PRIMER

Japanese criteria if the patient meets at least two of the magnetic resonance cholangiopancreatography
four clinical symptoms listed and shows signs of early (MRCP), in which secretin is used to stimulate pan­
chronic pancreatitis on imaging. creatic fluid secretion107. The secretin test can directly
measure exocrine pancreatic function and is useful as
Exocrine pancreatic insufficiency a reference method to evaluate new tests. To answer
Exocrine pancreatic deficiency is defined as a decreased clinical questions, a non-invasive function test such as
secretion of digestive enzymes and bicarbonate by the the pancreatic elastase test (the measurement of enzyme
pancreas regardless of its cause and may be mild or in the faeces) can be used (TABLE 1). However, none of
severe. Exocrine pancreatic insufficiency implies that the non-invasive pancreatic function tests is sensitive
the reduced pancreatic enzyme secretion is insufficient enough to diagnose slight-to-moderate exocrine pancre­
to maintain the normal digestion of nutrients. The main atic insufficiency reliably, and these tests are generally
causes of exocrine pancreatic insufficiency in adults are unnecessary for advanced disease108.
chronic pancreatitis, pancreatic carcinoma and a previ­ In addition to exocrine pancreatic insufficiency,
ous pancreas resection. Typical symptoms of exocrine other causes of malnutrition are pain-related reduction
pancreatic insufficiency include abdominal cramps, of food intake, continued alcohol consumption and an
bloating, flatulence, steatorrhoea and malnutrition105. increased metabolic rate3. Some studies suggest impaired
In patients with alcoholic chronic pancreatitis, clin­ absorption of fat-soluble vitamins in patients with
ically manifest exocrine pancreatic insufficiency usually mild-to-moderate exocrine pancreatic insufficiency due
develops approximately 10–15 years after the onset of to chronic pancreatitis. Although the prevalence of vita­
symptoms; this interval might be even longer in patients min D deficiency is high in patients with chronic pan­
with early-onset idiopathic or hereditary chronic pan­ creatitis, no difference in vitamin D levels was observed
creatitis. The late manifestation of exocrine pancre­ compared with age-matched and sex-matched controls
atic insufficiency, despite the destruction of pancreatic in several studies103,109–111. The increased risk of osteo­
­tissue in the early stages of the disease, is explained by porotic fractures in patients with chronic pancreatitis
the large functional reserve capacity of the pancreas of is likely to be multifactorial and can only partially be
>90–95%106. attributed to exocrine pancreatic deficiency and reduced
In most patients, there is a correlation between the absorption of vitamin D112.
extent of morphological changes visualized via imaging
and reduced exocrine pancreatic function3. MRI-based Endocrine pancreatic insufficiency
techniques allow semi-quantitative parameters to assess Pancreatogenic (type 3c) diabetes in chronic pancreati­
exocrine pancreatic function by determining fluid secre­ tis is characterized by a deficiency of insulin secretion
tion into the duodenum during a secretin-­enhanced and other hormones secreted by the cells of the islets of
Langerhans101,113. The risk of hypoglycaemia is increased
in patients with pancreatogenic diabetes, resulting in
a b increased mortality. Episodic hypoglycaemia occurs
L in up to 79% of patients with pancreatogenic diabetes,
* and severe hypoglycaemia occurs in up to 41%. Median
L
survival is 8.7 years after the diagnosis of pancreato­
* genic diabetes. Thus, diagnosis and follow‑up monitor­
* ing of diabetes mellitus is important and should be
L
­performed according to international guidelines114.

* 2 mm 2 mm
Histopathology
The morphological features of chronic pancreatitis vary
c d depending on aetiology and stage of disease115 (TABLE 2).
Macroscopic features of the pancreas in advanced-stage
classic (alcoholic) chronic pancreatitis are an irregular
contour, firm consistency and loss of lobulation. Dilated
and irregular ducts that sometimes contain stones can
be observed. Pseudocysts can be present in the pancreas
and can even extend beyond the normal pancreatic
900 μm 600 μm border. These changes can be patchy and may mimic a
­neoplasm on imaging 115.
Figure 3 | Histological characteristics of alcoholic chronic pancreatitis. a | Early-stage The key histological features are interlobular, intra­
disease with well-preserved pancreatic parenchyma with Nature Reviews
moderate | Disease Primers
perilobular, lobular and periductal fibrosis, atrophy of the acinar
interlobular and periductal fibrosis, and focal intralobular fibrosis (lower left). Lobuli are
parenchyma and duct distortion with intraluminal pro­
represented by ‘L’ and ducts by asterisks. b | Histological image of a cystically dilated duct
(centre) with flattened epithelium, periductal (white arrows) and perilobular fibrosis tein plugs and/or calcifications; squamous metaplasia
(asterisk), and a blood vessel with a thickened wall (black arrow). c | End-stage chronic of the duct epithelium may be observed (FIG. 3). Sparse
pancreatitis with complete atrophy of the acinar cells, paucicellular fibrosis with lympho­c ytic infiltrates are present in the interstitial
prominent islets (arrows) and nerves (arrowheads). d | Large interlobular duct with tissue, and also surrounding often enlarged periph­
squamous metaplasia. eral nerves. Other characteristic features include the

8 | ARTICLE NUMBER 17060 | VOLUME 3 www.nature.com/nrdp

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
 PRIMER

a b c

200 mm 400 μm 400 μm

d e f

400 μm 300 μm 400 μm

Figure 4 | Histological characteristics of autoimmune pancreatitis. a | Massive obliteration of the lobular architecture
Nature Reviews | Disease Primers
with diffuse fibrosis, abundant inflammatory infiltrates and stenosis of a medium-sized duct (asterisk) in autoimmune
pancreatitis type 1. b | Detail of storiform fibrosis in autoimmune pancreatitis type 1. c | Core needle biopsy sample
of a patient with autoimmune pancreatitis type 1 showing massive lymphoplasmacellular infiltration and fibrosis.
d | Immunohistochemistry for CD138 showing abundant plasma cell infiltrates with periductal localization (arrows)
in autoimmune pancreatitis type 1. e | Immunohistochemistry for IgG4 showing abundant IgG4‑positive cells in
autoimmune pancreatitis type 1. f | Granulocytic epithelial lesions of an interlobular medium-sized duct with a ruptured
epithelium (arrow) in autoimmune pancreatitis type 2.

presence of irregular clusters of islets and the presence features of type 1, except for numbers of IgG4‑positive
of thick-walled vessels. Owing to its recurrent nature, plasma cells. Notably, although it is known as auto­
chronic pancreatitis may show areas of acute paren­ immune pancreatitis, type 2 has no serological findings
chymal and fat necrosis, especially in the earlier stages suggesting an autoimmune aetiology 120–123.
of the disease. In this case, periductal and interlobular
fibrosis predominate over intralobular involvement, Imaging
and acinar tissue is still well preserved. Variations of the CT provides excellent information about morphological
histological features, such as those described in alco­ changes of the pancreas, such as pancreatic duct dila­
holic chronic pancreatitis, are also observed in heredi­ tation or strictures, atrophy, calcification and pseudo­
tary, metabolic, tropical, idiopathic and paraduodenal cystic changes (FIG. 5). The main drawbacks of CT are
chronic ­pancreatitis115–119 (TABLE 2). the limited visualization of the pancreatic ductal sys­
Autoimmune pancreatitis is the only type of chronic tem and the lack of sensitivity and specificity for mild
pancreatitis for which the aetiology can be determined forms of chronic pancreatitis, although the sensitivity
based on histological features120–122 (FIG.  4; TABLE  2). for the detection of calcifications is better than that
Macroscopically, both a focal, tumour-like appearance of MRI11,124,125. MRCP (FIG. 6) has a high sensitivity for
with surrounding normal parenchyma and a diffuse detecting strictures and dilatation in (the side branches
induration and enlargement of the pancreas can be of) the pancreatic duct. Images obtained by MRCP are
observed. Autoimmune pancreatitis type 1 belongs essentially identical to those obtained by ERCP (FIG. 6),
to the spectrum of IgG4‑sclerosing disease and is which was once used for diagnosis but is now restricted
character­ized by intense periductal inflammation, high to interventional procedures. Dynamic MRCP (FIG. 6),
numbers of IgG4‑positive plasma cells, inflammatory in which images are acquired every 30 or 60 seconds
changes of venules, sometimes in the form of an obliter­ over a 10‑minute period following intravenous secretin
ative phlebitis (inflammation of the venules typical in hormone injection, provides excellent spatial resolu­
autoimmune pancreatitis type 1), and by the so‑called tion and functional information on exocrine pancreatic
storiform fibrosis consisting of swirling collagen fibres, insufficiency 124,126,127. Thus, MRCP or dynamic MRCP
activated fibroblasts and myofibroblasts and inflamma­ can be considered as the first choice for the classifica­
tory cells. Dense infiltrates of eosinophils, extension of tion and staging of patients with chronic pancreatitis.
the inflammatory process beyond the borders of the The major limitation of MRCP is in the identification
pancreas, hyperplastic lymph follicles and arteritis may of parenchymal calcifications and small ductal calculi,
be additional features. Infiltrates of neutrophils within which is where CT is especially sensitive. Although
the acinar parenchyma and the ductal epithelium, some­ endoscopic ultrasonography has an important role in
times with the formation of small intraductal abscesses obtaining diagnostic biopsies and therapeutic inter­
(so‑called granulocytic epithelial lesions) are the hall­ ventions, the relatively poor interobserver agreement
marks of autoimmune pancreatitis type 2, which other­ for endoscopic ultrasonography limits its diagnostic
wise shares most of the above-described morphological accuracy for chronic pancreatitis128–130. In addition,

NATURE REVIEWS | DISEASE PRIMERS VOLUME 3 | ARTICLE NUMBER 17060 | 9

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
PRIMER

a b interventions (for example, whether it is responsive to

steroids). Rather complex classification criteria involve a
points system describing the severity of chronic pancre­
atitis. The Chronic Pancreatitis Prognosis Score (COPPS)
predicts individual short-term (12‑month) prognosis
using C‑reactive protein levels, thrombocyte count,
glyco­sylated haemoglobin levels (a marker of diabetes
mellitus), body mass index and pain levels. It allows
objective monitoring of chronic pancreatitis regarding
readmission to hospital, as well as length of hospital
stay 141. Finally, several scoring systems exist for pain,
Figure 5 | CT imaging in chronic pancreatitis. Post-contrast
NatureCT
Reviews | Disease
image of Primers
severe chronic such as visual analogue scales and the Izbicki pain score
pancreatitis shows ductal stones (white arrows in part a and part b) within the dilated (see REF. 3 for a comprehensive discussion).
pancreatic duct (orange arrow; part b).
Prevention
Few data are available when it comes to the prevention of
endoscopic ultrasonography will incorrectly overcall chronic pancreatitis. Alcohol consumption and smoking
early chronic pancreatitis124. Prospective evaluation is should be avoided and abstinence may slow the progres­
required for emerging innovative technologies, such sion of the disease142,143. Earlier surgical intervention can
as MRI elastography for the evaluation of tissue stiff­ also ameliorate the long-term consequences of pain, and
ness131,132, T1‑mapping (MRI) for the quantification endocrine and exocrine function144,145.
of fibrosis in the pancreas133, diffusion-weighted MRI
for the analysis of microscopic physical properties of Management
­tissues at a molecular level134,135, and pancreatic fluid flow The aim of management is to alleviate symptoms (most
dynamics imaging 136. commonly pain, followed by exocrine and endocrine
insufficiency) and to prevent further disease progres­
Scoring sion and disease-related complications, such as bile
Several scoring systems exist to evaluate the severity duct obstruction, gastric outlet obstruction and portal
of chronic pancreatitis. The Cambridge classification vein thrombosis.
for severity grading using ERCP137, and its adaptation for
other types of imaging is still used for diagnosing and Pain management
scoring chronic pancreatitis in adults103,124. The ABC The treatment options for pain associated with chronic
system uses a classification that consists of three stages pancreatitis are shown in BOX  4. Pain is most often
combining clinical criteria (pain, recurrent attacks of ­genuine, but there are several secondary causes that
pancreatitis, local complications, steatorrhoea and dia­ should always be considered, as they are easier to treat
betes mellitus) with imaging (ductal or parenchymal than g­ enuine pancreatic pain97 (BOX 2). Pathology such as
changes)138. The Rosemont classification can diagnose pancreatic duct stones and strictures are typically treated
chronic pancreatitis using only endoscopic ultrasono­ invasively, including by endoscopic stone removal and
graphy criteria139. The number of parameters correlates stenting, extracorporeal shock wave lithotripsy (ESWL)
with the severity of the disease as confirmed by histo­ and surgical resections, and drainage procedures146.
pathology. The M‑ANNHEIM classification attempts to Endoscopic treatment can be combined with ESWL
characterize patients according to aetiology, clinical stage in the presence of large (>4 mm in size) obstructive
and severity 140. The severity of the inflammatory reac­ stones147. Pain relief is seen in 60–80% of cases, but data
tion is evaluated using clinical symptoms and therapeutic must be interpreted carefully as no sham-controlled

a b c

Figure 6 | ERCP and MRCP imaging in chronic pancreatitis. a | Endoscopic retrograde cholangiopancreatography
Nature Reviews | Disease Primers
(ERCP) showing moderate chronic pancreatitis as evidenced by a slightly dilated duct (white arrows) and visible side
branches (orange arrows). b,c | Magnetic resonance cholangiopancreatography (MRCP) imaging shows abnormal
pancreatic duct (white arrows; part b) and enables the visualization of multiple abnormal side branches after secretin
injection (white arrows; part c). Note the bile duct in part b and part c (green arrows).

10 | ARTICLE NUMBER 17060 | VOLUME 3 www.nature.com/nrdp

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
 PRIMER

Box 4 | Different treatment modalities for pain in chronic pancreatitis

Before treatment is initiated, all secondary causes of pain should be • Celiac ganglion neurolysis (temporal neuronal block performed under
considered and treated appropriately. Not all treatments have been CT) and splanchnicectomy (surgical excision of part of the greater
evaluated in chronic pancreatitis, and evidence often comes from splanchnic nerve) aimed at pain suppression
randomized trials in other chronic pain conditions. • Spinal cord stimulation aimed at pain suppression
Risk factor modification • Intrathecal opioids aimed at pain suppression
• Alcohol abstinence Treatment of anxiety and depression
• Tobacco smoking cessation • Tricyclic antidepressants, selective serotonin reuptake inhibitors
• Dietary therapy (to improve the condition in general, including pain) and serotonin–noradrenaline reuptake inhibitors; these drugs also
have a direct effect on the pain system
Initiatives to minimize pressure in the duct system and parenchyma
• Cognitive–behavioural therapy, including sleep therapy
• Pancreatic rest by tube feeding into the jejunum or nutritional products
that do not activate cholecystokinin • Anxiolytics
• Uncoated pancreatic enzymes that may degrade cholecystokinin-­ Neuromodulation aimed at improving pain
releasing factor • Acupuncture
• Cholecystokinin inhibitors • Transcutaneous electrical nerve stimulation
• Proton pump inhibitors (to increase the efficacy of enzyme treatment) • Transcranial magnetic stimulation
Analgesics • Direct current stimulation
• Simple analgesics such as paracetamol and NSAIDs (combined with • Physiotherapy
proton pump inhibitors) Experimental therapies aimed at improving pain and health status
• Adjuvant analgesics; for example, gabapentinoids in general
• Weak or strong opioids plus laxatives • Ketamine, antipsychotics and clonidine
Invasive procedures • Radiotherapy of the pancreas
• Endoscopy • Nerve growth factor inhibitors
• Extracorporeal shock wave lithotripsy • Cannabinoids
• Surgery For details and further information, the reader is referred to www.pancreapedia.org.

studies have been carried out 148. For a detailed discus­ Exocrine and endocrine substitution
sion of the different techniques, the reader is referred Maldigestion-related symptoms (such as diarrhoea,
below and to REF. 148. steatorrhoea, weight loss, flatulence and abdominal
Some patients benefit from enzyme supplementation distention) and nutritional deficiencies are the main
and antioxidants for pain relief, but the results are vari­ consequences of exocrine pancreatic insufficiency,
able and inconclusive149–152. Analgesic treatment follows which can be avoided by appropriate pancreatic enzyme
the principles of the ‘pain relief ladder’ provided by the replacement therapy 166,167. Pancreatic enzymes (usually a
WHO, with the serial introduction of drugs with increas­ combination of lipase, amylase and protease) should be
ing analgesic potency (I: simple analgesics; II: weak administered orally together with each meal or snack168.
­opioids ± adjuvants; and III: strong opioids) that are Enteric-coated enzyme preparations in the form of
titrated until pain relief is obtained153. In most patients, microspheres or mini-microspheres improve digestion
simple analgesics such as paracetamol are not sufficient and the absorption of nutrients, relieve symptoms and
and tramadol is the preferred level II analgesic as it is in improve the nutritional status of patients169. Enzyme
some respects superior to morphine154. Opioid therapy is dose should be adjusted based on the volume, fat and
difficult owing to adverse effects, and rotation between calorific content of meals. As a guide, a starting dose of
the drugs is often needed155,156. Transdermal administra­ 40,000–50,000 United States Pharmacopeia unit (USP)
tion of opioids can be used for patients who have ­trouble or European Pharmacopeia of lipase with main meals,
with tablet ingestion157. Adjuvant analgesics include and one-half of that dose with snacks is generally recom­
antidepressants and anticonvulsants. In chronic pancre­ mended170. Enzyme dose should be further individ­ualized
atitis, only pregabalin has been investigated in a placebo-­ based on symptom response and objective evaluation of
controlled trial and it was found to provide moder­ate nutritional status.
but significant pain relief compared with p ­ lacebo158. Diabetes therapy in chronic pancreatitis should be
In difficult cases, other treatments such as thoraco­ individually tailored, not only to control for hypergly­
scopic splanchnic denervation, spinal cord s­ timulation, caemia but also to reduce the risk of hypoglycaemia101,113.
­radiation therapy, transcranial magnetic stimulation and An adequate and balanced food intake together with
acupuncture may have limited benefits159–163. Patients oral pancreatic enzymes and alcohol abstinence are
who have comorbid depression and anxiety 164 may required for appropriate glycaemic control101,113. Pain
benefit from cognitive–behavioural therapy and drugs treatment should also be optimized, as postprandial
with combined effects on pain and depression such as pain compromises food intake and hence glycaemic
serotonin–noradrenaline reuptake inhibitors3,165, but control. Insulin substitution is usually required for dia­
­multidisciplinary evaluation and follow‑up is mandatory. betes mellitus type 3c, and long-acting basal insulin

NATURE REVIEWS | DISEASE PRIMERS VOLUME 3 | ARTICLE NUMBER 17060 | 11

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
PRIMER

analogues together with on‑demand short-term insu­ that involved 41 patients found that, over 3 years, stent
lin is recommended. Controlling mild hyperglycaemia placement reduced pain by 35% compared with controls
with oral hypoglycaemic agents, such as metformin, but did not affect endocrine function174.
may be a valid approach101. Patients should be moni­ For the obstruction of the common bile duct, one
tored for the d
­ evelopment of retinopathy, nephropathy or more plastic stents can be inserted into the bile duct
and neuropathy. after biliary sphincterotomy to restore patency 175. More
recently, fully covered (membrane-covered metal or
Endoscopy polymer) self-expanding stents (FCSEMS) have become
Between 30% and 60% of patients with chronic pancre­ popular. Like stents in the pancreatic duct, stents in the
atitis will ultimately require some type of endoscopic or bile duct need to be regularly exchanged to prevent occlu­
surgical intervention for treatment 103. The indication for sion and cholangitis (that is, infection of the bile duct
endoscopic therapy is most often intractable pain, and system). A suggested exchange interval for plastic stents
treatment should be considered before patients become is 3 months, whereas FCSEMS remain open for 6 months
opiate-dependent. The most common indications for or longer 176,177. FCSEMS seem to improve outcome
endoscopic treatment are strictures of the pancreatic in case series178, non-randomized179 and randomized
duct, obstruction of the common bile duct and pan­ ­trials180, with a stricture resolution rate of 76–93% and a
creatic pseudocysts148. Several endoscopic procedures ­recurrence rate of strictures of only 14–15%.
are available, which differ in their potential benefits The third most common indication for endoscopic
and prognosis. intervention, following increased pressure in the pan­
When increased pressure in the pancreatic duct creatic duct or the obstruction of the bile duct, is pancre­
owing to a dilated duct with strictures is assumed to be atic pseudocysts, which develop in 20–40% of patients.
the cause of either pain or recurrent disease episodes, These patients should be treated when symptoms (most
endoscopic sphincterotomy is performed first. This often infection, pain, nausea and vomiting) or complica­
procedure, in which the muscular valve that controls tions develop (for example, compression of large blood
the flow of bile and pancreatic juice is cut, is usually ­vessels or the common bile duct, bleeding into pseudo­
followed by the insertion of one or more plastic stents cysts or gastrointestinal tract, or pancreatic pleural fis­
into the pancreatic duct 171. Whether the removal of tulas). When cysts remain asymptomatic, their size and
intraductal stones or their disintegration by prior clin­ical course can still warrant endoscopic drainage.
ESWL is of additional benefit is still being discussed If they measure >5 cm in diameter and do not regress
and is not necessarily supported by high-quality clinical spontane­ously over 6 (or 12) weeks, complications are
evidence172. Observational studies suggest that approx­ likely and drainage should be considered181. Drainage
imately two-thirds of patients experience some pain can be achieved either by inserting a transpapillary
reduction after endoscopic therapy involving sphinc­ plastic stent used for smaller cysts located closer to
terotomy and/or stent placement 173. One randomized the papilla182 or by transmural placement of pigtail or
study on stent placement with no endoscopic treatment self-expanding, lumen-apposing metal stents from the
stomach or the duodenum into the cyst cavity 183 (FIG. 7).
Whether drainage by plastic or lumen-apposing metal
a b stents leads to a better outcome or to fewer complications
is still being debated183. Depending on the location, com­
bined trans-sphincter, transmural or transabdominal
ultrasound and CT‑guided or l­aparoscopic p ­ rocedures
may be necessary.

Surgery
The aim of surgical intervention is to alleviate severe,
constant or recurrent abdominal pain and to cure or
c d prevent additional organ complications, including
bili­ary compression that results in jaundice, duodenal
compression that causes gastric outlet obstruction, or
vascular obstruction that potentially leads to portal
hypertension or portal vein thrombosis with subsequent
* varicose collaterals.
Evidence-based data regarding surgical versus endo­
scopic intervention are sparse and no studies have yet
been performed that are sham controlled. Two random­
ized controlled trials have been carried out comparing
Figure 7 | Endoscopic management of chronic pancreatitis. Nature Drainage
Reviews | of a long Primers
Disease
stenotic segment of the pancreatic main duct (arrows; part a), by the insertion of a plastic surgery and endoscopy in a specific subpopulation of
stent (arrows; part b). A pancreatic pseudocyst with solid content is drained via the patients with chronic pancreatitis; both trials showed that
posterior wall of the stomach with a self-expanding metal stent (arrows; part c) under the surgery was superior for short-term and/or long-term
guidance of endoscopic ultrasonography. CT scan showing near-complete resolution pain relief than endoscopy 184–186. However, these trials
of the pseudocyst (asterisk; part d) following the insertion of the stent (arrows; part d). have been criticized because of their small cohorts and

12 | ARTICLE NUMBER 17060 | VOLUME 3 www.nature.com/nrdp

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
 PRIMER

a b
c
Gallbladder

Stomach

Pancreas

Superior
mesenteric artery
Duodenum Superior
mesenteric vein

Figure 8 | Surgical management of chronic pancreatitis. a | Duodenum-preserving pancreatic Nature Reviews | Disease Primers
head resection.
Note that the pancreas is cut above the superior mesenteric vein and portal vein axis. b | Variant of the duodenum-
preserving pancreatic head resection. Note that the pancreas is not cut above the superior mesenteric vein or portal
vein axis and that the pancreatic duct is opened towards the pancreatic body and tail. c | Photograph of the intraoperative
situs (surgical site) following a variant duodenum-preserving pancreatic head resection. The resection area is circled.
The arrows mark the opened pancreatic duct.

specific subpopulations187, and the results have not led Partial or left pancreatectomies have also been
to a change in general clinical practice. One could argue c­ arried out for segmental dominant disease. Evidence-
that surgery — if necessary — should be carried out early based data regarding the optimal surgical strategy are
during the disease course to reduce further parenchymal sparse, as only a few relatively small randomized con­
damage and severe chronic pain syndromes144,145,188–192. trolled t­ rials have been carried out. Meta-analyses of
There are better outcomes for early versus later surgery these trials, as well as of prospective and retrospective
regarding pain management, and endocrine and exo­ studies198,203–205, have suggested potential advantages
crine function of the remnant pancreas, as well as better of duodenum-­preserving procedures with respect
yields of islet cells (in the case of auto-transplantation) to operating time and hospital stay, whereas better post­
and a lower risk of cancer development (in the case of operative weight gain and quality of life (QOL) and
total pancreatectomy)193–197.
As chronic pancreatitis reduced perioperative complications have not been
­usually presents with diverse morphological patterns, consistently shown. A comparison of different variants
symptoms and complications198, two main surgical of duodenum-preserving procedures did not reveal any
principles have been suggested, which often overlap in relevant differences.
individual surgical approaches: drainage to alleviate pan­
creatic duct pressure and resection to remove inflamed Quality of life
pancreatic tissue. Resection often focuses on the pancre­ Multidimensional QOL questionnaires are highly rele­vant
atic head, as most obstructive complications arise there. to assess disease severity. Previously, various question­
Drainage-only procedures, such as the Partington– naires have been used in chronic pancreatitis, includ­
Rochelle operation, drain the opened pancreatic duct ing the Medical Outcomes Short-Form Health Survey
by connecting it to a small bowel loop (longitu­dinal (SF‑36 or SF‑12)206–209 and the European Organization
pancreato-­jejunostomy)199. Classic resectional proce­ for Research and Treatment of Cancer Quality of Life
dures, such as the Whipple operation (­partial duodeno-­ Questionnaire210. Recently, the Pancreatitis Quality of
pancreatectomy), are carried out less frequently in favour Life Instrument has been introduced211.
of duodenum-preserving pancreatic head resections (the Studies using such questionnaires have all demon­
Beger procedure200) (FIG. 8a), with various variants of this strated that patients with chronic pancreatitis have
oper­ation being described198 (FIG.  8b,c). Duodenum- a substantially impaired QOL211–213. Compared with
preserving and spleen-preserving total pancreatec­ before the development of chronic pancreatitis, 74% of
tomy is indicated especially in patients with hereditary patients have had their work altered and only 37% were
­pancreatitis and/or end-stage disease188. employed with a substantial reduction in income214.
In mild chronic pancreatitis, total pancreatectomy QOL and symptom burden are strongly associated.
with auto-islet transplantation, a procedure in which Decreased QOL has been associated with low body
islet cells are isolated from the resected pancreas and weight, disease duration, unemployment, depression,
re‑implanted in the liver, is possible201. This procedure fatigue, and fear of future and extra-pancreatic com­
can be considered in patients with hereditary pancreati­ plications100,213,215. The association between QOL and
tis, but is controversial in patients with non-hereditary pain has been the most consistent finding, but the exact
chronic pancreatitis who show only minimal morpho­ findings are inconsistent and controversial. A large
logical alterations, as long-term graft failure remains multicentre study showed that patients with constant
a concern202. pain were reported to have a lower QOL than patients

NATURE REVIEWS | DISEASE PRIMERS VOLUME 3 | ARTICLE NUMBER 17060 | 13

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
PRIMER Nature Reviews | Disease Primers

Injury Injury or stress

At risk (Recurrent) acute Early CP Established CP End-stage CP

Susceptibility pancreatitis (RAP) CP biomarkers • Immune dysregulation • Fibrosis and/or sclerosis
factors Sentinel acute • Acinar dysfunction • Exocrine insufficiency
pancreatic event • Islet dysfunction • Diabetes mellitus type 3c
Asymptomatic then RAP • Pathological pain • Pain syndrome
• Metaplasia • PDAC
Susceptibility Symptomatic and
to recurrence Therapeutic supportive treatment
approaches
Resolve

Years Days Months Months to years Remainder of life

Figure 9 | A conceptual model of disease progression linked to the mechanistic definition NatureofReviews
chronic| Disease
pancreatitis.
Primers
This model organizes the genetic and environmental risk factors, the role of recurrent injury or stress, and the normal
and abnormal response to the injury–inflammation–resolution–regeneration sequence. CP, chronic pancreatitis;
PDAC, pancreatic ductal adenocarcinoma. Adapted with permission from REF. 1, Elsevier.

with intermittent pain207–209. However, studies in Poland dysplasia” (REF. 1). Thus, the common features could be
and Denmark showed no differences in temporal pain linked to the responses of specific cell types that display
patterns in QOL, but the pain intensity was closely a normal or abnormal response to injury or stress over
­correlated with QOL100,215. time. Atrophy suggests abnormal regeneration, fibrosis
suggests an abnormal type or duration of the immune
Outlook response, and pain can be subclassified by the type of
Definition injury, the response to injury and/or the duration
Although guidelines summarize the current approach to of injury with normal or abnormal types of adaptation.
diagnosing and managing chronic pancreatitis3, this dis­ Thus, knowledge of the underlying pathogenetic risk
ease needs to be redefined with a focus on essential path­ factors, with the stage and activity of disease, can be used
ways and informative biomarkers rather than relying on to predict future outcomes and targets for intervention.
end-stage pathology. The goal is to develop a rational Debate continues on the criteria for early chronic pan­
framework and approach for early diagnosis, classifica­ creatitis, and whether patients with multiple features
tion and prognosis. A new mechanistic definition should of chronic pancreatitis, except fibrosis, r­ epresent an
include two components, as a combination of these ­atypical form of classic chronic pancreatitis.
features defines the syndrome: the essence of chronic
pancreatitis and the characteristics of ­well-established Conceptual model of disease progression
chronic pancreatitis. Regarding the mechanistic definition, a conceptual
The essence of chronic pancreatitis is characterized model has been proposed of the progression of an
as “a pathologic fibro-inflammatory syndrome of the individ­ual through five stages, beginning with at risk
pancreas in individuals with genetic, environmen­ and progressing to end-stage disease1 (FIG. 9). This model
tal and/or other risk factors who develop persistent serves as a template to organize the genetic and environ­
pathologic responses to parenchymal injury or stress” mental risk factors, the role of recurrent injury or stress,
(REF. 1). This mechanistic definition emphasizes the and the normal and abnormal responses to the inflam­
fact that pancreatitis is rare and only occurs under matory environment. Importantly, this model highlights
extreme conditions, such as pathogenetic variants of the role of acute pancreatitis and recurrent acute pancre­
key susceptibility genes, strong environmental forces, atitis in the initiation of the cascade that leads to chronic
or other, rare, internal factors. Second, as most cases pancreatitis in the majority of patients, and the existence
of chronic pancreatitis are not congenital, consider­ of early-stage disease, with some features of chronic
able events must occur to cause injury and/or stress pancreatitis before the loss of significant function or
to parenchymal cells, resulting in injury signals and irreversible damage characteristic of more-advanced
inflammation. Third, a persistent pathological response disease. Incorrect diagnosis of early chronic pancreati­
is needed for the features of chronic pancreatitis to tis may have harmful consequences for the patient, as it
develop. Thus, the pathogenetic process is an aberration can lead to radical treatments for overlapping dis­orders
of the injury–inflammation–resolution–regeneration that are not true chronic pancreatitis124. However, imag­
sequence. ing evidence of fibrosis alone is not diagnostic of early
The characteristics of chronic pancreatitis have chronic pancreatitis. However, patients with recurrent
also been defined: “common features of established acute pancreatitis, who have strong risk factors (for
and advanced chronic pancreatitis include pancre­ example, alcohol abuse) plus biomarkers of chronic
atic ­atrophy, fibrosis, pain syndromes, duct distor­ pancreatitis (such as fibrosis), without other causes of
tion and strictures, calcifications, pancreatic exocrine fibrosis (such as long-standing diabetes ­mellitus) are
dysfunction, pancreatic endocrine dysfunction and likely to have true early chronic pancreatitis.

14 | ARTICLE NUMBER 17060 | VOLUME 3 www.nature.com/nrdp

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
 PRIMER

Classification As each patient seems to develop a different set of com­

The mechanistic definition allows chronic pancreatitis plications at different rates, each complication must be
to be classified by aetiology, stage and complications. isolated and viewed according to the underlying bio­
Aetiology can be complex, with multiple aetiological fac­ logical system (for example, nervous system or immune
tors converging to cause disease. The TIGAR‑O system response), the specific risk and aetiologies, and the
provides an accepted list of known aetiologies216 (BOX 1). mechanisms of progression. A deep understanding of
Chronic pancreatitis can be classified by stage, as out­ these systems will be the foundation for ‘personalized
lined in FIG. 9. No consensus currently exists on the crite­ medicine’ in the future. However, even in the absence of
ria for each stage, especially as the pathological features a complete understanding of the pathogenesis, we need
of the later stages are not surrogates of each other. For to move forwards in a concerted effort to develop causal
example, one cannot predict the pain experienced by the treatment strategies and to overcome the past century
patient using imaging studies217. The definition of early of symptomatic treatment. Further advances require an
chronic pancreatitis is also important, and a consensus accurate system of defining early chronic pancreatitis,
definition of this stage is needed. and a robust staging system that distinctly describes pro­
The classification of chronic pancreatitis by compli­ gressive disease that is supported by biomarkers that can
cation remains important among researchers because categorize the stage of disease. This will only be possi­
accurate classification of patients is needed to define ble with a combined effort from all the sub-specialities
primary end points for trials in a heterogeneous cohort. involved in the treatment of chronic pancreatitis.

1. Whitcomb, D. C. et al. Chronic pancreatitis: 16. Schwarzenberg, S. J. et al. Pediatric chronic 33. Yadav, D. et al. Alcohol consumption, cigarette
an international draft consensus proposal for a new pancreatitis is associated with genetic risk factors smoking, and the risk of recurrent acute and chronic
mechanistic definition. Pancreatology 16, 218–224 and substantial disease burden. J. Pediatr. 166, pancreatitis. Arch. Intern. Med. 169, 1035–1045
(2016). 890–896.e1 (2015). (2009).
2. Steer, M. L., Waxman, I. & Freedman, S. Chronic 17. Giefer, M. J. et al. Early-onset acute recurrent The North American Pancreatitis Study data
pancreatitis. N. Engl. J. Med. 332, 1482–1490 and chronic pancreatitis is associated with PRSS1 showing a threshold for alcoholic pancreatitis,
(1995). or CTRC gene mutations. J. Pediatr. 186, 95–100 the independent risk of smoking, and the additive
3. Löhr, J. M. et al. United European Gastroenterology (2017). effects of alcohol and smoking.
evidence-based guidelines for the diagnosis and 18. Masamune, A. et al. Nationwide epidemiological 34. Lin, Y., Tamakoshi, A., Hayakawa, T., Ogawa, M.
therapy of chronic pancreatitis (HaPanEU). survey of early chronic pancreatitis in Japan. & Ohno, Y. Associations of alcohol drinking and
United European Gastroenterol. J. 5, 153–199 (2017). J. Gastroenterol. 52, 992–1000 (2017). nutrient intake with chronic pancreatitis: findings from
Current state-of-the-art guidelines for the 19. Otsuki, M. Chronic pancreatitis in Japan: a case–control study in Japan. Am. J. Gastroenterol.
diagnosis and therapy of chronic pancreatitis. epidemiology, prognosis, diagnostic criteria, and 96, 2622–2627 (2001).
4. Muniraj, T., Aslanian, H. R., Farrell, J. & Jamidar, P. A. future problems. J. Gastroenterol. 38, 315–326 35. Takeyama, Y. Long-term prognosis of acute
Chronic pancreatitis, a comprehensive review and (2003). pancreatitis in Japan. Clin. Gastroenterol. Hepatol. 7,
update. Part I: epidemiology, etiology, risk factors, 20. Shimosegawa, T. et al. The revised Japanese clinical S15–S17 (2009).
genetics, pathophysiology, and clinical features. diagnostic criteria for chronic pancreatitis. 36. Rebours, V. et al. Smoking and the course of recurrent
Dis. Mon. 60, 530–550 (2014). J. Gastroenterol. 45, 584–591 (2010). acute and chronic alcoholic pancreatitis: a dose-
5. Peery, A. F. et al. Burden of gastrointestinal disease 21. Cote, G. A. et al. Alcohol and smoking as risk factors dependent relationship. Pancreas 41, 1219–1224
in the United States: 2012 update. Gastroenterology in an epidemiology study of patients with chronic (2012).
143, 1179–1187.e3 (2012). pancreatitis. Clin. Gastroenterol. Hepatol. 9, 37. Andriulli, A. et al. Smoking as a cofactor for causation
6. Yadav, D., Timmons, L., Benson, J. T., 266–273 (2011). of chronic pancreatitis: a meta-analysis. Pancreas 39,
Dierkhising, R. A. & Chari, S. T. Incidence, prevalence, 22. Frulloni, L. et al. Chronic pancreatitis: report from 1205–1210 (2010).
and survival of chronic pancreatitis: a population- a multicenter Italian survey (PanCroInfAISP) on 893 38. Wittel, U. A. et al. Chronic pancreatic inflammation
based study. Am. J. Gastroenterol. 106, 2192–2199 patients. Dig. Liver Dis. 41, 311–317 (2009). induced by environmental tobacco smoke inhalation
(2011). 23. Yang, A. L., Vadhavkar, S., Singh, G. & Omary, M. B. in rats. Am. J. Gastroenterol. 101, 148–159 (2006).
This is one of the few systematic, population-based Epidemiology of alcohol-related liver and pancreatic 39. Thrower, E. Pathologic cellular events in smoking-
studies on the epidemiology of chronic disease in the United States. Arch. Intern. Med. 168, related pancreatitis. Cancers 7, 723–735 (2015).
pancreatitis. 649–656 (2008). 40. Whitcomb, D. C. Genetic risk factors for pancreatic
7. Levy, P. et al. Estimation of the prevalence 24. Lowenfels, A. B. et al. Pancreatitis and the risk of disorders. Gastroenterology 144, 1292–1302
and incidence of chronic pancreatitis and its pancreatic cancer. International Pancreatitis Study (2013).
complications. Gastroenterol. Clin. Biol. 30, Group. N. Engl. J. Med. 328, 1433–1437 (1993). 41. Cohn, J. A. et al. Increased risk of idiopathic chronic
838–844 (2006). This is the first epidemiological study showing that pancreatitis in cystic fibrosis carriers. Hum. Mutat. 26,
8. Levy, P., Dominguez-Munoz, E., Imrie, C., Lohr, M. chronic pancreatitis is a risk factor for pancreatic 303–307 (2005).
& Maisonneuve, P. Epidemiology of chronic cancer. 42. Threadgold, J. et al. The N34S mutation of SPINK1
pancreatitis: burden of the disease and consequences. 25. Grocock, C. J. et al. The variable phenotype of (PSTI) is associated with a familial pattern of
United European Gastroenterol. J. 2, 345–354 (2014). the p. A16V mutation of cationic trypsinogen (PRSS1) idiopathic chronic pancreatitis but does not cause
9. Lowenfels, A. B. et al. Racial factors and the risk in pancreatitis families. Gut 59, 357–363 (2010). the disease. Gut 50, 675–681 (2002).
of chronic pancreatitis. Am. J. Gastroenterol. 94, 26. Howes, N. et al. Clinical and genetic characteristics of 43. Bertin, C. et al. Pancreas divisum is not a cause of
790–794 (1999). hereditary pancreatitis in Europe. Clin. Gastroenterol. pancreatitis by itself but acts as a partner of genetic
10. Yadav, D. & Lowenfels, A. B. The epidemiology of Hepatol. 2, 252–261 (2004). mutations. Am. J. Gastroenterol. 107, 311–317
pancreatitis and pancreatic cancer. Gastroenterology 27. Whitcomb, D. C. et al. Multicenter approach to (2012).
144, 1252–1261 (2013). recurrent acute and chronic pancreatitis in the 44. Okazaki, K. et al. International consensus for the
11. Majumder, S. & Chari, S. T. Chronic pancreatitis. United States: the North American Pancreatitis treatment of autoimmune pancreatitis. Pancreatology
Lancet 387, 1957–1966 (2016). Study 2 (NAPS2). Pancreatology 8, 520–531 (2008). 17, 1–6 (2016).
12. Hirota, M. et al. The seventh nationwide 28. Conwell, D. L. et al. Validation of demographics, 45. Sah, R. P. et al. Differences in clinical profile and
epidemiological survey for chronic pancreatitis etiology, and risk factors for chronic pancreatitis in the relapse rate of type 1 versus type 2 autoimmune
in Japan: clinical significance of smoking habit in USA: a report of the North American Pancreas Study pancreatitis. Gastroenterology 139, 140–148
Japanese patients. Pancreatology 14, 490–496 (NAPS) group. Dig. Dis. Sci. 62, 2133–2140 (2017). (2010).
(2014). 29. Apte, M. V., Pirola, R. C. & Wilson, J. S. Mechanisms 46. Shimosegawa, T. et al. International consensus
13. Wilcox, C. M. et al. Racial differences in the clinical of alcoholic pancreatitis. J. Gastroenterol. Hepatol. diagnostic criteria for autoimmune pancreatitis:
profile, causes, and outcome of chronic pancreatitis. 25, 1816–1826 (2010). guidelines of the International Association of
Am. J. Gastroenterol. 111, 1488–1496 (2016). 30. Nordback, I. et al. The recurrence of acute alcohol- Pancreatology. Pancreas 40, 352–358 (2011).
14. Romagnuolo, J. et al. Clinical profile, etiology, and associated pancreatitis can be reduced: a randomized 47. Rebours, V. et al. The natural history of hereditary
treatment of chronic pancreatitis in North American controlled trial. Gastroenterology 136, 848–855 pancreatitis: a national series. Gut 58, 97–103
women: analysis of a large multicenter cohort. (2009). (2009).
Pancreas 45, 934–940 (2016). 31. Pandol, S. J. & Raraty, M. Pathobiology of alcoholic 48. Masamune, A. Genetics of pancreatitis: the 2014
15. Yadav, D., Muddana, V. & O’Connell, M. pancreatitis. Pancreatology 7, 105–114 (2007). update. Tohoku J. Exp. Med. 232, 69–77 (2014).
Hospitalizations for chronic pancreatitis in Allegheny 32. Yadav, D. & Whitcomb, D. C. The role of alcohol and 49. Keim, V. et al. Clinical characterization of patients with
County, Pennsylvania, USA. Pancreatology 11, smoking in pancreatitis. Nat. Rev. Gastroenterol. hereditary pancreatitis and mutations in the cationic
546–552 (2011). Hepatol. 7, 131–145 (2010). trypsinogen gene. Am. J. Med. 111, 622–626 (2001).

NATURE REVIEWS | DISEASE PRIMERS VOLUME 3 | ARTICLE NUMBER 17060 | 15

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.