BBSs are a relatively common complication of liver transplantation. Extensive stripping of the blood vessels during the operation can cause ischemic injury [4] and hypertrophic change [3] promotes biliobiliary anastomotic strictures [5]. Although the optimal strategy for BBSs has not been determined, nonsurgical approaches are more popular. The insertion of MPSs with or without balloon dilatation is currently the treatment of choice [36]. The success rate of MPSs is approximately 70%–91% in patients with BBSs after deceased-donor liver transplantation and 60%–100% in those with BBSs after LDLT [36-39]. FCSEMS has a similar stricture resolution rate to MPS, but the number of ERCP sessions is fewer [1,40]. However, MPS and FCSEMS insertion cannot be applied when a guidewire cannot pass though the stricture.

The overall clinical success rate of MCA for biliobiliary strictures was reported to be 87.5%, and the recurrence rate was 7.1% (Table 1). The clinical success rates of MCA differ according to the etiology of the stricture and the treatment method in biliobiliary strictures [41-45]. Although surgery-related strictures and strictures due to stones have high resolution rates with endoscopic treatment, idiopathic and chronic pancreatitis-related strictures respond poorly to endoscopic management [46]. The clinical success rate of endoscopic treatment for biliobiliary strictures is 90% in postoperative strictures but only 65% in strictures due to chronic pancreatitis [46]. Moreover, the resolution rates of percutaneous treatment of biliobiliary strictures were reported to be from 61.4% to 90.9%.47-50 Although advances in conventional techniques have increased the clinical success rates, these methods are unsuccessful for severe stenotic strictures through which passage of a guidewire is impossible. Therefore, these methods cannot be used to resolve all biliobiliary strictures. Moreover, with conventional methods, the clinical resolution rates are lower than the technical success rates, because of stricture recurrence. Repeated pneumatic dilation can cause traumatic tissue injury and promote a fibrotic reaction resulting in the recurrence of strictures [51].

The early complication rates after MCA are not reported in most studies (Table 1). The main adverse event is mild cholangitis, which can be resolved with conservative treatment [22,23]. The only reported adverse event occurring from magnet approximation to the removal of the indwelling catheter is a slight fever [34]. Doppler ultrasound is performed to evaluate the vessel because the risk of vessel rupture is a concern during the early stages of MCA [17,19]. However, no blood vessel rupture has been reported to date. As the magnets gradually get closer to each other during a long period after magnet approximation, the intervening vessels are not compressed or ruptured. The mean duration of full magnet approximation is 53.3 days in biliobiliary strictures and 7–40 days in bilioenteric strictures [23].

No late adverse events and MCA procedure-related mortality have been reported during the follow-up. Adverse events related to magnets have not been reported because the magnets are aseptic devices that do not induce an inflammatory or immune reaction in the bile duct. Moreover, no adverse events directly related to other equipment used in MCA have been reported because the MCA procedure is performed using conventional ERCP techniques and a PTBD tract. Therefore, the MCA procedure seems to be safe in patients undergoing liver transplantation or with an immunocompromised status.

The length of the stricture is an important consideration for a successful MCA. The strictures are usually longer and the shape is more tortuous in LDLT recipients than in OLT recipients. Moreover, the distance between the approximated magnets is shorter for bilioenteric anastomoses (2–7 mm) than for biliobiliary anastomoses (2–15 mm). The technical factors influencing the effectiveness of MCA are stricture length, duration from LDLT to stricture occurrence, shape of the bile duct, and magnetic power of the magnets. The maximal stricture length for a successful MCA has not been confirmed and needs to be evaluated.

MCA creates a new fistula tract through tissue necrosis instead of simple dilation of fibrotic tissue in BBSs; thus, the recurrence rate seems to be lower than that with other conventional treatments. The possibility of elastic recoiling in the fistula formed by MCA may be lower. Moreover, a PTBD catheter is not needed and the risk of infection is low after recanalization with MCA. However, the long-term efficacy of MCA should be clarified by further large-scale and long-term follow-up studies.

As MCA is applied in biliary strictures that could not be resolved using conventional methods, it is difficult to compare its advantages and disadvantages with those of conventional methods, and it cannot replace conventional methods. Currently, MCA is not widely used because the magnets and equipment used in MCA are not commercially available, and cases that are indicated for MCA are relatively rare.

BBSs are more frequent after LDLT than after OLT [52-56]. Overall, the incidence of BBSs is 25%–32% after LDLT [57-61] and 5%–15% after OLT [39,60,62]. The higher incidence of post-LDLT strictures is due to technical aspects of the operation because ductal anastomosis between the donor and the recipient during LDLT is complicated. Moreover, the duct of the donor liver has many variants, including multiple bile ducts, a poor blood supplement, and a relatively short stump for the anastomosis [9,60,61,63].

The MCA process in post-OLT strictures is the same as in post-LDLT strictures (Fig. 3). The level of post-OLT strictures is more distal in the CBD than that of post-LDLT strictures. Post-OLT strictures are mid-level BBSs but post-LDLT strictures are high-level BBSs. In addition, intrahepatic ducts (IHDs) are more dilated but less angulated and tortuous in post-OLT strictures than in post-LDLT strictures. Therefore, MCA is more feasible and the success rate is high in post-OLT strictures (Fig. 3).

Bile duct injury during laparoscopic cholecystectomy has occurred with an incidence of 0.3%–0.6% [64]. The type of biliary injuries can be categorized according to the Bismuth [65], Strasberg [66], and Stewart-Way classifications [67]. The Stewart-Way classification categorizes biliary injuries into four types according to the mechanism and anatomy [67]. Stewart-Way class I injuries are immediately resolved during the operation. Stewart-Way class II injuries with stricture are managed with insertion of MPSs and FCSEMS. Stewart-Way class III and IV injuries can be treated using multidisciplinary methods [68]. A combined endoscopic-radiologic rendezvous technique is needed for complete transection of the main bile duct (Stewart-Way class III) and transection of a sectoral bile duct (Stewart-Way class IV). This technique can avoid surgical reintervention, and can reduce surgical morbidity and mortality. These endoscopic-radiologic rendezvous techniques for treating bile duct stenosis or complete obstruction are feasible only if the guidewire can pass through the stricture. MCA can be an alternative method for patients with Stewart-Way class III and IV ductal injuries that cannot be resolved using rendezvous techniques (Fig. 4). The procedure for MCA is similar to the procedures applied in patients with post-LDLT strictures.

Bile duct injury and stricture may also occur after other hepatobiliary operations including hepatectomy. Complex dissection, poorly defined anatomy, or management of intraoperative bleeding by clipping can cause early-phase stricture [69]. Late-phase stricture may result from ischemic injury in the bile ducts, and a stricture may not develop clinically until several years after the initial injury. Endoscopic treatment with MPS or FCSEMS insertion is a more preferred method for strictures after biliary surgery than percutaneous treatment, but multidisciplinary approaches involving endoscopists, radiologists, and surgeons are usually required to resolve these complicated BBSs. Usually, an endoscopic or percutaneous approach can be useful if the guidewire can pass through the BBS. However, MCA can be applied when these conventional methods cannot resolve the BBS. The technique for MCA in this situation is also similar to that used for post-LDLT strictures.

BBSs are caused by postoperative complications, especially after a Roux-en-Y reconstruction, which is the most common method of anastomosis in biliary surgery (Table 2). The mean distance between the approximated magnets was reported to be 4 mm (range, 2–7 mm), and the duration from magnet approximation to magnet removal was 7–40 days. Complete resolution of the obstruction was accomplished in 41 of 42 patients (97.6%), and no severe complications were observed. The recurrence rate was low (follow-up period: mean, 40 months; range, 2–53 months). One patient had restenosis at 6 days after the removal of the indwelling catheter, which was treated successfully and without difficulty with balloon dilation [34].

The difference in the MCA method between bilioenteric anastomosis strictures and biliobilary strictures is the route for magnet delivery [24,34,70]. The delivery route for magnets in biliobiliary strictures is mainly the percutaneous-peroral tract. However, the delivery routes for magnets in bilioenteric strictures are various and include the percutaneous-peroral tract, surgically formed percutaneous-enteric tract, or percutaneous- percutaneous tract. The method of magnet delivery to the percutaneous tract is similar to that described previously. With the peroral approach, a forward-viewing endoscope is used instead of an ERCP scope [23]. A peroral approach with the forward-viewing endoscope is sometimes difficult in patients with long and redundant afferent loops. Single-balloon enteroscopy is feasible in these situations [71]. If the endoscope cannot approach the stricture site, magnets can be delivered through a surgically formed percutaneous-enteric fistula [23]. When the left and right IHDs are anastomosed separately to the jejunum and the stricture occurs in the right IHD, two PTCS scopes can be applied for MCA [72]. One magnet is moved through the right IHD using one PTCS scope, and the other through the left IHD to approximate via the left IHD tract using another PTCS scope. The cases described above provide the possible methods of magnet delivery that can be applied in various situations.