Clin J Gastroenterol (2011) 4:249–254 DOI 10.1007/s12328-011-0229-6
CASE REPORT
Fulminant type of emphysematous pancreatitis has risk of massive hemorrhage Hirotake Komatsu • Hiroshi Yoshida • Hiroki Hayashi • Naoaki Sakata • Takanori Morikawa Tohru Onogawa • Fuyuhiko Motoi • Toshiki Rikiyama • Yu Katayose • Shinichi Egawa • Morihisa Hirota • Tooru Shimosegawa • Michiaki Unno
•
Received: 7 December 2010 / Accepted: 10 April 2011 / Published online: 19 June 2011 Ó Springer 2011
Abstract Emphysematous pancreatitis (air in the parenchyma) was previously considered an indication for surgery, but some recent studies have reported good clinical outcomes with non-operative management. As a step toward establishing a better treatment strategy, we report a case of fulminant pancreatitis with massive hemorrhage into the emphysematous space. A 75-year-old man was admitted with worsening abdominal pain with obstructive jaundice and renal failure 28 h after the onset. He was diagnosed as having emphysematous pancreatitis with slight pancreatic necrosis. Despite conservative treatment with intensive care, sudden cardiac and respiratory failure occurred, and he died 53 h after onset. The autopsy findings revealed biliary sludge and massive bleeding in the retroperitoneal space around the pancreas, suggesting that temporary obstruction of the bile duct with sludge induced emphysema and the hemorrhage rapidly spread into the broadened emphysematous space. Whereas conservative management has been thought to be appropriate in selected cases of emphysematous pancreatitis, when there is pancreatic emphysema in the early phase, a fulminant course tends to develop. Since there is a risk of massive bleeding into the emphysematous space, endoscopic or invasive H. Komatsu H. Yoshida H. Hayashi N. Sakata T. Morikawa T. Onogawa F. Motoi T. Rikiyama Y. Katayose S. Egawa M. Unno (&) Division of Hepato-Biliary Pancreatic Surgery, Department of Surgery, Tohoku University Graduate School of Medicine, 1-1 Seiryo-machi, Aoba-ku, Sendai, Miyagi 980-8574, Japan e-mail: [email protected] M. Hirota T. Shimosegawa Division of Gastroenterology, Tohoku University Graduate School of Medicine, 1-1 Seiryo-machi, Aoba-ku, Sendai, Miyagi 980-8574, Japan
drainage performed to collapse the emphysematous space could benefit the outcome. Keywords Acute necrotizing pancreatitis Emphysema Computed tomography (CT) Hemorrhagic shock
Introduction Emphysematous pancreatitis (EP) is a rare subtype of severe acute pancreatitis characterized by air in the pancreatic parenchyma. It has been considered an indication for surgical management including necrosectomy, but the outcome has been associated with high morbidity and mortality [1]. Several recent reports have suggested that EP could be managed with good outcome by conservative treatment with intensive care. However, these selected patients with better outcomes correlate with delayed gas formation [2, 3]. In contrast, when pancreatic gas is found in the early phase, a fulminant clinical course tends to develop [4, 5]. Thus, the surgical indication for EP has become controversial again. Recently, we experienced a fatal case of EP with a fulminant clinical course, in spite of intensive care. The aim of this case report is to reveal the risk factors based on the outcome and autopsy findings, and to clarify the best management for EP.
Case report A 75-year-old man was admitted with worsening abdominal pain with nausea and vomiting 28 h after onset. He had no history of alcohol abuse. He showed jaundice, oliguria, and slight fever (37.2°C) with tachypnea and tachycardia. Laboratory analysis revealed high levels of serum amylase
123
250
(1,661 U/l), white blood cells (9,800/ll), C-reactive protein (13.8 mg/ml), total bilirubin (6.3 mg/dl), aspartate aminotransferase (769 IU/l), alanine aminotransferase (585 IU/l), alkaline phosphatase (457 IU/l), c-glutamyl transpeptidase (616 IU/l), blood urea nitrogen (52 mg/dl) and creatinine (2.3 mg/dl), suggesting severe acute pancreatitis associated with obstructive jaundice and renal failure. Scores of Ranson’s criteria at admission were 3 (age in years [55 years, serum AST [250 IU/l and serum LDH [350 IU/l), and the Acute Physiology And Chronic Health Evaluation (APACHE) II score was 11 points. Chest X-ray in supine position revealed elevated diaphragm and heart enlargement (Fig. 1a). Abdominal X-ray depicted slightly dilated and paralytic small intestine, suggesting that inflammation affected bowel movement (Fig. 1b). Abdominal computed tomography (CT) showed intraparenchymal mottled gas throughout the pancreas (Fig. 2a). A small amount of free air was detected in the abdominal cavity (Fig. 2b). The CT Severity Index (CTSI) was 6 points, based on the presence of ‘‘gas bubbles’’ (4 points) and ‘‘0–30% necrosis’’ (2 points). Upper gastrointestinal endoscopy revealed neither ulcer nor perforation in the gastrointestinal tract. Furthermore, abdominal CT after the endoscopy showed no increase of free air in the abdominal cavity. These findings supported the suggestion that there was no perforation of the gastrointestinal tract and that the free air came through the disrupted pancreatic capsule. Therefore, we diagnosed the patient as EP. Though he had jaundice with high total bilirubin, there was no biliary dilatation or biliary stone on CT scan. Therefore, we did not perform endoscopic sphincterotomy or biliary stenting. Subsequently, we started non-surgical intensive treatment because his vital signs were relatively stable, and there was no sign of septic shock. Another reason for conservative treatment was that surgery would have been a burden on his hepatic and renal dysfunction. He remained stable hemodynamically in the intensive care unit under conservative treatment including sufficient fluid supply, intravenous administration of antibiotics, protease inhibitor and continuous hemodiafiltration (CHDF). Ten hours after starting the treatment, circulatory and respiratory failure occurred suddenly. Taking septic shock into consideration, respiratory support and endotoxin absorption therapy were administered. However, the patient died at 25 h after his admission (53 h after onset). Under informed consent of his family, an autopsy was performed. The whole pancreatic parenchyma showed hemorrhagic necrosis but no liquefaction (Fig. 3a). A large amount of clotted blood and ruptured mesenteric veins destructed by proteolysis were seen in the retropancreatic space, suggesting that hemorrhage occurred into the emphysematous space. Additionally, a small perforation was seen on the surface of the pancreatic body, through
123
Clin J Gastroenterol (2011) 4:249–254
Fig. 1 Chest X-ray and abdominal X-ray at admission. a Chest X-ray in supine position revealed elevated diaphragm and heart enlargement. b Abdominal X-ray depicted slightly dilated and paralytic small intestine, suggesting that inflammation weakened bowel movement
which air escaped to the abdominal cavity (Fig. 3b). No other perforation of the gastrointestinal tract was detected. We identified some sludge in the gallbladder, but no gallstones in the bile duct. Pancreatitis due to obstruction of the bile duct by sludge was considered as the final diagnosis in this case. Microbial culture of the autopsy specimen was not performed, and there was no growth of microbial organisms in the blood culture examination.
Discussion EP is quite rare, and few well-documented cases (only 18 cases had been published by 2009, Table 1) have been
Clin J Gastroenterol (2011) 4:249–254
published in the literature. Though the incidence rate of pancreas with intrapancreatic air is considered to be approximately 2% in patients with pancreatic disease
Fig. 2 Contrast-enhanced CT of the abdomen at admission. a Intraparechymal mottled gas was seen throughout the pancreas. b A small amount of free air was detected on the surface of the liver and by the gallbladder bed. Some effusion appeared in the left subphrenic space
251
[6, 7], the exact incidence rate of patients with emphysema in acute pancreatitis has not been clarified yet. Traditionally, EP has been regarded as the most severe form of pancreatitis, with high morbidities and mortality. Simchuk and colleagues revealed that a high CTSI score significantly correlated with death and need for necrosectomy in a series of 268 patients with acute pancreatitis [8]. In the CTSI scoring system, pancreatic emphysema is regarded as the most severe state. On the other hand, several recent reports support the notion that EP does not always constitute a severe entity of pancreatitis [2, 3]. Wig et al. [1] showed that there was no significant difference in mortality and incidence of multiple organ failure between EP and non-emphysematous infected pancreatic necrosis. Kvinlaug et al. [2] reported five consecutive cases of EP successfully managed with conservative intensive treatment that had favorable long-term outcomes and concluded that EP is a less aggressive form of acute pancreatitis. According to these results, the rationale for surgical intervention has again become controversial. The mechanism of EP is not well understood, which makes it difficult to select an appropriate treatment. Based on recent reports, EP may be divided into two subtypes: an aggressive type with acute onset and a less aggressive type with delayed onset. Most of the successful cases with nonsurgical treatment had delayed gas formation in the necrotic tissue [2]. The difference in the time of the gas appearance could be associated with the etiology. The proposed causes of emphysema include: (1) gas-forming bacteria infection; (2) fistula from the adjacent gastrointestinal tract; (3) reflux of enteric air through the papilla of Vater and pancreatic duct. Bacterial infection of the necrotic tissue can be induced in the late phase of necrotizing pancreatitis. Fistula formation between a capsule of necrosis and the gastrointestinal wall can also occur in the late phase because internal pressure in the capsulized necrosis gradually increases, leading to the fistula formation as a natural form of drainage [2].
Fig. 3 Autopsy findings. a Sagittal section of the pancreatic head. The whole pancreatic parenchyma showed hemorrhagic necrosis and few normal areas remained. b A small perforation was seen on the surface of the pancreatic body through which the air escaped to the abdominal cavity from the necrotic parenchyma of the pancreas
123
252
Clin J Gastroenterol (2011) 4:249–254
Table 1 Reported cases of emphysematous pancreatitis Case
Author
Year
Age
Gender
Duration from the onset to diagnosis as EP
Pancreatic necrosisb
Surgical treatments
Outcome
1
Morris [15]
1993
58
M
3 weeks
30–50%a
CT-guided percutaneous drainage
Improved
2
Daly [9]
1995
30
F
24 h
[50%a
1. Percutaneous drainage (for the first time)
Improved
2. Surgical drainage (for recurrence) 3
Birgisson [4]
2001
50
M
8h
30–50%a
1. Exploratory laparotomy
Improved
2. Drainage of the abscess 4
Bazan [16]
2003
78
M
1 week
[50%a
5
Stockinger [17]
2004
62
M
2 days
6
Ghidirim [18]
2005
26
M
7 days
7
De Silva [5]
2006
75
M
24 h
Improved
[50%a
Necrosectomy and cholecystectomy Total colectomy for ischemia due to mesenteric venous thrombosis and pancreatic necrosectomy
[50%a
1. Necrosectomy, open packing
Improved
[50%a
2. Scheduled repeated debridements 1. Lavage and drainage
Improved
Dead
2–8. Additional 7 times laparotomies 8
De Silva [5]
2006
67
M
2 days
\30%a
9
2007
72
F
ND
30–50%
Necrosectomy and closed lavage
Improved
2007
66
M
4 days
30–50%
Necrosectomy and open packing
Improved
2007
74
M
3 days
\30%
Necrosectomy and closed lavage
Improved
12
Sˇileikis [3] Sˇileikis [3] Sˇileikis [3] Sˇileikis [3]
2007
67
F
ND
\30%
None
Improved
13
Singh [19]
2008
65
ND
1 day
[50%a
Debridement and drainage
Improved
14
Kvinlaug [2]
2009
71
M
1 week
[50%
None
Improved
15 16
Kvinlaug [2] Kvinlaug [2]
2009 2009
61 60
M M
5 months 3 weeks
[50% [50%
None None
Improved Improved
17
Kvinlaug [2]
2009
77
M
3 weeks
30–50%
None
Improved
18
Kvinlaug [2]
2009
50
M
8 weeks
[50%
None
Improved
19
Present case
2010
75
M
28 h
\30%
None
Dead
1. Lavage and drainage
Improved
2. Lavage 10 11
ND not described a
Not described in the article and judged based on the attached CT image in the article
b
Necrotic area is divided into three classes: less than 30%, from 30 to 50% and more than 50%, according to the necrosis score using in the CTSI score
In contrast, pancreatic emphysema in the early phase could be induced by the reflux of enteric air into the pancreatic duct. One of the feasible causes for the reflux is a biliary stone or sludge in the common bile duct. When it forcibly passes through the sphincter of Oddi, the air in the duodenum can get into the pancreatic duct through the dilated sphincter. In this present patient, we could not detect any perforation or fistula of the gastrointestinal tract in the endoscopic and autopsy findings. Infection of the necrotic tissue leading to gas formation was not likely, because the ascites and the necrotic pancreas were not purulent, and the blood
123
culture was negative. Accordingly, we assumed that the etiology of EP in this case was reflux of enteric air because of the temporary obstruction with biliary sludge [9]. The important point is that emphysema itself does not necessarily imply a lethal state directly, but septicemia and massive hemorrhage into the emphysematous space that occur with emphysema can result in the fatal condition. This patient could be classified as having the aggressive type with high mortality, and one of the possible causes of death was hypovolemic shock due to hemorrhage around the pancreas. In non-emphysematous pancreatitis, hemorrhage into the retroperitoneal space tends to occur rather
Clin J Gastroenterol (2011) 4:249–254
slowly because of its limited space. On the other hand, especially in EP patients, acute hypovolemic shock due to massive hemorrhage might be a characteristic manifestation because hemorrhage into the broadened emphysematous space can easily spread. Therefore, not only arterial hemorrhage, but also venous hemorrhage with low pressure could be lethal in this particular situation. In fact, this patient showed no aneurysm upon CT scan and massive hemorrhage from the mesenteric vein was confirmed by the autopsy. Generally, both septicemia from pancreatic infection and hypovolemia from retroperitoneal hemorrhage are the major fatal complications in acute pancreatitis. Each of them is important as a cause of shock, but it is clinically difficult to distinguish between them. Certain kinds of drainage such as ENPD or minimally invasive surgical drainage might have been useful for making the correct diagnosis in the present case. Regarding surgical intervention, Navaneethan et al. [10] reviewed strategic approaches for necrotizing pancreatitis by minimally invasive techniques to access the necrotic space. Percutaneous interventional drainage, an endoscopic transgastric or transduodenal approach, laparoscopic debridement and a retroperitoneal approach were introduced as representative ways to access the necrotic space in the review. Although these approaches have their respective benefits, it has not been determined which technique is best. In addition, these techniques are limited to large medical centers whose staff physicians have sufficient expertise. Practically, drainage with endoscopic retrograde naso-pancreatic drainage (ENPD) is widely accepted and may be useful for EP. Shinozuka et al. [11] reported successful treatment using ENPD and pancreatic duct stent for patients with acute pancreatitis and pancreatic pseudocysts [12]. Although this procedure provides no evidence for EP, decompression in the pancreatic duct can be beneficial to the elimination of the emphysematous space where the massive bleeding can occur. Additionally, it can provide information about lethal massive hemorrhage, leading to the correct diagnosis promptly. Another feasible treatment is continuous regional arterial infusion (CRAI) of protease inhibitor and antibiotics. This treatment is advantageous especially for the necrotizing pancreatitis because necrosis and infection induce microcirculatory disorders, which lower the drug transition into the pancreatic parenchyma. CRAI therapy enables high drug concentrations in the pancreas and contributes to preventing the development of pancreatic infection [13, 14]. In this patient, we estimated that necrotic area was less than 30% of the pancreas on CT scan. Therefore, we did not perform CRAI therapy for him. In conclusion, we experienced the case of a patient with fulminant EP who could not be rescued. It should be
253
emphasized that EP can have quite different outcomes depending on the mechanisms of emphysema formation. Pancreatic emphysema in the early phase should be given special attention, even when pancreatic necrosis is slight, as in this case report. Massive hemorrhage into the emphysematous space may occur unexpectedly in this condition. This patient’s outcome implicated a significant principle for future research and treatment. Earlier endoscopic or percutaneous intervention may have benefit to diagnose correctly and to collapse the emphysema, which could contribute to improving the clinical course. Acknowledgments This study was partly supported by a Grant-inAid from the Research Committee of Intractable Pancreatic Diseases (Principal investigator: Tooru Shimosegawa) provided by the Ministry of Health, Labour, and Welfare of Japan.
References 1. Wig JD, Kochhar R, Bharathy KG, Kudari AK, Doley RP, Yadav TD, et al. Emphysematous pancreatitis. Radiological curiosity or a cause for concern? JOP. 2008;9:160–6. 2. Kvinlaug K, Kriegler S, Moser M. Emphysematous pancreatitis: a less aggressive form of infected pancreatic necrosis? Pancreas. 2009;38:667–71. 3. Sˇileikis A, Beisˇa V, Dementavicˇien_e J, Tamosˇi_enas A, Strupas K. Experience of the treatment of emphysematous necrotizing pancreatitis. Chir Gastroenterol. 2007;23:195–8. 4. Birgisson H, Stefansson T, Andresdottir A, Moller PH. Emphysematous pancreatitis. Eur J Surg. 2001;167:918–20. 5. De Silva NM, Windsor JA. Clostridium perfringens infection of pancreatic necrosis: absolute indication for early surgical intervention. ANZ J Surg. 2006;76:757–9. 6. Mendez G Jr, Isikoff MB. Significance of intrapancreatic gas demonstrated by CT: a review of nine cases. AJR Am J Roentgenol. 1979;132:59–62. 7. Torres WE, Clements JL Jr, Sones PJ, Knopf DR. Gas in the pancreatic bed without abscess. AJR Am J Roentgenol. 1981;137:1131–3. 8. Simchuk EJ, Traverso LW, Nukui Y, Kozarek RA. Computed tomography severity index is a predictor of outcomes for severe pancreatitis. Am J Surg. 2000;179:352–5. 9. Daly JJ Jr, Alderman DF, Conway WF. General case of the day. Emphysematous pancreatitis. Radiographics. 1995;15:489–92. 10. Navaneethan U, Vege SS, Chari ST, Baron TH. Minimally invasive techniques in pancreatic necrosis. Pancreas. 2009; 38:867–75. 11. Shinozuka N, Okada K, Torii T, Hirooka E, Tabuchi S, Aikawa K, et al. Endoscopic pancreatic duct drainage and stenting for acute pancreatitis and pancreatic cyst and abscess. J Hepatobiliary Pancreat Surg. 2007;14:569–74. 12. Matsuhashi T, Kikuyama M, Sasada Y, Ota Y, Nakahodo J. Endoscopic nasopancreatic drainage was effective on pancreatic abscess after severe acute pancreatitis with pneumoperitoneum: a case report (in Japanese with English abstract). Gastroenterol Endosc. 2009;51:1159–64. 13. Takeda K, Matsuno S, Ogawa M, Watanabe S, Atomi Y. Continuous regional arterial infusion (CRAI) therapy reduces the mortality rate of acute necrotizing pancreatitis: results of a cooperative survey in Japan. J Hepatobiliary Pancreat Surg. 2001;8:216–20.
123
254 14. Takeda K, Yamauchi J, Shibuya K, Sunamura M, Mikami Y, Matsuno S. Benefit of continuous regional arterial infusion of protease inhibitor and antibiotic in the management of acute necrotizing pancreatitis. Pancreatology. 2001;1:668–73. 15. Morris DL, Wilkinson LS, al Mokhtar N. Case report: emphysematous tuberculous pancreatitis diagnosis by ultrasound and computed tomography. Clin Radiol. 1993;48:286–7. 16. Bazan HA, Kim U. Images in clinical medicine. Emphysematous pancreatitis. N Engl J Med. 2003;349:e25.
123
Clin J Gastroenterol (2011) 4:249–254 17. Stockinger ZT, Corsetti RL. Pneumoperitoneum from gas gangrene of the pancreas: three unusual findings in a single case. J Gastrointest Surg. 2004;8:489–92. 18. Ghidirim G, Gagauz I, Misin I, Gutu E, Vozian M. Emphysematous necrotizing pancreatitis. Chirurgia (Bucur). 2005; 100:293–6. 19. Singh A. Acute emphysematous necrotizing pancreatitis. J Assoc Physicians India. 2008;56:519.
CASE REPORT
Fulminant type of emphysematous pancreatitis has risk of massive hemorrhage Hirotake Komatsu • Hiroshi Yoshida • Hiroki Hayashi • Naoaki Sakata • Takanori Morikawa Tohru Onogawa • Fuyuhiko Motoi • Toshiki Rikiyama • Yu Katayose • Shinichi Egawa • Morihisa Hirota • Tooru Shimosegawa • Michiaki Unno
•
Received: 7 December 2010 / Accepted: 10 April 2011 / Published online: 19 June 2011 Ó Springer 2011
Abstract Emphysematous pancreatitis (air in the parenchyma) was previously considered an indication for surgery, but some recent studies have reported good clinical outcomes with non-operative management. As a step toward establishing a better treatment strategy, we report a case of fulminant pancreatitis with massive hemorrhage into the emphysematous space. A 75-year-old man was admitted with worsening abdominal pain with obstructive jaundice and renal failure 28 h after the onset. He was diagnosed as having emphysematous pancreatitis with slight pancreatic necrosis. Despite conservative treatment with intensive care, sudden cardiac and respiratory failure occurred, and he died 53 h after onset. The autopsy findings revealed biliary sludge and massive bleeding in the retroperitoneal space around the pancreas, suggesting that temporary obstruction of the bile duct with sludge induced emphysema and the hemorrhage rapidly spread into the broadened emphysematous space. Whereas conservative management has been thought to be appropriate in selected cases of emphysematous pancreatitis, when there is pancreatic emphysema in the early phase, a fulminant course tends to develop. Since there is a risk of massive bleeding into the emphysematous space, endoscopic or invasive H. Komatsu H. Yoshida H. Hayashi N. Sakata T. Morikawa T. Onogawa F. Motoi T. Rikiyama Y. Katayose S. Egawa M. Unno (&) Division of Hepato-Biliary Pancreatic Surgery, Department of Surgery, Tohoku University Graduate School of Medicine, 1-1 Seiryo-machi, Aoba-ku, Sendai, Miyagi 980-8574, Japan e-mail: [email protected] M. Hirota T. Shimosegawa Division of Gastroenterology, Tohoku University Graduate School of Medicine, 1-1 Seiryo-machi, Aoba-ku, Sendai, Miyagi 980-8574, Japan
drainage performed to collapse the emphysematous space could benefit the outcome. Keywords Acute necrotizing pancreatitis Emphysema Computed tomography (CT) Hemorrhagic shock
Introduction Emphysematous pancreatitis (EP) is a rare subtype of severe acute pancreatitis characterized by air in the pancreatic parenchyma. It has been considered an indication for surgical management including necrosectomy, but the outcome has been associated with high morbidity and mortality [1]. Several recent reports have suggested that EP could be managed with good outcome by conservative treatment with intensive care. However, these selected patients with better outcomes correlate with delayed gas formation [2, 3]. In contrast, when pancreatic gas is found in the early phase, a fulminant clinical course tends to develop [4, 5]. Thus, the surgical indication for EP has become controversial again. Recently, we experienced a fatal case of EP with a fulminant clinical course, in spite of intensive care. The aim of this case report is to reveal the risk factors based on the outcome and autopsy findings, and to clarify the best management for EP.
Case report A 75-year-old man was admitted with worsening abdominal pain with nausea and vomiting 28 h after onset. He had no history of alcohol abuse. He showed jaundice, oliguria, and slight fever (37.2°C) with tachypnea and tachycardia. Laboratory analysis revealed high levels of serum amylase
123
250
(1,661 U/l), white blood cells (9,800/ll), C-reactive protein (13.8 mg/ml), total bilirubin (6.3 mg/dl), aspartate aminotransferase (769 IU/l), alanine aminotransferase (585 IU/l), alkaline phosphatase (457 IU/l), c-glutamyl transpeptidase (616 IU/l), blood urea nitrogen (52 mg/dl) and creatinine (2.3 mg/dl), suggesting severe acute pancreatitis associated with obstructive jaundice and renal failure. Scores of Ranson’s criteria at admission were 3 (age in years [55 years, serum AST [250 IU/l and serum LDH [350 IU/l), and the Acute Physiology And Chronic Health Evaluation (APACHE) II score was 11 points. Chest X-ray in supine position revealed elevated diaphragm and heart enlargement (Fig. 1a). Abdominal X-ray depicted slightly dilated and paralytic small intestine, suggesting that inflammation affected bowel movement (Fig. 1b). Abdominal computed tomography (CT) showed intraparenchymal mottled gas throughout the pancreas (Fig. 2a). A small amount of free air was detected in the abdominal cavity (Fig. 2b). The CT Severity Index (CTSI) was 6 points, based on the presence of ‘‘gas bubbles’’ (4 points) and ‘‘0–30% necrosis’’ (2 points). Upper gastrointestinal endoscopy revealed neither ulcer nor perforation in the gastrointestinal tract. Furthermore, abdominal CT after the endoscopy showed no increase of free air in the abdominal cavity. These findings supported the suggestion that there was no perforation of the gastrointestinal tract and that the free air came through the disrupted pancreatic capsule. Therefore, we diagnosed the patient as EP. Though he had jaundice with high total bilirubin, there was no biliary dilatation or biliary stone on CT scan. Therefore, we did not perform endoscopic sphincterotomy or biliary stenting. Subsequently, we started non-surgical intensive treatment because his vital signs were relatively stable, and there was no sign of septic shock. Another reason for conservative treatment was that surgery would have been a burden on his hepatic and renal dysfunction. He remained stable hemodynamically in the intensive care unit under conservative treatment including sufficient fluid supply, intravenous administration of antibiotics, protease inhibitor and continuous hemodiafiltration (CHDF). Ten hours after starting the treatment, circulatory and respiratory failure occurred suddenly. Taking septic shock into consideration, respiratory support and endotoxin absorption therapy were administered. However, the patient died at 25 h after his admission (53 h after onset). Under informed consent of his family, an autopsy was performed. The whole pancreatic parenchyma showed hemorrhagic necrosis but no liquefaction (Fig. 3a). A large amount of clotted blood and ruptured mesenteric veins destructed by proteolysis were seen in the retropancreatic space, suggesting that hemorrhage occurred into the emphysematous space. Additionally, a small perforation was seen on the surface of the pancreatic body, through
123
Clin J Gastroenterol (2011) 4:249–254
Fig. 1 Chest X-ray and abdominal X-ray at admission. a Chest X-ray in supine position revealed elevated diaphragm and heart enlargement. b Abdominal X-ray depicted slightly dilated and paralytic small intestine, suggesting that inflammation weakened bowel movement
which air escaped to the abdominal cavity (Fig. 3b). No other perforation of the gastrointestinal tract was detected. We identified some sludge in the gallbladder, but no gallstones in the bile duct. Pancreatitis due to obstruction of the bile duct by sludge was considered as the final diagnosis in this case. Microbial culture of the autopsy specimen was not performed, and there was no growth of microbial organisms in the blood culture examination.
Discussion EP is quite rare, and few well-documented cases (only 18 cases had been published by 2009, Table 1) have been
Clin J Gastroenterol (2011) 4:249–254
published in the literature. Though the incidence rate of pancreas with intrapancreatic air is considered to be approximately 2% in patients with pancreatic disease
Fig. 2 Contrast-enhanced CT of the abdomen at admission. a Intraparechymal mottled gas was seen throughout the pancreas. b A small amount of free air was detected on the surface of the liver and by the gallbladder bed. Some effusion appeared in the left subphrenic space
251
[6, 7], the exact incidence rate of patients with emphysema in acute pancreatitis has not been clarified yet. Traditionally, EP has been regarded as the most severe form of pancreatitis, with high morbidities and mortality. Simchuk and colleagues revealed that a high CTSI score significantly correlated with death and need for necrosectomy in a series of 268 patients with acute pancreatitis [8]. In the CTSI scoring system, pancreatic emphysema is regarded as the most severe state. On the other hand, several recent reports support the notion that EP does not always constitute a severe entity of pancreatitis [2, 3]. Wig et al. [1] showed that there was no significant difference in mortality and incidence of multiple organ failure between EP and non-emphysematous infected pancreatic necrosis. Kvinlaug et al. [2] reported five consecutive cases of EP successfully managed with conservative intensive treatment that had favorable long-term outcomes and concluded that EP is a less aggressive form of acute pancreatitis. According to these results, the rationale for surgical intervention has again become controversial. The mechanism of EP is not well understood, which makes it difficult to select an appropriate treatment. Based on recent reports, EP may be divided into two subtypes: an aggressive type with acute onset and a less aggressive type with delayed onset. Most of the successful cases with nonsurgical treatment had delayed gas formation in the necrotic tissue [2]. The difference in the time of the gas appearance could be associated with the etiology. The proposed causes of emphysema include: (1) gas-forming bacteria infection; (2) fistula from the adjacent gastrointestinal tract; (3) reflux of enteric air through the papilla of Vater and pancreatic duct. Bacterial infection of the necrotic tissue can be induced in the late phase of necrotizing pancreatitis. Fistula formation between a capsule of necrosis and the gastrointestinal wall can also occur in the late phase because internal pressure in the capsulized necrosis gradually increases, leading to the fistula formation as a natural form of drainage [2].
Fig. 3 Autopsy findings. a Sagittal section of the pancreatic head. The whole pancreatic parenchyma showed hemorrhagic necrosis and few normal areas remained. b A small perforation was seen on the surface of the pancreatic body through which the air escaped to the abdominal cavity from the necrotic parenchyma of the pancreas
123
252
Clin J Gastroenterol (2011) 4:249–254
Table 1 Reported cases of emphysematous pancreatitis Case
Author
Year
Age
Gender
Duration from the onset to diagnosis as EP
Pancreatic necrosisb
Surgical treatments
Outcome
1
Morris [15]
1993
58
M
3 weeks
30–50%a
CT-guided percutaneous drainage
Improved
2
Daly [9]
1995
30
F
24 h
[50%a
1. Percutaneous drainage (for the first time)
Improved
2. Surgical drainage (for recurrence) 3
Birgisson [4]
2001
50
M
8h
30–50%a
1. Exploratory laparotomy
Improved
2. Drainage of the abscess 4
Bazan [16]
2003
78
M
1 week
[50%a
5
Stockinger [17]
2004
62
M
2 days
6
Ghidirim [18]
2005
26
M
7 days
7
De Silva [5]
2006
75
M
24 h
Improved
[50%a
Necrosectomy and cholecystectomy Total colectomy for ischemia due to mesenteric venous thrombosis and pancreatic necrosectomy
[50%a
1. Necrosectomy, open packing
Improved
[50%a
2. Scheduled repeated debridements 1. Lavage and drainage
Improved
Dead
2–8. Additional 7 times laparotomies 8
De Silva [5]
2006
67
M
2 days
\30%a
9
2007
72
F
ND
30–50%
Necrosectomy and closed lavage
Improved
2007
66
M
4 days
30–50%
Necrosectomy and open packing
Improved
2007
74
M
3 days
\30%
Necrosectomy and closed lavage
Improved
12
Sˇileikis [3] Sˇileikis [3] Sˇileikis [3] Sˇileikis [3]
2007
67
F
ND
\30%
None
Improved
13
Singh [19]
2008
65
ND
1 day
[50%a
Debridement and drainage
Improved
14
Kvinlaug [2]
2009
71
M
1 week
[50%
None
Improved
15 16
Kvinlaug [2] Kvinlaug [2]
2009 2009
61 60
M M
5 months 3 weeks
[50% [50%
None None
Improved Improved
17
Kvinlaug [2]
2009
77
M
3 weeks
30–50%
None
Improved
18
Kvinlaug [2]
2009
50
M
8 weeks
[50%
None
Improved
19
Present case
2010
75
M
28 h
\30%
None
Dead
1. Lavage and drainage
Improved
2. Lavage 10 11
ND not described a
Not described in the article and judged based on the attached CT image in the article
b
Necrotic area is divided into three classes: less than 30%, from 30 to 50% and more than 50%, according to the necrosis score using in the CTSI score
In contrast, pancreatic emphysema in the early phase could be induced by the reflux of enteric air into the pancreatic duct. One of the feasible causes for the reflux is a biliary stone or sludge in the common bile duct. When it forcibly passes through the sphincter of Oddi, the air in the duodenum can get into the pancreatic duct through the dilated sphincter. In this present patient, we could not detect any perforation or fistula of the gastrointestinal tract in the endoscopic and autopsy findings. Infection of the necrotic tissue leading to gas formation was not likely, because the ascites and the necrotic pancreas were not purulent, and the blood
123
culture was negative. Accordingly, we assumed that the etiology of EP in this case was reflux of enteric air because of the temporary obstruction with biliary sludge [9]. The important point is that emphysema itself does not necessarily imply a lethal state directly, but septicemia and massive hemorrhage into the emphysematous space that occur with emphysema can result in the fatal condition. This patient could be classified as having the aggressive type with high mortality, and one of the possible causes of death was hypovolemic shock due to hemorrhage around the pancreas. In non-emphysematous pancreatitis, hemorrhage into the retroperitoneal space tends to occur rather
Clin J Gastroenterol (2011) 4:249–254
slowly because of its limited space. On the other hand, especially in EP patients, acute hypovolemic shock due to massive hemorrhage might be a characteristic manifestation because hemorrhage into the broadened emphysematous space can easily spread. Therefore, not only arterial hemorrhage, but also venous hemorrhage with low pressure could be lethal in this particular situation. In fact, this patient showed no aneurysm upon CT scan and massive hemorrhage from the mesenteric vein was confirmed by the autopsy. Generally, both septicemia from pancreatic infection and hypovolemia from retroperitoneal hemorrhage are the major fatal complications in acute pancreatitis. Each of them is important as a cause of shock, but it is clinically difficult to distinguish between them. Certain kinds of drainage such as ENPD or minimally invasive surgical drainage might have been useful for making the correct diagnosis in the present case. Regarding surgical intervention, Navaneethan et al. [10] reviewed strategic approaches for necrotizing pancreatitis by minimally invasive techniques to access the necrotic space. Percutaneous interventional drainage, an endoscopic transgastric or transduodenal approach, laparoscopic debridement and a retroperitoneal approach were introduced as representative ways to access the necrotic space in the review. Although these approaches have their respective benefits, it has not been determined which technique is best. In addition, these techniques are limited to large medical centers whose staff physicians have sufficient expertise. Practically, drainage with endoscopic retrograde naso-pancreatic drainage (ENPD) is widely accepted and may be useful for EP. Shinozuka et al. [11] reported successful treatment using ENPD and pancreatic duct stent for patients with acute pancreatitis and pancreatic pseudocysts [12]. Although this procedure provides no evidence for EP, decompression in the pancreatic duct can be beneficial to the elimination of the emphysematous space where the massive bleeding can occur. Additionally, it can provide information about lethal massive hemorrhage, leading to the correct diagnosis promptly. Another feasible treatment is continuous regional arterial infusion (CRAI) of protease inhibitor and antibiotics. This treatment is advantageous especially for the necrotizing pancreatitis because necrosis and infection induce microcirculatory disorders, which lower the drug transition into the pancreatic parenchyma. CRAI therapy enables high drug concentrations in the pancreas and contributes to preventing the development of pancreatic infection [13, 14]. In this patient, we estimated that necrotic area was less than 30% of the pancreas on CT scan. Therefore, we did not perform CRAI therapy for him. In conclusion, we experienced the case of a patient with fulminant EP who could not be rescued. It should be
253
emphasized that EP can have quite different outcomes depending on the mechanisms of emphysema formation. Pancreatic emphysema in the early phase should be given special attention, even when pancreatic necrosis is slight, as in this case report. Massive hemorrhage into the emphysematous space may occur unexpectedly in this condition. This patient’s outcome implicated a significant principle for future research and treatment. Earlier endoscopic or percutaneous intervention may have benefit to diagnose correctly and to collapse the emphysema, which could contribute to improving the clinical course. Acknowledgments This study was partly supported by a Grant-inAid from the Research Committee of Intractable Pancreatic Diseases (Principal investigator: Tooru Shimosegawa) provided by the Ministry of Health, Labour, and Welfare of Japan.
References 1. Wig JD, Kochhar R, Bharathy KG, Kudari AK, Doley RP, Yadav TD, et al. Emphysematous pancreatitis. Radiological curiosity or a cause for concern? JOP. 2008;9:160–6. 2. Kvinlaug K, Kriegler S, Moser M. Emphysematous pancreatitis: a less aggressive form of infected pancreatic necrosis? Pancreas. 2009;38:667–71. 3. Sˇileikis A, Beisˇa V, Dementavicˇien_e J, Tamosˇi_enas A, Strupas K. Experience of the treatment of emphysematous necrotizing pancreatitis. Chir Gastroenterol. 2007;23:195–8. 4. Birgisson H, Stefansson T, Andresdottir A, Moller PH. Emphysematous pancreatitis. Eur J Surg. 2001;167:918–20. 5. De Silva NM, Windsor JA. Clostridium perfringens infection of pancreatic necrosis: absolute indication for early surgical intervention. ANZ J Surg. 2006;76:757–9. 6. Mendez G Jr, Isikoff MB. Significance of intrapancreatic gas demonstrated by CT: a review of nine cases. AJR Am J Roentgenol. 1979;132:59–62. 7. Torres WE, Clements JL Jr, Sones PJ, Knopf DR. Gas in the pancreatic bed without abscess. AJR Am J Roentgenol. 1981;137:1131–3. 8. Simchuk EJ, Traverso LW, Nukui Y, Kozarek RA. Computed tomography severity index is a predictor of outcomes for severe pancreatitis. Am J Surg. 2000;179:352–5. 9. Daly JJ Jr, Alderman DF, Conway WF. General case of the day. Emphysematous pancreatitis. Radiographics. 1995;15:489–92. 10. Navaneethan U, Vege SS, Chari ST, Baron TH. Minimally invasive techniques in pancreatic necrosis. Pancreas. 2009; 38:867–75. 11. Shinozuka N, Okada K, Torii T, Hirooka E, Tabuchi S, Aikawa K, et al. Endoscopic pancreatic duct drainage and stenting for acute pancreatitis and pancreatic cyst and abscess. J Hepatobiliary Pancreat Surg. 2007;14:569–74. 12. Matsuhashi T, Kikuyama M, Sasada Y, Ota Y, Nakahodo J. Endoscopic nasopancreatic drainage was effective on pancreatic abscess after severe acute pancreatitis with pneumoperitoneum: a case report (in Japanese with English abstract). Gastroenterol Endosc. 2009;51:1159–64. 13. Takeda K, Matsuno S, Ogawa M, Watanabe S, Atomi Y. Continuous regional arterial infusion (CRAI) therapy reduces the mortality rate of acute necrotizing pancreatitis: results of a cooperative survey in Japan. J Hepatobiliary Pancreat Surg. 2001;8:216–20.
123
254 14. Takeda K, Yamauchi J, Shibuya K, Sunamura M, Mikami Y, Matsuno S. Benefit of continuous regional arterial infusion of protease inhibitor and antibiotic in the management of acute necrotizing pancreatitis. Pancreatology. 2001;1:668–73. 15. Morris DL, Wilkinson LS, al Mokhtar N. Case report: emphysematous tuberculous pancreatitis diagnosis by ultrasound and computed tomography. Clin Radiol. 1993;48:286–7. 16. Bazan HA, Kim U. Images in clinical medicine. Emphysematous pancreatitis. N Engl J Med. 2003;349:e25.
123
Clin J Gastroenterol (2011) 4:249–254 17. Stockinger ZT, Corsetti RL. Pneumoperitoneum from gas gangrene of the pancreas: three unusual findings in a single case. J Gastrointest Surg. 2004;8:489–92. 18. Ghidirim G, Gagauz I, Misin I, Gutu E, Vozian M. Emphysematous necrotizing pancreatitis. Chirurgia (Bucur). 2005; 100:293–6. 19. Singh A. Acute emphysematous necrotizing pancreatitis. J Assoc Physicians India. 2008;56:519.
