Surgeries involving the creation of bilioenteric anastomoses are performed in patients for a variety of reasons including primary repair of benign biliary strictures, tumor resection, and liver transplantation. Anastomotic strictures are generally thought to form due to fibrosis with resultant retraction, as opposed to non-anastomotic benign biliary strictures which often occur due to iatrogenic causes (most commonly during cholecystectomy), chronic biliary inflammation, or ischemia during hepatic transplant.¹

Benign stricturing of the bilioenteric anastomosis is significant problem, occurring in 8%–40% of cases of primary biliary repair, with biliary complications occurring in 11.5%–34% of liver transplants.^2,3 Postoperative strictures at the hepaticojejunostomy develop in 2.6% of patients who undergo pancreaticoduodenectomy at a high volume center.⁴ Repeat surgical interventions often result in poorer outcomes, with greater rates of morbidity and mortality correlating with the number of previous repairs.⁵ Endoscopic treatment of benign bilioenteric anastomotic strictures (BAS) is difficult in altered surgical enteric anatomy and these patients are often referred for percutaneous treatment.⁶ Unlike treatment of malignant strictures, these patients may have extended survival and are not good candidates for traditional self-expandable metallic stents^7,8 or prolonged percutaneous transhepatic drain placement.

Definitive percutaneous treatment of benign biliary strictures was first described by Molnar and Stockum⁹ in 1978, and again in a larger series by Mueller et al¹⁰ in 1986. Since this time there has been evolution of tools and technique. Current protocols for percutaneous treatment of BAS are varied in their technique and success rates. We aim to review some of the current techniques used for percutaneous treatment of bilioenteric strictures, examine short and long-term outcomes of the more common protocols, and comment on future directions of the procedure (Table 1).

The classic symptoms of cholestasis including jaundice and pruritus are generally the first evidence of the occurrence of a biliary stricture, anastomotic or otherwise. These symptoms often occur with the expected laboratory abnormalities including a rising bilirubin, alkaline phosphatase, and γ-glutamyltransferase (the cholestatic pattern), as well as mild increases in the transaminases.¹¹ In many cases the first sign of a biliary stricture is frank cholangitis, with presenting symptoms of fever, abdominal pain, jaundice, and rigors. Generally some form of medical imaging is then performed demonstrating intrahepatic biliary dilation with identification of a biliary stricture (Fig. 1).

Terminology can vary, though for the purposes of this review, each case in which balloon dilation is performed is referred to as a dilation procedure, of which several may occur in a single dilation session. Dilation sessions are concluded once the desired stricture diameter is attained. A single treatment consists of one or multiple dilation sessions separated by intervals of catheter stenting. The treatment is concluded once all catheters are removed from the patient. Should a stricture recur after conclusion of a treatment, it is considered a treatment failure.

Catheter-based upsizing

Percutaneous biliary interventions are among the most painful minimally invasive procedures that are routinely performed, necessitating expert sedation to maintain patient comfort. Most of the recent series conducting balloon dilations are able to effectively treat patients using moderate sedation with intravenous midazolam and fentanyl as for most interventional procedures. However, some interventionalists promote the use of general anesthesia for these procedures, citing not only the improved pain relief for the patient, but the ability to more aggressively dilate during the procedure.¹² Trambert et al¹³ suggest the use of 1% Li-docaine administered directly into the biliary tree before dilation as a useful adjunct to opiate analgesia. At our institution, general anesthesia is routinely used for biliary dilation and stenting procedures.

Access and staging

The first report by Molnar and Stockum⁹ described a three phase approach to percutaneous dilation of choledochoenterostomy strictures. Phase I involved percutaneous access and placement of an external drain, phase II involved placement of an internal/external drain, and phase III involved balloon dilation of the stricture (Fig. 2). The period of time between initial biliary drainage and the first dilation allows for an opportunity for the patient’s symptoms and lab values to normalize. This is especially important if the patient is actively cholangitic, as manipulation of a fresh intraparenchymal tract increases the exposure of hepatic venous blood to bile, with resultant increased risk of sepsis. Staging the procedure allows maturation of the percutaneous transhepatic tract. Most protocols for biliary dilation follow the general staged protocol initially described by Molnar and Stockum.⁹ The period of time between initial biliary drainage and balloon dilation varies, with some waiting a period of 1–4 days,^2,14 and others advocating a period of weeks.^2,14,15 Several of the large series evaluating percutaneous dilation over the last decade however routinely conduct their first balloon dilations on the initial day of access.^16–19 The study by Choo et al¹⁵ used the longest delay of 4–6 weeks between access and initial dilation, though did not have a significantly lower complication rate. However, treatment subjects in that study were exclusively post-transplant, confounding a direct comparison. Factors such as patient acuity, ease of percutaneous access, level of pre-stenotic biliary dilation, severity of stricture, ease of crossing the stricture with a guide wire, and various logistical and economic factors all play a role.

Balloon diameter

Due to the considerable fibrosis associated with anastomotic strictures, initial dilation is often difficult.¹⁰ Definitive treatment balloon diameter is determined based on the estimated normal duct diameter, persistence of the stricture, and concern for rupture. Because of stricture elasticity, some operators advocate for use of balloon diameters 25%–30% larger than the normal duct.^1,15 Most commonly 8–10 mm²⁰ are used, though balloon sizes up to 15 mm have been reported without an associated increase in complictions.¹ Short term recurrence of the stricture suggests that the balloon diameter used was too small or the balloon pressure was not great enough to fracture the scar.²¹ Of note, some interventionalists have opted to avoid balloon dilation altogether, instead utilizing sequential bougie dilation with good result.¹⁴

Balloon inflation pressure

As with dilation procedures elsewhere in the body, high-pressure low-compliance balloons are typically used for biliary stricture dilation. Most protocols do not specify a particular pressure for inflation. At our institution the balloon is expanded to its rated pressure in a gradual fashion, with use of a larger diameter balloon or possibly cutting balloon should the current balloon prove insufficient to dilate the stricture.

Balloon inflation duration

There is wide variation in the duration of balloon inflation during a given dilation procedure, ranging from 10 seconds to 12 hours.^15,21–23 In his initial description, Molnar and Stockum⁹ inflated their balloons for approximately 1–3 minutes, which remains the most commonly used inflation time today.^{9,12,16,19,20} There has been no documented difference in outcome between short and long dilation times, and several institutions do not specify a specific time on their protocols. Of the largest studies in the past decade, the shortest inflation time of 10–30 seconds was associated with a treatment failure rate of 17%,¹⁵ while the longest inflation time of 3 minutes was associated with a treatment failure rate of 23%.¹⁶ Overnight inflation for 8–12 hours with alternating 1 hour inflation and deflation periods has been described²³ but is cumbersome, costly, and is not widely used.

Stent catheter placement

Once the stenosis has been balloon dilated to an appropriate size, the tract is generally stented with an internal/external biliary catheter to allow healing of the fractured stricture while maintaining patency of the anastomosis and access to the biliary tree. Catheter sizes range from 8.5 F to 18 F (Fig. 3). Proponents of stenting with large caliber (16–18 F) catheters argue that the insertion of the large caliber catheter is more important to treatment than balloon dilation.²¹ This strategy is supported by the efficacy of endoscopic placement of multiple endoprostheses to dilate the stricture.²⁴ Detractors of this strategy have pointed out that dilating the entire biliary tract can cause undue morbidity and are not as well tolerated by patients.¹⁰ No direct comparison of catheter size on long-term treatment outcome has been performed. Of the most recent larger trials, Vos et al²⁰ used the smallest catheters (8.4 F) in their protocol, with a treatment failure rate of 27%. Choo et al¹⁵ used among the largest catheters (14–18 F) in their protocol, with a treatment failure rate of 17%¹⁵ suggesting that larger bore catheters may be superior.

Stent duration and repeat dilation

Once the stent catheter has been placed, it has to be maintained long enough that the ballooned stricture has time to heal around it. This is supported by the finding that anastomotic patency is better when surgically-placed stents are in place for longer than 1 year.⁵ Most of the major series over the past decade have used treatment durations of close to 1 year, with a few opting for slightly shorter stenting times of 3–6 months.^{2,14,16–18,25} Weber et al¹⁴ employed the longest treatment duration (mean, 19.9 months) with a treatment failure rate of 39%. Proponents of short-term stenting (up to one month) argue equivalent efficacy with less inconvenience to the patient.^10,19 The trial conducted by Cantwell et al¹⁹ used the shortest treatment duration (mean, 1.1 months), and observed the largest treatment failure rate of 44%. Some maintain the stenting catheter only between balloon dilation procedures, removing it as soon as maximal balloon dilation is achieved. Lee et al¹² used this approach (mean stenting duration of 10 days), with a treatment failure rate of only 7%; however, this was only in a small group of patients. Saad and colleagues^26,27 described a treatment protocol whereby if the stenosis has not been adequately treated after three dilation sessions (maximum stenting duration of 6 weeks), it is considered a treatment failure; this protocol saw a treatment failure rate of 64%.

Long term maintenance of an indwelling biliary stent catheter requires routine exchanges to ensure continued patency of the catheter, evaluate the stricture and perform repeat dilation when necessary. The most common interval for exchange and repeat cholangiography is 3 months, used in the majority of the most recent series. However some operators advocate more frequent exchanges, with some intervals on the order of a few days.¹⁹ Bonnel and Fingerhut² speculate that the greater success rate observed in their study (treatment failure rate of 15%) compared to other recent series may be due to their choice of a relatively shorter interval between exchanges/dilations of 1.5 months. It has also been speculated that longer time intervals between balloon dilations may result in more focal fibrosis at the stricture site due to interval healing.¹ At our institution, an exchange frequency of every 3 months is most commonly used.

Catheter-based upsizing

Some protocols use sequential stricture dilation with catheters, rather than balloon inflation, as the main mechanism of treatment. It is hypothesized that the longer-term catheter dilation allows healing of the stricture around the catheter, determining the overall end result.²¹ DePietro et al²⁵ conducted a recent trial using catheter upsizing every two weeks until an optimal size was reached (generally 16–18 F), at which point the catheter was maintained for 6–12 months. Balloon dilation was performed when an optimal size catheter could not be placed. They demonstrated a treatment failure rate of 33%. This protocol is similar to that used by Schumacher et al²⁸ and Weber et al,¹⁴ with treatment failure rates of 25.8% and 39%, respectively.

Minor complications related to percutaneous biliary interventions commonly include post-dilation fever and chills, tube displacement, pericatheter leakage, and gallstones.²⁵ Major complications include cholangitis/sepsis, bile leak, and hemorrhagic complications including pseudoaneurysm and hemobilia, some requiring catheter embolization. In the most recent series, rates of bile leak/biloma formation range from 0%–6.3%, while rates of hemorrhagic complications range from 0%–5% (with the exception of Choo et al¹⁵ which demonstrated a hemorrhagic complication rate of 16.6%). Direct comparison of complication rates is difficult as reporting and classification of complications is quite variable. To date, no specific procedural factor has been associated with a significant increase or decrease in complication rate.

Balloon dilation is usually performed until abolition of the waist is observed. How aggressively to upsize balloons is best judged on a case-by-case basis, but most agree that once the balloon waist has decreased to less than 20% of its initial size, dilation has been sufficient.¹⁵ The ease of passage of contrast is a useful adjunct in determining stricture patency and incomplete drainage after 5 minutes is an indication for continued dilation.²⁸ In the case of multiple dilation procedures in a single session, a dilated stricture may not show improvement on the immediate post-procedure cholangiogram due to edema within the duct immediately post-dilation.¹³ Significant improvement may not be seen until the pre-dilation cholangiogram during the subsequent procedure.

Determination of success of dilation treatment is often accompanied by a clinical ‘clamping’ trial. This consists of removal of catheter access across the stricture placement of a small caliber external biliary drain. This catheter is then capped and the patient monitored over a time course of a few weeks to a few months.^17,21 Signs of stricture recurrence including jaundice, abdominal pain, cholangitis, or laboratory abnormalities would be an indication for repeat cholangiogram through the maintained access. Should the patient remain asymptomatic for the treatment period, the catheter can be removed, either in clinic or after a final follow-up cholangiogram.

Once stenting has been maintained for longer than 1 year without desired clinical result, treatment is unlikely to be successful.²⁸ At this point, the patient is best either referred to surgery for attempt at repair, or maintained with permanent internal/external drainage. Endoscopic literature agrees that there is no benefit from stenting beyond 1 year.⁶ Some interventionalists are more persistent, suggesting that surgical intervention be pursued after drainage is attempted for 24 months.¹⁴ The protocol put forth by Saad²⁷ targets a < 30% residual stenosis to define success; if there is no improvement in any single dilation session or if < 30% residual stenosis cannot be achieved after an arbitrary maximum of three dilation sessions, the treatment is considered a failure. At our institution, we have found success with balloon dilation and catheter upsizing every 1–2 months, with no response after 4–5 dilations or 1 year considered a failure.

Before performing percutaneous balloon dilation, a full biochemical laboratory analysis including complete blood count, comprehensive metabolic profile including a liver panel, and a coagulation profile should be obtained. In accordance with the Society of Interventional Radiology guidelines, platelets should be maintained above 50,000 and the international normalized ratio below 1.5 if a new biliary tract is being created. Aspirin should be held for 5 days if a new biliary tract is being created; Clopidigrel should be held for 5 days regardless.²⁹ Biliary procedures are considered clean-contaminated, and as such antibiotic prophylaxis against predominantly gram negative organisms should be used. Common antibiotic choices include intravenous ceftriaxone, ampicilin/sulbactam, and clindamycin with an aminoglycoside.³⁰ Significant ascites should be drained prior to the procedure. Contraindications to balloon dilation include massive ascites, a shrunken cirrhotic liver which cannot be accessed safely (i.e., no available approach without traversing pleura or bowel), and uncorrectable coagulopathy.²²

Overnight observation is suggested when a new percutaneous access is being placed. Subsequent dilations and catheter exchanges can usually be performed on an outpatient basis. However, operators should remain vigilant for signs of sepsis or other complications which would necessitate a hospital admission.

There is no formal protocol for maintenance of the stenting catheter by the patient between dilation sessions. At our institution, daily flushing of the catheter with 10 mL of normal saline is used to ensure continued catheter patency. This requires adequate patient education and motivation. We also routinely prescribe prophylactic ursodiol to help solubilize cholesterol and lower the incidence of stone formation. The medication is prescribed at 300 mg by mouth twice daily, indefinitely.

BAS can develop several years after the creation of the anastomosis, with only 36% of recurrences developing during the first postoperative year.^31,32 Similarly, strictures can recur years after percutaneous treatment, making long-term patient follow up a priority. Recent studies by DePietro et al²⁵ and Janssen et al¹⁸ demonstrate the longest follow-up period, with recurrence rates rising from 5%–16% at 1 year to 28%–33% at 10 years.

Cutting balloons

Cutting balloons have become increasingly used in the context of benign biliary strictures, especially when strictures prove refractory to dilation with conventional angioplasty balloons. The embedded blades create small controlled incisions which score the stricture, allowing for a more effective subsequent dilation by the balloon via crack propagation. Case reports by Kakani et al³³ and Sheridan and Maclennan³⁴ have shown successful treatment using cutting balloons without the need for catheter stenting, and have shown continued patency for 10 months and 3 years respectively. Indeed, cutting balloons are often used at our own institution with good result (Fig. 4). A small series of 11 patients conducted by Atar et al³⁵ showed the utility of cutting balloon dilation followed by conventional balloon dilation in biliary and ureteral strictures, with technical success achieved in 82% without evidence of ductal rupture. They too did not employ long-term catheter stenting. Saad et al³⁶ conducted a trial involving 22 patients using a protocol featuring multiple sessions of cutting balloon dilation followed by conventional balloon dilation, separated by short-term small caliber catheter stenting. They demonstrated a technical success rate of 93% with this protocol, without evidence of biliary rupture. Minor hemobilia was encountered in 10%. This risk can be minimized by ensuring the diameter of the cutting balloon does not exceed the estimated normal diameter of the duct, reserving over-dilation for the conventional balloon.

Certain technical aspects must also be mentioned when using cutting balloons. First, because of the embedded blades, caution must be taken by the operator when handling the balloon catheter outside of the patient. Second, a slow inflation is recommended so as not to induce undue trauma to the duct. Third, a slow deflation is also recommended to allow for retraction of the blades into their grooves. Should the blades remain exposed as the catheter is retracted, severe complications may occur. Lastly, withdrawing the cutting balloon catheter through the sheath can result in splaying of the tip of the sheath. Thus when the sheath is eventually removed, careful inspection should be undertaken to ensure no fragments of the sheath are inadvertently left in the patient.^27,36

Self-expanding bare metal stents

Placement of bare metallic stents have little role in benign biliary strictures and are reserved for patients with < 2 year life expectancy or those expected to receive transplant relatively soon.^7,8

Covered self-expanding metal stents

Fully covered self-expanding metallic stents have been increasingly used to substitute for the placement of an internal/ external biliary catheter for anastomotic stenting after balloon dilation (Fig. 5). These stents allow for effective dilation of the anastomotic stricture without the issues brought about by large caliber internal/external catheters, including the necessity of dilating the entire tract and patient discomfort. The stent covering limits potential mucosal ingrowth and hyperplasia, factors which lead to occlusion and difficult retrieval in uncovered metallic stents.^37,38 These stents, however, do require retrieval via endoscopy, which as stated before is extremely technically challenging when having to navigate the altered post-surgical anatomy. We have found success at our institution using a rendezvous procedure for retrieval, where percutaneous access is maintained with a small caliber 8 F external biliary drain after initial stent placement. After 6 months of stenting, repeat cholangiogram and stent retrieval can be performed. Through the percutaneous access, a guidewire is advanced through the stent into the bowel where it is then grasped by the endoscope. The guidewire is subsequently retracted, pulling the endoscope into the biliary tract where the stent can then be retrieved (Fig. 6). In situations when this approach is unusable, techniques for percutaneous stent retrieval also exist.^39,40 While a dwell time of 6 months is used at our institution, no optimal dwell time has yet been established. Potential complications of retrievable stents include migration, side branch occlusion, and stent degradation with subsequent difficulty of removal.^40,41 Of note, the WallFlex covered stent is not yet approved in the United States for endoscopic retrieval.

Dissolvable stents

Dissolvable stents made from polydioxanone (Fig. 7) are becoming increasingly mentioned in the literature as an effective tool, allowing for long-term dilation without the need for removal.⁴² Hydrolytic degradation of the stent begins after approximately 4 weeks and results in complete absorption within 6 months. Case reports by Petrtýl et al⁴² and Jones et al⁴³ describe continued patency at 10-month and 2-year follow-up, respectively. The trial conducted by Mauri et al⁴⁴ demonstrated continued patency at a median follow up time of 16.5 months with complete dissolution of all implanted stents by 6 months. Biliary obstruction due to fragmentation of the stent is a potential complication resulting in transient cholangitis, though is likely less pronounced in the context of a bilioenteric anastomotic stricture.⁴⁰

Drug eluting stents

The benefits of paclitaxel-coated stents are well established in treating vascular stenosis.⁴⁵ Animal studies examining the use of paclitaxel-coated dissolvable stents in the context of BAS have demonstrated decreased granulation tissue, inhibited myofibroblast activity, and reduction in extracellular matrix over-deposition compared to controls, without significant difference in liver function.⁴⁶ Human trials have only been conducted in the context of malignant biliary strictures, demonstrating no significant difference in patency rates compared to non-drug eluting covered stents, though with a trend toward improved patency and survival for the drug eluting stents.⁴⁷ Additional work with other drugs and drug combinations including various antibiotics and chemotherapy agents is ongoing.⁴⁸

Percutaneous dilation remains a safe and effective approach to benign biliary strictures, especially in post-surgical strictures at the bilioenteric anastomosis. While numerous variations in procedural technique exist, the basic paradigm of percutaneous access followed by balloon dilation and catheter stenting is followed at most institutions. Differences in protocol are largely anecdotal, with no current evidence-based research demonstrating the superiority of any particular technique. However the use of large bore stent catheters have shown better success overall than small bore catheters. Further studies directly comparing different protocols will need to be carried out to determine which are most efficacious. However, multiple emerging techniques including the use of cutting balloons, retrievable stents, biodegradable stents, and drug-eluting stents have shown early promise, and should become valuable additions to the percutaneous dilation toolbox.