JOURNAL OF PATHOLOGY, VOL.
167: 421424 (1992)
PATHOGENESIS OF PANCREATIC PERILOBULAR NECROSIS IN PATIENTS WITH LIVER DISEASE YURO SHIBAYAMA, KAZUAKI HASHIMOTO AND KATSUJI NAKATA
Department of Pathology, Osaka Medical College, Takatsuki, Osaka, Japan Received 18 February 1992 Accepted 7 April 1992
SUMMARY We often see perilobular necrosis of the pancreas in patients with liver disease at autopsy. This study was undertaken to determine the frequency and the mechanism of development of pancreatic perilobular necrosis in patients with liver disease. Pancreatic perilobular necrosis was seen in 21 per cent of 261 autopsied patients: in 41 per cent of 73 autopsied patients with liver disease and in 13 per cent of 188 autopsied patients without liver disease. Moreover, splanchnic congestion was present in 90 per cent of 30 pancreatic perilobular necrosis patients with liver disease. These data indicate that patients with liver disease develop perilobular necrosis of the pancreas more often than patients without liver disease, and that the high frequency may be a sequela of splanchnic congestion; that is, congestion of the pancreas and endotoxaemia due to congestion of the gut. KEY
worm-Pancreatic perilobular necrosis, liver disease, splanchnic congestion, congestion of the pancreas, endotoxaemia
INTRODUCTION An anatomical classification of acute pancreatitis, which can be correlated with various clinical aetiologies, has been proposed: periductal, perilobular, and panlobular necrosis.' Perilobular necrosis, which is defined as necrosis and inflammation confined to the periphery of lobules, is commonly associated with hypotension, septic or cardiogenic shock, or cardiac surgery.' We have often seen perilobular necrosis of the pancreas at autopsy, but it is usually mild. Moreover, we have the impression that patients with liver disease are more likely to have perilobular necrosis of the pancreas than patients without liver disease. However, the frequency and the mechanism of the development of pancreatic perilobular necrosis in patients with liver disease have not been established. This autopsy study focuses on these points. Addressee for correspondence: Dr Y. Shibayama, Department of Pathology, Osaka Medical College, Daigaku-Cho, Takatsuki City, Osaka, Japan.
0022-341 7/92/080421-04 $07.00 0 1992 by John Wiley & Sons, Ltd
MATERIALS AND METHODS We reviewed 261 consecutive autopsies of adults ( > 16 years of age) performed at the Osaka Medical
College Hospital between June 1987 and March 1991. Cases of pancreatic perilobular necrosis were selected on the basis of a confirmed microscopic diagnosis. The presence of splanchnic congestion was determined by dilatation of the splanchnic vascular bed, congestion of the gut, and congestive splenomegaly . RESULTS The ages of the 261 autopsied patients ranged from 17 to 90 years (mean age 61 years). There were 147 men and 1 14 women. The diseases related to pancreatic perilobular necrosis and their incidence are shown in Table I. There were 55 patients with perilobular necrosis of the pancreas, and no periductal necrosis or panlobular necrosis. There was some overlapping of patients with pancreatic perilobular necrosis in
422
Y. SHIBAYAMA ETAL.
Table I-Diseases related to perilobular necrosis of the pancreas
Autopsied patients Liver cirrhosis with hepatocellular carcinoma without hepatocellular carcinoma Liver fibrosis with hepatocellular carcinoma without hepatocellular carcinoma Hepatic infarction Fulminant hepatitis Gallstones Serious bacterial infection Sepsis DIC Carcinoma of the pancreas Malignant tumour excluding carcinoma of the pancreas and hepatocellular carcinoma Cardiovascular disorders
Table I: four patients with liver fibrosis without hepatocellular carcinoma had malignant tumour excluding cancer of the pancreas and one patient with liver fibrosis had DIC; five patients with malignant tumour excluding cancer of the pancreas and hepatocellular carcinoma had DIC; one patient with hepatic infarction had cancer of the pancreas and serious bacterial infection; one patient with DIC had hepatic infarction. The ages of patients with pancreatic perilobular necrosis ranged from 19 to 83 years (mean age 61 years). There were 34 men and 21 women. In all those with pancreatic perilobular necrosis, small necrotic areas of the periphery of the lobules and fat tissue around the pancreas were observed microscopically (Fig. 1A). There were fibrin thrombi in capillaries and venules and an acute inflammatory cell reaction at the boundary between areas of necrotic and viable pancreas (Fig. 1B). Thirty ( 5 5 per cent) of 55 patients with pancreatic perilobular necrosis had liver disease. Pancreatic perilobular necrosis was seen in 30 (41 per cent) of 73 autopsied patients with liver disease and 25 (13 per cent) of 188 autopsied patients without liver disease. In patients with DIC, pancreatic perilobular necrosis was seen relatively often, but it was not found in patients with sepsis without DIC. Pancreatic perilobular necrosis was frequently seen
No. of patients
No. of patients with perilobular necrosis
261
55 (21 Yo)
21 30
8 (38%) 10 (33%)
2 10 6 4 12 12 10 9 111
1 (50?'0) 5 (50%) 3 (5ooLo) 3 (75%) 0 (0%) 4 (33%) 0 (0%) 7 (44%) 3 (33%) 19 (17%)
31
5 (16%)
16
in patients with serious bacterial infection, e.g., suppurative peritonitis and liver abscess, without DIC. It was found more often in patients with cancer of the pancreas than in patients with malignant tumour excluding cancer of the pancreas and hepatocellular carcinoma. Perilobular necrosis of the pancreas did not develop in patients with gallstones. The pathological and clinical correlations in 55 patients with perilobular necrosis of the pancreas are shown in Table 11. Splanchnic congestion and jaundice were frequent, while alcohol abuse was unusual. Tumour metastasis to the liver was observed in 9 of 19 pancreatic perilobular necrosis patients who died of malignant tumour, and 8 of them had splanchnic congestion. Jaundice was relatively common even in patients without liver disease.
DISCUSSION In this study we found that 13 per cent of autopsied patients without liver disease and 41 per cent of those with liver disease had perilobular necrosis of the pancreas, and that splanchnic congestion was present in 90 per cent of those with liver disease and
423
PANCREATIC PERILOBULAR NECROSIS IN LIVER DISEASE
Fig. I-Perilobular necrosis of the pancreas in a patient with liver cirrhosis. (A) A small necrotic area of the periphery of the lobule and fat tissue around the pancreas is seen. (B) High poweer view of the necrotic area shown in A. An acute inflammatory cell response is seen at the boundary between areas of necrotic and viable pancreas
Table 11-Pathological and clinical correlations in 55 patients with perilobular necrosis of the pancreas
Patients with perilobular necrosis Liver cirrhosis Liver fibrosis Hepatic infarction Fulminant hepatitis Serious bacterial infection DIC Carcinoma of the pancreas Malignant tumour excluding carcinoma of the pancreas and hepatocellular carcinoma Cardiovascular disorders
No. of patients
Splanchnic congestion
Jaundice
55 18 6 3 3 4 7 3 19
38 (69%) 17 (94%) 6 (lOOY0) 2 (67%) 2 (67%) 2 (50%) 5(71%) 3 (100%) 8 (42%)
29 (53%) 15 (83%) 3 (5O0/o) 2 (67%) 3 (1 00%) 2 (50%) 3 (43%) 2 (67%) 7 (37%)
5
3 (60%)
2 (40%)
Severe metastasis to the liver 10 (18%) -
~
1(33%) 9 (47%)
Alcohol abuse 3 (So/) l(6Yo) 2 (33%) 0 (OYO) 0 (0%) 0 (0%) 0 (0%) 0 (0%) 0 (0%)
0 (OYO)
424
Y. SHIBAYAMA E T A L .
pancreatic perilobular necrosis. These data suggest or suppurative peritonitis due to perforation of that splanchnic congestion, especially congestion of the gut. the pancreas and the gut, plays an important role in It has been reported that perilobular necrosis of the development of pancreatic perilobular necrosis the pancreas is not associated with either of the in patients with liver disease. We know that acute two common causes of pancreatitis: gallstones and pancreatitis can be induced by congestion of the alcohol abuse.' The results of the present study pancreas and endotoxaemia in rats, and that the support this finding. incidence and the degree of acute pancreatitis are In summary, this study shows that perilobular high when they co-exist.* It has been reported that necrosis of the pancreas frequently develops in congestion of the gut may lead to endotoxaemia in patients with liver disease, and that the development animals through diffusion of endotoxin from the may be associated with sequelae of splanchnic injured In patients with liver disease, such as congestion, namely congestion of the pancreas and liver cirrhosis, liver fibrosis, hepatic infarction, or endotoxaemia due to congestion of the gut. fulminant hepatitis, endotoxaemia is often a result of splanchnic congestion due to portal hypertension and of depressed reticuloendothelial f ~ n c t i o n . ~ - ' ~ REFERENCES These disorders may play a role in the development of perilobular necrosis in many patients with liver 1. Foulis AK. Histologic evidenceofinitiating factors in acute pancreatitis inman. JC/inPathol1980;33: 1125-1131. disease. 2. Shibayama Y. Pancreatic venous stasis and endotoxaemia as aetioJaundice was observed in many patients with logic factors in acute haemorrhagic pancreatitis. J Parhol 1987; 1 5 2 177-1 82. pancreatic perilobular necrosis in this study. It 3. Ravin HA, Rowley D, Jenkins C, Fine J, On the absorption of would be expected in patients with liver disease. bacterial endotoxin from the gastrointestinal tract of the normal and However, it was also seen in many patients with shocked animal. J Exp Med 1960; 112: 783-792. pancreatic perilobular necrosis and bacterial in- 4. Cans H, Matsumoto K. The escape of endotoxin from the intestine. Surg Gynecol Obstet 1974; 1 3 9 395-402. fection or cardiovascular disease, perhaps because 5. Nolan JP, HareDK, McDevitt JJ,AliMV. Invifrostudiesofintestinal of intrahepatic cholestasis due to end~toxaemia,'~ endotoxin absorption. Gasfroenterology 1977; 7 2 434-439. 6. Olofsson P, Nylander G, Olsson P. Endotoxin-transport routes and since endotoxaemia is known to be a factor leading kinetics in intestinal ischemia. Acra Chir Scand 1985; 151; 635-639. to jaundice in bacterial i n f e ~ t i o n . ' ~The , ' ~ exact 7. Wilkinson SP, Arroyo V, Gazzard BG, Moodie H, Williams R. Relation of renal impairment and haemorrhagic diathesis to endomechanism for the development of jaundice in toxaemia in fulminant hepatic failure. Lancet 1974; 1: 521-524. patients with heart disease is not certain,l6.l7 but 8. Tarao K, So K, Moroi T, ef a!. Detection of endotoxin in plasma congestive cardiac failure may lead to splanchnic and ascitic fluid of patients with cirrhosis: its clinical significance. Gasrroenfero1og.v1977; 73: 539-542. congestion or a decrease in the splanchnic blood 9. Magliulo E, Cruclani M, Dietz A, e t a / . Endotoxaemia in acute viral flow and subsequently induce endotoxaemia. 1981; 28: 299-303. hepatitis. Hepafo~a~trornferolog.~ Perilobular necrosis of the pancreas was observed 10. Gaeta GB, Perna P, Adinolfi LE, Utili R, Ruggiero G. Endotoxemia 104 patients with chronic liver disease: prevalence and in a series of in 44 per cent of autopsied patients with DIC in this significance. Digestion 1982; 23: 239-244. study. In such cases, fibrin thrombi were found in I I . Bigatello LM, Broitman SA, Fattori L, el a/. Endotoxemia, encephalopathy, and mortality in cirrhotic patients. Am J Gastroenferol1987; capillaries and venules at the boundary between 11-15, areas of necrotic and viable pancreas. This finding 12. 82: Lumsden AB, Henderson JM, Kutner MH. Endotoxin levels suggests that DIC itself can be a cause of pancreatic measured by a chromogenic assay in portal, hepatic and peripheral venous blood in patients with cirrhosis. Hepatology 1988; 8: 232-236. perilobular necrosis. However, in this study 13. Utili R, Abernathy CO, Zimmerrnan HJ. Cholestaticeffects of Eschersplanchnic congestion was seen in 71 per cent of irhia coliendotoxin on the isolated perfused rat liver. Gastroenterology DIC patients with perilobular necrosis, indicating 1976; 7 0 248-253. Zimmerman HJ, Fang M, Utili R, Seeff LB, Hoofnagle J. Jaundice that splanchnic congestion may be related to the 14. due to bacterial infection. Gastroenterology 1979; 77: 362-374. development of perilobular necrosis of the pancreas 15. Letlowitch JH. Bile ductularcholestasis: an ominous histopathologic sign related to sepsis and 'cholangitis lenta'. Hum Pathol 1982; 13: even in patients with DIC. 19-24, Perilobular necrosis of the pancreas was observed 16. Sherlock S . The liver in heart failure. Relation of anatomical, in 33 per cent of autopsied patients with severe functional, and circulatory changes. Br Hearr J 1951; 13: 273-293. bacterial infection. In these cases, endotoxaemia 17. Dunn GD, Hayes P, Breen KJ, Schenker S. The liver in congestive heart failure: a review. Am J MedSci 1973; 265 174-189. is a possible complication," because severe bac- 18. Olofsson P, Nylander G, Olsson P. Endotoxins: routes of transport in terial infection can originate from liver abscess experimental peritonitis. Am J Surg 1986; 151: 443-446.
167: 421424 (1992)
PATHOGENESIS OF PANCREATIC PERILOBULAR NECROSIS IN PATIENTS WITH LIVER DISEASE YURO SHIBAYAMA, KAZUAKI HASHIMOTO AND KATSUJI NAKATA
Department of Pathology, Osaka Medical College, Takatsuki, Osaka, Japan Received 18 February 1992 Accepted 7 April 1992
SUMMARY We often see perilobular necrosis of the pancreas in patients with liver disease at autopsy. This study was undertaken to determine the frequency and the mechanism of development of pancreatic perilobular necrosis in patients with liver disease. Pancreatic perilobular necrosis was seen in 21 per cent of 261 autopsied patients: in 41 per cent of 73 autopsied patients with liver disease and in 13 per cent of 188 autopsied patients without liver disease. Moreover, splanchnic congestion was present in 90 per cent of 30 pancreatic perilobular necrosis patients with liver disease. These data indicate that patients with liver disease develop perilobular necrosis of the pancreas more often than patients without liver disease, and that the high frequency may be a sequela of splanchnic congestion; that is, congestion of the pancreas and endotoxaemia due to congestion of the gut. KEY
worm-Pancreatic perilobular necrosis, liver disease, splanchnic congestion, congestion of the pancreas, endotoxaemia
INTRODUCTION An anatomical classification of acute pancreatitis, which can be correlated with various clinical aetiologies, has been proposed: periductal, perilobular, and panlobular necrosis.' Perilobular necrosis, which is defined as necrosis and inflammation confined to the periphery of lobules, is commonly associated with hypotension, septic or cardiogenic shock, or cardiac surgery.' We have often seen perilobular necrosis of the pancreas at autopsy, but it is usually mild. Moreover, we have the impression that patients with liver disease are more likely to have perilobular necrosis of the pancreas than patients without liver disease. However, the frequency and the mechanism of the development of pancreatic perilobular necrosis in patients with liver disease have not been established. This autopsy study focuses on these points. Addressee for correspondence: Dr Y. Shibayama, Department of Pathology, Osaka Medical College, Daigaku-Cho, Takatsuki City, Osaka, Japan.
0022-341 7/92/080421-04 $07.00 0 1992 by John Wiley & Sons, Ltd
MATERIALS AND METHODS We reviewed 261 consecutive autopsies of adults ( > 16 years of age) performed at the Osaka Medical
College Hospital between June 1987 and March 1991. Cases of pancreatic perilobular necrosis were selected on the basis of a confirmed microscopic diagnosis. The presence of splanchnic congestion was determined by dilatation of the splanchnic vascular bed, congestion of the gut, and congestive splenomegaly . RESULTS The ages of the 261 autopsied patients ranged from 17 to 90 years (mean age 61 years). There were 147 men and 1 14 women. The diseases related to pancreatic perilobular necrosis and their incidence are shown in Table I. There were 55 patients with perilobular necrosis of the pancreas, and no periductal necrosis or panlobular necrosis. There was some overlapping of patients with pancreatic perilobular necrosis in
422
Y. SHIBAYAMA ETAL.
Table I-Diseases related to perilobular necrosis of the pancreas
Autopsied patients Liver cirrhosis with hepatocellular carcinoma without hepatocellular carcinoma Liver fibrosis with hepatocellular carcinoma without hepatocellular carcinoma Hepatic infarction Fulminant hepatitis Gallstones Serious bacterial infection Sepsis DIC Carcinoma of the pancreas Malignant tumour excluding carcinoma of the pancreas and hepatocellular carcinoma Cardiovascular disorders
Table I: four patients with liver fibrosis without hepatocellular carcinoma had malignant tumour excluding cancer of the pancreas and one patient with liver fibrosis had DIC; five patients with malignant tumour excluding cancer of the pancreas and hepatocellular carcinoma had DIC; one patient with hepatic infarction had cancer of the pancreas and serious bacterial infection; one patient with DIC had hepatic infarction. The ages of patients with pancreatic perilobular necrosis ranged from 19 to 83 years (mean age 61 years). There were 34 men and 21 women. In all those with pancreatic perilobular necrosis, small necrotic areas of the periphery of the lobules and fat tissue around the pancreas were observed microscopically (Fig. 1A). There were fibrin thrombi in capillaries and venules and an acute inflammatory cell reaction at the boundary between areas of necrotic and viable pancreas (Fig. 1B). Thirty ( 5 5 per cent) of 55 patients with pancreatic perilobular necrosis had liver disease. Pancreatic perilobular necrosis was seen in 30 (41 per cent) of 73 autopsied patients with liver disease and 25 (13 per cent) of 188 autopsied patients without liver disease. In patients with DIC, pancreatic perilobular necrosis was seen relatively often, but it was not found in patients with sepsis without DIC. Pancreatic perilobular necrosis was frequently seen
No. of patients
No. of patients with perilobular necrosis
261
55 (21 Yo)
21 30
8 (38%) 10 (33%)
2 10 6 4 12 12 10 9 111
1 (50?'0) 5 (50%) 3 (5ooLo) 3 (75%) 0 (0%) 4 (33%) 0 (0%) 7 (44%) 3 (33%) 19 (17%)
31
5 (16%)
16
in patients with serious bacterial infection, e.g., suppurative peritonitis and liver abscess, without DIC. It was found more often in patients with cancer of the pancreas than in patients with malignant tumour excluding cancer of the pancreas and hepatocellular carcinoma. Perilobular necrosis of the pancreas did not develop in patients with gallstones. The pathological and clinical correlations in 55 patients with perilobular necrosis of the pancreas are shown in Table 11. Splanchnic congestion and jaundice were frequent, while alcohol abuse was unusual. Tumour metastasis to the liver was observed in 9 of 19 pancreatic perilobular necrosis patients who died of malignant tumour, and 8 of them had splanchnic congestion. Jaundice was relatively common even in patients without liver disease.
DISCUSSION In this study we found that 13 per cent of autopsied patients without liver disease and 41 per cent of those with liver disease had perilobular necrosis of the pancreas, and that splanchnic congestion was present in 90 per cent of those with liver disease and
423
PANCREATIC PERILOBULAR NECROSIS IN LIVER DISEASE
Fig. I-Perilobular necrosis of the pancreas in a patient with liver cirrhosis. (A) A small necrotic area of the periphery of the lobule and fat tissue around the pancreas is seen. (B) High poweer view of the necrotic area shown in A. An acute inflammatory cell response is seen at the boundary between areas of necrotic and viable pancreas
Table 11-Pathological and clinical correlations in 55 patients with perilobular necrosis of the pancreas
Patients with perilobular necrosis Liver cirrhosis Liver fibrosis Hepatic infarction Fulminant hepatitis Serious bacterial infection DIC Carcinoma of the pancreas Malignant tumour excluding carcinoma of the pancreas and hepatocellular carcinoma Cardiovascular disorders
No. of patients
Splanchnic congestion
Jaundice
55 18 6 3 3 4 7 3 19
38 (69%) 17 (94%) 6 (lOOY0) 2 (67%) 2 (67%) 2 (50%) 5(71%) 3 (100%) 8 (42%)
29 (53%) 15 (83%) 3 (5O0/o) 2 (67%) 3 (1 00%) 2 (50%) 3 (43%) 2 (67%) 7 (37%)
5
3 (60%)
2 (40%)
Severe metastasis to the liver 10 (18%) -
~
1(33%) 9 (47%)
Alcohol abuse 3 (So/) l(6Yo) 2 (33%) 0 (OYO) 0 (0%) 0 (0%) 0 (0%) 0 (0%) 0 (0%)
0 (OYO)
424
Y. SHIBAYAMA E T A L .
pancreatic perilobular necrosis. These data suggest or suppurative peritonitis due to perforation of that splanchnic congestion, especially congestion of the gut. the pancreas and the gut, plays an important role in It has been reported that perilobular necrosis of the development of pancreatic perilobular necrosis the pancreas is not associated with either of the in patients with liver disease. We know that acute two common causes of pancreatitis: gallstones and pancreatitis can be induced by congestion of the alcohol abuse.' The results of the present study pancreas and endotoxaemia in rats, and that the support this finding. incidence and the degree of acute pancreatitis are In summary, this study shows that perilobular high when they co-exist.* It has been reported that necrosis of the pancreas frequently develops in congestion of the gut may lead to endotoxaemia in patients with liver disease, and that the development animals through diffusion of endotoxin from the may be associated with sequelae of splanchnic injured In patients with liver disease, such as congestion, namely congestion of the pancreas and liver cirrhosis, liver fibrosis, hepatic infarction, or endotoxaemia due to congestion of the gut. fulminant hepatitis, endotoxaemia is often a result of splanchnic congestion due to portal hypertension and of depressed reticuloendothelial f ~ n c t i o n . ~ - ' ~ REFERENCES These disorders may play a role in the development of perilobular necrosis in many patients with liver 1. Foulis AK. Histologic evidenceofinitiating factors in acute pancreatitis inman. JC/inPathol1980;33: 1125-1131. disease. 2. Shibayama Y. Pancreatic venous stasis and endotoxaemia as aetioJaundice was observed in many patients with logic factors in acute haemorrhagic pancreatitis. J Parhol 1987; 1 5 2 177-1 82. pancreatic perilobular necrosis in this study. It 3. Ravin HA, Rowley D, Jenkins C, Fine J, On the absorption of would be expected in patients with liver disease. bacterial endotoxin from the gastrointestinal tract of the normal and However, it was also seen in many patients with shocked animal. J Exp Med 1960; 112: 783-792. pancreatic perilobular necrosis and bacterial in- 4. Cans H, Matsumoto K. The escape of endotoxin from the intestine. Surg Gynecol Obstet 1974; 1 3 9 395-402. fection or cardiovascular disease, perhaps because 5. Nolan JP, HareDK, McDevitt JJ,AliMV. Invifrostudiesofintestinal of intrahepatic cholestasis due to end~toxaemia,'~ endotoxin absorption. Gasfroenterology 1977; 7 2 434-439. 6. Olofsson P, Nylander G, Olsson P. Endotoxin-transport routes and since endotoxaemia is known to be a factor leading kinetics in intestinal ischemia. Acra Chir Scand 1985; 151; 635-639. to jaundice in bacterial i n f e ~ t i o n . ' ~The , ' ~ exact 7. Wilkinson SP, Arroyo V, Gazzard BG, Moodie H, Williams R. Relation of renal impairment and haemorrhagic diathesis to endomechanism for the development of jaundice in toxaemia in fulminant hepatic failure. Lancet 1974; 1: 521-524. patients with heart disease is not certain,l6.l7 but 8. Tarao K, So K, Moroi T, ef a!. Detection of endotoxin in plasma congestive cardiac failure may lead to splanchnic and ascitic fluid of patients with cirrhosis: its clinical significance. Gasrroenfero1og.v1977; 73: 539-542. congestion or a decrease in the splanchnic blood 9. Magliulo E, Cruclani M, Dietz A, e t a / . Endotoxaemia in acute viral flow and subsequently induce endotoxaemia. 1981; 28: 299-303. hepatitis. Hepafo~a~trornferolog.~ Perilobular necrosis of the pancreas was observed 10. Gaeta GB, Perna P, Adinolfi LE, Utili R, Ruggiero G. Endotoxemia 104 patients with chronic liver disease: prevalence and in a series of in 44 per cent of autopsied patients with DIC in this significance. Digestion 1982; 23: 239-244. study. In such cases, fibrin thrombi were found in I I . Bigatello LM, Broitman SA, Fattori L, el a/. Endotoxemia, encephalopathy, and mortality in cirrhotic patients. Am J Gastroenferol1987; capillaries and venules at the boundary between 11-15, areas of necrotic and viable pancreas. This finding 12. 82: Lumsden AB, Henderson JM, Kutner MH. Endotoxin levels suggests that DIC itself can be a cause of pancreatic measured by a chromogenic assay in portal, hepatic and peripheral venous blood in patients with cirrhosis. Hepatology 1988; 8: 232-236. perilobular necrosis. However, in this study 13. Utili R, Abernathy CO, Zimmerrnan HJ. Cholestaticeffects of Eschersplanchnic congestion was seen in 71 per cent of irhia coliendotoxin on the isolated perfused rat liver. Gastroenterology DIC patients with perilobular necrosis, indicating 1976; 7 0 248-253. Zimmerman HJ, Fang M, Utili R, Seeff LB, Hoofnagle J. Jaundice that splanchnic congestion may be related to the 14. due to bacterial infection. Gastroenterology 1979; 77: 362-374. development of perilobular necrosis of the pancreas 15. Letlowitch JH. Bile ductularcholestasis: an ominous histopathologic sign related to sepsis and 'cholangitis lenta'. Hum Pathol 1982; 13: even in patients with DIC. 19-24, Perilobular necrosis of the pancreas was observed 16. Sherlock S . The liver in heart failure. Relation of anatomical, in 33 per cent of autopsied patients with severe functional, and circulatory changes. Br Hearr J 1951; 13: 273-293. bacterial infection. In these cases, endotoxaemia 17. Dunn GD, Hayes P, Breen KJ, Schenker S. The liver in congestive heart failure: a review. Am J MedSci 1973; 265 174-189. is a possible complication," because severe bac- 18. Olofsson P, Nylander G, Olsson P. Endotoxins: routes of transport in terial infection can originate from liver abscess experimental peritonitis. Am J Surg 1986; 151: 443-446.
