Keywords Biodegradable polymer plug - Transhepatic tract embolization - Transhepatic biliary
drainage
Introduction
A variety of benign and malignant conditions require biliary drainage, either to relieve
biliary obstruction or to divert bile in cases of postoperative bile leakage. Radiologically
placed percutaneous transhepatic biliary drains (PTBDs) or surgically placed transhepatic
biliary drains (STBDs) are preferred when endoscopic biliary access is inappropriate
or fails.
In most cases, PTBD/STBDs can be safely removed when the drains have served their
purpose in a decompressed biliary system. However, in some cases, tract-related biliary,
infectious, or bleeding adverse events (AEs) occur leading to re-interventions and
prolonged care, which can affect the patients’ quality of life [1 ].
Preventive transhepatic tract embolization is one measure to reduce tract-related
AEs [2 ]. The few available studies on tract embolization after biliary interventions have
shown a reduction in bleeding AEs and post-procedural pain [1 ]
[3 ]. Prevention of peritoneal irritation caused by bleeding or bile leakage is assumed
to be the reason for pain relief [3 ]. Although there are no comparative studies available proving the effect of tract
embolization on bile leakage, the reported cumulative incidence of bile leaks after
PTBD/STBD removal with tract embolization was lower than without embolization [2 ]. Therefore, preventive tract embolization appears particularly beneficial in vulnerable
patients with potential risk factors for developing tract-related AEs.
Several embolic agents have been described to occlude the liver tract [4 ]
[5 ]
[6 ]
[7 ]
[8 ]
[9 ]
[10 ]
[11 ]
[12 ]. However, none of the available techniques is suitable for every clinical scenario,
as each method has specific drawbacks depending on the type of application and the
embolization material itself.
The objective of this feasibility study was to estimate the technical and clinical
success rates of liver tract embolization using a biodegradable polymer plug.
Materials and Methods
Patient characteristics and study design
A retrospective, observational, descriptive, and longitudinal single-center feasibility
study from a prospectively obtained database was conducted between February 2022 and
February 2023. The study was approved by the institutional review board, which raised
no fundamental ethical or legal concerns based on the information available (reference
number 20240111 01). STROBE guidelines were followed.
Adult patients were eligible for polymer plug embolization of the tract if they were
scheduled for routine PTBD/STBD removal and had at least one patient- or technique-related
risk factor for developing a tract-related AE. Two categories of risk factors have
been defined based on the literature and our clinical experience: First, risk factors
for bleeding AEs, i.e., coagulopathy, cirrhosis, central bile duct puncture, or previous
drain-related hemorrhage; second, risk factors for bile leakage, i.e., immature tracts
(catheter dwell time <2–3 weeks), large tract diameters, malignant obstruction, or
multilevel strictures [13 ]
[14 ]
[15 ]
[16 ]
[17 ].
If none of the above risk factors were present in the patients’ clinical and imaging
records, tract embolization with gelatin derivatives was performed according to our
institutional standard of care. [5 ]
Removal of the PTBD/STBD and plug embolization was contraindicated in cases where
clinical evidence pointed to persistent biliary congestion, the presence of a necrotic
cavity, biloma, or hematoma at the intended plug position, persistent postoperative
bile leak, uncontrolled or untreated infection, or a known intolerance to one of the
plug components.
Of the 27 patients scheduled for PTBD/STBD removal during the study period, 13 patients
(10 males; median age 61 years [IQR 9]) were enrolled. The 13 participants underwent
15 liver tract embolizations with 15 plugs (12 × 1 plug per patient and 3 plugs in
one patient with 3 bilioenteric anastomoses (BEA)). Indications for PTBD/STBD removal
included: planned endoscopic internalization, resolved postoperative bile leaks, healed
bilioenteric anastomosis, or widened benign biliary strictures. The median time between
the first biliary drainage and tract embolization was 90 days [IQR 34]. Eight patients
had non-dilated ducts and five patients had dilated ducts. Patient characteristics
are presented in [Table 1 ] ([Table 1 ]).
Table 1 Patient characteristics.
Patient
Age/gender
Underlying disease
Indication for PTBD/STBD
Insertion technique
Access side
Bile duct entry level
PTBD/STBD-related adverse events
Rationale for PTBD/STBD removal
Abbr. : BEA : Bilio-enteric anastomosis; CCC : Cholangiocarcinoma; CCE : Cholecystectomy; CRC : Colorectal cancer; ERCP : Endoscopic retrograde cholangiopancreaticography; F : Female; H.o. : History of; IBDI : Iatrogenic bile duct injury; L : Left liver lobe; LTx : Liver transplantation; M : Male; NET : Neuroendocrine tumor; P : Percutaneous/radiological; PsA : Pseudoaneurysm; PSC : Primary sclerosing cholangitis; PTBD : Percutaneous transhepatic biliary drainage; R : Right liver lobe; S : Surgical/intraoperative; STBD : Surgically placed transhepatic biliary drainage.
Bile duct entry/obstruction levels : 0 Central/right or left hepatic duct/bifurcation/BEA, 1 Lobar, 2 Segmental, 3 Subsegmental
1
59 M
Cholestasis + cholangitis;
CRC liver metastases
Failed ERCP
P
R
3
Endoscopic internalization
2
27 M
Cholestasis + cholangitis;
PSC recurrence after LTx
Failed ERCP
P
R
2
Strictures widened
3
56 F
CCC (Bismuth 3A)
Preparation for surgery
P
L
2
Endoscopic internalization
4
59 F
Cholangitis + biliary peritonitis;
IBDI after CCE for cholecystitis
Preparation for surgery
P
R
2
PsA + arterio-biliary fistula, shock
Bile leak resolved
5
63 M
BEA insufficiency after pancreatico-duodenectomy for pancreatic NET
BEA splinting +
decompression
P
R
0
PsA + arterio-biliary fistula, shock
Bile leak resolved
6
78 M
BEA insufficiency after extended surgery for CCC (Bismuth 3B)
BEA splinting +
decompression
S
R
1
Bile leak resolved
7
69 M
Duodenal stump fistula after surgery for cholecystitis and cholecysto-duodenal fistula
Bilio-duodenal decompression
P
R
2
Recurrent bleeding from PsA + arterio-biliary fistula, shock
Bile leak resolved
8
53 M
Biliary strictures + cholangitis;
IBDI after CCE for cholecystitis
Failed ERCP
P
R
2
Endoscopic internalization
9
66 M
CCC (Bismuth 3A)
BEA splinting
S
L
2
BEA healed
10
60 M
Cholangitis + biliary peritonitis;
IBDI after CCE for cholecystitis
BEA splinting
S
R
2
BEA healed
11
62 M
Cholangitis + cholelithiasis;
H.o. distal gastrectomy + Roux-Y reconstruction
Failed ERCP + liver hematoma
P
R
2
Bile stone extraction with internalization
12
71 M
BEA stricture;
H.o. pancreatico-duodenectomy for pancreatic cancer
Failed ERCP
P
R
2
BEA stricture widened
13
68 F
IBDI after CCE for cholecystitis;
CCC as incidental finding
BEA splinting (3 bile ducts)
S
R
2
BEA healed
Liver tract embolization technique and plug design
Written informed consent was obtained from all patients prior to the procedure. The
procedures were performed under local anesthesia in an angiography suite by a board-certified
interventionalist with 27 years of experience. Antibiotic treatment was not routinely
administered. For hepatic tract embolization, the PTBD/STBD was replaced by a braided
7 Fr introducer sheath. A contrast-enhanced tractogram was performed to determine
the tract diameter and biliary access point. When the tip of the sheath was located
anterior to the biliary access point, the sheath was flushed with 5–10 ml of a sodium
chloride solution pre-warmed to 37°C. In each case, 12 mm plugs were selected.
The pushable plug (IMPEDE-FX Embolization Plug, Shape Memorial Medical, Santa Clara,
CA, USA) consists of a self-expanding, biodegradable, low-density polyurethane polymer
and a platinum/iridium marker band ([Fig. 1 ]) [18 ]. The plug is available in sizes of 6 × 10, 8 × 10, and 12 × 15 (diameter × length
in mm). The proximal marker measures between 0.813 and 1.651 mm. During expansion,
the polymer reacts with bile, lymphatic fluid, or blood leaking from the liver parenchyma.
The presumed mechanism of plug action is to prevent or stop bleeding from the peribiliary
plexus as well as from inadvertently traversed blood vessels within the portal triad,
and to block the reflux of bile from the bile duct into the tract. On account of the
viscosity of bile and the low pressure within the biliary system, the polymer, with
average pore sizes ranging from 0.85 to 2.20 mm, creates multiple zones of stagnation
for the bile [19 ]. This allows the hepatic tract to recoil and heal. Plug expansion takes approximately
10 minutes [18 ]. Polymer degradation takes approximately 90–180 days [18 ].
Fig. 1 IMPEDE-FX polymer plug. a Plug introducer (arrowheads indicate the crimped plug). b Crimped polymer plug. Fully expanded plug shown in a lateral c and proximal view d .
As soon as the plug was released proximally to the biliary access point, the sheath
was slightly withdrawn. After 5 minutes, contrast medium was gently injected through
the sheath to check tract occlusion. The sheath was removed when the tract was completely
sealed ([Fig. 2 ]).
Fig. 2 Liver tract embolization after STBD via a long and steep access route. a Initial finding. Note the elevated diaphragm. The access point at the Glisson’s capsule
is marked with a suture (arrowhead). b Magnitude of the cholangiography. The repetitive motion of the elevated diaphragm
and the steep access contributed to a widening of the tract with pericatheter bile
leakage causing intrahepatic (asterisk) and extrahepatic (arrow) biloma. c Plug with radiopaque marker (asterisk) deployed through a kink-resistant 7 Fr sheath
(arrow). d Gentle injection of contrast medium (arrowhead) demonstrates complete occlusion of
the tract 5 minutes after plug deployment.
Post-procedure follow-up
Follow-up consisted of clinical and ultrasound evaluations at 24 hours, 3 months,
and 6 months. One patient did not undergo a pre-discharge ultrasound examination due
to his decision to leave the hospital prior to the scheduled examination. If indicated
by the underlying disease, computed tomography (CT) or magnetic resonance imaging
(MRI) was performed. Moreover, the clinical and imaging records of all patients were
reviewed to obtain further demographic, clinical, and imaging information about the
patients’ medical history and follow-up.
The median follow-up period was 10 months [IQR 9]. At 2 and 16 months of follow-up,
two patients died from progression of the underlying malignancy. One patient was lost
to follow-up after 4 months. [Table 2 ] summarizes procedure and follow-up data ([Table 2 ]).
Table 2 Procedure data and follow-up.
Patient
Plug number
PTBD/STBD diameter (Fr)
Catheter dwell time (days)
Dilated bile ducts
Plug-to-tract oversizing (%)
Technical success
Clinical success
Detected plug marker migration (days after implantation)
Tract obliterated within 6 months
Plug marker detectable on imaging (months)
Adverse events/findings during follow-up
Overall follow-up (months)
Abbr. : CCC : Cholangiocarcinoma; CRC : Colorectal cancer; Fr : French; N : No; NE : Not evaluated; PSC : Primary sclerosing cholangitis; PTBD : Percutaneous transhepatic biliary drainage; STBD : Surgically placed transhepatic biliary drainage; Y : Yes
1
1
8.5
65
Y
300
Y
Y
56
Y
10
The patient died 16 months after plug embolization due to CRC progression
16
2
2
14
242
Y
50
Y
Y
Y
15
Subcapsular biloma within the former liver tract 9 months after embolization due to
PSC progression
15
3
3
8.5
41
Y
100
Y
Y
Y
2
The patient died 2 months after plug embolization due to CCC progression
2
4
4
12
97
N
71
Y
Y
Y
6
6
5
5
12
62
N
71
Y
Y
Y
12
12
6
6
10.2
68
N
50
Y
Y
30
Y
1
20
7
7
10.2
44
N
71
Y
Y
Y
6
6
8
8
8.5
85
Y
33
Y
N
N
11
Bilio-cutaneous fistula
15
9
9
8.5
15
N
100
Y
Y
Y
4
Lost to follow-up after 4 months
4
10
10
10.2
10
N
71
Y
Y
Y
14
14
11
11
8.5
5
N
100
Y
Y
Y
6
6
12
12
14
460
Y
20
Y
N
1
Y
0
Bilio-cutaneous fistula
9
13
13
8.5
49
N
NE
Y
Y
130
Y
4
10
14
8.5
54
N
NE
Y
Y
286
Y
4
10
15
6
54
N
NE
Y
Y
123
Y
10
10
Study endpoints and definitions
The primary study endpoints were technical and clinical success. Technical success
was defined as successful plug deployment in the intended position. Clinical success
was defined as the absence of biliary AEs (biliocutaneous fistula, biloma, and choleperitoneum/cholethorax),
infectious AEs, or bleeding AEs within 6 months of follow-up. The grading of AEs was
conducted in accordance with the guidelines of the Society of Interventional Radiology
[20 ].
Additionally, to assess clinically inapparent hemorrhage or biliary obstruction, pre-
and post-procedural hemoglobin, hematocrit and bilirubin levels were compared. Secondary
endpoints included plug migration, plug oversizing, and plug visibility on ultrasound,
CT, and MRI.
Plug migration within 24 hours was defined as early migration. In addition, migration
and oversizing were analyzed regarding the presence of dilated or non-dilated ducts.
Bile ducts were considered dilated if the diameter of an intrahepatic peripheral bile
duct was wider than 2 mm or if the diameter of the bile duct exceeded that of the
accompanying portal vein. Bile duct and liver tract diameters were determined by periprocedural
tractography, cholangiography, and ultrasound. Plug oversizing was calculated with
respect to the diameters of the liver tract and the accessed bile duct.
Statistical analysis
Statistical analysis was performed using GraphPad Prism (version 10.1.2 (324), GraphPad
Software Inc., San Diego, CA, USA). Normal distribution was tested using the Shapiro-Wilk
test. Categorical variables are expressed as numbers with percentage. Normally distributed
variables are presented as mean and standard deviation (SD), non-normally distributed
variables are shown as median and interquartile range (IQR). Statistical testing was
performed using Fisher’s exact test and Wilcoxon matched-pairs signed rank test, where
appropriate. A p-value less than 0.05 was considered statistically significant.
Results
Technical and clinical success
The technical success rate was 100% (15/15 plugs). The clinical success rate was 84.6%
(11/13 patients). Two patients with dilated ducts developed biliocutaneous fistulas.
In both cases, the persistence of relevant biliary obstruction was clinically underestimated
prior to drain removal: One patient had a long-term (460 days) 14 Fr internal-external
biliary drain for the treatment of a BEA stricture due to scarring. After one week
of uneventful clamping of the external part of the drain, the anastomosis was considered
wide enough to allow adequate trans-anastomotic bile outflow into the efferent bowel
loop. However, a biliocutaneous fistula occurred after PTBD removal, indicating persistent
biliary congestion. In this case, bile leakage was present for 3 days until successful
endoscopic BEA stenting. The second patient with multilevel strictures due to cholangitis
had a biliocutaneous fistula after removal of a right-sided PTBD. Before removal,
the external part of the drain was disconnected without any signs of biliary obstruction.
In this case, the fistula persisted for 5 months because endoscopic stenting was successful
only in the left hepatic duct. Secretion from the fistula finally ceased after completion
of bilateral stenting. However, undulating subcutaneous swelling at the former percutaneous
access site for the next 10 months indicated persistence of the tract. Apart from
the two fistulas, there were no biliary, infectious, or bleeding AEs within 6 months
of follow-up.
Periprocedural hemoglobin and hematocrit levels were unremarkable, excluding clinically
occult bleeding. The median delta was 2% [IQR 2.2] (p=0.463) for hematocrit and 0.7
mg/dl [IQR 0.95] (p=0.673) for hemoglobin. Pre- and postintervention bilirubin levels
differed insignificantly by a median of 0.6 mg/dl [IQR 0.28] (p=0.502).
Beyond the 6-month study period, a post-liver transplant patient developed a subcapsular
biloma 9 months after plug embolization within the previously obliterated tract. At
that time, the patient had progressive cholestasis due to worsening of sclerosing
cholangitis within the graft. The plug marker was in a stable position. The biloma
resolved after endoscopic intervention.
Migration
Early plug migration within 24 hours occurred in the above patient who had a fistula
for 3 days until endoscopic biliary decompression. In the second clinical failure
case mentioned above, the plug marker was in a stable position. Five late plug marker
losses were noted. In these cases, the tracts healed without delay and no migration-related
biliary obstruction or tract reopening occurred. Four of the five late marker losses
were encountered in cases with non-dilated ducts at 30, 123, 130, and 286 days. 3
months after plug embolization, the fifth marker migrated into a dilated common hepatic
duct in a case with a malignant stenosis.
As a result, clinical failure occurred in 2 out of 13 patients including one case
of early plug migration resulting in a 15.4% rate of Grade 3 AEs.
Oversizing
Bile duct diameters at the PTBD entry point ranged from 3 to 10 mm (median 5 mm [IQR
1.37]). The diameter of the tract within the liver parenchyma ranged from 3 to 10
mm (median 7 mm [IQR 2]). The median plug-to-duct oversizing was 173% [IQR 60] and
the median plug-to-tract oversizing was 86% [IQR 50]. Regarding plug-to-tract oversizing
in patients with immediate tract closure, the median oversizing was 150% [IQR 125]
in the dilated subgroup and 76% [IQR 14.5] in the non-dilated subgroup. In the 2 cases
of clinical failure, the plug-to-tract oversizing was considerably lower (33% and
20%). There was no statistically significant association between clinical success
and the presence of non-dilated or dilated ducts (p=0.095). The association between
clinical success and more than 50% plug-to-tract oversizing was statistically significant
(p=0.015).
Visibility on imaging
The plug was visible as an echogenic structure on ultrasound ([Fig. 3 ]). The marker induced minimal beam hardening artifacts on CT images. The plug could
not be clearly identified on MRI, especially in the presence of aerobilia. MRI was
not affected by plug-related artifacts.
Fig. 3 Visibility of the plug on imaging. B-mode ultrasound images at 17 hours a , 3 months b , and 6 months c after plug embolization. The expanded polymer (white arrow) is visible in a stable
position as an echogenic structure within the progressively obliterating liver tract
(black arrows). A sharp reflex indicates the proximal marker (white arrowhead). d Transverse contrast-enhanced fat-saturated T1-weighted 3D Volumetric Interpolated
Breath-hold Examination magnetic resonance image 5 months after plug placement showing
a thin scar remnant (black arrows) of the healed liver tract. The plug does not induce
artifacts.
Discussion
The polymer plug was precisely deployed within the liver tract without inadvertent
propagation into the bile ducts, even in long and steep access routes of surgically
placed drains ([Fig. 2 ]). This is a potential advantage over liquid agents such as gelfoam slurry or glue
[2 ]
[5 ]. Contrary to glue, coils, or metal plugs, the plug polymer degrades. Temporary agents
for parenchymal tract embolization are considered superior to permanent materials
or agents [11 ]. As episodes of cholestasis or cholangitis are common after major hepatobiliary
surgery, treatment of iatrogenic biliary injury, or liver transplantation, a permanent
embolic material may serve as a trigger for chronic or recurrent infection [11 ]. In addition, unlike certain metallic devices, the polymer plug does not interfere
with future liver imaging.
In the present study, tractograms showed that the diameter of the tract could be considerably
larger than the size of the indwelling catheter. Steep access angles may contribute
to a continuous widening of the tract by repetitive diaphragmatic motion as shown
in [Fig. 2 ]. Therefore, to ensure adequate filling of the tract, the plug size should be determined
based on the diameter of the hepatic tract, rather than on the size of the indwelling
drain. To obtain a more precise estimation of the tract diameter, it is recommended
to perform a contrast-enhanced tractogram after the PTBD/STBD has been replaced over
a guidewire by an introducer sheath.
This is particularly important as the clinical failures observed were due to a combination
of suboptimal plug sizing and inadequate biliary decongestion. As previously reported,
biliary decompression is a prerequisite for successful and durable tract occlusion
[10 ]. Regardless of the device used, bile leakage cannot be permanently prevented if
the biliary obstruction persists or recurs. For this reason, the plug does not appear
to be suitable for patients with multilevel benign or malignant biliary strictures
where complete drainage of the biliary system is not feasible (e.g., sclerosing cholangitis
or Bismuth ≥3).
Based on this preliminary experience, the authors suggest substantial oversizing of
more than 50% for this particular plug in relation to the liver tract diameter. In
cases where sufficient plug-to-tract oversizing is not possible due to the unavailability
of appropriate plug sizes, tract embolization with more than one plug could allow
for sustained tract occlusion. However, this needs to be further investigated.
It should also be noted that in the present study the biliocutaneous fistulas occurred
after long periods of catheterization (85 and 460 days), as bile leaks are usually
more common after shorter catheter dwell times.
The mechanism of clinically inapparent late marker migration remains unclear. One
may speculate that this is caused by polymer degradation or inadequate plug integration,
which would be of concern as the plug may have negatively impacted the healing of
the tract. However, this needs to be studied further with a larger number of cases.
One may discuss whether the use of the plug in the liver tract is covered by the instructions
for use. The intended indication for use of the plug is to obstruct or reduce the
rate of blood flow in the peripheral vasculature [18 ]. The instructions do not explicitly state what type of vessel is meant or that the
plug has to be deployed exclusively in the lumen of a blood vessel. It is, therefore,
our considered opinion that the deployment of the plug in the transition zone between
the bile duct and the liver parenchyma, including the portal triad, where the bile
duct, peribiliary plexus, branches of the portal vein, and hepatic artery are in close
proximity, with the aim of either stopping or preventing bleeding complications and
bile leakage, could be considered a combined use in peripheral blood and bile vessels.
This is in accordance with the prevailing view in the hepatologic literature regarding
the portal triad as a functional vascular-biliary unit, with biliary and vascular
structures linked by a close anatomic and functional association [21 ]
[22 ].
In view of the study results, the plug should be used in appropriate cases after careful
consideration. In the authors’ opinion, it may be an option for select patients with
a definitively decompressed biliary system who are at risk for bleeding complications,
including central bile duct puncture with inadvertent passage through large blood
vessels and known erosion of the PTBD into accompanying blood vessels with previous
bleeding episodes. Patients with benign or malignant multilevel biliary strictures
who are at risk for persistent or recurrent bile flow obstruction are not good candidates
for liver tract embolization with the polymer plug.
This study has several limitations. It represents a single institution, non-randomized
preliminary experience with a small sample size and a heterogeneous patient population.
The efficacy of tract embolization cannot be clearly assessed due to the lack of a
control group without embolization.
Conclusion
Polymer plug embolization of the liver tract after PTBD/STBD is technically feasible.
The plug could be an option in select cases complicated by patient- or insertion technique-related
risk factors for the development of tract-related bleeding AEs where a non-permanent
embolic device that can be precisely delivered is desired. However, the plug does
not appear to be suitable for patients with multilevel strictures where complete drainage
of the biliary system is not possible. Further studies including a larger sample size
are needed for appropriate patient selection for embolization using this device.
Clinical relevance
The biodegradable polymer plug can be precisely delivered within the liver tract.
Plug sizing should be based on the diameter of the hepatic tract, rather than on the
size of the indwelling catheter to ensure adequate filling of the tract.
Adequate plug-to-tract oversizing and biliary decongestion are essential to achieve
successful and durable tract closure.
The plug appears unsuitable for patients with multilevel biliary strictures where
complete drainage of the biliary system is not feasible.
