Biliary Tract Tumors

Epidemiology

The prevalence of gallbladder cancer appears to be highest in South America, intermediate in Europe, and lower in the United States and United Kingdom. In the United States the incidence is 1 to 2 per 100,000, but incidence rates are as high as 22 per 100,000 in women in Delhi, India. Epidemiologic analysis suggests that processes promoting chronic gallbladder irritation and inflammation are also risk factors: history of biliary disease, Mirizzi syndrome, age, female gender, obesity, high carbohydrate diet, ethanol abuse, and tobacco abuse (all of which are associated with calculous biliary disease) have been associated with a higher risk of gallbladder cancer. Moreover, 69% to 86% of patients with gallbladder cancer have a personal history of gallstone disease. The presence of an abnormal pancreaticobiliary duct junction, thought to promote chronic biliary inflammation, has been associated with both choledochal cyst disease and gallbladder cancer. The incidence of gallbladder cancer in the so-called porcelain gallbladder, presumably resulting from chronic inflammation and calcification of the gallbladder wall, was once estimated to be as high as 61%; however, more contemporary analyses suggest that the correct figure is more likely between 7% and 25%.

The exact nature of the relationship between chronic inflammation and gallbladder tumorigenesis is unclear. It has been estimated that only 0.3% to 3% of patients with gallstones will develop gallbladder cancer, eliminating any theoretical benefit for prophylactic cholecystectomy (with the exception of porcelain gallbladder).

Anatomy

The anatomic relationships of the gallbladder to surrounding structures dictate the surgical strategies that must be used in its treatment. The gallbladder fossa, against which the fundus and body of the gallbladder lie, is found beneath the junction of hepatic segments IVB and V. As a result, the likelihood of direct hepatic invasion of gallbladder cancer typically mandates resection of these segments. The infundibulum of the gallbladder lies adjacent to the right portal pedicle within the porta hepatis; as a result, tumors arising in the infundibulum commonly invade the right portal pedicle and require a right trisectionectomy for complete surgical extirpation.

The thin gallbladder wall is composed of an inner mucosa, a thin lamina propria, and a single muscularis layer (unlike the two muscle layers that line most hollow viscera). The serosa of the gallbladder is typically opened during a standard cholecystectomy, with the avascular subserosal layer being used as the surgical plane of dissection; the ability of mucosally based tumors to microscopically invade across the serosa explains the high prevalence of positive resection margins after standard cholecystectomy for gallbladder cancer.

The lymphatic drainage of the gallbladder has been well characterized. The pattern of lymphatic flow appears to be directed initially toward the cystic and pericholedochal lymph nodes, then to the posterior pancreaticoduodenal, periportal, and common hepatic artery nodes within the hepatoduodenal ligament, and eventually to the celiac, aortocaval, and superior mesenteric artery nodes. There appears to be no ascending lymphatic drainage into the hilum of the liver. For this reason, meticulous lymphadenectomy within the hepatoduodenal ligament is a critical component of surgical strategy in the management of gallbladder cancer. Unfortunately, the potential for direct drainage from the pericholedochal nodes into the aortocaval nodes explains the difficulty of completely encompassing the extent of lymphatic involvement after surgical resection.

Pathology

Approximately 60% of gallbladder cancers arise in the fundus, with 30% arising from the body and 10% from the neck. Although it is likely that gallbladder cancer may follow the pathogenic sequence of mucosal dysplasia to carcinoma in situ to invasive cancer, it is unlikely that most gallbladder cancers arise from precursor adenomata.

Gallbladder cancers have been categorized as infiltrative, nodular, combined nodular-infiltrative, papillary, and combined papillary-infiltrative. Infiltrative tumors, which are the most common variety, initially appear as indurated areas of gallbladder wall thickening that spread into the subserosal plane, which is typically violated during routine cholecystectomy. Nodular tumors invade into adjacent pericholecystic structures early, but unlike infiltrative cancers, induce sharply defined borders that can facilitate curative resection. Papillary tumors tend to grow in a polypoid fashion, often filling into the lumen of the gallbladder with minimal wall invasion; as such, this variety of tumors tends to be associated with more favorable prognoses.

Microscopically, adenocarcinoma is the most common histologic subtype seen with gallbladder malignancies. Other histologic subtypes that have been reported include adenosquamous carcinoma, oat cell carcinoma, sarcoma, carcinoid, lymphoma, and melanoma. Histologic grading for gallbladder cancer, which has been recognized as a significant prognostic variable, is categorized from G1 (well differentiated) to G4 (undifferentiated); patients most commonly present with G3 (poorly differentiated) tumors.

The propensity of gallbladder cancer to penetrate beyond the single muscle layer of the gallbladder wall results in a high likelihood of tumor penetration into the liver, peritoneal cavity, and lymphovascular spaces at the time of diagnosis. Review of the literature suggests that only 10% of cases are confined to the gallbladder wall at the time of diagnosis; 59% exhibit direct invasion into hepatic parenchyma, 45% demonstrate lymph node metastases, and 20% present with distant extrahepatic metastases. Indeed, a high level of suspicion for occult intraperitoneal metastases should be maintained throughout the diagnostic process for patients with gallbladder cancer. The most common site of extraabdominal spread is the lungs, although pulmonary metastases are rare in the absence of extensive intraperitoneal disease.

Mutations and Targeted Therapies

In the coming years, relevant evidence for this section should grow dramatically. Traditionally, biliary and gallbladder cancers have been poorly researched and thereby mechanisms of carcinogenesis remain poorly understood. Earlier studies posited that gallbladder carcinoma develops via distinct pathways, either occurring de novo with predominant p53 alteration with low percentage of K-ras mutation, or via adenoma-carcinoma sequence in the absence of p53, K-ras, or adenomatous polyposis coli (APC) gene mutations. Although targeted therapies have yielded promising results in other more uniformly developed malignancies, it appears that a variety of heterogeneous pathways can result in gallbladder cancer development. This may explain lack of success in targeted agents against biliary tract cancers thus far because few have examined targeted therapies in relationship to mutations carried. Mutations identified thus far include BRAF mutations in 7 of 21 (33%) resected specimens, and interestingly K-ras and BRAF mutations were not identified in the same specimens. There is much left to learn, but targeted therapies do represent a promising avenue for future adjuvant therapies.

Diagnosis

Patients with gallbladder cancer may experience complaints that mimic those of benign biliary colic. Symptoms of persistent pain, weight loss, anorexia, jaundice, and a palpable right upper quadrant mass are typically indicative of advanced disease that is not amenable to surgical resection. A review of the Memorial Sloan Kettering Cancer Center (MSKCC) experience highlighted the observation that 95% of patients presenting with jaundice were ultimately noted to harbor unresectable disease.

Tumor markers provide limited assistance with diagnosis. In the presence of appropriate symptomatology, carcinoembryonic antigen (CEA) elevations greater than 4 ng/mL have been shown to predict gallbladder cancer with 50% sensitivity and 93% specificity. Similarly, elevations of carbohydrate antigen (CA) 19-9 greater than 20 U/mL are 79.4% sensitive and 79.2% specific. Radiographic findings on ultrasonography include the presence of a polypoid gallbladder mass (seen in 27% of gallbladder cancer cases) or an invasive gallbladder-based lesion (seen in 50% of cases); other sonographic findings consistent with gallbladder cancer include discontinuous gallbladder mucosa, echogenic mucosa, or submucosal echolucency. Computed tomographic (CT) findings seen in patients with gallbladder cancer include a mass filling the gallbladder lumen in 42% of cases, a polypoid mass in 26%, a mass in the region of the gallbladder fossa without a distinctly recognizable gallbladder in 26%, and diffuse wall thickening in 6% ( Fig. 112.1 ). Magnetic resonance imaging (MRI) and magnetic resonance cholangiopancreatography (MRCP) are especially accurate means of identifying small hepatic metastases and involvement of the common bile duct.

Despite the high frequency of nodal involvement, definitive preoperative identification of lymph node metastases is challenging. Enlarged benign inflammatory lymph nodes are commonly encountered at the time of laparotomy. Although the CT finding of ringlike or heterogeneous enhancement of a more than 10-mm large lymph node has been found to identify lymph node metastases with 89% accuracy, only 38% of nodal metastases are preoperatively identified by CT. Endoscopic ultrasonography may be useful for assessing peripancreatic and periportal adenopathy. Fluorodeoxyglucose positron emission tomography (PET) is useful for identifying distant metastases that may contraindicate surgical intervention, but PET efficacy for nodal metastases is limited.

The striking ability of disseminated gallbladder cancer cells to implant within needle tracts limits the utility of percutaneous core biopsy for diagnosis. Percutaneous fine-needle aspiration appears to have a lower incidence of needle tract seeding while providing satisfactory diagnostic accuracy, and it can be used in cases of surgically unresectable disease in which a definitive tissue diagnosis may direct nonoperative therapy. Cytologic analysis of bile samples collected either percutaneously or endoscopically is not often helpful for diagnosing gallbladder cancer, with suboptimal sensitivities of approximately 50%.

Staging

The most accurate predictor of outcome is tumor stage. With changes and improvements in surgical therapy, the impact of various staging criteria has evolved. For example, previous iterations of the American Joint Committee on Cancer (AJCC) staging system categorized patients with tumors extending into the liver as having unresectable stage IV disease. With the increased implementation of modern hepatic resection techniques, curative resection is now possible for this subset of patients with gallbladder cancer; the current AJCC system ( Table 112.1 ) now categorizes patients with disease invasive into the liver within stage III. In addition, nodal metastases found outside of the hepatoduodenal ligament (N2) have been shown to carry the same ominous prognostic weight as distant nonnodal metastases (M1) ; for this reason, the current AJCC system categorizes any patients with N2 disease within stage IV.

This emphasizes the importance of thorough lymphadenectomy. To reiterate, the cystic duct, common bile duct, hepatic artery, and periportal lymph nodes are considered N1 disease in the 7th edition of AJCC staging manual, whereas the paraaortic, paracaval, superior mesenteric artery, and celiac lymph nodes are N2 nodes. Standard lymphadenectomy for regional clearance is thus limited to the hepatoduodenal ligament. If N2 disease is suspected, most surgeons will sample nodes in the N2 distribution for staging purposes but will not completely clear these nodal basins in the interest of sparing the patient added morbidity. Imaging plays a critical role when considering surgical re-resection after incidentally found gallbladder cancer, or when planning surgery for suspected gallbladder cancer. Most authors advocate for some combination of CT, MRI, and PET.

Surgical Management

The standard template on which all gallbladder cancer operations should be based is the so-called radical or extended cholecystectomy. This consists of cholecystectomy with en bloc resection of a rim of segments IVB and V and lymphadenectomy of the cystic, pericholedochal, periportal, and posterior pancreaticoduodenal lymph nodes residing in the hepatoduodenal ligament and local aortocaval lymph nodes ( Fig. 112.2 ). Knowledge of a patient's tumor stage and familiarity with the general biologic proclivities of gallbladder cancer allow the surgeon to tailor surgical therapy to the individual oncologic needs of each patient. For example, lymphadenectomy can often be performed by simply skeletonizing the porta hepatis. However, in cases of prior dissection in which cicatricial changes in the porta hepatis might blur any distinction between tumor and postoperative changes, in patients with infundibular tumors extending into the region of the common bile duct or in very obese patients, resection of the extrahepatic biliary system with Roux-en-Y hepaticojejunostomy reconstruction may be necessary to accomplish a margin-negative resection and adequate lymphadenectomy.

In practice, the surgeon is often confronted with gallbladder cancer diagnosed incidentally on final pathology after a routine simple cholecystectomy. In the scenario in which the diagnosis of gallbladder cancer is unexpectedly made at the time of laparoscopy, the operating surgeon should discuss the situation with the patient's family and either convert to an open exploration for possible extended cholecystectomy with portal lymphadenectomy, or abort the procedure with subsequent reexploration or referral. The only exception to this is if the patient is at a specialty center where surgeons are well-versed in minimally invasive techniques and are capable of safely performing resection and lymphadenectomy laparoscopically or robotically.

In a more recent series, 66% of those presenting after incidental discovery were eligible for reexploration, and, of those, 17% had no evidence of residual disease. Ultimately, 62% of those reexplored underwent R0 or curative resections, but this represents only 41% of those presenting following incidental disease discovery. Still more discouraging, evidence suggests that the presence of residual disease identified on liver resection following cholecystectomy with incidentally discovered gallbladder cancer confers similar survival to patients with stage IV disease. This makes the case for prudent preoperative investigation prior to index operations and likely reflects the invariable violation of the subserosal plane during laparoscopic cholecystectomy that can result in inadvertent peritoneal seeding. This also underscores the importance of staging laparoscopy, which remains an effective means of identifying patients with unresectable gallbladder cancer.