Acute pancreatitis in marrow transplant patients - Nature

Bone Marrow Transplantation, (1997) 20, 1081–1086  1997 Stockton Press All rights reserved 0268–3369/97 $12.00

Acute pancreatitis in marrow transplant patients: prevalence at autopsy and risk factor analysis CW Ko1, T Gooley2, HG Schoch2, D Myerson3 , RC Hackman3, HM Shulman3, GE Sale3, SP Lee1 and GB McDonald 4 4

Gastroenterology/Hepatology, 2Clinical Statistics and 3Pathology Sections of the Fred Hutchinson Cancer Research Center and the Medical Service, VA Puget Sound Health Center System, and 1University of Washington School of Medicine, Seattle, WA, USA

Summary: Pancreatitis has been described as an infrequent complication of marrow transplantation. This study investigated the prevalence of pancreatitis at autopsy in marrow transplant patients and determined risk factors for its development. We reviewed consecutive autopsy reports from 1991 to 1993. Medical records and laboratory reports were reviewed for analysis of clinical variables. Autopsy findings and clinical variables were correlated with the autopsy diagnosis of pancreatitis. Pancreatitis was found in 51 of 184 (28%) patients at autopsy. Of those with pancreatitis, 35% had abdominal pain, 10% had measurements of serum pancreatic enzymes, and 20% had abdominal imaging studies in the week prior to death. By univariable analysis, risk factors associated with development of pancreatitis included clinical grades 3 and 4 GVHD, GVHD at autopsy, liver GVHD at autopsy, major infection at autopsy, and increasing days of survival. By multivariable analysis, independent risk factors for its development included any GVHD at autopsy, increasing length of survival after transplantation, and major infection at autopsy. We conclude that pancreatitis is a common but often subclinical complication of marrow transplantation. Its development may be associated with a high prevalence of biliary sludge and prolonged treatment of GVHD with cyclosporine and prednisone. Keywords: acute pancreatitis; bone marrow transplant; corticosteroids; graft-versus-host disease; hematopoietic stem cell transplantation; malignant disease

Although pancreatitis is a relatively common disorder in hospital-based populations, it has only recently been described as a complication of marrow transplantation, occurring in 3.5% of children undergoing this procedure at one institution.1 The presence of multiple other causes of abdominal pain; nausea, and vomiting (such as conditioning toxicity, infections, liver disease, or graft-versus-host disease (GVHD) of the intestinal tract) may make it difficult to Correspondence: Dr CW Ko, Gastroenterology/Hepatology Section [SC114], Fred Hutchinson Cancer Research Center, 1124 Columbia Street, Seattle, WA 98104, USA Received 14 April 1997; accepted 16 August 1997

recognize the symptoms of pancreatitis in this population. However, marrow transplant patients may be predisposed to developing pancreatitis due to the high prevalence of biliary sludge and sonographic biliary tract abnormalities, 2,3 use of irradiation and cytotoxic drugs in the conditioning regimen, and treatment with medications that have been associated with pancreatitis such as corticosteroids and cyclosporine.4–12 Infectious etiologies of pancreatitis are also possible because of the profound immunosuppression that develops after transplant. There are reports in marrow transplant patients of disseminated cytomegalovirus, adenovirus, and varicella zoster virus infections involving the pancreas and mimicking attacks of idiopathic pancreatitis.13–16 To estimate the prevalence of pancreatitis at autopsy in marrow transplant patients, we reviewed the histologic and gross findings from all patients autopsied at our center between January 1991 and December 1993. Clinical, radiologic, and laboratory data were also reviewed to analyze the frequency of pre-mortem signs and symptoms in patients with pancreatitis. Gallbladder contents were obtained at autopsy and analyzed chemically and microscopically for the presence of sludge, as previously described.17 To determine risk factors for its development, the frequency of clinical and therapeutic parameters in patients with pancreatitis was compared to those in patients without pancreatitis. Materials and methods Marrow transplant technique Patients undergoing marrow transplantation received conditioning therapy (chemotherapy with or without total body irradiation) as previously described.18 Recipients of allogeneic marrow were treated with prophylactic medications to prevent acute graft-versus-host disease, usually combinations of cyclosporine and methotrexate.19 Patients at risk for herpes virus infection were given prophylactic acyclovir or ganciclovir.20,21 For patients who developed acute graftversus-host disease, prednisone at 1–2 mg/kg/day was the initial therapy.22 Patient selection Between January 1991 and December 1993, 391 patients died at the Fred Hutchinson Cancer Research Center. Of

Pancreatitis in marrow transplant patients CW Ko et al

1082

these, 221 patients (56.5%) came to autopsy. Thirty-seven patients who had either limited autopsies that did not include the pancreas or who died prior to transplantation were excluded from analysis, leaving a final cohort of 184 patients (47% of total deaths) who had undergone marrow transplantation. Recipients of both autologous (n = 31) and allogeneic transplants (n = 153) were included in this analysis. At autopsy, we noted the presence or absence of pancreatitis (as defined below), graft-versus-host disease (GVHD), biliary tract abnormalities (gallbladder sludge or stones, common bile duct sludge or stones, cholecystitis), other liver diseases such as infection or veno-occlusive disease, or any major infection at the time of death. Patients with GVHD were also subclassified by the organs affected (skin, liver, or intestine). Definition of pancreatitis at autopsy Mild pancreatitis was defined as interstitial edema, scattered peripancreatic fat necrosis, and mild inflammation with scattered neutrophilic infiltration. Moderate pancreatitis was defined as moderate inflammation, mild acinar cell necrosis, and moderate peripancreatic fat necrosis. Severe pancreatitis was defined as presence of hemorrhage, widespread parenchymal necrosis, extensive peripancreatic fat necrosis, or pseudocyst formation. Patients in whom pancreatitis was felt to be a cause of death based on pathologic review were also noted.

of GVHD, prophylactic medications used for GVHD, and days of survival from the last transplant. Patients were categorized as being in remission or relapse at time of transplantation. Those with non-malignant disorders or CML in chronic phase were considered to be in remission. For total body irradiation, patients were categorized by whether or not any irradiation was given and by radiation dose (0 Gy, 13.2 Gy, 13.2 to 14.4 Gy, or .14.4 Gy). GVHD was examined in four separate ways – presence of any GVHD at autopsy, liver GVHD at autopsy, intestinal GVHD at autopsy, and overall clinical grade of GVHD (grades 0 to 4).23 Prophylactic medications for GVHD were grouped into the categories of none (autologous marrow recipients), any cyclosporine-containing regimen, or other regimens (generally containing methotrexate or corticosteroids). Liver disease at autopsy included veno-occlusive disease, GVHD, steatosis, cholestasis, recurrent neoplasm, congestion, or hepatic infection. Use of selected medications during the patients’ course was also reviewed, including the day relative to transplant on which the medications were started and the total number of days on which the medications were given. Medications analyzed included amphotericin, penicillins, cephalosporins, aminoglycosides, acyclovir, ganciclovir, omeprazole, trimethoprim-sulfamethoxazole, fluconazole, corticosteroids, cyclosporine, and anti-thymocyte globulin (ATG). Daily corticosteroid doses greater than or equal to the equivalent of 50 mg of hydrocortisone were counted in this analysis.

Clinical and laboratory manifestations of pancreatitis The medical records of patients with pancreatitis were examined for reports of abdominal pain in the time periods of 1 to 2 weeks prior to death and the week prior to death, respectively. Reports of any abdominal imaging studies (ultrasound, CT X-ray, or magnetic resonance) were also reviewed for visualization of the pancreas and pancreatic abnormalities. Laboratory reports for the years 1991 to 1993 were reviewed for measurements of total serum amylase, pancreatic amylase, or lipase. Measurements were considered abnormal if greater than the upper limit of normal for our laboratory. If these measurements were not made during a patient’s course, we assumed that pancreatitis was not suspected clinically. Analysis of measurements in three time periods were made: in the 4 weeks prior, in the 2 weeks prior, and in the week prior to death, to determine if pancreatitis was suspected in the immediate pre-mortem period. Analysis of risk factors for pancreatitis Risk factors examined included age, gender, underlying disease, disease status (relapse vs remission for malignant disorders), type of transplant (autologous vs allogeneic), whether undergoing first or second transplant, pre-transplant cytomegalovirus (CMV) serology of patient and donor, presence and dose of total body irradiation (TBI) during conditioning therapy, engraftment (by trilineage criteria), presence of major infection at autopsy, presence and site of involvement of graft-versus-host disease (GVHD) at autopsy, presence of any liver disease at autopsy, clinical grade

Analysis of gallbladder contents Gallbladder contents were obtained at autopsy by incision and drainage of the gallbladder into a sterile container. Samples were frozen until analysis could be performed. Gallbladder sludge was defined as viscous bile with suspended particles. Particles larger than 2 mm and which could not be crushed digitally were classified as gallstones.

Statistical methods In order to determine which, if any, factors were significantly associated with the probability of pancreatitis at autopsy, logistic regression was used to relate the probability of pancreatitis to various covariables. Factors that were statistically significant or suggestive of being associated with the finding of pancreatitis at autopsy from univariable models were used to build multivariable models. Only variables that were statistically significantly associated with or suggestive of being associated with the probability of pancreatitis at autopsy were included in the multivariable models. For the univariable and multivariable models, P values are two-sided and are derived from the Wald test. No adjustments for multiple comparisons were made in calculating the reported P values.


Results

seen, one with pancreatic pseudocysts on both ultrasound and CT, and one with an enlarged pancreas, peripancreatic edema and possible pseudocyst with hemorrhage by CT. In the week prior to death, 10 of the 51 (19.6%) patients with pancreatitis had abdominal imaging studies, all by ultrasound. One study did not image the pancreas adequately for analysis. Only one of nine evaluable patients had an abnormal imaging study in this time period, showing a mass in the head of the pancreas. All three patients with pancreatic abnormalities on imaging had severe pancreatitis. However, 15 patients with severe pancreatitis did not have abnormalities on imaging studies.

Prevalence of pancreatitis at autopsy Fifty-one of 184 (28%) patients whose pancreas was evaluable at autopsy had histologic evidence of pancreatitis. It was classified as severe in 22, moderate in four, and mild in 25 patients. Pancreatic infections were noted in five patients on tissue staining (two with bacteria, two with cytomegalovirus, one with aspergillus), but whether infection was the sole etiology of pancreatitis in these cases could not be determined. Pseudocyst formation was not documented in any patient at autopsy.

Pancreatitis as a cause of death

Clinical features of pancreatitis

Pancreatitis was felt to be a cause of death in five patients, all of whom had histologically severe pancreatitis. Pancreatitis was diagnosed or suspected ante-mortem in these patients, as serum amylase or lipase measurements within 2 weeks of death were abnormal in all cases. Imaging studies were also obtained in all, and were abnormal in three patients. Pseudocysts were documented by CT in two patients, and a possible pancreatic mass was seen by ultrasound in one patient.

Symptoms: Fifteen of 51 (29.4%) patients with autopsy-proven pancreatitis had abdominal pain documented in the time period 1 to 2 weeks prior to death. In the week prior to death, 18 of 51 (35.3%) patients had complained of pain. A higher proportion with moderate or severe pancreatitis than with mild pancreatitis had complained of abdominal pain. Laboratory measurements: Of the 51 patients with pancreatitis at autopsy, only 15 had at least one measurement of serum total amylase, pancreatic amylase, or lipase in the 4 weeks prior to death; only 11 had abnormal values (Table 1). Of 133 patients without evidence of pancreatitis at autopsy, 51 had at least one measurement of amylase or lipase in the 4 weeks prior to death; 13 had abnormal values (Table 1). A higher proportion of patients with severe pancreatitis than with mild or moderate pancreatitis had these measurements done (data not shown). All patients with severe pancreatitis had abnormal values. No patients with mild pancreatitis had values greater than twice the upper limit of normal.

Pancreatitis in autologous marrow recipients Thirty-one of the 184 patients included in this study had received an autologous marrow transplant. Of these patients, seven had breast cancer, 15 had multiple myeloma, and nine had non-Hodgkin’s lymphoma as their underlying diagnosis. Pancreatitis developed in six of these patients, and was classified as severe histologically in two, moderate in two, and mild in two patients. Pancreatitis was not felt to be the primary cause of death in any of these patients.

Imaging studies: In the time period 1 to 2 weeks prior to death, 24 of the 51 (47.0%) patients with pancreatitis had abdominal imaging studies, 13 by ultrasound, seven by CT, and four by both ultrasound and CT. Of these studies, 11 did not image the pancreas adequately for analysis. Only two of 13 evaluable patients had abnormalities of the pancreas Table 1

Risk factor analysis No pre-transplant demographic features predicted the development of pancreatitis (Table 2). No apparent difference in underlying diseases was found between patients

Laboratory measurements in patients with and without pancreatitis at autopsy

Patients with pancreatitis at autopsy (n = 51)

Patients without pancreatitis at autopsy (n = 133)

Interval prior to death (weeks)

No. patients with amylase, pancreatic amylase or lipase measurements

No. patients with abnormal measurements

2–4

3

2

1–2

7

6

,1

5

3

2–4 1–2 ,1

11 11 29

1 1 11

Range of abnormal measurements

amylase 230–1737 pancreatic amylase 133–1637 lipase 89 amylase 254–2143 pancreatic amylase 49–221 amylase 283–1380 pancreatic amylase 52–179 amylase 160 pancreatic amylase 114 amylase 443 amylase 93–334 pancreatic amylase 87–334 lipase 37–44

1083


1084

Table 2 autopsy

Univariable analyses of risk factors for acute pancreatitis at

Risk factor

Odds ratio

P value

95% CI

Age (years) (continuous variable)

1.01

0.444

(0.99, 1.03)

Gender Female Male

1.0 1.04

— 0.909

— (0.53, 2.03)

First transplant Second transplant

1.0 1.08

— 0.893

— (0.36, 3.23)

Disease status Relapse Remission

1.0 1.16

— 0.677

— (0.58, 2.35)

Patient CMV serostatus Seronegative Seropositive

1.0 1.08

— 0.827

— (0.55, 2.12)

Donor CMV serostatus Seronegative Seropositive

1.0 1.38

— 0.352

— (0.70, 2.73)

Autologous transplant Allogeneic transplant

1.0 0.89

— 0.792

— (0.38, 2.10)

GVHD prophylaxis None Any CSP

1.0 0.85

— 0.734

— (0.32, 2.22)

Other regimen

0.86

0.815

(0.24, 3.12)

Clinical grade of GVHD 0–I Autologous marrow recipient II–IV

1.0 3.22 3.15

— 0.062 0.027

— (0.94, 11.01) (1.14, 8.69)

Clinical grade of GVHD GVHD 0–II Autologous marrow recipient GVHD III–IV

1.0 2.50 2.91

— 0.090 0.011

— (0.87, 7.21) (1.28, 6.64)

Major infection at autopsy No Yes

1.0 2.00

— 0.60

— (0.97, 4.12)

GVHD at autopsy Absent Autologous marrow recipient Present

1.0 3.56 4.11

— 0.046 0.006

— (1.02, 12.39) (1.50, 11.26)

Liver GVHD at autopsy Absent Autologous marrow recipient Present

1.0 2.39 3.11

— 0.117 0.007

— (0.80, 7.09) (1.37, 7.08)

Liver disease at autopsy No Yes

1.0 2.12

— 0.192

— (0.69, 6.53)

1.007

0.007

(1.002, 1.012)

Days of survival after transplantation (continuous variable)

creatitis was seen in GVHD prophylaxis regimens. In this analysis, regimens were grouped into no prophylaxis, any cyclosporine-containing regimen, and other regimens (generally including methotrexate or corticosteroids). In patients with pancreatitis, there was a higher frequency of clinical grades 3 to 4 GVHD, any GVHD at autopsy, liver GVHD at autopsy, and major infection at autopsy. No obvious difference was found for the autopsy variables of gut GVHD or any liver disease (Table 2). All patients were found to have biliary sludge on microscopic analysis of gallbladder contents at autopsy. Of all patients autopsied, seven had gallstones, two had common bile duct sludge or stones, and three had cholecystitis at autopsy. Pancreatitis was noted in six of the seven patients with gallstones, one of the two with common bile duct sludge or stones, and all of the three patients with cholecystitis. Medication history was analyzed in two ways. The first method examined medications as indicator variables (whether or not a medication had been given regardless of the number of days over which it was administered). The second method included as a continuous variable the number of days on which a medication was received after adjusting for days survived. When examined as indicator variables in univariable analyses, the only medications that approached significance were fluconazole (P = 0.13) and omeprazole (P = 0.12). However, when analyzed as a continuous variable in univariable analyses, fluconazole was only marginally significant (P = 0.07), and omeprazole was not significant (P = 0.296). Data for the remaining medications are not shown. In a multivariable model, independent risk factors for development of pancreatitis were any GVHD at autopsy, major infection at autopsy, and increasing length of survival after transplantation (Table 3). When added to this model, the administration of omeprazole or fluconazole as indicator variables were statistically significant risk factors for development of pancreatitis (P = 0.031 and P = 0.013; respectively). However, when medications were analyzed as continuous variables, only fluconazole was retained as a marginally significant risk factor (P = 0.098) when examined in the model containing the factors listed in Table 3.

Table 3

Multivariable model of acute pancreatitis at autopsy

Risk factor

with and without pancreatitis (data not shown). The two groups were not statistically significantly different in the presence or dose of total body irradiation during conditioning (data not shown). Patients with no TBI in their regimen received chemotherapy only. There was not a statistically significant difference in incidence of pancreatitis among different chemotherapy regimens (data not shown). The mean time of survival from last transplant was statistically significantly longer in the pancreatitis cases. No apparent difference between patients with and without pan-

GVHD at autopsy None Autologous marrow recipient Any GVHD at autopsy Days of survival after transplantation (continuous variable) Major infection at autopsy No Yes

Odds ratio

P value

95% CI

1.0 3.88 3.66

— 0.036 0.015

— (1.09, 13.79) (1.29, 10.34)

1.006

0.017

(1.001, 1.011)

1.0 2.34

— 0.030

— (1.09, 5.05)


Discussion Marrow transplant patients have many potential factors that predispose them to pancreatitis, including a high prevalence of biliary sludge and treatment with medications such as corticosteroids that are associated with development of pancreatitis.4–6 In this series, we found an unexpectedly high prevalence of pancreatitis at autopsy. This prevalence may be higher than that found in the general population of patients following marrow transplant. In mild cases, pancreatitis was often subclinical, as shown by the relative infrequency of abdominal pain and laboratory or imaging studies. Severe cases more frequently had abdominal pain, abnormal laboratory values, or abnormal imaging tests. The major risk factor for pancreatitis was severe acute GVHD. Although severe GVHD was identified as a statistically significant risk factor for pancreatitis in this series, it may be the treatment of GVHD that predisposes patients to pancreatitis, rather than GVHD per se. For example, high-dose corticosteroids, often used to treat severe GVHD, may have contributed to the development of pancreatitis. Pancreatitis patients had a longer mean survival time, suggesting that prolonged, high-dose immunosuppressive therapy may be required to induce pancreatic toxicity. In fact, one study has shown high levels of cyclosporine can be found in the pancreas, adipose tissue, and liver after varying lengths of therapy.24 Other studies have found that cyclosporine may induce or worsen acute pancreatitis.10,11 In our study, however, neither cyclosporine nor corticosteroids were found to be statistically significant risk factors for development of pancreatitis. Widespread use of these medications in patients with and without pancreatitis may not allow their identification as risk factors in statistical analysis, and this does not completely exclude the possibility that their use in some way predisposes to the development of pancreatitis. However, the longer mean survival time in patients with pancreatitis could simply imply that such patients have a longer time period in which they could develop this complication. The prolonged use of immunosuppressive therapy also enables development of major infection, which may confound the association of pancreatitis and infection. Although major infection was seen in 75% of cases, pancreatic infection was seen in only 10% of those with pancreatitis. It is not clear whether infection was the primary cause of pancreatitis in these cases. Another finding which remains unexplained is the increased risk of pancreatitis in patients undergoing autologous marrow transplantation, who are not at risk for GVHD. As these patients do not receive prolonged immunosuppression, this cannot be implicated as a risk factor. We also could not explain this finding by differences in other variables such as conditioning regimens that might make recipients of autologous grafts different from recipients of allogeneic transplants. The histopathology of the pancreas in patients after bone marrow transplantation has been described.25 After high-dose chemotherapy, squamous metaplasia of the pancreatic ducts and dilatation of the pancreatic acini with oncocytic changes (eosinophilic cytoplasm and pyknotic nuclei) are seen most frequently histologically. Both squamous metaplasia and oncocytic changes were attributed to high-dose chemotherapy, as they were sometimes seen in patients receiving

chemotherapy for breast cancer. High-dose radiation can cause pancreatic toxicity, but this is usually in the form of vascular sclerosis and interstitial fibrosis.26,27 The doses required for these changes to develop far exceed the dose of TBI used in marrow transplantation. However, multidrug chemotherapy regimens may cause pancreatitis and the concomitant use of TBI and cytoreductive drugs may lower the threshold for TBI-related damage. The data, however, do not show a relation of pancreatitis to the conditioning regimen. Four cases of presumed pancreatic GVHD have been described in the literature.28 Findings included epithelial cell atypia and a mild lymphocytic infiltrate in the exocrine ductal system. Histologic changes in both these instances are distinct from those of acute pancreatitis,29 and no cases of acute pancreatitis were found in either the series of pancreatic GVHD or radiation-induced pancreatitis. Thus, the histologic changes seen in the current autopsy cohort cannot be explained by toxic effects of the conditioning regimen or by GVHD itself. However, we cannot rule out the possibility that more subtle immunologic alterations occur, causing subclinical pancreatic injury and predisposing to development of acute pancreatitis. Another possible confounding factor is the occurrence of ‘agonal’ pancreatitis, occurring in the immediate pre-mortem period, which could in part explain the high proportion of clinically unsuspected pancreatitis. Others have reported a high incidence of undiagnosed pancreatitis in autopsy series.30,31 However, this cannot easily be distinguished from non-‘agonal’ acute pancreatitis on histologic grounds. Gallbladder sludge was seen in all patients, with and without pancreatitis, at autopsy. Gallbladder sludge in the post-marrow transplant setting has been shown to be primarily composed of calcium bilirubinate.17 Sludge of this composition is often a result of gallbladder stasis, which frequently occurs in the post-transplant setting.2,3 In patients after solid organ transplant, pancreatitis has not been a significant complication. One series of patients after renal transplantation noted development of clinical acute pancreatitis in six of 228 (2.6%) who were also receiving cyclosporine and prednisone for immunosuppression.11 Steed et al32 described three out of 143 patients who died of pancreatitis after cardiac transplantation. The incidence of biliary sludge was not given in either report. In the nontransplant population, biliary sludge has been identified as an etiology in 74% of patients with previously classified ‘idiopathic’ pancreatitis.33 The presence of biliary sludge could be a factor in the development of acute pancreatitis in the marrow transplant population as well, but our data shows that all patients had gallbladder sludge at autopsy, irrespective of whether pancreatitis was present. The contribution of pancreatitis to the patients’ overall course is difficult to determine in an autopsy study. Pancreatitis that is mild and subclinical may have few clinical consequences in patients who are severely ill from other conditions. However, some patients in this study had severe pancreatitis histologically. It is well known that severe pancreatitis can lead to systemic complications, including respiratory distress, renal dysfunction, and multiorgan failure. Thus, unsuspected severe pancreatitis may have contributed substantially to the terminal course in some of these patients by causing or exacerbating multiorgan dysfunction. If patients with acute GVHD on prednisone therapy have a

1085


1086

deteriorating clinical course whose cause is not well defined, serum lipase or pancreatic amylase tests may allow early identification of pancreatitis and its complications. Risk factors such as biliary microcalculi and prednisone dose can be modified and complications such as pseudocysts treated. These therapies, however, will be of limited use in patients with severe acute GVHD that has not responded to either primary or secondary immunosuppressive therapy. Although the therapy for severe pancreatitis is generally supportive, interventions including ERCP or drainage of infected pancreatic necrosis have been shown to have a role in some patients. Thus, making a diagnosis pre-mortem may allow more aggessive therapy, and potentially alter the outcome in those with severe pancreatitis. In summary, pancreatitis was seen in 27% of autopsied marrow transplant patients. Many patients with pancreatitis did not have amylase or lipase levels measured or abdominal imaging studies performed, suggesting that the disease was subclinical. Risk factors for developing pancreatitis include increasing days of survival and any GVHD at autopsy. The high prevalence of biliary sludge may also contribute as an etiologic factor. Pancreatitis is one of many co-morbid events that occur in marrow transplant patients who receive immunosuppressive drugs for extended periods of time. Acknowledgements This work was supported by National Institutes of Health grants CA 1802, CA 47748, and CA 157704. Portions were published in abstract form (Gastroenterology 1995; 107: A1101) and presented in part at the annual meeting of the American Gastroenterological Association, May 1995, San Diego, California.

References 1 Werlin SL, Casper J, Antonsdon D, Calabro C. Pancreatitis associated with bone marrow transplantation in children. Bone Marrow Transplant 1992; 10: 65–69. 2 Frick MP, Snover DC, Feinberg SB et al. Sonography of the gallbladder in bone marrow transplant patients. Am J Gastroenterol 1984; 79: 122–127. 3 Teefey SA, Hollister MS, Lee SP et al. Gallbladder sludge formation after bone marrow transplant: sonographic observations. Abdom Imaging 1994; 19: 57–60. 4 Levine RA, McGuire RF. Corticosteroid-induced pancreatitis: a case report demonstrating recurrence with rechallenge. Am J Gastroenterol 1988; 83: 1161–1164. 5 Mallory A, Kern F. Drug-induced pancreatitis: a critical review. Gastroenterology 1980; 78: 813–820. 6 Carone FA, Liebow A. Acute pancreatitis lesions in patients treated with ACTH and adrenal corticoids. New Engl J Med 1957; 257: 690–697. 7 Alexander JA, Demetrius AJ, Gavaler JS et al. Pancreatitis following liver transplantation. Transplantation 1988; 45: 1062–1065. 8 Aziz S, Bergdahl L, Baldwin JC et al. Pancreatitis after cardiac and cardiopulmonary transplantation. Surgery 1985; 97: 653–660. 9 Uti T, Kimura T, Yamaguchi H et al. Acute pancreatitis induced by cyclosporin A under stimulation of the pancreas by cerulein. Pancreas 1993; 8: 693–699. 10 Echigo Y, Inoue K, Kogire M et al. Effects of cyclosporine and tacrolimus (FK506) on acute pancreatitis in mice. Arch Surg 1995; 130: 64–68.

11 Lorber MI, van Buren CT, Flechner SM et al. Hepatobiliary and pancreatic complications of cyclosporine therapy in 446 renal transplant recipients. Transplantation 1987; 43: 35–40. 12 Yoshimura N, Nakai I, Ohmori Y et al. Effect of cyclosporine on the endocrine and exocrine pancreas in kidney transplant recipients. Am J Kidney Dis 1988; 12: 11–17. 13 Niemann TH, Trigg ME, Winick N, Penick GD. Disseminated adenoviral infection presenting as acute pancreatitis. Hum Pathol 1993; 24: 1145–1148. 14 Schiller GJ, Nimer SD, Gajewski JL, Golde DW. Abdominal presentation of varicella-zoster infection in recipients of allogeneic bone marrow transplantation. Bone Marrow Transplant 1991; 7: 489–491. 15 Margreiter R, Schmid T, Dunser M et al. Cytomegalovirus (CMV) pancreatitis: a rare complication after pancreas transplantation. Transplant Proc 1991; 23: 1619–1622. 16 Parham DM. Post-transplantation pancreatitis associated with cytomegalovirus (report of a case). Hum Pathol 1981; 12: 663–665. 17 Ko CW, Murakami C, Sekijima et al. Chemical composition of gallbladder sludge in patients after marrow transplantation. Am J Gastroenterol 1996; 91: 1207–1210. 18 Peterson FB, Bearman SI. Preparative regimens and their toxicity. In: Forman SJ, Blume KG, Thomas ED (eds). Bone Marrow Transplantation Blackwell Scientific: Boston, 1994, pp 79–95. 19 Vogelsang G. Pharmacology and use of immunosuppressive agents after bone marrow transplantation. In: Forman SJ, Blume KG, Thomas ED (eds). Bone Marrow Transplantation. Blackwell Scientific: Boston, 1994; pp 114–123. 20 Burns WH. Herpes simplex virus. In: Forman SJ, Blume KG, Thomas ED (eds). Bone Marrow Transplantation. Blackwell Scientific: Boston, 1994, pp 404–411. 21 Zaia JA. Cytomegalovirus infection. In: Forman SJ, Blume KG, Thomas ED (eds): Bone Marrow Transplantation. Blackwell Scientific: Boston, 1994, pp 376–403. 22 Martin PJ, Schoch G, Fisher L et al. A retrospective analysis of therapy for acute graft-versus-host disease: initial treatment. Blood 1990; 76: 1464–1472. 23 Sullivan KM. Graft-versus-host disease. In: Forman SJ, Blume KG, Thomas ED (eds). Bone Marrow Transplantation. Blackwell Scientific: Boston, 1994, pp 339–362. 24 Ried M, Gibbons S, Kwok D et al. Cyclosporine levels in human tissues of patients treated for 1 week to 1 year. Transplant Proc 1983; 15: 2434–2443. 25 Washington K, Peters WS, Gottfried MR. Pathology of the pancreas in bone marrow transplant patients. Hum Pathol 1993; 24: 152–159. 26 Fajardo LF, Berthirong M. Radiation injury in surgical pathology. III. Salivary glands, pancreas, and skin. Am J Surg Pathol 1981; 5: 279–296. 27 Woodruff KH, Castro JR, Quivey JM et al. Postmortem exam of 22 pancreatic carcinoma patients treated with helium ion irradiation. Cancer 1984; 53: 420–425. 28 Foulis AK, Farquharson MA, Sale GE. The pancreas in acute graft-versus-host disease in man. Histopathol 1989; 14: 121–128. 29 Kloppel G, Maillet B. Pathology of acute and chronic pancreatitis. Pancreas 1993; 8: 659–670. 30 Wilson C, Imrie CW. Deaths from acute pancreatitis: why do we miss the diagnosis so frequently? Int J Pancreatol 1988; 3: 273–282. 31 Lanisch PG, Schirren CA, Kunze E. Undetected fatal acute pancreatitis: why is the disease so frequently overlooked? Am J Gastroenterol 1991; 86: 322–326. 32 Steed DL, Brown B, Reilly JJ et al. General surgical complications in heart and heart–lung transplantation. Surgery 1985; 98: 739–749. 33 Lee SP, Nicholls JF, Park HZ. Biliary sludge as a cause of acute pancreatitis. New Engl J Med 1992; 326: 589–593.


1087