Idiopathic Tumefactive Chronic Pancreatitis: Clinical Profile,Histology, and Natural History After Resection

DHIRAJ YADAV,* KENJI NOTAHARA,‡ THOMAS C. SMYRK,‡ JONATHAN E. CLAIN,*RANDALL K. PEARSON,* MICHAEL B. FARNELL,§ and SURESH T. CHARI**Division of Gastroenterology and Hepatology, ‡Division of Anatomic Pathology, §Division of Gastroenterologic and General Surgery,Mayo Clinic, Rochester, Minnesota

Background & Aims: Little is known about subjects withidiopathic tumefactive chronic pancreatitis (TCP), thatis, chronic pancreatitis whose clinical presentation, usu-ally with a mass or obstructive jaundice, suggests can-cer. Methods: We independently reviewed clinical dataand histology of 45 TCP (27 idiopathic, 18 alcohol in-duced) resected at Mayo Clinic (January 1985–March2001). Follow-up data were obtained from medicalrecords and mailed questionnaires. Results: Comparedwith alcoholic subjects, idiopathic TCP patients wereolder (58 � 2 vs. 48 � 3 yr, P < 0.001), had shortersymptom duration (median 3 vs. 24 wk, P < 0.001),were more likely to have no or mild abdominal pain(70% vs. 17%, P � 0.001), and were more often jaun-diced (67% vs. 33%, P � 0.02). Three distinct histologicpatterns were identified in TCP. Typical CP (n � 19)showed lobular atrophy, fat necrosis, and ductalchanges (dilatation, protein plugs, and stones). Lym-phoplasmacytic sclerosing pancreatitis (LPSP) (n � 14)was characterized by periductal lymphoplasmacytic in-filtration, obliterative phlebitis, and cholangitis withedema. Idiopathic duct-centric CP (IDCP) (n � 12) hadneutrophil-predominant lobular inflammation, withoutphlebitis. On correlation of clinical and histologic diag-nosis, 17 of 18 (94%) patients with alcohol-induced TCPhad typical CP, and 25 of 27 (93%) with idiopathic TCPhad LPSP or IDCP. LPSP and IDCP were indistinguish-able clinically except for higher incidence of jaundice inLPSP (93% vs. 42%, P � 0.005). In idiopathic TCP norecurrence of symptoms was observed after resection(median follow-up 49 mo). Conclusions: Idiopathic TCPis clinically and histologically distinct from alcohol-in-duced TCP. It is unclear whether LPSP and IDCP, 2 uniquepatterns of histologic injury observed in idiopathic TCP, arepart of the spectrum of the same disease or represent 2 ormore different entities. Resection of mass prevents recur-rence of symptoms in idiopathic TCP.

Patients presenting with a pancreatic mass or obstruc-tive jaundice often undergo pancreatic resection be-

cause of suspicion of cancer. In most patients in whompancreatic cancer is suspected, the combination of clin-

ical information, imaging studies, and, if necessary, fine-needle aspiration of the mass, helps to establish thediagnosis of pancreatic cancer preoperatively. In approx-imately 5% of resections for suspected pancreatic cancer,a histologic diagnosis of chronic pancreatitis is madepostoperatively.1,2

It is generally believed that the different etiologies ofchronic pancreatitis produce similar pathologic find-ings.3 This is probably because most patients studiedhave had alcohol-induced chronic pancreatitis or haveend-stage chronic pancreatitis, in which extensive fibro-sis replaces the glandular tissue and obscures potentialetiologic clues. Recently, several reports have described adistinct histology characterized by prominent lympho-plasmacytic infiltration in a subset of patients presentingwith a pancreatic mass.4–12 Several names have been usedto describe this entity (e.g., autoimmune pancreatitis,5,6

chronic inflammatory sclerosis of the pancreas,4 lym-phoplasmacytic sclerosing pancreatitis,11 and non–alco-hol-induced duct destructive chronic pancreatitis).7

We identified 45 patients from the Mayo Clinic Roch-ester Surgical Pathology database who underwent pancre-atic resection for suspicion of pancreatic cancer owing to amass and/or obstructive jaundice and on histology werediagnosed to have chronic pancreatitis. For this study, wedefine such patients as having tumefactive chronic pancre-atitis (TCP). The aims of our study were to compare theclinical and histologic features and long-term follow-up ofresected tumefactive chronic pancreatitis.

MethodsThe Mayo Foundation Institutional Review Board ap-

proved the study and the contact of consenting patients, either bytelephone or mail. In accordance with a Minnesota state statute,13

Abbreviations used in this paper: CP, chronic pancreatitis; TCP,tumefactive chronic pancreatitis; LPSP, lymphoplasmacytic sclerosingpancreatitis; IDCP, idiopathic duct-centric pancreatitis.

© 2003 by the American Gastroenterological Association1542-3565/03/$30.00

doi:10.1053/jcgh.2003.50016

CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2003;1:129–135

2 individuals who declined to authorize the use of their medicalrecords in research were excluded from further review.

Patient Acquisition

From the Surgical Pathology database at the MayoClinic, Rochester, we retrospectively identified 245 patientswho underwent pancreatic resection between January 1985and March 2001 and had a final pathologic diagnosis ofchronic pancreatitis. Forty-five of the 245 patients underwentpancreatic resection for suspicion of pancreatic cancer owing toa mass and/or obstructive jaundice and were defined as havingTCP. We independently reviewed the medical records andpathologic sections of these 45 patients and separate clinicaland pathologic diagnoses were established for each patient.

Clinical Data

The medical records of these 45 patients were reviewedto collect data on demographics, clinical presentation, labora-tory evaluation, imaging studies, and intraoperative findings.We also noted personal and family history of autoimmunediseases. Pain was evaluated as documented in the patient’smedical record or the patient’s response to a questionnaire onfollow-up and graded as none, mild (abdominal discomfort orpain not requiring analgesics or not disabling), moderate (con-trolled with nonnarcotic analgesics), or severe (excruciating,disabling pain requiring narcotic analgesics). We noted theduration of pain and other symptoms in all patients. We notedthe findings on imaging studies (ultrasound of the abdomen,computed tomography scan, endoscopic retrograde cholangio-pancreatography, or endoscopic ultrasound). Imaging studiesperformed at outside institutions were reviewed at the MayoClinic at the time of their initial evaluation.

Clinical Classification of TumefactiveChronic Pancreatitis

Clinically, patients were classified as alcoholic if therewas history of alcohol use of 50 g/day or greater for 5 years ormore, and as idiopathic if no known etiology for chronicpancreatitis was identified after a thorough evaluation. Twopatients with inflammatory bowel disease, one each withCrohn’s disease and ulcerative colitis, were classified as idio-pathic TCP.

Follow-up

Follow-up information was obtained from medicalrecords and patient correspondence. Additionally, 34 of 40(85%) patients who had surgery performed before 1998 com-pleted a structured questionnaire sent out in 1998. The ques-tions related to symptoms of pancreatic disease (abdominalpain, use of narcotics, jaundice, hospitalizations for pancreas-related problems or other conditions, exocrine and endocrineinsufficiency, and any other medical problems after surgery).

Histologic Examination

Two independent pathologists (T.C.S. and K.N.), un-aware of the clinical data, reviewed the slides of all patients.

The resected materials were fixed in formalin and embedded inparaffin. H&E stain was performed in all patients. For arepresentative block of the cases with periductal infiltration,elastic (von Giesen) stain and immunohistochemical studieswere performed.

Analyses

We used the Student t test for all continuous variablesand �2 analysis for categoric variables. The values are repre-sented as mean � standard error of mean and a P value of lessthan 0.05 was considered significant.

ResultsClinical Presentation

Of 45 TCP patients who underwent pancreaticresection, 27 were classified as idiopathic and 18 asalcohol-induced TCP. Two patients with inflammatorybowel disease (one each with Crohn’s disease and ulcer-ative colitis) were included in the idiopathic TCP group.None of the TCP patients had a personal or familyhistory of autoimmune disorder. Compared with alcohol-induced TCP, idiopathic TCP patients were older (58 �2 vs. 48 � 3 yr; P � 0.001), had a shorter duration ofsymptoms (median 3 vs. 24 wk, P � 0.001), were morelikely to have no or mild abdominal pain (70% vs. 17%,P � 0.001), and more often were jaundiced (67% vs.33%, P � 0.02) (Table 1).

A pancreatic mass was seen in 40 of the 45 cases (89%)either on imaging studies (35 of 40 cases) or at surgery(5 of 40 cases) (Table 1). The mass was located in thepancreatic head in all but one patient who had a massinvolving the body of the pancreas. More patients withidiopathic TCP (9 of 27 or 33%) were noted to havediffuse enlargement of the pancreas (6 on imaging stud-ies and 3 at the time of surgery) compared with alcohol-induced TCP (1 of 18 or 5%; P � 0.02; Table 1).

The most frequent findings on endoscopic retrogradecholangiopancreatography, performed in 28 (62%) pa-tients, were strictures in the bile duct, pancreatic duct, orboth ducts (Table 1). Pancreaticoduodenectomy was per-formed in 43 of 45 cases, one patient had distal pancre-atectomy and one had total pancreatectomy (Table 1).None of the patients were treated with steroids pre- orpostoperatively.

Follow-up

Three patients died during follow-up evaluation(2 in the idiopathic and 1 in the alcohol-induced TCPgroup) owing to unrelated causes (Table 2). In onepatient with idiopathic TCP, pain persisted even aftersurgery and worsened over the years despite several celiacplexus blocks and octreotide injections. Computed to-

130 YADAV ET AL. CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 1, No. 2

Table 1. Demographics, Clinical Features, and Imaging Studies in Tumefactive Chronic Pancreatitis (TCP)

VariablesIdiopathic TCP

(n � 27)Alcoholic TCP

(n � 18) P value

DemographicsAge (yr) 58 � 2 48 � 3 0.009Sex (% males) 63 72 ns

Clinical featuresDuration of symptoms in weeks (median) 3 24 0.005Pain

% Yes 44 95 —Pain � 8 weeks 15 55 0.0003

Pain severity (% None or mild) 70 17 0.02Jaundice (% Yes) 67 33 —Proven/Suspected CP (% Yes) None 39 0.02Diabetes/Abnormal glucose tolerance (% Yes) 33 11 nsWeight loss (% Yes) 52 50 ns

Findings in imaging studies (%)Mass lesion (Total) 85 95 ns

Detected on imaging studies 83 94 nsDetected only at surgery 17 6 ns

Diffuse enlargement of pancreas (on imaging or at surgery) 33 5 0.03ERCP

Performed n (% Yes) 16 (59) 11 (66)Findings (N)

Normal CBD/PD 1 1 nsDistal CBD stricture 8 2 nsPD stricture 3 3 nsDouble duct sign 4 5 ns

Type of surgery (N) nsPD:DP:TP 25:1:1 18:0:0

CP, chronic pancreatitis; ERCP, endoscopic retrograde cholangiopancreatography; PD, pancreatico-duodenectomy; DP, distal pancreatectomy;TP, total pancreatectomy; ns, not significant.

Table 2. Follow-up Data in TCP Patients

Idiopathic TCP (n � 27) Alcoholic TCP (n � 18)

Follow-up available n (% Yes) 22 (81) 14 (78)Follow-up (mo)

Mean � SEM 63 � 10 64 � 15Median 49 36Follow-up � 3 years n (%) 19 (70) 50Follow-up � 5 years n (%) 12 (44) 44

Abdominal pain (n)Present before surgery 12 (44) 17 (94)Pain continued after surgery 1 (4) 3 (17)Need for narcotics 1 (4) 2 (11)

Diabetes (n)Diabetes/glucose intolerance before

surgery9 (33) 2 (11)

Developed after surgery 4 (15) 6 (33)No diabetes after surgery 10 (37) 6 (33)Unknown 4 (15) 4 (22)

Recurrence of pancreatic mass None NoneExocrine insufficiency after surgery n (%) 10 (37) 4 (22)Medical problems on follow-up (n) Lymphoma—1

Sclerosing cholangitisa—2Thrombocytopenic purpura—1End-stage renal disease—1Peptic ulcer disease—3

Stricture anastomotic site—1Recurrent GI bleeding

(anastomotic ulcer)—11Peptic ulcer disease—2

aPancreatic histology: LPSP in one and IDCP in one.

March 2003 TUMEFACTIVE CHRONIC PANCREATITIS 131

mography scans on follow-up only showed an atrophic-appearing remnant pancreas. In the alcohol-induced TCPgroup, pain relief was seen in 14 of 17 (82%) patients.Three patients continued to have pain and were takingnarcotics, although at reduced doses. Recurrence of pan-creatic mass was not seen in any of the patients over thefollow-up period.

Histopathologic Examination andClinicopathologic Correlation

Pathologically, 3 distinct histologic patterns wereobserved (Table 3): (1) lymphoplasmacytic sclerosingpancreatitis (LPSP) (Figure 1) in 14, (2) idiopathic duct-centric chronic pancreatitis (IDCP) (Figure 2) in 12, and(3) typical chronic pancreatitis (alcohol induced) in 19.

Of 27 patients with idiopathic TCP, 25 (93%) hadeither LPSP or IDCP on histology, and 17 of 18 (94%)patients with alcohol-induced TCP had typical CP onhistology. Both patients with idiopathic TCP who hadfeatures of typical CP had chronic and constant pain.

In 42 of 45 (93%) patients, idiopathic TCP could bedifferentiated easily from alcohol-induced TCP. None of

the features of alcohol-induced pancreatitis were seen inLPSP and IDCP. Clinically, the only difference seenbetween patients with LPSP and IDCP was the incidenceof jaundice. Although 13 of 14 (93%) patients withLPSP had jaundice, it was seen only in 5 of 12 (42%)cases of IDCP (P � 0.005).

DiscussionIn this study we show that idiopathic TCP is

clinically and histologically distinct from alcohol-in-duced TCP. The histologic findings of lymphoplasma-cytic infiltration seen in idiopathic TCP resemble thosedescribed previously by a variety of different namesincluding autoimmune and sclerosing pancreatitis.4–12

In idiopathic TCP, pancreatic resection prevented recur-rence of symptoms.

To our knowledge, the term tumefactive chronic pancre-atitis has not been used in the literature before. Inflam-matory pseudotumor is a descriptive term that alreadyhas been used for different lesions including the inflam-matory myofibroblastic tumor.14 We chose the term

Table 3. Comparison of Histology of LPSP (Figure 1), IDCP (Figure 2), and Typical CP

LPSP IDCP Typical CP

General description Ill-defined fibrosing process involvespancreatic ducts, veins, andcommon bile duct. Prominentinflammatory infiltrate in fibroustissue easily seen on low power.

Patchy inflammatory infiltrate involvesmainly lobules and pancreaticducts and is less marked in fibrousand adipose tissue.

Destruction of acinartissue andreplacement withextensive fibrosis.

Infiltrate Predominantly lymphoplasmacyticwith few eosinophils and rarepolymorphonuclear cells.

Predominantly neutrophilic.Polymorphonuclear cells formsignificant part of inflammatoryinfiltrate with occasionalmicroabscesses.

Lymphocytes, plasmacells, andmacrophages.

Pancreatic ducts Inflammation surrounds pancreaticduct but lumen of the ducts ispatent. Necrosis of pancreaticductal epithelium rare.

Inflammatory infiltrate involves entirewall of pancreatic ducts withepithelial destruction commonlyseen (sometimes with obstructionof lumen).

Pancreatic ductsenlarged anddistorted with fibroticwalls.

Intraductal proteinplugs and stones

No No Frequent

Parenchyma Lymphoplasmacytic infiltrationinvolves and replaces exocrinepancreas.

Patchy inflammatory infiltrate mainlyinvolves lobules.

Lobular atrophy withfibrosis and minimalmononuclearinfiltrate.

Vein Obliterative phlebitis (organizedobstruction of veins inassociation with denselymphoplasmacytic infiltration) inand around pancreascharacteristic (seen in all cases).

Obliterative phlebitis rarely seen. Obliterative phlebitisnot seen.

Artery Arterial involvement rarely seen. Endarteritis obliterans seen in onecase.

No arterial involvement.

Pseudocysts No No YesPeripancreatic fat Fibroinflammatory process extends

to peripancreatic region.Inflammation limited to the pancreas. Peripancreatic fat

necrosis frequent.

LPSP, lymphoplasmacytic chronic pancreatitis; IDCP, idiopathic duct-centric chronic pancreatitis; CP, chronic pancreatitis.

132 YADAV ET AL. CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 1, No. 2

tumefactive because it describes the clinical presentation ofthe patients in this study without committing to etiol-ogy. The diversity in the histology in these patients alsomeant that terms such as sclerosing pancreatitis, chronicinflammatory pancreatic sclerosis, or non–alcohol-induced ductdestructive pancreatitis could not be used to describe allpatients.

A number of previous reports have anecdotally and insmall case series described patients presenting with apancreatic mass who have a unique form of chronicpancreatitis. Presence of increased gamma globulins anda nonspecific increase in various antibody titers and aresponse to steroids has prompted the use of the termautoimmune pancreatitis.5,6,15–20 Nomenclatures based onhistology include nonsclerosing pancreatitis, non–alcohol-induced duct destructive pancreatitis, and chronic inflammatory

sclerosis of the pancreas.4,7,11 Though the terms autoimmuneand sclerosing pancreatitis have been used interchange-ably,18 in the absence of specific markers it is unclear ifall these investigators are describing the same disease.

Do the idiopathic TCP patients described in our studyhave the same disease as that described by the Japanese assclerosing or autoimmune pancreatitis? This question isdifficult to answer with certainty because reports ofsclerosing or autoimmune pancreatitis have used sero-logic criteria and response to steroids to make the diag-nosis, and histology generally is not available.18,21,22 Onthe other hand, all our patients are histologically wellcharacterized, but because of lack of clinical suspicion ofautoimmune pancreatitis and absence of associated auto-immune disorders, none of the patients reported in ourseries had serologic parameters measured. Also, at pre-sentation, none of our patients had disorders known to beassociated with autoimmune pancreatitis such asSjogren’s syndrome or primary sclerosing pancreatitis.However, the histology of idiopathic TCP described inour study resembled that described in sclerosing pancre-atitis,11,23 suggesting that our patients probably havesclerosing pancreatitis.

In our study, subjects with alcohol-related pancreatitiswere differentiated clinically from those with idiopathicTCP not only by their history of alcohol abuse but alsoby their age, duration of symptoms, and severity of pain.Although the presence of cancer could not be excludedconclusively preoperatively, the predominant indicationfor surgery in patients with alcohol-induced TCP wasintractable abdominal pain.

On the other hand, idiopathic TCP and pancreaticductal cancer have very similar demographics and clinicalpresentation. In our study, the older age at presentation,

Figure 1. Lymphoplasmacytic sclerosing pancreatitis. (A) Fibrosis and mononuclear cell infiltration in lobule and beneath epithelium which isintact. (B) Obliterative phlebitis showing dense inflammation of the vein. The artery is uninvolved.

Figure 2. Idiopathic duct-centric chronic pancreatitis. Low-power viewof an interlobular duct showing dense periductal and intraepithelialinflammatory infiltrate with epithelial erosion.

March 2003 TUMEFACTIVE CHRONIC PANCREATITIS 133

short duration of symptoms, and presence of obstructivejaundice led to a preoperative diagnosis of pancreaticcancer in all patients with idiopathic TCP. Because thepancreatic mass appeared resectable and the concern forpresence of cancer was high, all patients underwent pan-creatic resection.

The histologic findings described in this study are notentirely new. The histologic findings seen in our patientswith alcohol-related pancreatitis3 and the entities ofLPSP and IDCP all have been reported previously.7,11,23

In our study, the histology of alcohol-related pancreatitishad distinct features that readily allowed differentiationfrom idiopathic TCP. These included lobular atrophy, fatnecrosis, and ductal changes of dilatation, protein plugs,and intraductal stones, features absent in idiopathic TCP.Two distinct histologic patterns were seen among pa-tients with idiopathic TCP: LPSP and IDCP. Clinically,LPSP and IDCP were indistinguishable except for ahigher prevalence of jaundice in LPSP. Histologically,however, LPSP had diffuse fibrosis with lymphoplasma-cytic infiltration and obliterative phlebitis, whereasIDCP had neutrophil-predominant lobular inflammationand a duct destructive infiltrate, without obliterativephlebitis. Several of the features of IDCP seen in ourseries were similar to those described by Ectors et al.7

Obliterative phlebitis was described only in 1 of their 10patients, and that patient had Sjogren’s syndrome. The 2patients in our study, who had inflammatory boweldisease, had IDCP. It is unclear whether LPSP and IDCPare part of the spectrum of the same disease or represent2 or more different entities. None of the features oftypical chronic pancreatitis (ductal dilatation, intraduc-tal/parenchymal stones, pseudocysts, and so forth) wereseen in either LPSP or IDCP.

For this study we only examined patients who pre-sented with suspicion of cancer (i.e., with mass or ob-structive jaundice). Although we have not performed asystematic survey of all resected chronic pancreatitis,outside this study we have not observed LPSP or IDCP inpatients who clinically had typical chronic pancreatitis orhad tumor-induced obstructive pancreatitis. However,one patient in this study with alcohol-induced TCP didhave LPSP on histology. The specificity of the differenthistologic patterns needs further study.

What is the best course of treatment for a patient withidiopathic TCP? Reports from Japan suggest that ifdiagnosed prospectively, this condition is amenable tosteroid therapy.6,24,25 Because the ability to diagnoseLPSP or IDCP on pancreatic biopsy specimens is limitedowing to small sample size, the crucial question we arefaced with is, can the diagnosis be made conclusively

without histology? Japanese researchers suggest that pa-tients who meet serologic and imaging criteria can betreated with steroids without histologic confirma-tion.24,25 Although we were unable to confirm this, wedo show that surgical resection prevents recurrence ofpancreas-related symptoms on long-term follow-up. Inaddition to the surgically treated patients, we have man-aged successfully at least 2 patients who met criteria forsclerosing pancreatitis with observation alone. At least inthese 2 patients the disease had a self-limited course andhas not recurred on follow-up.

We recommend, based on review of literature andlimited personal experience, that a trial of steroid therapywould be reasonable in patients with typical features ofsclerosing pancreatitis26 such as diffusely enlarged sau-sage-shaped pancreas, elevated immunoglobulin G4 and� globulins, and endoscopic retrograde cholangiopancre-atography showing diffusely irregular and attenuatedpancreatic duct. In patients without typical features ofsclerosing pancreatitis, surgical resection would be pref-erable because it will relieve symptoms while avoidingthe risk of misdiagnosing resectable pancreatic cancer.

Recurrences of pancreatitis have been reported aftermedical treatment of autoimmune pancreatitis.18 Thereis one report of recurrence after surgical resection ofsclerosing pancreatitis.27 In our study none of the pa-tients had recurrence of pancreatitis after resection aftera median follow-up of 49 months. The fact that many ofour patients were diabetic or had steatorrhea postopera-tively suggests that the disease is diffuse even thoughsymptoms may resolve with partial pancreatic resection.A patient of ours treated medically has residual milddiabetes and steatorrhea suggesting burn out of thepancreas even with conservative management. It is un-clear if steroid therapy will protect the remaining pan-creas from destruction after resection.

In summary, idiopathic TCP is clinically and histo-logically different from alcohol-induced TCP. LPSP andIDCP are unique patterns of histologic injury observed inidiopathic TCP. It is unclear whether LPSP and IDCPare part of the spectrum of the same disease or represent2 or more different entities. The etiology of idiopathicTCP is unclear, but resection of mass prevents recurrenceof symptoms.

References1. Smith CD, Behrns KE, van Heerden JA, Sarr MG. Radical pancre-

atoduodenectomy for misdiagnosed pancreatic mass. Br J Surg1994;81:585–589.

2. van Gulik TM, Reeders JW, Bosma A, Moojen TM, Smits NJ,Allema JH, Rauws EA, Offerhaus GJ, Obertop H, Gouma DJ.Incidence and clinical findings of benign, inflammatory disease in

134 YADAV ET AL. CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 1, No. 2

patients resected for presumed pancreatic head cancer. Gastroi-ntest Endosc 1997;46:417– 423.

3. Forsmark C. Chronic pancreatitis. In: Felman M, Fiedman LS,Sleisenger MH, eds. Sleisenger and Fordtran’s gastrointestinaland liver disease. Volume 1. 7th ed. Philadelphia: Saunders,2002:943–969.

4. Sarles HSJ, Muratore R, Guien C. Chronic inflammatory sclerosisof the pancreas—an autonomous pancreatic disease? Am J DigDis 1961;6:688–698.

5. Yoshida K, Toki F, Takeuchi T, Watanabe S, Shiratori K, et al.Chronic pancreatitis caused by an autoimmune abnormality. Pro-posal of the concept of autoimmune pancreatitis. Dig Dis Sci1995;40:1561–1568.

6. Ito T, Nakano I, Koyanagi S, Miyahara T, Migita Y, et al. Autoim-mune pancreatitis as a new clinical entity. Three cases of auto-immune pancreatitis with effective steroid therapy. Dig Dis Sci1997;42:1458–1468.

7. Ectors N, Maillet B, Aerts R, Geboes K, Donner A, et al. Non-alcoholic duct destructive chronic pancreatitis. Gut 1997;41:263–268.

8. Marrano D, Gullo L, Casadei R, Santini D, Leone O, et al. Anunusual case of chronic pancreatitis of possible immune origin.Pancreas 1996;12:202–204.

9. Sood S, Fossard DP, Shorrock K. Chronic sclerosing pancreatitisin Sjogren’s syndrome: a case report. Pancreas 1995;10:419–421.

10. Scully KA, Li SC, Hebert JC, Trainer TD. The characteristic ap-pearance of non-alcoholic duct destructive chronic pancreatitis: areport of 2 cases. Arch Pathol Lab Med 2000;124:1535–1538.

11. Kawaguchi K, Koike M, Tsuruta K, Okamoto A, Tabata I, et al.Lymphoplasmacytic sclerosing pancreatitis with cholangitis: avariant of primary sclerosing cholangitis extensively involving pan-creas. Hum Pathol 1991;22:387–395.

12. Levey JM, Mathai J. Diffuse pancreatic fibrosis: an uncommonfeature of multifocal idiopathic fibrosclerosis. Am J Gastroenterol1998;93:640–642.

13. Melton L 3rd. The threat to medical-records research. N EnglJ Med 1997;337:1466–1470.

14. Dehner LP, Coffin CM. Idiopathic fibrosclerotic disorders andother inflammatory pseudotumors. Semin Diagn Pathol 1998;15:161–173.

15. Nakano STI, Kitamura K, Watahiki H, Iinuma Y, Takenaka M.

Vanishing tumor of the abdomen in patient with Sjogren’s syn-drome. Am J Dig Dis 1978;23:75s–79s.

16. Horiuchi A, Kaneko T, Yamamura N, Nagata A, Nakamura T, et al.Autoimmune chronic pancreatitis simulating pancreatic lym-phoma. Am J Gastroenterol 1996;91:2607–2609.

17. Kino-Ohsaki J, Nishimori I, Morita M, Okazaki K, Yamamoto Y, etal. Serum antibodies to carbonic anhydrase I and II in patientswith idiopathic chronic pancreatitis and Sjogren’s syndrome. Gas-troenterology 1996;110:1579–1586.

18. Hamano H, Kawa S, Horiuchi A, Unno H, Furuya N, et al. Highserum IgG4 concentrations in patients with sclerosing pancreati-tis. N Engl J Med 2001;344:732–738.

19. Erkelens GW, Vleggaar FP, Lesterhuis W, van Buuren HR, van derWerf SD. Sclerosing pancreato-cholangitis responsive to steroidtherapy. Lancet 1999;354:43–44.

20. Taniguchi T, Seko S, Azuma K, Tamegai M, Nishida O, et al.Autoimmune pancreatitis detected as a mass in the tail of thepancreas. J Gastroenterol Hepatol 2000;15:461–464.

21. Procacci C, Carbognin G, Biasiutti C, Frulloni L, Bicego E, et al.Autoimmune pancreatitis: possibilities of CT characterization.Pancreatology 2001;1:246–253.

22. Horiuchi A, Kawa S, Hamano H, Ochi Y, Kiyosawa K. Sclerosingpancreato-cholangitis responsive to corticosteroid therapy: re-port of 2 case reports and review. Gastrointest Endosc 2001;53:518–522.

23. Furukawa N, Muranaka T, Yasumori K, Matsubayashi R, Ha-yashida K, et al. Autoimmune pancreatitis: radiologic findings inthree histologically proven cases. J Comput Assist Tomogr 1998;22:880–883.

24. Horiuchi A, Kawa S, Hamano H, Hayama M, Ota H, et al. ERCPfeatures in 27 patients with autoimmune pancreatitis. Gastroin-test Endosc 2002;55:494–499.

25. Okazaki K, Uchida K, Chiba T. Recent concept of autoimmune-related pancreatitis. J Gastroenterol 2001;36:293–302.

26. Okazaki K, Chiba T. Autoimmune related pancreatitis. Gut 2002;51:1–4.

27. Motoo Y, Minamoto T, Watanabe H, Sakai J, Okai T, et al.Sclerosing pancreatitis showing rapidly progressive changes withrecurrent mass formation. Int J Pancreatol 1997;21:85–90.

Address requests for reprints to: Suresh T. Chari, M.D., 200 FirstStreet SW, Mayo Clinic, Rochester, Minnesota 55905. e-mail:chari.suresh@mayo.edu; fax: (507) 284-5486.

March 2003 TUMEFACTIVE CHRONIC PANCREATITIS 135