Postcholecystectomy benign biliary stricture: Surgery is the gold standard
Sadiq S Sikora
Chief, Department of Surgical Gastroenterology, Manipal Institute of Liver and Digestive Diseases, Manipal Hospital, Bangalore, India
|Date of Web Publication||13-Apr-2012|
Sadiq S Sikora
MS, FRCS, FACS, Chief, Surgical Gastroenterology, Manipal Institute of Liver and Digestive Diseases, Manipal Hospital, Bangalore
Source of Support: None, Conflict of Interest: None
Post cholecystectomy bile duct strictures present a challenge to the treating physicians. Advancement in skills and technology offers alternative treatment modalities to the standard surgical repair. Contemporary series of surgical repair by experienced surgeons report excellent long-term results with <5% restricture rates. Endoscopic therapy is conceptually flawed, is not applicable to all patients, requires prolonged duration of treatment with multiple interventions. Surgical repair by an experienced surgeon is the "Gold Standard" of care in management of postcholecystectomy bile duct strictures.
Keywords: Benign iliary tricture, roux-en-Y hepatico-jejunostomy, endotherapy
|How to cite this article:|
Sikora SS. Postcholecystectomy benign biliary stricture: Surgery is the gold standard. J Dig Endosc 2012;3, Suppl S1:36-9
|How to cite this URL:|
Sikora SS. Postcholecystectomy benign biliary stricture: Surgery is the gold standard. J Dig Endosc [serial online] 2012 [cited 2018 Mar 10];3, Suppl S1:36-9. Available from: http://www.jdeonline.in/text.asp?2012/3/5/36/95029
It has been estimated that 0.1-0.2% of patients develop a biliary stricture after open cholecystectomy.  Since the introduction of laparoscopic cholecystectomy in 1985, the rates of bile duct injury with resultant stenosis from intraoperative injury have decreased as the surgeons have gained greater experience with the procedure. Currently, the biliary stricture rates vary from 0.2% to 0.7%, although some reports indicate figures as high as 2.8%. ,,,
Traditionally, surgical repair of biliary strictures is the standard of care. Surgical treatment involves anastomosing a Roux-en-Y loop of jejunum to the healthy, vascularized and unscarred part of the bile duct. Conventional surgical wisdom dictates avoiding the scarred and unhealthy part of the stricture for anastomosis in order to prevent restricture and facilitate healing Roux-en-Y hepatico-jejunostomy is a one-time, effective and efficient method of treating biliary obstruction due to postoperative bile duct strictures. The hospital stay for most patients ranges from 5 to 10 days. 
Success with endotherapy is not possible because endoscopic stenting is conceptually flawed - the endobiliary stent forcibly dilates the scarred, unhealthy and ischemic part of the bile duct. The process of healing of an injury goes through a cascade of angiogenesis, epithelialization and connective tissue proliferation with eventual remodeling and healing [Figure 1].  In the event of forcible dilatation (as in the dilatation of the stricture) there is bleeding and hematoma formation, scarred tissues are hypoxic and thus there is a proliferative scarring in the process of wound healing resulting in stricture formation [Figure 2]. Thus, long-term patency in the presence of scarring and unhealthy mucosa as in endoscopic treatment is unlikely. Surgical repair, in contrast, results in a mature scar when performed without tension in healthy well-vascularized tissues [Figure 3].
Few studies have directly compared the results of surgical and endoscopic management of benign biliary strictures. The available data come primarily from nonrandomized retrospective series. Davids et al. compared 66 patients treated endoscopically with 35 patients treated surgically.  Endoscopic therapy consisted of placement of a single 10-Fr plastic stent after biliary sphincterotomy followed 6 weeks later by exchange for two 10-Fr stents. These stents were exchanged every 3 months for a period of 1 year. Early complications were more common in the surgical group (26% vs. 8%), but complications during the treatment period were more common in the endoscopic group (27% vs. 0%). The rates of recurrence of strictures were similar in the two groups (17% each) at 50 and 42 months, respectively. Tocchi et al. reported good or excellent results in postcholecystectomy strictures at 60 months in 17 of 22 patients treated surgically and 16 of 20 patients treated endoscopically. 
In a long-term follow-up study of patients treated with endoscopic therapy followed over 9 years,  only 59% completed the treatment and eventually 47% were stricture-free with a relatively short stent removal to restricture period of 2.6 months. In a multicenter French study of postcholecystectomy bile duct strictures, of 99 patients 65 had a successful stent insertion. Of these, only 45(69%) had a partial or total success (45% of the total cases) with only 35 of 65 (54%) with long-term success. 
A more aggressive endoscopic approach may yield better results. The protocol used by Costamagna et al.  consisted of endoscopic dilatation followed by insertion of as many 10-Fr plastic biliary stents as could be fitted into the duct. Three months later, these stents were removed and replaced by an increased number of similar stents till the stricture was obliterated on occlusion cholangiogram. Of the 42 patients who completed the protocol, 2 died of unrelated causes; of the remaining 40 patients, 39 were asymptomatic with normal liver function test and normal abdominal ultrasonography at 49 months, and one patient had developed cholangitis. Dragnov et al. further supported the idea in their series of 29 patients who were treated with sequential insertion of multiple plastic biliary stents; 62% could be treated successfully at follow-up of 48 months.  Therapy failed in 38% of patients, primarily in those with hilar stricture or related to chronic pancreatitis.
Despite all the publications of endoscopic therapy [Table 1], there is a lot of variation as regards the treatment strategy. There is no consensus in terms of a) the number of stents used which varies from 1-6, b) the number and frequency of procedures in a patient, which varies from 2 to 8, and c) duration of treatment which varies from 12 to 24 months. It is also not clear from the data whether a blanket more is better in all parameters mentioned is true or not. Moreover, most endoscopic series have a case selection bias with majority of series includes only patients where the continuity of the duct is maintained, possibly Bismuth type I/II strictures (50-80%).
|Table 1: Endoscopic treatment of postcholecystectomy benign biliary stricture|
Click here to view
The major argument against surgical treatment is the high morbidity of 15-25%, mortality of 6-13% and restructure rate between 17 and 25% over a long-term follow-up. This comparison is unfair and biased since the surgical series report all cases of biliary strictures with 50% and more cases categorized as Bismuth Type III-V in most contemporary series. Moreover, the contemporary series from experienced centers have low mortality (0.6-4%) and restricture (5-13%) rates [Table 2]. So the outcomes of worst case surgical series with difficult strictures are compared to the best case Type I/II strictures in endoscopical series, to suggest that both treatment modalities are comparable. In a subanalysis of Type I/II strictures in our series, the morbidity was 5%, mortality was 1.3%, hospital stay of average 6 days and restricture rate of 3.4% over a follow-up of 5 years.
|Table 2: Surgical treatment of postcholecystectomy benign biliary stricture|
Click here to view
Long-term stenting as recommended in most protocols is associated with significant complications in the form of stent migration and or fever, jaundice, pain and cholangitis leading to secondary biliary cirrhosis. Moreover, morphological changes in the form of narrowing of the ductal lumen, thick walls and increased grade of fibrosis and inflammation of the bile duct have been documented by us.  Therefore, the ill effects of long-term stenting need to be documented by close follow-up in this group of patients.
Although, compared to surgery, endoscopic treatment has the advantage of being less 'invasive' but it is less effective, needs multiple hospital admissions, and is certainly not suitable for all patients. In patients with strictures affecting the region of biliary bifurcation, complete transections and in those with significant loss of length of bile duct with or without associated ischemia, endoscopic stenting has a high chance of failure. Endoscopic management is ideal for patients who have bile leaks from cystic duct stump or from small lateral injuries of the bile duct. Endoscopic treatment may also be feasible in the subgroup of patients with portal hypertension, severe associated medical diseases precluding surgery provided that the anatomy is conducive for the intervention.
There are no randomized controlled trials comparing surgical and endoscopic treatment of postoperative biliary strictures. Comparisons are difficult because of variable criteria for selecting patients, varying duration of follow-up, and absence of uniform criteria for defining successful treatment. In a retrospective study of 157 patients with postoperative biliary strictures, endoscopic stenting was compared with surgery.  Intention-to-treat analysis showed that surgery provides better long-term outcome because patients with total obstruction are not amenable to endoscopic stenting. In a case control study of 42 patients with postcholecystectomy stricture, as compared to endoscopic treatment surgery gave better long-term results.  Assessment of quality of life after bile duct injury has also shown that duration of treatment is an important independent prognostic factor, with prolonged treatment, like repeated endoscopic stenting, having significant psychological impact. 
In summary, it is fair to conclude that surgical repair is the "Gold standard" for patients with postcholecystectomy benign biliary stricture, especially so, for the subset of patients (Bismuth Type I/II) being considered for endoscopic therapy. In experienced centers the results of surgical repair are excellent with more than 90% success rate over long-term follow-up. Endoscopic therapy is certainly feasible in a subset of patients with favorable anatomy or those who are high risk for surgery due to associated medical conditions, with the caveat that prolonged duration of treatment and multiple interventions are the norm with long-term success rates ranging from 47 to 70%.
|1.||Deziel DJ. Complications of cholecystectomy. Incidence, clinical manifestations, and diagnosis. Surg Clin North Am 1994;74:809-23. |
|2.||A prospective analysis of 1518 laparoscopic cholecystectomies. The Southern Surgeons Club. N Engl J Med 1991;324:1073-8. |
|3.||Wherry DC, Marohn MR, Malanoski MP, Hetz SP, Rich NM. An external audit of laparoscopic cholecystectomy in the steady state performed in medical treatment facilities of the Department of Defense. Ann Surg 1996;224:145-54. |
|4.||Adamsen S, Hansen OH, Funch-Jensen P, Schulze S, Stage JG, Wara P. Bile duct injury during laparoscopic cholecystectomy: A prospective nationwide series. J Am Coll Surg 1997;184:571-8. |
|5.||Windsor JA, Pong J. Laparoscopic biliary injury: More than a learning curve problem. Aust N Z J Surg 1998;68:186-9. |
|6.||Sikora SS, Pottakkat B, Srikanth G, Kumar A, Saxena R, Kapoor VK. Postcholecystectomy benign biliary strictures - long-term results. Dig Surg 2006; 23:304-12. |
|7.||Davids PH, Tanka AK, Rauws EA, van Gulik TM, van Leeuwen DJ, de Wit LT, et al. Benign biliary strictures. Surgery or endoscopy? Ann Surg 1993;217:237-43. |
|8.||Tocchi A, Mazzoni G, Liotta G, Costa G, Lepre L, Miccini M, et al. Management of benign biliary strictures: Biliary enteric anastomosis vs endoscopic stenting. Arch Surg 2000;135:153-7. |
|9.||Bergman JJ, Burgemeister L, Bruno MJ, Rauws EA, Gouma DJ, Tytgat GN, et al. Long-term follow-up after biliary stent placement for postoperative bile duct stenosis. Gastrointest Endosc 2001;54:154-61. |
|10.||Kassab C, Prat F, Liguory C, Meduri B, Ducot B, Fritsch J, et al. Endoscopic management of post-laparoscopic cholecystectomy biliary strictures. Long-term outcome in a multicenter study. Gastroenterol Clin Biol 2006;30:124-9. |
|11.||Costamagna G, Pandolfi M, Mutignani M, Spada C, Perri V. Long-term results of endoscopic management of postoperative bile duct strictures with increasing numbers of stents. Gastrointest Endosc 2001;54:162-8. |
|12.||Draganov P, Hoffman B, Marsh W, Cotton P, Cunningham J. Long-term outcome in patients with benign biliary strictures treated endoscopically with multiple stents. Gastrointest Endosc 2002;55:680-6. |
|13.||Vitale GC, Tran TC, Davis BR, Vitale M, Vitale D, Larson G. Endoscopic management of postcholecystectomy bile duct strictures. J Am Coll Surg 2008;206:918-23; discussion 924-5. |
|14.||Lillemoe KD, Melton GB, Cameron JL, Pitt HA, Campbell KA, Talamini MA, et al. Postoperative bile duct strictures: Management and outcome in the 1990s. Ann Surg 2000;232:430-41. |
|15.||Walsh RM, Henderson JM, Vogt DP, Brown N. Long-term outcome of biliary reconstruction for bile duct injuries from laparoscopic cholecystectomies. Surgery 2007;142:450-6; discussion 456-7. |
|16.||Chapman WC, Havely A, Blumgart LH, Benjamin IS. Postcholecystectomy bile duct strictures: Management and outcome in 130 patients. Arch Surg 1995;130:597-602; discussion 602-4. |
|17.||Wagholikar GD, Sikora SS, Pandey R, Prasad KK, Kumar A, Saxena R, et al. Morphological changes in bile ducts following preoperative biliary stenting. Indian J Gastroenterol 2003;22:166-9. |
|18.||De Palma GD, Persico G, Sottile R, Puzziello A, Iuliano G, Salvati V, et al. Surgery or endoscopy for treatment of postcholecystectomy bile duct strictures? Am J Surg 2003;185:532-5. |
|19.||Boerma D, Rauws EA, Keuleman YC, Bergman JJ, Obertop H, Huibregtse K, et al. Impaired quality of life 5 years after bile duct injury during laparoscopic cholecystectomy. A prospective analysis. Ann Surg 2001;234:750-7. |
[Figure 1], [Figure 2], [Figure 3]
[Table 1], [Table 2]
Klinik für Allgemein, Viszeral- und Transplantationschirurgie, Medizinische Hochschule Hannover, Carl-Neuberg-Straβe 1, 30625 Hanover, Germany
Copyright © 2011 Hüseyin Bektas et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background. There is only limited evidence available to justify generalized clinical classification and treatment recommendations for iatrogenic bile duct lesions. Methods. Data of 93 patients with iatrogenic bile duct lesions was evaluated retrospectively to analyse the variety of encountered lesions with the Hanover classification and its impact on surgical treatment and outcomes. Results. Bile duct lesions combined with vascular lesions were observed in 20 patients (21.5%). 18 of these patients were treated with additional partial hepatectomy while the majority were treated by hepaticojejunostomy alone (). Concomitant injury to the right hepatic artery resulted in additional right anatomical hemihepatectomy in 10 of 18 cases. 8 of 12 cases with type A lesions were treated with drainage alone or direct suture of the bile leak while 2 patients with a C2 lesion required a Whipple’s procedure. Observed congruence between originally proposed lesion-type-specific treatment and actually performed treatment was 66–100% dependent on the category of lesion type. Hospital mortality was 3.2% (). Conclusions. The Hannover classification may be helpful to standardize the systematic description of iatrogenic bile duct lesions in order to establish evidence-based and lesion-type-specific treatment recommendations.
Intraoperative injury of the ductus hepatocholedochus (DHC) or hepatic duct is one of the most severe complications in gallbladder surgery. The literature reports an incidence rate of 0.5–0.8% after laparoscopic cholecystectomy and an incidence rate of 0.2–0.3% after open surgical cholecystectomy [2–7]. These lesions are typically demanding for the surgeon as the pattern of injury may be complicated, for example, by concomitant vascular injuries. The treatment of such lesions is associated with a high rate of complications [1, 8]. The choice of surgical reconstruction and the timing of surgical repair are decisive for the long-term course . Numerous surgical and interventional treatment modalities that are available require close interdisciplinary cooperation of gastroenterologists, radiologists, and surgeons [3, 5, 9–13].
We performed a retrospective study to demonstrate the variety of injury patterns and the subsequent therapeutic concepts for the treatment of iatrogenic bile duct lesions and their outcome. For this purpose we used the Hanover classification for iatrogenic bile duct lesions including concomitant vascular lesions  and reevaluated its clinical application and significance in a larger series of cases.
2. Patients and Methods
Data of 93 patients who were treated for iatrogenic bile duct lesions at our institution were analysed retrospectively by chart review. For the follow-up survey a questionnaire was sent to patients and general practitioners (GPs). The questionnaire for the GPs included questions on the clinical condition of the patient and cholangitis-specific laboratory parameters during follow-up. Cholangitis was defined as a number of symptoms including fever, chills, increased infection parameters, and cholestasis values.
Patients were categorised using a new classification (Hanover classification), for which we were able to demonstrate clear advantages over the published conventional classification systems . The new classification covers lesions localised above the bifurcation of the common bile duct in more detail, thus categorising patients, who were otherwise not included in any other published classification system so far. The Hanover Classification was used as described before  (see also Figures 1–5).
Figure 1: Shown is an illustration of an iatrogenic bile duct lesion which is characterized by peripheral bile leakage with connection to the main bile duct system. According to the Hanover Classification and as described and shown previously , such a lesion would be labelled as a type A lesion (permission to use this figure has been obtained from the publisher).
Figure 2: Shown is an illustration of an iatrogenic bile duct lesion which is characterized by stenosis of the common bile duct (ductus hepatocholedochus, DHC) which may be caused by a clip. According to the Hanover Classification and as described and shown previously , such a lesion would be labelled as a type B lesion (permission to use this figure has been obtained from the publisher).
Figure 3: Shown is an illustration of an iatrogenic bile duct lesion which is characterized by tangential injury of the common bile duct (ductus hepatocholedochus, DHC) with or without additional vascular injury. According to the Hanover Classification and as described and shown previously , such a lesion would be labelled as a type C lesion (permission to use this figure has been obtained from the publisher).
Figure 4: Shown is an illustration of an iatrogenic bile duct lesion which is characterized by complete transsection of the common bile duct (ductus hepatocholedochus, DHC) with or without additional vascular injury. According to the Hanover Classification and as described and shown previously , such a lesion would be labelled as a type D lesion (permission to use this figure has been obtained from the publisher).
Figure 5: Shown is an illustration of an iatrogenic bile duct lesion which is characterized by strictures of the common bile duct (ductus hepatocholedochus, DHC). According to the Hanover Classification and as described and shown previously  such a lesion would be labelled as a type E lesion (permission to use this figure has been obtained from the publisher).
The classification of bile duct lesions, their treatment, and outcome were analysed in order to demonstrate the variety of encountered injury patterns and subsequent therapeutic concepts and their outcome in a larger series and in order to evaluate the clinical value of the Hanover Classification.
The gender distribution among the 93 patients was 67 females versus 26 males. The mean age was 61 years (14–79 years.); the mean follow-up period was 53 months (2–172 months).
In 3 patients the lesions occurred in our clinic; 90 patients were referred from other hospitals after an iatrogenic bile duct lesion had been diagnosed. Iatrogenic bile duct lesions occurred after laparoscopic cholecystectomy (), after open cholecystectomy (), and after ERCP (). Indications for cholecystectomy were symptomatic cholecystolithiasis (), chronic cholecystitis (), acute cholecystitis (), contracted gallbladder (), gallbladder empyema (), and in two cases bile duct perforation during an ERCP without concomitant cholecystitis.
In 38 cases the lesion was noted immediately during the primary intervention due to bile leakage. In 23 of these 38 cases the operating surgeon mentioned possible explanations for the lesions in his operative reports: misinterpretations of the anatomy (13 cases), difficult anatomical situations (4 cases), and bile duct anatomical variance (6 cases).
Intraoperative cholangiography (IOC) was performed in 24 cases. However, in 13 cases the lesion remained unnoticed despite the IOC. According to the operative reports which were available for this study, the cystic artery was positively identified prior to transsection in 54 cases while in 8 cases the surgeon could not positively identify the cystic artery. In 59 cases the cystic duct was also positively identified; however, 11 surgeons reported that positive identification had not been possible prior to transsection (see Table 1).
Table 1: Shown is the frequency of identification of the cystic duct (d. cysticus) and/or the cystic artery (a. cystica) prior to transsection during cholecystectomy which was followed by the diagnosis of an iatrogenic bile duct lesion. The frequencies of identification were determined in this study with the available operating reports.
Most lesions were categorised as type D lesions according to the Hanover Classification (). In 12 patients a tangential type C lesion was apparent (see Figure 6).
Figure 6: Categorisation of patients with bile duct lesions and concomitant vascular injuries according to the Hanover Classification (d = A. hep. dex; s: a. hep. sin; p: a. hep. prop.; c: a. cystica; pv: portal vein; com: a. hepatica communis; DHC: common bile duct, ductus hepatocholedochus).
In 20 cases bile duct lesions were associated with additional vascular injuries. Arterial injury was found in 19 cases and injury of the portal vein in 4 cases. In 3 of 4 cases with portal vein injury additional injury of the right hepatic artery was found. In one case the tangential injury of the DHC was associated with a vascular lesion of the right hepatic artery (C2d lesion according to the Hannover Classification). In 6 cases with complete transsection of the DHC below, the bifurcation concomitant vascular injuries to the right hepatic artery were found (D2d lesion). In 4 cases with complete transsection at the level of the hepatic duct bifurcation additional injuries to either the right hepatic artery (D3d, ) or the portal vein (D3pv, ) or to both the portal vein and the right hepatic artery (D3d + pv, ) were found. Complete transsection of the DHC above the bifurcation combined with additional vascular injuries was evident in 8 cases (D4d, ; D4d + pv, ) (see also Figure 6 and Figures 1, 2, 3, 4, 5). 18 of the 20 cases with additional vascular injuries were treated by surgery including liver resection.
Concomitant injury to the right hepatic artery (C2d lesion ; D2d lesion ; D3d lesion ; D4d lesion ; D3d + pv lesion , D4d + pv lesion ) resulted in additional right hemihepatectomy in 10 of 18 cases. 8 of 12 cases with type A lesions were treated with drainage alone or direct suture of the bile leak while 2 patients with a C2 lesion required a Whipple’s procedure (see Table 3).
Only 8 of 93 patients primarily underwent endoscopic therapy only. However, during the later course surgery was indicated in all of these cases due to development of strictures, cholestasis, or recurrent cholangitis. Table 2 shows data of the primary and the definitive interventions performed in our clinic in order to treat the iatrogenic lesion.
Table 2: Therapeutic methods and results of 93 patients with bile duct reconstruction after iatrogenic bile duct injuries.
Table 3: Shown is a summary of specifically proposed initial surgical approaches for different types of bile duct lesions as classified by the Hanover classification versus the actually performed primary or secondary surgical treatment in our study.
In our hospital first-line operative therapy of iatrogenic bile duct lesions was undertaken in 72 patients 0–11 days after the initial injury. Further 21 patients were referred to our hospital for surgical therapy with long-term complications like anastomotic stricture or bile duct stenosis after iatrogenic bile duct lesions (range: 1–15 years after bile duct injury) (for details see Table 2). Among all 93 patients treated in our hospital, hepaticojejunostomy was performed in 53 patients, one patient was treated with arterial reconstruction by primary suture plus hepaticojejunostomy, 14 patients received a hemihepatectomy plus hepaticojejunostomy (one of these procedures was already done prior to referral to us for biliary stricture), nine patients were treated with re-hepaticojejunostomy, four patients were treated by liver resection alone (right hemihepatectomy two cases, segmental liver resection two cases), two patients were operated with a Whipple’s procedure due to intrapancreatic injury to the bile duct, one patient was treated by drainage only, three patients were treated with exploratory laparotomy and adhesiolysis, five patients were treated by direct suture of the bile duct lesion after exploratory laparotomy, and in two patients the simple removal of a clip was sufficient after surgical exploration.
Early complications during the first hospitalisation in our institution requiring urgent operative revision occurred in 19 patients (20.5%) (Table 2). One patient needed urgent operative revision due to postoperative thrombosis of the hepatic artery and the portal vein followed by urgent liver transplantation. Other early complications leading to urgent operative revisions included secondary haemorrhage, bile leakage, anastomotic insufficiency of the hepaticojejunostomy, peritonitis, and duodenal perforation (see Table 2). The mean duration of hospitalisation was 16 days (3–116 days) with a mean stay in the ICU of 2 days (0–116 days).
12 patients required additional reconstructive surgery in our clinic during long-term follow-up, 4 patients needed re-hepaticojejunostomy, and 7 patients had to undergo herniotomy due to an incisional hernia. Liver transplantation was performed in one other case due to chronic secondary sclerosing cholangitis. Further, closure of a tracheostoma was performed in two cases, partial resection of the liver due to recurrent cholangitis following injury of the right hepatic artery and adhesiolysis due to adhesive ileus was necessary in one patient each. Long-term follow-up data of 63 patients was available.
Diagnosis and therapy of iatrogenic bile duct lesions are a challenge for the surgeon [5, 8, 9, 11, 13, 14]. Less than 50% of these lesions are detected and treated adequately during cholecystectomy. The majority of lesions are noticed at a later stage during hospitalisation or later due to their imminent sequelae which may become apparent sometimes months after the cholecystectomy has been performed [6, 9, 15–17]. In 41% () of our cases lesions were detected while the cholecystectomy was being performed; other centres report similar numbers [15, 16].
Our series with a broad variety of different injury patterns of the central bile ducts with and without concomitant vascular involvement clearly demonstrates a great variance in the scope, extent, and invasiveness of surgical interventions for the treatment of iatrogenic bile duct lesions and highlights the amount of complexity in specific care. Taken together it appears obvious that the complexity of heterogeneity in bile duct lesions and their therapy requires a systematic approach for which we have developed the Hanover Classification in order to help the clinician to develop a rationale for decision making in these patients  (see Table 3).
We attempted to investigate to what degree the primary therapy of iatrogenic bile duct lesions was actually in line with the initial procedures that we have proposed for specific types of bile duct lesions according to the Hanover Classification as outlined in our previous publication (see Table 3). It is important to note in this context that the vast majority of primary interventions were carried out prior to referral to us. Interestingly, for type A, type B, type C, type D and type E lesions the previously proposed therapy and the actual primary intervention were identical in 62%, 75%, 90%, 72%, and 100% of cases, respectively (see Table 3). We assume that the majority of surgeons in our area follow the treatment proposals for different iatrogenic bile duct lesions as outlined in our proposed Hanover Classification. Still many patients were referred to us with clinical problems after an initial surgical attempt to treat the bile duct lesion locally first. We believe therefore that the treatment of complicated iatrogenic biliary lesions frequently requires specialist surgical experience in hepatobiliary surgery. We assume that a large but unknown proportion of cases with iatrogenic bile duct lesions were treated successfully locally and were therefore not referred to us or any other centre. We believe that this assumption warrants further investigation.
The frequency of iatrogenic bile duct lesions with additional vascular lesions is reported to be 11–32% [18, 19]. In our collective concomitant vascular injuries were evident in 20 of 93 patients (21.5%). Apart from four cases with portal vein injury, most vascular injuries affected the right hepatic artery. Detection and adequate treatment of these concomitant injuries are essential for the long-term course as the main blood supply to the bile duct system is from the right hepatic artery. Alves et al. reported that they observed no significant differences in the long-term course of patients with postoperative biliary complications, either with or without arterial lesions . In contrast, Schmidt et al. showed that injury of the right hepatic artery increases the risk for the development of biliary complications . In our view injury of the hepatic artery has to be seen as potentially life threatening. It was found that in all three patients who died during hospitalisation a concomitant injury to the common hepatic artery had been diagnosed. We advocate therefore a preoperative angio-CT of the liver. This ensures that the planned operation can be adapted as necessary if there is evidence of a concomitant vascular injury. Intraoperative identification of the hilar structures is frequently seriously complicated by previous infection and previous surgery. In most cases injury of the hepatic artery (often the right hepatic artery) necessitates partial resection of the liver and frequently due to decreased blood supply to the central bile ducts also additional resection of the bifurcation of the common bile duct as well. In our cohort 18 of 20 patients had to undergo partial resection of the liver due to a concomitant vascular lesion, usually to the right hepatic artery. In these cases angio-CT showed a remarkable demarcation and intraoperative inspection a visibly impaired arterial perfusion of the right liver lobe. In one female patient partial resection of the liver was not performed because intrahepatic arterial perfusion was not significantly impaired as demonstrated by angio-CT results as well as intraoperative inspection. Apparently, sufficient collaterals ensured adequate arterial perfusion of the right liver lobe. Generally, the right hepatic artery supplies the bile ducts of the right liver lobe and segment IV. In our experience injury to the right hepatic artery virtually always results in secondary cholangitis with its consecutive complications. In our series the indication for partial resection of the liver is a result of questionable perfusion of the bile duct. In our view, arterial perfusion of the central bile ducts appears also to be a prerequisite for the healing process of biliodigestive anastomosis. In our view end-to-end reconstruction of the hepatic artery should only be considered in cases with fresh arterial lesions. This approach requires immediate detection of vascular injuries.
At least in our view none of the conventional classifications of iatrogenic bile duct lesions is able to differentiate the extent of iatrogenic bile duct lesions in sufficient detail to provide clear treatment guidelines based on comparative studies. This lack of relevant detail includes the lack of consideration for the variance of possible lesion combinations, including vascular injuries [3, 5, 9, 20, 21]. Until today the most frequently cited classification is the Strasberg classification, which is based mainly on the classification according to Bismuth . The Bismuth classification was developed to describe the degree of bile duct lesions in terms of fixed strictures following open cholecystectomy . The advantage of the Strasberg classification is the comprehensive demonstration of bile duct lesions. In our view, the subclassification of detailed lesions of aberrant right bile ducts is inappropriate as this anatomic variance is evident in only 2% of the normal population [3, 22]. For the reasons outlined above we consider the coincidence of concomitant vascular injuries with a frequency rate of 11–32% as much more relevant for proper therapeutic decision making as well as for prognostic considerations. The Strasberg classification does not consider additional vascular lesions. The Neuhaus classification also refrains from classifying concomitant vascular injuries, but characterises in more detail the extent and localisation of bile duct lesions as compared to the Strasberg classification. However, the Neuhaus classification also does not include the extent of lesions including the level of the lesion, for example, a lesion above the hepatic duct bifurcation . In comparison, the classification developed by Siewert includes concomitant vascular injuries, but has weaknesses in the detailed description of bile duct lesions. Therefore, classification of some lesions, for instance of an anatomically aberrant right bile duct, is not always possible. Basically, the same applies to the Steward-Way classification [18, 20, 21].
To categorise our patients, we used the Hanover Classification, which we have developed and validated. Figures 1–5 illustrate this classification. This classification permits us to look at the complete extent of the lesion including possible additional vascular injuries. As this classification also comprises lesions above the bifurcation, we were able to categorise patients, who were not categorised in any of the existing classifications so far. In our collective 17 patients could be classified in this category (D4 lesions), for example, extensive lesions of the bile duct above the bifurcation as well as accidental resections of the bifurcation. Surgical treatment of this type of injury pattern is particularly demanding.
The majority of patients treated in our institution remained symptom-free during follow-up. But in further 19 cases mostly minor surgical interventions were required during the first hospitalisation in our institution. This correlates with data published by other large centres [6, 8, 19]. We consider therefore that our long-term results reflect the adequacy of the actually chosen definitive surgical treatments in our series.
As we have demonstrated here the extent of iatrogenic bile duct lesions is very variable. As this series shows, most iatrogenic bile duct lesions that are referred to a tertiary referral centre are usually very serious complications. The Hannover Classification may be helpful to standardize the systematic description of these lesions in order to establish evidence-based generalized lesion-type-specific treatment recommendations. It must be underlined that this current study is biased by the fact that it was performed at a referral centre for hepatobiliary surgery and may therefore not reflect how iatrogenic bile duct injuries are managed at the national level.
Conflict of Interests
Drs. H. Bektas, M. Kleine, A. Tamac, J. Klempnauer, and H. Schrem have no conflict of interests or financial ties to disclose.