Introduction
Catheter-based radiofrequency ablation (RFA) delivered during endoscopic retrograde
cholangiopancreatography (ERCP) has recently emerged as a possible therapeutic option
within the bile duct [1]
[2]
[3]
[4]. Intrabiliary extension of neoplasm remains an important challenge in the endoscopic
eradication of complex ampullary lesions [5]
[6], and RFA may represent a viable treatment adjunct for this problem. Recently, the
use of RFA at the ampulla and within the distal bile duct has been described [7]
[8]. Herein we present 4 cases assessing the technical feasibility, safety, and treatment
outcomes of RFA employed at the time of ERCP to treat ampullary lesions with intraductal
extension.
Case Reports
The study was conducted at the Medical University of South Carolina (MUSC) from July
1, 2014 through October 1, 2015. After institutional review board approval, we retrospectively
identified eligible adult subjects through the MUSC endoscopy report database (Endoworks,
Olympus America, Center Valley, PA) by searching for reports that contained the keywords
“radiofrequency ablation (RFA)” and “endoscopic retrograde cholangiopancreatography
(ERCP)”. We excluded patients who underwent RFA of a stricture not associated with an ampullary
lesion. We collected relevant clinical, histologic, and endoscopic data on all eligible
subjects.
All procedures were performed by an experienced pancreaticobiliary endoscopist under
general anesthesia using a side-viewing duodenoscope. Included patients had undergone
histologic evaluation of their ampulla prior to treatment. Ampullary resection was
performed either en bloc or in piecemeal fashion by delivering electrosurgical current through a snare with
or without prior submucosal lift. Intraductal extension of the lesion was assessed
cholangiographically ([Fig. 1]) and/or visually ([Fig. 2]). In some cases, a biliary sphincterotomy extension and papillary balloon dilation
was performed to expose the inside of the terminal bile duct for assessment and therapy.
Ablative therapy was delivered using a standard Argon Plasma Coagulation (APC) probe
(ERBE USA Inc., Mariette, GA) at a flow rate of 1.0 L/min to 1.2 L/min and 30 to 40
maximum watts (W) and/or the Habib EndoHPB RFA bipolar cautery probe (EMcision United
Kingdom, London, United Kingdom) at 10 W for 60 to 90 seconds, extrapolating from
manufacturer’s recommendations of 7 to 10 W × 120 seconds [9]. Given the proximity to the pancreatic orifice and the benign nature of the target
lesions, a shorter duration of treatment was chosen. In general, APC was reserved
for treating exposed target tissue in the duodenum or very distal duct, whereas RFA
was reserved for treating hidden or difficult to access tissue within the duct. All
patients undergoing RFA received a temporary pancreatic stent (5 Fr, 2 – 5 cm) and
rectal indomethacin to reduce the risk of post-ERCP pancreatitis (PEP), as well as
a plastic endobiliary prostheses to prevent biliary obstruction and cholangitis.
Fig. 1 Cholangiogram showing a filling defect in the distal bile duct (arrow) representing
bulky intraductal extension of an ampullary adenoma.
Fig. 2 Endoscopic view of the papilla after ampullectomy demonstrating intraductal extension
of the adenoma (arrow).
Technical success was defined as the ability to successfully position the RFA probe
across the biliary orifice and deliver thermal energy to the region of the papilla
and terminal bile duct, resulting in coagulation of the visualized target areas. Clinical
success was defined as endoscopic absence of polypoid or adenomatous-appearing tissue
at the treatment site and histologic absence of neoplasm based on extensive follow-up
biopsies from the papilla, pancreaticobiliary septum, biliary orifice, and distal
bile duct. When the distal bile duct was not fully exposed by prior sphincterotomy,
a pediatric biopsy forceps was introduced into the distal duct to acquire tissue.
We intended to repeat RFA sessions until visual and histologic clearance was observed.
Patient demographics, procedure indications, and treatment outcomes are listed in
[Table 1]. Four eligible patients were identified, all of whom were men with a mean age of
63 years (range 54 – 84). Three patients (75 %) had a history of familial adenomatous
polyposis (FAP). Three patients were treated for ampullary adenoma and 1 for ampullary
adenoma with a focus of adenocarcinoma (he declined surgical evaluation).
Table 1
Characteristics and outcomes of included cases.
|
Patient
|
1
|
2
|
3
|
4
|
|
Age
|
54
|
84
|
54
|
58
|
|
Gender
|
Male
|
Male
|
Male
|
Male
|
|
FAP
|
Yes
|
No
|
Yes
|
Yes
|
|
Histology
|
Adenoma
|
Adenoma with HGD/IMC
|
Adenoma
|
Adenoma
|
|
Sphincterotomy
|
Yes
|
Yes
|
Yes
|
Yes
|
|
Ampullectomy
|
Yes
|
Yes
|
Yes
|
Yes
|
|
APC sessions
|
2
|
2
|
2
|
4
|
|
RFA sessions, mean sec (range)
|
1, 80
|
3, 80 (70 – 90)
|
1, 75
|
1, 80
|
|
Follow up, days
|
56
|
105
|
38
|
51
|
|
Complications
|
No
|
No
|
Bile duct stricture
|
No
|
|
Treatment success
|
Yes
|
No
|
Yes
|
Yes
|
|
Recurrence
|
No
|
Developed cancer
|
No
|
No
|
FAP, familial adenomatous polyposis; LGD, low-grade dysplasia; HGD, high-grade dysplasia;
IMC, intramucosal cancer; APC, argon plasma coagulation; RFA, radiofrequency ablation;
PEP, post-endoscopic retrograde cholangiopancreatography pancreatitis; EUS, endoscopic
ultrasound
[Video 1] presents a synopsis of 2 representative cases. All RFA procedures were technically
successful, resulting in a perceptible tissue effect ([Fig. 3]). RFA was performed immediately following endoscopic resection in 1 case and during
a subsequent session in the remaining cases. The mean number of RFA sessions per patient
was 1.5 (range 1 – 3). All patients were discharged uneventfully after the procedure
without any immediate adverse events (AEs). One patient developed obstructive jaundice
due to a fibro-inflammatory bile duct stricture at the level of prior RFA that manifested
3 days after biliary stent removal (approximately 6 weeks after the RFA) and has required
ongoing endobiliary stent therapy in excess of 3 months. No other AEs have been observed.
During the follow-up period, 3 patients had visual and histologic evidence of complete
eradication; the patient with a focus of adenocarcinoma who declined surgery developed
overt invasive ampullary cancer.
This footage consists of 2 video clips demonstrating catheter-based RFA of intraductal
ampullary adenoma.
Fig. 3 Endoscopic view of the papilla after ampullectomy and intraductal radiofrequency
ablation.
Discussion
Although endoscopic ampullectomy is the preferred treatment for noninvasive ampullary
lesions with a success rate reported as high as 92 % [10], biliary extension of neoplasm represents a significant obstacle to endoscopic eradication.
Exposure and eversion of the adenoma through a biliary sphincterotomy to allow resection
or ablation has been described in amenable cases [5]
[11]
[12]. However, broad adenomatous involvement of the distal bile is associated with limited
treatment success (< 50 %) and has been considered an indication for surgical resection
[5]. Based on its ease of use and the ability to precisely position the probe within
the distal duct, radiofrequency ablation may represent the first viable treatment
adjunct for this challenging scenario. To date, only single case reports of RFA for
benign ampullary lesions have been described; we aimed to expand our understanding
of this technology by presenting our experience in 5 patients.
Catheter-based RFA was technically successful in all cases, and based on short-term
follow up in a small sample, may be safe and clinically effective. However, because
RFA induces thermal injury and subsequent necrosis of the bile duct wall and beyond,
several safety concerns exist. First, while RFA has been associated with a favorable
safety profile when applied to malignant biliary strictures [1]
[2]
[3]
[4], it remains unclear whether RFA in the intra-pancreatic portion of the bile duct
without the protective buffer of a surrounding tumor – especially in the vicinity
of the pancreatic orifice – will be associated with an increased risk of pancreatitis.
Until additional data on the risk of post-ERCP pancreatitis in this context are available,
prophylactic pancreatic stent placement seems reasonable. If the pancreatic and biliary
orifices are in close proximity, especially if adenoma appears to involve the pancreaticobiliary
septum, it may be best to perform the RFA adjacent to a guidewire which has already
been placed in the pancreatic duct (subsequently guaranteeing pancreatic access for
stent placement) rather than adjacent to a plastic pancreatic stent which may be damaged
or even fractured during RFA.
Another safety concern is the development of clinically important post-RFA biliary
strictures that occurred in 1 of our patients, akin to what has been observed in the
esophagus after RFA of Barrett’s epithelium [13]. This concern is particularly relevant in the context of benign ampullary disease
in which patients do not typically undergo long-term stent placement, as is the case
when RFA is performed for palliation of malignant strictures. Along these lines, until
additional data are available, we have attempted to minimize RFA across the cystic
duct takeoff to avoid thermal injury-related obstruction of the cystic duct, which
has intentionally been induced by electrohydraulic lithotripsy to treat refractory
bile leak [14].
In our series, RFA appears to have provided effective adjunctive therapy in all 4
cases of benign pathology but was ineffective in the setting of early adenocarcinoma,
underscoring the concept that surgical resection remains first-line therapy for ampullary
cancer (our patient declined surgery and chemoradiation). Despite the apparent effectiveness
for benign lesions, it is important to consider that intrabiliary extension is often
nodular in nature, leading to heterogeneous contact between the RFA probe and the
target tissue; this may lead to incomplete therapy and/or an increased risk of buried
neoplasm as is the concern when RFA is used to treat nodular Barrett’s esophagus.
Moreover, it can be technically challenging to ensure circumferential contact of the
probe and the target tissue within a dilated bile duct, even when luminal air is suctioned
to induce collapse of the duct around the probe. In these cases, a balloon-based RFA
device that flattens nodular tissue and maximizes treatment contact may be of value.
An additional consideration is that the proximal extent of neoplasm is often difficult
to assess cholangiographically and the role of cholangioscopy to guide probe placement
should be further explored. Prospective studies are necessary to evaluate these issues
and determine the long-term effectiveness of this modality.
In summary, catheter-based RFA after endoscopic resection of ampullary lesions that
extend up the bile duct is technically feasible. Concerns regarding injury to the
pancreas and bile duct as well as incomplete treatment of nodular target tissue exist
and will be addressed by additional clinical experience and research.
