- Email: [email protected]
The clinical and histopathological effects of pancreatic duct occlusion in experimental acute pancreatitis in dogs
JOURNAL
OF SURGICAL
RESEARCH
34,
164- 170 ( 1983)
The Clinical and Histopathological Effects of Pancreatic Duct Occlusion in Experimental Acute Pancreatitis in Dogs R. J. A.
ESTOURGIE,
S.
H. YAP,
*J
U . J. G.
VAN HAELST,~
AND
H. H. M. DE BOER
Department of Surgery, *Division of Gastrointestinal and Liver Disease, Department of Medicine, and TDepartment of Pathology, St. Radboud Hospital, University of Nijmegen, 6500 HB Nijmegen, The Netherlands Submitted for publication
August 20, 1982
A continuous production of significant pancreatic enzymes, which are thought to be responsible for the maintenance of the digesting process, is frequently found in fulminant necrotizing pancreatitis. Since the medical therapies known to be effective are based upon the rationale of slowing pancreatic secretion, a simple measure which permits the “burning out” of residual pancreatic tissue might therefore have a therapeutic value. In this study, 2 hr after the induction of acute hemorrhagic pancreatitis, 5 dogs (Group 1) were treated with 1.5 ml Ethibloc injected into the pancreatic duct; 5 other animals (Group II) were given 1.5 ml saline; Group III (5 dogs) had no treatment. All animals in Group II and 4 of the 5 animals in Group III expired within 8 days postoperatively. In contrast, 4 of 5 animals from Group I survived. Although some of the biochemical parameters showed significant changes after the induction of acute pancreatitis, no differences were seen between the three groups. In the expired animals, the picture of histological examination was that of a fulminant acute hemorrhagic pancreatitis of the left lobe. In the survival dogs although normal pancreatic tissue was present in the right lobe at necropsy at intervals, there was always a pancreatic atrophy of the left lobe and striking adhesions with the surrounding tissues suggesting the severity of the disease in the acute phase. These findings suggest that pancreatic duct occlusion causing the exocrine secretion to stop may have beneficial effects in the treatment of acute fulminant pancreatitis in the acute phase and may improve the survival rate.
INTRODUCTION
Despite the significant progress in the area of diagnostic aids and intensive care treatment in the past decade, the mortality rate of acute hemorrhagic necrotizing pancreatitis is still high. Standard textbooks cited mortality rates of 40 to 50% [8]. The major antecedent factors associated with pancreatitis have been shown to be alcoholism and gallstones. However, the mechanism by which these antecedents trigger pancreatic inflammation and the factors determining the progression of “mild” to “fulminant” necrotizing pancreatitis are still obscure. Once triggered, the pathogenesis of tissue injury is brought about by a multitude of agents. Activated pancreatic proteolytic enzymes are known to be among the agents for digesting the pancreas and surrounding tissues as al’ To whom correspondence
should be addressed.
0022-4804/83/020164-07$01.50/0 Copyright Q 1983 by Academic Press, Inc. All rights ofreprdutiion in any form resewed.
164
ready suggested by Chiari in 1896 [4], but they may also be responsible for activation of other enzymes that may provoke some of the systemic effects of pancreatitis [5]. The primary cause of death in the early phase of acute pancreatitis after an attack appears to be the cardiovascular collapse, postulated to be mediated by circulating vasoactive substances [6, 121. In experimental pancreatitis these vasoactive toxins are found in abundance in the peritoneal exsudate and are readily absorbed into the bloodstream [ 1, 241. Since continuous production of pancreatic exsudate has been found frequently in acute necrotizing pancreatitis, a total pancreatectomy in the early phase has been suggested to afford protection [9, 141.However, radical dissection of pancreas and surrounding tissues in this acute illness has a high mortality rate. In this regard we thought if we were able to reduce the production of pancreatic enzymes by a simple technique,
ESTOURGIE
ET AL.:
the deteriorated effects of toxic pancreatic exsudate may be reversed. Our present study was designed to observe the effects of pancreatic duct occlusion by means of intraductal administration of polymerized agents, a mixture of zein and sodium-amidotrizoat-tetrahydrate (Ethibloc, see under Materials and Methods) on the clinical features and the histopathological changes in dogs with experimentally induced acute pancreatitis.
DUCT
OCCLUSION
165
dogs were kept continously under anesthesia. Two hours after the infusion of the bile-trypsin mixture, the duodenum was reopened and the major pancreatic duct was recannulated. The dogs were then divided in 3 groups. The first group was treated with a 1.5-ml mixture of zein and sodium-amidotrizoat-tetrahydrate (Ethibloc) injected into the pancreatic duct by hand for several minutes; the second group was given 1.5 ml saline intraductally; and the third group had no treatment and served a control group. Ethibloc was purchased from Ethicon MATERIALS AND METHODS Gmbh Norderstedt, West Germany. It conFifteen female beagle dogs maintained on tains 2 10 mg zein, 162 mg sodium-amidoa standard kennel diet were used in this trizoat-tetrahydrate, 145 mg oleum papavstudy. They were anesthetised with Thiopen- et-is, and 6 mg propylene glycol/ml. It has tal Sodium 15 mg/kg body wt, 30 min after been used for mechanical embolisation or a 5-mg Droperidol and 0.1-mg Fentanyl pre- occlusion, it polymerizes immediately in wet medication given intramuscularly. The body conditions and hardens to a firm rubbery weight ranged from 7.4 to 12.9 kg. After an material. The occlusion gel is resorbed after endotracheal intubation, controlled respira- 10 days in the pancreatic duct. tion (Engstrom respirator system 300) took After withdrawal of the cannula, the duoplace with a mixture of Oz, N20, and halo- denotomy was closed with one layer of No. thane. During the operation all dogs had an 4-O silk interrupted inverting sutures and the intravenous infusion of 500 ml saline. There abdominal wall was closed in layers. Water was a continuous registration of arterial was allowed on the first day following opblood pressure, ECG, pulse rate, and body eration and a standard kennel diet after 24 temperature. hr. In cases of dehydration and frequent Under sterile conditions the duodenum vomiting, fluid was given either orally using and attached pancreas were exposed through a stomach tube (milk) or intravenously (glua midline upper abdominal incision. Bile was cose 5%). collected aseptically by puncturing the gallBefore and immediately after the surgical bladder and stored at 4°C. The minor pan- procedure blood was drawn for gas analysis creatic duct beyond the duodenum was dou- and determination of blood sugar, serum amble ligated and cut. Through a 3-4-cm-long ylase, albumin, calcium, leucocyte count, antimesenteric duodenotomy the major pan- hemoglobin, hematocrit, transaminases creatic duct was then cannulated using a (GOT, GPT), lactate dehydrogenase (LDH), polyethylene cannula Longdwel 20 gauge. urea nitrogen, and creatinine. In the survivUnder a constant pressure of 40 cm HZ0 a ing dogs repeated biochemical estimations mixture of 1 ml bile/kg body wt, containing were made several times during the first week 2000 BAEE units of trypsin (T 8253 Sigma and weekly thereafter for 4 weeks. The enChemical Co.) in each mililiter of bile (the docrine function of the pancreas was also mixture of bile and trypsin had been incu- assessedat intervals in the survivors by meabated at 37°C for 16 hr), was infused in the suring the intravenous glucose tolerance tests pancreatic duct. The mean infusion time in and the circulating plasma insulin (radio imthis study was 27 min. After the intraductal munoassay) after the glucose overload. infusion, the cannula was withdrawn and the At the time of death an autopsy was perduodenotomy was temporally closed. The formed on all the dogs. The pancreas and
166
JOURNAL
OF SURGICAL
RESEARCH:
duodenum were dissected for histological examination. Special attentions were made at the sites of macroscopic changes, however, microscopic examinations were also performed in the right as well as in the left lobe, the papilla, and in the central part of the pancreas. RESULTS
Almost immediately after the induction of acute pancreatitis by infusion of the incubated bile-trypsin mixture into the pancreatic duct, the surface of the pancreas showed edema with brownish pigmentation. The color of the pancreatic tissue usually became darker than the bile administered to induce the pancreatitis. After a period of 30 min or so, the pancreas could be seento weep a bloody fluid from its surface into the abdominal cavity. Areas of ecchymotic, petechial, and frank hemorrhage could also be observed in the gland and peripancreatic tissue after 1 hr. There were also exudations present along the blood vesselsin omentum and mesenterium of the duodenum. Before the laparotomy incision could be closed, the left lobe of the pancreas was usually very dark and identification of lobuli was sometimes difficult. Table 1 shows the fatality rate of our model of acute pancreatitis as described under Materials and Methods and the effect of intraductal injection of saline and of Ethibloc in these experimental animals. In the first 10
TABLE
1
THE INFLUENCE OF PANCREATIC DUCT OCCLUSION WITH ETHIBL~C AND THE EFFECT OF INTRADU~TAL ADMINISTRATION OF SALINE ON THE MORTALITY RATE OF EXPERIMENTAL ACUTE PANCREATITIS IN Does Group I (Ethibloc) II (Saline) III (No treatment) Total
Survivals 4 0 1 5
Nonsurvivals 1 5 4 10
Total 5 5 5 15
VOL. 34, NO. 2, FEBRUARY TABLE
1983 2
BKXHEMICAL ANALYSIS OF BLEND COLLECI%D 8 HR BEFORE DEATH OF THE ANIMAL FROM THE ETHIBLOC-TREATED GROUP
Hemoglobin (mmol/l) Hematocrit (l/l) Glucose (mmol/l) Urea-N (mmol/l) Creatinine (pmol/l) Calcium (mmol/l) Albumin (g/l) Amylase (u/l) GOT (u/l) GPT (u/l) LDH (u/l)
Obtained finding
Normal value
13.9 0.68 0.30 15.0 668.0 2.46 23.0 25,900 19 41 264
8-11 0.35-0.50 4.6-6.3 <6.6 t68 2.30-2.80 25-33 13000 <40 160 <180
days after operation, all dogs, if they survived, were very ill with weakness,occasional vomiting, and anorexia. All animals in Group II (saline) and 4 of the 5 animals in Group III (control group) expired within 8 days postoperatively. No obvious difference was observed in the survival time between Groups II and III, however, one animal from Group III survived. After the complete recovery and intravenous feeding in the first week after operation, the dog was sacrificed on the tenth day for necropsy. In contrast to Groups II and III, 4 of the 5 animals of Group I with experimentally induced acute pancreatitis treated by pancreatic duct occlusion survived. In these animals long-term follow up biochemical studies including intravenous glucose tolerance test and plasma insulin measurements had been performed at intervals. These dogs were killed at 42, 7 1, 162, and 2 19 days after the operation for histopathological examination. One of the Ethibloc-treated animals expired 28 hr after the operation. From the retrospective results of biochemical investigation of blood collected 20 hr after the operation as demonstrated in Table 2 the severe hypoglycemia was probably responsible for the acute coma and resulted in the death of this animal.
167
ESTOURGIE ET AL.: DUCT OCCLUSION O-amylase “/I x 1000
30 n-
t
t
1
2
3
4
3 hours after operation before operation
5
6
7
8
9 10 11 12 days after operation
FIG. 1. The course of serum amylase activity in the surviving dogs. These animals were treated with Ethibloc intraductally 2 hr after the induction of acute hemorrhagic pancreatitis.
Biochemical Investigation For the biochemical examinations of the three groups of animals, the blood was collected in all animals before and 1 hr after the end of the operation (i.e., 3 hr after the induction of pancreatitis). Although some of the parameters showed statistically significant changes (Student’s t test) after the induction of acute pancreatitis, no obvious differences were seenbetween the three groups. Plasma calcium, serum albumin, and leucocyte counts were decreased after the induction of pancreatitis in all animals, while the serum amylase activity and GOT were higher. Further investigation on the differencesof the biochemical parameters between the three groups was not possible because 36 hr after the operation there were only 2 dogs from Groups II and III alive. Figure 1 illustrates the course of serum amylase activity in the surviving dogs. The peak values were found 2 days after the operation. After 7
days, the serum amylase returned to approaching normal range. To investigate the endocrine function of the pancreas in the surviving animals, intravenous glucose tolerance testsand plasma insulin responseafter the intravenous glucose administration (0.5 mg/kg body wt injected for 2 min) were performed at intervals ranging from 1 to 31 weeks postoperatively. The results (data not shown) were found to be within normal limits except in one dog. In this animal, the peak value of insulin was found to be significantly lower; however, the blood sugar was normal and there was no evidence of glucosuria.
Histopathological Findings At autopsy of the 7 animals expired within 36 hr after the operation (1 from Group I, 3 from Group II, and 3 from Group III), the left lobe as well as the right lobe of the pancreas were always involved by the disease.
168
JOURNAL OF SURGICAL RESEARCH: VOL. 34, NO. 2, FEBRUARY 1983
However, the hemorrhagic necrosis of the right lobe was only moderate or absent in some dogs. The abdominal cavities were uniformly filled with quantities of bloody ascitic fluid in all animals. The pancreas was dark green to black and there was peripancreatic fat necrosis. The gross picture was that of the fulminating acute hemorrhagic pancreatitis. Microscopically there were interstitial and acinar hemorrhage, edema, necrosis of exocrine and endocrine cells, fat necrosis, and there was only slight leucocytic infiltration in the interstitial space. Although there were congestions in some sections, the blood vesselswere usually normal. Except in the ethiblot-treated animals, the main pancreatic ducts and the papilla were also not affected. In these animals, no differences in histopathological findings could be found between the saline-treated and untreated groups. In the Ethibloc-treated animals the ducts were filled with an eosinophilic mass (Ethiobloc). Disruption of the epithelial lining cells was evident. There were also periductal leucocytic infiltrations. These findings were also found in healthy dogs treated with duct occlusion with Ethibloc. The autopsy of the animals that expired between 36 hr and 8 days after operation (2 from Group II and 1 from Group III) showed evidences of peripancreatic fat necrosis and inflammation. However, interstitial fibrosis, replacement of pancreatic gland by fibrous tissue and extensive adhesions had also been observed. In cases of surviving dogs (4 from Group I and 1 from Group III) sacrificed at intervals ranging from 10 days to 32 weeks after operation, a pancreatic atrophy of the left lobe was found. Fat necrosis was not seen macroscopically. The pancreas had been replaced by fibrous tissue. There were striking adhesions around the pancreas and in the peripancreatic tissue. However, surviving segmentsof normal pancreas (exocrine and endocrine cells) were usually present in the right lobe and in the central part of the “residual pancreatic tissue.” In all animals of Ethibloc-treated group, no occlusive material in pancreatic ducts was observed.
DISCUSSION
In studying the effect of therapy in experimental pancreatitis, it is necessaryto utilize a suitable model, which would develop pancreatitis uniformly with the same prognosis. In our present study, we have used the model of acute, necrotizing pancreatitis as described by Musa et al. [ 161. However, to avoid the rupturing of the duct system in our experiments, we infused the bile-trypsine mixture under a constant pressure of 40 cm HZ0 instead of injecting by hand for 15 min. The mixture had been incubated for 16 hr at 37°C as described by Elliott et al. [ 111. Although all animals in the studies of Musa and Elliott were dead after the induction of pancreatitis, one dog from our control group (Group III, no treatment) given the same amount of trypsinized bile, survived. This finding confirms that the method of administration of bile-trypsin mixture plays also an important role in determining the absolute lethality in an acute pancreatitis model [ 11, 161. In a retrospective study [ 181 in patients with acute pancreatitis, it has been shown that a high mortality rate was associated with biochemical parameters on admission including age, blood sugar, WBC, LDH, and SGOT. Falling of hematocrit, arterial PO? and serum calcium levels, an anion base deficit, increased BUN, and fluid sequestration in the first 48 hr after admission worsened the prognosis. There is no medical therapy that has been proved by control studies to alter favorably the mortality rate of acute necrotizing pancreatitis. Surgery has a legitimate role in the management of acute pancreatitis, especially for the debridement of necrotic tissue or for drainage of pancreatic exsudate or abscesses and pseudocysts. Drainage of the peritoneal exsudate to remove the vasoactive toxins which are responsible for the cardiovascular collapse resulted in clinical improvement in experimental animals [20, 211 as well as in patient studies [ 19, 231. A total pancreatectomy in early phase has also been suggested to afford protection (9, 141.However, radical
ESTOURGIE ET AL.: DUCT OCCLUSION
dissection of pancreas and surrounding tissues has a high mortality rate. Eggink [lo] reported a series of 17 patients with a mortality rate of 7 1% after segmental left resection. Necrosis of the pancreas parenchyma is the predominant finding in acute necrotizing pancreatitis; however, a total necrosis is only found in 10% of the cases.In most patients, when the surface of the whole pancreas is necrotic, there are still segmentsof vital tissue present in the gland, especially around the main pancreatic duct [3]. Since the medical therapies known to be effective are based upon the rationale of slowing pancreatic secretion (putting the pancreas at rest) and a continued production of significant pancreatic enzymes is frequently found in fulminant necrotizing pancreatitis, a simple measure which permits the “burning out” of residual pancreatic tissue might therefore be effective to alter the natural history of this diseasewith poor prognosis. Total duct obstruction with resultant exocrine atrophy offers a theoretically appealing alternative to most of the direct pancreatic procedures to stop pancreatic enzyme production. It is well known that total duct obstruction leads to exocrine atrophy without pancreatitis and perhaps with preservation of endocrine elements. The advent of the acrylate glue and of Ethiobloc (alcoholic prolamin solution) has made possible a new technique of total pancreatic duct obstruction. The use of these substancesto obstruct the pancreatic duct in healthy experimental animals has been reported [2, 7, 15, 171. In these studies, a noninflammatory atrophy has been produced with preservation of the endocrine gland. Encouraging results and evidences of clinical improvement have also been obtained in patients with chronic pancreatitis [ 13, 15, 221. In our present study, we demonstrate that injection of Ethibloc in the pancreatic duct for a total occlusion, 2 hr after the induction of acute pancreatitis in dogs, resulted in an improvement of the survival rate. One of the 5 animals, however, expired 28 hr after the operation. The prox-
169
imate cause of death in this animal was probably not directly related to the toxic cardiovascular collapse as found in the expired control dogs. The observation of the changesof pancreas and surrounding tissue after induction of the pancreatitis and before the laparotomy incisions were closed, suggeststhat the severe inflammatory process was uniformly found in the experimental animals in the acute phase. Despite the fact that the gross picture at necropsy of expired animals was that of acute fulminating pancreatitis, segments of normal pancreatic tissue especially in the right lobe has been found in surviving dogs. As has been found in normal dogs after duct occlusion with Ethibloc, the pancreas of these animals was replaced largely by fibrous tissue. However, the presence of adhesions around the pancreas and in the peripancreatic tissue was very striking compared to those found in normal dogs treated with Ethibloc. The finding that the right lobe of the pancreas in all dogs had little pathologic alterations is due to the fact of the special anatomy of the pancreas in dog and the method of inducing pancreatitis in our experiments. The pancreas in the dog is boomerang shaped. The cannula for infusion of the bile-trysin mixture and for injection of saline or Ethibloc lies always in the left lobe; the infused fluid came therefore only retrograde or through narrow anastomotic ducts into the right lobe. These findings are supported by the results of experiments in healthy animals treated with Ethibloc (to be reported in a separate paper). Although further studies and many details in the mechanism remain to be established, the results of our present study suggest that pancreatic duct occlusion causing exocrine secretion to stop may have beneficial effects in the treatment of acute fulminant pancreatitis in the acute phase and may improve the survival rate. ACKNOWLEDGMENT The authors thank Dr. J. C. M. Hafkenscheid for performing the biochemical analysis.
170
JOURNAL OF SURGICAL RESEARCH: VOL. 34, NO. 2, FEBRUARY 1983
REFERENCES 1. Anderson, M. C., Schoenfeld, F. B., Iams, W. B., and Suwa, M. Circulatory changes in acute pancreatitis. Surg. Clin. North Amer. 47: 121, 1961. 2. Baumgartner, D., Hanson, S. L., and Sutherland, D. E. R. Endocrine function and histology of pancreatic segments after duct occlusion with ethibloc or neoprene: A comparative study. Eur. Surg. Res. 13: 33, 1981. 3. Becker, V. Akute Pankreatitis, Morfologie, Pathogenese, Prognose. Chirurg 51: 351, 1980. 4. Chiari, H. Uber Selbstverdauung des menschlichen Pankreas. Z. Heilk. 17: 69, 1896. 5. Creutzfeldt, W., and Schmidt, H. Aetiology and pathogenesis of pancreatitis (current concepts). Stand. J. Gastroenterol. Suppl. 6: 47, 1910. 6. Donaldson, L. A., Williams, R. W., and Schenk, W. G. Experimental pancreatitis: Effect of plasma and dextran on pancreatic blood flow. Surgery 84: 313, 1978. I. Dubemard, J. M., Traeger, J., Neyra, P., Touraine, J-L., Devonec, M., Blanc-Brunat, N., and Ruitton, A. Une nouvelle m&hode de pr6paration due greffon pancr6atique en vue de la transplantation. Chirurgie 104: 242, 1978. 8. Dtlrr, G. H. K. Acute pancreatitis. In H. T. Howat and H. Sarles (Ed%), The Exocrine Pancreas. London: Saunders, Chap. 15. 1979. 9. Edelmann, G., and Boutelier, Ph. Le traitement des pancreatites aigti6s necrotisantes par l’ablation chirurgicale precose des portions nbcros&. Chirurgie loo: 155, 1974. 10. Eggink, W. F. De operatieve behandling van pankreatitis. Thesis. Rotterdam 1981. 11. Elliott, D. W., Williams, R. D., and Zollinger, R. M. Alterations in the pancreatic resistanceto bile in the pathogenesis of acute pancreatitis. Ann. Surg. 146: 669, 1957. 12. Ellison, E. C., Pappas, T. N., Johnson, J. A., Fabri,
P. J., and Carey, L. C. Demonstration and characterization of the hemoconcentrating effect of as-
citic fluid that accumulates during hemorrhagic pancreatitis. J. Surg. Res. 30: 24 1, 1981. 13. Gebhardt, Ch., and Gall, F. B. PartiBle Duodenopankreatektomie mit intraoperativer Pankreasschwanzverijdung bei chronicher Pankreatitis. Langenbecks Arch. Chir. 353: 57, 1980. 14. Henry, L. G., and Condon, R. E. Ablative surgery for necrotizing pancreatitis. Am. J. Surg. 131: 125, 1916.
15. Little, J. M., Lauer, C., and Hogg, J. Pancreatic duct obstruction with an acrylate glue. A new method for producing pancreatic exocrine atrophy. Surgery 81: 243, 1977.
16. Musa, B. E., Nelson, A. W., Gilette, E. L., Ferguson, H. L., and Lumb, W. V. A model to study acute pancreatitis in the dog. J. Surg. Res. 21: 5 1, 1976. 17. Papachristou, D. N., D’Agostino, H., and Fortner, J. G. Ligation of the pancreatic duct in pancreatectomy. &it. J. Surg. 67: 260, 1980. 18. Ranson, J. H. C., Rilkind, K. M., and Turner, J. W. Prognostic signs and nonoperative peritoneal lavage in acute pancreatitis. Surg. Gynecol. Obstet. 143: 209, 1976.
19. Ranson, J. H. C. Acute pancreatitis. Curr. Probl. Surg. 16: 11, 1979. 20. Rodgers, R. E., and Carey, L. C. Peritoneal lavage in experimental pancreatitis in dogs. Amer. J. Surg. 111: 192, 1966. 21. Rosato, E. F., Mullis, W. F., and Rosato, F. E. Peri-
toneal lavage therapy in hemorrhagic pancreatitis. Surgery 74: 106, 1913. 22. R&ch, W., Phillip, J., and Gebhardt, Ch. Endoscopic duct obstruction in chronic pancreatitis. Endoscopy 1: 43, 1979. 23. Stone, H. H., and Fabian, T. C. Peritoneal dialysis in the treatment of acute alcoholic pancreatitis. Surg. Gynecol. Obstet. 150: 818, 1980. 24. Takada, Y., Appert, H. E., and Howard, J. M. Vascular permeability induced by pancreatic exsudate formed during acute pancreatitis in dogs, Surg. Gynecol. Obstet. 143: 779, 1976.
OF SURGICAL
RESEARCH
34,
164- 170 ( 1983)
The Clinical and Histopathological Effects of Pancreatic Duct Occlusion in Experimental Acute Pancreatitis in Dogs R. J. A.
ESTOURGIE,
S.
H. YAP,
*J
U . J. G.
VAN HAELST,~
AND
H. H. M. DE BOER
Department of Surgery, *Division of Gastrointestinal and Liver Disease, Department of Medicine, and TDepartment of Pathology, St. Radboud Hospital, University of Nijmegen, 6500 HB Nijmegen, The Netherlands Submitted for publication
August 20, 1982
A continuous production of significant pancreatic enzymes, which are thought to be responsible for the maintenance of the digesting process, is frequently found in fulminant necrotizing pancreatitis. Since the medical therapies known to be effective are based upon the rationale of slowing pancreatic secretion, a simple measure which permits the “burning out” of residual pancreatic tissue might therefore have a therapeutic value. In this study, 2 hr after the induction of acute hemorrhagic pancreatitis, 5 dogs (Group 1) were treated with 1.5 ml Ethibloc injected into the pancreatic duct; 5 other animals (Group II) were given 1.5 ml saline; Group III (5 dogs) had no treatment. All animals in Group II and 4 of the 5 animals in Group III expired within 8 days postoperatively. In contrast, 4 of 5 animals from Group I survived. Although some of the biochemical parameters showed significant changes after the induction of acute pancreatitis, no differences were seen between the three groups. In the expired animals, the picture of histological examination was that of a fulminant acute hemorrhagic pancreatitis of the left lobe. In the survival dogs although normal pancreatic tissue was present in the right lobe at necropsy at intervals, there was always a pancreatic atrophy of the left lobe and striking adhesions with the surrounding tissues suggesting the severity of the disease in the acute phase. These findings suggest that pancreatic duct occlusion causing the exocrine secretion to stop may have beneficial effects in the treatment of acute fulminant pancreatitis in the acute phase and may improve the survival rate.
INTRODUCTION
Despite the significant progress in the area of diagnostic aids and intensive care treatment in the past decade, the mortality rate of acute hemorrhagic necrotizing pancreatitis is still high. Standard textbooks cited mortality rates of 40 to 50% [8]. The major antecedent factors associated with pancreatitis have been shown to be alcoholism and gallstones. However, the mechanism by which these antecedents trigger pancreatic inflammation and the factors determining the progression of “mild” to “fulminant” necrotizing pancreatitis are still obscure. Once triggered, the pathogenesis of tissue injury is brought about by a multitude of agents. Activated pancreatic proteolytic enzymes are known to be among the agents for digesting the pancreas and surrounding tissues as al’ To whom correspondence
should be addressed.
0022-4804/83/020164-07$01.50/0 Copyright Q 1983 by Academic Press, Inc. All rights ofreprdutiion in any form resewed.
164
ready suggested by Chiari in 1896 [4], but they may also be responsible for activation of other enzymes that may provoke some of the systemic effects of pancreatitis [5]. The primary cause of death in the early phase of acute pancreatitis after an attack appears to be the cardiovascular collapse, postulated to be mediated by circulating vasoactive substances [6, 121. In experimental pancreatitis these vasoactive toxins are found in abundance in the peritoneal exsudate and are readily absorbed into the bloodstream [ 1, 241. Since continuous production of pancreatic exsudate has been found frequently in acute necrotizing pancreatitis, a total pancreatectomy in the early phase has been suggested to afford protection [9, 141.However, radical dissection of pancreas and surrounding tissues in this acute illness has a high mortality rate. In this regard we thought if we were able to reduce the production of pancreatic enzymes by a simple technique,
ESTOURGIE
ET AL.:
the deteriorated effects of toxic pancreatic exsudate may be reversed. Our present study was designed to observe the effects of pancreatic duct occlusion by means of intraductal administration of polymerized agents, a mixture of zein and sodium-amidotrizoat-tetrahydrate (Ethibloc, see under Materials and Methods) on the clinical features and the histopathological changes in dogs with experimentally induced acute pancreatitis.
DUCT
OCCLUSION
165
dogs were kept continously under anesthesia. Two hours after the infusion of the bile-trypsin mixture, the duodenum was reopened and the major pancreatic duct was recannulated. The dogs were then divided in 3 groups. The first group was treated with a 1.5-ml mixture of zein and sodium-amidotrizoat-tetrahydrate (Ethibloc) injected into the pancreatic duct by hand for several minutes; the second group was given 1.5 ml saline intraductally; and the third group had no treatment and served a control group. Ethibloc was purchased from Ethicon MATERIALS AND METHODS Gmbh Norderstedt, West Germany. It conFifteen female beagle dogs maintained on tains 2 10 mg zein, 162 mg sodium-amidoa standard kennel diet were used in this trizoat-tetrahydrate, 145 mg oleum papavstudy. They were anesthetised with Thiopen- et-is, and 6 mg propylene glycol/ml. It has tal Sodium 15 mg/kg body wt, 30 min after been used for mechanical embolisation or a 5-mg Droperidol and 0.1-mg Fentanyl pre- occlusion, it polymerizes immediately in wet medication given intramuscularly. The body conditions and hardens to a firm rubbery weight ranged from 7.4 to 12.9 kg. After an material. The occlusion gel is resorbed after endotracheal intubation, controlled respira- 10 days in the pancreatic duct. tion (Engstrom respirator system 300) took After withdrawal of the cannula, the duoplace with a mixture of Oz, N20, and halo- denotomy was closed with one layer of No. thane. During the operation all dogs had an 4-O silk interrupted inverting sutures and the intravenous infusion of 500 ml saline. There abdominal wall was closed in layers. Water was a continuous registration of arterial was allowed on the first day following opblood pressure, ECG, pulse rate, and body eration and a standard kennel diet after 24 temperature. hr. In cases of dehydration and frequent Under sterile conditions the duodenum vomiting, fluid was given either orally using and attached pancreas were exposed through a stomach tube (milk) or intravenously (glua midline upper abdominal incision. Bile was cose 5%). collected aseptically by puncturing the gallBefore and immediately after the surgical bladder and stored at 4°C. The minor pan- procedure blood was drawn for gas analysis creatic duct beyond the duodenum was dou- and determination of blood sugar, serum amble ligated and cut. Through a 3-4-cm-long ylase, albumin, calcium, leucocyte count, antimesenteric duodenotomy the major pan- hemoglobin, hematocrit, transaminases creatic duct was then cannulated using a (GOT, GPT), lactate dehydrogenase (LDH), polyethylene cannula Longdwel 20 gauge. urea nitrogen, and creatinine. In the survivUnder a constant pressure of 40 cm HZ0 a ing dogs repeated biochemical estimations mixture of 1 ml bile/kg body wt, containing were made several times during the first week 2000 BAEE units of trypsin (T 8253 Sigma and weekly thereafter for 4 weeks. The enChemical Co.) in each mililiter of bile (the docrine function of the pancreas was also mixture of bile and trypsin had been incu- assessedat intervals in the survivors by meabated at 37°C for 16 hr), was infused in the suring the intravenous glucose tolerance tests pancreatic duct. The mean infusion time in and the circulating plasma insulin (radio imthis study was 27 min. After the intraductal munoassay) after the glucose overload. infusion, the cannula was withdrawn and the At the time of death an autopsy was perduodenotomy was temporally closed. The formed on all the dogs. The pancreas and
166
JOURNAL
OF SURGICAL
RESEARCH:
duodenum were dissected for histological examination. Special attentions were made at the sites of macroscopic changes, however, microscopic examinations were also performed in the right as well as in the left lobe, the papilla, and in the central part of the pancreas. RESULTS
Almost immediately after the induction of acute pancreatitis by infusion of the incubated bile-trypsin mixture into the pancreatic duct, the surface of the pancreas showed edema with brownish pigmentation. The color of the pancreatic tissue usually became darker than the bile administered to induce the pancreatitis. After a period of 30 min or so, the pancreas could be seento weep a bloody fluid from its surface into the abdominal cavity. Areas of ecchymotic, petechial, and frank hemorrhage could also be observed in the gland and peripancreatic tissue after 1 hr. There were also exudations present along the blood vesselsin omentum and mesenterium of the duodenum. Before the laparotomy incision could be closed, the left lobe of the pancreas was usually very dark and identification of lobuli was sometimes difficult. Table 1 shows the fatality rate of our model of acute pancreatitis as described under Materials and Methods and the effect of intraductal injection of saline and of Ethibloc in these experimental animals. In the first 10
TABLE
1
THE INFLUENCE OF PANCREATIC DUCT OCCLUSION WITH ETHIBL~C AND THE EFFECT OF INTRADU~TAL ADMINISTRATION OF SALINE ON THE MORTALITY RATE OF EXPERIMENTAL ACUTE PANCREATITIS IN Does Group I (Ethibloc) II (Saline) III (No treatment) Total
Survivals 4 0 1 5
Nonsurvivals 1 5 4 10
Total 5 5 5 15
VOL. 34, NO. 2, FEBRUARY TABLE
1983 2
BKXHEMICAL ANALYSIS OF BLEND COLLECI%D 8 HR BEFORE DEATH OF THE ANIMAL FROM THE ETHIBLOC-TREATED GROUP
Hemoglobin (mmol/l) Hematocrit (l/l) Glucose (mmol/l) Urea-N (mmol/l) Creatinine (pmol/l) Calcium (mmol/l) Albumin (g/l) Amylase (u/l) GOT (u/l) GPT (u/l) LDH (u/l)
Obtained finding
Normal value
13.9 0.68 0.30 15.0 668.0 2.46 23.0 25,900 19 41 264
8-11 0.35-0.50 4.6-6.3 <6.6 t68 2.30-2.80 25-33 13000 <40 160 <180
days after operation, all dogs, if they survived, were very ill with weakness,occasional vomiting, and anorexia. All animals in Group II (saline) and 4 of the 5 animals in Group III (control group) expired within 8 days postoperatively. No obvious difference was observed in the survival time between Groups II and III, however, one animal from Group III survived. After the complete recovery and intravenous feeding in the first week after operation, the dog was sacrificed on the tenth day for necropsy. In contrast to Groups II and III, 4 of the 5 animals of Group I with experimentally induced acute pancreatitis treated by pancreatic duct occlusion survived. In these animals long-term follow up biochemical studies including intravenous glucose tolerance test and plasma insulin measurements had been performed at intervals. These dogs were killed at 42, 7 1, 162, and 2 19 days after the operation for histopathological examination. One of the Ethibloc-treated animals expired 28 hr after the operation. From the retrospective results of biochemical investigation of blood collected 20 hr after the operation as demonstrated in Table 2 the severe hypoglycemia was probably responsible for the acute coma and resulted in the death of this animal.
167
ESTOURGIE ET AL.: DUCT OCCLUSION O-amylase “/I x 1000
30 n-
t
t
1
2
3
4
3 hours after operation before operation
5
6
7
8
9 10 11 12 days after operation
FIG. 1. The course of serum amylase activity in the surviving dogs. These animals were treated with Ethibloc intraductally 2 hr after the induction of acute hemorrhagic pancreatitis.
Biochemical Investigation For the biochemical examinations of the three groups of animals, the blood was collected in all animals before and 1 hr after the end of the operation (i.e., 3 hr after the induction of pancreatitis). Although some of the parameters showed statistically significant changes (Student’s t test) after the induction of acute pancreatitis, no obvious differences were seenbetween the three groups. Plasma calcium, serum albumin, and leucocyte counts were decreased after the induction of pancreatitis in all animals, while the serum amylase activity and GOT were higher. Further investigation on the differencesof the biochemical parameters between the three groups was not possible because 36 hr after the operation there were only 2 dogs from Groups II and III alive. Figure 1 illustrates the course of serum amylase activity in the surviving dogs. The peak values were found 2 days after the operation. After 7
days, the serum amylase returned to approaching normal range. To investigate the endocrine function of the pancreas in the surviving animals, intravenous glucose tolerance testsand plasma insulin responseafter the intravenous glucose administration (0.5 mg/kg body wt injected for 2 min) were performed at intervals ranging from 1 to 31 weeks postoperatively. The results (data not shown) were found to be within normal limits except in one dog. In this animal, the peak value of insulin was found to be significantly lower; however, the blood sugar was normal and there was no evidence of glucosuria.
Histopathological Findings At autopsy of the 7 animals expired within 36 hr after the operation (1 from Group I, 3 from Group II, and 3 from Group III), the left lobe as well as the right lobe of the pancreas were always involved by the disease.
168
JOURNAL OF SURGICAL RESEARCH: VOL. 34, NO. 2, FEBRUARY 1983
However, the hemorrhagic necrosis of the right lobe was only moderate or absent in some dogs. The abdominal cavities were uniformly filled with quantities of bloody ascitic fluid in all animals. The pancreas was dark green to black and there was peripancreatic fat necrosis. The gross picture was that of the fulminating acute hemorrhagic pancreatitis. Microscopically there were interstitial and acinar hemorrhage, edema, necrosis of exocrine and endocrine cells, fat necrosis, and there was only slight leucocytic infiltration in the interstitial space. Although there were congestions in some sections, the blood vesselswere usually normal. Except in the ethiblot-treated animals, the main pancreatic ducts and the papilla were also not affected. In these animals, no differences in histopathological findings could be found between the saline-treated and untreated groups. In the Ethibloc-treated animals the ducts were filled with an eosinophilic mass (Ethiobloc). Disruption of the epithelial lining cells was evident. There were also periductal leucocytic infiltrations. These findings were also found in healthy dogs treated with duct occlusion with Ethibloc. The autopsy of the animals that expired between 36 hr and 8 days after operation (2 from Group II and 1 from Group III) showed evidences of peripancreatic fat necrosis and inflammation. However, interstitial fibrosis, replacement of pancreatic gland by fibrous tissue and extensive adhesions had also been observed. In cases of surviving dogs (4 from Group I and 1 from Group III) sacrificed at intervals ranging from 10 days to 32 weeks after operation, a pancreatic atrophy of the left lobe was found. Fat necrosis was not seen macroscopically. The pancreas had been replaced by fibrous tissue. There were striking adhesions around the pancreas and in the peripancreatic tissue. However, surviving segmentsof normal pancreas (exocrine and endocrine cells) were usually present in the right lobe and in the central part of the “residual pancreatic tissue.” In all animals of Ethibloc-treated group, no occlusive material in pancreatic ducts was observed.
DISCUSSION
In studying the effect of therapy in experimental pancreatitis, it is necessaryto utilize a suitable model, which would develop pancreatitis uniformly with the same prognosis. In our present study, we have used the model of acute, necrotizing pancreatitis as described by Musa et al. [ 161. However, to avoid the rupturing of the duct system in our experiments, we infused the bile-trypsine mixture under a constant pressure of 40 cm HZ0 instead of injecting by hand for 15 min. The mixture had been incubated for 16 hr at 37°C as described by Elliott et al. [ 111. Although all animals in the studies of Musa and Elliott were dead after the induction of pancreatitis, one dog from our control group (Group III, no treatment) given the same amount of trypsinized bile, survived. This finding confirms that the method of administration of bile-trypsin mixture plays also an important role in determining the absolute lethality in an acute pancreatitis model [ 11, 161. In a retrospective study [ 181 in patients with acute pancreatitis, it has been shown that a high mortality rate was associated with biochemical parameters on admission including age, blood sugar, WBC, LDH, and SGOT. Falling of hematocrit, arterial PO? and serum calcium levels, an anion base deficit, increased BUN, and fluid sequestration in the first 48 hr after admission worsened the prognosis. There is no medical therapy that has been proved by control studies to alter favorably the mortality rate of acute necrotizing pancreatitis. Surgery has a legitimate role in the management of acute pancreatitis, especially for the debridement of necrotic tissue or for drainage of pancreatic exsudate or abscesses and pseudocysts. Drainage of the peritoneal exsudate to remove the vasoactive toxins which are responsible for the cardiovascular collapse resulted in clinical improvement in experimental animals [20, 211 as well as in patient studies [ 19, 231. A total pancreatectomy in early phase has also been suggested to afford protection (9, 141.However, radical
ESTOURGIE ET AL.: DUCT OCCLUSION
dissection of pancreas and surrounding tissues has a high mortality rate. Eggink [lo] reported a series of 17 patients with a mortality rate of 7 1% after segmental left resection. Necrosis of the pancreas parenchyma is the predominant finding in acute necrotizing pancreatitis; however, a total necrosis is only found in 10% of the cases.In most patients, when the surface of the whole pancreas is necrotic, there are still segmentsof vital tissue present in the gland, especially around the main pancreatic duct [3]. Since the medical therapies known to be effective are based upon the rationale of slowing pancreatic secretion (putting the pancreas at rest) and a continued production of significant pancreatic enzymes is frequently found in fulminant necrotizing pancreatitis, a simple measure which permits the “burning out” of residual pancreatic tissue might therefore be effective to alter the natural history of this diseasewith poor prognosis. Total duct obstruction with resultant exocrine atrophy offers a theoretically appealing alternative to most of the direct pancreatic procedures to stop pancreatic enzyme production. It is well known that total duct obstruction leads to exocrine atrophy without pancreatitis and perhaps with preservation of endocrine elements. The advent of the acrylate glue and of Ethiobloc (alcoholic prolamin solution) has made possible a new technique of total pancreatic duct obstruction. The use of these substancesto obstruct the pancreatic duct in healthy experimental animals has been reported [2, 7, 15, 171. In these studies, a noninflammatory atrophy has been produced with preservation of the endocrine gland. Encouraging results and evidences of clinical improvement have also been obtained in patients with chronic pancreatitis [ 13, 15, 221. In our present study, we demonstrate that injection of Ethibloc in the pancreatic duct for a total occlusion, 2 hr after the induction of acute pancreatitis in dogs, resulted in an improvement of the survival rate. One of the 5 animals, however, expired 28 hr after the operation. The prox-
169
imate cause of death in this animal was probably not directly related to the toxic cardiovascular collapse as found in the expired control dogs. The observation of the changesof pancreas and surrounding tissue after induction of the pancreatitis and before the laparotomy incisions were closed, suggeststhat the severe inflammatory process was uniformly found in the experimental animals in the acute phase. Despite the fact that the gross picture at necropsy of expired animals was that of acute fulminating pancreatitis, segments of normal pancreatic tissue especially in the right lobe has been found in surviving dogs. As has been found in normal dogs after duct occlusion with Ethibloc, the pancreas of these animals was replaced largely by fibrous tissue. However, the presence of adhesions around the pancreas and in the peripancreatic tissue was very striking compared to those found in normal dogs treated with Ethibloc. The finding that the right lobe of the pancreas in all dogs had little pathologic alterations is due to the fact of the special anatomy of the pancreas in dog and the method of inducing pancreatitis in our experiments. The pancreas in the dog is boomerang shaped. The cannula for infusion of the bile-trysin mixture and for injection of saline or Ethibloc lies always in the left lobe; the infused fluid came therefore only retrograde or through narrow anastomotic ducts into the right lobe. These findings are supported by the results of experiments in healthy animals treated with Ethibloc (to be reported in a separate paper). Although further studies and many details in the mechanism remain to be established, the results of our present study suggest that pancreatic duct occlusion causing exocrine secretion to stop may have beneficial effects in the treatment of acute fulminant pancreatitis in the acute phase and may improve the survival rate. ACKNOWLEDGMENT The authors thank Dr. J. C. M. Hafkenscheid for performing the biochemical analysis.
170
JOURNAL OF SURGICAL RESEARCH: VOL. 34, NO. 2, FEBRUARY 1983
REFERENCES 1. Anderson, M. C., Schoenfeld, F. B., Iams, W. B., and Suwa, M. Circulatory changes in acute pancreatitis. Surg. Clin. North Amer. 47: 121, 1961. 2. Baumgartner, D., Hanson, S. L., and Sutherland, D. E. R. Endocrine function and histology of pancreatic segments after duct occlusion with ethibloc or neoprene: A comparative study. Eur. Surg. Res. 13: 33, 1981. 3. Becker, V. Akute Pankreatitis, Morfologie, Pathogenese, Prognose. Chirurg 51: 351, 1980. 4. Chiari, H. Uber Selbstverdauung des menschlichen Pankreas. Z. Heilk. 17: 69, 1896. 5. Creutzfeldt, W., and Schmidt, H. Aetiology and pathogenesis of pancreatitis (current concepts). Stand. J. Gastroenterol. Suppl. 6: 47, 1910. 6. Donaldson, L. A., Williams, R. W., and Schenk, W. G. Experimental pancreatitis: Effect of plasma and dextran on pancreatic blood flow. Surgery 84: 313, 1978. I. Dubemard, J. M., Traeger, J., Neyra, P., Touraine, J-L., Devonec, M., Blanc-Brunat, N., and Ruitton, A. Une nouvelle m&hode de pr6paration due greffon pancr6atique en vue de la transplantation. Chirurgie 104: 242, 1978. 8. Dtlrr, G. H. K. Acute pancreatitis. In H. T. Howat and H. Sarles (Ed%), The Exocrine Pancreas. London: Saunders, Chap. 15. 1979. 9. Edelmann, G., and Boutelier, Ph. Le traitement des pancreatites aigti6s necrotisantes par l’ablation chirurgicale precose des portions nbcros&. Chirurgie loo: 155, 1974. 10. Eggink, W. F. De operatieve behandling van pankreatitis. Thesis. Rotterdam 1981. 11. Elliott, D. W., Williams, R. D., and Zollinger, R. M. Alterations in the pancreatic resistanceto bile in the pathogenesis of acute pancreatitis. Ann. Surg. 146: 669, 1957. 12. Ellison, E. C., Pappas, T. N., Johnson, J. A., Fabri,
P. J., and Carey, L. C. Demonstration and characterization of the hemoconcentrating effect of as-
citic fluid that accumulates during hemorrhagic pancreatitis. J. Surg. Res. 30: 24 1, 1981. 13. Gebhardt, Ch., and Gall, F. B. PartiBle Duodenopankreatektomie mit intraoperativer Pankreasschwanzverijdung bei chronicher Pankreatitis. Langenbecks Arch. Chir. 353: 57, 1980. 14. Henry, L. G., and Condon, R. E. Ablative surgery for necrotizing pancreatitis. Am. J. Surg. 131: 125, 1916.
15. Little, J. M., Lauer, C., and Hogg, J. Pancreatic duct obstruction with an acrylate glue. A new method for producing pancreatic exocrine atrophy. Surgery 81: 243, 1977.
16. Musa, B. E., Nelson, A. W., Gilette, E. L., Ferguson, H. L., and Lumb, W. V. A model to study acute pancreatitis in the dog. J. Surg. Res. 21: 5 1, 1976. 17. Papachristou, D. N., D’Agostino, H., and Fortner, J. G. Ligation of the pancreatic duct in pancreatectomy. &it. J. Surg. 67: 260, 1980. 18. Ranson, J. H. C., Rilkind, K. M., and Turner, J. W. Prognostic signs and nonoperative peritoneal lavage in acute pancreatitis. Surg. Gynecol. Obstet. 143: 209, 1976.
19. Ranson, J. H. C. Acute pancreatitis. Curr. Probl. Surg. 16: 11, 1979. 20. Rodgers, R. E., and Carey, L. C. Peritoneal lavage in experimental pancreatitis in dogs. Amer. J. Surg. 111: 192, 1966. 21. Rosato, E. F., Mullis, W. F., and Rosato, F. E. Peri-
toneal lavage therapy in hemorrhagic pancreatitis. Surgery 74: 106, 1913. 22. R&ch, W., Phillip, J., and Gebhardt, Ch. Endoscopic duct obstruction in chronic pancreatitis. Endoscopy 1: 43, 1979. 23. Stone, H. H., and Fabian, T. C. Peritoneal dialysis in the treatment of acute alcoholic pancreatitis. Surg. Gynecol. Obstet. 150: 818, 1980. 24. Takada, Y., Appert, H. E., and Howard, J. M. Vascular permeability induced by pancreatic exsudate formed during acute pancreatitis in dogs, Surg. Gynecol. Obstet. 143: 779, 1976.