Keywords
Strictures - ERC topics - Pancreatobiliary (ERCP/PTCD) - Diagnostic ERC
Introduction
Pathological diagnosis plays an important role in determining the appropriate treatment
strategy for patients with biliary strictures. However, establishing a definitive
diagnosis is challenging. Endoscopic retrograde cholangiopancreatography (ERCP) is
the most common technique used to obtain biliary tissue samples for pathological diagnosis
[1]. Transpapillary biliary forceps biopsy (TBFB) and brush cytology under fluoroscopy
are conventional ERCP methods; however, their diagnostic yield is suboptimal, with
a diagnostic sensitivity for malignancy of 33% to 88% and 18% to 80%, respectively
[2]
[3]
[4]
[5]. In addition, TBFB requires repeated insertion of biopsy forceps through the major
duodenal papilla to achieve higher sensitivity [6]
[7]. Careless handling of biliary forceps may cause injury to the major duodenal papilla
or perforation of the bile duct because the tip of the biopsy forceps is stiffer than
that of the standard ERCP catheter. Moreover, frequent attempts to insert the device
into the bile duct require additional time. These factors increase the risk of post-ERCP
pancreatitis (PEP). PEP occurs in 3.5% to 9.7% of patients after ERCP; furthermore,
in 0.1% to 0.7% of patients who develop PEP, it can be fatal [8]
[9]
[10]. Thus, an alternative tissue sampling method with a higher sensitivity and safer
profile is required.
A TBFB method using an existing biliary stent delivery system or a biliary dilator
has been reported [11]. Using these methods, adequate specimens can be obtained for pathological evaluation
without leading to PEP [11]. However, these devices require modification by endoscopists for use in TBFB, which
is a major disadvantage. Recently, a novel, dedicated sheath device for TBFB (Endosheather;
Piolax, Kanagawa, Japan) was launched. It consisted of a tapered inner catheter and
an outer sheath. After removing the inner catheter following the insertion of the
device into the bile duct, biopsy forceps can be inserted into the targeted bile duct
lesion through the outer sheath. A retrospective single-center study evaluated TBFBs
using this novel delivery device in 14 patients and revealed high diagnostic yields
of 90% and 92.3% in sensitivity and accuracy, respectively, making it a promising
method for TBFB [12]. However, there have been no comparative studies of TBFB using this novel sheath
device and conventional TBFB. Therefore, in this retrospective study, we compared
these two techniques and evaluated the efficacy and safety of the delivery device.
Patients and methods
Study design
We retrospectively analyzed consecutive patients with suspected bile duct cancer who
underwent TBFB using Radial Jaw 4 Pediatric Biopsy Forceps (Boston Scientific Japan,
Tokyo, Japan) and FB-39Q-1 forceps (Olympus, Tokyo, Japan) at the Shizuoka Cancer
Center Hospital between January 2020 and December 2021.More than 500 ERCP procedures
are performed annually. During the study period, conventional TBFB or TBFB using a
novel sheath device (Endosheather) was performed to obtain histological evidence of
biliary stricture. Regarding the selection of TBFB methods, conventional TBFB was
used until March 2021; thereafter, the novel sheath device was utilized. We excluded
patients with surgically altered anatomy other than Billroth I reconstruction and
with insufficient data. This study was approved by the Institutional Review Board
(J2022–14) and performed in accordance with the principles of the Declaration of Helsinki.
A dedicated delivery system
The novel sheath device consists of a tapered inner catheter with a 1.13-mm tip diameter
and a 2.44-mm outer sheath (7.2F) with a radiopaque marker ([Fig. 1]). After inserting the device into the bile duct over the guidewire, biopsy forceps
up to a diameter of 1.9 mm were inserted through the outer sheath following withdrawal
of the inner catheter. In addition, this device is compatible with a 0.035-inch guidewire.
Fig. 1 Image of the novel delivery device. a The device has an inner catheter with a tapered tip (red arrowhead) and an outer
sheath. There is almost no gap between the outer sheath and the inner catheter (blue
arrow). b The inner catheter can be removed from the outer sheath. c The biopsy forceps (Radial Jaw 4) are inserted through the outer sheath.
ERCP and TBFB
ERCP and TBFB were performed by three endoscopists: one with more than 10 years of
experience and two with more than 5 years but less than 10 years of experience with
ERCP. ERCP was performed using a side-viewing duodenoscope (TJF-Q290V, TJF-260V, or
JF-260V; Olympus Medical Systems, Tokyo, Japan). After selective biliary cannulation
using wire-guided cannulation, a 0.025-inch guidewire (VisiGlide2; Olympus Medical
Systems, Tokyo, Japan) was inserted into the bile duct, and the biliary stricture
was detected by cholangiography. Endoscopic sphincterotomy (EST) was performed when
there were no contraindications or a history of EST. Regarding conventional TBFB,
we generally used either of the two biopsy forceps (Radial Jaw 4 Pediatric Biopsy
Forceps; 2.0 mm outer diameter, 5.4-mm opening width and effective length 160 cm,
Boston Scientific Japan, or FB-39Q-1 forceps; 1.95-mm opening width, Olympus Medical
Systems, Tokyo, Japan). When there was difficulty in inserting Radial Jaw 4 into the
bile duct, FB-39Q-1 forceps were used instead because the latter has more flexibility.
Biopsy forceps were inserted into the bile duct alongside the guidewire, and bile
duct tissue was obtained from the biliary stricture. In principle, biopsies were performed
at least three times. When using the novel sheath device, the delivery device was
inserted into the biliary stricture over the guidewire, and the inner catheter and
guidewire were removed, leaving the outer sheath in place ([Fig. 2]
a). Thereafter, biopsy forceps (Radial Jaw 4 Pediatric Biopsy Forceps; 2.0-mm outer
diameter, 5.4-mm opening width, and effective length 240 cm, Boston Scientific Japan,
Tokyo, Japan) were inserted into the outer sheath, and bile duct tissue was obtained
([Fig. 2]
b). Five biopsies were performed because the insertion of biopsy forceps into the bile
duct was easier compared with conventional TBFB. The selection of biliary drainage
methods, such as endoscopic nasobiliary drainage and endoscopic biliary stenting,
was performed depending on patient condition. A prophylactic pancreatic stent was
inserted at endoscopist discretion. Diclofenac suppositories and ulinastatin were
administered to prevent PEP when there were no contraindications. Fluoroscopic and
endoscopic videos during ERCP were recorded for all patients.
Fig. 2 A cholangiogram image obtained using the novel delivery device. a The outer sheath tip is located above the bile duct stricture (yellow ahead) after
removing the inner catheter. b Biopsy forceps (Radial Jaw 4) (white ahead) are inserted into the bile duct through
the outer sheath to obtain the bile duct specimen.
Pathological assessment of the sample obtained by TBFB and final diagnosis
In the pathological reports, histological classifications were divided into five categories:
inadequate, benign, atypical, suspicious for adenocarcinoma, and adenocarcinoma, which
were performed by an experienced pathologist (KS). In this study, an adequate sample
was defined as a specimen that included tumor cells or epithelial cells with stroma
and diagnosed as a benign, atypical, suspicious, or typical adenocarcinoma. When they
were classified as suspicious for adenocarcinoma or adenocarcinoma, the samples were
categorized as positive for malignancy; however, samples that were considered benign
or atypical were categorized as negative for malignancy. The final diagnosis was based
on the surgical diagnosis of a resected specimen, positive for malignancy at TBFB
diagnosis with a compatible clinical course of more than 6 months, or negative for
malignancy at TBFB diagnosis with spontaneous resolution or a lack of deterioration
on imaging findings for a clinical follow-up time of at least 6 months. An accurate
diagnosis using TBFB was defined as positive for malignancy with a final diagnosis
of malignant disease and negative for malignancy with a final diagnosis of benign
disease.
Outcome measurements
We defined the conventional and Endosheather groups as patients who underwent TBFB
using the conventional method and Endosheather, respectively. Patient data were collected
from electronic medical records and endoscopy databases. We evaluated the sensitivity,
specificity, positive predictive value (PPV), negative predictive value (NPV), and
accuracy of TBFB for pathological diagnosis and compared the conventional and Endosheather
methods. Furthermore, the rates of adequate sample collection, failed insertion of
forceps into the bile duct per attempt, biopsy time per biopsy, biopsy time per ERCP,
procedure time, and adverse events (AEs) were compared between the two groups. Regarding
failed attempts of the forceps into the bile duct, defined as when the biopsy forceps
were not inserted into the bile duct, two gastroenterologists (H.I. and F.N.; Japanese
Gastroenterological Endoscopy Society board-certified members) reevaluated it by examining
the video of ERCP for the study. When there was a discrepancy between opinions, a
decision was made after a discussion between them. Biopsy time per ERCP was measured
from the first attempt to insert the biopsy forceps to the end of the last biopsy.
Biopsy time per biopsy was defined as biopsy time per ERCP divided by the total number
of biopsies. Procedure time was measured from endoscope insertion to withdrawal. AEs
were defined and graded according to the American Society for Gastrointestinal Endoscopy
Severity Grading system [13].
Statistical analyses
Continuous variables were presented as medians and interquartile ranges and were compared
using the Mann-Whitney U-test. Categorical variables were shown as proportions and
were compared using Fisher’s exact test. Sensitivity, specificity, PPV, NPV, and accuracy
were compared using the χ2 test and Fisher’s exact test, where appropriate. P < 0.05 was considered to be statistically significant. All analyses were performed
using R version 3.4.1 (The R Foundation for Statistical Computing, Vienna, Austria).
Results
Patient characteristics
During the study period, 122 ERCP cases with suspected bile duct cancer underwent
TBFB. After excluding seven ERCP cases (surgically altered anatomy, n = 3; insufficient
data, n = 4), 115 ERCP procedures were analyzed in this study (Conventional group,
n = 76; Endosheather group, n = 39). Patient characteristics are shown in [Table 1]. No significant differences were observed between the two groups.
Table 1 Patient characteristics.
|
Measure
|
Conventional
n = 76
|
Endosheather
n = 39
|
P value
|
|
ERCP, endoscopic retrograde cholangiopancreatography; EST, endoscopic sphincterotomy;
PGW, pancreatic guidewire; ENBD, endoscopic nasobiliary drainage; PEP, post-ERCP pancreatitis;
IQR, interquartile range.
|
|
Age, years, median (IQR)
|
71 (65–77)
|
72 (65.8–77)
|
0.25
|
|
Location of biliary stricture, n (%)
|
|
|
54 (71.1)
|
26 (66.7)
|
0.67
|
|
|
22 (28.9)
|
13 (33.3)
|
|
|
Length of biliary stricture, mm, median (IQR)
|
16 (12–22)
|
16.5 (10–22.8)
|
0.43
|
|
Cannulation time, min, (median [IQR])
|
1.25 (0.3–4)
|
1 (0.5–2.5)
|
0.08
|
|
Naïve major duodenal papilla, n (%)
|
45 (59.2)
|
20 (51.3)
|
0.43
|
|
EST history
|
28 (36.8)
|
10 (25.6)
|
0.30
|
|
ERCP-related procedure, n (%)
|
|
|
13 (17.1)
|
3 (7.7)
|
0.28
|
|
|
2 (2.6)
|
0 (0)
|
0.55
|
|
|
48 (63.2)
|
29 (74.4)
|
0.30
|
|
|
|
|
0.24
|
|
|
21 (27.6)
|
16 (41)
|
|
|
|
4 (5.3)
|
0 (0)
|
|
|
|
45 (59.2)
|
22 (56.4)
|
|
|
|
6 (7.9)
|
1 (2.6)
|
|
|
Pancreatic stent placement for the prevention of PEP
|
4 (5.3)
|
1 (2.6)
|
0.66
|
|
Final diagnosis, n (%)
|
|
|
63 (82.9)
|
36 (92.3)
|
0.26
|
|
|
55 (72.4)
|
33 (84.6)
|
|
|
|
8 (10.5)
|
3 (7.7)
|
|
|
|
13 (17.1)
|
3 (7.7)
|
|
|
Surgery
|
23 (30.3)
|
16 (41)
|
0.34
|
Procedure outcomes of bile duct biopsy
The median number of biopsies was three and five in the conventional and Endosheather
groups, respectively (P < 0.001) ([Table 2]). Rates of failed insertion of the forceps into the bile duct per attempt were 28.3%
(95/336) and 0% in the conventional and Endosheather groups, respectively (P < 0.01). Biopsy times per ERCP were 220 and 296.5 sec in the conventional and Endosheather
groups (P = 0.56), respectively, whereas biopsy time per biopsy of the conventional group was
significantly longer than that of the Endosheather group (68.3 vs. 54 sec, P = 0.03).
Table 2 Study outcomes of TBFB.
|
Measure
|
Conventional
n = 76
|
Endosheather
n = 39
|
P value
|
|
TBFB, transpapillary biliary forceps biopsy; ERCP, endoscopic retrograde cholangiopancreatography;
IQR, interquartile range.
|
|
Type of forceps for TBFB, n
|
|
|
25 (32.9)
|
0 (0)
|
< 0.01
|
|
|
51 (67.1)
|
39 (100)
|
|
|
Number of biopsies, n
|
|
|
3
|
5
|
< 0.01
|
|
|
3–3
|
3–5
|
|
|
|
1–6
|
2–10
|
|
|
Total number of biopsies, n
|
241
|
225
|
|
|
Total number of attempts for insertion of forceps into the bile duct, n
|
336
|
225
|
|
|
Total failed attempts of the forceps into the bile duct, n
|
95
|
0
|
|
|
Rate of failed attempts of the forceps into the bile duct, % (n/N)
|
28.3 (95/336)
|
0 (0/225)
|
< 0.01
|
|
Biopsy time per one biopsy, sec median (IQR)
|
68.3 (53.4–103)
|
54 (42–69.6)
|
0.03
|
|
Biopsy time per ERCP, sec, median (IQR)
|
220 (160–330)
|
296.5 (218.5–375)
|
0.56
|
|
Procedure time, min, median (IQR)
|
25 (19–37.5)
|
24 (19–33)
|
0.7
|
Diagnostic yields of biliary forceps biopsy
Diagnostic yields of biliary forceps biopsies are summarized in [Table 3]. The rate of obtaining adequate samples was significantly lower in the conventional
group than in the Endosheather group (72.4% [55/76] vs. 89.7% [35/39], P = 0.03). Sensitivity of the conventional group was significantly lower than that
of the Endosheather group (66.7% vs. 88.9%, P = 0.02). There were no differences observed in specificity, PPV, NPV, or accuracy
between the two groups.
Table 3 Diagnostic yield of TBFB.
|
Conventional n = 76
|
Endosheather n = 39
|
|
TBFB, transpapillary biliary forceps biopsy; CI, confidence interval; PPV, positive
predictive value; NPV, negative predictive value; NS, not significant.
|
|
Final diagnosis
|
|
Final diagnosis
|
|
Malignant
|
Benign
|
|
Malignant
|
Benign
|
|
TBFB diagnosis
|
Malignant
|
42
|
0
|
TBFB diagnosis
|
Malignant
|
32
|
0
|
|
Benign
|
1
|
12
|
Benign
|
1
|
2
|
|
Inadequate
|
20
|
1
|
Inadequate
|
3
|
1
|
|
Measure
|
|
|
|
P value
|
|
Sensitivity %, (n/N)
|
66.7 (42/63)
|
88.9 (32/36)
|
0.02
|
|
|
53.7–78
|
73.9–96.9
|
|
|
Hilar
|
70 (31/47)
|
87 (20/23)
|
0.08
|
|
Distal
|
68.8 (11/16)
|
92.3 (12/13)
|
0.18
|
|
Specificity %, (n/N)
|
92.3 (12/13)
|
66.7 (2/3)
|
0.35
|
|
|
64–99.8
|
9.4–99.2
|
|
|
Hilar
|
100 (5/5)
|
66.7 (2/3)
|
0.38
|
|
Distal
|
87.5 (7/8)
|
0
|
NS
|
|
PPV %, (n/N)
|
100% (42/42)
|
100 (32/32)
|
NS
|
|
|
91.6–100
|
89.1–100
|
|
|
Hilar
|
100 (31/31)
|
100 (20/20)
|
NS
|
|
Distal
|
100 (11/11)
|
100 (12/12)
|
NS
|
|
NPV %, (n/N)
|
92.3 (12/13)
|
66.7 (2/3)
|
0.35
|
|
|
64–99.8
|
9.4–99.2
|
|
|
Hilar
|
83.3 (5/6)
|
66.7 (2/3)
|
1
|
|
Distal
|
100 (7/7)
|
0
|
NS
|
|
Accuracy %, (n/N)
|
71.1 (54/76)
|
87.2 (34/39)
|
0.06
|
|
|
59.5–80.9
|
72.6–95.7
|
|
|
Hilar
|
69.2 (36/52)
|
84.6 (22/26)
|
0.18
|
|
Distal
|
75 (18/24)
|
92.3 (12/13)
|
0.39
|
Adverse events
Procedure-related AEs are summarized in [Table 4]. Pancreatitis and cholangitis occurred in 3.9% and 1.3% of patients in the conventional
group, respectively; however, these events did not occur in the Endosheather group.
Table 4 Adverse events.
|
Measure
|
Conventional
n = 76
|
Endosheather
n = 39
|
P value
|
|
NS, not significant.
|
|
Overall, n (%)
|
4 (5.2)
|
0 (0)
|
0.30
|
|
Pancreatitis, n (%)
|
3 (3.9)
|
0 (0)
|
0.55
|
|
Cholangitis, n (%)
|
1 (1.3)
|
0 (0)
|
1.00
|
|
Bleeding, n (%)
|
0 (0)
|
0 (0)
|
NS
|
|
Perforation, n (%)
|
0 (0)
|
0 (0)
|
NS
|
Discussion
In this study, we compared outcomes of conventional TBFB with those of TBFB using
a novel sheath device. TBFB with a novel sheath device exhibited higher sensitivity
than conventional TBFB.
A retrospective study evaluating TBFB with a pusher tube of an existing biliary stent
delivery system revealed that the rate of obtaining samples, including submucosal
tissue, was 91.4%, whereas another retrospective study using a biliary dilation catheter
demonstrated that sensitivity, specificity, and accuracy were 87.5%, 100%, and 93.7%,
respectively [14]
[15]. Furthermore, a retrospective, single-arm study evaluating the same sheath device
used in our study revealed high diagnostic yields (sensitivity, 90%; specificity,
100%; and accuracy, 92.3%). Our study also demonstrated satisfactory diagnostic ability
of the novel sheath device, in line with previous studies. In addition, we compared
results with those of the conventional TBFB, and no study has compared these two methods.
A significantly higher rate of adequate sampling and higher sensitivity were observed
when using the novel sheath device.
Insertion of biopsy forceps into the bile duct can be challenging. However, using
the novel sheath device, biopsy forceps can be inserted easily into the bile duct,
and once inserted, it can be repeated, in line with our study. This resulted in a
shorter biopsy time per biopsy and a lower rate of failed insertion of the forceps
into the bile duct in the Endosheather group. We presumed that reducing the number
of failed insertions of forceps into the bile duct shortened the time required for
biopsy.
Although there was no significant difference in the rate of AEs, pancreatitis was
observed only in the conventional group. We speculated that the small sample size
was the reason for the lack of statistical difference. However, considering the reduction
in failed insertion of forceps into the bile duct, it is intuitively understandable
that risk of pancreatitis was reduced. We presumed that the favorable outcome resulted
from the ability to obtain tissue from the same lesion repeatedly and effortlessly
through the outer sheath. In recent years, precision medicine has become increasingly
important in bile duct cancer, and it is necessary to obtain a larger sample size
for comprehensive genome profiling using next-generation sequencing [16]. Because increasing the number of biopsies is desirable to increase the amount of
tissue obtained, TBFB using the novel sheath device can also be a promising technique
to safely perform TBFB in a shorter time for comprehensive genome profiling.
Nonetheless, this study has some limitations. First, some bias was inevitable owing
to its retrospective nature. However, enrolling consecutive patients and the choice
of two methods (conventional TBFB and TBFB using Endosheather), which were performed
at different times, probably reduced selection bias. Second, the number of biopsies
performed differed between the two groups. For conventional TBFB, to consider the
balance of diagnostic yield and safety, we decided that the number of biopsies was
set to at least three, although we knew that five or more biopsies were reported to
be an independent predictive factor for a positive cancer diagnosis of malignant biliary
stricture in a retrospective study [17]. Therefore, if more biopsies had been performed, the conventional method might have
had higher sensitivity. However, this may also lead to a higher incidence of pancreatitis.
Whether our approach can be adopted as a conventional method in all facilities remains
unclear, but we believe that there are not many institutions where biopsy forceps
are consistently inserted into the bile duct for tissue acquisition, five times in
every procedure. In that regard, this conventional method might be more realistic.
Third, this was a single-center study with a small sample size; thus, a multicenter
study with a larger sample size is required to confirm our results.
Conclusions
In conclusion, TBFB using a novel dedicated sheath device had a higher diagnostic
yield without failed insertion into the bile duct than conventional TBFB. This novel
sheath device can also facilitate TBFB and produce higher sensitivity.
