Pancreatic Duct Obstruction Is an Aggravating Factor in the Canine Model of Chronic Alcoholic Pancreatitis

Transcription

1 GASTROENTEROLOGY 1998;115: Pancreatic Duct Obstruction Is an Aggravating Factor in the Canine Model of Chronic Alcoholic Pancreatitis TSUNEO TANAKA,* YOSHIO MIURA,* YASUHIRO MATSUGU,* YASUYUKI ICHIBA,* HISAO ITO, and KIYOHIKO DOHI* *Department of Surgery II, Hiroshima University School of Medicine, Hiroshima; and Department of Pathology I, Tottori University School of Medicine, Tottori, Japan Background & Aims: Chronic alcoholic pancreatitis occurs in only a limited number of heavy drinkers. Other factors than alcohol are necessary for the occurrence of chronic alcoholic pancreatitis. The aim of this study was to examine whether pancreatic duct obstruction resulted in increased alcohol-induced parenchymal cell damage. Methods: Four groups of adult mongrel dogs were used. In group A, 2.0 g kg 1 day 1 of ethanol was administered via a gastric cannula. In group O, after ligation of the minor pancreatic duct, a polyethylene tube was inserted transduodenally into the major duct. In group AO, the protocols used in groups A and O were combined. Laparotomy was repeated after 3 months in each group. Results: Three of the 9 dogs in group AO had pancreatic calculi in the main pancreatic duct. Moderate interlobular fibrosis, parenchymal cell loss, and inflammatory cell infiltration resembling human chronic alcoholic pancreatitis were observed in group AO. Little change was observed in groups A and O. Exocrine function assessed by secretin test in group AO was significantly reduced. Total protein, hexosamine, and calcium contents of the pancreatic juice in group AO were significantly increased. Conclusions: Pancreatic duct obstruction is an aggravating factor in chronic alcoholic pancreatitis. The pathophysiology underlying chronic alcoholic pancreatitis (CAP) remains unknown. The lack of a useful experimental model for CAP may be one of the reasons why the mechanism is unknown, although many experimental models 1 3 have been reported. Sarles et al. 1 gave 20% ethanol to rats for 2 years, and Tiscornia et al. 2 administered ethanol to dogs for 3 years. However, only initial CAP lesions were produced in these animal models. 1 3 In humans, the period of alcohol intake necessary for the development of CAP has been estimated to be more than 10 years. Taking into consideration the life span of experimental animals such as rats and dogs, it is therefore impossible to prepare animal CAP models identical in histopathology to that of human CAP. Sarles et al. 1 reported that protein plugs were present in all the small pancreatic ducts in the initial lesions of human CAP, and Bordalo et al. 4 reported that ductal hypertrophy and reduplication caused obstruction of the small pancreatic ducts. Therefore, it is clear that pancreatic duct obstruction plays an important role in the pathogenesis of CAP. We have endeavored to produce an experimental model of chronic pancreatitis 5 7 and have adopted the combination theory 6 as the basis for the development of new experimental models of chronic pancreatitis. This theory proposes that persistent parenchymal cell damage is essential for the development of chronic pancreatitis, and that pancreatic duct obstruction exacerbates the parenchymal cell damage. The appropriate combinations of these two factors are believed to be important in the development of chronic pancreatitis models that are representative of the various stages and grades of this disease. We have already succeeded in producing an animal model that has a histological profile closely resembling that of human chronic pancreatitis. 6 This model was developed by combining chronic ischemia with pancreatic duct obstruction. Our experimental observations have subsequently suggested that pancreatic duct obstruction aggravates pancreatic lesions due to the presence of chronic ischemia. Although the pancreatic parenchymal cell damage produced by alcohol is less severe than that produced by ischemia, duct obstruction may nonetheless induce congestion and accumulation of pancreatic juice. This in turn may increase the cell damage caused by alcohol consumption. The aim of the present study was to determine whether pancreatic duct obstruction resulted in increased parenchymal cell damage after alcohol administration. A model of alcohol administration with incomplete pancreatic duct obstruction was developed, and the pancreas was studied using pancreatic function tests, pancreatic juice Abbreviation used in this paper: CAP, chronic alcoholic pancreatitis by the American Gastroenterological Association /98/$3.00

2 November 1998 MODEL OF CHRONIC ALCOHOLIC PANCREATITIS 1249 analysis, pancreatic tissue blood flow rates, pancreatography, and histopathologic examination. Materials and Methods Animals and Experimental Design Thirty-six adult mongrel dogs weighing kg were used. The animals were fed standard dog chow (Oriental Yeast, Tokyo), and the food was withheld for 24 hours before the experiment. Surgical manipulations were performed aseptically under anesthesia with 20 mg/kg sodium pentobarbital (Nembutal). These animals were subdivided into four groups. The control group (n 9) underwent laparotomy only. In group A (n 9), after creation of a gastric fistula, 2.0 g kg 1 day 1 ethanol was administered via a gastric cannula. In group O (n 9), after ligation of the minor pancreatic duct, the duodenum was incised to insert a polyethylene tube (0.4 mm ID, 1.7 mm OD, length of 7 mm) into the major duct. In group AO (n 9), both of the surgical manipulations described for groups A and O were performed (Figure 1). The laparotomy was repeated after 3 months in each group. Pancreatic Function Tests Serum amylase levels in each group were examined for 3 months using the blue starch method (Amylase Test A; Pharmacia Diagnostics, Stockholm, Sweden). After a duodenotomy, the previously inserted tube was removed and a new tube (5F) was inserted via the papilla. The minor ducts of the control and ethanol-treated dogs were ligated before the exocrine function tests. After 30 minutes of basal secretion collection, secretin (SecrepanR; Eizai, Tokyo, Japan) was drip-infused intravenously at 3 U kg 1 h 1. The pancreatic juice was then collected at intervals of 10 minutes for 60 minutes. The flow rate of the pancreatic juice was also measured, and the bicarbonate concentration was determined using a Natelson s manometer. The amylase concentration was determined by the blue starch method. Figure 1. Schema of the preparation performed in alcohol administration with incomplete duct obstruction. Intravenous glucose tolerance test was performed by injecting 0.5 g/kg of glucose. Blood was collected at 5-, 10-, 15-, 30- and 60-minute intervals, and serum glucose levels were measured by a Glucose Analyzer 23A (Yellow Springs Instrument, Yellow Springs, OH). Pancreatic Juice Analysis The pancreatic juice collected during the secretin test was analyzed. The total protein content was determined by the Lowry method, 8 hexosamine content by the Winzler method, 9 and calcium content by the orthocresolphthalein complex method. 10 Pancreatic Tissue Blood Flow Blood flow rates in both lobes were measured using an electrolytic blood flow meter (RBF-1, Biochemical Science, Tokyo, Japan) before the secretin test. Pancreatography After resection of the pancreas and duodenum en masse, pancreatography was performed immediately. The contrast medium (microbarium) was introduced from a height of 10 cm to equalize the pressure as much as possible. The extracted pancreas was then placed on an X-Omat TL film (Kodak, Rochester, NY) for roentgenography at 300 ma, 28 kv(p), 3.0 seconds, and at a focal-film distance of 60 cm. Histological Findings For the histopathologic examination, 8 tissue samples were collected from both the left and right lobes of all dogs and were stained with H&E. A pathologist (H.I.), who works in another institution, evaluated the slides in a blinded manner. Fibrosis, parenchymal cell loss, inflammatory cell infiltration, fat replacement, and tubular complex formation were divided into three grades. The classification of the histopathologic findings was performed according to Howard and Nedwich as follows 11 : (1) minimal pancreatitis, incipient coarse lobulation as a result of increased interstitial connective tissue fibrosis with relatively well-preserved acini and minimal infiltration by inflammatory cells; (2) moderate pancreatitis, a progressive increase in interstitial fibrosis with a further loss of acini; and (3) advanced pancreatitis, marked fibrosis with a proliferation of thick bundles replacing the acinar parenchyma. Statistical Analysis The results are expressed as means SEM. Mann Whitney U test and Fisher s exact P value test were used to compare data from two experimental groups. The results were considered to be statistically significant at P 0.05.

3 1250 TANAKA ET AL. GASTROENTEROLOGY Vol. 115, No. 5 Results Macroscopic Findings No changes were detected in group A. Mild hardening of the pancreas was observed in group O. The hardening of the pancreas was more advanced in group AO than in group O. Three of the 9 dogs in group AO (not significantly different from controls, P ) had pancreatic calculi in the main pancreatic duct (Figure 2). These calculi contained more than 97% CaCO 3. Pancreatic Function Tests Serum amylase. No significant changes in serum amylase levels were observed at any time in any of the groups. Secretin test. In group A, no significant differences in secretin test results were recognized. In group O, amylase output was significantly reduced compared with control. In group AO, flow rate, bicarbonate concentration, and amylase output were significantly reduced compared with the control (Figure 3). Endocrine function test. No significant differences in the intravenous glucose tolerance test results were found among the experimental groups. Pancreatic Juice Analysis In group A, total protein content was significantly increased compared with control. In group O, no significant changes were recognized. In group AO, the total protein content, hexosamine content, and calcium content were significantly increased compared with control (Figure 4). Figure 3. Secretin test in each group. Cont., control group; A, alcohol administration group; O, duct obstruction group; AO, group A O. *P 0.05; **P Pancreatic Tissue Blood Flow The pancreatic tissue blood flow rate in group AO was significantly decreased compared with the control group (Figure 5). Pancreatography In the control group and group A, the pancreatic duct was distinctly visible up to the fourth branch, and no abnormalities were detected. In group O, the main pancreatic duct was diffusely dilated with mild irregularities and the tertiary and quaternary bifurcations were not distinctly delineated. In group AO, the main duct was even more dilated with clear irregularities and the number of branches was decreased. Histopathologic Findings The control group and group A (Figure 6A) showed no evidence of histological changes. Mild periductal fibrosis was recognized in group O (Figure 6B), but the degree of fibrosis was less than 10%. In group AO, the fibrosis extended into the interlobular region, and Figure 2. Pancreatic calculi (arrow) in group AO (alcohol administration with duct obstruction). Figure 4. Pancreatic juice analysis in each group. Cont., control group; A, alcohol administration group; O, duct obstruction group; AO, group A O. *P 0.05; **P 0.01.

4 November 1998 MODEL OF CHRONIC ALCOHOLIC PANCREATITIS 1251 Figure 5. Pancreatic tissue blood flow in each group. Cont., control group; A, alcohol administration group; O, duct obstruction group; AO, group A O. **P parenchymal cell loss with chronic inflammatory cell infiltration was also observed (Figure 6C). Furthermore, the degree of fibrosis in group AO animals was 30% 40%. A summary of the histological findings is shown in Table 1. Using Howard s classification, 11 the histology of the AO group corresponded to moderate pancreatitis. There were individual differences in the extent of pancreatic fibrosis in each dog. Discussion The histopathology of chronic pancreatitis is characterized by fibrosis, parenchymal cell loss, and inflammatory cell infiltration into the pancreas. The fibrosis in the AO group was mainly characterized by interlobular fibrosis. In the group with pancreatic duct obstruction (group O), only mild fibrosis in the periphery of the pancreatic duct could be seen. By coadministering alcohol (group AO), interlobular fibrosis was induced. Suda et al., 12,13 reported that the pancreatic fibrosis in alcoholics was characterized by intralobular fibrosis, and that the fibrosis in patients with CAP and clinical symptoms was characterized by interlobular fibrosis, with little fibrosis within the lobules. They also suggested that the pathogeneses of both types of fibrosis are different, which is in agreement with the results of our present experiment. The degree of pancreatic fibrosis in group AO (30% 40%) was greater than in groups C (0%), A (0%), and O ( 10%). In other words, the degree of pancreatic fibrosis induced by alcohol administration was enhanced through obstruction of the pancreatic duct. In a study by Sarles et al., 14 secretin was given to dogs given alcohol for more than 2 years; the investigators reported an increase in the flow rate of the pancreatic juices and amylase output. However, Yahata 15 did not find any changes in exocrine pancreatic function when the dogs were given alcohol for 18 months. In the present study, reductions in three parameters were found: the Figure 6. Photomicrograph of a dog in (A ) alcohol administration group (group A), (B) duct obstruction group (group O), and (C) group AO (original magnifications: A and B,40 ;C,20 ).

5 1252 TANAKA ET AL. GASTROENTEROLOGY Vol. 115, No. 5 Table 1. Summary of Histological Findings in Each Group Groups Fibrosis Periductal Interlobular Intralobular Parenchymal cell loss Inflammatory cell infiltration Fat replacement Control A O AO A, alcohol administration group; O, duct obstruction group; AO, group A O;, none;, slightly;, moderate. Tubular complex flow rate of the pancreatic juice, bicarbonate concentration, and amylase output. These changes were noted after only 3 months of investigation, and thus this also fits the criteria for diagnosing exocrine pancreatic function in human CAP. However, on the basis of the present investigation, it is not possible to determine which of these three factors started to decrease first, and whether there exists a state of hypersecretion. Many reports have stated that, clinically, the reduction in exocrine function observed in CAP starts initially with a reduction in bicarbonate or amylase output. Based on the results of the pancreatic juice analysis, only group AO had an increase in protein concentration. This suggests that when pancreatic function is impaired, protein concentration increases reciprocally with the reduction in the amount of pancreatic juice. The hexosamine concentration was also increased in groups A and AO, which received alcohol. Mucoproteins are one of the protective mechanisms of the body against deleterious substances and act as a barrier even in the pancreatic duct. Alcohol is excreted into the pancreatic juice as it is in the blood, 16 and therefore an increase in hexosamine concentration may occur to protect a barrier in the duct. Geobell et al. 17 reported that the form of calcium secretion in the pancreas was driven by enzymatic proteins and that regardless of the form of secretion, there was secretion of calcium from the pancreatic wall. The increase in calcium content that was observed in group AO could be caused by an impairment of the pancreatic wall, because there was no increase in the enzymatic protein levels. On the other hand, Reber et al. 18 reported an increase in the permeability of the pancreatic duct, and that the cause of this increase was assumed to be an impairment of the barrier mechanism of the pancreatic wall induced by alcohol. However, in their experiment, there was no increase in calcium concentration after either administration of alcohol only or stenosis of the pancreatic duct only. For calcium concentration to increase, a combination of both treatments appears to be essential. Group AO in the present study is the first report of experimental pancreatolithiasis with CAP. A mechanism of the formation of these stones is proposed as follows. Alcohol intake causes an increase in mucoprotein concentration in the pancreatic juice, and the viscosity of the pancreatic juice is thus enhanced. If an obstruction of the pancreatic duct system occurs simultaneously, then the alcohol in the pancreatic juice will be retained for a longer period in the pancreatic duct. This will further increase the concentration of mucoproteins, and hence a protein plug will be formed, thereby rendering the pancreatic duct system more vulnerable to closure. Because there is also an increase in calcium concentration, calcium is deposited on the protein plug and the condition progresses to pancreatolithiasis. We have advocated, 19,20 and verified experimentally, 19 that one important cause of chronic pancreatitis is a reduction in pancreatic tissue blood flow. 19 The results on the quantity of pancreatic blood flow in the present CAP model are of great interest with respect to the mechanism of the onset of CAP. It is known that by increasing the alcohol concentration in the blood, the hepatic blood flow is also increased but the pancreatic blood flow is significantly reduced. 21 In addition, Widdison et al. 22 examined a model of obstructive chronic pancreatitis and reported that an increase in interstitial pressure induced a reduction in pancreatic tissue blood flow, and that coadministration of alcohol further enhanced and prolonged this reduction in blood flow. Takahashi and Taginuma 23 classified the impairment of blood flow into two categories: a central injury and a peripheral injury. If there is stenosis of the intrapancreatic artery, which is the terminal artery, then central injury occurs and pancreatic infarction is manifested. On the other hand, if there is stenosis of the central artery, peripheral blood flow is decreased. In the present model, the results suggest that peripheral injury may have occurred without pancreatic infarction. Because the pancreatic blood flow was measured before secretin stimulation in this study, the effects of secretin stimulation were unclear. However, blood flow was shown to be unaffected by physiological doses of secretin in the conscious dog. 24 Based on the results of this study, alcohol appears to be a very weak inducer of pancreatic parenchymal tissue impairment. However, it has also been demonstrated that the addition of mild stenosis of the pancreatic duct results in greater pancreatic injury of several orders of magnitude

6 November 1998 MODEL OF CHRONIC ALCOHOLIC PANCREATITIS 1253 in potency. We have suggested two mechanisms by which this occurs. First, the concentration of alcohol in the pancreatic juice rapidly reaches the same concentration as in the blood, and the addition of even mild stenosis of the pancreatic duct causes a retention of alcohol in the pancreatic juice over long periods of time. As a result, the barrier function of the wall of the pancreatic duct is impaired, thus causing a change in the properties of the pancreatic juice. Second, there is obstruction of the pancreatic duct in addition to alcohol administration, and the quantity of pancreatic tissue blood flow will be reduced; in other words, peripheral blood flow is reduced. This results in the induction of interlobular fibrosis of the pancreas, with inflammatory cell infiltration. Sarles et al. 1 hypothesized that in the mechanisms responsible for the onset of CAP, the protein plugs first block the small pancreatic ducts. However, we have suggested that for these protein plugs to obstruct the small pancreatic ducts, obstruction of the main pancreatic duct system, which is the retention of pancreatic juice, is necessary (the new big duct theory 25 ). In addition, it is assumed that a dysfunction of the papilla of Vater is probably present in human CAP. Clinical cases similar to our experimental model have already been reported. In cases of pancreas divisum, the dorsal pancreatic duct (Santorini duct) is known to be more prone to the development of relative stenosis 26 because it is surrounded by a larger volume of pancreatic parenchymal tissue than the accessory papilla. Therefore, obstructive pancreatitis occurs more frequently in the dorsal pancreas. It has also been reported that when alcohol is ingested in cases of pancreas divisum, the region near the dorsal duct develops moderate to severe CAP, whereas the region near the ventral duct undergoes fewer changes. 27 This clinical observation supports both our experimental results and our hypothesis. References 1. Sarles H, Lebreuil G, Tasso F, Figarella C, Clemente F, Devaux MA, Fagonde B, Payan H. A comparison of alcoholic pancreatitis in rat and man. Gut 1971;12: Tiscornia O, Palasciano G, Sarles H. Effects of chronic ethanol administration on canine exocrine pancreatic secretion. Digestion 1974;11: Singh M, Lasure MM, Bockman DE. Pancreatic acinar cell function and morphology in rats chronically fed an ethanol diet. Gastroenterology 1982;82: Bordalo O, Gonelves D, Noronha M. Newer concept for the pathogenesis of chronic alcoholic pancreatitis. Am J Gastroenterol 1977;68: Tanaka T, Ichiba Y, Fujii Y, Kodama O, Dohi K. Clinical and experimental study of pancreatic exocrine function after pancreaticoduodenectomy for periampullary carcinoma. Surg Gynecol Obstet 1988;166: Tanaka T, Ichiba Y, Fujii Y, Ito H, Kodama O, Dohi K. New canine model of chronic pancreatitis due to chronic ischemia with incomplete pancreatic duct obstruction. Digestion 1988;41: Tanaka T, Ichiba Y, Miura Y, Ito H, Dohi K. Canine model of chronic pancreatitis due to chronic ischemia. Digestion 1994;55: Lowry OH, Rosebrough NJ, Farr AL, Randall RJ. Protein measurement with the Folin phenol reagent. J Biol Chem 1951;193: Winzler RJ. Determination of serum glycoprotein. Methods Biochem Anal 1955;2: Connerty HV, Briggs A. Determination of serum calcium by means of orthocresolphthalein complex. Am J Clin Pathol 1966;45: Howard JM, Nedwich A. Correlation of the histologic observations and operative findings in patients with chronic pancreatitis. Surg Gynecol Obstet 1971;132: Suda K, Shiotsu H, Nakamura T. Pancreatic fibrosis in patients with chronic alcohol abuse and correlation with alcoholic pancreatitis. Am J Gastroenterol 1994;89: Suda K, Takase M, Takei K. Histopathologic study on coexistent pathologic states in pancreatic fibrosis in patients with chronic alcohol abuse. Two distinct pathologic fibrosis entities with different mechanisms. Pancreas 1996;12: Sarles H, Tiscornia O, Palasciano G. Chronic alcoholism and canine exocrine pancreas secretion. Gastroenterology 1977;72: Yahata K. Pathogenesis of early pancreatic lesions in chronic alcoholic pancreatitis; experimental study on long-term alcoholfed dogs. J Jpn Panc Soc 1986;1: Beck IT, Paloschi GB, Dinda PK, Beck M. Effect of intragastric administration of alcohol on the ethanol concentrations and osmolality of pancreatic juice, bile, and portal and peripheral blood. Gastroenterology 1974;67: Geobell AM, Nan LW, Fischer R. Effect of ethanol on pancreatic and biliary secretions in humans. Dig Dis 1973;18: Reber HA, Roberts C, Way LW. The pancreatic duct mucosal barrier. Am J Surg 1979;137: Tanaka T, Ichiba Y, Miura Y, Itoh H, Dohi K. Pathogenesis of chronic pancreatitis. Am J Gastroenterol 1993;88: Tanaka T, Ichiba Y, Miura Y, Itoh H, Dohi K. Is ischemia an etiological factor in chronic pancreatitis (letter)? Am J Gastroenterol 1991;86: Friedman HS, Lowery R, Shaughnessy E, Scorza J. The effects of ethanol on pancreatic blood flow in awake and anesthetized dogs. Proc Soc Exp Biol Med 1983;174: Widdison AL, Alvarez C, Schwarz M, Reber HA. The influence of ethanol on pancreatic blood flow in cats with chronic pancreatitis. Surgery 1992;112: Takahashi T, Yaginuma N. Ischemic injury of the human pancreas. Its basic patterns correlated with the pancreatic microvasculature. Pathol Res Pract 1985;179: Chung RS, Safaie-Shirazi S. The effect of secretin on pancreatic blood flow in the awake and anesthetized dog. Proc Soc Exp Biol Med 1983;173: Tanaka T, Ichiba Y, Miura Y, Ito H, Dohi K. Pathogenesis of chronic alcoholic pancreatitis. Am J Gastroenterol 1990;85: Gregg JA. Pancreas divisum: its association with pancreatitis. Am J Surg 1977;134: Cotton PB. Congenital anomaly of pancreas divisum as cause of obstructive pain and pancreatitis. Gut 1980;21: Received November 10, Accepted August 11, Address requests for reprints to: Tsuneo Tanaka, M.D., Department of Surgery II, Hiroshima University School of Medicine, Kasumi, Minamiku, Hiroshima 734, Japan. Fax: (81)