Introduction
During the last several decades, many advances in technology have rendered peroral
cholangioscopy (POC) a useful diagnostic and therapeutic technique. POC is conducted
during endoscopic retrograde cholangiopancreatography (ERCP) in one of three ways:
with a dual-operator dedicated (“mother – daughter”) cholangioscopic system, with
a single-operator catheter-based cholangioscopic system (SOC), or directly with an
ultraslim endoscope or slim gastroscope. The procedures vary with respect to number
of operators, maneuverability, image quality, and method of access, resulting in variable
success rates.
POC is most commonly used for treating difficult bile duct stones with electrohydraulic
lithotripsy or laser lithotripsy or for directly visualizing and/or sampling indeterminate
biliary strictures. Other indications and reported uses for POC include, but are not
limited to, placing a guidewire during ERCP, monitoring primary sclerosing cholangitis,
facilitating stent placement for biliary drainage, assessing the extent of biliary
malignancy before surgery, and staging and ablating biliary tumors [1]
[2]
[3]
[4]. POC is a safe procedure associated with a low adverse event rate. Variable results
have been published in regard to its efficacy and safety for these indications [5]. As such, the aim of this study was to perform a systematic review and meta-analysis
to assess (i) the overall clinical efficacy of POC for the therapy of difficult bile
duct stones, (ii) the accuracy of POC for diagnosing indeterminate biliary strictures,
and (iii) the overall adverse event rate of POC.
Patients and methods
This review and meta-analysis was performed in accordance with the Preferred Reporting
Items for Systematic Reviews and Meta-Analysis (PRISMA) statement [6].
Information sources and medical literature search
A search for eligible publications was conducted via Ovid Medline, the Cochrane Library,
and Scopus with the following key words: cholangiopancreatoscopy, choledochoscopy,
pancreatocholangioscopy, cholangioscopy, and pancreatoscopy. Two authors (P. K. and
S. K.) independently conducted a medical literature search and screened the resulting
studies for inclusion. One reviewer (P. K.) extracted data from all studies that met
inclusion criteria and stored relevant data in an Excel (Microsoft, Redmond, Washington,
USA) database, and a second reviewer (S. K.) performed a second pass of data entry.
A third reviewer (S. W.) resolved any discrepancies. EndNote X7 (Thomson Reuters,
New York, New York, USA) was used for reference management.
Eligibility criteria
For the systematic review, our search included all clinical studies evaluating POC
until December 2014.
Inclusion criteria were as follows: (i) studies that investigated POC for the removal
of difficult bile duct stones, (ii) studies that investigated POC and its ability
to help diagnose indeterminate biliary strictures, (iii) studies that enrolled more
than 10 participants, and (iv) full-text articles in English. Notably, difficult bile
duct stones were most often defined as stones that could not be removed via conventional
methods (ERCP with standard extraction balloons, baskets, or lithotriptors; large
endoscopic papillary balloon dilation). Indeterminate biliary strictures were most
often defined as strictures that could not be definitively diagnosed with conventional
ERCP sampling techniques (brushings, intraductal biopsy).
Exclusion criteria were as follows: (i) case reports, (ii) abstracts, (iii) reviews,
(iv) letters to authors or editors, (v) studies evaluating percutaneous cholangioscopy,
(vi) animal studies, and (vii) studies evaluating pancreatoscopy only.
Quality assessment
A modified Newcastle-Ottawa Scale [7] was employed to assess the methodological quality of each study included in this
review. The studies were divided into two groups: those in which biliary stone removal
was an indication for POC and those in which POC was used for the diagnosis of indeterminate
strictures; it should be noted that these two groups of studies are not mutually exclusive.
The scale assessed the following for “Selection” criteria: (i) representativeness
of the exposed cohort, (ii) ascertainment of exposure, and (iii) demonstration that
the outcome of interest was not present at the start of the study. The scale also
assessed the following for “Outcome” criteria: (i) assessment by record linkage; (ii)
follow-up length, which was determined to be an average follow-up in the study of
at least 6 months for both the evaluation of recurrent stones and clinical follow-up
for indeterminate strictures; and (iii) percentage of patients lost to follow-up,
which was determined to be less than 15 %. Follow-up length and percentage of patients
who were lost to follow-up were not used for studies evaluating biliary stone clearance
because these factors are not commonly assessed in patients after stone removal.
Thus, according to the modified Newcastle-Ottawa Scale that was used, studies evaluating
outcomes of POC for difficult bile duct stones could receive a maximum of four points,
and studies evaluating outcomes of POC for indeterminate strictures could receive
a maximum of six points. Any question regarding the allocation of points for each
study was discussed by three reviewers (P. K., S. K., and S. W.).
List of items and data collected
The following data elements were extracted (if available) from each study included
in the review: (i) publication year; (ii) number of centers involved (single center
or multicenter); (iii) setting (university, multicenter, or community); (iv) study
design (prospective, retrospective, or randomized controlled trial); (v) type of cholangioscopy
(peroral dual-operator dedicated cholangioscope, peroral catheter-based cholangioscope
[SpyGlass; Boston Scientific, Natick, Massachusetts, USA], direct peroral cholangioscope
or ultraslim endoscope); (vi) study focus (stones, strictures, or both); (vii) sample
size; (viii) number of POC procedures attempted; (ix) POC technical success rate (i. e.,
number of successful POC procedures divided by number attempted POC procedures); (x)
adverse event rate; (xi) number of patients lost to follow up; and (xii) follow-up
period (mean).
For studies evaluating the outcomes of POC for difficult bile duct stones, additional
data included the following: (i) number of patients undergoing stone removal (denominator
for stone clearance rate); (ii) stone clearance rate (rate of complete stone clearance,
not including partial clearance); (iii) average number of stones per patient (mean);
(iv) average stone size in millimeters (mean); (v) location of more than 75 % of stones
(extrahepatic, intrahepatic, cystic, or mixed); (vi) stone removal technique (cholangioscopy-assisted
basket or balloon, electrohydraulic lithotripsy, laser lithotripsy, or multiple methods);
and (vii) stone recurrence rate.
For studies in which the outcomes of POC for indeterminate strictures were determined
by visual impression only, additional relevant data included the following: (i) number
of patients involved in the diagnostic study (denominator for accuracy), (ii) number
of patients with true malignant disease (denominator for sensitivity), (iii) number
of patients with true benign disease (denominator for specificity), (iv) sensitivity,
(v) specificity, (vi) positive predictive value, (vii) negative predictive value,
and (viii) accuracy.
For studies in which the outcomes of POC for indeterminate strictures were determined
by directed tissue sampling, additional relevant data included the following: (i)
number of patients or biopsy samples involved in the diagnostic study (denominator
for accuracy), (ii) mean number of biopsy samples per patient/procedure, (iii) number
of patients with true malignant disease (denominator for sensitivity), (iv) number
of patients with true benign disease (denominator for specificity), (v) sensitivity,
(vi) specificity, (vii) positive predictive value, (viii) negative predictive value,
and (ix) accuracy.
Outcomes measured
The primary outcomes for studies evaluating POC for difficult bile duct stone included
the following: (i) technical success rate (ability to achieve selective bile duct
access), (ii) stone clearance rate, and (iii) stone recurrence rate. The primary outcomes
for studies evaluating POC for indeterminate strictures included the following: (i)
technical success rate (ability to achieve selective bile duct access), (ii) accuracy
(both visual and directed tissue sampling), (iii) sensitivity (both visual and directed
tissue sampling), and (iv) specificity (both visual and directed tissue sampling).
The overall adverse event rate related to POC was determined.
Statistical analysis and summary measures
Comprehensive Meta-Analysis Software v2.0 (Biostat, Englewood, New Jersey, USA) was
used for all formal meta-analyses (when the number of studies was more than five)
to obtain summary estimates of proportions (stone clearance rate, technical success
rates, stone recurrence rate, adverse event rates, sensitivities, specificities, and
accuracy rates). Because of the assumption of inherently different study scenarios
and study populations, a random effects model for all analyses was assumed. Heterogeneity
across studies via a chi-squared test on the Q-statistic with appropriate degrees
of freedom (dependent on outcome because not all studies uniformly reported all outcomes
of interest) and the estimated measure of excess-to-total variation (I
2) across studies for each outcome of interest were also calculated. In instances in
which the degrees of freedom were sufficiently large and there was significant evidence
of between-study variation (i. e., heterogeneity), meta-regression to examine potential
sources of between-study variation was performed.
Publication bias was assessed via funnel plots and Egger’s test on the regression
intercept for these plots. In instances of significant evidence of publication bias
(P < 0.05), imputed studies were used to create adjusted summary estimates for each
measure. Other factors, such as differences in trial quality and true study heterogeneity,
could produce asymmetry in funnel plots.
Results
Literature search and included studies
The outlined search strategy resulted in the identification of a total of 1028 studies.
Based on the defined inclusion and exclusion criteria, a total of 49 studies [8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
[34]
[35]
[36]
[37]
[38]
[39]
[40]
[41]
[42]
[43]
[44]
[45]
[46]
[47]
[48]
[49]
[50]
[51]
[52]
[53]
[54]
[55]
[56] were included in the analysis ([Fig. 1]). Of the 49 studies evaluated, 33 contained data on difficult bile duct stones ( [Table 1]) and 29 studies contained data on indeterminate strictures ([Table 2]); there were 20 studies focusing only on difficult bile duct stones, 16 studies
only on indeterminate strictures, and 13 studies on both.
Fig. 1 Flow chart of the selection of relevant studies. POC, peroral cholangioscopy.
Table 1
Characteristics of the stone studies included in a systematic review and meta-analysis
of the efficacy of peroral cholangioscopy for difficult bile duct stones and indeterminate
strictures.
|
First author
|
Year
|
Setting
|
Study design
|
Type of POC
|
Sample size, n
|
Technical success rate
|
Patients undergoing stone removal, n
|
Stone clearance rate
|
Stones per patient, mean, n
|
Stone size, mean, mm
|
Location of > 75 % of stones
|
Stone removal method
|
Stone recurrence rate
|
Complication/adverse event rate
|
Patients lost to follow-up, n
|
NOS score
|
|
Akerman
|
2012
|
Single
|
Retrospective
|
Catheter-based
|
34
|
0.97
|
11
|
0.64
|
NR
|
NR
|
NR
|
EHL
|
NR
|
0
|
NR
|
4
|
|
Alameel
|
2013
|
Single
|
Prospective
|
Catheter-based
|
30
|
NR
|
10
|
0.9
|
NR
|
NR
|
NR
|
EHL
|
NR
|
0.05
|
0
|
4
|
|
Arya
|
2004
|
Multicenter
|
Retrospective
|
Mother – daughter
|
94
|
NR
|
94
|
0.9
|
1.92
|
0
|
Mixed
|
EHL
|
0.04
|
0.18
|
NR
|
4
|
|
Awadallah
|
2006
|
Single
|
Prospective
|
Mother – daughter
|
41
|
NR
|
9
|
0.78
|
NR
|
NR
|
Mixed
|
EHL
|
NR
|
0.05
|
1
|
4
|
|
Chen
|
2011
|
Multicenter
|
Prospective
|
Catheter-based
|
297
|
0.983
|
66
|
0.92
|
NR
|
NR
|
Extrahepatic
|
Laser lithotripsy
|
NR
|
0.075
|
20
|
4
|
|
Chen
|
2007
|
Multicenter
|
Prospective
|
Catheter-based
|
35
|
NR
|
9
|
1
|
NR
|
NR
|
NR
|
Multiple methods
|
NR
|
0.06
|
0
|
4
|
|
Draganov
|
2011
|
Single
|
Prospective
|
Catheter-based
|
75
|
0.933
|
26
|
0.923
|
3.55
|
16.52
|
NR
|
EHL
|
NR
|
0.048
|
0
|
4
|
|
Farnik
|
2014
|
Multicenter
|
Retrospective
|
Ultraslim endoscope
|
89
|
0.885
|
23
|
NR
|
NR
|
NR
|
NR
|
Multiple methods
|
NR
|
0.077
|
NR
|
3
|
|
Farrell
|
2005
|
Single
|
Prospective
|
Catheter-based
|
75
|
NR
|
26
|
1
|
NR
|
20
|
Mixed
|
EHL
|
NR
|
0
|
NR
|
4
|
|
Fishman
|
2009
|
Single
|
Retrospective
|
Catheter-based
|
128
|
NR
|
41
|
0.87
|
NR
|
NR
|
NR
|
EHL
|
NR
|
0
|
NR
|
4
|
|
Huang
|
2013
|
Single
|
Prospective
|
Ultraslim endoscope
|
22
|
0.82
|
5
|
1
|
NR
|
13.4
|
NR
|
POC-assisted basket
|
0.182
|
0
|
0
|
4
|
|
Itoi
|
2012
|
Single
|
Retrospective
|
Ultraslim endoscope
|
24
|
NR
|
8
|
1
|
NR
|
12
|
Intrahepatic
|
POC-assisted basket
|
NR
|
0
|
0
|
4
|
|
Itoi
|
2010
|
Single
|
Retrospective
|
Mother – daughter
|
108
|
NR
|
26
|
1
|
2.4
|
14.6
|
NR
|
Multiple methods
|
NR
|
0
|
NR
|
4
|
|
Itoi
|
2014
|
Multicenter
|
Prospective
|
Ultraslim endoscope
|
41
|
0.83
|
8
|
1
|
NR
|
NR
|
NR
|
Multiple methods
|
NR
|
0.048
|
NR
|
4
|
|
Jakobs
|
2007
|
Single
|
Prospective
|
Mother – daughter
|
89
|
NR
|
17
|
0.824
|
NR
|
22
|
NR
|
Laser lithotripsy
|
NR
|
0
|
NR
|
3
|
|
Jakobs
|
1996
|
Single
|
Prospective
|
Mother – daughter
|
30
|
NR
|
10
|
0.83
|
2.7
|
18
|
Mixed
|
Laser lithotripsy
|
NR
|
NR
|
NR
|
4
|
|
Kalaitzakis
|
2012
|
Multicenter
|
Retrospective
|
Catheter-based
|
165
|
0.95
|
33
|
0.73
|
NR
|
18
|
Extrahepatic
|
Multiple methods
|
NR
|
0.09
|
4
|
4
|
|
Kim
|
2011
|
Single
|
Prospective
|
Ultraslim endoscope
|
13
|
0.923
|
13
|
0.923
|
2.4
|
20.9
|
NR
|
Laser lithotripsy
|
NR
|
0.077
|
0
|
4
|
|
Lee TY
|
2012
|
Single
|
Prospective
|
Ultraslim endoscope
|
10
|
NR
|
10
|
0.9
|
2.3
|
19
|
Extrahepatic
|
Laser lithotripsy
|
NR
|
0.1
|
0
|
4
|
|
Lee YN
|
2012
|
Single
|
Prospective
|
Ultraslim endoscope
|
48
|
0.958
|
13
|
0.846
|
2.6
|
16.7
|
Extrahepatic
|
POC-assisted basket
|
NR
|
0
|
0
|
4
|
|
Maydeo
|
2011
|
Single
|
Prospective
|
Catheter-based
|
64
|
NR
|
60
|
1
|
1.5
|
23.4
|
Extrahepatic
|
Laser lithotripsy
|
NR
|
0.133
|
0
|
4
|
|
Meves
|
2014
|
Single
|
Prospective
|
Ultraslim endoscope
|
84
|
0.87
|
11
|
1
|
NR
|
NR
|
NR
|
Multiple methods
|
NR
|
0.12
|
NR
|
4
|
|
Moon
|
2009
|
Single
|
Prospective
|
Ultraslim endoscope
|
18
|
0.944
|
18
|
0.89
|
2.3
|
23.2
|
Extrahepatic
|
Multiple methods
|
NR
|
0
|
0
|
4
|
|
Moon
|
2009
|
Single
|
Prospective
|
Ultraslim endoscope
|
29
|
0.78
|
4
|
1
|
NR
|
NR
|
NR
|
Multiple methods
|
NR
|
0
|
NR
|
4
|
|
Mori
|
2012
|
Single
|
Prospective
|
Ultraslim endoscope
|
40
|
0.925
|
13
|
1
|
NR
|
NR
|
NR
|
Multiple methods
|
NR
|
0
|
NR
|
4
|
|
Neuhaus
|
1993
|
Single
|
Prospective
|
Mother – daughter
|
35
|
NR
|
12
|
0.83
|
NR
|
20
|
Extrahepatic
|
Laser lithotripsy
|
NR
|
0
|
NR
|
4
|
|
Patel
|
2014
|
Multicenter
|
Prospective
|
Catheter-based
|
69
|
NR
|
69
|
0.97
|
NR
|
NR
|
Extrahepatic
|
Laser lithotripsy
|
NR
|
0.041
|
0
|
4
|
|
Piraka
|
2007
|
Single
|
Prospective
|
Mother – daughter
|
32
|
NR
|
32
|
0.81
|
NR
|
12
|
Mixed
|
EHL
|
0.18
|
0.038
|
4
|
4
|
|
Pohl
|
2013
|
Single
|
RCT
|
Mixed
|
60
|
0.88
|
NR
|
NR
|
NR
|
NR
|
NR
|
Multiple methods
|
NR
|
0.117
|
0
|
3
|
|
Sauer
|
2013
|
Single
|
Retrospective
|
Mixed
|
20
|
NR
|
20
|
0.9
|
2.2
|
22
|
Extrahepatic
|
Laser lithotripsy
|
NR
|
0.25
|
NR
|
4
|
|
Sepe
|
2012
|
Single
|
Retrospective
|
Catheter-based
|
13
|
NR
|
13
|
0.769
|
NR
|
8
|
Cystic
|
EHL
|
0.077
|
0
|
NR
|
4
|
|
Tsuyuguchi
|
2011
|
Single
|
Prospective
|
Mother – daughter
|
122
|
NR
|
122
|
0.959
|
2.9
|
17
|
NR
|
Multiple methods
|
0.161
|
NR
|
6
|
3
|
|
Tsuyuguchi
|
2000
|
Single
|
Retrospective
|
Mother – daughter
|
25
|
0.92
|
22
|
0.82
|
NR
|
20
|
NR
|
Multiple methods
|
0.18
|
0.16
|
1
|
4
|
POC, peroral cholangioscopy; NR, not reported; EHL, electrohydraulic lithotripsy;
NOS, Newcastle – Ottawa Scale.
Table 2
Characteristics of the stricture studies included in a systematic review and meta-analysis
of the efficacy of peroral cholangioscopy for difficult bile duct stones and indeterminate
strictures.
|
First author
|
Year
|
Setting
|
Study design
|
Type of POC
|
Sample size
|
Technical success rate
|
Patients involved (VISUAL), n
|
Stricture sensitivity (VISUAL)
|
Stricture specificity (VISUAL)
|
Stricture accuracy (VISUAL)
|
Patients involved (BIOPSY), n
|
Biopsy samples per patient, mean, n
|
Stricture sensitivity (BIOPSY)
|
Stricture specificity (BIOPSY)
|
Stricture accuracy (BIOPSY)
|
Complication/adverse event rate
|
Patients lost to follow-up, n
|
Duration of follow-up, mean, mo
|
NOS score
|
|
Akerman
|
2012
|
Single
|
Retrospective
|
Catheter-based
|
34
|
0.97
|
0
|
NR
|
NR
|
NR
|
0
|
NR
|
NR
|
NR
|
NR
|
0
|
NR
|
0
|
3
|
|
Alameel
|
2013
|
Single
|
Prospective
|
Catheter-based
|
30
|
NR
|
19
|
0.83
|
0.84
|
0.84
|
16
|
NR
|
0.4
|
1
|
0.81
|
0.05
|
0
|
5
|
5
|
|
Albert
|
2011
|
Single
|
Prospective
|
Ultraslim endoscope
|
22
|
0.88
|
0
|
NR
|
NR
|
NR
|
0
|
NR
|
NR
|
NR
|
NR
|
0.045
|
NR
|
0
|
3
|
|
Awadallah
|
2006
|
Single
|
Prospective
|
Mother – daughter
|
41
|
NR
|
0
|
NR
|
NR
|
NR
|
0
|
NR
|
NR
|
NR
|
NR
|
0.05
|
1
|
0
|
5
|
|
Chen
|
2011
|
Multicenter
|
Prospective
|
Catheter-based
|
297
|
0.983
|
95
|
0.78
|
0.82
|
0.8
|
95
|
3
|
0.49
|
0.98
|
0.75
|
0.075
|
20
|
> 6
|
6
|
|
Chen
|
2007
|
Multicenter
|
Prospective
|
Catheter-based
|
35
|
NR
|
20
|
1
|
0.77
|
0.85
|
20
|
4.5
|
0.71
|
1
|
0.9
|
0.06
|
0
|
> 6
|
6
|
|
Draganov
|
2011
|
Single
|
Prospective
|
Catheter-based
|
75
|
0.933
|
0
|
NR
|
NR
|
NR
|
0
|
NR
|
NR
|
NR
|
NR
|
0.048
|
0
|
0
|
3
|
|
Draganov
|
2012
|
Single
|
Prospective
|
Catheter-based
|
26
|
1
|
0
|
NR
|
NR
|
NR
|
26
|
NR
|
0.765
|
1
|
0.846
|
0.077
|
0
|
21.78
|
6
|
|
Farnik
|
2014
|
Multicenter
|
Retrospective
|
Ultraslim endoscope
|
89
|
0.885
|
0
|
NR
|
NR
|
NR
|
0
|
NR
|
NR
|
NR
|
NR
|
0.077
|
NR
|
0
|
3
|
|
Fishman
|
2009
|
Single
|
Retrospective
|
Catheter-based
|
128
|
NR
|
0
|
NR
|
NR
|
NR
|
0
|
NR
|
NR
|
NR
|
NR
|
0
|
NR
|
0
|
3
|
|
Fukuda
|
2005
|
Single
|
Retrospective
|
Mother – daughter
|
97
|
1
|
76
|
1
|
0.87
|
0.934
|
0
|
NR
|
NR
|
NR
|
NR
|
0.02
|
NR
|
> 12
|
6
|
|
Hartman
|
2012
|
Single
|
Retrospective
|
Catheter-based
|
89
|
NR
|
15
|
0.88
|
0.86
|
0.87
|
29
|
3
|
0.57
|
1
|
0.78
|
NR
|
3
|
23
|
5
|
|
Itoi
|
2014
|
Multicenter
|
Prospective
|
Ultraslim endoscope
|
41
|
0.83
|
0
|
NR
|
NR
|
NR
|
0
|
NR
|
NR
|
NR
|
NR
|
0.048
|
NR
|
0
|
3
|
|
Itoi
|
2010
|
Multicenter
|
Retrospective
|
Mother – daughter
|
144
|
NR
|
0
|
NR
|
NR
|
NR
|
0
|
1.6
|
NR
|
NR
|
NR
|
0.07
|
0
|
> 12
|
6
|
|
Kalaitzakis
|
2012
|
Multicenter
|
Retrospective
|
Catheter-based
|
165
|
0.95
|
0
|
NR
|
NR
|
NR
|
49
|
3
|
0.62
|
1
|
0.84
|
0.09
|
4
|
15
|
5
|
|
Khan
|
2013
|
Single
|
Retrospective
|
NA
|
66
|
NR
|
0
|
NR
|
NR
|
NR
|
66
|
NR
|
0.487
|
0.963
|
0.68
|
NR
|
0
|
0
|
3
|
|
Liu
|
2014
|
Multicenter
|
Retrospective
|
Catheter-based
|
25
|
NR
|
0
|
NR
|
NR
|
NR
|
0
|
NR
|
NR
|
NR
|
NR
|
0
|
NR
|
0
|
4
|
|
Manta
|
2013
|
Single
|
Prospective
|
Catheter-based
|
52
|
1
|
0
|
NR
|
NR
|
NR
|
42
|
NR
|
0.88
|
0.94
|
0.9
|
0.038
|
0
|
24
|
6
|
|
Meves
|
2014
|
Single
|
Prospective
|
Ultraslim endoscope
|
84
|
0.87
|
0
|
NR
|
NR
|
NR
|
26
|
NR
|
0.895
|
NR
|
NR
|
0.12
|
NR
|
0
|
4
|
|
Moon
|
2009
|
Single
|
Prospective
|
Ultraslim endoscope
|
29
|
0.78
|
0
|
NR
|
NR
|
NR
|
0
|
NR
|
NR
|
NR
|
NR
|
0
|
NR
|
0
|
3
|
|
Nguyen
|
2013
|
Single
|
Prospective
|
Catheter-based
|
40
|
0.947
|
0
|
NR
|
NR
|
NR
|
18
|
NR
|
NR
|
NR
|
0.89
|
0.05
|
0
|
22
|
6
|
|
Nishikawa
|
2013
|
Single
|
Prospective
|
Mother – daughter
|
33
|
1
|
33
|
1
|
0.917
|
0.97
|
33
|
2.39
|
0.381
|
1
|
0.606
|
0.06
|
0
|
> 12
|
6
|
|
Osanai
|
2013
|
Multicenter
|
Prospective
|
Mother – daughter
|
87
|
1
|
38
|
0.964
|
0.8
|
0.921
|
35
|
2.4
|
0.815
|
1
|
0.857
|
0.069
|
0
|
> 12
|
6
|
|
Pohl
|
2013
|
Single
|
RCT
|
Mixed
|
60
|
0.88
|
0
|
NR
|
NR
|
NR
|
0
|
NR
|
NR
|
NR
|
NR
|
0.117
|
0
|
6
|
6
|
|
Ramchandani
|
2011
|
Single
|
Prospective
|
Catheter-based
|
36
|
1
|
36
|
0.95
|
0.79
|
0.89
|
33
|
3.5
|
0.82
|
0.82
|
0.82
|
0.083
|
0
|
> 6
|
6
|
|
Shah
|
2006
|
Single
|
Prospective
|
Mother – daughter
|
62
|
NR
|
0
|
NR
|
NR
|
NR
|
0
|
NR
|
NR
|
NR
|
NR
|
0.056
|
4
|
12.4
|
6
|
|
Siddiqui
|
2012
|
Single
|
Retrospective
|
Catheter-based
|
30
|
NR
|
0
|
NR
|
NR
|
NR
|
30
|
NR
|
0.77
|
NR
|
NR
|
0.033
|
0
|
> 6
|
6
|
|
Tischendorf
|
2006
|
Single
|
Prospective
|
Mother – daughter
|
53
|
1
|
53
|
0.92
|
0.93
|
0.93
|
0
|
NR
|
NR
|
NR
|
NR
|
0
|
0
|
37
|
6
|
|
Woo
|
2014
|
Single
|
Retrospective
|
Catheter-based
|
32
|
NR
|
31
|
1
|
0.9
|
0.967
|
19
|
2.84
|
0.642
|
1
|
0.736
|
0.094
|
0
|
> 6
|
6
|
POC, peroral cholangioscopy; NR, not reported; NA, not applicable; NOS, Newcastle – Ottawa
Scale.
Efficacy of peroral cholangioscopy for difficult bile duct stones
The overall estimated stone clearance rate (n = 31 studies) was 88 % (95 % confidence
interval [95CI] 85 % – 91 %), without significant evidence of heterogeneity (P = 0.09, I
2 = 26.14) ([Fig. 2]). There was evidence of publication bias (P = 0.0466) in this analysis. Imputed values would fall below the estimated mean rate
with larger standard errors, and the adjusted stone clearance rate according to the
trim and fill method of Duval and Tweedie [57] is 85 % (95 %CI 82 % – 88 %). Study year, study design, stone size, stone location,
number of stones, and type of POC had no impact on stone clearance rates based on
meta-regression analysis with regard to stone clearance.
Fig. 2 Forest plot of studies reporting bile duct stone clearance rate with peroral cholangioscopy.
Pooled clearance rate was 88 % (95 % confidence interval [CI] 85 % – 91 %).
The estimated stone recurrence rate (n = 6 studies) was 13 % (95 %CI 7 % – 20 %) ( [Fig. 3]) with no evidence of heterogeneity (P = 0.13, I
2 = 40.09) or publication bias (P = 0.55). The estimated technical success rate (n = 15 studies) was 91 % (95 %CI 88 % – 94 %)
( [Fig. 4]), with evidence of heterogeneity (P < 0.01, I
2 = 61.72). Meta-regression identified a significant association between the type of
POC used and technical success rates, with SOC demonstrating higher technical success
rates compared with other methods (P < 0.01) ([Fig. 5]).
Fig. 3 Forest plot of studies reporting stone recurrence rate after clearance by peroral
cholangioscopy. Pooled recurrence rate was 13 % (95 % confidence interval [CI] 7 % – 20 %).
Fig. 4 Forest plot of studies reporting technical success rate of peroral cholangioscopy
for stone-related indications. Pooled success rate was 91 % (95 % confidence interval
[CI] 88 % – 94 %).
Fig. 5 Relationship between technical success rate for stone-related indications and type
of peroral cholangioscopy (POC). Single-operator catheter-based cholangiography had
a higher rate of technical success for stone-related indications compared with other
methods.
Efficacy of peroral cholangioscopy for indeterminate strictures
The diagnostic characteristics of POC for visual impression were as follows ([Table 3]): accuracy (n = 10 studies), 89 % (95 %CI 84 % – 93 %) ([Fig. 6]); sensitivity (n = 9 studies), 93 % (95 %CI 85 % – 97 %); specificity (n = 9 studies),
85 % (95 %CI 79 % – 89 %). In each case, there was no significant evidence of heterogeneity.
The diagnostic characteristics of POC for directed tissue sampling were as follows
([Table 3]): accuracy (n = 13 studies), 79 % (95 %CI 74 % – 84 %) ( [Fig.7]); sensitivity (n = 12 studies), 69 % (95 %CI 57 % – 78 %); specificity (n = 10 studies),
94 % (95 %CI 89 % – 97 %). Meta-regression identified a significant association between
the type of POC used and visual accuracy (P < 0.01) and between the type of POC used and visual sensitivity (P = 0.01), with dual-operator cholangioscopy having higher rates compared with SOC.
There was a potential trend toward an association between the number of biopsies and
accuracy (P = 0.077) such that an increased number of biopsies was associated with increased
accuracy. The estimated technical success rate (n = 18 studies) was 94 % (95 %CI 90 % – 96 %)
([Fig. 8]), with significant evidence of heterogeneity (P < 0.011, I
2 = 67.39).
Table 3
Efficacy and safety of peroral cholangioscopy for the removal of bile duct stones
and the diagnosis of indeterminate strictures.
|
Estimated
|
95 % CI
|
I
2
|
Heterogeneity?
(P value)
|
Publication bias?
(P value)
|
|
Stones
|
|
|
|
|
|
|
Clearance rate
|
88 %
|
85 % – 91 %
|
26.14
|
No (0.09)
|
Yes (0.05)
|
|
Recurrence rate
|
13 %
|
7 % – 20 %
|
40.09
|
No (0.14)
|
No (0.56)
|
|
Technical success rate
|
91 %
|
88 % – 94 %
|
61.72
|
Yes ( < 0.01)
|
No (0.32)
|
|
Strictures
|
|
|
|
|
|
|
Visual accuracy
|
89 %
|
84 % – 93 %
|
35.21
|
No (0.13)
|
Yes (0.01)
|
|
Visual sensitivity
|
93 %
|
85 % – 97 %
|
38.46
|
No (0.11)
|
Yes ( < 0.01)
|
|
Visual specificity
|
85 %
|
79 % – 89 %
|
0
|
No (0.84)
|
No (0.50)
|
|
Biopsy accuracy
|
79 %
|
74 % – 84 %
|
19.12
|
No (0.09)
|
Yes (0.01)
|
|
Biopsy sensitivity
|
69 %
|
57 % – 78 %
|
97.97
|
Yes ( < 0.01)
|
No (0.07)
|
|
Biopsy specificity
|
94 %
|
89 % – 97 %
|
0
|
No (0.88)
|
No (0.18)
|
|
Technical success rate
|
94 %
|
90 % – 96 %
|
67.39
|
Yes ( < 0.01)
|
Yes ( < 0.01)
|
|
Adverse event rate
|
|
|
|
|
|
|
Overall
|
7 %
|
6 % – 9 %
|
32.36
|
Yes (0.02)
|
Yes ( < 0.01)
|
|
Pancreatitis
|
2 %
|
2 % – 3 %
|
0
|
No (0.99)
|
Yes ( < 0.01)
|
|
Cholangitis
|
4 %
|
3 % – 5 %
|
25.55
|
No (0.06)
|
Yes ( < 0.01)
|
|
Perforation
|
1 %
|
1 % – 2 %
|
0
|
No (0.99)
|
No (0.73)
|
|
Other events
|
3 %
|
2 % – 4 %
|
37.74
|
Yes (0.01)
|
Yes ( < 0.01)
|
|
Serious events
|
1 %
|
1 % – 2 %
|
0
|
No (0.99)
|
No (0.28)
|
CI, confidence interval.
Fig. 6 Forest plot of studies reporting visual accuracy of peroral cholangioscopy in diagnosing
indeterminate biliary strictures. Pooled accuracy rate was 89 % (95 % confidence interval
[CI] 84 % – 93 %).
Fig. 7 Forest plot of studies reporting biopsy accuracy of peroral cholangioscopy in diagnosing
indeterminate biliary strictures. Pooled accuracy rate was 79 % (95 % confidence interval
[CI] 74 % – 94 %).
Fig. 8 Forest plot of studies reporting technical success rate of peroral cholangioscopy
for stricture-related indications. Pooled success rate was 94 % (95 % confidence interval
[CI] 90 % – 96 %).
Adverse events of peroral cholangioscopy
The estimated overall adverse event rate was 7 % (95 %CI 6 % – 9 %) ( [Fig.9]). The estimated rates of pancreatitis, cholangitis, perforation, and other adverse
events were 2 % (95 %CI 2 % – 3 %), 4 % (95 %CI 3 % – 5 %), 1 % (95 %CI 1 % – 2 %),
and 3 % (95 %CI 2 % – 4 %), respectively. The estimated rate of severe adverse events
was 1 % (95 %CI 1 % – 2 %).
Fig. 9 Forest plot of studies reporting overall adverse event rates of peroral cholangioscopy.
Pooled event rate was 7 % (95 % confidence interval [CI] 6 % – 9 %).
Discussion
POC has become a valuable tool for the treatment of difficult bile duct stones and
the evaluation of indeterminate strictures. Despite increasing clinical use, there
are very limited composite data evaluating its efficacy and safety. The aims of this
study were to systematically review and analyze the efficacy of POC for difficult
bile duct stones and indeterminate biliary strictures. The results of this systematic
review and meta-analysis demonstrate a high stone clearance rate with the use of POC
for difficult bile duct stones (88 %, 95 %CI 85 % – 91 %). Similarly, POC showed an
accuracy of 89 % (95 %CI 84 % – 93 %) for visual impression of indeterminate biliary
strictures and of 79 % (95 %CI 74 % – 84 %) for directed tissue sampling. Finally,
POC was noted to have an overall low adverse event rate (7 %, 95 %CI 6 % – 9 %).
This analysis found that the accuracy of the visual impression was greater than biopsy-related
accuracy, likely because of the high sensitivity of visual impression and poor sensitivity
of biopsies. Currently, there is no standardized classification system used to help
make a visual diagnosis of malignancy. However, studies evaluating POC for visual
impression used characteristics such as the presence of irregular mucosa, an intraductal
mass, or a tumor vessel to qualify a lesion as malignant, as these findings are often
suggestive of malignancy [9]
[14]
[20]
[43]
[44]
[48]
[53]
[56]. It should be noted, however, that the data on the diagnostic characteristics of
these individual characteristics are limited at the present time. Given the low specificity
of visual impression, it cannot be used alone to confirm a diagnosis. This analysis
also found that SOC systems had a significantly reduced sensitivity for visual impression
when compared with dual-operator cholangioscopes. This is likely due to the fact that
SOC systems provide a fiberoptic image that is of poorer quality than the digital
image obtained with dual-operator cholangioscopes.
The suboptimal biopsy-related accuracy of POC was attributed to low overall sensitivity.
This highlights the technical challenges of sampling indeterminate biliary strictures
and calls for an improvement in tissue acquisition techniques. Our analysis found
a statistically insignificant but potential trend toward greater accuracy with an
increased number of biopsies. As suggested by Kalaitzakis et al. [29], taking more biopsy samples may result in an increased sensitivity (and potentially
accuracy) for making a histological diagnosis. The high sensitivity of visual impression
and high specificity of POC-directed biopsy make a combined approach, rather than
the individual use of each, likely the most helpful method for making a diagnosis
of malignancy.
Two meta-analyses [58]
[59] have assessed the efficacy and diagnostic performance of SOC for indeterminate biliary
strictures. One study [58] concluded that visual impression is useful for detecting a malignant lesion, and
the other [59] that SOC biopsies have a moderate sensitivity for diagnosing malignant strictures.
Both studies revealed that SOC is useful in confirming a malignant diagnosis because
of its high specificity. One notable difference in this meta-analysis is that the
studies involved looked at all types of POC and were not limited to SOC. However,
the data from this meta-analysis are in concordance with those of the aforementioned
meta-analyses in that they reveal a high sensitivity of visual impression for the
detection of malignant strictures and a high specificity associated with biopsy that
can be useful in the confirmation of a malignant diagnosis.
POC appears to be a relatively safe procedure with a very low rate of serious events
(1 %, 95 %CI 1 % – 2 %). The data obtained in this systematic review and meta-analysis
provide point estimates of adverse events that may be used in discussions with patients
before a procedure. Notably, the patients undergoing POC have failed ERCP; this may
be because they have more difficult anatomy or unusual lesions that require more manipulation.
As such, there is a component of selection bias when patients are chosen to undergo
POC. A recent study [60], completed in Sweden based on a national registry, reported that the risk for intra-
and post-procedural adverse events is significantly increased when a patient undergoes
POC in conjunction with ERCP, as opposed to ERCP alone. However, the study also noted
that in a multivariate analysis that adjusted for confounders, the risk for pancreatitis
and cholangitis was not increased. Of note, a systematic survey evaluating the incidence
rates of post-ERCP complications [61] revealed an ERCP complication rate of approximately 6.85 %, with a severe event
rate of approximately 1.67 %. These figures are comparable with the adverse event
rates for POC estimated in this meta-analysis. Overall, it is clear that further research
and data comparing POC with ERCP alone or with EUS are needed to compare the rates
of adverse events and determine whether there is an increased adverse event rate with
POC.
Limitations to this analysis included study heterogeneity and variability in the type
of POC used. The studies had various patient populations, and the procedures were
completed by using various methods of POC as well as differing instruments within
each method. Furthermore, interoperator variability cannot be accounted for. Also,
the definition of adverse event varied from study to study and accounted only for
what was reported by the authors of each study. For example, some studies documented
minor bleeding and considered it an adverse event, whereas others did not. It should
also be noted that are various types of difficult stones – large stones, confluence
stones, impacted stones, etc. Although the meta-regression found no association between
the size and location of stones, confluence stones and impacted stones were not specifically
addressed in most studies. Therefore, they could not be distinctly evaluated in this
analysis. Finally, it is important to make a distinction between filling defects caused
by malignant strictures and filling defects caused by extrinsic compression/factors.
Unfortunately, information on the latter was often very limited and not made distinct
in the literature. Thus, the use of POC for detecting malignancy in filling defects
caused by external compression or other factors could not be analyzed in this study.
POC is a safe and effective adjunctive tool with ERCP for the evaluation of bile duct
strictures and for the treatment of bile duct stones when conventional methods have
failed. Despite the increasing utilization of POC and technical advances such as the
recently introduced digital single-operator cholangioscope, the current systematic
review and meta-analysis confirm the paucity of high level evidence supporting the
use of POC. Prospective, controlled clinical trials are needed to further elucidate
the precise role of POC and develop criteria that can be used to standardize the diagnosis
and treatment of pancreaticobiliary diseases.
