- Email: [email protected]
Magnetic pancreaticobiliary stents and retrieval system: obviating the need for repeat endoscopy (with video)
NEW METHODS: Experimental Endoscopy
Magnetic pancreaticobiliary stents and retrieval system: obviating the need for repeat endoscopy (with video) Marvin Ryou, MD,1 Padraig Cantillon-Murphy, PhD,5 Sohail N. Shaikh, MD,4 Dan Azagury, MD,3 Michele B. Ryan, MS,1 Jeffrey H. Lang, ScD,2 Christopher C. Thompson, MD, MSc, FACG, FASGE1 Boston, Cambridge, Massachusetts; Tucson, Arizona; Cork, Ireland
Background: Plastic stents are routinely placed in the pancreaticobiliary system to facilitate drainage. A second endoscopy is often required for stent removal. We have developed magnetic pancreaticobiliary stents that can be removed by using an external hand-held magnet, thereby obviating the need for a second endoscopy. Objective: To develop and test magnetic pancreaticobiliary stents and retrieval system in ex-vivo and in-vivo porcine models. Setting: Animal laboratory. Design: Benchtop and animal study. Animals: 5 pigs. Interventions: Design: Computer simulations determined both the optimal design of cylindrical magnets attached to the distal aspect of existing plastic stents and the optimal design of the external hand-held magnet. Benchtop ex-vivo experiments measured magnetic force to validate the design. In-vivo analysis: In 5 Yorkshire pigs, magnetic stents were deployed into the common bile duct by using a conventional duodenoscope. An external hand-held magnet was applied for stent removal. Stent insertion and removal times were recorded. Main Outcome Measurements: Technical feasibility. Results: Magnetic stents of varying lengths and calibers were successfully created. In ex-vivo testing, the capture distance was 10.0 cm. During in-vivo testing, the magnetic stents were inserted and removed easily. The mean insertion and removal times were 3.2 minutes and 33 seconds, respectively. Limitations: Animal study, small numbers. Conclusions: Magnetic pancreaticobiliary stents and associated retrieval system were successfully designed and tested in the acute porcine model. An external, noninvasive means of stent removal potentially obviates the need for a second endoscopy, which could represent a major gain both for patients and in health care savings.
Plastic stents are routinely placed in the pancreaticobiliary tree to promote temporary drainage. Some applications require only stent removal, not necessarily repeated ERCP. For example,
plastic biliary stents are placed for postsurgical bile leaks1,2 and do not require repeated ERCP if the leak is clinically resolved.3 Similarly, pancreatic stents are placed for prophylaxis of post-
DISCLOSURE: The authors disclosed no financial relationships relevant to this publication.
doi:10.1016/j.gie.2011.09.051
Use your mobile device to scan this QR code and watch the article video. Download a free QR code scanner by searching ‘QR Scanner’ in your mobile device’s app store. Copyright © 2012 by the American Society for Gastrointestinal Endoscopy 0016-5107/$36.00
888 GASTROINTESTINAL ENDOSCOPY Volume 75, No. 4 : 2012
Received July 28, 2011, Accepted September 30, 2011. Current affiliations: Division of Gastroenterology (1) and Department of Surgery (3), Brigham & Women’s Hospital, Harvard Medical School, Boston, Massachusetts, USA; Department of Electrical Engineering and Computer Science, Massachusetts Institute of Technology, Cambridge, Massachusetts, USA (2); Section of Gastroenterology, University of Arizona College of Medicine, University Medical Center, Tuscon, Arizona, USA (4), Department of Electrical Engineering, University College Cork, Cork, Ireland (5). Reprint requests: Christopher Thompson, MD, Division of Gastroenterology, 75 Francis St, Boston, MA 02115.
www.giejournal.org
Ryou et al
Magnetic pancreaticobiliary stents and retrieval system
Take-home Message ●
Figure 1. A, Magnetic stent deployment. B, Noninvasive retrieval.
ERCP pancreatitis, particularly in certain high-risk populations.4 In these circumstances, the rate of spontaneous stent dislodgement is highly variable,5 and a more controlled process of stent removal may be desirable. We have developed a noninvasive retrieval system for the removal of magnetic pancreaticobiliary stents. One iteration of this system uses an external magnet for noninvasive removal of magnetically tipped stents, thereby obviating the need for a second endoscopy. The stent is extracted into the duodenum and passed spontaneously through the gastrointestinal tract (Fig. 1). Another iteration of this system is a magnetically tipped catheter to retrieve proximally migrated or buried stents. The aim of this proof-of-principle study was to develop and test a noninvasive retrieval system for magnetic pancreaticobiliary stents in ex-vivo and in-vivo porcine models.
METHODS AND MATERIALS Magnetic retrieval system The stents were magnetically tipped Geenen pancreatic stents (Cook Medical, Inc., Bloomington, IN); 3F ⫻3 cm, www.giejournal.org
A novel noninvasive retrieval system for magnetic pancreaticobiliary stents was successfully demonstrated in the porcine model. This type of system may obviate the need for follow-up endoscopy in certain populations.
5F ⫻5 cm, and 7F ⫻5 cm stents were assembled, both with and without internal flanges. These were tipped with neodymium-iron-boron ring magnets (N-42 grade, axially magnetized, outer diameter 3.175 mm, internal diameter 1.5875 mm; KJ Magnetics, Jamison, PA) stacked to achieve varying lengths (9.5 mm–16.9 mm) (Fig. 2A). Attachment of magnets to stents was achieved by using medical-grade adhesive (Loctite 4541 PrismGel, Henkel Corp., Dusseldorf, Germany). The external hand-held magnet was a 3.8-pound (1.78-kg) neodymium-iron-boron disk magnet (cylindrical N-42 grade with axial magnetization; KJ Magnetics, Jamison, PA) (Fig. 2B). Important design parameters for the magnetic retrieval system included the following: (1) magnetically tipped stents that could be delivered into the pancreaticobiliary tree via a conventional duodenoscope (ie, with a 4.2-mm instrument channel), (2) an arbitrarily chosen 5-pound maximum for an external hand-held magnet, and (3) a target capture distance of 10 cm to simulate a nonobese body habitus. A more thorough description of rational device design, computer modeling, and bench-top ex-vivo experiments to measure forces and validate design has been previously detailed6 and is also summarized in the appendix (available online at www.giejournal.org).
In-vivo testing in acute porcine laboratory studies Animal laboratory studies were conducted at Concord Biomedical Sciences and Emerging Technologies (Lexington, MA) after approval from the Institutional Animal Care and Use Committee. In-vivo porcine studies were performed in 5 Yorkshire pigs weighing approximately 35 to 40 kg. After induction of anesthesia by using an intramuscular injection of Telazol 4.4 mg/kg, xylazine 1.1 mg/kg, and atropine 0.05 mg/kg (all medications from Fort Dodge Animal Health, Fort Dodge, IA), the pigs were intubated, and anesthesia was maintained with isoflurane 1% to 4%. Magnetic stents were deployed into the common bile duct by using a conventional duodenoscope (Olympus TJF-160F; Olympus America Inc., Center Valley, PA). Biliary cannulation was achieved by using a soft-tipped 0.035 Jagwire (Boston Scientific Corporation, Natick, MA) through a ProForma tapered-tip catheter (ConMed Corporation, Largo, FL). Biliary cannulation was chosen instead of pancreatic duct cannulation because the porcine pancreatic duct in juvenile pigs is very small and has a separate, more distal Volume 75, No. 4 : 2012 GASTROINTESTINAL ENDOSCOPY 889
Magnetic pancreaticobiliary stents and retrieval system
Ryou et al
TABLE 1. Ex-vivo capture distances with varying magnetic ring lengths
Length (cm)
Distance (cm)
Force (mN)
Internal flange (ⴙ/-)
9.5
2.5 ⫾ 0.1
680
⫹
9.5
9.0 ⫾ 0.3
59.2
⫺
9.5
9.0 ⫾ 0.6
59.2
⫺
12.7
4.0 ⫾ 0.1
361.1
⫹
12.7
10.0 ⫾ 0.4
44.1
⫺
12.7
9.0 ⫾ 0.4
59.2
⫺
15.9
4.0 ⫹ 0.1
361.1
⫹
15.9
10.0 ⫾ 0.3
44.1
⫺
Figure 2. A, Magnetic stent. B, External hand-held magnet.
orifice.7 Two pigs received a 5F ⫻ 5 cm stent, and 3 pigs received a 5F ⫻ 3 cm stent. Each stent was tipped with a 9.5⫺mm long cylindrical magnetic extension (weight, 0.42 g), and the internal flanges were removed. With fluoroscopic and endoscopic guidance, the external handheld magnet was applied to the anterior abdominal surface for stent removal. The time of stent insertion (time from stent insertion into instrument channel to removal of Jagwire from bile duct and stent) and the time of stent removal (time from application of external magnet to endoscopic visualization of proximal aspect of stent within duodenal lumen) were recorded. Fluoroscopic and endoscopic footage were recorded. Animals were euthanized at the conclusion of the procedure.
RESULTS Ex-vivo testing Table 1 shows the ex-vivo capture distances of various magnetic ring lengths interacting with the 3.8-pound external magnet (Table 1). Both 5F ⫻ 3 cm and 5F ⫻ 5 cm stents were used in these experiments (Fig. 3). Surprisingly, a magnetic ring length of 9.5 mm without internal flanges achieved a capture distance of 9.0 cm, which was very close to the target capture distance of 10 cm. Increases in length to 12.7 mm and 15.9 mm yielded minimal increases in capture distance. Therefore, it was decided to use 5Fr stents (both 3 cm and 5 cm) without internal flanges and with a magnetic ring length of 9.5 mm for subsequent acute animal laboratory studies. 890 GASTROINTESTINAL ENDOSCOPY Volume 75, No. 4 : 2012
Figure 3. Ex-vivo tests to measure capture distance (15.9-mm magnet assembly shown [longest tested]).
Acute animal laboratory studies As shown in the video (available online at www. giejournal.org), upon successful location of the porcine ampulla, the mean time for stent insertion into the bile duct was 3.2 minutes (range, 2.75 minutes–5.8 minutes). The mechanics of stent deployment were easy and not technically different from the mechanics of deploying a nonmodified stent. The stents were inserted until the proximal portion of the magnetic rings contacted the ampulla (Fig. 4). The mean removal time was 33 seconds (range, 25 seconds– 44 seconds). Fluoroscopy revealed an approximate capture distance of 6 cm in all pigs (within the specifications of the system design). Capture was immediate upon application of the external magnet. Maneuvering of the external magnet was similar for all animals. The stents were completely extracted into the duodenum in all animals as corroborated by endoscopic visualization. There was no evidence of mucosal or tissue trauma during stent removal. www.giejournal.org
Ryou et al
Magnetic pancreaticobiliary stents and retrieval system
Figure 4. A-C, In-vivo stent deployment. D-F, Removal.
DISCUSSION This proof-of-concept animal study successfully demonstrated a magnetic retrieval system for pancreaticobiliary stents. Magnetically tipped stents were deployed in a conventional manner through a duodenoscope. The stents were then fully extracted into the duodenum by using a hand-held external magnet. The magnetic retrieval system targets applications of temporary stenting that do not necessarily require repeat ERCP. Examples include pancreatic duct stenting for prophylaxis of post-ERCP pancreatitis in certain high-risk groups (sphincterotomy for sphincter of Oddi dysfunction, ampullectomy, pancreatic sphincterotomy, difficult cannulation)4 and for postcholecystectomy biliary leak.3 The main study limitation was that it was an acute animal study. To minimize the effects of anatomic differences, our modified pancreatic stents were inserted into the porcine common bile duct (the pancreatic duct in juvenile pigs is extremely small and has a separate distal orifice from the biliary orifice7). Consequently, maneuvering of the external magnet for stent extraction will likely be different for stents deployed in the pancreatic duct and different for human anatomy. However, anatomic differences likely do not influence feasibility, as evidenced by the successful use of external magnets to guide magnetically tipped enteral feeding tubes through the pylorus and duodenal sweep in human studies.8 Other parameters also require further study. For example, how will stent retrieval be affected by long-term retention or modifications to stent size? Internal flanges in particular will be an important feature to include in subsequent investigations to prevent spontaneous dislodgement, especially for applications that require placement for many days to weeks. www.giejournal.org
Extrapolation of our results to larger stents is unclear. A 5Fr stent was used in this study because it appeared to be the most suitable size for our particular animal model. Successful demonstration in larger-diameter stents may require larger pigs or a different animal model altogether. Although magnets have been shown to be safe for use in GI endoscopy for foreign body retrieval,9 treatment of bile duct stenosis,10 and retraction during endoscopic submucosal dissection,11 the magnetic component of the retrieval system may lead to certain restrictions in implementation. For example, this retrieval system may not be applicable to patients with pacemakers and automatic implantable cardioverter-defibrillators, and magnetic resonance imaging may not be performed while the stents are in place. In addition, the magnetic retrieval system may be difficult to use in morbidly obese patients, given the excessive capture distance. However, in bench-top models, the current prototype achieved a capture distance of 10 cm in spite of an arbitrarily chosen limitation of 5 pounds for the external magnet. It would certainly be possible to scale up the power of the retrieval system with stronger magnets or an electromagnet. Finally, there is the theoretic risk of accidentally pushing in the stent if the external magnet were accidentally oriented to cause magnet–magnet repulsion. However, this risk can be mitigated by retaining external flanges. Future work will focus on several goals: increase in magnetic strength, determination of prescribed motions and maneuvers of the external magnet for reproducible stent removal in human anatomy, development of a realtime feedback for capture without the use of fluoroscopy, and development of a through-the-scope retrieval catheter for migrated magnetic stents. Volume 75, No. 4 : 2012 GASTROINTESTINAL ENDOSCOPY 891
Magnetic pancreaticobiliary stents and retrieval system
In conclusion, a magnetic stent retrieval system is technically feasible in the porcine model. This may potentially obviate the need for a second endoscopy and represent a major gain both for patients and in health care savings. REFERENCES 1. Mergener K, Strobel JC, Suhocki P, et al. The role of ERCP in diagnosis and management of accessory bile duct leaks after cholecystectomy. Gastrointest Endosc 1999;50:527-31. 2. Sandha GS, Bourke MJ, Haber GB, et al. Endoscopic therapy for bile leak based on a new classification: results in 207 patients. Gastrointest Endosc 2004;60:567-74. 3. Coelho-Prabhu N, Baron TH. Assessment of need for repeat ERCP during biliary stent removal after clinical resolution of post-cholecystectomy bile leak. Am J Gastroenterol 2010;105:100-5. 4. Freeman ML. Pancreatic stents for prevention of post-endoscopic retrograde cholangiopancreatography pancreatitis. Clin Gastroenterol Hepatol 2007;5:1354-65.
Ryou et al
5. Rashdan A, Fogel EL, McHenry L, et al. Improved stent characteristics for prophylaxis of post-ERCP pancreatitis. Clin Gastroenterol Hepatol 2004; 2:322-9. 6. Cantillon-Murphy P, Ryou M, Shaikh SN, et al. A magnetic retrieval system for stents in the pancreaticobiliary tree. IEEE Trans Biomed Eng 2010;57:2018-25. 7. Mullen Y, Taura Y, Nagata, et al. Swine as a model for pancreatic beta-cell transplantation. In: Swindle MM, ed. Swine as Models in Biomedical Research. Ames, IA: Iowa State University Press; 1992:1634. 8. Gabriel SA, Ackermann RJ. Placement of nasoenteral feeding tubes using external magnetic guidance. J Parenter Enteral Nutr 2004;28:119-22. 9. Paulson E, Jaffe R. Metallic foreign bodies in the stomach: fluoroscopic removal with a magnetic orogastric tube. Radiology 1990;174:191-4. 10. Muraoka N, Uematsu H, Yamanouchi E, et al. Yamanouchi magnetic compression anastomosis for bilioenteric anastomotic stricture after living-donor liver transplantation. J Vasc Interv Radiol 2005;16: 1263-7. 11. Gotoda T, Oda I, Tamakawa K, et al. Prospective clinical trial of magneticanchor guided endoscopic submucosal dissection for large early gastric cancer (with videos). Gastrointest Endosc 2009;69:10-15.
GIE on Facebook GIE now has a Facebook page. Fans will receive news, updates, and links to author interviews, podcasts, articles, and tables of contents. Search on Facebook for “GIE: Gastrointestinal Endoscopy” and become a fan.
892 GASTROINTESTINAL ENDOSCOPY Volume 75, No. 4 : 2012
www.giejournal.org
Ryou et al
APPENDIX System requirements Design of the overall magnetic system was guided by the following requirements and restrictions: (1) The modified stent would have to be deployed in conventional fashion through a duodenoscope. (2) The modified stent would have to fit down a 4.2 mm instrument channel. (3) As an extension of the stent, the attached magnets should also feature a lumen to provide continued drainage. (4) The external hand-held magnet would have an arbitrarily chosen 4-pound weight restriction in order to feel comfortable for most operators. (5) The overall system should provide sufficient capture force at 10 cm (an approximate estimate of separation distance in a non-obese human body habitus) to dislodge the stent completely from the duct.
Design optimization The dimensions of the hand-held magnet were optimized for a fixed weight and a cylindrical shape (to minimize edges and corners) via numerical simulation using MATLAB (Mathworks, Inc., Natick, MA). For the magnet
www.giejournal.org
Magnetic pancreaticobiliary stents and retrieval system
attachments, axially-magnetized ring magnets were chosen because of Requirements #1 and #3 listed previously. A ring of outer diameter 3.175 mm and inner diameter 1.5875 mm was chosen as this was this was the largest, off-the-shelf commercially available dimensions that could fit down an instrument channel diameter of 4.2 mm. Attractive force between an optimized cylindrical hand-held magnet and various ring magnet assembly lengths (6.35 mm, 9.525 mm, 12.7 mm, 15.9 mm) were modeled using MATLAB.
Ex vivo validation tests Measurements of attractive force between various magnetic ring lengths and the hand-held magnet were corroborated experimentally using an aluminum bending-beam force gauge as beam deflection correlated to a known load.6 Ring lengths of 9.525 mm, 12.7 mm, and 15.9 mm were attached to Geenen 5F ⫻ 3 cm and 5F ⫻ 5 cm stents. In ex vivo tissue experiments, these modified stents were inserted into the porcine bile duct. The external hand-held magnet was incrementally advanced towards the stent until it completely dislodged. Capture distance was subsequently recorded.
Volume 75, No. 4 : 2012
GASTROINTESTINAL ENDOSCOPY
892.e1
Magnetic pancreaticobiliary stents and retrieval system: obviating the need for repeat endoscopy (with video) Marvin Ryou, MD,1 Padraig Cantillon-Murphy, PhD,5 Sohail N. Shaikh, MD,4 Dan Azagury, MD,3 Michele B. Ryan, MS,1 Jeffrey H. Lang, ScD,2 Christopher C. Thompson, MD, MSc, FACG, FASGE1 Boston, Cambridge, Massachusetts; Tucson, Arizona; Cork, Ireland
Background: Plastic stents are routinely placed in the pancreaticobiliary system to facilitate drainage. A second endoscopy is often required for stent removal. We have developed magnetic pancreaticobiliary stents that can be removed by using an external hand-held magnet, thereby obviating the need for a second endoscopy. Objective: To develop and test magnetic pancreaticobiliary stents and retrieval system in ex-vivo and in-vivo porcine models. Setting: Animal laboratory. Design: Benchtop and animal study. Animals: 5 pigs. Interventions: Design: Computer simulations determined both the optimal design of cylindrical magnets attached to the distal aspect of existing plastic stents and the optimal design of the external hand-held magnet. Benchtop ex-vivo experiments measured magnetic force to validate the design. In-vivo analysis: In 5 Yorkshire pigs, magnetic stents were deployed into the common bile duct by using a conventional duodenoscope. An external hand-held magnet was applied for stent removal. Stent insertion and removal times were recorded. Main Outcome Measurements: Technical feasibility. Results: Magnetic stents of varying lengths and calibers were successfully created. In ex-vivo testing, the capture distance was 10.0 cm. During in-vivo testing, the magnetic stents were inserted and removed easily. The mean insertion and removal times were 3.2 minutes and 33 seconds, respectively. Limitations: Animal study, small numbers. Conclusions: Magnetic pancreaticobiliary stents and associated retrieval system were successfully designed and tested in the acute porcine model. An external, noninvasive means of stent removal potentially obviates the need for a second endoscopy, which could represent a major gain both for patients and in health care savings.
Plastic stents are routinely placed in the pancreaticobiliary tree to promote temporary drainage. Some applications require only stent removal, not necessarily repeated ERCP. For example,
plastic biliary stents are placed for postsurgical bile leaks1,2 and do not require repeated ERCP if the leak is clinically resolved.3 Similarly, pancreatic stents are placed for prophylaxis of post-
DISCLOSURE: The authors disclosed no financial relationships relevant to this publication.
doi:10.1016/j.gie.2011.09.051
Use your mobile device to scan this QR code and watch the article video. Download a free QR code scanner by searching ‘QR Scanner’ in your mobile device’s app store. Copyright © 2012 by the American Society for Gastrointestinal Endoscopy 0016-5107/$36.00
888 GASTROINTESTINAL ENDOSCOPY Volume 75, No. 4 : 2012
Received July 28, 2011, Accepted September 30, 2011. Current affiliations: Division of Gastroenterology (1) and Department of Surgery (3), Brigham & Women’s Hospital, Harvard Medical School, Boston, Massachusetts, USA; Department of Electrical Engineering and Computer Science, Massachusetts Institute of Technology, Cambridge, Massachusetts, USA (2); Section of Gastroenterology, University of Arizona College of Medicine, University Medical Center, Tuscon, Arizona, USA (4), Department of Electrical Engineering, University College Cork, Cork, Ireland (5). Reprint requests: Christopher Thompson, MD, Division of Gastroenterology, 75 Francis St, Boston, MA 02115.
www.giejournal.org
Ryou et al
Magnetic pancreaticobiliary stents and retrieval system
Take-home Message ●
Figure 1. A, Magnetic stent deployment. B, Noninvasive retrieval.
ERCP pancreatitis, particularly in certain high-risk populations.4 In these circumstances, the rate of spontaneous stent dislodgement is highly variable,5 and a more controlled process of stent removal may be desirable. We have developed a noninvasive retrieval system for the removal of magnetic pancreaticobiliary stents. One iteration of this system uses an external magnet for noninvasive removal of magnetically tipped stents, thereby obviating the need for a second endoscopy. The stent is extracted into the duodenum and passed spontaneously through the gastrointestinal tract (Fig. 1). Another iteration of this system is a magnetically tipped catheter to retrieve proximally migrated or buried stents. The aim of this proof-of-principle study was to develop and test a noninvasive retrieval system for magnetic pancreaticobiliary stents in ex-vivo and in-vivo porcine models.
METHODS AND MATERIALS Magnetic retrieval system The stents were magnetically tipped Geenen pancreatic stents (Cook Medical, Inc., Bloomington, IN); 3F ⫻3 cm, www.giejournal.org
A novel noninvasive retrieval system for magnetic pancreaticobiliary stents was successfully demonstrated in the porcine model. This type of system may obviate the need for follow-up endoscopy in certain populations.
5F ⫻5 cm, and 7F ⫻5 cm stents were assembled, both with and without internal flanges. These were tipped with neodymium-iron-boron ring magnets (N-42 grade, axially magnetized, outer diameter 3.175 mm, internal diameter 1.5875 mm; KJ Magnetics, Jamison, PA) stacked to achieve varying lengths (9.5 mm–16.9 mm) (Fig. 2A). Attachment of magnets to stents was achieved by using medical-grade adhesive (Loctite 4541 PrismGel, Henkel Corp., Dusseldorf, Germany). The external hand-held magnet was a 3.8-pound (1.78-kg) neodymium-iron-boron disk magnet (cylindrical N-42 grade with axial magnetization; KJ Magnetics, Jamison, PA) (Fig. 2B). Important design parameters for the magnetic retrieval system included the following: (1) magnetically tipped stents that could be delivered into the pancreaticobiliary tree via a conventional duodenoscope (ie, with a 4.2-mm instrument channel), (2) an arbitrarily chosen 5-pound maximum for an external hand-held magnet, and (3) a target capture distance of 10 cm to simulate a nonobese body habitus. A more thorough description of rational device design, computer modeling, and bench-top ex-vivo experiments to measure forces and validate design has been previously detailed6 and is also summarized in the appendix (available online at www.giejournal.org).
In-vivo testing in acute porcine laboratory studies Animal laboratory studies were conducted at Concord Biomedical Sciences and Emerging Technologies (Lexington, MA) after approval from the Institutional Animal Care and Use Committee. In-vivo porcine studies were performed in 5 Yorkshire pigs weighing approximately 35 to 40 kg. After induction of anesthesia by using an intramuscular injection of Telazol 4.4 mg/kg, xylazine 1.1 mg/kg, and atropine 0.05 mg/kg (all medications from Fort Dodge Animal Health, Fort Dodge, IA), the pigs were intubated, and anesthesia was maintained with isoflurane 1% to 4%. Magnetic stents were deployed into the common bile duct by using a conventional duodenoscope (Olympus TJF-160F; Olympus America Inc., Center Valley, PA). Biliary cannulation was achieved by using a soft-tipped 0.035 Jagwire (Boston Scientific Corporation, Natick, MA) through a ProForma tapered-tip catheter (ConMed Corporation, Largo, FL). Biliary cannulation was chosen instead of pancreatic duct cannulation because the porcine pancreatic duct in juvenile pigs is very small and has a separate, more distal Volume 75, No. 4 : 2012 GASTROINTESTINAL ENDOSCOPY 889
Magnetic pancreaticobiliary stents and retrieval system
Ryou et al
TABLE 1. Ex-vivo capture distances with varying magnetic ring lengths
Length (cm)
Distance (cm)
Force (mN)
Internal flange (ⴙ/-)
9.5
2.5 ⫾ 0.1
680
⫹
9.5
9.0 ⫾ 0.3
59.2
⫺
9.5
9.0 ⫾ 0.6
59.2
⫺
12.7
4.0 ⫾ 0.1
361.1
⫹
12.7
10.0 ⫾ 0.4
44.1
⫺
12.7
9.0 ⫾ 0.4
59.2
⫺
15.9
4.0 ⫹ 0.1
361.1
⫹
15.9
10.0 ⫾ 0.3
44.1
⫺
Figure 2. A, Magnetic stent. B, External hand-held magnet.
orifice.7 Two pigs received a 5F ⫻ 5 cm stent, and 3 pigs received a 5F ⫻ 3 cm stent. Each stent was tipped with a 9.5⫺mm long cylindrical magnetic extension (weight, 0.42 g), and the internal flanges were removed. With fluoroscopic and endoscopic guidance, the external handheld magnet was applied to the anterior abdominal surface for stent removal. The time of stent insertion (time from stent insertion into instrument channel to removal of Jagwire from bile duct and stent) and the time of stent removal (time from application of external magnet to endoscopic visualization of proximal aspect of stent within duodenal lumen) were recorded. Fluoroscopic and endoscopic footage were recorded. Animals were euthanized at the conclusion of the procedure.
RESULTS Ex-vivo testing Table 1 shows the ex-vivo capture distances of various magnetic ring lengths interacting with the 3.8-pound external magnet (Table 1). Both 5F ⫻ 3 cm and 5F ⫻ 5 cm stents were used in these experiments (Fig. 3). Surprisingly, a magnetic ring length of 9.5 mm without internal flanges achieved a capture distance of 9.0 cm, which was very close to the target capture distance of 10 cm. Increases in length to 12.7 mm and 15.9 mm yielded minimal increases in capture distance. Therefore, it was decided to use 5Fr stents (both 3 cm and 5 cm) without internal flanges and with a magnetic ring length of 9.5 mm for subsequent acute animal laboratory studies. 890 GASTROINTESTINAL ENDOSCOPY Volume 75, No. 4 : 2012
Figure 3. Ex-vivo tests to measure capture distance (15.9-mm magnet assembly shown [longest tested]).
Acute animal laboratory studies As shown in the video (available online at www. giejournal.org), upon successful location of the porcine ampulla, the mean time for stent insertion into the bile duct was 3.2 minutes (range, 2.75 minutes–5.8 minutes). The mechanics of stent deployment were easy and not technically different from the mechanics of deploying a nonmodified stent. The stents were inserted until the proximal portion of the magnetic rings contacted the ampulla (Fig. 4). The mean removal time was 33 seconds (range, 25 seconds– 44 seconds). Fluoroscopy revealed an approximate capture distance of 6 cm in all pigs (within the specifications of the system design). Capture was immediate upon application of the external magnet. Maneuvering of the external magnet was similar for all animals. The stents were completely extracted into the duodenum in all animals as corroborated by endoscopic visualization. There was no evidence of mucosal or tissue trauma during stent removal. www.giejournal.org
Ryou et al
Magnetic pancreaticobiliary stents and retrieval system
Figure 4. A-C, In-vivo stent deployment. D-F, Removal.
DISCUSSION This proof-of-concept animal study successfully demonstrated a magnetic retrieval system for pancreaticobiliary stents. Magnetically tipped stents were deployed in a conventional manner through a duodenoscope. The stents were then fully extracted into the duodenum by using a hand-held external magnet. The magnetic retrieval system targets applications of temporary stenting that do not necessarily require repeat ERCP. Examples include pancreatic duct stenting for prophylaxis of post-ERCP pancreatitis in certain high-risk groups (sphincterotomy for sphincter of Oddi dysfunction, ampullectomy, pancreatic sphincterotomy, difficult cannulation)4 and for postcholecystectomy biliary leak.3 The main study limitation was that it was an acute animal study. To minimize the effects of anatomic differences, our modified pancreatic stents were inserted into the porcine common bile duct (the pancreatic duct in juvenile pigs is extremely small and has a separate distal orifice from the biliary orifice7). Consequently, maneuvering of the external magnet for stent extraction will likely be different for stents deployed in the pancreatic duct and different for human anatomy. However, anatomic differences likely do not influence feasibility, as evidenced by the successful use of external magnets to guide magnetically tipped enteral feeding tubes through the pylorus and duodenal sweep in human studies.8 Other parameters also require further study. For example, how will stent retrieval be affected by long-term retention or modifications to stent size? Internal flanges in particular will be an important feature to include in subsequent investigations to prevent spontaneous dislodgement, especially for applications that require placement for many days to weeks. www.giejournal.org
Extrapolation of our results to larger stents is unclear. A 5Fr stent was used in this study because it appeared to be the most suitable size for our particular animal model. Successful demonstration in larger-diameter stents may require larger pigs or a different animal model altogether. Although magnets have been shown to be safe for use in GI endoscopy for foreign body retrieval,9 treatment of bile duct stenosis,10 and retraction during endoscopic submucosal dissection,11 the magnetic component of the retrieval system may lead to certain restrictions in implementation. For example, this retrieval system may not be applicable to patients with pacemakers and automatic implantable cardioverter-defibrillators, and magnetic resonance imaging may not be performed while the stents are in place. In addition, the magnetic retrieval system may be difficult to use in morbidly obese patients, given the excessive capture distance. However, in bench-top models, the current prototype achieved a capture distance of 10 cm in spite of an arbitrarily chosen limitation of 5 pounds for the external magnet. It would certainly be possible to scale up the power of the retrieval system with stronger magnets or an electromagnet. Finally, there is the theoretic risk of accidentally pushing in the stent if the external magnet were accidentally oriented to cause magnet–magnet repulsion. However, this risk can be mitigated by retaining external flanges. Future work will focus on several goals: increase in magnetic strength, determination of prescribed motions and maneuvers of the external magnet for reproducible stent removal in human anatomy, development of a realtime feedback for capture without the use of fluoroscopy, and development of a through-the-scope retrieval catheter for migrated magnetic stents. Volume 75, No. 4 : 2012 GASTROINTESTINAL ENDOSCOPY 891
Magnetic pancreaticobiliary stents and retrieval system
In conclusion, a magnetic stent retrieval system is technically feasible in the porcine model. This may potentially obviate the need for a second endoscopy and represent a major gain both for patients and in health care savings. REFERENCES 1. Mergener K, Strobel JC, Suhocki P, et al. The role of ERCP in diagnosis and management of accessory bile duct leaks after cholecystectomy. Gastrointest Endosc 1999;50:527-31. 2. Sandha GS, Bourke MJ, Haber GB, et al. Endoscopic therapy for bile leak based on a new classification: results in 207 patients. Gastrointest Endosc 2004;60:567-74. 3. Coelho-Prabhu N, Baron TH. Assessment of need for repeat ERCP during biliary stent removal after clinical resolution of post-cholecystectomy bile leak. Am J Gastroenterol 2010;105:100-5. 4. Freeman ML. Pancreatic stents for prevention of post-endoscopic retrograde cholangiopancreatography pancreatitis. Clin Gastroenterol Hepatol 2007;5:1354-65.
Ryou et al
5. Rashdan A, Fogel EL, McHenry L, et al. Improved stent characteristics for prophylaxis of post-ERCP pancreatitis. Clin Gastroenterol Hepatol 2004; 2:322-9. 6. Cantillon-Murphy P, Ryou M, Shaikh SN, et al. A magnetic retrieval system for stents in the pancreaticobiliary tree. IEEE Trans Biomed Eng 2010;57:2018-25. 7. Mullen Y, Taura Y, Nagata, et al. Swine as a model for pancreatic beta-cell transplantation. In: Swindle MM, ed. Swine as Models in Biomedical Research. Ames, IA: Iowa State University Press; 1992:1634. 8. Gabriel SA, Ackermann RJ. Placement of nasoenteral feeding tubes using external magnetic guidance. J Parenter Enteral Nutr 2004;28:119-22. 9. Paulson E, Jaffe R. Metallic foreign bodies in the stomach: fluoroscopic removal with a magnetic orogastric tube. Radiology 1990;174:191-4. 10. Muraoka N, Uematsu H, Yamanouchi E, et al. Yamanouchi magnetic compression anastomosis for bilioenteric anastomotic stricture after living-donor liver transplantation. J Vasc Interv Radiol 2005;16: 1263-7. 11. Gotoda T, Oda I, Tamakawa K, et al. Prospective clinical trial of magneticanchor guided endoscopic submucosal dissection for large early gastric cancer (with videos). Gastrointest Endosc 2009;69:10-15.
GIE on Facebook GIE now has a Facebook page. Fans will receive news, updates, and links to author interviews, podcasts, articles, and tables of contents. Search on Facebook for “GIE: Gastrointestinal Endoscopy” and become a fan.
892 GASTROINTESTINAL ENDOSCOPY Volume 75, No. 4 : 2012
www.giejournal.org
Ryou et al
APPENDIX System requirements Design of the overall magnetic system was guided by the following requirements and restrictions: (1) The modified stent would have to be deployed in conventional fashion through a duodenoscope. (2) The modified stent would have to fit down a 4.2 mm instrument channel. (3) As an extension of the stent, the attached magnets should also feature a lumen to provide continued drainage. (4) The external hand-held magnet would have an arbitrarily chosen 4-pound weight restriction in order to feel comfortable for most operators. (5) The overall system should provide sufficient capture force at 10 cm (an approximate estimate of separation distance in a non-obese human body habitus) to dislodge the stent completely from the duct.
Design optimization The dimensions of the hand-held magnet were optimized for a fixed weight and a cylindrical shape (to minimize edges and corners) via numerical simulation using MATLAB (Mathworks, Inc., Natick, MA). For the magnet
www.giejournal.org
Magnetic pancreaticobiliary stents and retrieval system
attachments, axially-magnetized ring magnets were chosen because of Requirements #1 and #3 listed previously. A ring of outer diameter 3.175 mm and inner diameter 1.5875 mm was chosen as this was this was the largest, off-the-shelf commercially available dimensions that could fit down an instrument channel diameter of 4.2 mm. Attractive force between an optimized cylindrical hand-held magnet and various ring magnet assembly lengths (6.35 mm, 9.525 mm, 12.7 mm, 15.9 mm) were modeled using MATLAB.
Ex vivo validation tests Measurements of attractive force between various magnetic ring lengths and the hand-held magnet were corroborated experimentally using an aluminum bending-beam force gauge as beam deflection correlated to a known load.6 Ring lengths of 9.525 mm, 12.7 mm, and 15.9 mm were attached to Geenen 5F ⫻ 3 cm and 5F ⫻ 5 cm stents. In ex vivo tissue experiments, these modified stents were inserted into the porcine bile duct. The external hand-held magnet was incrementally advanced towards the stent until it completely dislodged. Capture distance was subsequently recorded.
Volume 75, No. 4 : 2012
GASTROINTESTINAL ENDOSCOPY
892.e1