Received: 12 March 2019 Revised: 22 April 2019 Accepted: 16 July 2019

DOI: 10.1111/1751-2980.12801

ORIGINAL ARTICLE

Pancreatic fat content may increase the risk of imaging

progression in low-risk branch duct intraductal papillary
mucinous neoplasm

Kazuhiro Kashiwagi1 | Kazuhiro Minami2 | Takashi Seino1 | Kenro Hirata2 |

Eisuke Iwasaki2 | Nagamu Inoue1 | Yasushi Iwao1 | Takanori Kanai2

1
Center for Preventive Medicine, Keio
University Hospital, Tokyo, Japan
2
Department of Internal Medicine, Division of
Gastroenterology and Hepatology,
Department of Internal Medicine, Keio
University School of Medicine, Tokyo, Japan
Correspondence
Kazuhiro Kashiwagi, Center for Preventive
Medicine, Keio University Hospital,
Shinanomachi 35, Shinjyuku-ku, Tokyo, Japan.
Email: kashiwagi.z2@keio.jp
Funding information
The authors received no specific funding for
this work.
Objective: To identify risk factors of imaging progression (increase in cyst size or
main pancreatic duct size, or a new mural nodule) in low-risk branch duct intraductal
papillary mucinous neoplasm (BD-IPMN), including obesity-related factors such as
pancreatic fat content.
Methods: Our hospital databases were searched for patients who had completed health
checkup, including upper abdominal magnetic resonance imaging (MRI) over 48 months
(August 2012 to July 2016). Individuals with BD-IPMN without worrisome features and
high-risk stigmata who underwent surveillance with at least one follow-up MRI,
irrespective of the follow-up period, were included. Pancreatic computed tomography
attenuation indexes were defined as the difference between the pancreas and spleen
attenuation (P - S) and the pancreas to spleen attenuation ratio (P/S).
Results: Among 75 patients diagnosed as having low-risk BD-IPMN, during a median
follow-up of 36 months, 11 (15%) had imaging progression in cyst size, including two
with worrisome features. A multivariate logistic analysis showed that the initial cyst size
and both indexes (P - S, or P/S) were significantly associated with imaging progression in
IPMN, respectively (Model 1: odds ratio [OR] 1.188, 95% confidence interval
[CI] 1.060-1.331, P = 0.003; OR 0.871, 95% CI 0.776-0.977, P = 0.019; Model 2:
OR 1.186, 95% CI 1.064-1.322, P = 0.002; OR 0.002, 95% CI 0.000-0.970, P = 0.049).
Conclusions: Pancreatic fat content and initial cyst size were significantly associated
with imaging progression in low-risk BD-IPMN. Revisions of international consensus
Fukuoka guidelines might be customized based on initial cyst size and pancreatic fat
content.

KEYWORDS
computed tomography attenuation index, imaging progression, initial cyst size, pancreatic fat
content, pancreatic intraductal neoplasms, pancreatic intraductal papillary mucinous
neoplasms

© 2019 Chinese Medical Association Shanghai Branch, Chinese Society of Gastroenterology, Renji Hospital Affiliated to Shanghai Jiaotong University School of
Medicine and John Wiley & Sons Australia, Ltd

J Dig Dis. 2019;20:557–562. wileyonlinelibrary.com/journal/cdd 557

558
1 | I N T RO D UC T I O N
Intraductal papillary mucinous neoplasms (IPMN) can present with high-
grade dysplasia or cancer, and are at a risk of developing concomitant
pancreatic cancer.1 Branch duct (BD)-IPMN have been demonstrated to
have a lower malignancy rate (25%) than main duct (MD)-IPMN or mixed
IPMN.2,3 Revisions to the Fukuoka guidelines for the management of
BD-IPMN recommend that patients with high-risk stigmata should
undergo surgery and those with worrisome features should undergo an
endoscopic ultrasound (EUS).2 For the patients having BD-IPMN but
without high-risk stigmata or worrisome features (low-risk BD-IPMN),
who comprise most cases of IPMN, management may depend on the ini-
2,4-9
tial cyst size. Several studies have reported factors that are associated
with malignant IPMN, which include features of the cyst (size >3 cm, the
presence of a mural nodule, main pancreatic duct dilatation, and growth
rate in size) as well as patients' factors (symptoms, presence of diabetes
mellitus, or smoking status). Although the main concern of IPMN is the
10
onset of malignancy, Crippa et al recently reported that 52% of the
281 patients, including 159 with BD-IPMN and high-risk stigmata or
worrisome features, had imaging progression such as an increase in cyst
size, main pancreatic duct (MPD) dilatation, new nodules, or presence of
new or recurrent symptoms during a median follow-up period of
51 months.
Pancreatic steatosis is a common condition with a prevalence of
16%-31% in Asian populations.11,12 Hori et al13 observed that fatty
infiltration in the pancreas might be independently associated with an
increased risk of developing pancreatic cancer. A recent study also
showed that the presence of pancreatic intraepithelial neoplasia was
associated with intralobular fatty infiltration (odds ratio [OR] 17.86,
95% confidence interval [CI] 4.94-88.12, P < 0.01), independent of body
mass index (BMI) or visceral adipose tissues.14 Additionally, a recent sys-
tematic review demonstrated that pancreatic steatosis was associated
with a significant increase in the risk of metabolic syndrome (OR 2.37,
95% CI 2.07-2.71, P < 0.0001), hypertension (OR 1.67, 95% CI
1.32-2.10, P < 0.0001), and diabetes mellitus (OR 2.08, 95% CI
1.44-3.00, P = 0.0001).15 Whether obesity-related factors, including BMI
and visceral adipose tissue, or related metabolic complications are associ-
ated with imaging progression in low-risk BD-IPMN has been investi-
gated in few studies. Only one study16 has reported the relationship
between life-style habits and imaging progression in low-risk IPMN,
although pancreatic cancer has been proven to be associated with
17
smoking as well as obesity and inactivity.
In this study, we aimed to identify factors associated with imaging
progression of low-risk BD-IPMN, including obesity-related factors,
such as pancreatic fat content as measured by the computed tomog-
raphy (CT) attenuation indexes in the individuals who undertook a
medical checkup.
2 | PATI ENT S AND M ET HODS
The collection of registry data was approved by the Institutional
Review Board at Keio University Hospital (no. 20160398).
KASHIWAGI ET AL.
2.1 | Study design and participant enrollment
This was a retrospective study of patients diagnosed as having IPMN after
a health checkup, including an initial upper abdominal magnetic resonance
imaging (MRI), with magnetic resonance cholangiopancreatography
(MRCP) as an optional examination, at our hospital from August 2012 to
July 2016, as reported previously.18 Patients with BD-IPMN but without
worrisome features or high-risk stigmata who had undergone surveillance
with at least one follow-up MRI examination, irrespective of their follow-
up period, were included. The incidence of the two outcomes was
assessed over the entire follow-up period by comparing the images of the
initial MRI with those of the last follow-up MRI, in accordance with a
report by Girometti et al19: (a) imaging progression, defined as any increase
in number, size, or appearance, including an increase in the size of the
MPD and/or the occurrence of a mural nodule in ≥1 follow-up examina-
tion; and (b) alert findings, defined as the occurrence of worrisome fea-
tures and/or high-risk stigmata. Therefore, imaging progression included
cases of alert findings.
2.2 | Measurement of visceral adipose tissue and
pancreatic fat content by CT
Non-enhanced CT for screening chest lesions was performed with a
single scan at the umbilicus level to measure visceral adipose tissues.
We also used the CT scan for the examination of the pancreas to esti-
mate the average fat content of the pancreas. A region of interest of
about 1.0 cm2 was selected from the pancreatic head, body, and tail,
respectively, according to the method described by Kim et al.20 The
density of pancreatic steatosis was expressed by using Hounsfield
units (HU). Pancreatic CT attenuation indexes were defined as the dif-
ference between pancreatic and the splenic attenuation (P - S) and
the pancreas to spleen attenuation ratio (P/S), using that of the spleen
as an internal control. Figure 1A shows CT imaging of a 73-year-old
man with low-risk BD-IPMN at initial diagnosis. Two circles in the CT
image indicate regions of interest in the body of the pancreas (35.1
HU) and the spleen (49.5 HU), respectively. Thus, P - S and P/S were
calculated as −14.4 and 0.71, respectively.
2.3 | Data collection
After fasting overnight, all participants underwent a peripheral blood test
including items related to glucose and lipid levels. Their height, waist cir-
cumference (measured at the umbilical level) and body weight were mea-
sured and recorded, and BMI was calculated. Metabolic syndrome was
diagnosed according to the Japanese criteria,21 including the presence of
central obesity (defined as a waist circumference ≥85 cm for men
or ≥90 cm for women). The participants were classified into non-smoker,
past smoker and current smoker based on their cigarette smoking status.
Habitual drinking was defined as alcohol intake >20 g daily for men or
>10 g for women.22 Habitual exercise was defined as exercise for at least
30 minutes per time and ≥2 per week.
KASHIWAGI ET AL. 559

2.4 | Statistical analysis
Statistical analyses were performed by using the SPSS version 25.0
for Windows (IBM, Armonk, NY, USA). Continuous variables were
expressed as mean ± standard deviation or median and range, and
were analyzed using the Student's t-test or the Mann-Whitney U test
when appropriate. While categorical variables were expressed as
numbers and percentages and were analyzed using the Fisher's exact
test. Statistical significance was set at a two-sided P value of less than
0.05. A multivariate logistic regression analysis was performed to
determine whether the CT attenuation indexes, obesity-related
factors, or lifestyle, were associated with imaging progression in low-
risk BD-IPMN, and OR and 95% CI were calculated.
3 | RESULTS
3.1 | Overall view of imaging progression in patients
with low-risk IPMN
Altogether 98 patients were diagnosed with IPMN. After excluding two
with worrisome features and 21 without follow-up MRI results,
75 patients with low-risk BD-IPMN were included in the final analysis.
None of the patients had a history of pancreatic disease including chronic
pancreatitis. Their mean age was 67.6 ± 9.1 years, and 53 (70.7%) were
men and 17 (22.7%) were found to have metabolic syndrome. The
median period of surveillance was 36 months (range 6-67 mo).
Among the 75 participants, imaging progression was observed in
11 (14.7%) with an increase in cyst size, including one participant with an
increase in the size of the MPD over the entire follow-up period (mean:
43.9 mo; median: 46 mo). The characteristics of these eleven patients are
summarized in Table 1. Worrisome features were found in only two (2.7%)
out of all the 75 individuals. All subjects undertook EUS when the size of the
cyst grew more than 20 mm during the follow-up period. However, none
underwent surgery or died of pancreatic disease. Figure 1B shows a repre-
sentative initial MRCP image of BD-IPMN at the pancreatic head, which was
initially 28 mm in maximum diameter but grew to 35 mm after 38 months,
which then progressed to IPMN with worrisome features (Figure 1C).

F I G U R E 1 Low-risk intraductal papillary mucinous neoplasm
(IPMN) at the pancreatic head at initial diagnosis in a 73-year-old
man. A, The density of steatosis of (1) the pancreatic body (35.1 HU)
and (2) the spleen (49.5 HU), respectively. Thus, pancreas and spleen
attenuation index, and the pancreas and spleen attenuation ratio are
calculated as −14.4 and 0.71, respectively. B, The initial magnetic
resonance cholangiopancreatography imaging shows branch duct-
IPMN at the pancreatic head, with an initial maximum diameter of
28 mm. C, It increased to 35 mm after 38 months, which then
progressed to IPMN with worrisome features
3.2 | Clinical characteristics of the imaging
progression and non-imaging progression groups
A univariate analysis was performed for the comparison in the clinical
characteristics between the imaging progression and non-imaging pro-
gression groups (Table 2). No significant differences were identified in
the mean follow-up period (43.9 ± 15.3 mo vs 36.5 ± 14.7 mo,
P = 0.127), the presence of obesity-related factors, or that of metabolic-
related complications between the imaging progression and the non-

TABLE 1 Overall view of imaging progression in low-risk branch duct intraductal papillary mucinous neoplasm

Age (y)/sex Cyst size (mm)/location at baseline MRI Follow-up (mo) Cyst size at a follow-up MRI (mm)
80/M 13/pancreatic body 34 19
78/M 15/pancreatic head 30 25
66/M 23/pancreatic head 60 25
74/M 20/pancreatic head (pancreatic duct caliber: 4 mm) 67 28 (6)
66/F 16/pancreatic body 34 20
82/M 16/pancreatic head 46 18
73/F 15/pancreatic head 15 20
62/M 10/pancreatic head 48 15
73/M 28/pancreatic head 40 35
70/M 14/pancreatic tail 60 17
75/M 9/pancreatic tail 49 12

Abbreviations: F, female; M, male; MRI, magnetic resonance imaging.

560 KASHIWAGI ET AL.

TABLE 2 Univariate analysis of clinical characteristics in imaging progression group and non-progression group

Progression Non-progression
Characteristics (n = 11) (n = 64) P value
Age, y (mean ± SD) 72.6 ± 6.2 66.7 ± 9.3 0.046a
Male sex, n (%) 9 (81.8) 44 (68.8) 0.491b
2
BMI, kg/m (mean ± SD) 23.2 ± 2.3 22.7 ± 2.8 0.577a
Visceral adipose tissue, cm2 (mean ± SD) 118.1 ± 51.8 98.6 ± 41.2 0.167a
Metabolic syndrome, n (%) 4 (36.4) 13 (20.3) 0.211b
Hypertension, n (%) 5 (45.5) 13 (20.3) 0.082b
Diabetes mellitus, n (%) 4 (36.4) 9 (14.1) 0.090b
Hyperlipidemia, n (%) 6 (54.5) 17 (26.6) 0.070b
Fatty liver, n (%) 5 (45.5) 18 (28.1) 0.209b
CT index (mean ± SD) P-S −11.1 ± 7.4 −5.5 ± 6.5 0.012c
P/S 0.80 ± 0.10 0.87 ± 0.12 0.048a
High-density lipoprotein, mg/dL (mean ± SD) 50.8 ± 13.8 57.3 ± 15.1 0.187c
Low-density lipoprotein, mg/dL (mean ± SD) 104.1 ± 32.6 117.5 ± 29.0 0.168a
Triglyceride, mg/dL (mean ± SD) 99.6 ± 31.4 114.9 ± 67.5 0.464c
Lipoprotein(a), mg/dL (mean ± SD) 20.5 ± 18.8 19.0 ± 23.7 0.849c
Apolipoprotein-A1, mg/dL (mean ± SD) 137.7 ± 27.5 148.4 ± 20.3 0.131a
Apolipoprotein-B, mg/dL (mean ± SD) 79.3 ± 19.7 86.1 ± 18.9 0.272a
Fasting blood sugar, mg/dL (mean ± SD) 119.6 ± 26.8 108.9 ± 22.1 0.158c
Hemoglobin A1c, % (mean ± SD) 6.2 ± 1.0 5.9 ± 0.5 0.352c
Insulin, μg/mL (mean ± SD) 5.5 ± 2.3 5.3 ± 3.6 0.878c
HOMA-IR (mean ± SD) 1.6 ± 0.8 1.5 ± 1.2 0.697c
HOMA-β, % (mean ± SD) 40.0 ± 22.7 42.5 ± 24.4 0.758
p-amylase, U/L (mean ± SD) 35.6 ± 8.8 35.3 ± 10.3 0.933c
Non-smoker, n (%) 4 (36.4) 38 (59.4) 0.138b
Past smoker, n (%) 5 (45.5) 18 (28.1) 0.209b
Current smoker, n (%) 2 (18.2) 8 (12.5) 0.452b
Habitual drinker, n (%) 2 (18.2) 8 (12.5) 0.452b
Habitual exercise, n (%) 3 (27.3) 29 (45.3) 0.218b
Cyst size, mm (mean ± SD) 16.5 ± 6.0 9.3 ± 5.6 0.000c
Cyst location (head/body/tail), n (%) 6/3/2 (54.5/27.3/18.2) 29/23/12 (45.3/35.9/18.8) 0.910b
Follow-up, mo (mean ± SD) 43.9 ± 15.3 36.5 ± 14.7 0.127c
a
Student's t-test.
b
Fisher's exact test.
c
Mann-Whitney U test.
Bold text indicates statistical significance.
Habitual drinking was defined as alcohol intake >20 g daily for men and >10 g for women.
Habitual exercise was defined as exercise ≥30 min/d and ≥ twice/wk.
Abbreviations: BMI, body mass index; CT, computed tomography; HOMA-IR, homeostasis model assessment insulin resistance; P, pancreas; S, spleen;
P - S, the difference between the pancreas and spleen attenuation; P/S; pancreas-to-spleen attenuation ratio; SD, standard deviation.

imaging progression groups. Additionally, neither smoking status nor
habitual drinking was related to the progression of low-risk BD-IPMN.
On the other hand, in terms of mean age, the progression group was
significantly older than the non-progression group (72.6 ± 6.2 y vs
66.7 ± 9.3 y, P = 0.046) and the initial cyst size was larger in the former
group than in the latter (16.5 ± 6.0 mm vs 9.3 ± 5.6 mm, P = 0.000).
More interestingly, both pancreatic CT attenuation indexes (P - S and
P/S) were significantly lower in the progression group than in the non-
progression group (−11.1 ± 7.4 vs −5.5 ± 6.5, P = 0.012; and 0.80
± 0.10 vs 0.87 ± 0.12, P = 0.048; Figure S1). In the multivariate logistic
regression analysis using two models (Table 3), the initial cyst size (Model
1: OR 1.188, 95% CI 1.060-1.331, P = 0.003; Model 2: OR 1.186, 95%
CI 1.064-1.322, P = 0.002) and the indexes P - S (Model 1: OR 0.871,
95% CI 0.776-0.977, P = 0.019) or P/S (Model 2: OR 0.002, 95% CI
0.000-0.970, P = 0.049) had a significant relationship with imaging pro-
gression in low-risk BD-IPMN.
KASHIWAGI ET AL.
T A B L E 3 Multivariate logistic regression analysis of clinical
characteristics in imaging progression group and non-progression
group
Model 1
Clinical characteristics OR (95% CI) P value
Age (y) 1.000 0.159
Cyst size (mm) 1.188 (1.060–1.331) 0.003
P-S 0.871 (0.776–0.977) 0.019
Model 2
Clinical characteristics OR (95% CI) P value
Age (y) 1.000 0.115
Cyst size (mm) 1.186 (1.064–1.322) 0.002
P/S 0.002 (0.000–0.970) 0.049
Bold text indicates statistical significance.
CI, confidence interval; OR, odds ratio; P/S, pancreas-to-spleen attenuation
ratio; P - S, the difference between the pancreas and spleen attenuation.
4 | DISCUSSION
Revisions to the Fukuoka guidelines2 recommend that low-risk BD-IPMN
should be followed up by CT, MRI, and/or EUS at appropriate intervals,
based on the size of each initial cyst. The main concern is the develop-
ment of malignancy in the cyst and many studies have reported the natu-
ral history of malignant progression of low-risk BD-IPMN.2,4-9 However,
the majority of incidental IPMN in clinical setting are small in size and lack
currently defined high-risk features. A meta-analysis23 of 20 studies only
included patients with low-risk BD-IPMN showed that disease progres-
sion occurred only in 28% of them, mainly as an increase in cyst size
(21%) during an average observation period ranging from 29 to
77 months. Moreover, Han et al24 recently reported that worrisome fea-
tures were observed in only 209 (15.3%) patients with BD-IPMN during a
median observation period of 61 months, including cyst size ≥3 cm
(8.0%), cyst wall thickening (3.7%), MPD dilatation (5.6%), and a mural
nodule (3.1%). Approximately 15% of any size increase in our study is
lower than those in the two reports, which may be due to the smaller cyst
size at initial diagnosis and/or a shorter follow-up period in our cohort.
It is also essential to identify factors associated with disease progres-
sion to malignancy in BD-IPMN. Several studies2,4-9 have reported that
those factors included cystic features (cyst size, the presence of a mural
nodule, MPD dilatation, and growth rate in size) and patient's factors
(symptoms, diabetes mellitus, smoking, or obesity). However, few data are
available on risk factors of imaging progression in low-risk BD-IPMN.16,19
Gausman et al 16
concluded that an initial cyst size ≥2 cm, multifocality of
cystic lesions, the patient's past history of prostate cancer, and of smoking
could independently predict the onset of new worrisome features. How-
ever, Girometti et al19 reported that the clinical or baseline features of
images under MRCP did not predict imaging evolution or the development
of alert findings. The well-known risk factors for pancreatic cancer include
IPMN as well as smoking, diabetes mellitus, a family history of pancreatic
cancer and obesity.17 Fukuda et al25 reported that the density of fatty pan-
creas measured by CT had an independent association with pancreatic
cancer. Very recently, we reported that pancreatic fat content determined
561
by CT had a significant relationship with the presence of IPMN.18 We then
attempted to investigate whether there was a correlation between imag-
ing progression and obesity-related factors or metabolic-related complica-
tions, including pancreatic fat content, in asymptomatic subjects with low-
risk BD-IPMN. This study shows that obesity-related factors and
metabolic-related complications including diabetes mellitus had no associ-
ation with the imaging progression of BD-IPMN. Most significantly, the
pancreatic CT attenuation indexes (P - S and P/S) were independently
related to the imaging progression of BD-IPMN. These results suggest that
pancreatic steatosis may promote the development of malignant IPMN, in
the same way that non-alcoholic fatty liver diseases cause liver cancer in
pathophysiology. Our study also supports the notion that there is signifi-
cant association of initial cyst size with imaging progression of BD-IPMN,
which is in line with the current consensus guideline2 classifying low-risk
BD-IPMN into three categories by initial cyst size as one surveillance strat-
egy. Further classification by predictors of imaging progression, including
pancreatic fat content, may contribute to a more appropriate scheduling
of imaging follow-up of BD-IPMN, and the cost effectiveness of prolonged
follow-up procedures.
The main limitation of this study was the relatively small number
of participants enrolled and the mean follow-up period may be too
short to depict the natural history of BD-IPMN, because small
changes in the appearances (ie, any change in number or size) of inci-
dental BD-IPMN are not uncommon and occur over wide time inter-
vals, as shown in recent reports.9,20 Moreover, the number of MRI
and the timing of the follow-up were not uniform, as the design of this
study was retrospective. Finally, pancreatic steatosis was diagnosed
neither by pathological examination nor MRI-based techniques, which
are considered to be more accurate than CT.
In conclusion, initial size of the cyst and the pancreatic fat content
determined by CT were significantly related to imaging progression,
including alert findings in low-risk BD-IPMN. Pancreatic steatosis
itself may promote the progression of BD-IPMN, and the proposed
surveillance guideline could be based on not only the initial cyst size
but also on pancreatic fat content. Further prospective studies with
MRI are needed to determine whether pancreatic steatosis may
increase the number of new IPMN or the size of known IPMN, and
whether it may promote malignant IPMN.
ACKNOWLEDG MENT
We would like to thank Yoshinori Sugino, MD, PhD, and Suketaka
Momoshima, MD, PhD for their advices on imaging diagnosis.
CONFLIC T OF INT ER E ST
All the authors declare no conflict of interest.
OR CID
Kazuhiro Kashiwagi https://orcid.org/0000-0002-6301-1466
562
RE FE R ENC E S
1. Hruban RH, Maitra A, Kern SE, Goggins M. Precursors to pancreatic
cancer. Gastroenterol Clin North Am. 2007;36(4):831-849.
2. Tanaka M, Fernández-Del Castillo C, Kamisawa T, et al. Revisions
of international consensus Fukuoka guidelines for the manage-
ment of IPMN of the pancreas. Pancreatology. 2017;17(5):
738-753.
3. Fong ZV, Ferrone CR, Lillemoe KD, Fernández-Del Castillo C. Intra-
ductal papillary mucinous neoplasm of the pancreas: current state
of the art and ongoing controversies. Ann Surg. 2016;263(5):
908-917.
4. Anand N, Sampath K, Wu BU. Cyst features and risk of malignancy in
intraductal papillary mucinous neoplasms of the pancreas: a meta-
analysis. Clin Gastroenterol Hepatol. 2013;11(8):913-921.
5. Sultana A, Jackson R, Tim G, et al. What is the best way to identify
malignant transformation within pancreatic IPMN: a systematic
review and meta-analyses. Clin Transl Gastroenterol. 2015;6(12):e130.
https://doi.org/10.1038/ctg.2015.60.
6. Kang MJ, Jang JY, Kim SJ, et al. Cyst growth rate predicts malignancy
in patients with branch duct intraductal papillary mucinous neo-
plasms. Clin Gastroenterol Hepatol. 2011;9(1):87-93.
7. Ohno E, Hirooka Y, Kawashima H, et al. Natural history of pancreatic
cystic lesions: a multicenter prospective observational study for eval-
uating the risk of pancreatic cancer. J Gastroenterol Hepatol. 2018;33
(1):320-328.
8. Morales-Oyavide V, Mino-Kenudson M, Ferrone CR, et al. Diabetes
mellitus in intraductal papillary mucinous neoplasm of the pancreas is
associated with high-grade dysplasia and invasive carcinoma. Pan-
creatology. 2017;17(6):920-926.
9. Chang YT, Tien YW, Jeng YM, et al. Overweight increases the risk of
malignancy inpatients with pancreatic mucinous cystic neoplasms.
Medicine (Baltimore). 2015;94(20):e797. https://doi.org/10.1097/MD.
0000000000000797.
10. Crippa S, Bassi C, Salvia R, et al. Low progression of intraductal papil-
lary mucinous neoplasms with disturbing features and high-risk stig-
mata undergoing non-operative management: a mid-term follow-up
analysis. Gut. 2017;66(3):495-506.
11. Wang CY, Ou HY, Chen MF, Chang TC, Chang CJ. Enigmatic ectopic
fat: prevalence of nonalcoholic fatty pancreas disease and its associ-
ated factors in a Chinese population. J Am Heart Assoc. 2014;3(1):
e000297. https://doi.org/10.1161/JAHA.113.000297.
12. Zhou J, Li ML, Zhang DD, et al. The correlation between pancreatic
steatosis and metabolic syndrome in a Chinese population. Pan-
creatology. 2016;16(4):578-583.
13. Hori M, Takahashi M, Hiraoka N, et al. Association of pancre-
atic fatty infiltration with pancreatic ductal adenocarcinoma. Clin
Transl Gastroenterol. 2014;5(3):e53. https://doi.org/10.1038/ctg.
2014.5.
14. Rebours V, Gaujoux S, d'Assignies G, et al. Obesity and fatty pancre-
atic infiltration are risk factors for pancreatic precancerous lesions
(PanIN). Clin Cancer Res. 2015;21(15):3522-3528.
KASHIWAGI ET AL.
15. Singh RG, Yoon HD, Wu LM, Lu J, Plank LD, Petrov MS. Ectopic fat
accumulation in the pancreas and its clinical relevance: a systematic
review, meta-analysis, and meta-regression. Metabolism. 2017;69:1-13.
16. Gausman V, Kandel P, Van Riet PA, et al. Predictors of progression
among low-risk intraductal papillary mucinous neoplasms in a multi-
center surveillance cohort. Pancreas. 2018;47(4):471-476.
17. Ryan DP, Hong TS, Bardeesy N. Pancreatic adenocarcinoma. N Engl J
Med. 2014;371(11):1039-1049.
18. Kashiwagi K, Seino T, Fukuhara S, et al. Pancreatic fat content detected
by computed tomography and its significant relationship with intraductal
papillary mucinous neoplasm. Pancreas. 2018;47(9):1087-1092.
19. Girometti R, Pravisani R, Intini SG, et al. Evolution of incidental
branch-duct intraductal papillary mucinous neoplasms of the pan-
creas: a study with magnetic resonance imaging
cholangiopancreatography. World J Gastoenterol. 2016;22(43):9562-
9570.
20. Kim SY, Kim H, Cho JY, et al. Qualitative assessment of pancreatic fat
by using unenhanced CT: pathologic correlation and clinical implica-
tions. Radiology. 2014;271(1):104-112.
21. Yamagishi K, Iso H. The criteria for metabolic syndrome and the
national health screening and education system in Japan. Epidemiol
Health. 2017;39:e2017003. https://doi.org/10.4178/epih.e2017003.
22. Farrell GC, Chitturi S, Lau GK, Sollano JD. Guidelines for the assess-
ment and management of non-alcoholic fatty liver disease in the
Asia-Pacific region: executive summary. J Gastroenterol Hepatol.
2007;22(6):775-777.
23. Crippa S, Capurso G, Cammà C, Fave GD, Fernández-Del Castillo C,
Falconi M. Risk of pancreatic malignancy and mortality in branch-duct
IPMN undergoing surveillance: a systematic review and meta-analy-
sis. Dig Liver Dis. 2016;48(5):473-479.
24. Han Y, Lee H, Kang JS, et al. Progression of pancreatic branch duct
intraductal papillary mucinous neoplasm associates with cyst size.
Gastroenterology. 2018;154(3):576-584.
25. Fukuda Y, Yamada D, Eguchi H, et al. CT density in the pancreas is a
promising imaging predictor for pancreatic ductal adenocarcinoma.
Ann Surg Oncol. 2017;24(9):2762-2769.
SUPPORTING INF ORMATION
Additional supporting information may be found online in the
Supporting Information section at the end of this article.
How to cite this article: Kashiwagi K, Minami K, Seino T,
et al. Pancreatic fat content may increase the risk of imaging
progression in low-risk branch duct intraductal papillary
mucinous neoplasm. J Dig Dis. 2019;20:557–562. https://doi.
org/10.1111/1751-2980.12801