 carcinomas, 1 case was classified as suspicious for malignancy, and 1 case was nondiagnostic because of the absence of a cellular component. The authors concluded that pancreatic cyst fluid cytologic analysis is useful in differentiating mucinous from nonmucinous pancreatic cysts and may provide definitive evidence of malignancy. In some cases, serous cystadenoma can be diagnosed based on cytologic analysis. An inflammatory smear without epithelial cells suggests a pseudocyst, but these findings are nonspecific, as a similar pattern may occur when a cystic neoplasm undergoes degenerative changes. Therefore, pseudocyst remains a diagnosis of exclusion.

Failure of symptomatic relief after pancreaticojejunal decompression for chronic pancreatitis. Strategies for salvage

OBJECTIVE

To evaluate causes of intractable recurrent pain following pancreaticojejunostomy for chronic pancreatitis and to evaluate treatment strategies aimed at lasting pain relief.

DESIGN

Case series.

SETTING

Tertiary care referral center.

PATIENTS

Fifteen selected patients having severe pain associated with chronic pancreatitis with onset 0 to 60 months (median, 5 months) following pancreaticojejunostomy. Each patient underwent computed tomography and endoscopic retrograde cholangiopancreatography. Two patients (13%) were found to have pancreatic cancer, two (13%) had inadequate pancreatic duct decompression, two (13%) had biliary stenosis, and 10 (67%) had presumed neuropathy in the pancreatic head.

INTERVENTIONS

Fourteen (93%) of the 15 patients underwent the following reoperations: distal pancreatectomy and splenectomy (one patient), extension of the pancreaticojejunostomy and choledochojejunostomy (one patient), biliary stenting followed by choledochojejunostomy (one patient), and Whipple-type resection of the pancreatic head (14 patients). Two patients subsequently underwent a completion pancreatectomy.

MAIN OUTCOME MEASURES

Pain relief, functional capacity, and pancreatic exocrine and endocrine status were determined. The median follow-up after reoperation was 39 months.

RESULTS

Of the 14 patients who underwent reoperation, 13 were long-term survivors. One died of pancreatic cancer. Ten of the other 13 have had satisfactory-to-excellent relief of pain, with resumption of a normal level of function. Of the 10 previously euglycemic patients who underwent pancreatic head resection, eight remain free of diabetes mellitus to date.

CONCLUSIONS

The causes of recurrent or persistent pain following pancreaticojejunal decompression for chronic pancreatitis are complex and include neuropathic changes, residual or evolving pancreatic and biliary duct obstruction, and unrecognized pancreatic cancer. Acceptable outcomes can usually be achieved by following a treatment strategy aimed at addressing identified residual disease while maximally preserving pancreatic tissue.

A rat model to study hypercalcemia-induced acute pancreatitis

Hypercalcemia causes acute pancreatitis in humans, a phenomenon reproduced experimentally in cats and guinea pigs. Because the rat is the most frequently used animal for the study of experimental pancreatitis, the present studies were performed to evaluate the effects of hypercalcemia in the rat. In in vitro studies, pancreatic lobules were prepared from fasted Wistar rats (200-250 g) and incubated in HEPES bicarbonate-buffered medium (pH 7.4) containing 0, 0.6, 1.2, 2.5, 5, and 10 mM CaCl2 with or without carbachol 10(-6) M. Amylase was measured in the medium after 30 min to 3 h, and expressed as percent of total amylase. In in vivo studies, fasted male Wistar rats (300-400 g) received calcium (CaCl2; 0.6 mmol/kgh) into the tail vein for 12 h. Control animals received NaCl 0.9% infusion. Histologic slides (H&E-stained) were evaluated in a blinded fashion. Pancreatic lobules showed a higher basal amylase output when incubated in higher calcium medium. The largest, significant difference (2.6-fold) was between 0.6 and 5 mM medium CaCl2 (p < 0.05). Carbachol-stimulated amylase release was again higher with increasing medium calcium with the most pronounced difference (1.3-fold) between 0.6 and 2.5 mM CaCl2 (p < 0.05). In vivo calcium-treated animals showed accumulation of zymogen granules in the cytoplasm, cytoplasmic vacuolization, focal acinar cell depolarization, acinar necrosis, and edema. Calcium causes amylase release from rat pancreatic lobules in vitro. Higher medium calcium levels both significantly increase amylase release from unstimulated and carbachol stimulated lobules. Twelve-hour in vivo calcium infusion leads to accumulation of zymogen granules in acinar cells and acinar injury.(ABSTRACT TRUNCATED AT 250 WORDS)

[Trypsinogen activation peptides in acute pancreatitis. Experimental data and clinical implications]

Trypsinogen activation peptides (TAP) can be found in patients with acute pancreatitis (AP) as well as in different models of experimental AP. First experience has suggested that early elevation of TAP indicates development of necrotizing AP and that the amount of TAP correlates with the extent of acinar cell necrosis and mortality. It is however unknown whether TAP similarly assess severity of AP in the later course of the disease. The present study monitores TAP in plasma, urine and ascites during the initial development of pancreatic injury and correlates the amount of TAP and the extent of pancreatic necrosis over 48 h in a rodent model of AP. While there was no elevation of TAP in control animals or animals with edemantous AP, significant amounts of TAP im plasma were found as early as 30 min following induction of severe necrotizing AP. Serum amylase returned to normal values within 24 h, TAP maintained at high levels until the end of the 48 h observation period. During the first 24 h TAP paralleled the development of acinar cell necrosis, but decreased thereafter despite further progression of pancreatic injury. Our results provide further evidence suggesting that TAP initially precede morphological changes in AP. Early in the course of the disease TAP parallel development of pancreatic injury.(ABSTRACT TRUNCATED AT 250 WORDS)

Expression of CA 72-4 (TAG-72) in the fluid contents of pancreatic cysts. A new marker to distinguish malignant pancreatic cystic tumors from benign neoplasms and pseudocysts

OBJECTIVE

The authors evaluated cyst fluid CA 72-4 as a tumor marker in the differential diagnosis of pancreatic cystic lesions.

SUMMARY BACKGROUND DATA

Pancreatic cystic lesions include inflammatory pseudocysts, serious cystadenomas, and mucinous tumors. Mucinous tumors can be further subdivided into mucinous cystadenocarcinomas and premalignant mucinous cystic neoplasms. The clinical and radiologic features of these lesions are unreliable to make a preoperative diagnosis of these diagnostically difficult lesions. Analysis of aspirated cyst fluid was proposed as an aid to making the preoperative differential diagnosis. Currently, a number of parameters have been reported as useful markers in cyst fluid aspirates, including the tumor markers carcinoembryonic antigen and CA 15.3, enzymes (amylase, lipase, and amylase isoenzymes), relative viscosity, and cytologic analysis. However, owing to the rarity of pancreatic cystic tumors, experience with cyst fluid analysis is limited. To define additional markers that might be useful in the differential diagnosis of pancreatic cysts, the authors measured the tumor-associated glycoprotein 72 (TAG-72) in aspirates from 19 pancreatic cystic lesions.

METHODS

Cyst fluid from 19 pancreatic cysts was obtained by needle aspiration. The tumor marker TAG-72 was measured by a commercial (CA 72-4) immunoassay.

RESULTS

Cyst fluid CA 72-4 levels in mucinous cystadenocarcinomas were markedly elevated (mean, 10,027 U/mL; range, 780 to 34,853 U/mL) compared with that in pseudocysts (mean, 3.8 U/mL; range, < 3 to 5.7 U/mL) and serous cystadenomas (mean and range, < 3 U/mL; p < 0.001). The level of CA 72-4 in benign mucinous cystic neoplasms was intermediate (mean, 44.2 U/mL; range, < 3 to 137 U/mL), but it was statistically different from either carcinomas (p = 0.009) or benign cysts (p < 0.001).

Expression of CA 15.3 protein in the cyst contents distinguishes benign from malignant pancreatic mucinous cystic neoplasms

BACKGROUND

Inflammatory pseudocysts, serous cystadenomas, and mucinous cystic tumors comprise most cystic lesions of the pancreas. The mucinous tumors include mucinous cystic neoplasms, which are benign on histologic examination but have the potential for malignant transformation, and mucinous cystadenocarcinomas. Accurate preoperative classification of pancreatic cysts with clinical and radiologic criteria is often impossible, and as a result many cystic tumors are inappropriately treated as pseudocysts. Analysis of percutaneous needle-aspirated cyst fluid for tumor markers, enzymes, viscosity, and cytologic study has been proposed as a useful modality to distinguish mucinous from nonmucinous cystic lesions. However, no reliable cyst fluid parameter distinguishes benign from malignant mucinous tumors.

METHODS

The concentration of the tumor marker CA 15-3 was measured by immunoassay in cyst fluid from six pseudocysts, five serous cystadenomas, three mucinous cystic neoplasms, and six mucinous cystadenocarcinomas.

RESULTS

The concentration of CA 15-3 in the cyst fluid of mucinous cystadenocarcinomas was higher (mean, 178 units/ml; range, 40 to 392) than that of mucinous cystic neoplasms (mean, 4.7 units/ml; range, 0 to 14), serous cystadenomas (mean, 9.2 units/ml; range, 0 to 32), and pseudocysts (mean, 15.3 units/ml; range, 0 to 66). An upper cutoff value of 30 units/ml distinguished mucinous cystic neoplasms from mucinous cystadenocarcinomas with 100% sensitivity and 100% specificity (p = 0.01).

CONCLUSIONS

Production of CA 15-3 appears to coincide with malignant transformation in pancreatic mucinous cystic neoplasms. We conclude that measurement of CA 15-3 levels in the cyst fluid is useful in the differentiation of benign from malignant pancreatic mucinous cysts.

Intravenous contrast medium accentuates the severity of acute necrotizing pancreatitis in the rat

BACKGROUND/AIMS

Contrast-enhanced computed tomography (CECT) is used to show areas of decreased pancreatic perfusion in severe acute pancreatitis (AP). To evaluate possible adverse effects of the contrast medium (CM) on the course of AP, the impact of intravenous CM in AP of graded severity in the rat was studied.

METHODS

Pancreatitis of three levels of severity was induced in Sprague-Dawley rats with intravenous cerulein hyperstimulation plus time- and pressure-controlled intraductal infusion of saline or glycodeoxycholic acid. At 7 hours, control and pancreatitis animals received intravenous ionic CM, nonionic CM, or saline. The principal outcome measures were 24-hour survival, trypsinogen activation peptides (TAP) in ascites, and histological acinar necrosis score.

RESULTS

There was no measurable effect of CM on the index features in control animals or animals with mild or moderate AP. In severe AP, CM caused a significant increase in mortality, ascites TAP, and necrosis score.

CONCLUSIONS

Intravenous CM increases pancreatic injury when administered early in the course of severe experimental AP. Because CM may convert borderline ischemia to irreversible necrosis, CECT performed early in pancreatitis to show poor perfusion and predict areas of necrosis may depict a self-fulfilling prophecy. Early CECT should be reconsidered and perhaps avoided.

Variability in cyst fluid carcinoembryonic antigen level, fluid viscosity, amylase content, and cytologic findings among multiple loculi of a pancreatic mucinous cystic neoplasm

The procedure of percutaneous aspiration and analysis of cyst contents has been advocated to provide a preoperative diagnosis of pancreatic cystic lesions (pseudocysts and cystic tumors), but it is not known whether variation in the contents of separate loculi of a multilocular neoplasm might misrepresent the identity of the tumor. The authors measured the cyst fluid carcinoembryonic antigen (CEA) level, fluid viscosity, and amylase content and performed cytologic analysis on aspirates rates from ten different loculi of a single mucinous cystic neoplasm of the pancreas. The CEA levels were highly variable (median, 6,326 ng/mL; range, 962-64,670 ng/mL) but in all cases were diagnostic of a mucinous tumor. Fluid relative viscosity values were also variable (median, 2.4; range, 1.3-10+) but diagnostic in eight of nine aspirates. The amylase content in all of the loculi was low (< 91 U/L), and values were consistent with a cystic tumor. Cytologic analysis showed diagnostic mucin-secreting epithelial cells in nine of ten loculi. Although cytologic examination was nondiagnostic in one loculus, there were no false-positive results for malignancy. The combination of all four tests would not have resulted in misclassification of any of the tumors. The authors conclude that the characteristics of the contents of different loculi of pancreatic cystic neoplasms may be variable, but the use of a combination of tests still ensures accurate diagnosis.

Cystic neoplasms of the pancreas: evaluation with endoscopic retrograde pancreatography

Images from endoscopic retrograde pancreatography performed in 38 patients with cystic pancreatic neoplasms were reviewed to determine whether they might allow differentiation of malignant from benign lesions. Twenty-three patients had histologically proved benign lesions; 15 had frankly malignant lesions. Images were classified into four groups on the basis of the appearance of the pancreatic duct: (a) normal, (b) showing narrowing or draping; (c) showing communication with a cystic space, and (d) showing obstruction. Of the 23 patients with benign lesions, eight had normal studies, 11 had studies showing narrowing or draping, three had studies showing communication, and one had a study showing obstruction. Of the 15 patients with malignant lesions, one had a normal study, three had studies showing narrowing or draping, two had studies showing communication, and nine had studies showing obstruction. Overall, there was a significant difference in the rates of pancreatographic findings for benign and malignant tumors (P = .001). Although no finding proved to be an absolute indicator of benign or malignant disease, obstruction occurred with 60% of malignant tumors and was seen with only one benign lesion (4%).

Evaluation and validation of microsphere technique for determination of pancreatic blood flow

The purpose of this study was to evaluate the radiolabeled microsphere technique for pancreatic blood flow measurements. Using a canine model with an isolated pancreatic circulation, we assessed the shunting of 11- and 15-microns-diam microspheres in the pancreas, correlated pancreatic blood flow measurements obtained with the microsphere technique with those made with an ultrasonic flow probe, and determined the effects of high doses of microspheres on pancreatic blood flow and its measurement. Microspheres of 11 microns demonstrate significant shunting through the pancreatic microcirculation with underestimation of pancreatic blood flow of approximately 10% compared with results obtained with 15-microns microspheres. There is a close linear relationship between flow results obtained with 15-microns microspheres and with an ultrasonic flow probe for both the resting (r = 0.85) and the secretin-stimulated pancreas (r = 0.97). Left atrial injections of very large doses of 15-microns microspheres (50 x 10(6)) caused an acute transient reduction of pancreatic blood flow (to 39% of baseline flow) with a return to baseline values within 2 min. The accuracy of flow results and absence of physiological changes after repeated injections support the use of 15-microns microspheres for pancreatic blood flow measurements.

Portal vein thrombosis after elective splenectomy. An underappreciated, potentially lethal syndrome

We describe seven cases of thrombosis of the portal and splenic vein after elective splenectomy. The diagnosis was initially unrecognized in all patients and was confused with biliary sepsis (three cases), postoperative pancreatitis (three cases), or pulmonary emboli (one case). Two patients in whom the diagnosis of portal vein thrombosis was not made within 3 days of disease onset died. In the five survivors, the diagnosis was based on clinical suspicion, confirmed with color flow Doppler ultrasonography or computed tomography with intravenous contrast material, and treated with thrombolytic agents, anticoagulants, and antibiotics. In two patients, splenic vein thrombus was visualized on initial postoperative imaging studies and the thrombus subsequently extended into the portal vein. Portal vein thrombosis should be considered in patients with fever and abdominal complaints after splenectomy. Urgent treatment with thrombolysis and anticoagulants may preserve bowel integrity and be lifesaving.

Intraluminal pancreatic candidiasis presenting as recurrent pancreatitis

An 88-year-old man with closely spaced attacks of acute pancreatitis who was found to have ductal changes of chronic pancreatitis with multiple noncalcified intraluminal filling defects during endoscopic retrograde pancreatography is presented. These defects proved to be fungus balls made up of Candida albicans. He was treated with longitudinal pancreaticojejunostomy and oral fluconazole and has since remained recurrence free (30 months). It is suggested that Candida superinfection may occur in a chronically dilated pancreatic duct and may contribute to symptomatic recurrent inflammation of the pancreas.

Portal vein thrombosis after elective splenectomy. An underappreciated, potentially lethal syndrome

We describe seven cases of thrombosis of the portal and splenic vein after elective splenectomy. The diagnosis was initially unrecognized in all patients and was confused with biliary sepsis (three cases), postoperative pancreatitis (three cases), or pulmonary emboli (one case). Two patients in whom the diagnosis of portal vein thrombosis was not made within 3 days of disease onset died. In the five survivors, the diagnosis was based on clinical suspicion, confirmed with color flow Doppler ultrasonography or computed tomography with intravenous contrast material, and treated with thrombolytic agents, anticoagulants, and antibiotics. In two patients, splenic vein thrombus was visualized on initial postoperative imaging studies and the thrombus subsequently extended into the portal vein. Portal vein thrombosis should be considered in patients with fever and abdominal complaints after splenectomy. Urgent treatment with thrombolysis and anticoagulants may preserve bowel integrity and be lifesaving.

Gut macromolecular permeability in pancreatitis correlates with severity of disease in rats

BACKGROUND

Increased intestinal macromolecular permeability could allow absorption of substances from the bowel into the systemic circulation and contribute to multiple organ system failure.

METHODS

Mild, intermediate, and severe grades of pancreatitis were induced in rats using intravenous caerulein and intraductal glycodeoxycholic acid. [14C]polyethylene glycol (molecular weight, 3350 daltons; 1.1 microCi/142 mg) was instilled into the distal duodenum. At 24 hours, the animals were killed, ascitic fluid was collected for trypsinogen activation peptide measurement, and pancreatic specimens were collected and scored for based on the degree of necrosis, inflammation, and hemorrhage.

RESULTS

Gut permeability to polyethylene glycol 3350 (PEG 3350) was increased in animals with early experimental pancreatitis (5.4% +/- 1.2%, n = 20) when compared with control animals (1.8% +/- 0.2%; n = 6) (P = 0.0005). Furthermore, intestinal macromolecular permeability to PEG 3350 correlated with severity of disease as predicted by the method of induction of pancreatitis (P = 0.0003), the histological findings (P = 0.0002), and total ascitic trypsinogen activation peptides content (P = 0.029).

CONCLUSIONS

Increased gut permeability in experimental pancreatitis can be correlated with pancreatitis severity.

Factors associated with successful laparoscopic cholecystectomy for acute cholecystitis

OBJECTIVE

This article determined which preoperative data correlated with successful completion of a laparoscopic cholecystectomy in patients with acute cholecystitis.

SUMMARY BACKGROUND DATA

Although laparoscopic cholecystectomy is the procedure of choice in chronic cholecystitis, its use in acute cholecystitis may be associated with higher costs and complication rates. It is not known which patients with acute cholecystitis are likely to require conversion to open cholecystectomy based on preoperative data or if a cooling-off period with medical therapy can diminish inflammation and increase the chance of successful laparoscopic cholecystectomy.

METHODS

All laparoscopic cholecystectomies done by the authors between 10/90 and 2/92 were reviewed. Data on cases of acute cholecystitis were prospectively collected on standardized data forms.

RESULTS

Twenty of 281 laparoscopic cholecystectomies were done for acute cholecystitis; 7/20 patients with acute cholecystitis required conversion to open cholecystectomy compared with 6/281 patients undergoing elective operation for chronic cholecystitis. In patients with acute cholecystitis the interval from admission to cholecystectomy in the successful cases was 0.6 days vs. 5 days in the cases requiring conversion to open cholecystectomy (p = .01). Cases requiring conversion to open cholecystectomy also had higher WBC (14.0 vs. 9.0, p < .05), alkaline phosphatase (206 vs. 81, p < .02, and APACHE II scores (10.6 vs. 5.1, p < .05). Ultrasonographic findings such as gallbladder distention, wall thickness, and pericholecystic fluid did not correlate with the success of laparoscopic cholecystectomy. Patients converted from laparoscopic to open cholecystectomy required more operating room time (120 min vs. 87 min, p < .01) and more postop hospital days (6 vs. 2, p < .001).

CONCLUSIONS

Laparoscopic cholecystectomy for acute cholecystitis should be done immediately after the diagnosis is established because delaying surgery allows inflammation to become more intense, thus increasing the technical difficulty of laparoscopic cholecystectomy.

Resection margins in carcinoma of the head of the pancreas. Implications for radiation therapy

A retrospective review of the pathology and clinical course of 72 patients undergoing resection of carcinoma of the head of the pancreas was undertaken to identify the frequency of tumor involvement at standard surgical transection margins (stomach, duodenum, pancreas, and bile duct) as well as the peripancreatic soft tissue margin and the potential clinical significance of these findings. Of 72 patients undergoing resection, 37 patients (51%) were found to have tumor extension to the surgical margins. The most commonly involved margin was peripancreatic soft tissue (27 patients) followed by pancreatic transection line (14 patients) and bile duct transection line (4 patients). For 37 patients with tumor present at a resection margin, there were no survivors beyond 41 months. No difference in survival or local control was seen between 14 patients receiving postoperative radiation therapy and 5-fluorouracil (5-FU) compared with 23 patients not receiving additional treatment. In contrast, the 5-year actuarial survival and local control of 35 patients undergoing resection without tumor invasion to a resection margin was 22% and 43%, respectively. The 5-year survival and local control of 16 patients receiving adjuvant radiation therapy and 5-FU was 29% and 42%, respectively, whereas these figures were 18% and 31% for 19 patients not receiving adjuvant therapy (p > 0.10). Because residual local tumor after resection is common, preoperative radiation therapy may be beneficial in this disease. It should minimize the risk of dissemination during operative manipulation and facilitate a curative resection by promoting tumor regression. Because local failure rates approach 60% after resection and adjuvant therapy even in cases having clear resection margins, intraoperative radiation therapy to the tumor bed at the time of resection also might be considered. Protocols evaluating the feasibility and efficacy of preoperative radiation therapy and resection with intraoperative radiation therapy for patients with pancreatic cancer are underway.