Surgical resection of localized cholangiocarcinoma

INTRODUCTION — Cholangiocarcinomas are rare malignancies arising from the epithelial cells of the intrahepatic and extrahepatic bile ducts. Surgical resection can be offered to patients in whom disease appears to be localized and potentially resectable and is individualized according to the location of the tumor within the biliary tree: intrahepatic, perihilar, or distal. True resectability can often only be determined at the time of exploration.

Surgical resection of localized cholangiocarcinoma is reviewed here. The clinical evaluation, diagnosis, and treatment, including when to consider orthotopic liver transplantation, are discussed elsewhere. (See "Epidemiology, risk factors, anatomy, and pathology of cholangiocarcinoma" and "Clinical manifestations and diagnosis of cholangiocarcinoma" and "Adjuvant and neoadjuvant therapy for localized cholangiocarcinoma" and "Treatment options for locally advanced, unresectable, but nonmetastatic cholangiocarcinoma" and "Systemic therapy for advanced cholangiocarcinoma".)

CLASSIFICATION OF BILIARY TRACT CANCERS AND STAGING — Biliary tract cancers are classified according to their locations along the biliary tracts (figure 1) and the combined American Joint Committee on Cancer (AJCC)/Union for international Cancer Control (UICC) Tumor, Node, Metastasis (TNM) cancer staging system (table 1 and table 2 and table 3). These are reviewed in detail elsewhere. (See "Epidemiology, risk factors, anatomy, and pathology of cholangiocarcinoma", section on 'Anatomy, tumor classification, and staging'.)

Unfortunately, neither the Bismuth-Corlette classification nor the AJCC's TNM staging accurately assesses resectability, and true resectability may be ultimately determined only at surgical exploration. (See 'Surgical approach' below.)

A preoperative clinical staging system that accurately assesses resectability would be of value clinically. Such a classification, the Blumgart staging system, has been proposed that is based upon biliary tumor extent, the presence or absence of portal vein involvement, and the presence or absence of hepatic lobar hypertrophy [1,2]. In a series of 376 patients diagnosed with a perihilar cholangiocarcinoma whose disease could be adequately staged, this clinical T staging system accurately predicted resectability, metastatic disease, and the likelihood of a microscopically complete (R0) resection [3]. Independent confirmation of these results is needed.

SURGICAL MANAGEMENT FOR CURE — In general, patients with cholangiocarcinoma have a poor prognosis, with an average five-year survival rate in modern series of 5 to 10 percent for node-positive disease. Surgery provides the only possibility for a cure. Distal cholangiocarcinomas (figure 2) have the rates of highest resectability, while proximal (both intrahepatic and perihilar) tumors have the lowest. In one large series, the resectability rates for distal, intrahepatic, and perihilar lesions were 91, 60, and 56 percent, respectively [4].

Even in patients who undergo potentially curative resection, tumor-free margins can be obtained in only 20 to 40 percent of proximal and 50 percent of distal tumors [1]. These numbers are even lower if a proximal tumor-free margin of at least 5 mm is considered to constitute a curative procedure [5].

However, the current applicability of these data is uncertain. Most of the case series in the surgical literature (even those reported in the last few years) span 10 to 30 years, a time period over which much has changed in cancer and surgical therapy [6]. Resectability rates for cholangiocarcinomas have increased over time, due in part to more aggressive operative strategies and broadened criteria for resectability. As a result, caution should be exercised in the interpretation of outcomes derived from past literature.

Criteria for resectability — The traditional guidelines for resectability of cholangiocarcinoma in the United States include [7,8]:

●Absence of retropancreatic and paraceliac nodal metastases or distant liver metastases

●Absence of invasion of the main portal vein or main hepatic artery (although some centers support en bloc resection with vascular reconstruction [9,10])

●Absence of extrahepatic adjacent organ invasion

●Absence of disseminated disease

Additional criteria are specific to tumor location. As an example, radiographic criteria that suggest local unresectability of perihilar tumors include bilateral hepatic duct involvement up to secondary radicles, encasement or occlusion of the main portal vein proximal to its bifurcation, atrophy of one liver lobe with encasement of the contralateral portal vein branch, atrophy of one liver lobe with contralateral secondary biliary radicle involvement, or involvement of bilateral hepatic arteries [1,11]. These criteria are traditional, and many experienced centers will offer resection to selected patients who may have been deemed unresectable in the past. An example of this would be resection and reconstruction of an involved portal vein segment or resection of local diaphragmatic invasion.

Careful evaluation with radiologic imaging, such as with high-resolution computed tomography (CT) or magnetic resonance imaging (MRI), is particularly important in determining resectability of cholangiocarcinoma [12]. However, true resectability can sometimes only be determined at surgery, particularly with perihilar tumors [13]. Laparoscopy can be useful in ruling out unrecognized peritoneal disease, but perihilar tumors often extend into the liver and major vascular structures, complicating the preoperative evaluation of resectability. Thus, surgical exploration to determine resectability is appropriate when radiological studies cannot rule out resection.

Prognostic factors — The major prognostic factors are margin status, vascular invasion, and lymph node metastases, as reflected by the TNM staging criteria (table 1 and table 2 and table 3). A discussion of prognosis stratified according to the T stage, N stage, and the prognostic stage groupings at all three sites is provided elsewhere. (See "Epidemiology, risk factors, anatomy, and pathology of cholangiocarcinoma", section on 'TNM staging classifications'.)

PREOPERATIVE ASSESSMENT

Preoperative imaging — Imaging is used to diagnose and accurately stage the tumor and may include computed tomography (CT) of the abdomen or cholangiopancreatography (magnetic resonance, endoscopic, transhepatic), and possibly positive emission tomography (PET). (See "Clinical manifestations and diagnosis of cholangiocarcinoma", section on 'The staging workup'.)

Preoperative biliary decompression — Whether preoperative biliary decompression should be performed in patients with cholangiocarcinoma who present with obstructive jaundice is controversial [14,15]. The best method (endoscopic versus percutaneous transhepatic) by which to perform preoperative biliary drainage is also debated [16-18]. (See "Treatment options for locally advanced, unresectable, but nonmetastatic cholangiocarcinoma", section on 'Patients with obstructive jaundice'.)

Most centers selectively perform nonoperative biliary drainage in patients with perihilar cholangiocarcinoma who have a serum bilirubin level over 10 mg/dL, deferring definitive operative intervention until bilirubin levels are less than 2 to 3 mg/dL. However, if stent placement is planned, high-quality imaging, which is necessary to assess unresectability (CT, MRI, endoscopic retrograde cholangiopancreatogram [ERCP], magnetic resonance cholangiopancreatography [MRCP]), should be accomplished in advance since the stent may introduce artifacts that may impede radiographic assessment. Intraoperatively, some surgeons also find the presence of a biliary stent to be a hindrance to determining the proximal tumor extent. On the other hand, cholestasis, liver dysfunction, and biliary cirrhosis can develop rapidly with unrelieved obstruction, and these are main contributors to postoperative morbidity and mortality following surgical resection [13]. Resection of hilar cholangiocarcinoma essentially always requires a concomitant major liver resection, which may require effective preoperative drainage of the planned future remnant liver depending on the remnant volume [13,19]. (See 'Perioperative morbidity and mortality' below.)

Although experimental studies in jaundiced animals suggest that preoperative biliary drainage improves surgical outcomes, clinical studies report variable benefit from preoperative biliary decompression in terms of morbidity and mortality. One study has examined the impact of major liver resection in jaundiced patients [20]. Twenty consecutive patients with obstructive jaundice (14 cholangiocarcinomas, five gallbladder cancers, one hepatocellular cancer) who were to undergo major liver resection without preoperative biliary drainage were matched (for age, tumor size, type of liver resection, and vascular occlusion) with 27 nonjaundiced patients with normal underlying liver selected from a computer bank of 261 patients undergoing liver resection for a variety of reasons. Although there were no significant differences in mortality (5 versus 0 percent) or the incidence of postoperative liver failure (5 versus 0 percent), postoperative morbidity rates (mainly resulting from subphrenic collections and bile leaks) were significantly higher in the jaundiced patients (50 versus 15 percent). These data support a potential benefit for preoperative drainage. Another study supports routine drainage of future liver remnant when the predicted future liver remnant volume is less than 30 percent [19].

However, a meta-analysis of 11 studies addressing the benefit of preoperative biliary drainage in jaundiced patients with perihilar cholangiocarcinoma came to the following conclusions [21]:

●There was no difference in death rate or length of postoperative stay with and without preoperative biliary decompression.

●Overall postoperative complication rates and infectious complication rates were significantly adversely affected by preoperative biliary drainage as compared with surgery without preoperative biliary drainage.

●In the absence of evidence for a clinical benefit, preoperative biliary decompression in jaundiced patients with perihilar cholangiocarcinoma planned for surgery should not be routinely performed. Randomized trials with large sample size and optimal biliary drainage techniques are needed.

Another issue to consider is the possibility for percutaneous catheter tract recurrence. Catheter tract recurrence rates up to 6 percent have been reported in patients with cholangiocarcinoma [22,23]. In one review, survival of 23 patients with catheter tract recurrence was poorer than that of the 422 patients without recurrence (median 22.8 versus 27.3 months) [22]. In another review of 441 patients, for which six patients had catheter recurrence, the median time to recurrence was 11 months. Five of the patients died of systemic recurrence at a median of 3.9 months after detection of catheter tract seeding [24].

These unresolved issues as to the benefit of preoperative biliary drainage and the best means of achieving biliary decompression have led to differing approaches. Some surgeons proceed directly to laparotomy without preoperative biliary drainage [20,25]. However, there is often uncertainty as to resectability as well as the timetable of surgical evaluation and operative management in patients presenting with jaundice (frequently to gastroenterologists, who perform endoscopic stenting). As a practical issue, endoscopic stents are often placed to alleviate jaundice while these issues are being settled. If an endoscopic stent cannot be placed for whatever reason, a percutaneous approach is generally tried. (See "Endoscopic stenting for malignant biliary obstruction".)

For patients with perihilar or intrahepatic cholangiocarcinoma who have a serum bilirubin level over 10 mg/dL, we suggest preoperative drainage of the future liver remnant and defer definitive operative intervention until bilirubin levels are under 3 mg/dL. For those patients who are thought to be potentially resectable, laparoscopic staging can be accomplished shortly after the drainage procedure, while the bilirubin is decreasing to a safe level that will permit definitive resection. If extrahepatic disease or nonresectable tumor is found, curative resection is not possible, and alternative management strategies can be considered.

Preoperative portal vein embolization — Because a histologically negative resection margin is important for the outcome, preoperative portal vein embolization (PVE) has been used to increase the limits of safe hepatic resection [17,26-31]. The intent of PVE is to induce lobar hypertrophy in patients who are predicted to have an inadequate future liver remnant. Preoperative PVE may permit a margin-negative resection in a patient who otherwise would be considered unresectable because of concerns about insufficient postoperative residual liver volume. Preoperative portal vein embolization, including criteria for an adequate future liver remnant volume, is discussed in detail elsewhere. (See "Overview of hepatic resection", section on 'Preoperative PVE and other alternatives'.)

SURGICAL APPROACH — Surgical resection is individualized according to the location of the tumor within the biliary tree: intrahepatic, perihilar, or distal.

Due to vascular involvement that often cannot be detected effectively by imaging techniques, true resectability can often only be determined at the time of exploration. Laparoscopy often is not able to determine vascular involvement but may be useful in ruling out small peritoneal implants before proceeding with a laparotomy incision. (See "Clinical manifestations and diagnosis of cholangiocarcinoma", section on 'Staging laparoscopy'.)

Intrahepatic cholangiocarcinoma — Intrahepatic cholangiocarcinoma is usually treated by hepatic resection to achieve negative resection margins [32]. However, even in appropriately selected patients, curative resection with tumor-negative margins can be achieved in fewer than 30 percent of patients [33-35]. (See "Overview of hepatic resection" and "Open hepatic resection techniques".)

There are few data regarding the benefits of lymphadenectomy for intrahepatic cholangiocarcinoma [36]. Although it is clear that lymph node involvement is an important prognostic factor, lymphadenectomy does not appear to provide proven therapeutic benefit, and there is a lack of consensus as to whether or not it should be routinely performed [37-41]. The risk of performing portal lymph node dissection such as common bile duct devascularization should be weighed against the perceived benefit. The presence of grossly positive porta hepatis lymph nodes portends such a poor prognosis that resection should only be considered in highly selected cases [37]. (See "Adjuvant and neoadjuvant therapy for localized cholangiocarcinoma", section on 'Prognostic factors'.)

Thus, we suggest selective performance of portal lymphadenectomy. (See "Adjuvant and neoadjuvant therapy for localized cholangiocarcinoma", section on 'Intrahepatic cholangiocarcinoma'.)

●For a peripheral localized tumor, we do not perform a routine portal lymph node dissection.

●For more centrally located tumors in which an extrahepatic bile duct resection is required to maximize the width of the resection margin, we recommend portal lymphadenectomy at the same time.

Perihilar cholangiocarcinoma — For perihilar cholangiocarcinomas (eg, a Klatskin tumor), bile duct resection alone leads to high local recurrence rates due to early involvement of the confluence of the hepatic ducts (figure 2) and the caudate lobe branches [42,43]. The addition of a hepatic resection has improved resectability rates [43-46]. However, curative resection is possible in fewer than one-half of patients, and the majority do not achieve long-term disease control [47-52]. (See "Bile duct resection and reconstruction".)

At present, surgical treatment of perihilar cholangiocarcinoma depends upon the Bismuth-Corlette classification (figure 1). (See 'Classification of biliary tract cancers and staging' above.)

●For type I and II lesions, the procedure is en bloc resection of the extrahepatic bile ducts and gallbladder with 5 to 10 mm bile duct margins and a regional lymphadenectomy with Roux-en-Y hepaticojejunostomy reconstruction (see "Bile duct resection and reconstruction"). The addition of an en bloc hepatic lobectomy or trisectionectomy is usually required to achieve adequate negative bile duct margins (5 to 10 mm). For type II tumors, the caudate branches are often involved, and the caudate lobe must be resected. Clinically, it is often difficult to distinguish between types I and II. Most surgeons recommend treating unclear cases as a type II tumor. (See "Open hepatic resection techniques".)

●In addition to the above operations, type III tumors usually require hepatic lobectomy or trisectionectomy [7,42,51,53-57]. Since type II and III lesions often involve the ducts of the caudate lobe, many surgeons recommend routine caudate lobectomy [7,56,57]. (See "Open hepatic resection techniques".)

●Type III and type IV tumors are amenable to potentially curative resection in centers with expertise in these procedures. Aggressive techniques such as multiple hepatic segment resection with portal vein resection (hilar en bloc resection) to achieve negative margins should not be a contraindication to resection [58-60]. (See "Open hepatic resection techniques".)

Highly selected patients with type II, III, and IV tumors that are not resectable due to vascular invasion or primary sclerosing cholangitis may be candidates for liver transplantation. Selection criteria are discussed elsewhere. (See "Adjuvant and neoadjuvant therapy for localized cholangiocarcinoma", section on 'Prior to orthotopic liver transplantation'.)

Distal cholangiocarcinoma — Distal lesions are usually treated with pancreaticoduodenectomy. (See "Surgical resection of lesions of the head of the pancreas".)

PERIOPERATIVE MORBIDITY AND MORTALITY — Perioperative morbidity and mortality have improved with refinements in techniques and perioperative surgical management [6,61]. In a review of 574 resections for perihilar cholangiocarcinoma over a 34 year period, perioperative mortality decreased significantly from 11.1 percent (before 1990) to 1.4 percent. Perioperative morbidity also decreased but remains high at approximately 40 percent. A benchmark for resecting perihilar cholangiocarcinoma has been established by 24 expert centers [62].

Perioperative mortality rates following resection of cholangiocarcinoma parallel the complexity of the procedure needed to manage the condition, related to its associated complications. In a review of the American College of Surgeons National Surgical Quality Improvement Program, overall mortality among 839 patients was 5.4 percent [63]. Mortality for patients with no postoperative complications was <2 percent for the entire cohort and each individual procedure. Mortality was highest at 12 percent for tumors that were managed with hepatectomy and biliary-enteric anastomosis and lowest for those requiring only biliary-enteric anastomosis at 1.2 percent. For any hepatic resection, the extent of liver dysfunction is an important factor that increases postoperative morbidity and mortality. (See "Overview of hepatic resection".)

In an international study of 687 patients undergoing curative-intent resection of intrahepatic cholangiocarcinoma, only 25.5 percent achieved "textbook" outcome (negative margins, no perioperative transfusion, no postoperative surgical complications, no prolonged length of stay, no 30 day readmissions or mortality) [64]. Being <60 years of age, no preoperative jaundice, no neoadjuvant chemotherapy, T1a/b disease, N0 status, and no bile duct resection were independently associated with achieving a textbook outcome after surgery.

RECURRENCE — Following complete surgical resection, the most common relapse pattern is local [65-67], although few local recurrences are isolated. The main risk factors for recurrence include histologically positive margins and lymph node involvement. (See 'Prognostic factors' above.)

Distant metastases (typically hepatic or peritoneal) are also frequent as the initial site of disease recurrence [65,68] but not as common as following resection of gallbladder cancer. In one series, the initial recurrence involved a distant site in 41 percent of patients with a perihilar cholangiocarcinoma (as compared with 85 percent of patients with gallbladder cancer) [66]. Nevertheless, the high rate of distant recurrence provides the rationale for exploring adjuvant systemic chemotherapy as well as radiation therapy (RT)-based regimens for extrahepatic cholangiocarcinomas. (See "Adjuvant and neoadjuvant therapy for localized cholangiocarcinoma", section on 'Adjuvant therapy'.)

LONG-TERM OUTCOMES — Among patients who undergo potentially curative resection for cholangiocarcinoma, long-term outcomes vary according to location and stage of the primary lesion, extent of surgery, associated comorbidities, and treatment-related complications. (See "Adjuvant and neoadjuvant therapy for localized cholangiocarcinoma", section on 'Overview of surgical treatment and prognosis'.)

SUMMARY AND RECOMMENDATIONS

●Localized cholangiocarcinoma – Patients with cholangiocarcinoma require individualized management. The following general treatment principles are applicable to patients whose disease appears to be localized and potentially resectable. (See 'Introduction' above and 'Surgical management for cure' above.)

●Criteria for resectability – For patients with suspected cholangiocarcinoma who do not have clear radiologic or physiologic (ie, liver dysfunction) evidence of unresectability, we recommend surgical exploration (Grade 1B). For patients with radiographic evidence of invasion of the main portal vein or main hepatic artery, we suggest exploration and an attempt at resection at a center with hepatobiliary surgery expertise, provided a margin-negative resection can be achieved with vascular resection and reconstruction of the planned future liver remnant (Grade 2C). The finding of nodal metastases beyond the hepatoduodenal ligament (ie, celiac or periaortic nodes) negates any benefits and contraindicates resection. (See 'Criteria for resectability' above.)

●Preoperative biliary decompression – For patients who present with obstructive jaundice, whether preoperative biliary drainage should be accomplished prior to laparotomy is controversial. For patients with perihilar tumors who have a serum bilirubin level over 10 mg/dL, we suggest preoperative biliary drainage of the future liver remnant (Grade 2C). We defer definitive operative intervention until the serum bilirubin is less than 3 mg/dL. Other surgeons may disagree with our approach, preferring to proceed directly to surgery in all patients regardless of the serum bilirubin level. (See 'Preoperative biliary decompression' above.)

●Preoperative portal vein embolization – For patients requiring hepatic resection who have a predicted future liver remnant volume that is inadequate, we suggest portal vein embolization (PVE) to induce lobar hypertrophy prior to resection (Grade 2C). (See 'Preoperative portal vein embolization' above and "Overview of hepatic resection", section on 'Preoperative PVE and other alternatives'.)