Introduction
Endoscopic drainage of malignant biliary obstruction using self-expandable metal stents
(SEMSs) is a widely used standard procedure to treat obstructive jaundice which enables
chemotherapy and improves patients’ symptoms [1]
[2]
[3]
[4]
[5]. Covered SEMSs (CSEMSs) may prevent tumor ingrowth more effectively than uncovered
SEMSs (USEMSs) [6]. In patients with malignant biliary obstruction, Isayama et al. found that the time
to recurrent biliary obstruction (RBO) was longer with CSEMS than with USEMS [5]. In patients with biliary obstruction caused by pancreatic carcinoma, Kitano et
al. reported that duration of patency was longer with CSEMS than with USEMS [7].
However, several meta-analyses reported that CSEMS has a higher risk of migration
than USEMS, despite prevention of ingrowth [1]
[2]
[3]
[8]. Mukai et al. developed a 12-mm-diameter fully-covered self-expandable metal stent
(FCSEMS) to prevent RBO, but it resulted in several cases of migration [9]. Therefore, we evaluated the efficiency of a 12-mm-diameter covered self-expandable-end,
bare metal stent (CSEEMS) in patients with malignant distal biliary obstruction for
preventing RBO.
Patients and methods
We retrospectively evaluated 99 patients with unresectable malignant distal biliary
obstructions
treated with covered biliary metallic stents at Fujita Health University Hospital.
Of the 99
patients, 33 underwent placement of 12-mm-diameter CSEEMS (Tae Woong Medical, Seoul,
Korea)
between June 2015 and April 2017 (12-mm-CSEEMS group) ([Fig. 1]) and 66 underwent 10-mm-diameter FCSEMS (Wallflex biliary RX stent, Boston Scientific,
Natick, Massachusetts) placement between January 2010 and July 2015 (10-mm-FCSEMS
group).
Fig. 1 A 12-mm-diameter covered self-expandable-end bare metal stent (CSEEMS). This SEMS
is made of nitinol wire and covered with a silicone membrane. The proximal 10 mm is
uncovered, and 5 mm of the distal end is flared to prevent migration. The area of
12-mm CSEEMS is 1.44-fold larger than that of the 10-mm SEMS.
The endpoint of this study was RBO with SEMS, or patients’ death, whichever was earlier.
The patients survived during the observation period were considered as censored cases.
Before inserting these metal stents, carcinoma was diagnosed by cytology, biopsy,
or endoscopic ultrasound-guided fine-needle aspiration. If diagnosis by tissue biopsy
or cytology was not possible, enhanced computed tomography (CT) or magnetic resonance
imaging was used. We initially performed drainage using a plastic stent and then switched
the plastic stent with a 12-mm CSEEMS or 10-mm FCSEMS after confirming that there
was no indication for surgery and that the patients had good life expectancy. Thereafter,
the patients were treated with chemotherapy or optimal supportive care.
Eligibility criteria
Patients who were age ≥ 20 years and those with a life expectancy ≥ 3 month, an Eastern
Cooperative Oncology Group Performance Status (ECOG-PS) < 4 and diagnosed with distal
biliary obstruction caused by an unresectable malignancy were included. Patients with
ECOG-PS ≥ 4, massive ascites, an intestinal obstruction distal to the ampulla, and
prior biliary SEMS placement and those who were unable to give informed consent for
SEMS replacements were excluded.
Ethical affairs
The study protocol was approved by the Institutional Review Board of Fujita Health
University Hospital (HM16-059) and was carried out following the ethical principles
of the Declaration of Helsinki.
SEMS used in the study
A 12-mm CSEEMS is made of nitinol wire and covered with a silicone membrane, with
the proximal
10 mm uncovered and distal 5-mm flared ends designed to prevent migration. The area
of 12-mm
CSEEMS was 1.44-fold larger than that of 10-mm FCSEMS. A 12-mm CSEEMS is available
in lengths of
6, 7, and 8 cm and is equipped with a 9-Fr standard delivery device. For 12-mm CSEEMS,
the axial
force (AF) at a 20-mm distance from the bending point was 0.29 N and the radial force
(RF) measured at a 4-mm diameter was 4.5 N, as previously described [10]. A 12-mm CSEEMS was newly manufactured just before this study ([Fig.1]).
A 10-mm FCSEMS is made of nitinol wire and covered with a silicone membrane, with
both ends flared. A 10-mm FCSEMS is available in lengths of 4, 6, and 8 cm and equipped
with an 8.5-Fr standard delivery device. The AF of this stent at a 20-mm distance
from the bending point was 0.65N, and RF measured at a 4-mm diameter was 4.7 N.
Procedures
SEMS was inserted during endoscopic retrograde cholangiopancreatography (ERCP) by
two experienced investigators using a standard duodenoscope (TJF-260V; Olympus, Tokyo,
Japan). Sphincterotomy was done before stent insertion in all cases. The length of
the SEMS was determined by the primary endoscopist, and the distal end of the SEMS
was located in the duodenum.
RBO and adverse events
We followed up all patients at least once a month and examined their clinical findings
and biochemical parameters of hepatobiliary functions and inflammation, such as aspartate
aminotransferase, alanine aminotransferase, alkaline phosphatase, γ-glutamyl transpeptidase,
total and direct bilirubin, white blood cell count, and C-reactive protein. CT scanning
or abdominal ultrasound was carried out at least once every 2 or 3 months until a
patient’s death. RBO and adverse events and their severity were defined according
to the Tokyo Criteria 2014 [11]. RBO was defined as an occlusion or migration, and TRBO as the interval between
SEMS placement and RBO or patients’ death, whichever was earlier.
Statistical analysis
The Kaplan – Meier method was used to evaluate overall survival (OS), with living
patients censored at the last date of follow-up (October 31, 2017). TRBO was also
estimated by the Kaplan – Meier method, with patients who had not experienced RBO
censored at the end of the study (October 31, 2017). The hazard ratios of prognostic
factors for OS and TRBO were estimated by a Cox proportional hazards model, which
included age, sex, clinical stage, chemotherapy, prior drainage, and stent types.
Continuous variables were compared using the Mann – Whitney U test and Fisherʼs exact
test for categorical variables. All analyses were done using StatFlex version 6.0
for windows (StatFlex, Osaka, Japan).
Results
Patient characteristics, outcomes, and survival
Clinical characteristics were not significantly different between the 12-mm-CSEEMS
group and the
10-mm-FCSEMS group ([Table 1]). In all 99 patients,
placements of 12-mm CSEEMS and 10-mm FCSEMS were technically successful. Median OS
of
12-mm-CSEEMS group was 232 days (range, 35 – 814 days), and 27 patients (81.8 %) died
and six
patients (18.2 %) were still alive by the end of the study ([Fig. 2]).
Table 1
Characteristics of patients in the 12-mm-CSEEMS and 10-mm-FCSEMS groups.
|
Patientsʼ groups
|
12-mm-CSEEMS group (n = 33)
|
10-mm-FCSEMS group (n = 66)
|
P value
|
|
Age, years, median (range)
|
75 (61 – 92)
|
71 (36 – 95)
|
0.200
|
|
Sex, n (%)
|
|
|
17 (51.5)
|
35 (53.0)
|
0.887
|
|
Length of stricture, mm, median (range)
|
44 (16 – 72)
|
52 (18 – 74)
|
0.870
|
|
Length of stent, n (%)
|
0.643[1]
|
|
|
0 (0)
|
1 (1.5)
|
|
|
|
9 (27.3)
|
20 (30.3)
|
|
|
|
12 (36.4)
|
0 (0)
|
|
|
|
12 (36.4)
|
45 (68.2)
|
|
|
Etiology, n (%)
|
|
|
27 (81.8)
|
51 (77.2)
|
0.602[2]
|
|
|
3 (9.1)
|
11 (16.7)
|
|
|
|
1 (3.0)
|
2 (3.0)
|
|
|
|
1 (3.0)
|
0 (0)
|
|
|
|
0 (0)
|
2 (3.0)
|
|
|
|
1 (3.0)
|
0 (0)
|
|
|
Clinical stage, n (%)
|
0.697
|
|
|
|
|
2 (6.1)
|
2 (3.0)
|
|
|
|
6 (18.2)
|
15 (22.7)
|
|
|
|
25 (75.8)
|
49 (74.2)
|
|
|
Chemotherapy, n (%)
|
20 (60.6)
|
38 (57.6)
|
0.943
|
|
Best supportive care, n (%)
|
10 (30.3)
|
28 (42.4)
|
0.342
|
CSEEMS, covered, self-expandable end bare metal stent; FCSEMS, fully-covered self-expandable
metal stent
1 Length of stent (4 – 6 cm or 7 – 8 cm) were compared between 12-mm-CSEEMS group and
10-mm-FCSEMS group.
2 Etiology (pancreatic cancer or other diseases) were compared between 12-mm-CSEEMS
group and 10-mm-FCSEMS group.
Fig. 2 Overall survival in the 12-mm-CSEEMS group and 10-mm-FCSEMS group tended to be significantly
different by Kaplan-Meier analysis (log rank, P = 0.0809). Median OS in the 12-mm-CSEEMS group was 232 days and median OS in the
10-mm-FCSEMS group was 169.5 days.
Median OS of the 10-mm-FCSEMS-group was 169.5 days (range, 21 – 1019 days), and all
66 patients
died by the end of the study ([Fig. 2]). OS was tended to be
significantly different between 12-mm-CSEEMS group and 10-mm-FCSEMS group (P = 0.081,
[Fig. 2]). Univariate Cox analysis demonstrated that risk of mortality was lower in patients
with chemotherapy (HR, 0.610; 95 % CI, 0.4041 – 0.92560; P = 0.020), and it tended to be lower in patients with clinical stage II or III disease
(HR, 0.647; 95 % CI, 0.49376 – 1.03775; P = 0.071) and in the 12-mm-CEEMS group (HR, 0. 667; 95 % CI, 0.42377 – 1.04927; P = 0.080). Multivariate Cox hazard analysis demonstrated that risk of mortality was
lower in the females (HR, 1.974; 95 % CI, 1.23762 – 3.14849; P = 0.004), in patients with clinical stage of II or III disease (HR, 0.417; 95 % CI,
0.24050 – 0.72313; P = 0.002) and in the 12-mm-CSEEMS group (HR, 0.592; 95 % CI, 0.36340 – 0.96495; P = 0.044) ([Table 2]).
Table 2
Univariate and multivariate Cox hazard analyses of OS.
|
Variables
|
Univariate analysis
|
Multivariate analysis
|
|
Hazard ratio
|
95 % CI
|
P value
|
Hazard ratio
|
95 % CI
|
P value
|
|
Age (years)
|
1.009
|
0.99109 – 1.02624
|
0.340
|
1.0103
|
0.98964 – 1.0318
|
0.330
|
|
Sex, male
|
1.512
|
0.99728 – 2.29347
|
0.052
|
1.974
|
1.23762 – 3.14849
|
0.004
|
|
Primary disease (pancreatic cancer)
|
1.092
|
1.09198 – 1.85604
|
0.745
|
|
|
|
|
Clinical stage (II and III)
|
0.647
|
0.40376 – 103775
|
0.071
|
0.417
|
0.24050 – 0.72313
|
0.002
|
|
Chemotherapy
|
0.610
|
0.4041 – 0.92560
|
0.020
|
0.744
|
0.45211 – 1.22419
|
0.245
|
|
12-mm CSEEMS
|
0.667
|
0.42377 – 1.04927
|
0.080
|
0.592
|
0.36340 – 0.96495
|
0.044
|
OS, overall survival; CSEEMS, covered, self-expandable end bare metal stent
RBO and TRBO
In the 12-mm-CSEEMS group, RBO occurred in three patients (9.1 %) on days 132,155
and 505 by food impaction in one (3.0 %) and tumor ingrowth at the covered part of
the stent in two (6.1 %) ([Table 3]).
Table 3
Recurrent biliary obstruction and adverse events in 12-mm-CSEEMS and 10-mm-FCSEMS
groups.
|
12-mm-CSEEMS group (n = 33)
|
10-mm-CSEMS group (n = 66)
|
P value
|
|
Recurrent biliary obstruction, n (%)
|
3 (9.1)
|
|
29 (43.9)
|
|
0.001
|
|
Occlusion, n (%)
|
3 (9.1)
|
|
22 (33.3)
|
|
0.009
|
|
|
1 (3.0)
|
|
2 (3.0)
|
|
1.000
|
|
|
0
|
|
13 (19.7)
|
|
0.009
|
|
|
2 (6.1)
|
|
1 (1.5 )
|
|
0.549
|
|
|
0
|
|
5 (7.6)
|
|
0.166
|
|
|
0
|
|
0
|
|
1.000
|
|
|
0
|
|
1 (1.5)
|
|
1.000
|
|
Migration, n (%)
|
0
|
|
7 (10.6)
|
|
0.092
|
|
|
0
|
|
3 (4.5)
|
|
0.549
|
|
|
0
|
|
4 (6.1)
|
|
0.298
|
|
Adverse events, n (%)
|
|
Early adverse events (≤ 30 days)
|
2 (6.1)
|
|
5 (7.6)
|
|
1.000
|
|
|
0
|
|
1 (1.6)
|
on day 7
|
1.000
|
|
|
0
|
|
3 (4.5)
|
on day 1
|
0.298
|
|
|
0
|
|
1 (1.6)
|
on day 1
|
1.000
|
|
|
1 (3.0)
|
on day 1
|
0
|
0.333
|
|
|
|
1 (3.0)
|
on day 27
|
0
|
|
0.333
|
|
Late adverse events (≥ 31 days)
|
3 (9.1)
|
|
8 (12.1)
|
|
0.747
|
|
|
1 (3.0)
|
on day 77
|
1 (1.6)
|
on day 32
|
1.000
|
|
|
2 (6.1)
|
on days 116 and 151
|
7 (10.6)
|
on days 82, 108, 116, 132, 146, 172 and 196
|
0.714
|
CSEEMS, covered, self-expandable end bare metal stent; FCSEMS, fully-covered self-expandable
metal stent
In the 10-mm-FCSEMS group, RBO occurred in 29 patients (43.9 %) by food impaction
in two (3.0 %),
sludge formation in 13 (19.7 %), tumor ingrowth in one (1.5 %), tumor overgrowth in
five
(7.6 %), kinking in one (1.5 %), distal migration in three (4.5 %) and proximal migration
in
four (6.1 %) ([Table 3]). TRBO in the 12-mm-CSEEMS
group was significantly longer than that in the 10-mm-FCSEMS group (log rank, P = 0.001). Median TRBO in the 12-mm-CSEEMS group was 232 days and median TRBO in the
10-mm-FCSEMS group was 139.5 days ([Fig. 3]).
Fig. 3 Cumulative TRBO was significantly longer in the 12-mm-CSEEMS group than in the 10-mm-FCSEMS
group by Kaplan-Meier analysis (log rank, P = 0.0012). Median TRBO in the 12-mm-CSEEMS group was 232 days and median TRBO in
the 10-mm-FCSEMS group was 139.5 days.
Univariate Cox analysis ([Table 4]) demonstrated that risk of RBO was significantly lower in the 12-mm-CSEEMS group
(HR, 0.449; 95 % CI, 0.27967 – 0.72215; P = 0.001) than in 10-mm-FCSEMS group and chemotherapy also decreased risk of RBO (HR,
0.429; 95 % CI, 0.27665 – 0.66392; P < 0.001).
Table 4
Univariate and multivariate Cox hazard analyses of TRBO.
|
Variables
|
Univariate analysis
|
Multivariate analysis
|
|
Hazard ratio
|
95 % CI
|
P value
|
Hazard ratio
|
95 % CI
|
P value
|
|
Age (years)
|
1.012
|
0.99379 – 1.03151
|
0.192
|
1.000
|
0.98209 – 1.01913
|
0.963
|
|
Sex, male
|
1.183
|
0.85001 – 1.93025
|
0.2367
|
1.189
|
0.78224 – 1.80824
|
0.891
|
|
Primary disease (pancreatic cancer)
|
0.880
|
0.52295 – 1.48068
|
0.6300
|
|
|
|
|
Clinical stage (II and III)
|
0.711
|
0.44559 – 1.13394
|
0.1520
|
|
|
|
|
Chemotherapy
|
0.429
|
0.27665 – 0.66392
|
0.0001
|
0.453
|
0.27791 – 0.73974
|
0.002
|
|
12-mm CSEEMS
|
0.449
|
0.27967 – 0.72215
|
0.0009
|
0.458
|
0.28395 – 0.73744
|
0.001
|
TRBO, time to recurrent biliary obstruction; CSEEMS, covered, self-expandable end
bare metal stent
Multivariate Cox hazard analysis also demonstrated that risk of RBO was significantly
lower in the 12-mm-CSEEMS group than in the 10-mm-FCSEMS group (HR, 0.458; 95 % CI,
0.28395 – 0.73744; P = 0.001) and chemotherapy decreased risk of RBO (HR, 0.453; 95 % CI, 0.27791 – 0.73974;
P = 0.002).
Early adverse events (≤ 30 days)
In the 12-mm-CSEEMS group, there were no cases of post-endoscopic retrograde cholangiopancreatography
pancreatitis (PEP), while one patient (3.0 %) experienced abdominal pain on day 1
and one patient (3.0 %) experienced non-occlusion cholangitis on Day 27 ([Table 3]). In the 10-mm-FCSEMS group, cholecystitis occurred in one patient on Day 7 (1.6 %),
PEP occurred in three patients (4.5 %) and hyperamylasemia in one (1.6 %) ([Table 3]).There were no bleeding events in either of the two groups.
Late adverse events (≥ 31 days)
In the 12-mm-CSEEMS-group, acute cholecystitis occurred in one patient on Day 77 (3.0 %)
and non-occlusion moderate cholangitis occurred in two patients (6.1 %) on Days 116
and 151 ([Table 3]). In the 10-mm-FCSEMS group, acute cholecystitis occurred in one on Day 32 (1.6 %)
and non-occlusion cholangitis occurred in seven patients (10.6 %) ([Table 3]).
Discussion
Endoscopic drainage of the common bile duct using SEMS is an effective and widely
performed treatment for unresectable malignant biliary obstruction. For patients with
unresectable tumors, SEMS placement maintains biliary flow, relieves jaundice, improves
quality of life, and facilitates delivery of consecutive chemotherapy.
In this study, 12-mm CSEEMS showed a longer TRBO compared with 10-mm FCSEMS. TRBO
was significantly longer in the 12-mm-CSEEMS group than in the 10-mm-FCSEMS group
(log rank, P = 0.001) and both univariate (HR, 0.449; 95 % CI, 0.27967 – 0.72215; P = 0.001) and multivariate (HR, 0.458; 95 % CI, 0.28395 – 0.73744; P = 0.001) Cox hazard analysis found that 12-mm CSEEMS was associated with a significantly
lower risk of RBO. In the 12-mm-CSEEMS group, median TRBO was 232 days and was equal
to median OS, on the other hand, median TRBO was 139.5 days, and the median OS was
169.5 days in 10-mm-FCSEMS group.
Because the time of treatment differed between the two groups, patients with pancreatic
cancer in the 12-mm-CSEEMS group were treated with newly developed chemotherapy, while
those in the 10-mm-FCSEMS received an older chemotherapy regimen [12]
[13]. In the 10-mm FCSEMS-group, 30 out of 38 patients (78.9 %) undergoing chemotherapy
had pancreatic cancer, of whom FOLFILINOX was done in three cases, GnP in one, GEM
in 24 and S-1 in two cases. On the other hand, in the 12-mm-CSEEMS group, 18 out of
20 patients (90 %) undergoing chemotherapy had pancreatic cancer, of whom FOLFILINOX
was done in one case, GnP in 14, GEM in tw and S-1 in one patient. Thus, tumors in
the 12-mm-CSEEMS group may have been more effectively controlled than those in the
10-mm-FCSEMS group. The longer TRBO in the 12-mm-CSEEMS group may be affected by the
difference in chemotherapy regimen. Thus, a further prospective study is needed to
compare TRBO between the two groups.
A meta-analysis of RCT reported better stent patency with CSEMS than with USEMS [6]. It also reported that risk of migration was greater with CSEMS and that there were
no differences between CSEMS and USEMS in occurrence of adverse events such as pancreatitis
or cholecystitis. Other meta-analyses of CSEMS and USEMS found no benefit for CSEMS
[14]
[15]
[16]. In our study, stent patency rate at 6 months was 91.7 % with 12-mm CSEEMS, and
we did not experience stent migration. With the 10-mm-diameter partially-covering
SEMS, stent migration occurred in 7.8 % of patients over 1 year in the WATCH study
[17]. In that study, 10-mm FCSEMS migrated in seven patients (10.6 %) during the observation
period of 12 to 410 days. We believe that the 12-mm CSEEMS proximally bare is effective
for prevention of migration and the larger-caliber style appears to be effective for
preventing occlusion.
Compared with USEMS, FCSEMS has the possibility of removal. CSEEMSs were not removed
because the proximal end bare might injure the bile duct. In two RBO cases caused
by tumor in-growth in the 12-mm-CSEEMS group, stent-in-stent placement was performed
and in one RBO case caused by food impaction, cleaning was performed. In these three
cases, no further RBO was experienced.
In the 12-mm-CSEEMS group, two patients experienced tumor ingrowth (2/33, 6.1 %) at
the covered portion of the stent. Therefore, an improvement in the membrane may be
needed.
The low incidence of non-occlusion cholangitis (3/33, 9.1 %) associated with 12-mm
CSEEMS was satisfactory and similar to the 5.7 % associated with 10-mm CSEMS reported
in the WATCH study [17].
Pancreatitis did not occur in any patients with 12-mm CSEEMS. Kawakubo et al. [18] reported that a high AF and primary diseases other than pancreatic cancer were risk
factors for pancreatitis after SEMS placement. Pancreatitis occurred in three patients
with 10-mm FCSEMS (3/66, 4.5 %), all of whom were treated conservatively. This result
was consistent with occurrence of pancreatitis of 5.9 % associated with CSEMSs [18]. We experienced one case of abdominal pain in a patient with a bile duct diameter
< 7 mm and a 12-mm CSEEMS. Pain was controlled with medication and disappeared after
1 day.
One patient (1/33, 3.0 %) in the 12-mm-CSEEMS group experienced moderate cholecystitis
as defined by the Tokyo 2014 criteria [11]. That patient had gallstones, and the tumor involved the orifice of the cystic duct.
That patient gradually improved with temporary percutaneous intervention. Similar
findings have been previously reported [19]
[20] and in this series, one patient with 10-mm FCSEMS and tumor involvement of the orifice
of the cystic duct and a gallstone experienced cholecystitis (1/66, 1.6 %) as a late
adverse event.
Limitations of this study include a small non-randomized patient sample. However,
because there have been no reports on this 12-mm CSEEMS, we think that it is meaningful
to report this study which demonstrates the superiority of this device.
Conclusion
The 12-mm CSEEMS showed a longer TRBO compared with the widely used 10-mm FCSEMS,
with a similar incidence of adverse events. Therefore, this stent may be safe and
effective for managing malignant distal biliary obstruction. A randomized controlled
trial comparing the novel 12-mm CSEEMS with a conventional 10-mm CSEMS is planned
to assess possible superiority.
