Endoscopic diagnosis and surveillance of PSC
Diagnosis of PSC
1. ESGE/EASL recommend that, as the primary diagnostic modality for PSC, magnetic
resonance cholangiography (MRC) should be preferred over endoscopic retrograde cholangiopancreatography
(ERCP).
Moderate quality evidence, strong recommendation.
Although ERCP has been regarded as the standard of reference in diagnosing PSC, MRC
is now recommended as a first-line noninvasive imaging method for patients with suspected
PSC that offers comparable accuracy (except in early-stage PSC restricted to intrahepatic
bile ducts, and in the rare cases of contraindications to MRC) [8]
[9]
[10]
[11]
[12]. A meta-analysis based on 6 studies using ERCP as a reference method concluded that
MRC has high sensitivity and specificity (0.86 and 0.94, respectively) for the diagnosis
of PSC [13]. According to a decision model comparing different approaches in the work-up of
patients with suspected PSC [14], the strategy of initial MRC, followed by ERCP only in selected cases (e. g. ambiguous
MRC findings), is the most cost-effective approach [14]
[15].
The ductographic features defining PSC are described below but a number of other diseases
of the
biliary tree may present similar features ([Table 1]). The specificity of the cholangiographic features of PSC without the additional
diagnostic clinical and biochemical clues is poor [16].
Table 1
Classification of secondary sclerosing cholangitis and conditions that may mimic primary
sclerosing cholangitis on cholangiography.
|
Infection
|
Bacterial/parasitic cholangitis
|
|
Recurrent pyogenic cholangitis
|
|
Immunodeficiency-related (infections)
|
Congenital immunodeficiency
|
|
Acquired immunodeficiency (e. g. HIV)
|
|
Combined immunodeficiencies
|
|
Angioimmunoblastic lymphadenopathy
|
|
Mechanical/toxic
|
Cholelithiasis/choledocholithiasis
|
|
Surgical bile duct trauma
|
|
Intra-arterial chemotherapy
|
|
Drug-induced sclerosing cholangitis
|
|
Ischemic
|
Vascular trauma
|
|
Hepatic allograft arterial insufficiency
|
|
Paroxysmal nocturnal hemoglobinuria
|
|
Other pancreaticobiliary disease
|
Cystic fibrosis
|
|
Sclerosing cholangitis of critical illness
|
|
ABCB4-associated cholangiopathy
|
|
Chronic pancreatitis
|
|
Systemic inflammatory diseases
|
IgG4-associated systemic disease
|
|
Hypereosinophilic syndrome
|
|
Sarcoidosis
|
|
Graft-versus-host disease
|
|
Potentially mimicking on cholangiography
|
Langerhans cell histiocytosis
|
|
Systemic mastocytosis
|
|
Caroli’s disease
|
|
Congenital hepatic fibrosis
|
|
Other types of ductal plate abnormalities
|
|
Hodgkin’s disease
|
|
Cholangitis glandularis proliferans
|
|
Neoplastic/metastatic disease
|
|
Amyloidosis
|
|
Hepatic allograft rejection
|
HIV, human immunodeficiency virus; IgG4, immunoglobulin G4.
Of note, the visualization of the distal common bile duct and the peripheral intrahepatic
ducts is still suboptimal using MRC [10]
[12]. One study has suggested that a numerical score calculated on the basis of three-dimensional
MRC may predict progression of bile duct changes, but the study lacked ERCP reference
[17]. A diagnostic MRC, because of its very high specificity for the diagnosis of PSC
when diagnostic clinical and biochemical clues are present, obviates a confirmatory
ERCP unless therapeutic procedures or ductal sampling are indicated [13]
[18].
2. ESGE/EASL suggest that ERCP can be considered if MRC plus liver biopsy is equivocal
or contraindicated in patients with persisting clinical suspicion of PSC. The risks
of ERCP have to be weighed against the potential benefit with regard to surveillance
and treatment recommendations.
Low quality evidence, weak recommendation.
Whether or not to perform ERCP in patients with high quality normal findings at MRC
depends on the level of clinical suspicion for PSC and on the impact of the diagnosis
on patient management and prognosis. ERCP is regarded as unnecessary in patients with
a low level of clinical suspicion, but it could be considered in patients with an
intermediate or high level of clinical suspicion, as suggested by a meta-analysis
of MRC diagnostic performance [13]. However, this meta-analysis included only studies performed prior to 2007. The
continuous improvement in MRC quality due to use of higher magnetic fields, as exemplified
by the ability to visualize third- and fourth-order intrahepatic ducts as well as
the availability of three-dimensional image acquisition, is likely to further decrease
the probability of abnormal ERCP findings in patients with normal MRC results. In
addition, as detailed reports including the clinical, biochemical, and histological
characteristics and outcomes of these patients with negative MRC but positive ERCP
findings are lacking, the clinical benefit of ERCP can be questioned in this setting.
If high quality MRC images are not available, or in equivocal cases, it is reasonable
to consider patient referral to centers with known technical expertise with MRC as
a first step [19], followed by liver biopsy. If high quality MRC images and liver biopsy still cannot
definitely exclude or confirm the presence of PSC, ERCP can be considered in patients
with persisting clinical suspicion for the diagnosis, to take advantage of the filling
pressure obtained by the balloon occlusion and the slight superiority as to visualization
of the extrahepatic bile ducts.
Ductographic criteria for PSC
The first ERCP criteria for ductographic changes in PSC were published in 1984 by
Li-Yeng &
Goldberg [20]. Typical changes seen in PSC consist of minor irregularities of duct contour and
local narrowing with pre-stenotic dilatation (type I), threadlike narrowings alternating
with normal caliber of bile ducts or slight dilatation (type II), multiple strictures
with saccular dilatations (type III), and the most advanced changes consisting of
advanced ductal narrowing with resultant lack of filling of the peripheral ducts (type
IV). The classification has later been modified by Majojie et al. [21] and Ponsioen et al. [22]
[23]. The classification of Ponsioen et al. [23] has been validated and shown to correlate with patient prognosis ([Table 2]). Another type of classification is based on evaluation of the grade, length, and
extent of strictures, the degree of bile duct dilatation, and the distribution of
lesions [24].
Table 2
Amsterdam classification of cholangiographic changes in primary sclerosing cholangitis
(PSC) [23].
|
Type
|
Intrahepatic
|
Extrahepatic
|
|
0
|
No visible abnormalities
|
No visible abnormalities
|
|
I
|
Multiple caliber changes; minimal dilatation
|
Slight irregularities of duct contour; no stricture
|
|
II
|
Multiple strictures; saccular dilatations, decreased arborization
|
Segmental strictures
|
|
III
|
Only central branches filled despite adequate filling pressure; severe pruning
|
Strictures of almost entire length of duct
|
|
IV
|
–
|
Extremely irregular margins; diverticulum-like outpouchings
|
None of the ductographic criteria published are specific for PSC and the findings
must be
interpreted in the context of patient demographic data and the clinical features.
Review by teams with expertise in complex biliary disease is often useful, as multiple
secondary causes of sclerosing cholangitis must be considered [25] ([Table 3]).
Table 3
Characteristic cholangiographic features in primary sclerosing cholangitis (PSC) and
other ductal diseases.
|
Diagnosis
|
Main cholangiographic features
|
|
PSC
|
Multifocal intrahepatic and extrahepatic bile duct strictures (“beaded” appearance),
slight biliary dilatation, diverticular outpouchings, “pruned tree” appearance at
chronic stage
|
|
Ascending cholangitis
|
Multiple intrahepatic bile duct strictures, stones, biliary abscesses
|
|
Ischemic cholangitis
|
Proximal intrahepatic bile duct strictures, bile duct necrosis, biliomas, abscesses,
biliary cast
|
|
Caustic cholangitis
|
Localized intrahepatic bile duct strictures, irregularities of bile duct wall
|
|
AIDS-related cholangitis
|
Stricture of the distal common bile duct, papillitis, acalculous cholecystitis
|
|
IgG4-related cholangitis
|
Multifocal central bile duct strictures, bile duct wall thickening with visible lumen,
pancreatic abnormalities compatible with autoimmune pancreatitis
|
|
Portal biliopathy
|
Central and extrahepatic bile duct irregularities
|
Unusual cholangiographic features
Some PSC patients may present with cystic dilatations of intrahepatic bile ducts simulating
Caroli’s disease [10]. Of note, the fusiform and small cystic dilatations of intrahepatic (mostly peripheral)
bile ducts, as observed in patients with congenital hepatic fibrosis and autosomal
recessive polycystic kidney disease, should not be misdiagnosed as PSC [11].
Another differential diagnosis is the peculiar cholangiographic phenotype of adult
forms of ABCB4 /MDR3 deficiency which may be characterized by large unifocal or multifocal
spindle-shaped intrahepatic bile duct dilatations with or without apparent bile duct
stenosis [12]
[26]. This diagnosis should be suspected on familial clustering of excessive gallstone
disease and often a history of prior cholecystectomy at age < 40 years and associated
intrahepatic cholestasis of pregnancy, and is confirmed by ABCB4 genotyping.
3. For the diagnosis of PSC, ESGE/EASL do not suggest routine use of endoscopic techniques
other than ERCP (i. e., endoscopic ultrasound including intraductal ultrasound [IDUS],
cholangioscopy, confocal endomicroscopy).
Weak recommendation, low quality evidence.
In the diagnosis of PSC there is no established role for endoscopic techniques beyond
ERCP, e. g. brush cytology, ductal biopsy, cholangioscopy, or confocal laser endomicroscopy.
In selected cases with suspected extrahepatic disease and inconclusive MRC findings,
endoscopic ultrasound (including IDUS) and elastography may add information on common
bile duct strictures, wall thickening, and liver fibrosis stage [27]
[28]
[29]
[30].
ERCP in established PSC
4. ESGE/EASL suggest that a dominant stricture at ERCP should be defined as a stenosis
with a diameter of ≤ 1.5 mm in the common bile duct and/or ≤ 1.0 mm in an hepatic
duct within 2 cm of the main hepatic confluence.
Weak recommendation, low quality evidence.
Deciding on the clinical impact of a bile duct stricture may be challenging. The “dominant
stricture” denomination arose alongside the term “major stricture” early in the history
of endoscopic management of PSC [31]. The “major” or “dominant” stricture terms were initially used more broadly, pertaining
to strictures of the common bile duct and right and left bifurcation of the hepatic
ducts (extrahepatic PSC lesions), since these were found to be more prone to clinical
events than intrahepatic strictures [31]
[32]. The precise definition of a dominant stricture was introduced by Stiehl et al.
in 2002 for use in endoscopic studies as a severity measure [33]
[34], although it employs a somewhat arbitrary value, depending, for example, on filling
pressure. A number of endoscopic studies, both before and after 2002, do not apply
the diameter criterion strictly when determining a dominant stricture [35]
[36], and focus on suspected clinical relevance. Determination of the clinical significance
and potential benefit from endoscopic interventions should therefore not be based
on this definition alone, and the decision for intervention rather considered as a
compound clinical decision.
Multiple dominant strictures can be found in the same patient (12 % in the study by
Bjornsson et al.) [34].
Of note, the ERCP definition of a dominant stricture is usually considered to be not
applicable to MRC, in particular in the extrahepatic ducts, given the insufficient
spatial resolution of MRC [10]
[17] and the lack of the hydrostatic pressure that is present during ERCP.
A complete occlusion cholangiogram should generally be obtained if an ERCP is performed,
because it adds little risk to the ERCP, decreases variability, and may reveal that
a dominant stricture suspected at MRC is indeed not a stricture [37].
5. ESGE/EASL suggest ERCP and ductal sampling (brush cytology, endobiliary biopsies)
should be considered in established PSC in the case of: (i) clinically relevant or
worsening symptoms (jaundice, cholangitis, pruritus); (ii) rapid increase of cholestatic
enzyme levels; or (iii) new dominant stricture or progression of existing dominant
strictures identified at MRC in the context of appropriate clinical findings.
Weak recommendation, low quality evidence.
ERCP can be indicated in patients with a confirmed diagnosis of PSC when changes in
clinical, laboratory, and radiological findings occur during the course of the disease.
The purpose is to make an assessment of the likelihood of the presence of biliary
dysplasia as a risk factor for cholangiocarcinoma (CCA) and to identify biliary strictures
amenable to intervention.
(i) Clinical events
In the early stage of PSC, dominant biliary strictures are usually asymptomatic. Exacerbation
of jaundice (not related to liver failure), episodes of fever and chills suggestive
of cholangitis, or worsening of pruritus are indications for ERCP for the treatment
of dominant strictures and to perform ductal brush sampling to exclude malignancy
[8]
[38]. Worsening pain in the right upper abdominal quadrant, fatigue, and weight loss
also need careful evaluation.
(ii) Laboratory results
Serum laboratory tests are neither sensitive nor specific enough to evaluate PSC progression
[38], but in the case of rapid increase of serum bilirubin levels and/or cholestatic
liver enzymes (serum ALP, serum GGT) ERCP is indicated [6], especially in patients with a diagnosis of clinically significant hilar or extrahepatic
strictures on MRC. Elevation of serum CA19-9 in PSC patients has an unsatisfactory
sensitivity (14 %) and positive predictive value (PPV) (67 %) for the diagnosis of
CCA [36]
[38]
[39], and is not helpful in selecting patients for ERCP.
(iii) Progression/new-onset clinically significant strictures on MRC
Progressive intrahepatic or extrahepatic bile duct dilatation on imaging studies (ultrasound
or MRC) is an indication for ERCP with ductal sampling [6]. A careful evaluation of new-onset dominant strictures in PSC is recommended, because
of the increased risk of CCA in this situation.
In detail, a stricture that is disproportionately severe relative to others, concomitant
biliary filling defects, marked biliary dilatation (≥ 2 cm for the common bile duct,
≥ 1 cm for the right or left intrahepatic ducts, ≥ 5 mm for other intrahepatic ducts)
suggests CCA [40]. Conversely, this risk was low in patients without dominant strictures according
to a 25-year experience [41]. Abnormal cytological findings, such as suspicion of malignancy or aneuploid DNA
findings need a close follow-up by ERCP with repeated sampling, unless urgent liver
transplantation is considered to be warranted.
The utility of ERCP in handling dominant strictures was shown in a prospective study
[42] on 171 PSC patients followed for 20 years: repeated endoscopic therapy was associated
with a transplant-free survival of 81 % at 5 years and 52 % at 10 years after initial
endoscopic therapy. In this population, a 6 % CCA rate was found in patients with
dominant strictures.
6. ESGE/EASL suggest that, in patients with an established diagnosis of PSC, MRC should
be considered before therapeutic ERCP.
Weak recommendation, low quality evidence.
MRC may be useful to confirm the indication, to exclude focal parenchymal changes,
and to give the clinicians performing the ERCP imaging-based guidance to minimize
the risk of complications. Regarding MRC in established PSC, a retrospective single-center
study reported a 76 % accuracy of MRC in the diagnosis of CCA complicating PSC [40]. For these reasons, patients with an established diagnosis of PSC should have an
MRC examination in their clinical records [13]
[43].
7. ESGE/EASL suggest performing endoscopic treatment with concomitant ductal sampling
(brush cytology, endobiliary biopsies) of suspected significant strictures identified
at MRC in PSC patients who present with symptoms likely to improve following endoscopic
treatment.
Strong recommendation, low quality evidence.
Selected series reporting on endoscopic treatment in PSC patients are summarized in
[Table 4]; none of these compared performance versus no performance of endoscopic treatment
for dominant stricture. The benefits reported following dilation of dominant stricture
included short-term improvement of symptoms and of liver biochemical test results,
as well as a longer liver transplantation-free survival compared to that predicted
using the Mayo clinical risk model. Similar findings have also been reported in several
smaller case series [32]
[47]
[48]
[49]
[50].
Table 4
Selected series reporting on endoscopic treatment of dominant strictures in primary
sclerosing cholangitis.
|
First author,
Year [ref.]
|
Study design
|
Patients, n
|
Intervention
|
Outcomes
|
Results
|
|
1. Dilation ± stenting
|
|
Gotthardt,
2010 [42]
(Extension of Stiehl 2002 study [33])
|
Prospective
|
96
(ALP > 2 × ULN)
|
Balloon dilation
(8 mm in CBD, 6 – 8 mm for IHBD), plus stent in 5 patients with severe cholestasis
and bacterial cholangitis
|
Short-term improvement in cholestasis
Liver transplantation-free survival
Complications
|
-
At 2 weeks, mean bilirubin level significantly decreased (by 56 %)
-
Improvement in symptoms and liver transplantation-free survival
-
Comparison with Mayo model not reported (5-year and 10-year liver transplantation-free
survival, 81 % and 52 %)
-
Overall complication rate, 3.8 %
|
|
Gluck,
2008 [35]
|
Retrospective
|
84
Symptomatic patients
|
Balloon dilation and stenting (70 % and 51 % of patients, respectively)
|
Liver transplantation-free survival
|
-
Higher proportion of patients alive with no liver transplantation at 3 and 4 years
than predicted using Mayo model (P < 0.05); at 1 and 2 years survival similar to Mayo prediction
-
Adverse events in 21 therapeutic ERCPs (7.2 % of 291 procedures, 25 % of patients)
|
|
Stiehl,
2002 [33]
|
Prospective
|
52
(ALP > 2 × ULN)
|
Balloon dilation
(8 mm in CBD, 6 – 8 mm for IHBD), plus stent in 5 patients with severe cholestasis
and bacterial cholangitis
|
Bilirubin and liver enzymes 2 weeks after dilation
Symptoms
Liver transplantation-free survival
|
-
At 2 weeks, significant decrease in liver enzymes and bilirubin
-
Improvement of jaundice in 24 /24 and of pruritus in 12/13 patients
-
Longer liver transplantation-free survival than predicted using 1992 Mayo model (P < 0.0001)
|
|
Baluyut,
2001 [44]
|
Retrospective
|
56 with symptoms
7 without symptoms
|
Balloon dilation (4 – 12 mm, n = 61)
Once per year, with stent if no significant radiological improvement following dilation
(n = 33)
|
Liver transplantation-free survival
Complication rate
|
|
|
2. Stenting
|
|
Ponsioen,
1999 [36]
|
Retrospective
|
32
Symptomatic patients with successful stenting for dominant stricture
|
1-week stenting (10-Fr stent) with no balloon dilation
|
2-month symptomatic and biochemical improvement,
Actuarial curve of re-intervention-free patients
|
-
Improvement of symptoms in 83 %
-
Significant decrease in bilirubin (44 % had increased conjugated bilirubin at baseline)
and cholestasis enzymes
-
Re-intervention-free patients (actuarial): 60 % at 3 years
|
|
van Milligen de Wit,
1996 [45]
|
Retrospective
|
25
With symptoms or progression of serum tests for cholestasis
|
Stenting for a median of 3 months (plus 8-mm dilation in 3 patients)
|
Change in symptoms and biochemical tests within 6 months following stent insertion
Adverse events
|
-
Improvement of symptoms in 76 %
-
Significant decrease in bilirubin (52 % had increased bilirubin at baseline) and serum
tests for cholestasis
-
32 episodes of cholangitis/jaundice related to stent clogging
|
|
3. Dilation vs. dilation + stenting
|
|
Kaya,
2001 [46]
|
Retrospective
|
71
with symptoms
|
Balloon dilation
(4 – 8 mm, n = 34) vs.
Balloon dilation with
3 – 4-month stenting (n = 37)
Intervention via PTBD in 0/34 of balloon group vs. 23 /37 of stent group
|
Biochemical course up to 24 months
|
-
Both strategies improved liver biochemistry; fever resolved only in the dilation without
stent group. No additional benefit of stenting after balloon dilation
-
More complications in stent vs. dilation alone group (P = 0.001)
-
More complications in PTBD vs. ERCP group (P < 0.001)
-
(No multivariate analysis)
|
ALP, alkaline phosphatases; ULN, upper limit of normal values; CBD, common bile duct;
IHBD, intrahepatic bile duct; ERCP, endoscopic retrograde cholangiopancreatography;
PTBD, percutaneous transhepatic biliary drainage.
The main criticisms of these studies are as follows:
-
The Mayo clinical risk model was not designed to evaluate patients with dominant stricture;
specifically, many patients underwent therapeutic ERCP because of elevated bilirubin,
which is part of the Mayo risk score and went down in most patients after the intervention.
Hence, baseline Mayo risk score was not determined in a steady-state situation.
-
Serum test results for cholestasis may spontaneously fluctuate in patients with PSC
complicated
or not with a dominant stricture. In 125 PSC patients, Bjornsson et al. reported changes
in serum ALP and serum bilirubin from baseline up to 12 months following ERCP. As
patients with dominant stricture received no stricture dilation, the authors stated
that “If our patients had been consequently dilated or stented the decrease in bilirubin
and clinical features at follow-up would have been attributed to endoscopic therapy”
[34]. However, in that study, the variations reported in ALP and in total serum bilirubin
after versus before ERCP were not significant, in contrast with various studies listed
in [Table 4] that used dominant stricture dilation/stenting. Also, it was not clear on what basis
these patients were treated conservatively, while others did receive endoscopic therapy.
Other limitations of most studies listed in [Table 4] include retrospective design, selection bias, and reporting of results for a mixture
of treatments, namely dilation with and without stenting of dominant strictures as
well as, in a minority of patients, treatment with ursodeoxycholic acid started during
follow-up.
A critical issue is that potential benefits must be weighed against the certain risks
of therapeutic ERCP in patients with no other therapeutic option except liver transplantation.
Symptoms likely to improve following dominant stricture treatment generally include
pruritus, pain, cholangitis, and jaundice in patients with a significant (≥ 20 %)
increase in cholestasis, while in patients with end-stage liver disease, only cholangitis
is expected to improve.
Finally, patients with advanced liver disease with cirrhosis may not benefit from
endoscopic treatment. Ahrendt et al. reported no change in serum bilirubin at 1 year
following endoscopic and/or percutaneous stricture dilation in 10 patients with cirrhosis
and a baseline serum bilirubin ≥ 5 mg/dL [51]. Death following endoscopic balloon dilation of dominant stricture has been reported
in a patient with PSC and end-stage liver disease [46]. Diagnostic ERCP was followed by deterioration of cholestasis in 7 of 8 patients
with more advanced PSC at biopsy (Ludwig stage III or IV) versus 1 of 7 with less
advanced disease (Ludwig stage I or II) [52].
Balloon dilation versus stent therapy
8. ESGE/EASL suggest that the choice between stenting and balloon dilation should
be left to the endoscopist’s discretion.
Weak recommendation, low quality evidence.
Results from selected series reporting on endoscopic treatment of dominant strictures
in PSC are
summarized in [Table 4]. Of note: (i) in the majority of studies that reported on balloon dilation for dominant
stricture, stents were inserted in a minority of patients; (ii) a significant improvement
in liver transplantation-free survival compared with the Mayo model has been reported
only with balloon dilation; and (iii) the perforation rate has been higher with stenting
compared with balloon dilation.
A single retrospective study compared balloon dilation versus balloon dilation combined
with stenting for dominant stricture in PSC patients (n = 34 and n = 37, respectively)
[46]. The “balloon dilation alone” group was treated by endoscopic means only, while
23 patients (62 %) in the “stenting” group underwent percutaneous treatment because
of failed endoscopic access and/or dominant stricture dilation. Serum bilirubin decreased
similarly in both groups of patients, but more procedures and more complications were
recorded in the stent versus the balloon dilation group (median number of procedures
per patient, 5.0 vs. 2.1, respectively; patients with complications, 54 % vs. 15 %,
respectively). Complications included bile duct perforation in 7 patients (10 %),
5 of whom were in the stent group. However, it is difficult to draw conclusions because
of the different access routes used (percutaneous in 62 % in the stent group vs. 0
in the balloon dilation group), a selection bias due to more severe stricture in the
stent group, and the long stenting duration used (mean 3 months) putting the patient
at high risk for stent clogging and cholangitis. A short stenting duration (see recommendation
13) is currently the standard of care.
The European multicenter randomized DILSTENT trial comparing single-balloon dilatation
versus short-term stenting was prematurely stopped recently after a planned interim
analysis. Preliminary results show no differences in outcome, but a significantly
higher serious adverse event rate in the stent group (Dr. C.Y. Ponsioen, personal
communication).
Role of sphincterotomy
9. ESGE/EASL recommend weighing the anticipated benefits of biliary papillotomy/sphincterotomy
against its risks on a case-by-case basis.
Strong recommendation, moderate quality evidence.
Biliary papillotomy/sphincterotomy should be considered especially after difficult
cannulation
Strong recommendation, low quality evidence.
Biliary sphincterotomy was performed routinely as part of the endoscopic treatment
of dominant stricture in some studies [46] while its use was restricted to specific cases such as stone extraction and difficulties
in stent insertion in other studies. For example, in 32 PSC patients treated with
stents for dominant stricture, sphincterotomy was performed in 12 patients (38 %)
[36] while in another study of dominant stricture dilation with/without stenting, sphincterotomy
was performed in 63 % of 63 patients [44].
Generally, biliary sphincterotomy is not recommended as a routine procedure prior
to biliary stenting because of the associated risks as demonstrated in randomized
controlled trials (RCTs) [53]. However, if cannulation is difficult, biliary sphincterotomy is advised, bearing
in mind that these patients are likely to require multiple procedures. Many endoscopists
prefer a small sphincterotomy in PSC in order to avoid ascending cholangitis.
Specifically in PSC, biliary sphincterotomy was independently associated with an increased
risk of short-term adverse events in two retrospective studies (odds ratios [OR] 4.7
and 5.0) [54]
[55] while previous biliary papillotomy/sphincterotomy was protective for subsequent
ERCPs [54]. Therefore experienced endoscopists perform biliary sphincterotomy in patients with
difficult cannulation in whom ERCP is likely to be repeated during follow-up.
Balloon dilation
10. ESGE/EASL suggest selecting a balloon caliber of up to the maximum caliber of
the ducts delimiting the stricture.
Weak recommendation, low quality evidence.
11. ESGE/EASL suggest repeating dilation of relapsing dominant stricture if: (i) the
dominant stricture is regarded as the cause of recurrent symptoms (cholangitis, pruritus)
or of significant increase in cholestasis; and (ii) the patient’s response to previous
dilations has been satisfactory.
Weak recommendation, very low quality evidence.
There are no comparative data on the optimal dilation scheme or balloon diameter for
treating dominant strictures. In the largest prospective study (500 endoscopic balloon
dilations in 96 patients), the authors performed stepwise dominant stricture dilation
up to diameters of 8 mm and 6 – 8 mm in the common bile duct and the hepatic ducts,
respectively [42]. Bile duct diameter upstream and downstream of the dominant stricture should be
taken into account for selecting the balloon diameter to avoid dilating to more than
the duct diameter. Balloon dilations are usually repeated at intervals of 1 to 4 weeks
up to technical success, for an average of 2 – 3 balloon dilations [33]
[42]
[50]. Technical success has been defined as complete balloon inflation within the dominant
stricture with no waist observed fluoroscopically, followed by the unobstructed passage
of contrast medium through the dilated biliary segment to the duodenum [42]
[50]. Using this technique, bile duct perforation was reported in 0.2 % of dominant stricture
dilations (1 % of patients) [42]. In contrast, another study that used balloons of diameter 4 – 12 mm for dilation
reported dilation-related biliary perforations in 3.5 % of procedures [44].
Repeat balloon dilation during follow-up after initial treatment (usually consisting
of several ERCPs) has been mentioned in some studies, but no results of the repeat
dilation, in terms of clinical or biochemical improvement, have been reported [33]
[50].
Stent therapy
12. ESGE/EASL suggest selecting a single 10-Fr stent for dominant stricture in the
extrahepatic ducts or two 7-Fr stents for hilar strictures extending into the left
or right hepatic duct (final stent diameters in the case of stepwise stenting)
Weak recommendation, very low quality evidence.
In all large studies of endoscopic treatment for dominant stricture, plastic stents
measuring 7 to 10 Fr in diameter have been used, with no reported comparison of the
results obtained with various stent diameters. Specifically, the Amsterdam group aimed
at inserting a single 10-Fr stent, and if this was not possible at first attempt,
it was preceded by 1-week stenting with a 7-Fr stent or insertion of a nasobiliary
catheter [36]
[56]. The Mayo group used 7 – 10-Fr stents at the endoscopist’s discretion [46]. The Indianapolis group did not mention the diameter of stents used [44]. Two 7-Fr stents have typically been used in patients with multiple bilateral dominant
strictures, and in patients with a hilar stricture extending into the left or right
hepatic duct in order to avoid temporary obstruction of the contralateral biliary
system. In general, the stent caliber and length must be adapted to the specific biliary
tree configuration.
In other diseases, studies have shown that polyethylene stents provide better short-term
(1-month) patency than Teflon models and that, in the long term, 10-Fr models provide
longer biliary patency compared with thinner ones (11.5-Fr models do not provide longer
patency) [53].
With respect to balloon dilation prior to stenting, it is currently unclear whether
balloon dilation is beneficial before stent placement.
Duration of stenting
13. ESGE/EASL suggest that stents used for treating dominant stricture should be removed
1 – 2 weeks following insertion.
Weak recommendation, low quality evidence.
No comparison of various stenting durations has been identified in studies reporting
on stenting for dominant stricture. A short stenting duration is currently favored
because stents tend to clog rapidly in PSC patients and similar efficacy results have
been reported with short (1 – 2 weeks) versus standard (8 – 12 weeks) stenting duration.
Specifically, a retrospective study of short-term stenting (mean duration 11 days)
in 32 symptomatic PSC patients with dominant stricture showed, at 2 months, a symptomatic
improvement in 83 % of the patients as well as a significant improvement of cholestasis
test results; at 1 and 3 years, actuarial analysis showed that 80 % and 60 % of patients,
respectively, would not require re-intervention [36]. Stent dysfunction was not reported in this study but two patients treated by stent
removal developed hydrops of the gallbladder. The same group of authors had previously
reported similar efficacy results with 3-month stenting in 25 patients with symptomatic
dominant stricture but, in that study, unscheduled stent exchange had to be performed
on 32 occasions because of suspected stent clogging (cholangitis n = 23, jaundice
n = 9) [45].
All studies mentioned focused on clinical and serum liver tests, not radiological
data, to assess the short-term effect of therapeutic ERCP [36]
[45]
[46]
[56]. Endoscopic treatment has been repeated in a sizeable proportion of patients. For
example, with long median stenting periods (3 months), the median number of repeated
ERCPs per patient ranged between 3 and 5 during follow-up periods of 29 and 22 months
in two studies [45]
[46], while following a short stenting period (mean 11 days) repeat ERCP rates at 1 and
3 years after treatment were estimated at 20 % and 40 %, respectively [36]. Other details about repeated treatments were not reported.
In many centers, stents are removed during an esophagogastroduodenoscopy without biliary
opacification in PSC patients.
Complications of endoscopic therapy
14. ESGE/EASL suggest that ERCP in PSC patients should be undertaken by experienced
pancreaticobiliary endoscopists.
Strong recommendation, very low quality evidence.
Several studies have evaluated the risk of complications in PSC patients undergoing
ERCP [33, 35, 44, 49, 54, 55, 57 – 62] ( [Table 5]). ERCP carries an increased risk for complications in the context of PSC, especially
pancreatitis, cholangitis, and extravasation of contrast, although not all studies
have documented such an increased risk in PSC [59]
[62]. In a systematic survey [63] of post-ERCP complications associated with various indications for ERCP, including
21 prospective studies and 16 855 patients, the total complication rate was 6.85 %
(95 %CI 6.46 % – 7.24 %). Pancreatitis occurred in 585 patients (3.47 %, 95 %CI 3.19 % – 3.75 %).
In another large retrospective single-center study [47], with 11 497 procedures over 12 years, the total complication rate was 4.0 % and
pancreatitis occurred in 3.6 %. The overall risk of adverse events in patients with
PSC has varied in different, much smaller studies, from 1.8 % to 18.4 % [33]
[35]
[44]
[49]
[55]
[57]
[58]
[59]
[60]
[61]
[62], which is higher than reported for other indications [47]
[63].
Table 5
Complications of endoscopic retrograde cholangiopancreatography (ERCP) in primary
sclerosing cholangitis (PSC) patients.
|
First author,
Year [ref]
Country
|
Study design
|
Patients/ERCPs
|
Complications, % of procedures
|
|
Total
|
Pancreatitis
|
Cholangitis
|
|
Lee,
1995 [49]
USA
|
Retrospective
|
53/175
|
13.7
|
7
|
8
|
|
van den Hazel,
2000 [57]
The Netherlands
|
Retrospective
|
83/106
|
9
|
3
|
2
|
|
Baluyut,
2001 [44]
USA
|
Retrospective
|
63/63
|
1.8
|
1.26
|
0.6
|
|
Stiehl,
2002 [33]
Germany
|
Retrospective
|
106/ERCP yearly, median 5 years
|
9
|
5.2
|
3.3
|
|
Enns,
2003 [58]
Canada
|
Retrospective
|
104 patients
|
17
|
5
|
7.5
|
|
Gluck,
2008 [35]
USA
|
Retrospective
|
106/317
|
7.3
|
3.8
|
0.95
|
|
Etzel,
2008 [62]
USA
|
Retrospective
|
PSC: 30/85
Non-PSC: 45 /70
|
12.9
8.6
|
2.4
2.9
|
5.9
1.4
|
|
Bangarulingam
2009 [59]
USA
|
Retrospective
|
PSC: 168
Non-PSC: 981
|
11
8
|
5
4
|
3.6
0.2
|
|
Alkhatib,
2011 [60]
USA
|
Retrospective
|
75/185
|
8
|
5
|
1
|
|
Ismail,
2012 [54]
Finland
|
Retrospective
|
441/441
|
9
|
7
|
–
|
|
Navaneethan
2015 [55]
USA
|
Retrospective
|
294/697
|
4.3
|
1.2
|
2.4
|
|
von Seth
2015 [61]
Sweden
|
Retrospective, national registry study
|
PSC: 141/141
Non-PSC: 8791
|
18.4
7.3
|
7.8
3.2
|
7.1
2.1
|
Retraction of the papilla and an altered, more difficult position of the endoscope
due to hypertrophy of the left liver lobe may be encountered during ERCP in PSC patients.
Whether this actually influences cannulation success rates has not been investigated
by specific studies. Cohort studies describing PSC patients provide only limited details
on cannulation difficulties, with failure rates of 0 % to 6 % [33]
[36]
[41]
[49]
[50]
[57]
[62]
[64]
[65]
[66]. Furthermore, there is likely a selection bias since most retrospective series describing
the results of endoscopic treatment have the initiation of therapy as prerequisite,
therefore potentially excluding cannulation failures.
The largest series is the study by Ismail et al. [54]. In this retrospective review of 441 ERCP procedures over a 3-year time period,
primary cannulation success was 88.2 %. Of note, in 137 patients (37.8 %) a previous
biliary sphincterotomy had been performed. Pancreatic sphincterotomy as an access
technique was used in 11.8 % and freehand needle-knife sphincterotomy in a further
2.5 %. The primary failure rate was 0.5 %. These figures suggest that cannulation
in PSC patients may indeed be more difficult than in other types of patients.
Post-ERCP pancreatitis
15. ESGE/EASL recommends routine rectal administration of 100 mg of diclofenac or
indomethacin immediately before or after ERCP in all patients without contraindication.
In addition to this, in the case of high risk for post-ERCP pancreatitis, the placement
of a 5-Fr prophylactic pancreatic stent should be considered.
Strong recommendation, high quality evidence.
Post-ERCP pancreatitis (PEP) is the most common and feared complication associated
with ERCP. The risk for PEP in PSC varies from 1 % to 7 %, although the diagnostic
criteria vary between studies [67]. Although the quality of the evidence is low, the factors increasing the risk for
PEP are probably not different in PSC patients from those in the general population:
female sex (OR 2.6, P = 0.015) and a guidewire in the pancreatic duct (OR 8.2, P < 0.01). Presence of a native papilla increases the risk whereas previous sphincterotomy
decreases it [54], suggesting that pre-emptive endoscopic papillotomy might be warranted in PSC patients
where repeat procedures might be anticipated. This has however yet to be proven.
Prolonged papilla contact time, as well as therapeutic procedures such as biliary
brush cytology, sphincterotomy, stenting, and dilation, are associated with increased
risk of PEP. Precut biliary and pancreatic sphincterotomy is markedly associated with
PEP [54], possibly reflecting the difficult cannulation and prolonged procedure time. A recent
Cochrane analysis comparing the contrast-assisted with the guidewire-assisted cannulation
technique showed that the guidewire technique both increased the primary cannulation
rate and reduced the risk of PEP, and it appears to be the most appropriate first-line
cannulation technique [68].
Rectal nonsteroidal anti-inflammatory drugs (NSAIDs). In its 2014 update to a Guideline on the prophylaxis of PEP, ESGE recommends routine
rectal administration of 100 mg of diclofenac or indomethacin immediately before or
after ERCP in all patients undergoing ERCP who were without contraindication to NSAIDs
[69]. The recommendation was supported by the results of six meta-analyses published
between 2009 and 2014 that compared NSAIDs versus placebo administration for prophylaxis
of post-ERCP pancreatitis. These meta-analyses concordantly showed the benefit of
NSAIDs in preventing either mild or moderate/severe PEP. These results were further
supported by subsequent meta-analyses [70]
[71] and the cost-efficiency of this approach has been demonstrated [72]. This recommendation applies to PSC patients.
Pancreatic stenting: The ESGE 2014 recommendation on prophylactic pancreatic stenting was supported by:
(i) three meta-analyses of RCTs that showed a significant reduction in the incidence
and the severity of PEP when prophylactic pancreatic stenting was used; and (ii) a
study showing that pancreatic stent placement is cost-effective only in patients/procedures
at high risk for post-ERCP pancreatitis.
The following conditions relevant to PSC are considered to represent high risk for
PEP: precut biliary sphincterotomy, pancreatic guidewire-assisted biliary cannulation,
endoscopic balloon sphincteroplasty, pancreatic sphincterotomy, and presence of more
than three of the following risk factors: female gender, previous pancreatitis, younger
age, nondilated extrahepatic bile ducts, absence of chronic pancreatitis, normal serum
bilirubin, duration of cannulation attempts > 10 min, > 1 pancreatic guidewire passage,
pancreatic injection, failure to clear bile duct stones, intraductal ultrasound.
16. ESGE/EASL suggest routine administration of prophylactic antibiotics before ERCP
in patients with PSC.
Strong recommendation, low quality evidence.
Bacterial cholangitis and bacteriobilia are a not infrequent finding among patients
with PSC. In studies evaluating the complications of ERCP in PSC the risk for cholangitis
has varied from 0.25 % to 8 % [33]
[35]
[44]
[49]
[54]
[55]
[57]
[58]
[59]
[60]
[61]
[62] depending on, among other items, the criteria used to define cholangitis. The use
of prophylactic antibiotics varies markedly between studies, in terms of prevalence,
type of antibiotic, and duration of administration (from 1 oral dose before the procedure
to 1-week dosing afterwards). In a Cochrane meta-analysis (9 RCTs, 1573 patients),
the prophylactic use of antibiotics was shown to prevent cholangitis (relative risk
[RR] 0.54, 95 %CI 0.33 – 0.91), septicemia (RR 0.35, 95 %CI 0.11 – 1.11), bacteremia
(RR 0.50, 95 %CI 0.33 – 0.78), and pancreatitis (RR 0.54, 95 %CI 0.29 – 1.00). It
was concluded that prophylactic antibiotics reduce bacteremia and seem to prevent
cholangitis and septicemia in patients undergoing elective ERCP [73]. Our recommendation is in line with the American Society for Gastrointestinal Endoscopy
(ASGE) recommendation to prescribe antibiotic prophylaxis in procedures where drainage
achieved at ERCP is incomplete or achieved with difficulty, such as in PSC [74]. Bile fluid sampling could be considered during ERCP, to guide antibiotic treatment
in case cholangitis occurs despite the prophylaxis [75].
PSC and cholangiocarcinoma
17. EASL/ESGE recommend that cholangiocarcinoma (CCA) should be suspected in any patient
with worsening cholestasis, weight loss, raised serum CA19-9, and/or new or progressive
dominant stricture, particularly with an associated enhancing mass lesion.
Strong recommendation, moderate quality evidence.
18. A raised serum CA19-9 may support the diagnosis of CCA, but has a poor specificity.
Weak recommendation, low quality evidence.
PSC is associated with a markedly increased risk for CCA with a lifetime risk of 10 % – 20 %
[76]
[77], or up to 400-fold compared with the general population [78]. CCA represents a common cause of death among PSC patients [79], whereby 27 % – 50 % of all CCAs are detected within 1 year of a PSC diagnosis [41]
[78]
[80] depending on the indications for ERCP.
CCA should be suspected in PSC patients experiencing rapid deterioration of liver
function test findings, increasing jaundice, weight loss, and abdominal pain. However,
the development of such a clinical trend may also suggest an advanced form of CCA.
An observational study performed in the US on 230 patients affected by PSC, 23 of
whom had CCA, showed no major differences in clinical features between patients without
CCA and those with CCA at an earlier stage [40].
Increased serum CA19-9 levels have been reported to indicate the development of CCA
in PSC patients. Cutoff levels of 129 or 100 U/mL detected CCA with high sensitivity
(nearly 80 %) and specificity (nearly 100 %) [81], but only in advanced cases of CCA. These data are in contrast with other observations
that showed that one third of PSC patients with high CA19-9 levels did not have CCA
[82]
[83]. In a recent study performed on 433 PSC patients, 41 of whom had biliary malignancy,
the use of FUT2 /3 genotype-dependent cutoff values for CA19-9 improved sensitivity
and reduced the number of false-positive results [84]. In a study screening for biliary dysplasia using ERCP and brush cytology, serum
CA19-9 had no prognostic value for biliary dysplasia or CCA [37].
Currently, there are no definite radiologic features that indicate CCA in a PSC patient,
although the detection of a dominant stricture by MRC may be suggestive for CCA. However,
50 % of PSC patients experience a dominant stricture and its absence does not rule
out CCA. In a cohort of 230 patients, ultrasound, computed tomography (CT), and MRCP
were found to have high specificity but low sensitivity (10 % – 32 %) [40].
ERCP findings indicative of CCA
Dominant strictures are frequent in PSC [42] and do not per se indicate development of a malignancy. In a large single-center
study, CCA was seen in 6/95 dominant strictures (6 %). In general it could be inferred
that the chance of any dominant stricture of harboring a CCA is around 5 %. Most CCAs
develop in the perihilar region or in extrahepatic bile ducts, and are reachable with
a cytological brush. In a large series of patients with CCA [85], 50 % had perihilar cancers, 42 % had distal cancers, and only 8 % were intrahepatic
CCAs. No specific imaging features have been found to differentiate benign strictures
from malignant ones. Based on ERCP findings only, it is not possible to exclude CCA
from benign strictures caused by PSC, and the diagnosis always requires additional
techniques such as imaging or biliary cytology or histology.
19. ESGE/EASL recommend ductal sampling (brush cytology, endobiliary biopsies) as
part of the initial investigation for the diagnosis and staging of suspected CCA in
patients with PSC.
Strong recommendation, high quality evidence.
20. ESGE/EASL suggest that fluorescence in situ hybridization (FISH) or equivalent
chromosomal assessments are considered in patients with suspected CCA when brush cytology
results are equivocal.
Weak recommendation, low quality evidence.
21. ESGE/EASL suggest that additional investigations such as cholangioscopy, endoscopic
ultrasound, and probe-based confocal laser endomicroscopy (pCLE) may be useful in
selected cases.
Weak recommendation, low quality evidence.
Brush cytology
Bile duct brushing is the most common method for tissue sampling in patients with
PSC for detecting inflammation, biliary dysplasia or CCA ( [Table 6], [Table 7]). In a recent meta-analysis (11 studies, 747 patients) [95], the pooled diagnostic values of bile duct brushing for diagnosis of CCA in patients
with PSC were: sensitivity 43 % (95 %CI 35 % – 52 %), specificity 97 % (95 % – 98 %),
PPV 78.2 % (63.6 % – 86.7 %), and negative predictive value (NPV) 87.2 % (85.4 % – 89.1 %).
The authors concluded that bile duct brushing is a simple and highly specific technique
for detecting CCA in patients with PSC. However, the modest sensitivity from bile
duct brushing precludes its utility as a diagnostic tool for early detection of CCA
in patients with PSC. In a recent study of 261 mostly asymptomatic (81 %) patients
with PSC, who had been referred for their first ERC to confirm the diagnosis and to
screen for biliary dysplasia with systematic bile duct brushings, 43 % were found
to have advanced disease, and malignant/suspicious cytology was present in 6.9 % [37].
Table 6
Detection of biliary malignancy in primary sclerosing cholangitis (PSC) using brush
cytology.
|
First author,
Year [ref]
|
Study design
|
Intervention
|
Participants, n
|
Outcomes
|
Results
|
|
Sensitivity
|
Specificity
|
PPV
|
NPV
|
|
Ponsioen,
1999 [86]
|
Prospective
|
ERCP with brush cytology from dominant strictures
|
43
|
Detection of malignancy/CCA
|
60 %
|
89 %
|
59 %
|
89 %
|
|
Lindberg,
2002 [87]
|
Prospective
|
Brush cytology + DNA flow cytometry from biliary strictures
|
57
|
Detection of malignancy/CCA
|
71 %
|
100 %
|
NA
|
NA
|
|
Siqueira,
2002 [88]
|
Retrospective
|
Brush cytology from bile ducts
|
151
|
Detection of malignancy/CCA
|
46.4 %
|
100 %
|
NA
|
NA
|
|
Lal,
2004 [89]
|
Retrospective
|
Brush cytology from bile ducts
|
21
|
Detection of malignancy/CCA
|
67 %
|
94 %,
|
NA
|
NA
|
|
Furmanczyk,
2005 [90]
|
Retrospective
|
Brush cytology from bile ducts
|
51
|
Detection of malignancy/CCA
|
62.5 %
|
100 %
|
NA
|
NA
|
|
Boberg,
2006 [91]
|
Prospective
|
Brush cytology from biliary strictures
|
61
|
Detection of malignancy/CCA
|
100 %
|
84 %
|
68 %
|
100 %
|
|
Moff,
2006 [92]
|
Retrospective
|
Brush cytology from bile ducts
|
47
|
Detection of malignancy/CCA
|
50 %
|
91 %
|
NA
|
NA
|
|
Moreno Luna,
2006 [93]
|
Prospective
|
Brush cytology from biliary strictures
|
86 PSC
|
Detection of malignancy/CCA
|
18 %
|
100 %
|
100 %
|
83 %
|
|
Charatcharoenwitthaya, 2008, [40]
|
Prospective
|
Brush cytology from biliary strictures
|
230
|
Detection of malignancy/CCA
|
8 %
|
100 %
|
100 %
|
89 %
|
|
Levy,
2008 [39]
|
Prospective
|
Brush cytology from biliary strictures
|
32 PSC
|
Detection of malignancy/CCA
|
7 %
|
100 %
|
NA
|
NA
|
|
Halme,
2012 [94]
|
Retrospective
|
Brush cytology from bile ducts
|
102
|
Detection of dysplasia/CCA
|
46 %
|
88 %
|
86 %
|
52 %
|
CCA, cholangiocarcinoma; PPV, positive predictive value; NPV, negative predictive
value; ERCP, endoscopic retrograde cholangiopancreatography; NA, not available.
Table 7
Detection of biliary malignancy in primary sclerosing cholangitis (PSC) using brush
cytology: meta-analyses and reviews.
|
First author, Year [ref]
|
Study design
|
Intervention
|
Patients,n
|
Outcomes
|
Results
|
Comments
|
|
Sensitivity, %
|
Specificity, %
|
Positive likelihood ratio (PLR)
|
Negative likelihood ratio (NLR)
|
|
Trikudanathan, 2014 [95]
|
Meta-analysis including 11 studies (prospective and retrospective)
|
Bile duct brushing
|
747
|
Diagnostic yield of bile duct brushing in diagnosing CCA in PSC strictures
|
43 %
|
97 %
|
8.87
|
0.56
|
The moderate sensitivity in detecting CCA, precludes its utility as a surveillance
tool for early diagnosis of CCA.
|
|
Navaneethan, 2014 [96]
|
Meta-analysis including 4 studies (prospective and retrospective)
|
FISH
|
629
|
Diagnostic yield of FISH in diagnosing CCA in PSC strictures
|
31 %
|
71 %
|
1.19
|
0.95
|
FISH positivity has reasonable diagnostic accuracy; however, the specificity is poor.
|
|
Meta-analysis including 6 studies (prospective and retrospective)
|
FISH polysomy
|
690
|
Diagnostic yield of FISH polysomy in diagnosing CCA in PSC strictures
|
51 %
|
93 %
|
6.81
|
0.56
|
FISH polysomy is highly specific; however, it has limited sensitivity.
|
|
Navaneethan, 2014 [97]
|
Meta-analysis including 9 studies (prospective and retrospective)
|
Intraductal biopsy
|
730
|
Diagnostic yield of intraductal biopsies performed during ERCP
|
48 %
|
99 %
|
18.9
|
0.54
|
Limited sensitivity
|
|
Meta-analysis including 9 studies (prospective and retrospective)
|
Brush cytology
|
730
|
Diagnostic yield of brush cytology performed during ERCP
|
45 %
|
99 %
|
15.7
|
0.54
|
Limited sensitivity
|
|
Meta-analysis including 6 studies (prospective and retrospective)
|
Intraductal biopsy AND
Brush cytology
|
628
|
Diagnostic yield of both brush cytology and intraductal biopsies performed during
ERCP
|
59 %
|
100 %
|
53.8
|
0.42
|
Brushings and biopsy are comparable and have limited sensitivity.
|
|
Walker,
2007 [98]
|
Systematic review
|
MRI
|
NA
|
Diagnosing cholangiocarcinoma in PSC
|
NA
|
|
|
|
Lack of evidence
|
|
Systematic review
|
CT
|
45
|
Diagnosing cholangiocarcinoma in PSC
|
82 %
|
80 %
|
4.10
|
0.25
|
CT provides good sensitivity and specificity in detecting biliary tract carcinoma
complicating PSC.
|
FISH, fluorescence in situ hybridization: CCA, cholangiocarcinoma; ERCP, endoscopic
retrograde cholangiopancreatography; MRI, magnetic resonance imaging; CT, computed
tomography; NA, not available.
Addition of FISH analysis of cytology specimens enhanced the sensitivity for detecting
CCA in patients with PSC in several patient series [39]
[40]
[93]
[94]. The ideal modality (e. g. FISH vs. digital image analysis vs. flow cytometry) and
the appropriate threshold values for markers assessed by each of these modalities
have not been robustly established, and this makes meta-analysis of available data
challenging [96]. For this reason, chromosomal assessments can so far only be recommended in equivocal
cases [96]. As DNA technologies evolve, new markers are likely to emerge.
Ductal biopsy
Ductal biopsy has been shown to improve sensitivity, specificity, and accuracy in
diagnosing CCA compared to brush cytology alone [99]. Since the sampling area for ductal biopsies is limited, complementary biliary brushings
should be considered in all patients. In published studies the sensitivity for the
detection of CCA by ductal biopsy varies from 30 % to 88 % and the specificity from
97 % to 100 % [100]. Combined brush cytology and biopsy has a sensitivity varying from 47 % to 86 %
and specificity from 97 % to 100 %. A retrospective study [100] assessed the accuracy of triple modality testing, namely brush cytology, biopsy,
and FISH, and their combinations, in one patient group, and the accuracy of brush
cytology alone in a separate patient group. It demonstrated that brush cytology alone
had a sensitivity of 42 %, specificity 100 %, PPV 100 %, and NPV 88 %. The triple
sample assessment modality markedly improved the overall sensitivity (82 %), with
similar specificity (100 %), PPV (100 %), and NPV (87 %).
Cholangioscopy
Peroral cholangioscopy (POCS) allows direct visualization of extrahepatic bile duct
strictures. The recent development of video-based systems provides better image resolution
and offers clearer views than fiberoptic cholangioscopy. Compared to ERC and tissue
sampling, POCS was shown to improve diagnostic accuracy [101]
[102]
[103]. However, these studies were not focused on CCA in PSC patients.
Single-operator cholangioscopy (SpyGlass) is gaining popularity, primarily for stone
treatment and assessment of indeterminate strictures. Its utility in PSC was studied
in a recent case series [104], with visual assessment and targeted biopsies of 64 strictures in 47 patients. Only
1 of 3 patients with CCA were diagnosed by the ERCP procedure. It is likely that newer
digital versions of this instrument (e. g. SpyGlass DS) will perform better, at least
in terms of visual diagnostics.
Other techniques
Other techniques such as intraductal ultrasonography and confocal laser endomicroscopy
have shown potential utility in the diagnosis of CCA in PSC, but are not established
in routine clinical practice. Regular endoscopic ultrasonography with sampling of
detectable masses or locoregional lymph nodes is advocated by some, but such sampling
is also regarded as a contraindication to liver transplantation in some centers; thus
any such sampling should be discussed with local multidisciplinary teams.
Endoscopic surveillance of PSC-associated inflammatory bowel disease (IBD)
The relationship between PSC and IBD is well established [105]. The prevalence of IBD in patients with established PSC varies widely, but is reported
as 80 % in Scandinavian countries [106]. The often asymptomatic phenotype of IBD means that prevalence data are strongly
influenced by the level of proactive search for the disease. The typical scenario
was for IBD to precede the presentation of PSC. However, the clinical presentation
of IBD is variable, and the disease may be subclinical or asymptomatic for years [107] and is nowadays often diagnosed after the recognition of the liver disease. Notably,
IBD may have been present for an unknown period of time when PSC is diagnosed. The
increased risk of colon cancer in PSC-associated IBD [108]
[109] hence makes it crucial to perform a full ileocolonoscopy at the time of PSC diagnosis
in all patients. As to the diagnosis of IBD per se, complete ileocolonoscopy is critical
since rectal sparing, as well as right-sided involvement, is frequent in these patients
[8].
Timing of screening
22. ESGE/EASL recommend screening ileocolonoscopy at the time of PSC diagnosis.
Strong recommendation, high quality evidence.
If IBD is documented endoscopically or histologically, annual surveillance colonoscopies
are warranted.
Strong recommendation, low quality evidence.
23. ESGE/EASL suggest that if no IBD is documented, the next ileocolonoscopy should
be considered at 5 years or whenever bowel complaints suggestive of IBD occur.
Weak recommendation, low quality evidence.
Based on initial screening, subsequent surveillance can be planned. If IBD is documented,
annual colonoscopies are warranted [6]
[110] since it has been shown that PSC-IBD patients whose colorectal cancer (CRC) is detected
in a surveillance program have a significantly lower risk of CRC-related mortality
as compared to non-surveilled patients [78]. If not, repeat colonoscopy should be done with the occurrence of symptoms suggestive
of IBD, or of elevated F-calprotectin, or otherwise at 3 – 5 years [111], although this recommendation lacks any scientific evidence beyond extrapolation
from general IBD recommendations [112].
Endoscopic modality
24. For screening for the presence of IBD, EASL/ESGE recommend ileocolonoscopy with
four-quadrant biopsies from all colonic segments and the terminal ileum.
Strong recommendation, low quality evidence.
25. For dysplasia surveillance of PSC-associated IBD, EASL/ESGE recommend ileocolonoscopy
with dye-based chromoendoscopy with targeted biopsies.
Strong recommendation, low quality evidence.
PSC-associated colitis seems to be distinctive from other IBD: colitis is predominant
in the right colon [113] and colon cancer is typically right-sided [114]. Lack of inflammation in the rectum (“rectal sparing”) is reported in some studies
but less frequently observed in others [3]. Endoscopic surveillance of PSC-associated colitis is presumed to increase the chance
of early detection of dysplasia or malignancy [115].
Screening for IBD at diagnosis of PSC is best performed by high definition ileocolonoscopy
with four-quadrant biopsies from all colonic segments and the terminal ileum. Biopsies
should be taken at the index endoscopy even without macroscopic signs of inflammation
[111]
[116]
[117].
In established PSC-IBD, ileocolonoscopy with dye-based chromoendoscopy (0.1 % methylene
blue or 0.1 % – 0.5 % indigo carmine) with targeted biopsies is required for neoplasia
surveillance of PSC-associated IBD. In appropriately trained hands, in the situation
of quiescent disease activity and adequate bowel preparation, nontargeted four-quadrant
biopsies can be abandoned [118]. This approach is also endorsed by the European Crohn’s and Colitis Organisation
(ECCO) [112]. It should be noted that there are no studies on colonic neoplasia surveillance
specifically in the setting of PSC-associated IBD.
Routine use of pancolonic chromoendoscopy with targeted biopsies for neoplasia surveillance
in patients with long-standing colitis (disease duration of > 8 years) increased the
proportion of patients found with dysplasia by a factor of 2.1 – 3.3 compared to standard
definition videocolonoscopy. For the detection of patients with neoplasia, the pooled
incremental yield of conventional chromoendoscopy with random biopsies over standard
white-light endoscopy with random biopsies was 7 % (95 %CI 3.2 % – 11.3 %) [119]. The benefit of conventional chromoendoscopy over white-light endoscopy with latest-generation
high definition colonoscopes is unknown to date.
Handling of polyps and colorectal dysplasia
26. ESGE/EASL recommend endoscopic resection of any visible lesions and assessment
of the surrounding mucosa. We recommend proctocolectomy in the case of dysplasia in
the surrounding mucosa, or when the lesion cannot be completely resected. Otherwise,
repeat colonoscopy and close follow-up is warranted.
Strong recommendation, low quality evidence.
27. In the case of invisible lesions with high grade dysplasia (HGD) confirmed by
two expert pathologists, proctocolectomy should be advised
Strong recommendation, low quality evidence.
28. In the case of invisible lesions with low grade dysplasia (LGD) confirmed by two
expert pathologists, repeat colonoscopy after 3 months with chromoendoscopy is recommended.
Strong recommendation, low quality evidence.
Colorectal cancer (CRC) risk is significantly increased in patients with coexisting
IBD and PSC. A meta-analysis of 11 studies concluded that patients with ulcerative
colitis and PSC were at increased risk of developing CRC compared to patients with
ulcerative colitis alone (OR 4.09; 95 %CI 2.89 – 5.76) [109]. A recent large population-based study in the Netherlands found a 9-fold increased
risk of developing CRC in PSC-ulcerative colitis patients, compared to the age- and
gender-matched population (standardized incidence ratio [SIR] 8.6, 95 %CI 3.5 – 17.7),
and a 10-fold increased risk, compared to ulcerative colitis controls (ratio of SIRs
9.8, 95 %CI 1.9 – 96.6) [78].
Most dysplasia is visible at colonoscopy [120]
[121]. On the other hand, invisible dysplastic lesions can also be diagnosed by random
biopsies during surveillance. According to the IBD Dysplasia Morphology Study Group
[122], dysplasia is subdivided into LGD and HGD.
Recent ECCO guidelines state that a visible lesion with dysplasia should be completely
resected endoscopically irrespective of the grade of dysplasia or the location relative
to the inflamed mucosal areas [112]. Subsequently, the surrounding mucosa (around the visible lesion) should be examined
(with chromoendoscopy-guided targeted biopsies or random biopsies if chromoendoscopy
is not available). If endoscopic resection is incomplete or impossible, or if dysplasia
is detected in the surrounding mucosa, total proctocolectomy is recommended.
In the case of invisible lesions with LGD, urgent repeat chromoendoscopy should be
performed, to eventually identify a well-circumscribed lesion and/or perform additional
random biopsies. If the presence of LGD is confirmed, there is no clear consensus
regarding management; proctocolectomy or surveillance could be recommended. Actually,
two studies revealed a significant 5-year progression rate (33 % – 54 %) of LGD to
HGD [123]
[124], whereas others showed low progression rates [125]
[126]. Finally, in the case of invisible lesions with HGD or adenocarcinoma, total proctocolectomy
is indicated.
This Guideline from ESGE and EASL represents a consensus of best practice based on
the available evidence at the time of preparation. The recommendations might not apply
in all situations and should be interpreted in the light of specific clinical situations
and resource availability. Further controlled clinical studies may be needed to clarify
aspects of the Guideline, and revision may be necessary as new data appear. Clinical
considerations may justify a course of action at variance to these recommendations.
This ESGE/EASL Guideline is intended to be an educational device to provide information
that may assist endoscopists in providing care to patients. It is not a set of rules
and should not be construed as establishing a legal standard of care or as encouraging,
advocating, requiring, or discouraging any particular treatment.
