Bile duct injury during cholecystectomy

Summary

Laparoskopická cholecystektomie (LC) je léčbou volby u žlučových kamenů, avšak je spojená se zvýšeným rizikem poranění žlučových cest (bile duct injury − BDI). Pokud je BDI detekováno během LC, může se řešit ihned, pokud je dostupný hepatobiliární chirurg, ale nejjednodušší a nejbezpečnější postup pro všeobecného chirurga je umístění drénů do podjaterní krajiny a převedení akutní BDI na řízenou zevní biliární píštěl (external biliary fistula − EBF). Většina BDI jsou diagnostikované až v pooperačním období, kdy vzniká biliární leak. Léčbou je drenáž perkutánním katetrem a endoskopické zavedení stentu do žlučovodu; časná reparace se nedoporučuje. Reparace ve formě hepatikojejunostomie (HJ) by měla být provedena hepatobiliárním chirurgem s odstupem 4−6 týdnů poté, co se uzavře EBF. BDI jsou častou příčinou zdravotnicko-právních žalob a významně zatěžují náklady ve zdravotnictví. Většině BDI se dá předejít dodržováním principů bezpečné cholecystektomie.

Klíčová slova:
poranění žlučových cest − biliární leak – biliární píštěl − biliární striktura – cholecystektomie − hepatikojejunostomie

Introduction

Gall stone (GS) disease is prevalent all over the world – age-standardized necropsy prevalence of gall stones in former Czechoslovakia was 13.2% in men and 23.3% in women [1]. The median prevalence of GS ranges from 5.9 to 21.9% in Europe [2]. Cholecystectomy is indicated for symptomatic GS. Cholecystectomy rates ranged from 62 to 213 per 100,000 population in Europe in the 1990s [2]. Bile duct injury (BDI) is one of the complications of cholecystectomy. Laparoscopy has become the approach of choice for performing cholecystectomy. Risk of BDI was low (0.1−0.2%) during open cholecystectomy but is at least 2−3 times higher (0.4−0.6%) during laparoscopic cholecystectomy (LC). BDIs are under-reported in individual/hospital series; incidence is usually higher in national databases – Gall-Riks reported a 1.5% incidence of BDI in 51,041 cholecystectomies performed in Sweden during 2005−2010 [22]. The incidence of major BDI in 418,214 cholecystectomies (83% LC) in the National Health Service (NHS) UK between 2000 and 2009 was 0.4% [17]. It was thought that this higher risk of BDI during LC is a part of the learning curve and will decrease with time but that does not seem to have happened. The incidence of BDI in Sweden was 0.40%, 0.32% and 0.47% in 1989−1990, 1991−1995 and 1996−2001, respectively [24]. The higher incidence of BDI is an inherent risk of LC; also, BDI occurs in the hands of surgeons with adequate training, enough experience, as well as with a high volume of patients [4].

Etiology

Aberrant biliary ductal anatomy and difficult pathology e.g. acute cholecystitis and Mirizzi’s syndrome are frequently blamed (by the surgeon) as the cause of the BDI but an illusion of visual perception is the commonest cause of BDI during LC [25]. Most BDIs occur during a so-called uneventful cholecystectomy. The classical LC BDI is excision of a segment of the CBD which is misidentified as the cystic duct and dissected, looped, clipped and divided. Desperate attempts to control bleeding in the Calot’s triangle with clips or electrocautery are also a frequent cause of BDI. Dissection of the GB neck first and demonstration of the critical view of safety (21) is an important step to avoid a BDI during LC. Intra/per operative cholangiography (IOC/POC) does not reduce the risk of BDI but helps in earlier identification during operation.

“Injuries to the bile ducts are nearly always the results of misadventures during operation.” Grey Turner Lancet 1944;243:621−2

“Combination of inexperience and difficult GB or easy GB and overconfidence is a ripe setting for a BDI during LC.”

Intraoperative management

BDI during LC is usually missed during the operation and the surgeon considers the operation successfull – only 170 (23%) of 747 BDIs caused during 51,041 cholecystectomies in Sweden were detected intraoperatively [22]. Presence of bile during cholecystectomy is a warning – the surgeon should stop (no more dissection), look (from where the bile is coming) and only then proceed. If a BDI is recognized during the cholecystectomy itself, management depends not only on the type/ extent of the injury but also the expertise and experience of the injuring surgeon. A small (<3 mm) duct in the GB bed may safely be clipped but a large (>3 mm) duct in the hilum needs repair. If the injuring surgeon is a biliary surgeon with expertise and experience in performing reconstructive biliary surgery (i.e. hepatico-jejunostomy HJ) or if help of a biliary surgeon can be immediately obtained within the hospital or even from outside the hospital, as an outreach service [16], on-table (immediate) repair can be performed – this can be in the form of a suture closure (for a lateral injury) or end-to-end anastomosis or HJ (for complete transection). End-to-end anastomosis is an option only if the CBD is divided and there is no excision (loss of segment) – deReuver [7] reported a 91% stricture free result at a follow up of 7 years in 56 patients who underwent end-to-end anastomosis (supplemented by endoscopic dilation in case the anastomosis restrictured). If the injuring surgeon is a general/ laparoscopic surgeon with no expertise and experience of performing reconstructive biliary surgery, as is likely to be the case in most instances, repair should NOT be attempted. Two large bore drains should be introduced through the laparoscopic ports (subcostal and paraumbilical) and placed in the subhepatic fossa so as to prevent any bile collection and to convert the acute BDI into a controlled external biliary fistula (EBF) and the patient should be referred to a biliary center. Hepaticostomy (placing a tube in the divided proximal duct) is not recommended as attempts to manipulate a tube into an undilated duct may cause further injury. Clipping/ ligature of the divided proximal duct with intent to prevent bile leak does not work as the clipped duct sloughs off in a large number of cases [12].

“To injure is human, to refer divine”

Presentation

The majority of the BDIs are not even suspected during the LC and symptoms and signs of post-cholecystectomy bile leak are very non specific – a high index of suspicion is, therefore, required for early diagnosis of BDI. On the day following LC, if the patient is not well, has unstable vitals or an unsettled abdomen, bile leak should be suspected and investigated. Ultrasonography (US) will show a fluid collection which can be aspirated to confirm bile – any bile collection should be drained by percutaneous catheter placement; surgical intervention is rarely required for drainage of bile. An isotope hepato-biliary scan/magnetic resonance cholangiogram MRC will delineate biliary ductal continuity in the presence of which an endoscopic stent can be placed to reduce the ongoing bile leak. This again will convert the acute BDI into a controlled EBF. Non surgical (endoscopic and radiologic) intervention may be all that is required for some BDIs e.g. subvesical and cystic duct leaks and partial/ lateral BDI. Bile leak will stop and the BDI may heal without any sequelae. Uncontrolled bile leak can, however, result in complications such as biloma and bile peritonitis which may even cause death due to sepsis and multiple organ dysfunction syndrome (MODS).

“Bile duct injuries are unfortunately not rare, and often turn out to be tragedies”. Grey Turner Lancet 1944;243:621-622

Vascular (hepatic artery/ portal vein) injuries are frequently associated with BDI caused during LC. They were present in 84 (32%) of 261 BDIs [18]. A divided hepatic artery recognized during cholecystectomy may be repaired by a vascular surgeon but there is no role of repair of a vascular injury recognized postoperatively. A major vascular injury may lead to massive liver necrosis and acute liver failure. A partial arterial injury may result in the formation of a pseudoaneurysm which may rupture into the biliary tree to cause hemobilia; angioembolisation is the treatment of choice. In the long term, a vascular injury may result in atrophy-hypertrophy of the liver.

Timing of repair

Early (postoperative) repair is NOT recommended as the ducts are undilated and periductal tissues are inflamed due to bile leak. A nationwide review of 139 early (median 5 days) repairs done at 5 HPB centers in Denmark reported 4% mortality, 36% morbidity and 30% restricture rate at follow up of 102 months; 23/42 restrictures could be managed with percutaneous balloon dilation but 19 required re-HJ [19]. The only exception where an early repair can be performed is a ligated/clipped duct with no bile leak and no sepsis.

The acute BDI/ EBF may evolve into a benign biliary stricture (BBS) which should be repaired at least 4−6 weeks after the injury once the EBF has closed and the proximal bile ducts have dilated. In our experience, more than 80% of EBF closed in case of partial injury (n=66) while only 54% of EBF closed in case of complete injury (n=95) over a period of 5−7 weeks. BBS formed in about half (24/46) of the cases of partial injury but in almost all (73/76) cases of complete injury (unpublished data). If the EBF does not close even after waiting for an adequately long (3−6 months) time, repair may have to be performed in the presence of an ongoing fistula (and undilated ducts). Repair should always be performed by a biliary surgeon, as repair in the hands of a non-biliary general or laparoscopic surgeon is associated with a high failure (restricture) rate [26].

Investigation

A complete proximal cholangiogram is a must before repair is attempted – this invariably is in the form of an MRC. Percutaneous transhepatic cholangiography (PTC) is rarely indicated, if so then as a part of percutaneous transhepatic biliary drainage (PTBD) to control uncontrolled cholangitis. Vascular injury should be looked for in all patients with BDI by MR angiography (MRA), CT angiography (CTA) or Doppler US. The associated vascular injury, however, does not alter the management of the BDI except that an early repair should not be performed in the presence of a vascular injury. An upper GI endoscopy should be done to look for evidence of portal hypertension (varices, gastropathy) due to secondary biliary cirrhosis (SBC) in patients with a long standing (>6 months) injury repair interval.

The most commonly used classification for acute BDI is Strasberg’s [20] while that for BBS is Bismuth’s [5].

Management

Preoperative biliary drainage to reduce the serum bilirubin level is rarely indicated before repair of BBS. The author uses it only when serum bilirubin is very high (>20−25 mg/ dL), especially if associated with coagulopathy. PTBD may, however, be required to control acute cholangitis. Immediate preoperative transhepatic biliary catheterization (PTBC) is always performed before repair of a high (Bismuth Type IV) or difficult (Bismuth Type V) BBS.

The repair of a BBS is in the form of a Roux-en-Y HJ. The technique of a good HJ includes a generous long right subcostal incision, table mounted self retaining retractor for the costal margin, careful lysis of adhesions between parietes, liver, duodenum and colon, exposure of the hepatic helium, identification of the proximal (dilated) bile duct, lowering of the hilar plate at the base of segment IV so as to expose the extrahepatic left hepatic duct (LHD), ductotomy extending into the LHD, creation of an adequately long (45−60 cm) Roux loop of jejunum and a wide mucosa to mucosa biliary enteric anastomosis performed in a single layer with interrupted fine (4−0) long acting absorbable (e.g. PDS) sutures; use of transanastomotic stents is controversial – while the Johns Hopkins Hospital group routinely uses transanastomotic stents [15], the author uses them selectively in patients with high (Bismuth Type IV) BBS and/or undilated ducts. If used, the stents should be retained for a longer period (6−12 months).

In some patients with SBC and portal hypertension, direct access to the hepatic hilum for HJ may be hazardous due to the presence of high pressure collaterals in the hepato-duodenal ligament. A porto-systemic shunt should be performed first to reduce the portal venous pressure to be followed by HJ after 6−8 weeks. Some of these patients may even be candidates for a liver transplant.

Hepatectomy (usually right) may be required either in the acute setting (for liver necrosis due to associated vascular injury) or in a high (Bismuth Type IV) or recurrent stricture, especially when associated with a vascular injury and/ or for atrophy-hypertrophy [11]. There are several reports of patients with BDI, especially with vascular injury, requiring a liver transplant – either as an emergency for acute liver failure or electively for SBC and portal hypertension [13].

While endoscopic intervention plays a very important role in the management of acute BDI and EBF to reduce bile leak, its role in the management of an established BBS is limited. Endoscopic management includes balloon dilation of the stricture and placement of multiple stents with frequent exchanges. This is suitable for partial strictures with intact ductal continuity and provides better results in low (Bismuth Type I and II) strictures. In a large series of 96 BBS treated by endoscopic stenting – a median of 1.9 (1−4) stents were required for 12 (2−96) months; over a follow up of 6 (0−20) years, success was achieved in 76% patients; the success rate increased to 82% after additional treatments [23]. Endoscopic management may also be preferred in patients with BDI and portal hypertension due to SBC. Endoscopic balloon dilation is the treatment of choice for a strictured end-to-end repair of a divided CBD. Fully covered self expanding metal stents have recently been used but they are more effective in biliary strictures in chronic pancreatitis than in post cholecystectomy BBS [8].

Follow up

All patients with BDI, especially those who have undergone HJ for BBS, require regular follow up with liver function tests (LFT) and US. Isotope hepato-biliary scan is a useful investigation to evaluate the HJ. In case of suspicion of anastomotic (recurrent) stricture, MRC is indicated. Treatment of choice for anastomotic stricture is percutaneous transhepatic balloon dilation. Most anastomotic strictures occur within 2 years of HJ but can occur even after many years, thus necessitating the need for long, preferably lifelong, follow up. A meta-analysis of 6 studies including 831 patients, revealed that BDI patients were more likely to have reduced long term mental health-related quality of life as compared to LC patients [10].

Medico legal issues

BDI during cholecystectomy is one of the leading reasons for a medical malpractice suit. Immediate non-specialist repair is an independent predictor of a possible litigation. BDI was responsible for 41% of 300 claims related to LC in England between 1995 and 2008; BDI resulted in 86% of successful claims [9] and average payment for a successful claim was 102,827 GBP [3].

BDI is not only a potentially fatal complication for the patient and a possible medico legal suit for the surgeon, but also a heavy burden on the health care system [6].

The injuring surgeon, if not a biliary surgeon, should avoid the temptation to repair the BDI. Repair by a non-biliary surgeon is not only less likely to be successful but is also less cost effective [6] and more likely to result in a successful litigation [14]. An unsuccessful attempt at repair makes the future repair, even if done by a biliary surgeon, more difficult and less successful.

Conclusions

All BDIs are iatrogenic; most result in morbidity and may even cause death. Many BDIs can be prevented by following the techniques of safe cholecystectomy. Management of a BDI and its sequelae i.e. EBF and BBS, depends on both the type and extent of the injury and the expertise and experience of the surgeon. While immediate (on-table) repair can be performed by a biliary surgeon, early (postoperative) repair is not recommended. Therapeutic endoscopy and interventional radiology play an important role in the early management of a BDI and EBF but have a limited role in the management of a BBS. Repair, in the form of HJ, should be delayed and performed by a biliary surgeon.

Conflict of Interests

The author declares that he has no conflict of interest in connection with the emergence of and that the article was not published in any other journal.

Vinay K Kapoor FRCS

Professor of Surgical Gastroenterology

Sanjay Gandhi Post-graduate Institute of Medical Sciences (SGPGIMS)

Lucknow UP 226014 INDIA

e-mail: vkkapoor.india@gmail.com