1. CLINICAL REVIEW
Diagnostic Criteria for Autoimmune Pancreatitis
Terumi Kamisawa, MD, PhD
In North America, about 2.5% of pancreatoduode-
Abstract: Autoimmune pancreatitis (AIP), a particular type of nectomies are performed in AIP cases that have been
pancreatitis, is thought to have an autoimmune etiology; it is incorrectly diagnosed as pancreatic cancer2; between 21%
recognized as a distinct entity worldwide. AIP has many and 23% of pancreatoduodenectomies performed for
characteristic clinical, serologic, morphologic, and histopatho- benign conditions are done for AIP.2,3 As there is
logic features. In the absence of a diagnostic serologic marker currently no diagnostic serologic marker for AIP, AIP
for AIP, AIP should be diagnosed on the basis of combination should be diagnosed on the basis of presence of a
of characteristic findings. AIP responds dramatically to steroid combination of abnormalities unique to AIP. In 2002, the
therapy; thus, accurate diagnosis of AIP can avoid unnecessary Japan Pancreas Society established the ‘‘Diagnostic
laparotomy or resection. It is important not to misdiagnose Criteria for Autoimmune Pancreatitis,’’4,5 which were
pancreatic cancer as AIP, and not to misdiagnose AIP as revised in 2006.6 In 2006, 2 new sets of diagnostic criteria
pancreatic cancer. Currently, 3 sets of major diagnostic criteria for AIP were proposed, 1 in Korea7 and 1 in the United
for AIP have been proposed in Japan, Korea, and the United States.8 Thus, at present, there are 3 major sets of
States. The Japanese criteria are based on the minimum diagnostic criteria for AIP. This review discusses and
consensus features of AIP and aim to avoid misdiagnosis of compares the potential limitations of these 3 sets of
malignancy. When response to steroid therapy is added to the diagnostic criteria for AIP on the basis of our recent
criteria, the diagnostic sensitivity is increased. However, the use treatment of AIP cases and suggests an improved set of
of a steroid trial in cases where differentiation from malignancy diagnostic criteria.
is an issue may result in delaying pancreatic cancer surgery,
which could lead to cancer progression in several cases. Thus,
given that AIP is an IgG4-related systemic disease, an additional JAPANESE DIAGNOSTIC CRITERIA
criterion can be recommended to the Japanese diagnostic
During the past decade, many AIP cases have been
criteria: IgG4-immunostaining of biopsied extrapancreatic
reported in Japan. These cases were found to have many
lesions such as the major duodenal papilla, the bile duct, or
common clinicopathologic features. On the basis of these
the minor salivary gland. It is also time for an international
data, the Japan Pancreas Society published the ‘‘Diag-
consensus on AIP.
nostic Criteria for Autoimmune Pancreatitis, 2002,’’
Key Words: autoimmune pancreatitis, diagnostic criteria, steroid which included 3 items: (1) radiologic imaging showing
therapy, IgG4 diffuse enlargement of the pancreas and diffuse irregular
narrowing of the main pancreatic duct (more than one
(J Clin Gastroenterol 2008;42:404–407) third the length of the entire pancreas); (2) laboratory
data demonstrating abnormally elevated levels of serum
gammaglobulin or IgG, or the presence of autoantibo-
A utoimmune pancreatitis (AIP) is a particular type of
pancreatitis that is thought to have an autoimmune
etiology. Since Yoshida et al1 proposed AIP as a
dies; and (3) histologic examination of the pancreas
showing lymphoplasmacytic infiltration and fibrosis.4,5
The diagnosis of AIP is made when either all 3 criteria are
diagnostic entity in 1995, many cases of AIP have been present or criterion 1 together with either criterion 2 or
reported in Western countries, and also in Japan. During criterion 3 is present. The presence of the imaging
the past 10 years, a number of new clinicopathologic criterion is essential to make the diagnosis of AIP. These
aspects of AIP have been clarified, and AIP is now criteria are based on the minimum consensus features of
considered to be a discrete entity worldwide. It is AIP to minimize the risk of misdiagnosing pancreatic
important not to misdiagnose AIP as pancreatic cancer. cancer.
With the accumulation of more AIP cases, it has
From the Department of Internal Medicine, Tokyo Metropolitan
become clear that, in several cases strongly suspected of
Komagome Hospital, Tokyo, Japan. having AIP, the degree of narrowing of the main
The author declares no conflict of interest. pancreatic duct is less than one third of the entire
Supported by Research for Intractable Disease of the Pancreas, Ministry pancreas. Furthermore, serum IgG4 levels are rather
of Health, Labour and Welfare of Japan. significantly and specifically elevated in AIP patients.
Reprints: Terumi Kamisawa, MD, PhD, Department of Internal Medicine,
Tokyo Metropolitan Komagome Hospital, 3-18-22 Honkomagome, Thus, the 2002 criteria were revised in 2006.6 The 2006
Bunkyo-ku, Tokyo 113-8677, Japan (e-mail: kamisawa@cick.jp). criteria deleted the requirement that ‘‘more than one third
Copyright r 2008 by Lippincott Williams & Wilkins of the entire pancreas’’ be involved, allowing segmental
404 J Clin Gastroenterol Volume 42, Number 4, April 2008

2. J Clin Gastroenterol Volume 42, Number 4, April 2008 Diagnostic Criteria for Autoimmune Pancreatitis
TABLE 1. Clinical Diagnostic Criteria for AIP (2006) in Japan6 TABLE 2. Diagnostic criteria for AIP in Asan Medical Center7
1. Diffuse or segmental narrowing of the main pancreatic duct with Criterion I. Pancreatic imaging (essential)
irregular wall and diffuse or localized enlargement of the pancreas by (1) CT: Diffuse enlargement (swelling) of the pancreas and
imaging studies, such as abdominal ultrasonography, CT, and (2) ERCP: Diffuse or segmental irregular narrowing of the main
magnetic resonance imaging pancreatic duct
2. High serum g-globulin, IgG or IgG4, or the presence of Criterion II. Laboratory findings
autoantibodies, such as antinuclear antibodies and rheumatoid factor (1) Elevated levels of IgG and/or IgG4, or
3. Marked interlobular fibrosis and prominent infiltration of (2) Detected autoantibodies
lymphocytes and plasma cells in the periductal area, occasionally with Criterion III. Histopathologic findings
lymphoid follicles in the pancreas Fibrosis and lymphoplasmacytic infiltration
Criterion IV. Association of other postulated autoimmune disease
Proposed by the Research Committee of Intractable Diseases of the Pancreas Definite: I+II+III+IV or
supported by the Japanese Ministry of Health, Labour, and Welfare, and Japan I+II+III or I+III
Pancreas Society.
Diagnosis of AIP is established when criterion 1, together with criterion 2 and/
Probable: I+IV Rediagnosed as definite if response to
or 3, are fulfilled. the steroid is present.
However, it is necessary to exclude malignant diseases such as pancreatic or Possible: I only Rediagnosed as definite if response to
biliary cancers. the steroid is present
Patients are stratified by evidence strength for AIP into (1) definite, (2)
probable, and (3) possible AIP.
AIP cases to be diagnosed. Furthermore, the 2006 criteria
included elevation of the serum IgG4 level as a diagnostic
factor. Finally, the 2006 criteria stressed the need to be the gold standard for diagnosing AIP; many AIP
exclude malignant diseases such as pancreatic or biliary patients at Mayo Clinic are diagnosed using the histologic
cancers, before making the diagnosis of AIP (Table 1). criteria (Table 3).
KOREAN DIAGNOSTIC CRITERIA OUR CASES
7 During the past 8 years, we clinically diagnosed 27
The Korean diagnostic criteria proposed by Asan
patients (21 males, 6 females) as having AIP who have
Medical Center in 2006 stratified the strength of the
been followed long-term. The patients’ average age was
evidence for AIP into ‘‘definite,’’ ‘‘probable,’’ and
‘‘possible.’’ In addition to the Japanese criteria, the
Korean criteria include the patient’s response to steroid TABLE 3. Diagnostic Criteria for AIP (The HISORt Criteria,
therapy and the presence of extrapancreatic lesions as Mayo Clinic)8
diagnostic criteria. The definite AIP category includes Category Criteria
almost the same criteria as the Japanese criteria, whereas
Histology At least one of the following:
patients who have only typical features of AIP on (1) Periductal lymphoplasmacytic infiltrate with
diagnostic imaging are diagnosed as possible AIP. AIP obliterative phlebitis and storiform fibrosis
patients frequently have various extrapancreatic lesions, (LPSP)
including sclerosing cholangitis, sialadenitis, or retro- (2) Lymphoplasmacytic infiltrate with storiform
fibrosis showing abundant (^10 cells/HPF)
peritoneal fibrosis. When extrapancreatic lesions occur IgG4-positive cells
along with typical pancreatic imaging findings of AIP, the Pancreatic Typical: diffusely enlarged gland with delayed (rim)
index of suspicion becomes higher, and the patients are imaging enhancement; diffusely irregular, attenuated main
designated as possible AIP. The diagnosis of AIP is pancreatic duct
confirmed by the patient’s response to steroid therapy, Others*: Focal pancreatic mass/enlargement; focal
pancreatic duct stricture; pancreatic atrophy;
which involves dramatic resolution of narrowing of the pancreatic calcification; or pancreatitis
pancreatic duct; such patients are then reassigned to the Serology Elevated serum IgG4 level (normal, 8-140 mg/dL)
definite AIP category (Table 2). Other organ Hilar/intrahepatic biliary strictures, persistent distal
involvementw biliary stricture, parotid/lacrimal gland
involvement, mediastinal lymphadenopathy,
USA DIAGNOSTIC CRITERIA retroperitoneal fibrosis
In the United States, diagnostic criteria were Response to Resolution/marked improvement of pancreatic/
steroid extrapancreatic manifestation with steroid
proposed by Mayo Clinic in 2006.8 On the basis of these therapyz therapy
criteria, AIP can be diagnosed in patients with Z1 of the
following criteria: (1) diagnostic histology showing *With negative work-up for known etiologies for pancreatic disease, especially
pancreatic/biliary cancer.
lymphoplasmacytic sclerosing pancreatitis or abundant wRadiologic evidence of organ involvement can be confirmed by biopsy
IgG4-positive cell infiltration of pancreatic lymphoplas- showing lymphoplasmacytic infiltrate with abundant IgG4-positive cells or its
macytic infiltrate; (2) characteristic imaging on computed resolution/improvement with steroid therapy.
zSteroid therapy should be given only to patients with negative work-up for
tomography (CT) and pancreatography with elevated known etiologies for pancreatic disease and only to those in whom response can be
serum IgG4 levels; (3) response to steroid therapy of objectively assessed. It should not be used as a substitute for a thorough search for
etiology.
pancreatic and/or extrapancreatic lesions of AIP. The LPSP indicates lymphoplasmacytic sclerosing pancreatitis.
Mayo Clinic criteria stress that histologic criteria should
r 2008 Lippincott Williams Wilkins 405

3. Kamisawa J Clin Gastroenterol Volume 42, Number 4, April 2008
65.5 years (range, 29 to 83 y). Symptoms on presentation COMPARISON OF THE 3 SETS OF CRITERIA
included: painless obstructive jaundice in 22 (81%) The revised Japanese criteria4,5 require that the
patients; exacerbation of preexisting diabetes mellitus in characteristic imaging findings of AIP be present, and
3; pancreatic swelling on screening ultrasound in 2; and that at least 1 of the laboratory criterion or the
general malaise in 1. Mild abdominal pain was present at histopathologic criterion be present. The Korean criter-
onset in 6 patients, and 14 patients (52%) had diabetes ia,6 in addition to the Japanese criteria, include the
mellitus. The diagnoses of diabetes mellitus and AIP were response to steroid therapy and the presence of extra-
made simultaneously in 9 patients, and 5 patients pancreatic lesions. Using the Mayo Clinic criteria,7 AIP
developed an exacerbation of their preexisting diabetes can be diagnosed: solely on the basis of diagnostic
mellitus at the onset of AIP. On CT and ultrasonography, pancreatic histology; by the presence of both the imaging
swelling of the pancreas was noted in 27 patients. On criteria and serum IgG4 level elevation; or by response to
endoscopic retrograde cholangiopancreatography, 19 steroid therapy with elevated serum IgG4 levels and/or
patients had diffuse narrowing of the main pancreatic other organ involvement. In the 27 AIP cases that we
duct, and 9 patients had segmental narrowing of the main recently diagnosed based on the overall clinical picture,
pancreatic duct (head: n = 6, body: n = 1, tail: n = 2). the sensitivity of the Japanese, Korean, and Mayo Clinic
Two patients had a segmental change of the pancreatic criteria was 78% (21/27), 96% (26/27), and 85% (23/27),
tail at first, which progressed to diffuse pancreatic respectively. On the basis of the Japanese criteria,4,5 5
changes, 6 and 12 months later, respectively. Serum seronegative cases fulfilling the imaging criterion without
gammaglobulin, IgG, and IgG4 levels were elevated in histologic examination, and 1 case without pancreatic
50% (13/26), 44% (12/27), and 80% (20/25) of patients, enlargement could not be diagnosed. On the basis of
respectively. Autoantibodies were present in 35% Korean criteria,6 only 1 case without pancreatic enlarge-
(9/26) of patients including antinuclear antigen (7/26) ment could not be diagnosed. On the basis of the Mayo
and rheumatoid factor (5/26). Histologic findings of the Clinic criteria,7 2 seronegative cases with other organ
pancreas biopsied under abdominal ultrasound guidance involvement who responded to steroid therapy could be
included lymphoplasmacytic infiltration with fibrosis, diagnosed in addition to the cases that fulfilled the
compatible with AIP (2/2). Stenosis of the lower bile Japanese criteria.
duct was detected in 23 cases. Extrapancreatic lesions In the absence of a diagnostic serologic marker for
were found in 12 cases (44%): swelling of the AIP, AIP must be diagnosed on the basis of the presence
salivary glands (n = 5), retroperitoneal fibrosis (n = 4), of a combination of abnormalities unique to AIP. It is
thickening of the gallbladder wall (n = 4), swelling important not to misdiagnose pancreatic cancer as AIP,
of the cervical or mediastinal lymph nodes (n = 3), and also not to misdiagnose AIP as pancreatic cancer.
stenosis of the intrahepatic bile duct (n = 3), swelling The Japanese criteria are based on the minimum
of the lacrimal glands (n = 1), pulmonary pseudotumor consensus features of AIP to minimize the risk of
(n = 1), and hypophysitis (n = 1). Abundant infiltration misdiagnosing malignancy. When response to steroid
of IgG4-positive plasma cell and fibrosis were detected therapy is added to the criteria, the diagnostic sensitivity
in swollen salivary glands, a retroperitoneal mass, and is increased. Korean investigators recommend a diagnos-
a pulmonary pseudotumor. Swollen cervical lymph tic trial of steroid therapy in cases that do not fulfill the
nodes were also infiltrated by many IgG4-positive plasma Japanese criteria.6 As relief of narrowing of the pancreatic
cells. duct with steroid administration can be seen as early as 2
A total of 26 patients received steroid therapy. weeks after steroid therapy in AIP cases, and it does not
Indications for steroid therapy included: stenosis of the occur in pancreatic cancer cases, the Korean investigators
bile duct (n = 23), diffuse enlargement of the pancreas advocate a short trial of steroid therapy to differentiate
(n = 2), and associated retroperitoneal fibrosis (n = 2). AIP from pancreatic cancer.6 We also agree that a trial of
The most common regimen involved induction with oral steroid therapy can be used to assist in making the
prednisolone 30 mg/d for 2 weeks, followed by tapering diagnosis when it is used appropriately. However, as
of the dose (5 mg every 1 to 2 wk). Medication was generalist physicians who are not pancreatologists use the
discontinued in 4 patients, but the other 22 patients criteria, it is possible that the facile use of steroid trials
required continued maintenance therapy with predniso- will result in delaying pancreatic cancer surgery, which
lone 2.5 to 5 mg/d. All patients had resolution or could lead to cancer progression in some cases.
improvement clinically, morphologically, and serologi- AIP has characteristic histologic findings in the
cally in the extrapancreatic involvement and also in the pancreas.9,10 Thus, histology may be a gold standard for
pancreatic lesion. One patient relapsed during mainte- diagnosis,11 especially in association with IgG4-immu-
nance therapy, but the biliary stenosis improved with a nostaining. However, an adequate pancreatic core biopsy
temporary dose increase. On follow-up, the prognosis was can be obtained in only limited number of institutes. On
found to be good except in one 83-year-old male who died the other hand, rather specific pancreatic imaging on CT,
of esophageal cancer 3 years after developing AIP. Two ultrasonography, and endoscopic retrograde cholangio-
patients developed pancreatic stones during follow-up, pancreatography is widely available and more clinically
but pancreatic exocrine function did not deteriorate in applicable. Thus, the importance of characteristic pan-
any of the patients. creatic imaging findings, as diagnostic criteria should be
406 r 2008 Lippincott Williams Wilkins

4. J Clin Gastroenterol Volume 42, Number 4, April 2008 Diagnostic Criteria for Autoimmune Pancreatitis
stressed. It has become clear that some patients develop REFERENCES
pancreatic calcification or stone formation during follow- 1. Yoshida K, Toki F, Takeuchi T, et al. Chronic pancreatitis caused
up.12 Some AIP patients in the postacute phase may be by an autoimmune abnormality. Proposal of the concept of
autoimmune pancreatitis. Dig Dis Sci. 1995;40:1561–1568.
found to have an atrophic pancreas or pancreatic 2. Hardacre JM, Iacobuzio-Donahue CA, Sohn TA, et al. Results of
calcification, which are characteristics that are seen in pancreaticoduodenectomy for lymphoplasmacytic sclerosing pan-
patients with ordinary chronic pancreatitis. It will be creatitis. Ann Surg. 2003;237:853–859.
necessary in the future to incorporate atypical imaging 3. Abraham SC, Wilentz RE, Yeo CJ, et al. Pancreaticoduodenectomy
cases, as in the Mayo Clinic criteria. (Whipple resections) in patients without malignancy: are they all
‘‘chronic pancreatitis’’? Am J Surg Pathol. 2003;27:110–120.
Serologic positivity varies by the number of auto- 4. Members of the Criteria Committee for Autoimmune Pancreatitis of
antibodies that are measured. Detection rates tend to the Japan Pancreas Society. Diagnostic criteria for autoimmune
increase as more autoantibodies are assessed. Detection pancreatitis by the Japan Pancreas Society (in Japanese). Suizou
rates for antinuclear antibody and rheumatoid factor, (J Jpn Pan Soc). 2002;17:585–587.
which can be easily measured, are reported to be 43% to 5. Pearson RK, Longnecker DS, Chari ST, et al. Controversies in
clinical pancreatology. Autoimmune pancreatitis: does it exist?
75% and 53% to 71%, respectively, in AIP patients.13–15 Pancreas. 2003;27:1–13.
Serum IgG4 levels are rather significantly and specifically 6. Okazaki K, Kawa S, Kamisawa T, et al. Clinical diagnostic criteria
elevated in AIP patients.16 However, in recent reports, of autoimmune pancreatitis: revised proposal. J Gastroenterol. 2006;
the sensitivity of elevated serum IgG4 levels is 63% 41:626–631.
7. Kim KP, Kim MH, Kim JC, et al. Diagnostic criteria for
to 68%.17–19 Of note, recently, elevated serum IgG4
autoimmune chronic pancreatitis revised. World J Gastroenterol.
levels have been reported in a patient with pancreatic 2006;12:2487–2496.
cancer.20 8. Chari ST, Smyrk TC, Levy MJ, et al. Autoimmune pancreatitis:
On the basis of histologic and immunohistochemical diagnosis using histology, imaging, serology, other organ involve-
examination of various organs in AIP patients, dense ment and response to steroids. Clin Gastroenterol Hepatol. 2006;4:
1010–1016.
infiltration of IgG4-positive plasma cells, and also CD4- 9. Kamisawa T, Wakabayashi T, Sawabu N. Autoimmune pancreatitis
positive or CD8-positive T lymphocytes, and fibrosis have in young patients. J Clin Gastroenterol. 2006;40:847–850.
been observed in peripancreatic retroperitoneal tissue, the 10. Kamisawa T, Okamoto A. Autoimmune pancreatitis: proposal of
bile duct wall, the gallbladder wall, the periportal area of IgG4-related sclerosing disease. J Gastroenterol. 2006;41:613–625.
the liver, the salivary glands, and also in the pan- 11. Chari ST, Echelmeyer S. Can histopathology be the ‘‘Gold
Standard’’ for diagnosing autoimmune pancreatitis? Gastroenterol-
creas.10,21–23 All extrapancreatic lesions present in pa- ogy. 2005;129:2118–2120.
tients with AIP, such as sclerosing cholangitis, sclerosing 12. Takayama M, Hamano H, Ochi Y, et al. Recurrent attacks of
sialadenitis, or retroperitoneal fibrosis, show infiltration autoimmune pancreatitis result in pancreatic stone formation. Am
of abundant IgG4-positive plasma cells, but such infiltra- J Gastroenterol. 2004;99:932–937.
tion is not detected in patients with primary sclerosing 13. Horiuchi A, Kawa S, Hamano H, et al. ERCP features in 27 patients
with autoimmune pancreatitis. Gastrointest Endosc. 2002;55:
cholangitis, Sjogren syndrome, sialolithiasis, chronic 494–499.
alcoholic pancreatitis, or pancreatic cancer. Both pan- 14. Uchida K, Okazaki K, Asada M, et al. Case of chronic pancreatitis
creatic and extrapancreatic lesions of AIP respond well to involving an autoimmune mechanism that extended to retroper-
steroid therapy. Therefore, we proposed the existence of a itoneal fibrosis. Pancreas. 2003;26:92–94.
novel clinicopathologic entity, an IgG4-related sclerosing 15. Kamisawa T, Egawa N, Nakajima H, et al. Clinical difficulties in the
differentiation of autoimmune pancreatitis and pancreatic carcino-
disease.10,23 AIP and its extrapancreatic lesions appear to ma. Am J Gastroenterol. 2003;98:2694–2699.
be the lesions that involved in this systemic disease. In 16. Hamano H, Kawa S, Horiuchi A, et al. High serum IgG4
some cases, only 1 or 2 organs are clinically involved, concentrations in patients with sclerosing pancreatitis. N Engl
whereas in other cases, 3 or 4 organs are affected. Given J Med. 2001;344:732–738.
17. Okazaki K. Autoimmune pancreatitis is increasing in Japan.
the view that this is an IgG4-related systemic disease, we
Gastroenterology. 2003;125:1557–1558.
would recommend IgG4-immunostaining of biopsied 18. Kim KP, Kim MH, Song MH, et al. Autoimmune chronic
extrapancreatic lesions, such as the major duodenal pancreatitis. Am J Gastroenterol. 2004;99:1605–1616.
papilla,24 the bile duct, or the minor salivary gland, as 19. Kamisawa T, Okamoto A, Funata N. Clinicopathological features
an additional criterion for the Japanese diagnostic of autoimmune pancreatitis in relation to elevation of serum IgG4.
Pancreas. 2005;31:28–31.
criteria. 20. Kamisawa T, Chen PY, Tu Y, et al. Pancreatic cancer with a serum
In conclusion, as there is no diagnostic serologic IgG4 concentration. World J Gastroenterol. 2006;12:6225–6228.
marker for AIP, the diagnosis of AIP must be made on 21. Kamisawa T, Funata N, Hayashi Y, et al. Close relationship
the presence of a unique set of characteristics. In general, between autoimmune pancreatitis and multifocal fibrosclerosis. Gut.
diagnostic criteria should have high sensitivity and 2003;52:683–687.
22. Kamisawa T, Egawa N, Nakajima H, et al. Extrapancreatic lesions
specificity; they should be at least invasive as possible in autoimmune pancreatitis. J Clin Gastroenterol. 2005;39:904–907.
and clinically applicable across a wide range of medical 23. Kamisawa T, Nakajima H, Egawa N, et al. IgG4-related sclerosing
settings. Thus, we would recommend IgG4-immunostain- disease incorporating sclerosing pancreatitis, cholangitis, sialadenitis
ing of biopsied extrapancreatic lesions, such as the major and retroperitoneal fibrosis with lymphadenopathy. Pancreatology.
duodenal papilla, the bile duct, or the minor salivary 2006;6:132–137.
24. Kamisawa T, Yu Y, Nakajima H, et al. Usefulness of biopsying
gland, as an additional criterion for the Japanese the major duodenal papilla to diagnose autoimmune pancreatitis:
diagnostic criteria. Furthermore, it is time for interna- a prospective study using IgG4-immunostaining. World J Gastro-
tional consensus on AIP. enterol. 2006;12:2031–2033.
r 2008 Lippincott Williams Wilkins 407