This copy is for personal use only. To order printed copies, contact reprints@rsna.

org
1965

Gastrointestinal Imaging
Chronic Pancreatitis or Pancre-
atic Tumor? A Problem-solving
Approach
Kristy Marie Wolske, MD
Janardhana Ponnatapura, MD Certain inflammatory pancreatic abnormalities may mimic pan-
Orpheus Kolokythas, MD creatic ductal adenocarcinoma at imaging, which precludes accu-
Lauren M. B. Burke, MD rate preoperative diagnosis and may lead to unnecessary surgery.
Rafel Tappouni, MD Inflammatory conditions that may appear masslike include mass-
Neeraj Lalwani, MD forming chronic pancreatitis, focal autoimmune pancreatitis, and
paraduodenal pancreatitis or “groove pancreatitis.” In addition,
Abbreviations: AIP = autoimmune pancre- obstructive chronic pancreatitis can mimic an obstructing ampul-
atitis, CA = carbohydrate antigen, IPMN = in- lary mass or main duct intraductal papillary mucinous neoplasm.
traductal papillary mucinous neoplasm, OCP =
obstructive chronic pancreatitis, PDAC = pan- Secondary imaging features such as the duct-penetrating sign, bili-
creatic ductal adenocarcinoma, PDP = paraduo- ary or main pancreatic duct skip strictures, a capsulelike rim, the
denal pancreatitis, SMA = superior mesenteric
artery, SMV = superior mesenteric vein pancreatic duct-to-parenchyma ratio, displaced calcifications in pa-
tients with chronic calcific pancreatitis, the “double duct” sign, and
RadioGraphics 2019; 39:1965–1982
vessel encasement or displacement can help to suggest the possibil-
https://doi.org/10.1148/rg.2019190011 ity of an inflammatory mass or a neoplastic process. An awareness
Content Codes: of the secondary signs that favor a diagnosis of malignant or inflam-
From the Departments of Radiology of Wake matory lesions in the pancreas can help the radiologist to perform
Forest Baptist Medical Center, 1 Medical Center the differential diagnosis and determine the degree of suspicion for
Blvd, Winston-Salem, NC 27157 (K.M.W., J.P.,
R.T., N.L.); University of Washington, Seattle, malignancy. Repeat biopsy or surgical resection may be necessary
Wash (O.K.); and University of North Carolina to achieve an accurate diagnosis and prevent unnecessary surgery
at Chapel Hill, Chapel Hill, NC (L.M.B.B.).
Presented as an education exhibit at the 2018
for inflammatory conditions.
RSNA Annual Meeting. Received February 4,
2019; revision requested March 25 and received Online supplemental material and DICOM image stacks are available
April 22; accepted April 29. For this journal-based for this article.
SA-CME activity, the author R.T. has provided
disclosures; all other authors, the editor, and the ©
RSNA, 2019 • radiographics.rsna.org
reviewers have disclosed no relevant relationships.
Address correspondence to N.L. (e-mail:
[email protected]).
©
RSNA, 2019
Introduction
Pancreatic ductal adenocarcinoma (PDAC) is the fourth leading
SA-CME Learning Objectives cause of death from cancer among men and women in the United
After completing this journal-based SA-CME States, with an overall mean 5-year survival rate of 8% and a mean
activity, participants will be able to: survival rate for localized disease of 32% of patients (1). In 2018,
■■Identify common inflammatory condi-
more than 55 000 new cases of PDAC were diagnosed, and more
tions that mimic PDAC.
than 44 000 people died of the disease (1). Although the Whipple
■■Describe imaging features that help
in differentiating benign inflammatory procedure offers an improved survival rate for patients with local-
masses from PDAC. ized resectable disease (mortality rate, <5%), morbidity from the
■■Discuss the role of various imaging Whipple procedure can be as high as 40%–50%, which makes
modalities in the visualization and dif- preoperative identification of nonneoplastic pancreatic masses
ferentiation of inflammatory pancreatic important (2,3). Equally important is avoiding delays in diagnosis
masses and pancreatic neoplasms.
that result in progression of undiagnosed PDACs that may rapidly
See rsna.org/learning-center-rg.
become unresectable, leading to high mortality rates.
Three inflammatory entities commonly mimic PDAC at imaging:
mass-forming chronic pancreatitis, focal autoimmune pancreatitis
(AIP), and paraduodenal pancreatitis (PDP) or “groove pancreatitis”
(4). In addition, obstructive chronic pancreatitis (OCP), with diffuse
ductal dilatation and diffuse pancreatic parenchymal atrophy, can
mimic ampullary masses or an intraductal papillary mucinous neo-
plasm (IPMN) of the main duct at imaging. In 5%–35% of Whipple
1966 November-December 2019 radiographics.rsna.org

Teaching Points and therefore cannot exclude the diagnosis of

malignancy. In addition, serum CA 19-9 may not
■■ Smooth narrowing of the pancreatic duct that traverses a
pancreatic mass without abrupt obstruction strongly favors be elevated until the disease is advanced (9).
the diagnosis of an inflammatory mass. The percutaneous false-negative biopsy rate
■■ Pancreatic parenchymal calcifications are commonly seen in for PDAC is as high as 60%, which further
chronic pancreatitis, and peripheral displacement of calcifi- complicates the clinical management of pan-
cations can indicate a new underlying malignancy that may creatic masses. (12,13). Endoscopic US-guided
otherwise be imaging occult or poorly visualized because of fine-needle aspiration has higher sensitivity but
gland heterogeneity from chronic pancreatitis.
has a variable false-negative rate and has issues
■■ In the absence of the pathognomonic capsulelike rim, visual-
with nondiagnostic sampling (14). In a study by
ization of multiple pancreatic or biliary strictures rather than
a solitary stricture and absence of substantial dilatation of the Fritscher-Ravens et al (15) in patients with normal
side branches or upstream pancreatic duct may be helpful pancreatic parenchyma and a focal mass, endo-
clues that a focal mass in the pancreas represents focal AIP. scopic US-guided fine-needle aspiration achieved
Similar to other benign focal inflammatory lesions in the sensitivity of 89%, specificity of 100%, and accu-
pancreas, AIP often shows the duct-penetrating sign, which
racy of 92%. However, in the same study, sensitiv-
further supports a benign diagnosis. Extrapancreatic autoim-
mune disease may be present. ity for the detection of PDAC with endoscopic
■■ Displacement rather than encasement of the common bile
US-guided fine-needle aspiration was markedly
duct and/or the gastroduodenal artery is highly suggestive of lower when a mass was present in patients with
the diagnosis of PDP rather than PDAC. chronic pancreatitis (sensitivity, 54%; specific-
■■ Dilatation of both the common bile duct and the main pan- ity, 100%; and accuracy, 91%) (15). With these
creatic duct, or the double duct sign, is highly suggestive of limitations, imaging continues to be important
an underlying neoplasm. Although uncommon, inflammatory to the identification of alternative diagnoses and
disease can produce this finding, which mimics malignancy.
the guidance of clinical management when initial
biopsy results are negative.

procedures, the final pathologic diagnosis is a PDAC: A Brief Review

nonneoplastic inflammatory pancreatic mass (5). PDAC represents 94% of cancers arising from the
Although some patients with chronic inflamma- exocrine pancreas (1) and is a hypovascular mass
tory masses undergo the Whipple procedure for with extensive fibrosis at histopathologic examina-
symptom management, others undergo pancreatic tion (16), and these are also features of chronic
resection for high clinical suspicion for malignancy focal inflammatory lesions. This histologic overlap
and concern that biopsies have yielded false-nega- may help to explain the difficulty in differentiating
tive results. between the two entities.
Secondary imaging features such as the duct- At multiphase multidetector CT, classic
penetrating sign, the presence of biliary or main PDAC is a hypoattenuating hypoenhancing mass
pancreatic duct skip strictures, a capsulelike rim, with ill-defined margins. Multiple secondary
the pancreatic duct-to-parenchyma ratio, displaced signs of PDAC have been described and com-
calcifications in patients with chronic calcific monly include abrupt pancreatic duct cutoff,
pancreatitis, the double duct sign, and vessel upstream pancreatic duct dilatation, upstream
encasement or displacement may help to diagnose pancreatic parenchymal atrophy, and decreased
an inflammatory condition that may respond to enhancement in the distal pancreatic paren-
medical management or PDAC, even when initial chyma (17). Ductal dilatation is seen in 80% of
biopsy results showed a benign abnormality. tumors in the head of the pancreas and in 50%
of tumors in the body of the pancreas (16,18). In
Clinical Background addition to ductal dilatation, the classic finding
Imaging findings, clinical presentation, risk fac- of upstream parenchymal atrophy is present in
tors, and laboratory findings overlap in patients the majority of tumors (16). Decreased distal pa-
with pancreatic inflammatory processes and those renchymal enhancement was seen in almost 47%
with a pancreatic neoplasm. A history of alco- of patients in a study by Tamada et al (17).
hol abuse and recurrent abdominal pain for at At MRI, the classic appearance of PDAC is a
least 2 years is indicative of chronic pancreatitis; hypointense mass at fat-suppressed T1-weighted
however, chronic pancreatitis is also a known risk MRI and variable signal intensity at T2-weighted
factor for the development of PDAC, and the two MRI (19). At contrast-material–enhanced fat-
entities can coexist in 1.8%–4% of patients (6–8). suppressed T1-weighted MRI, PDAC typically
Serum tumor markers also overlap. A serum CA shows hypoenhancement, which is particularly evi-
19-9 level greater than 40 U/mL is 90% specific dent during the arterial phase (19,20). Diffusion
for a pancreatic malignancy (9–11). Unfortu- restriction improves detection of small, ill-defined
nately, serum CA 19-9 has only 81% sensitivity isointense masses from the normal pancreatic
RG  •  Volume 39  Number 7 Wolske et al  1967

parenchyma or adjacent obstructive pancreatitis duct-penetrating sign, segment strictures longer

(21,22). Diffusion-weighted MRI allows detec- than 6 mm, and decreased attenuation in the
tion of pancreatic masses with similar sensitivity to upstream pancreatic parenchyma (32).
that of gadolinium-enhanced MRI (23,24). PDAC
often shows increased signal intensity at diffusion- Mass-forming Chronic Pancreatitis
weighted MRI (b >500 sec/mm2) with relatively
low apparent diffusion coefficients (25). However, Background
diffusion restriction is not exclusive to PDAC In cases of chronic pancreatitis, 27%–50% of
and can be seen in conditions with inflammatory patients present with a localized mass or mass-
causes (eg, AIP) (26). forming pancreatitis, and 71% of focal lesions
Isoattenuating masses that are not visual- manifest in the pancreatic head (33,34). The
ized discretely at routine multidetector CT are typical patient history and risk factors includ-
particularly difficult to diagnose, and they ac- ing abdominal pain, weight loss, nausea, and
count for up to 10% of PDACs (27–29). Studies jaundice commonly overlap with the patient
(30,31) have suggested that dual-energy CT may history and risk factors of PDAC. CT may only
offer improved visualization of small PDACs. allow accurate differentiation of mass-forming
Specifically, low-voltage monoenergetic images pancreatitis from PDAC in 77% of cases (35).
that are reconstructed with noise reduction algo-
rithms show improved tissue contrast compared Imaging Features of
with polychromatic images and may allow better Mass-forming Pancreatitis
visualization of small pancreatic tumors. At multidetector CT, both mass-forming pan-
In the absence of a visualized mass at CT, creatitis and PDAC are typically hypoattenuat-
focal pancreatic duct dilatation, mild peripancre- ing and hypoenhancing. The primary imaging
atic infiltration, peritumoral cysts, or soft tissue features of mass-forming pancreatitis include
around the celiac artery suggest the possibility of an iso- or hypointense mass at T1-weighted
an occult malignancy (29,32). When a second- MRI and an iso- or hyperintense mass at
ary sign is present without a visible mass, further T2-weighted MRI, which are findings that
investigation with MRI or endoscopic US should are similar to those of PDAC. Fat-suppressed
be recommended (16). T1-weighted MRI sequences can provide
Additional findings including peritumoral increased lesion conspicuity compared with
cysts; persistent hypoenhancement during the non–fat-saturation sequences, and mass-form-
arterial, portal venous, and delayed phases ing pancreatitis, similar to PDAC, typically
of contrast-enhanced MRI; and marked T1- is hypoenhancing at arterial phase contrast-
weighted hypointensity similar to that of the enhanced MRI.
renal medulla also were seen more commonly in Imaging findings of chronic pancreatitis
PDAC than in focal chronic pancreatitis (22). such as parenchymal calcifications and pseu-
In particular, peritumoral cysts, which appear docysts should raise the possibility that a focal
as small irregular oval or linear areas of fluid mass represents focal inflammatory change
signal intensity in the periphery of or adjacent (33,36,37). Additional findings favoring mass-
to a mass, are highly suggestive of malignancy. forming pancreatitis over malignancy include
In their study, Lee et al (22) saw peritumoral the duct-penetrating sign, collateral branch
cysts in 84.5% of patients with PDAC and in duct dilatation, and a pancreatic duct-to-pa-
15.5% of those with focal chronic pancreatitis, renchyma ratio of less than 0.34 (4,38–40).
with an odds ratio for association of peritumoral
cysts with PDAC of 28.85. In comparison, Duct-penetrating sign.—Smooth narrowing of
abrupt pancreatic duct cutoff, which is widely the pancreatic duct that traverses a pancreatic
considered a finding that is highly suspicious for mass without abrupt obstruction strongly favors
PDAC, had an odds ratio of 9.07 in the same the diagnosis of an inflammatory mass (39). The
study. The cause of peritumoral cysts is uncer- duct-penetrating sign is said to be present when
tain, but they are thought to be secondary to the main pancreatic duct in the mass is seen in
obstruction of the side branches of the pancre- its entirety, without obstruction. The duct may
atic duct from intraductal PDAC or adjacent be smoothly narrowed or of normal caliber.
desmoplasia. The duct-penetrating sign is 96% specific for an
In the evaluation of unexplained pancreatic inflammatory pancreatic mass, with sensitivity
duct strictures with upstream ductal dilatation, of 85% and accuracy of 94% (39). Although the
secondary signs also can be helpful. Findings of duct-penetrating sign can be seen at CT, it is
malignant stricture include duct enhancement at better visualized at MR cholangiopancreatogra-
the transition site, abrupt transition without the phy (Fig 1).
1968 November-December 2019 radiographics.rsna.org

Figure 1.  Imaging findings that favor diagnosis of an inflammatory condition (or mass-forming pancre-
atitis) rather than malignancy. (a) Coronal T2-weighted MR image through the pancreatic head shows a
heterogeneous masslike lesion (black arrows). The proximal main pancreatic duct has a decreased caliber and
can be seen traversing through the mass (white arrows). (b) MR cholangiopancreatogram clearly shows the
narrowed main pancreatic duct in the masslike lesion (arrows), which is consistent with the duct-penetrating
sign. Prominent side branches are seen along the course of the distal pancreatic duct (arrowheads).

Collateral Duct Dilatation.—The presence of peripheral displacement of calcifications can

collateral duct or side-branch dilatation in the indicate a new underlying malignancy (Fig 2b)
uninvolved or distal pancreas further supports that may otherwise be imaging occult or poorly
an inflammatory cause (41,42). Dilatation of visualized because of gland heterogeneity from
pancreatic duct side branches is thought to be chronic pancreatitis (44,46).
secondary to a traction effect from fibrosis in
a patient with chronic pancreatitis, rather than The Double Duct Sign.—The junction of the
mass effect from a neoplasm, where duct oblit- pancreatic duct and the bile duct may be focally
eration would be expected (4). dilated in 70% of cases (ampulla of Vater) and
eventually may be open in the major duodenal
Imaging Findings that Favor Malignancy papilla (papilla of Vater). Therefore, an obstruc-
Certain imaging features such as a duct-to- tion at the papilla or peripapillary region may give
parenchyma ratio greater than 0.34, displaced rise to dilatation of both ducts (47). Simultaneous
calcifications in a patient with chronic calcific dilatation of the common bile duct and pancreatic
pancreatitis, the double duct sign, vessel encase- duct, known as the double duct sign (48,49), favors
ment, vessel deformity, and a superior mesenteric the diagnosis of a malignancy (Fig 2c).
artery (SMA) to superior mesenteric vein (SMV) Classically, the double duct sign is seen in
ratio of greater than 1 favor the diagnosis of PDACs involving the head of the pancreas and
PDAC (43–45). is seen in up to 77% of pancreatic head PDACs
(37,50). However, the double duct sign is non-
Duct-to-Parenchyma Ratio.—The pancreatic specific and can develop secondary to inflamma-
duct-to-parenchyma ratio also can be helpful in tory processes in the pancreas (eg, AIP) as well as
determining the degree of suspicion for malig- other nonmalignant conditions (49).
nancy (41). At endoscopic US, a pancreatic duct-
to-parenchyma ratio of greater than 0.34 strongly Vessel Encasement and Vessel Deformity.—Soft-
favors the diagnosis of PDAC (40). In patients tissue attenuation that encases the adjacent vas-
with a pancreatic duct-to-parenchyma ratio of culature is highly suggestive of the extraglandular
greater than 0.34, there is marked upstream pan- spread of PDAC and is crucial in determining
creatic ductal dilatation with marked parenchymal tumor resectability. In addition to vessel encase-
atrophy, which are the imaging hallmarks of PDAC ment, vessel caliber changes also are seen fre-
(Fig 2a). Relatively mild ductal dilatation with quently, including circumferential vessel narrow-
mild upstream parenchymal atrophy (pancreatic ing, occlusion, and vessel deformity. The SMV
duct-to-parenchyma ratio, <0.34) raises the possi- teardrop sign, a teardrop-shaped deformity of the
bility of an inflammatory nonneoplastic cause (40). SMV, is highly indicative of SMV encasement
(43) (Fig 2d). Although uncommon, inflamma-
Displaced Calcifications in Chronic Pancre- tory conditions such as AIP may result in loss of
atitis.—Pancreatic parenchymal calcifications fat planes with peripancreatic vessels or may have
are commonly seen in chronic pancreatitis, and associated sclerosing mesenteritis or retroperito-
RG  •  Volume 39  Number 7 Wolske et al  1969

Figure 2.  Imaging findings that favor diagnosis of a malignancy rather than an inflammatory condition.
(a) Coronal CT image shows a duct-to-parenchyma ratio (maximum diameter of the diffusely dilated main
pancreatic duct [*] and the overlying atrophic parenchyma [arrows]) of greater than 0.5). (b) Axial CT image
shows diffuse calcifications in the background parenchyma and peripheral displacement of calcifications by a
focal hypoattenuating lesion (dotted circle) in the pancreatic body. (c) MR cholangiopancreatogram shows the
double duct sign, or dilatation of both the pancreatic duct (double arrows) and the common bile duct (single
arrow). (d) Axial CT image shows the teardrop sign (arrows), a teardrop-shaped deformity of the SMV due to
vascular encasement. Note the loss of fat in the perivascular space. (e) Axial CT image shows the SMA-to-SMV
ratio, or the decreased caliber of the SMV (arrowhead) (almost the same size as the SMA [arrow]), of greater than
or equal to 1.0. Note the loss of fat in the perivascular space.

neal fibrosis mimicking extraglandular spread of diagnosis of malignancy rather than an inflamma-
pancreatic cancer (39,51). tory process in the pancreas (44,45).
The release of vasoactive substances in pancre-
SMA-to-SMV Ratio.—Enlargement of the SMA atitis is thought to increase regional blood flow,
relative to the SMV with a ratio greater than or resulting in increased diameter of the distensible
equal to 1.0 is an additional sign favoring the SMV compared with the adjacent SMA. Although
1970  November-December 2019 radiographics.rsna.org

it is not well understood, the increase in the caliber supports the diagnosis of an inflammatory mass
of the SMA in patients with PDAC may be due to (39,59,60). In cases of ductal dilatation without
increased resistance to blood flow in the pancreas a visualized mass, MRI can provide better lesion
secondary to the presence of malignancy, or it may detection than that of multidetector CT (61).
be due to wall thickening from perivascular inva- The better soft-tissue contrast of MRI compared
sion. The SMA-to-SMV ratio is not 100% sensi- with that of multidetector CT can help in the
tive or specific, but it can be considered a suspi- identification and differentiation of small pancre-
cious finding for PDAC if it is greater than 1 or if atic tumors from other benign masslike lesions
other signs also support the diagnosis. such as noninflammatory nonneoplastic pan-
creatic head hypertrophy and benign focal fatty
Imaging Modalities for Mass-forming infiltration of the pancreas (61). MR elastography
Pancreatitis can also help to differentiate malignancy from
At transabdominal US, it is difficult to distinguish chronic inflammation. Liu et al (62) showed that
mass-forming chronic pancreatitis from PDAC a combination of 40-Hz MR elastography and
because the imaging features overlap, and masses dynamic contrast-enhanced MR imaging may
may be obscured by overlying bowel gas or the help in the diagnosis of malignancy. They con-
patient’s body habitus (52,53). Endoscopic US cluded that an enveloping neoplasm may exhibit
offers improved visualization of masses and is higher stiffness than does chronic inflammation
used routinely to perform targeted biopsies. In because of obvious fibrous tissue infiltration.
masses with negative fine-needle aspiration re- However, these findings require more study.
sults, contrast-enhanced endoscopic US may help
to differentiate chronic pancreatitis from PDAC PDAC versus Mass-forming Pancreatitis:
(54). A meta-analysis (55) of contrast-enhanced A Practical Approach
transabdominal and endoscopic US for evalua- None of the secondary signs of PDAC and
tion of pancreatic masses found sensitivity of 89% mass-forming pancreatitis have 100% sensitivity
and specificity of 84% for the diagnosis of PDAC or specificity. A diagnosis can be favored if more
when the lesion had a hypovascular hypoen- secondary imaging signs are present to support
hanced appearance. In the diagnosis of neoplastic either a benign inflammatory process or a neoplas-
lesions, sensitivity was 95% and specificity was tic process (Table 1; Table E1). Figures 3–6 show
72% (55). Studies suggest that contrast-enhanced representative patients and provide a practical
endoscopic US–guided fine-needle aspiration can approach to diagnosis.
help to identify hypoenhanced hypovascular areas
for targeted biopsy and to avoid areas of prob- Autoimmune Pancreatitis
able necrosis (56). Similarly, US elastography has
shown promise for diagnosis of pancreatic can- Background
cers, with sensitivity of 89.5% (57). Elastography Clinically, symptoms of AIP can mimic those of
combined with guided fine-needle aspiration may PDAC, with development of obstructive jaundice,
further improve results, with accuracy of 94.4%, acute or chronic abdominal pain, weight loss, and
sensitivity of 93.4%, and specificity of 100% (58). steatorrhea (63). AIP has classically been defined
Overall, US elastography and contrast-enhanced by elevated immunoglobulin 4 levels. Patients
endoscopic US can help to identify better sites for with PDAC occasionally show elevated immuno-
targeted fine-needle aspiration and to avoid false- globulin 4 levels, although they are uncommon.
negative results and necrotic areas. Elevated carcinoembryonic antigen and CA 19-9
Multidetector CT provides better visualiza- levels are rarely seen in patients with AIP (64).
tion of pancreatic parenchymal calcifications than Two types of AIP have been described (Table
do US and MRI and allows better assessment of 2): type 1 lymphoplasmacytic sclerosing pancre-
findings of calcification displacement due to a atitis and type 2 idiopathic duct-centric pancre-
mass arising in a patient with chronic inflamma- atitis (65–69). Type 1 disease is characterized by
tion. Multidetector CT also provides excellent elevated serum immunoglobulin 4 levels and is
visualization of pancreatic parenchymal atrophy, strongly (60%) associated with extrapancreatic
duct structure, vessel encasement, and peripan- autoimmune processes (65–67). The pancreatic
creatic or metastatic spread of disease and allows involvement is typically diffuse (more than 60%),
assessment for staging and resection. and approximately 85% of patients develop
Compared with multidetector CT, MRI can diabetes, which may be a clue to the diagnosis.
be helpful for problem solving. MR cholangio- Symptoms typically improve with administration
pancreatography, especially with secretin, can of steroid therapy. Type 2 AIP can be more dif-
be used to assess the main pancreatic duct for ficult to diagnose clinically, and the imaging find-
the duct-penetrating sign, the presence of which ings can closely mimic PDAC (70). Type 2 AIP
RG  •  Volume 39  Number 7 Wolske et al  1971

Table 1: Summary of Imaging Findings of Mass-forming Pancreatitis and PDAC

Mass-forming
Imaging Finding Pancreatitis PDAC Comments
Duct-penetrating May be Typically Reliable sign of a benign abnormality (specificity, 96%;
sign present absent sensitivity, 85%), seen with mass-forming pancreatitis, AIP,
and PDP
Collateral duct May be Typically Inflammation causes traction over the side branches and
dilatation present absent dilatation
Duct-to-parenchyma Typically May be Smooth pancreatic ductal dilatation with an atrophic overly-
ratio >0.34 absent present ing parenchyma; reliable sign of malignancy (specificity,
97%; sensitivity, 94%)
Displaced calcifica- Typically May be Mass evolving from preexisting chronic pancreatitis displaces
tions absent present the calcification toward the periphery
Double duct sign Typically May be Peripapillary obstruction (specificity, 63%–80%; sensitivity,
absent present 50%–76%)
SMA-to-SMV ratio Typically May be Peritumoral fatty infiltration may lead to deformity and
>1 absent present decreased caliber of the SMV. Usually, the SMV is larger
in diameter than the SMA. This sign, along with other sup-
portive imaging signs, may help in making a diagnosis.
Vessel encasement or Typically May be Occasionally, AIP can have perivascular inflammatory
deformity absent present stranding mimicking PDAC
Note.—The online supplemental version of the table (Table E1) includes reference images.

may manifest with normal serum immunoglobu- In the absence of diffuse disease, the imaging
lin 4 levels, and focal pancreatic involvement that appearance of focal AIP can closely mimic that
closely mimics malignancy is seen in up to 85% of a neoplasm. The focal inflammatory mass may
of cases. Extrapancreatic solid organ autoimmune have indistinct margins, periglandular inflamma-
disease is also uncommon in type 2 AIP, with the tion may mimic extrapancreatic spread of ma-
exception of inflammatory bowel disease, which lignancy (39), and involvement of the common
can be seen in 30% of cases and may be a clue to bile duct could result in a nonmalignant cause of
the diagnosis (Fig 7b). the double duct sign (74,75). In the absence of
In addition to extensive overlap of clinical and the pathognomonic capsulelike rim, visualization
imaging features, a high false-positive biopsy rate of multiple pancreatic or biliary strictures rather
for carcinoma further contributes to unneces- than a solitary stricture and absence of substantial
sary surgical management in cases of AIP. At dilatation of the side branches or upstream pan-
histologic evaluation, the rate of false-positive creatic duct may be helpful clues that a focal mass
or suspected carcinoma is as high as 32% for in the pancreas represents focal AIP (76). Similar
biopsies and 41% for fine-needle specimens (71). to other benign focal inflammatory lesions in the
Although AIP is uncommon, accounting for only pancreas, AIP often shows the duct-penetrating
1.9%–6.6% of cases of chronic pancreatitis, it is sign, which further supports a benign diagnosis
the diagnosis in almost 25% of Whipple proce- (77). Extrapancreatic autoimmune disease may be
dures performed for nonneoplastic causes (72). present (Fig 8).

Imaging Features of AIP Imaging Modalities for AIP

In the diffuse form of AIP, type 1, the classic imag- Multidetector CT readily shows the classic
ing findings include diffuse sausage-like enlarge- capsulelike rim seen in AIP, if it is present (Fig
ment of the gland, a smooth contour, and loss of 7a). Invasion of malignancy in the extraglandular
normal pancreatic lobulations. A low-attenuation planes or vessel encasement are readily assessed
halo or capsulelike rim is pathognomonic for with multidetector CT. Occasionally, inflamma-
AIP type 1 when present (Fig 7). Homogeneous tion of the periglandular fatty tissue may mimic
contrast enhancement in both early and delayed extraglandular malignant soft-tissue invasion in
phases is common (65,73). Reactive lymphade- patients with AIP, and this remains an imaging
nopathy may be present, but calcifications and pitfall in the attempt to distinguish AIP from
pseudocysts are typically absent. Areas of stenosis PDAC (39,51).
can develop in the pancreatic duct and the distal The classic halo or fibrotic rim seen in AIP
common bile duct because of inflammation. is hypointense at T2-weighted MRI and shows
1972  November-December 2019 radiographics.rsna.org

Figure 3.  Mass-forming chronic pancreatitis in the pancreatic head in a 49-year-old woman. (a) Axial CT image
shows a dilated main pancreatic duct (). The overlying parenchyma is not atrophic and is well maintained. The duct-
to-parenchyma ratio is less than 0.5. Scattered calcifications are seen in the parenchyma (arrowheads). (b) Coronal CT
image shows a masslike lesion through the pancreatic head (circle). There is abrupt narrowing of the pancreatic duct,
which penetrates through the parenchyma (arrow).

Figure 4.  PDAC of the pancreatic head in a

56-year-old man. (a) Axial CT image shows
that the SMV (arrowhead) and SMA (black
arrow) are of comparable size and that there
is perivascular soft-tissue encasement (white
arrow). The outline of the SMV is deformed.
(b) Axial CT image shows a locally infiltrative
hypoattenuating mass in the pancreatic head
that is invading the gastroduodenal artery
(arrow). (c) Axial CT image shows the normal
caliber of the distal pancreatic duct and the
well-maintained overlying parenchyma (ar-
rowhead). Note the dilated distal common
bile duct (*). No background parenchymal
calcifications are shown.

weak enhancement at contrast-enhanced MRI to the diagnosis of an underlying autoimmune

(Fig 9a, 9b) (39). MR cholangiopancreatogra- inflammatory process (26,78).
phy shows either irregular stenosis in the main Studies have also suggested a role for diffu-
pancreatic duct or the duct-penetrating sign. sion-weighted MRI in the detection and diagno-
Multiple pancreatic duct strictures or a concomi- sis of AIP and for monitoring response to therapy
tant common bile stricture is highly suggestive of (59). For apparent diffusion coefficients of less
AIP rather than malignancy and is best detected than 1.26 { 10-3 mm2 per second, Hur et al (73)
at MR cholangiopancreatography (77). Stud- found sensitivity of 83% for the diagnosis of AIP
ies have shown improved visualization of the and specificity of 79% for differentiating AIP
duct-penetrating sign and improved detection from PDAC. Diffusion-weighted imaging alone
of multiple pancreatic duct strictures at secretin does not allow AIP to be distinguished reliably
MR cholangiopancreatography (60). Incidental from PDAC, and additional more specific imag-
findings of inflammatory bowel disease (Fig 7b) ing should be performed, when it is available.
or secondary sclerosing cholangitis with multiple Combining markedly decreased apparent dif-
biliary strictures may provide additional clues fusion coefficients with the presence of additional
RG  •  Volume 39  Number 7 Wolske et al  1973

Figure 5.  Mass-forming chronic pancreatitis with a pseudocyst in a 43-year-old man. (a, b) Axial CT images
show a diffusely prominent main pancreatic duct (double arrows) and a dilated common bile duct (black arrow
in b) representing the double duct sign. The overlying parenchyma is well maintained, and the duct-to-paren-
chyma ratio is less than 0.5. (c) Axial CT image shows scattered foci of calcifications in the pancreatic head (ar-
rows). A cystic lesion in the pancreatic head shows hemorrhagic content, which is consistent with a pseudocyst.
(d) Axial CT image shows the pancreatic head with a masslike enlargement (circle).

Figure 6.  PDAC and chronic pancreatitis in a 49-year-

old woman. (a) Axial CT image shows an ill-defined and
slightly hypoattenuating mass in the pancreatic head
(circle), peripheral displacement of calcifications (yel-
low arrow), an SMA-to-SMV (red and blue arrows, re-
spectively) ratio of almost 1, and a teardrop sign in the
SMV. The perivascular fat planes are invaded by the mass
(white arrow). (b) Coronal CT image shows the dilated
pancreatic duct () and side branches (arrowheads).
(c) Axial CT image shows the dilated common bile duct
(double arrows).
1974  November-December 2019 radiographics.rsna.org

Table 2: Types of AIP

Type 1: Lymphoplasmacytic sclerosing pancreatitis
  Positive immunoglobulin 4 tissue staining results
  Elevated serum immunoglobulin 4 level
  Extrapancreatic organ involvement is common (60% of cases)
  Inflammatory bowel disease in only 2%–6% of cases
  Older patients (age usually >60 years), more common in men than in women
  Pancreatic involvement is diffuse in 60% and focal in 40% of patients
  Obstructive jaundice in 75% of cases
  Patients may relapse after steroid therapy is completed
Type 2: Idiopathic duct-centric chronic pancreatitis
  Immunoglobulin 4 tissue staining results are often negative
  No elevation in serum immunoglobulin 4 level
  No extrapancreatic solid organ involvement
  Inflammatory bowel disease in 30% of cases
  Younger patients (mean age, 43 years), equally common in men and women
  Focal pancreatic lesion at imaging in 85% of cases
  Obstructive jaundice in 50% of cases
  Patients rarely relapse after steroid therapy is completed

Figure 7.  Type 2 AIP in a 40-year-old man. (a) Contrast-enhanced coronal CT image through the pancreatic head shows a
focal pancreatic head mass (circle). Note the smooth outline, the loss of normal pancreatic lobulations, and the hypoattenuating
fibrotic rim (arrow). (b) CT image through the sigmoid colon shows mural thickening and associated hyperenhancement. Associ-
ated pericolonic hyperemia and vascular congestion (black arrows) and lymphadenopathy (white arrow) are also seen. Colonos-
copy and biopsy findings were consistent with ulcerative colitis. (Full DICOM image stacks are available online for Fig 7a and 7b.)

signs can be helpful in diagnosis of AIP. For differ- under investigation. For example, early research
entiating AIP from PDAC, studies have found that shows a significant difference in time–signal-
the capsulelike rim sign has specificity as high as intensity curves, with PDAC typically showing
97%–100%; however, sensitivity is only 29% (73). type 2 time–signal-intensity curves with contrast
Similarly, skip strictures in the common bile duct enhancement followed by slow progressive en-
and the main pancreatic duct have been shown hancement, while focal chronic pancreatitis typi-
to have 100% specificity for AIP, but sensitivity is cally shows type 3 time–signal-intensity curves,
low, at 33% and 44%, respectively (73). with fast enhancement followed by a signal
Perfusion-weighted MRI is also being inves- intensity plateau. Additional parameters including
tigated for the assessment of focal pancreatic extravascular extracellular space volume fraction
lesions. The perfusion technique most widely analysis were also found to be helpful for differ-
used in abdominal imaging is dynamic contrast- entiation of chronic pancreatitis from PDAC, but
enhanced MRI (59). Various quantitative param- further studies are needed to validate results from
eters that may offer additional clues for differen- initial small trials (59,79,80). Figure 10 shows a
tiation of focal pancreatic lesions are currently diagnostic approach when AIP is suspected.
RG  •  Volume 39  Number 7 Wolske et al  1975

Figure 8.  Type 2 AIP in a 42-year-old man. (a) Axial T2-

weighted MR image through the proximal pancreas shows
an ill-defined intermediate-signal-intensity masslike lesion
(arrows). (b) Coronal T2-weighted MR image through the
pancreatic head shows the narrowed caliber of the pancreatic
duct traversing through the same region (arrow). Note co-
existing focal stricture involving the distal common bile duct
(arrowhead). (c) Axial contrast-enhanced delayed-phase MR
image through the same region shows focal masslike lesion
(arrowheads). The rest of the pancreas is unremarkable.

Figure 9.  Type 1 AIP in a 35-year-old woman. (a) Axial T2-weighted MR image shows a hypointense fibrotic
rim (arrows). (b) Axial contrast-enhanced MR image shows a homogeneously enhancing pancreas and a non-
enhancing fibrotic rim (arrows).

Management of AIP a pseudotumor that may extend into the adjacent

The mainstay of therapy is administration of cor- pancreatic head, mimicking locally invasive PDAC
ticosteroids. Relapse is frequent in type 1 AIP but of the pancreatic head (82). Three distinct sub-
rare in type 2 AIP. Patients with steroid-refractory types of PDP have been described (82), each with
cases should be treated with immunomodulators different imaging and histopathologic features. The
in conjunction with steroids or rituximab (70). solid tumoral type, or PDP type 1, manifests as a
solid pseudotumor with minimal cystic change.
Paraduodenal Pancreatitis Cysts may comprise less than 50% of the mass or
may be completely absent. As described by Muraki
Background et al (82), type 1 PDP may manifest as a solid-ap-
PDP, also known as groove pancreatitis or cystic pearing sheetlike mass in the pancreaticoduodenal
duodenal dystrophy, is a focal form of pancreatitis groove or a more rounded expansile lesion involv-
centered at the pancreaticoduodenal groove (81). ing the pancreatic head. The expansile solid form,
The resulting inflammation or fibrosis can form or subtype 1B, is particularly difficult to distinguish
1976  November-December 2019 radiographics.rsna.org

Figure 10.  Diagnostic approach in a case of suspected AIP versus PDAC (+ signifies focal mass is present). AdenoCa = PDAC,
fAP = focal AIP, MPD = main pancreatic duct.

from PDAC or a peripapillary neoplasm, given its creaticoduodenal groove, with involvement of the
round configuration, its solid appearance due to expected region of the accessory duct or acces-
microcystic change or absent cysts, and the possi- sory ampulla and medial duodenal wall thick-
bility of associated pancreatic parenchymal atrophy ening (81,82) (Fig 11). Cystic change should
reported in 44% of patients with PDP type 1B prompt consideration of PDP as the underlying
in the study by Muraki et al (82). Much easier to diagnosis, with the understanding that up to
identify because of the presence of cystic change in 20% of patients have no visible cysts in the le-
the pancreaticoduodenal groove is the cyst-forming sion (59). A sheetlike mass or sandwich sign, with
subtype of PDP, or type 2 PDP, in which the le- a linear mass centered in the groove, is highly
sions are predominantly cystic, with cysts account- suggestive of type 1A PDP, even in the absence
ing for greater than 80% of the lesion (82). The of visible cysts. A solid round tumoral subtype
ill-defined type, or type 3 PDP, is not like a mass of PDP (type 1B “rice ball pattern” as described
and is, therefore, less likely to mimic malignancy. by Muraki et al [82]) exists and, in the absence
At resection of nonneoplastic pancreatic le- of cystic change in the groove, is not reliably
sions, PDP was found in approximately 27% of distinguishable from a peripapillary neoplasm
cases, and as many as two-thirds of patients had or PDAC of the pancreatic head. The absence of
a presurgical diagnosis of pancreatic or peripapil- biliary dilatation or a lack of substantial pancre-
lary cancer (82). Patients with PDP frequently atic parenchymal atrophy may suggest a nonneo-
experience severe chronic pain or duodenal outlet plastic diagnosis (80). In addition, the presence
obstruction. Studies indicate that sustainable of the duct-penetrating sign, thickening of the
symptom relief can be achieved in more than medial duodenal wall, or widening of the distance
two-thirds of patients who undergo medical or between the ampulla and duodenal lumen may
endoscopic treatment including sphincterotomy, suggest the diagnosis of PDP (88). Displacement
pancreatic stent placement, or cyst drainage rather than encasement of the common bile duct
(82–86). Although studies have shown that the and/or the gastroduodenal artery is highly sugges-
Whipple procedure allows successful management tive of the diagnosis of PDP rather than PDAC
of pain and improves quality of life scores in most (4). In addition to the radiologic findings, clinical
patients with PDP, given the morbidity associated correlation may be helpful, because elevated CA
with the procedure, accurate preoperative diagno- 19-9 and overt jaundice are uncommon in PDP
sis preserves the opportunity for a trial of medical and are highly suggestive of PDAC.(88).
or endoscopic management (81,87).
Imaging Modalities for PDP
Imaging Features of PDP At transabdominal US, PDP and PDAC may have
Key imaging features that suggest consideration similar appearances. However, at endoscopic US,
of PDP in the differential diagnosis include a identification of tiny cysts in the pancreaticoduo-
solid or solid and cystic mass centered at the pan- denal groove can suggest the diagnosis of PDP.
RG  •  Volume 39  Number 7 Wolske et al  1977

Figure 11.  Type 1 solid PDP (groove pancreatitis) in a 39-year-old man. (a, b) Axial CT images show
an ill-defined hypoattenuating masslike lesion in the duodenopancreatic groove (arrows). Scattered pa-
renchymal calcifications appear in the pancreatic head. (c) Coronal CT image shows the extent of the
masslike lesion in the duodenopancreatic groove (arrows) and associated cystic changes in the duodenal
wall (). The distal common bile duct is partially visible. (d) MR cholangiopancreatogram shows focal nar-
rowing of the distal common bile duct (arrow) and subtle prominence of the side branches (arrowheads).
Note the widening of the duodenopancreatic groove and cystic changes in the duodenal wall (d).

At multidetector CT, PDP mimics PDAC that distance between the ampulla and the duodenal
involves the pancreaticoduodenal groove. PDP is lumen. Extensive fibrosis in the medial duodenal
hypointense during the arterial phase, with progres- wall in PDP can lead to a nonmalignant cause
sive late phase enhancement due to fibrosis, which of the double duct sign. In addition, MR chol-
is similar to PDAC. Visualization of a fat plane sepa- angiopancreatography may allow visualization of
rating the mass from the pancreas may help exclude subtle dilatation of the side branches or collateral
the pancreatic origin of the lesion. Multidetector duct, suggesting a chronic fibrotic or inflamma-
CT can help in determining whether the adjacent tory process rather than malignancy. Secretin-en-
distal common bile duct or adjacent vessels such hanced MR cholangiopancreatography may im-
as the gastroduodenal artery are encased, which prove visualization of the duct-penetrating sign in
supports the diagnosis of malignancy, or displaced, cases of PDP (91). Figure 13 shows an approach
which supports the diagnosis of PDP (89,90). to use when diagnosis of PDP is suspected.
Typically, the pancreaticoduodenal mass
in patients with PDP is iso- or hypointense at Obstructive Chronic Pancreatitis
T1-weighted MRI and iso- or hyperintense at
T2-weighted MRI and can mimic PDAC. At T2- Background
weighted MRI, microcysts in the mass that were OCP is a subtype of chronic pancreatitis where the
not detected at multidetector CT suggest the duct demonstrates a uniform contour and diffusely
diagnosis of PDP (Fig 12). dilated appearance (92). At pathologic examina-
At MR cholangiopancreatography, PDP may tion, OCP shows periductal fibrosis and subse-
cause smooth narrowing of the distal common quent ductal dilatation. Diffuse ductal changes may
bile duct and pancreatic duct, with widened be secondary to chronic inflammatory stenosis of
1978  November-December 2019 radiographics.rsna.org

Figure 12.  Type 1 solid PDP (groove pancreatitis) in a 41-year-old man. (a) Axial fat-suppressed T2-weighted
MR image shows an ill-defined masslike thickening of the medial duodenal wall, with intermediate signal in-
tensity (arrows). At presentation, the patient was found to have duodenal outlet obstruction. Note the dilated
stomach (). (b) Axial fat-suppressed T2-weighted MR image shows a predominantly solid masslike area in the
pancreatic head and microcystic changes along the duodenal wall (arrow).

Figure 13.  Diagnostic approach in a case of suspected solid variant PDP (groove pancreatitis) or PDAC. AdenoCa =
adenocarcinoma, CBD = common bile duct, GDA = gastroduodenal artery, MDP = main pancreatic duct.

the papilla after repeated microtrauma due to the Imaging Modalities for OCP
passage of stones and/or biliary sand, which leads Identifying small juxtapapillary neoplasms can
to fibrosis and narrowing of the sphincter of Oddi prove difficult at imaging, especially at CT, even
and, eventually, the onset of OCP (46). Although with contrast material administration, after which
protein inspissation may occur, pathologic changes the tumor may appear isoattenuating when com-
in the ductal mucosa and calcification are infre- pared with the pancreatic parenchyma.
quent. At gross examination, the pancreatic duct is Few studies have shown that US and endoscopic
dilated and there is variable parenchymal atrophy. US are more sensitive than CT in identifying small
At imaging, OCP may mimic juxtapapillary neo- solid peripapillary or ductal lesions (94,95). Endo-
plasms (causing papillary obstruction) and main- scopic US-guided fine-needle aspiration has been
duct IPMNs) (Table 3)(4). proposed as an option when a mass is detected at
US or CT (96). Contrast-enhanced MRI seems to
OCP and Juxtapapillary Neoplasms have the advantage over CT because of its better
Slowly growing peripapillary tumors that cause soft-tissue contrast definition (Fig 14a, 14b).
obstruction and diffuse ductal dilatation may Dilatation of both the common bile duct and
mimic OCP. These tumors include benign (eg, the main pancreatic duct, or the double duct sign, is
adenoma) and malignant (eg, PDAC and acinar highly suggestive of an underlying neoplasm. Al-
carcinoma) neoplasms of the pancreas (93). though uncommon, inflammatory disease can pro-
RG  •  Volume 39  Number 7 Wolske et al  1979

Table 3: Differentiating among OCP, IPMN, and Juxtapapillary Neoplasms at Imaging

Imaging Features OCP IPMN Juxtapapillary Neoplasms

Gland atrophy Present or absent Present/absent Strongly present
Parenchymal calcification Strongly present Present/absent Present or absent
Solid enhancing mass Absent Absent Present
Double duct sign Present or absent Absent Strongly present
Protrusion of major papilla into duodenum Absent Strongly present Present or absent, with
enhancing mass
Nodular enhancement along the wall of the Absent Strongly present Absent
pancreatic duct
Cystic ectasia of the branch ducts Absent Strongly present Absent
Mucinous deposits within the ductal lumen Absent Strongly present Absent

Figure 14. Peripapillary neo-

plasm (adenoma) in a 42-year-old
man. (a) MR cholangiopancreato-
gram shows a dilated biliary tree, a
prominent pancreatic duct, and an
ill-defined filling defect in the distal
common bile duct (arrow). (b) Ax-
ial contrast-enhanced T1-weighted
MR image shows a heterogeneous
enhancing nodular lesion at the
papilla (arrow).

duce this finding, which mimics malignancy (Fig hypoattenuating at multidetector CT, hyperintense
15a). Stones located in the main duct outlet at the at T2-weighted MRI, and hypointense at T1-
papilla occasionally can be a nonmalignant cause of weighted MRI (4). Thick mucinous deposits (if
the double duct sign. In this case, MR cholangio- visualized) and papillary proliferations can appear
pancreatography is the most sensitive and specific similar at T2-weighted MRI (Fig 15a), as hy-
technique for diagnosis, and endoscopic retrograde pointense deposits along the duct wall. However,
cholangiopancreatography has a therapeutic role. at contrast-enhanced imaging, unlike mucinous
deposits, the papillary proliferations enhance and
OCP and IPMN can help confirm the imaging diagnosis of IPMN
Dilatation of the main pancreatic duct in the ab- (Fig 15b) (63). In addition to nodular enhance-
sence of an obstructing mass should suggest the ment along the duct wall, cystic ectasia of branch
diagnosis of main duct IPMN. Main duct IPMN ducts also favors the diagnosis of IPMN. Although
has malignant potential and may require surgery, branch ducts may be dilated in OCP, they do not
which makes it clinically important to distinguish usually have a cystic structure.
IPMN from OCP (4,97). Even in the absence of enhancing nodules along
the duct wall, protrusion of the major papilla into
Imaging of IPMN the duodenum (the “fish-eye” appearance classi-
At cross-sectional imaging, both OCP and IPMN cally described at endoscopy) is pathognomonic
may have atrophic changes in the pancreatic for IPMN (Fig 15c). Ampullary PDAC also can
parenchyma in addition to main duct dilatation. cause papillary protrusion into the duodenum but
Stones are frequently seen in OCP but are un- can be differentiated from IPMN by the presence
common in IPMN (98). Mucinous deposits in of an enhancing mass at the ampulla at cross-sec-
the dilated main pancreatic duct that are seen in tional imaging to avoid misdiagnosis.
patients with IPMN are usually homogenous and Optical coherence tomography permits an
have a similar appearance to that of ductal dilata- evaluation of tissue microstructures by means of
tion seen in patients with OCP: hypoechoic at US, a high-resolution probe introduced into the main
1980 November-December 2019 radiographics.rsna.org

Figure 15.  Main duct IPMN in a 53-year-old man. (a) Coronal

T2-weighted MR image shows hypointense filling defects in the
dilated main pancreatic duct (arrows). (b) Axial contrast-enhanced
T1-weighted MR image shows nodular mural enhancement along
the dilated main pancreatic duct (arrows). (c) Endoscopic image
through the papillary region shows the classic protrusion of the
major papilla into the duodenum (the “fish-eye” appearance).

pancreatic duct through an endoscopic retrograde

cholangiopancreatographic catheter. Optical coher-
ence tomography has shown high (almost 100%)
accuracy for detection of malignant tissue com-
pared with that of brush cytology (67.7%) (99).

Conclusion
Secondary imaging features may help to differ-
entiate inflammatory from neoplastic masses in
the pancreas. Imaging findings in correlation with Disclosures of Conflicts of Interest.—R.T. Activities related to
serum CA 19-9 levels can help to guide clinical the present article: disclosed no relevant relationships. Activities
care and appropriate surgical treatment. How- not related to the present article: consultancy for Behold AI. Other
activities: disclosed no relevant relationships.
ever, no imaging feature or combination of fea-
tures is 100% sensitive or specific for malignancy. References
The clinical and imaging features of neoplastic 1. American Society of Clinical Oncology. Pancreatic Cancer:
and nonneoplastic pancreatic masses overlap, Statistics. https://www.cancer.net/cancer-types/pancreatic-
and the entities can coexist in 1.8%–4% of cases. cancer/statistics. Published 2018. Updated Janusary 2019.
Accessed September 23, 2019.
Chronic inflammation of the pancreas is a risk 2. Zakaria HM, Mohamed A, Alsebaey A, Omar H, Elazab D,
factor for developing PDAC, and background Gaballa NK. Prognostic factors following pancreaticoduo-
denectomy for pancreatic ductal adenocarcinoma. Int Surg
inflammatory changes can mask underlying ma- J 2018;5(12):3887–3882.
lignancy and reduce the sensitivity and accuracy 3. Klein F, Jacob D, Bahra M, et al. Prognostic factors for
of endoscopic US-guided fine-needle aspiration. long-term survival in patients with ampullary carcinoma: the
results of a 15-year observation period after pancreaticoduo-
Further complicating image interpretation are denectomy. HPB Surg 2014;2014:970234.
cases of imaging-occult PDAC. 4. Guarise A, Faccioli N, Morana G, Megibow AJ. Chronic
In conclusion, the imaging features of inflam- Pancreatitis vs Pancreatic Tumors. In: Balthazar EJ, Megibow
AJ, Pozzi Mucelli R, eds. Imaging of the Pancreas. Berlin,
matory masses and PDAC at CT and MRI have Germany: Springer, 2009; 329–369.
substantial overlap because both entities cause 5. Kennedy T, Preczewski L, Stocker SJ, et al. Incidence of benign
extensive fibrotic change. Therefore, secondary inflammatory disease in patients undergoing Whipple proce-
dure for clinically suspected carcinoma: a single-institution
imaging findings such as the duct-penetrating sign, experience. Am J Surg 2006;191(3):437–441.
biliary or main pancreatic duct skip strictures, 6. Leung TK, Lee CM, Wang FC, Chen HC, Wang HJ. Dif-
ficulty with diagnosis of malignant pancreatic neoplasms
a capsulelike rim, the pancreatic duct-to-paren- coexisting with chronic pancreatitis. World J Gastroenterol
chyma ratio, displaced calcifications in chronic cal- 2005;11(32):5075–5078.
cific pancreatitis, the double duct sign, and vessel 7. Tajima Y, Kuroki T, Tsutsumi R, Isomoto I, Uetani M, Kane-
matsu T. Pancreatic carcinoma coexisting with chronic pancre-
encasement or displacement can provide valuable atitis versus tumor-forming pancreatitis: diagnostic utility of the
clues in the differential diagnosis of pancreatic time-signal intensity curve from dynamic contrast-enhanced
masses and may help in avoiding unnecessary MR imaging. World J Gastroenterol 2007;13(6):858–865.
8. Lalwani N, Mannelli L, Ganeshan DM, et al. Uncommon
surgical morbidity and mortality for inflammatory pancreatic tumors and pseudotumors. Abdom Imaging
conditions and delays in diagnosis of PDAC. 2015;40(1):167–180.
RG  •  Volume 39  Number 7 Wolske et al  1981

9. Nazli O, Bozdag AD, Tansug T, Kir R, Kaymak E. The a noise-optimized monoenergetic reconstruction algorithm.
diagnostic importance of CEA and CA 19-9 for the early Eur J Radiol 2015;84(11):2052–2058.
diagnosis of pancreatic carcinoma. Hepatogastroenterology 32. Kim SW, Kim SH, Lee DH, et al. Isolated Main Pancre-
2000;47(36):1750–1752. atic Duct Dilatation: CT Differentiation Between Benign
10. Ballehaninna UK, Chamberlain RS. Serum CA 19-9 as a and Malignant Causes. AJR Am J Roentgenol 2017;209
Biomarker for Pancreatic Cancer-A Comprehensive Review. (5):1046–1055.
Indian J Surg Oncol 2011;2(2):88–100. 33. Kim T, Murakami T, Takamura M, et al. Pancreatic mass
11. Xing H, Wang J, Wang Y, et al. Diagnostic Value of CA due to chronic pancreatitis: correlation of CT and MR imag-
19-9 and Carcinoembryonic Antigen for Pancreatic Can- ing features with pathologic findings. AJR Am J Roentgenol
cer: A Meta-Analysis. Gastroenterol Res Pract 2018 Nov 2001;177(2):367–371.
21;2018:8704751. 34. Shinozaki M, Saisho H, Tokinaga K, et al. Tumor-forming
12. Parsons L Jr, Palmer CH. How accurate is fine-needle pancreatitis on ultrasonography: follow-up study of the
biopsy in malignant neoplasia of the pancreas? Arch Surg images and relation to clinical features [in Japanese] Jpn J
1989;124(6):681–683. Med Ultrason 1987;14(3):189–198. https://www.jsum.or.jp/
13. Rodriguez J, Kasberg C, Nipper M, Schoolar J, Riggs MW, journals/1364.
Dyck WP. CT-guided needle biopsy of the pancreas: a retro- 35. Liao Q, Zhao YP, Wu WW, Li BL, Li JY. Diagnosis and
spective analysis of diagnostic accuracy. Am J Gastroenterol treatment of chronic pancreatitis. Hepatobiliary Pancreat
1992;87(11):1610–1613. Dis Int 2003;2(3):445–448.
14. Lewitowicz P, Matykiewicz J, Heciak J, et al. Percutaneous 36. Luetmer PH, Stephens DH, Ward EM. Chronic pan-
fine needle biopsy in pancreatic tumors: a study of 42 cases. creatitis: reassessment with current CT. Radiology
Gastroenterol Res Pract 2012;2012:908963. 1989;171(2):353–357.
15. Fritscher-Ravens A, Brand L, Knöfel WT, et al. Compari- 37. Choueiri NE, Balci NC, Alkaade S, Burton FR. Advanced
son of endoscopic ultrasound-guided fine needle aspira- imaging of chronic pancreatitis. Curr Gastroenterol Rep
tion for focal pancreatic lesions in patients with normal 2010;12(2):114–120.
parenchyma and chronic pancreatitis. Am J Gastroenterol 38. Karasawa E, Goldberg HI, Moss AA, Federle MP, London
2002;97(11):2768–2775. SS. CT pancreatogram in carcinoma of the pancreas and
16. Frampas E, Morla O, Regenet N, Eugène T, Dupas B, chronic pancreatitis. Radiology 1983;148(2):489–493.
Meurette G. A solid pancreatic mass: tumour or inflamma- 39. Ichikawa T, Sou H, Araki T, et al. Duct-penetrating sign
tion? Diagn Interv Imaging 2013;94(7-8):741–755. at MRCP: usefulness for differentiating inflammatory
17. Tamada T, Ito K, Kanomata N, et al. Pancreatic adenocarci- pancreatic mass from pancreatic carcinomas. Radiology
nomas without secondary signs on multiphasic multidetector 2001;221(1):107–116.
CT: association with clinical and histopathologic features. 40. Eloubeidi MA, Luz LP, Tamhane A, Khan M, Buxbaum
Eur Radiol 2016;26(3):646–655. JL. Ratio of pancreatic duct caliber to width of pancreatic
18. Freeny PC, Marks WM, Ryan JA, Traverso LW. Pancreatic gland by endosonography is predictive of pancreatic cancer.
ductal adenocarcinoma: diagnosis and staging with dynamic Pancreas 2013;42(4):670–679.
CT. Radiology 1988;166(1 Pt 1):125–133. 41. Lin E, Alexander D. Focal Chronic Pancreatitis versus Pan-
19. Tamm EP, Bhosale PR, Vikram R, de Almeida Marcal LP, creatic Cancer. In: Lin EC, Escott EJ, Garg KD, Bleicher
Balachandran A. Imaging of pancreatic ductal adenocarci- AG, Alexander D, eds. Practical Differential Diagnosis for
noma: State of the art. World J Radiol 2013;5(3):98–105. CT and MRI. New York, NY: Thieme, 2008; 212.
20. Chandarana H, Babb J, Macari M. Signal characteristic and 42. Busireddy KK, AlObaidy M, Ramalho M, et al. Pancreatitis-
enhancement patterns of pancreatic adenocarcinoma: evalu- imaging approach. World J Gastrointest Pathophysiol
ation with dynamic gadolinium enhanced MRI. Clin Radiol 2014;5(3):252–270.
2007;62(9):876–883. 43. Hough TJ, Raptopoulos V, Siewert B, Matthews JB.
21. Park MJ, Kim YK, Choi SY, Rhim H, Lee WJ, Choi D. Preop- Teardrop superior mesenteric vein: CT sign for unresect-
erative detection of small pancreatic carcinoma: value of adding able carcinoma of the pancreas. AJR Am J Roentgenol
diffusion-weighted imaging to conventional MR imaging for 1999;173(6):1509–1512.
improving confidence level. Radiology 2014;273(2):433–443. 44. Elmas N. The role of diagnostic radiology in pancreatitis.
22. Lee JH, Min JH, Kim YK, et al. Usefulness of non-contrast MR Eur J Radiol 2001;38(2):120–132.
imaging in distinguishing pancreatic ductal adenocarcinoma 45. Elmas N, Oran I, Oyar O, Ozer H. A new criterion in dif-
from focal pancreatitis. Clin Imaging 2019;55:132–139. ferentiation of pancreatitis and pancreatic carcinoma: artery-
23. Kartalis N, Lindholm TL, Aspelin P, Permert J, Albiin N. to-vein ratio using the superior mesenteric vessels. Abdom
Diffusion-weighted magnetic resonance imaging of pancreas Imaging 1996;21(4):331–333.
tumours. Eur Radiol 2009;19(8):1981–1990. 46. Graziani R, Tapparelli M, Malagò R, et al. The various imaging
24. Ichikawa T, Erturk SM, Motosugi U, et al. High-b value aspects of chronic pancreatitis. JOP 2005;6(1,Suppl):73–88.
diffusion-weighted MRI for detecting pancreatic ad- 47. Horiguchi S, Kamisawa T. Major duodenal papilla and its
enocarcinoma: preliminary results. AJR Am J Roentgenol normal anatomy. Dig Surg 2010;27(2):90–93.
2007;188(2):409–414. 48. Ahualli J. The double duct sign. Radiology 2007;
25. Wang Y, Miller FH, Chen ZE, et al. Diffusion-weighted MR 244(1):314–315.
imaging of solid and cystic lesions of the pancreas. Radio- 49. Oterdoom LH, van Weyenberg SJ, de Boer NK. Double-duct
Graphics 2011;31(3):E47–E64. sign: do not forget the gallstones. J Gastrointestin Liver Dis
26. Sahani DV, Kalva SP, Farrell J, et al. Autoimmune pancre- 2013;22(4):447–450.
atitis: imaging features. Radiology 2004;233(2):345–352. 50. Fulcher AS, Turner MAMR. MR pancreatography: a use-
27. Ishigami K, Yoshimitsu K, Irie H, et al. Diagnostic value of ful tool for evaluating pancreatic disorders. RadioGraphics
the delayed phase image for iso-attenuating pancreatic carci- 1999;19(1):5–24; discussion 41–44; quiz 148–149.
nomas in the pancreatic parenchymal phase on multidetector 51. Ohara H, Nakazawa T, Sano H, et al. Systemic extrapancreatic
computed tomography. Eur J Radiol 2009;69(1):139–146. lesions associated with autoimmune pancreatitis. Pancreas
28. Prokesch RW, Chow LC, Beaulieu CF, Bammer R, Jeffrey RB 2005;31(3):232–237.
Jr. Isoattenuating pancreatic adenocarcinoma at multi-detec- 52. Maringhini A, Ciambra M, Raimondo M, et al. Clinical pre-
tor row CT: secondary signs. Radiology 2002;224(3):764–768. sentation and ultrasonography in the diagnosis of pancreatic
29. Kim JH, Park SH, Yu ES, et al. Visually isoattenuating pan- cancer. Pancreas 1993;8(2):146–150.
creatic adenocarcinoma at dynamic-enhanced CT: frequency, 53. Karlson BM, Ekbom A, Lindgren PG, Källskog V, Rastad J.
clinical and pathologic characteristics, and diagnosis at imaging Abdominal US for diagnosis of pancreatic tumor: prospective
examinations. Radiology 2010;257(1):87–96. cohort analysis. Radiology 1999;213(1):107–111.
30. Beer L, Toepker M, Ba-Ssalamah A, et al. Objective and sub- 54. Saftoiu A, Vilmann P, Bhutani MS. The role of contrast-
jective comparison of virtual monoenergetic vs. polychromatic enhanced endoscopic ultrasound in pancreatic adenocarci-
images in patients with pancreatic ductal adenocarcinoma. noma. Endosc Ultrasound 2016;5(6):368–372.
Eur Radiol 2019;29(7):3617–3625. 55. D’Onofrio M, Biagioli E, Gerardi C, et al. Diagnostic per-
31. Frellesen C, Fessler F, Hardie AD, et al. Dual-energy CT of formance of contrast-enhanced ultrasound (CEUS) and
the pancreas: improved carcinoma-to-pancreas contrast with contrast-enhanced endoscopic ultrasound (ECEUS) for the
1982 November-December 2019 radiographics.rsna.org

differentiation of pancreatic lesions: a systematic review and diagnosis between malignant and benign conditions at
meta-analysis. Ultraschall Med 2014;35(6):515–521. secretin-enhanced MRCP. Clin Imaging 2017;41:137–143.
56. Seicean A, Badea R, Moldovan-Pop A, et al. Harmonic 78. Manfredi R, Frulloni L, Mantovani W, Bonatti M, Graziani
Contrast-Enhanced Endoscopic Ultrasonography for the R, Pozzi Mucelli R. Autoimmune pancreatitis: pancreatic and
Guidance of Fine-Needle Aspiration in Solid Pancreatic extrapancreatic MR imaging-MR cholangiopancreatography
Masses. Ultraschall Med 2017;38(2):174–182. findings at diagnosis, after steroid therapy, and at recurrence.
57. Dyrla P, Gil J, Niemczyk S, Saracyn M, Kosik K, Czarkowski Radiology 2011;260(2):428–436.
S, Lubas A. Elastography in the Diagnosis of Pancreatic 79. Donati F, Boraschi P, Cervelli R, et al. 3 T MR perfu-
Malignancies. Adv Exp Med Biol 2019;1133:41–48. sion of solid pancreatic lesions using dynamic contrast-
58. Facciorusso A, Martina M, Buccino RV, Nacchiero MC, enhanced DISCO sequence: Usefulness of qualitative and
Muscatiello N. Diagnostic accuracy of fine-needle aspiration quantitative analyses in a pilot study. Magn Reson Imaging
of solid pancreatic lesions guided by endoscopic ultrasound 2019;59:105–113.
elastography. Ann Gastroenterol 2018;31(4):513–518. 80. Kim JH, Lee JM, Park JH, et al. Solid pancreatic lesions:
59. Siddiqui N, Vendrami CL, Chatterjee A, Miller FH. Advanced characterization by using timing bolus dynamic contrast-
MR Imaging Techniques for Pancreas Imaging. Magn Reson enhanced MR imaging assessment--a preliminary study.
Imaging Clin N Am 2018;26(3):323–344. Radiology 2013;266(1):185–196.
60. Manfredi R, Pozzi Mucelli R. Secretin-enhanced MR Imaging 81. Kalb B, Martin DR, Sarmiento JM, et al. Paraduodenal pan-
of the Pancreas. Radiology 2016;279(1):29–43. creatitis: clinical performance of MR imaging in distinguishing
61. Raman SP, Horton KM, Fishman EK. Multimodality imaging from carcinoma. Radiology 2013;269(2):475–481.
of pancreatic cancer-computed tomography, magnetic reso- 82. Muraki T, Kim GE, Reid MD, et al. Paraduodenal Pan-
nance imaging, and positron emission tomography. Cancer creatitis: Imaging and Pathologic Correlation of 47 Cases
J 2012;18(6):511–522. Elucidates Distinct Subtypes and the Factors Involved in its
62. Liu Y, Wang M, Ji R, Cang L, Gao F, Shi Y. Differentiation Etiopathogenesis. Am J Surg Pathol 2017;41(10):1347–1363.
of pancreatic ductal adenocarcinoma from inflammatory mass: 83. Claudon M, Verain AL, Bigard MA, et al. Cyst formation in
added value of magnetic resonance elastography. Clin Radiol gastric heterotopic pancreas: report of two cases. Radiology
2018;73(10):865–872. 1988;169(3):659–660.
63. Hammami M, Noomen F, Toumi O, et al. Autoimmune 84. Mohl W, Hero-Gross R, Feifel G, et al. Groove pancreatitis:
pancreatitis mimicking pancreatic cancer. N Am J Med Sci an important differential diagnosis to malignant stenosis of
2011;3(11):520–523. the duodenum. Dig Dis Sci 2001;46(5):1034–1038.
64. Ghazale A, Chari ST, Smyrk TC, et al. Value of serum 85. Procacci C, Graziani R, Zamboni G, et al. Cystic dystro-
IgG4 in the diagnosis of autoimmune pancreatitis and in phy of the duodenal wall: radiologic findings. Radiology
distinguishing it from pancreatic cancer. Am J Gastroenterol 1997;205(3):741–747.
2007;102(8):1646–1653. 86. Arvanitakis M, Rigaux J, Toussaint E, et al. Endotherapy for
65. Kamisawa T, Egawa N, Nakajima H, Tsuruta K, Okamoto paraduodenal pancreatitis: a large retrospective case series.
A, Kamata N. Clinical difficulties in the differentiation of Endoscopy 2014;46(7):580–587.
autoimmune pancreatitis and pancreatic carcinoma. Am J 87. Casetti L, Bassi C, Salvia R, et al. “Paraduodenal” pancreatitis:
Gastroenterol 2003;98(12):2694–2699. results of surgery on 58 consecutives patients from a single
66. Okazaki K, Chiba T. Autoimmune related pancreatitis. Gut institution. World J Surg 2009;33(12):2664–2669.
2002;51(1):1–4. 88. Mittal PK, Harri P, Nandwana S, et al. Paraduodenal pancre-
67. Zamboni G, Lüttges J, Capelli P, et al. Histopathological atitis: benign and malignant mimics at MRI. Abdom Radiol
features of diagnostic and clinical relevance in autoimmune (NY) 2017;42(11):2652–2674.
pancreatitis: a study on 53 resection specimens and 9 biopsy 89. Gabata T, Matsui O, Kadoya M, et al. Small pancreatic adeno-
specimens. Virchows Arch 2004;445(6):552–563. carcinomas: efficacy of MR imaging with fat suppression and
68. Zhang L, Chari S, Smyrk TC, et al. Autoimmune pancre- gadolinium enhancement. Radiology 1994;193(3):683–688.
atitis (AIP) type 1 and type 2: an international consensus 90. Gabata T, Kadoya M, Terayama N, Sanada J, Kobayashi S,
study on histopathologic diagnostic criteria. Pancreas Matsui O. Groove pancreatic carcinomas: radiological and
2011;40(8):1172–1179. pathological findings. Eur Radiol 2003;13(7):1679–1684.
69. Madhani K, Farrell JJ. Management of Autoimmune Pancre- 91. Chamokova B, Bastati N, Poetter-Lang S, et al. The clini-
atitis. Gastrointest Endosc Clin N Am 2018;28(4):493–519. cal value of secretin-enhanced MRCP in the functional and
70. Lee LK, Sahani DV. Autoimmune pancreatitis in the context morphological assessment of pancreatic diseases. Br J Radiol
of IgG4-related disease: review of imaging findings. World J 2018;91(1084):20170677.
Gastroenterol 2014;20(41):15177–15189. 92. Witt H, Apte MV, Keim V, Wilson JS. Chronic pancreatitis:
71. Learn PA, Grossman EB, Do RK, et al. Pitfalls in avoid- challenges and advances in pathogenesis, genetics, diagnosis,
ing operation for autoimmune pancreatitis. Surgery and therapy. Gastroenterology 2007;132(4):1557–1573.
2011;150(5):968–974. 93. Lévy P, Ruszniewski P. Chronic pancreatitis [in French]. Rev
72. Abraham SC, Wilentz RE, Yeo CJ, et al. Pancreaticoduo- Prat 2002;52(9):997–1000.
denectomy (Whipple resections) in patients without malig- 94. Minniti S, Bruno C, Biasiutti C, et al. Sonography versus
nancy: are they all ‘chronic pancreatitis’? Am J Surg Pathol helical CT in identification and staging of pancreatic ductal
2003;27(1):110–120. adenocarcinoma. J Clin Ultrasound 2003;31(4):175–182.
73. Hur BY, Lee JM, Lee JE, et al. Magnetic resonance imaging 95. Yasuda K, Mukai H, Nakajima M. Endoscopic ultrasonog-
findings of the mass-forming type of autoimmune pancreatitis: raphy diagnosis of pancreatic cancer. Gastrointest Endosc
comparison with pancreatic adenocarcinoma. J Magn Reson Clin N Am 1995;5(4):699–712.
Imaging 2012;36(1):188–197. 96. Gerstenmaier JF, Malone DE. Mass lesions in chronic pan-
74. Bennett GL, Hann LE. Pancreatic ultrasonography. Surg creatitis: benign or malignant? An “evidence-based practice”
Clin North Am 2001;81(2):259–281. approach. Abdom Imaging 2011;36(5):569–577.
75. Menges M, Lerch MM, Zeitz M. The double duct sign in 97. Tenner S, Carr-Locke DL, Banks PA, et al. Intraductal
patients with malignant and benign pancreatic lesions. Gas- mucin-hypersecreting neoplasm “mucinous ductal ectasia”:
trointest Endosc 2000;52(1):74–77. endoscopic recognition and management. Am J Gastroenterol
76. Negrelli R, Manfredi R, Pedrinolla B, et al. Pancreatic duct 1996;91(12):2548–2554.
abnormalities in focal autoimmune pancreatitis: MR/MRCP 98. Ito K, Koike S, Matsunaga N. MR imaging of pancreatic
imaging findings. Eur Radiol 2015;25(2):359–367. diseases. Eur J Radiol 2001;38(2):78–93.
77. Boninsegna E, Manfredi R, Negrelli R, Avesani G, Mehrabi 99. Perkins JD. Optical coherence tomography: expanding use
S, Pozzi Mucelli R. Pancreatic duct stenosis: Differential in the bile duct. Liver Transpl 2007;13(5):765–768.

TM
This journal-based SA-CME activity has been approved for AMA PRA Category 1 Credit . See rsna.org/learning-center-rg.