Clinical practice of basin-shaped hepaticojejunostomy following hilar resection of stage III/IV hilar cholangiocarcinoma

Hilar cholangiocarcinoma (HCCA) is the most common malignant tumor affecting the extrahepatic bile duct, including the total hepatic ducts and the right and left hepatic ducts, accounting for two-thirds of biliary tract tumors [1].

HCCA was first described by Klatskin in 1965, based on its distinctive clinical and pathological features [2]. The Bismuth-Corlette classification stratifies patients with HCCA into 4 types based on the extent of biliary involvement by the tumor [3]. Bismuth type IV tumors show extension into the secondary biliary radicals on both sides of the liver.

Resection of the involved intra- and extrahepatic bile ducts, as well as the associated hepatic lobe and caudate lobe, is the standard of care for patients with resectable tumors [4]. Over the past two decades, major hepatectomy or en-bloc resection with (extended) hemihepatectomy including the caudate lobe has been established as the mainstay of stage III and part of stage IV HCCA treatment if the range of invasion is mainly on one side of the primary bile duct. However, for the remaining stage IV HCCA patients with extended bilateral primary bile duct invasion, (extended) hemihepatectomy cannot achieve the curative criteria, while segment IV/extended IV is the best option to resect the tumor-invaded bile duct completely with R0 margins [5‐7]. Although resection segment 4(or plus resection segment 1) could achieve a curative resection and abundant liver preservation, the very practical problem was that resection would leave multiple biliary stumps (usually more than 4, including secondary or tertiary biliary stumps).

In the conventional method, the multiple stumps are anastomosed one by one, or alternatively, the major branches are anastomosed, while the minor ones are ligated. These methods usually result in unsatisfactory anastomoses, leading to an elevated risk of complications and extended surgical time.

Non-synchronized with the improvement of resectability rates [8], the results remain unsatisfactory when followed by conventional methods for bilioenteric anastomosis, with high morbidity rates, ranging from 40 to 68%, and in-hospital mortality rates higher than those reported after liver resection for all other diseases [9]. The main cause of such high postoperative morbidity and mortality rates is septic complications [10], primarily due to reflux or leakage through the bilioenteric anastomosis following removal of the tumor. Since the main cause of postoperative complications is unsatisfactory bilioenteric anastomosis, the risk is even higher in patients with Bismuth types IV HCCA. Due to their extensive invasion range, Bismuth type IV HCCA patients always require resection segment 4 + 1 (including the caudate lobe) resection, which leaves multiple bile duct stumps to address. Thus, an improved method of biliary-enteric anastomosis for this situation is a pressing need.

Basin-shaped biliary-enteric anastomosis was first performed in 1987 to address iatrogenic bile duct injury and multiple bile duct stones followed by stenosis of the bile duct. The key procedure includes the following: 1) excision of the injured or constrictive bile duct; 2) incision and modification of bile stumps into one or several common openings to shape a basin with the jejunum opening for anastomosis; and 3) connection of the jejunum to the modified “basin shaped” opening and adjacent connective tissue of the Glisson pedicle by continuous sutures. The original intent was to take advantage of the wide-shaped anastomosis, avoiding multiple suturing of the iatrogenically caused (or constrictive) bile duct stumps one by one, which is almost impossible in many situations. Due to the improved techniques and science of HCCA surgery, more patients with stage IV HCCA have the opportunity for curative resection, resulting in more situations with multiple and adjacent stumps. Based on similar conditions, we practiced and improved the basin-shaped anastomosis method in HCCA resection. The biliary stumps were directly inserted into the basin-shaped anastomosis without modification.

In this study, we performed basin-shaped biliary-enteric anastomosis after resection segment 4(or plus resection segment 1) and high hilar resection for Bismuth type III/IV HCCA, and we observed and analyzed the safety and effects for these patients.

The main purpose of this study was to introduce and practice basin-shaped hepaticojejunostomy, followed by segment IV/extended IV and high hilar resection of stage IV hilar cholangiocarcinoma, in which the tumor invaded the bilateral secondary (tertiary) bile ducts. The outcomes data revealed that the basin-shaped hepaticojejunostomy was safe and effective for patients with stage III/IV hilar cholangiocarcinoma after resection segment 4(or plus resection segment 1) and high hilar resection.

Hilar cholangiocarcinoma is the most common malignant tumor affecting the extrahepatic bile duct [15], constituting 46–97% of all bile duct cancers [16, 17]. Due to the relatively slow growth and concealed location of the tumor, most patients do not present with any symptoms until the late stage, with unresectable tumors. Therefore, most patients have a poor prognosis, except for very few patients with indolent and slow-growing hilar cholangiocarcinoma, who can attain long-term survival [18]. While palliative procedures, such as photodynamic therapy (PDT) and transarterial chemoembolization, have been reported to prolong survival [19, 20], the only potentially curative therapy for hilar cholangiocarcinoma is R0 resection [21]. Development of a more aggressive approach to HCCA, involving radical biliary resection with segmental hepatic resection inclusive of the caudate lobe, [22‐25] resulted in dramatic improvements in survival and decreases in local tumor recurrence. However, the postoperative morbidity and complication rates remained high [26, 27]. This contradiction was caused by the asynchrony of improved resection techniques with relatively underdeveloped reconstruction techniques.

Because the main goal of surgical resection is to achieve adequate resection margins,[24] stage III/IV HCCA usually requires left- or right-sided hepatectomy, and it always requires caudate lobe resection [28]. Such resections can provide a curative prognosis, but they are associated with increased difficulty of biliary reconstruction because the resection leaves more biliary stumps. The management of these remaining biliary stumps can greatly affect surgical outcomes and patient recovery.

In contrast, the common postoperative complications of HCCA include intra-abdominal hemorrhage, hemorrhage of the digestive tract, bile duct leaks and biliary infection. Most of these complications are related to the hepato-enteric anastomosis [29].

Therefore, the exploration of new methods to improve the postoperative complication rate is worthwhile. Thus, we assessed whether an improved hepaticojejunostomy might simplify the procedure, shorten the operation time and decrease the postoperative complication rate.

With the conventional method, biliary continuity is restored with a Roux-en-Y of jejunum loop, and the hepato-enteric anastomosis is performed in an end-to-side manner one by one [28]. This method is effective for 1–3 stumps; for more than 4 stumps, this one-by-one anastomosis approach is time consuming, laborious and prone to errors and omissions. Although, in most hilar resection operations, the biliary stumps numbered fewer than 4, for patients with stage III/IV HCCA who underwent resection segment 4(or plus resection segment 1), the biliary stumps always numbered more than 4, and the conventional method was never helpful.

In response to this situation, we practiced the basin-shaped anastomosis method, which sutures the jejunum and the sheath of the Glisson pedicle around the biliary stumps to shape a basin, after which the biliary stumps are inserted into the basin-shaped anastomosis, avoiding the one-by-one procedure.

The Glisson capsule extends into the liver as tissue sheaths around the hepatic ducts, hepatic arteries, and portal tributaries [30]. Primary, secondary and tertiary biliary ducts are surrounded by the Glisson’s capsule completely and are connected to each other in common tunnels. Thus, suturing of the peripheral Glisson’s capsule surrounding the biliary duct will include all of the passing biliary duct in the common tunnel and will shape the “basin” for our biliary-enteric anastomosis technique. The methods of suturing the ligament surrouding the biliary ducts for hepaticojejunostomy have been introduced and practiced in different studies, and the safety and efficacy have been demonstrated and verified [31‐33]. Similar to but different from these methods, we made some improvements with the aim of reducing the postoperative complication rate: 1. Each biliary stump was inserted a polyvinyl chloride tube as a stent. The secondary and tertiary biliary stumps were always tiny and easy to constrict by inflammatory edema, and the inserted stent could help to keep the tunnels open and unobstructed. The stent will come loose and fall off after regression of the local inflammation and will finally be excreted through the intestine. The lost stent is not new technique, but was rare seen in research about hepatic-portal to jejunum anastomosis. We tried to apply this technique to our hepaticojejunostomy to observe and discuss if it could be an improvement for basin-shaped hepaticojejunostomy. In our study, 17 cases were applied the stent while 3 cases had 3 biliary radicals, 12 cases had 4 biliary radicals and 2 cases had 5 biliary radicals. The decision was made according to the lumen size of biliary radicals, a stent was inserted if the lumen of biliary radical was not too tight nor too loose. According to our standard, the use of stent were mostly occurred in cases with less number but larger lumen size of biliary radicals, the internal control group had critically selective bias to the stent group. As the results had approved, all complications of bile leakage and biliary infection were found in the no-stent group, as well as hemorrhage. 2. Different from previous methods that suture the peripheral Glisson’s capsule to surround the biliary stumps, our method sutures the biliary wall together with the peripheral Glisson’s capsule to form the wall of the anastomosis, except for some parts for which there were no biliary stumps along the suture path. This method could fix the biliary stumps to prevent motion, and it works together with the inserted stent to keep the tunnels open. Compared with previous studies [5, 6] associated with stage III/IV HCCA, our improved method demonstrated advantages of a lower complication rate and shorter operation time. The reasons could be the following: 1. When multiple bile duct stumps were addressed with the basin-shaped method, all of the stumps were inserted into one common tunnel; the single anastomosis not only saved time but also allowed for sufficient space to ensure exact and complete suturing. 2. Furthermore, the bile ducts were anastomosed in their original locations, this avoiding the retraction and pulling that can increase the suture tension and the risk of laceration. With the conventional method of suturing the anastomosis one by one, the completed anastomosis will affect or even occupy the space for the next anastomosis, rendering the operation more difficult and uncertain. 3. Moreover, multiple anastomoses cannot ensure avoidance of  tension for every anastomosis; later sutures can retract previous ones to increase the tension, which can cause tearing of the anastomosis.

The limitations of this study were defined by its small sample size and lack of randomized controls. Further prospective studies, with large sample sizes and randomized controls, are needed to demonstrate the effectiveness. Furthermore, with this surgical technique for tumor disease, the risk of tumor recurrence remains unknown for its short follow-up observation period. Prolonged follow-up durations for the observation of tumor recurrence are needed to demonstrate oncologic safety.

Basin-shaped biliary-enteric anastomosis for patients with stage III/IV HCCA following IV/extended IV segment resection was safe and effective. Advantages in postoperative complications and operation time could be present with this method.