Impact of neoadjuvant chemoradiation on the tumor burden before liver transplantation for unresectable cholangiocarcinoma

Authors


  • See Editorial on Page 509

Abstract

The very early experience with liver transplantation (LT) for cholangiocarcinoma (CC) was dismal because of the poor survival outcomes and the high recurrence rates. However, LT for CC in conjunction with neoadjuvant chemoradiation recently has shown encouraging results, although the data are extremely limited. At our institution between 2001 and 2008, 22 CC patients underwent protocol orthotopic LT at a median age of 45 years (range = 24-63 years). At a median follow-up of 601.5 days (range = 111-1388 days), the median survival time of the cohort was 3.3 years. The 1-, 2-, and 3-year Kaplan-Meier survival probabilities were 90%, 70%, and 63%, respectively, whereas the historical 5-year survival rates were 0% to 18% for intrahepatic CC and 23% to 26% for extrahepatic CC when patients underwent transplantation without neoadjuvant therapy. These encouraging survival rates for patients with this type of tumor, which is difficult to diagnose and treat, are no less significant when they are compared to the national 1- and 3-year survival rates (86% and 68%, respectively) of patients undergoing deceased donor LT for malignant neoplasms of the liver (as reported by the United Network for Organ Sharing). In our series, disease recurrence was significantly associated with a larger residual tumor [6.3 versus 2.0 cm (mean values), P = 0.008] and with a shorter waiting time for LT after the chemoradiation protocol [18 versus 56 days (mean values), P = 0.04]. Our LT protocol for CC was found to be promising for patients with truly extrahepatic CC and for patients within stages I to IIB of the American Joint Committee on Cancer Staging system (100% survival at a median follow-up of 2.2 years), but the results were notably poor for patients with stage III extrahepatic CC (median survival = 1.2 years). These observations highlight the need for accurate preoperative staging of CC for ideal LT recipient selection and the importance of a low tumor burden and a longer wait after neoadjuvant therapy. More effective chemoradiation regimens for reducing the tumor burden and the appropriate timing of LT after neoadjuvant chemoradiation require further research. Liver Transpl, 2012. © 2012 AASLD.

Cholangiocarcinoma (CC), a devastating bile duct cancer, is often diagnosed late in the disease process and accounts for 10% to 20% of deaths from hepatobiliary malignancies.1 Currently available chemotherapeutic regimens neither are curative nor have a survival benefit. Surgical resection for CC patients without lymph node metastases and vascular invasion, when it is anatomically practical, is the current treatment of choice. However, the 5-year survival rates are 22% to 40% for intrahepatic CC, 11% to 41% for hilar CC, and 27% to 37% for distal extrahepatic CC.2-11 This bleak outcome is attributable in part to the difficulty in diagnosing CC early.

The very early experience with liver transplantation (LT) for CC was dismal because of the poor survival outcomes and high recurrence rates. The 5-year survival rates varied from 0% to 18% for intrahepatic CC and from 23% to 26% for extrahepatic CC.12-15 Despite improvements in diagnostic and surgical techniques, recent reports have not been encouraging. A recent German study that was published in 2008 surveyed 25 transplant centers, and it reported 3- and 5-year survival rates of 42% and 31%, respectively, after LT for hilar CC.16 CC recurred in 34% of the patients. When rigorous screening was applied to patient selection, the 3- and 5-year survival rates improved to 57% and 48%, respectively. Another German study that was published in 2008 reported a median survival time of 32.2 months after LT for intrahepatic CC.17 Two of the 10 patients who underwent LT died of tumor recurrence 18 and 21 months after LT, respectively. The 1-, 3-, and 5-year survival rates were 70%, 50%, and 33%, respectively.

Experimental protocols involving neoadjuvant chemoradiation therapy before LT for meticulously screened and selected patients with extrahepatic CC have yielded optimistic results. In 1987, the University of Nebraska Medical Center paved the way for neoadjuvant chemoradiation therapy before LT. The protocol included biliary brachytherapy and 5-fluorouracil (5FU) until LT. Eleven patients underwent LT. When the study was published in 2002, 5 of the 11 patients were still alive with no evidence of tumor recurrence with a median follow-up of 7.5 years.18 Another protocol was the Mayo Clinic Rochester protocol, which included external-beam radiation therapy concurrent with 5FU chemotherapy and followed by brachytherapy and chemotherapy with capecitabine before LT for carefully screened and selected patients. The screening included explorative laparotomy for staging before LT. Since the initiation of the protocol in 1993, 90 patients underwent LT. At the time of its publication in 2008, 71 patients were alive, and 19 had died (8 because of recurrent CC). The 1-, 3-, and 5-year survival rates were 90%, 80%, and 71%, respectively.19

LT protocols with neoadjuvant chemoradiation for CC have shown encouraging results in comparison with existing conventional treatments; however, additional experience from other centers is lacking. In this article, we report our institution's experience with LT for CC after neoadjuvant chemoradiation therapy in select CC patients. Our institution's neoadjuvant LT protocol is an adaptation of the Mayo Clinic Rochester protocol from 1998 to the present.

Abbreviations:

5FU, 5-fluorouracil; AJCCS, American Joint Committee on Cancer Staging; CA19-9, carbohydrate antigen 19-9; CC, cholangiocarcinoma; CI, confidence interval; CT, computed tomography; DIA, digital image analysis; ERCP, endoscopic retrograde cholangiography; EUS, endoscopic ultrasound; FISH, fluorescence in situ hybridization; HA, hepatic artery; LT, liver transplantation; MELD, Model for End-Stage Liver Disease; MRI, magnetic resonance imaging; PSC, primary sclerosing cholangitis; TNM, tumor-node-metastasis; UC, ulcerative colitis; UNOS, United Network for Organ Sharing.

PATIENTS AND METHODS

Mayo Clinic Florida LT Protocol for CC

Diagnosis

The morphology of CC varies from a mass-forming lesion to periductal infiltration and intraductal growth. The diagnosis can be difficult to establish. The exact localization of the CC tumor and the determination of its extent have a significant impact on the management outcome. The classification of intrahepatic CC and extrahepatic CC (including perihilar and distal segment CC) is based on a previous report.20 According to our protocol, patients who are suspected of having CC undergo a meticulous evaluation, which includes the following: tumor markers [carbohydrate antigen 19-9 (CA19-9), carcinoembryonic antigen, cancer antigen 125, and alpha-fetoprotein]; abdominal imaging with multiplanar, multisequence magnetic resonance imaging (MRI) and/or protocol 3-phase liver computed tomography (CT); endoscopic retrograde cholangiography (ERCP) with biliary epithelial brushings for cytology studies; fluorescence in situ hybridization (FISH); and digital image analysis (DIA). The metastatic survey includes a CT scan of the chest, CT or MRI scans of the abdomen and pelvis, a bone scan, and a positron emission tomography scan for select cases. Since 2005, we have performed linear array and radial endoscopic ultrasound (EUS) assessments and aspiration of hilar or celiac lymph nodes for metastases. Cytology and FISH analyses of the lymph node aspirate are performed. DIA was used in the past but is not currently used. Aspiration of the primary tumor is strictly avoided because of the possibility of seeding along the biopsy track and spreading the disease.

Our institution's definitive diagnostic criteria for CC are listed in Table 1. Cases with an indeterminate diagnosis are followed closely with repeat FISH (DIA in the past), cytology studies, and CA19-9 studies at 2- to 3-month intervals.

Table 1. Diagnostic Criteria for CC
Definitive Diagnostic Criteria for CC
Biopsy (transluminal) positive for cancer
Positive or suspicious brush cytology findings
Mass lesion on cross-sectional imaging consistent with CC
Stricture appearing to be malignant and CA19-9 level > 100 U/mL and/or FISH polysomy
Indeterminate Diagnostic Criteria for CC
FISH trisomy (7 or 3)
Dysplasia
FISH polysomy in the absence of a stricture appearing to be malignant
Stricture appearing to be malignant in the absence of a mass lesion, positive cytology findings, biopsy findings, elevated CA19-9 levels, or FISH polysomy

According to the protocol, CC patients are considered for the neoadjuvant therapy if they have not previously undergone surgical attempts at CC tumor resection and have no evidence of vascular invasion or extrahepatic metastases (including positive lymph nodes). Vascular encasement is not a contraindication.

Neoadjuvant Therapy

Radiation therapy is administered initially at a dosage of 4050 to 4500 cGy in 27 to 30 fractions (each fraction = 150 cGy); 2 fractions are given per day for 2 to 3 weeks, and the fractions are separated by 6 hours. This is followed by iridium-based brachytherapy (2000-3000 cGy) for 2 days to 1 week when it is feasible. If this is not feasible, a radiation boost is achieved with 600 to 1500 cGy in 4 to 10 fractions (each fraction = 150 cGy); 2 fractions are given per day for 2 days to 1 week, and the fractions are separated by 6 hours. Intravenous 5FU at 500 mg/m2/day is the initial chemotherapy, and it is administered within 2 hours of the radiation therapy for 3 days during the first week. 5FU at 225 mg/m2/day is infused again simultaneously with brachytherapy and is continued 7 days a week (5 weeks on and 1 week off) until LT. Recently, capecitabine has been used in place of 5FU while the patient is awaiting LT.

With the availability of EUS, at the completion of the pretransplant neoadjuvant chemoradiation therapy, an EUS assessment of regional lymph nodes with biopsy for cytology and FISH studies (and DIA in the past) is repeated. If these studies are negative for metastases, the patient is listed for LT. Then, while the patient is awaiting LT, tumor marker studies, abdominal imaging (MRI), and CT scans of the chest are performed every 3 months until transplantation, and bone scans and EUS assessments of regional lymph nodes are performed every 6 months.

At our center between 1998 and 2008, before the recently established United Network for Organ Sharing (UNOS) regional requirement (policy 3.6.4.5.2, which was approved at a meeting of the UNOS board of directors on November 17, 2009), we did not perform staging laparotomy before LT. We based the decision on the absence of metastases beyond the perihilar bile duct according to EUS and CT/MRI. Staging laparotomy was performed at the time of LT.

Surgical Technique

At the time of LT, a thorough exploration is performed. Any suspicious lesion beyond the porta hepatis is biopsied and searched for evidence of metastasis. In the absence of metastasis, we carry out the porta hepatis dissection immediately above the head of the pancreas. The distal bile duct is mobilized and transected. The distal margin is submitted to a pathological examination. If the distal bile duct margin is positive for CC, a pancreaticoduodenectomy is completed in conjunction with LT. Otherwise, the distal bile duct is oversewn, and we proceed with LT. LT is performed orthotopically with the piggyback technique with a deceased donor graft without venovenous bypass, as we previously reported.21 The portal vein is primarily reconstructed. The hepatic artery (HA) is reconstructed in a primary fashion except in the absence of a usable recipient HA. An infrarenal aortic conduit is then used for arterial inflow. Biliary reconstruction is completed with Roux-en-Y choledochojejunostomy, and a biliary tube is placed via the Roux limb. When the bile duct distal margin is positive for a residual tumor, we perform a standard Kausch-Whipple procedure (proximal pancreaticoduodenectomy including partial gastrectomy) with subsequent duct-to-mucosa pancreaticojejunostomy and Roux limb biliary and alimentary reconstruction.22, 23

Immunosuppression

Immunosuppression induction is performed according to our institution's standard protocol, which includes methyl prednisolone, mycophenolate mofetil, and tacrolimus. Immunosuppresion is maintained with a combination of prednisone (which is tapered over 3 to 4 months) and tacrolimus. Mycophenolate mofetil is discontinued when the tacrolimus serum level becomes therapeutic.

Follow-Up

Our LT recipients are usually followed at set intervals with clinical visits and laboratory studies, which include blood work, cross-sectional body imaging, and reports from the referring physicians. In general, we follow our post-LT patients closely for 21 days after LT before we release them to go home. They are seen again at 4 months and then annually. Patients who have undergone LT for CC are typically followed up at 4, 8, 12, 18, and 24 months, and they are evaluated at each visit for tumor recurrence or metastases via tumor markers and whole body cross-sectional imaging (preferably CT scans of the chest, MRI of the abdomen and pelvis, and bone scans).

Data Collection

With institutional review board approval, patients who underwent LT according to the institution's neoadjuvant protocol for CC were identified in our LT database. The cohort's electronic medical charts were reviewed for pertinent data. The following information was abstracted: each patient's demographics, blood group, and past medical history; the presentation, site, and size of the CC; the mode of diagnosis and confirmation [tumor markers, radiographic imaging, ERCP, EUS, biliary brushing cytology, FISH and DIA, and tumor histology (if biopsy/aspiration was performed)]; details regarding chemotherapy and radiation therapy; and the liver fibrosis stage. Each patient's LT surgical report was reviewed for surgical findings and techniques. The gross and microscopic examination findings of each explant [the site and size; details about the adjacent viscera and perineural, vascular, and lymph node invasion; and the tumor-node-metastasis (TNM) stage] were recorded.

Statistics

Fisher's exact test and t tests were used for comparing independent proportions and means, respectively; a 2-tailed P value less than 0.05 was considered statistically significant. Cumulative survival probabilities with 95% confidence intervals (CIs) were estimated with the Kaplan-Meier method, and log-rank tests were used for comparisons of subgroup survival curves; a P value less than 0.05 was considered statistically significant.

RESULTS

At our institution, deceased donor LT was performed 1557 times between 2001 and 2008. An LT protocol with neoadjuvant chemoradiation was initiated for CC patients in late 2000. All 22 enrolled patients (17 males and 5 females) completed the protocol and underwent orthotopic LT between 2001 and 2008 for CC that was considered to be nonmetastatic on the basis of noninvasive diagnostics. The statistics of the cohort are summarized in Table 2.

Table 2. Summary Statistics of the Study Cohort (n = 22)
Caucasian race (%)68
Male sex (%)77
PSC (%)77
 UC (%)53
Presentation: biliary mass (%)59
 Extrahepatic location (%)85
 Median size (cm)2.4 (1-5.5)
Biliary cytology (%) 
 Definite CC55
 Indeterminate CC45
Fibrosis (%)68
Cirrhosis (%)14
Brachytherapy (%)55
Median age at LT (years)45 (24-63)
Median time from neoadjuvant therapy to LT (days)24.5 (2-153)
Median actual MELD score at LT8 (6-19)

The patients were predominantly Caucasian (n = 15); the rest were Hispanic (n = 3), African American (n = 2), or Asian (n = 2). They had a past medical history that was negative for exposure to hepatitis B or C, but 77% (17/22) had primary sclerosing cholangitis (PSC); a choledochal cyst was resected for 1 patient many years ago. Among the PSC patients, 53% (9/17) had ulcerative colitis (UC), and 33% of these UC patients underwent colectomy because of dysplastic changes. No patient had active UC at the time of LT. Overall, only 14% of the cohort members had cirrhosis.

Sixty percent of the patients presented with a biliary mass; the mass was extrahepatic in 85% of these patients, and the median tumor size was 2.4 cm (range = 1-5.5 cm). The remaining members of the cohort presented with biliary strictures appearing to be malignant. The biliary cytology findings were definitively positive for CC in 55% of the cases, and the rest met the previously reviewed indeterminate criteria.

All 22 patients in the cohort received chemotherapy. Although 55% (12/22) received additional brachytherapy, the rest received an external radiation boost according to the protocol.

LT

The study cohort underwent LT 24.5 days (median; range = 2-153 days) after the completion of neoadjuvant therapy and 166.5 days (median; range = 76-478 days) after the diagnosis of CC. The donors were predominantly male (55% or 12/22), and their median age was 65 years (range = 17-87 years).

Arterial anatomy of the grafts was as follows: 2 of the grafts had replaced left hepatic artery (HA), 2 had accessory HA, 1 right accessory HA, and 1 replaced right and left HA. An aortic conduit (a donor iliac artery interposition graft to the infrarenal aorta) was used in 9 cases, a common HA to common HA anastomosis was used in 8 cases, and a donor celiac axis to the recipient common HA anastomosis was used in 5 cases. All LT recipients underwent primary portal vein reconstruction. Five recipients required the Whipple procedure at the time of LT.

Seventy-seven percent of the explanted livers (17/22) were found to have residual CC; 11 of these patients had extrahepatic CC, and 5 had intrahepatic CC. One patient had both intrahepatic and extrahepatic residual CC. There was no residual CC on 5 explants. The median size of the residual CC tumors was 4 cm [range = 0.2-14 cm (including the linear length where periductal infiltrating CC was present)]. Eleven of the 12 patients who met the definitive diagnostic criteria for CC before the initiation of the neoadjuvant protocol and LT had absolute evidence of CC on their corresponding liver explants. In fact, 60% of the patients whose CC was considered histologically indeterminate before LT also showed clear evidence of CC in their corresponding explants. Three patients were found to have metastases on resected specimens (a solitary metastatic peritoneal nodule in 1 patient and CC spreading to the adjacent small bowel in 2 patients). Overall, 10 explants showed evidence of perineural, vascular, and lymph node involvement in various combinations [vascular involvement (1 patient), lymph node involvement (3), and perineural involvement (10)]. The TNM stages according to the explants were T1N0M0 (7 patients), T0N0M0 (5), T2N0M0 (3), T4N0M0 (3), T3N1M0 (2), T3N0M0 (1), and T1N1M0 (1).

The study cohort had a median follow-up of 601.5 days (range = 111-1388 days). One patient was lost to follow-up 1326 days after LT. The median survival time of the study cohort was 3.3 years, and the 1-, 2-, and 3-year survival probabilities according to the Kaplan-Meier analysis were 90% (95% CI = 69%-98%), 70% (95% CI = 47%-87%), and 63% (95% CI = 40%-81%), respectively. Among the patients with evident residual CC on the explants (77% or 17/22), 11 also had lymphatic, vascular, perineural, or adjacent visceral invasion, and they were found to have 1- and 3-year survival probabilities of 88% (95% CI = 62%-97%) and 52% (95% CI = 28%-76%), respectively, whereas 100% survival was noted among the LT recipients without residual CC or metastases (n = 5) on the explants. Forty-one percent of the study cohort members (9/22) died 484 days (median; range = 111-1388 days) after LT, but 14% (3/22) actually died from transplant and infectious complications unrelated to CC recurrence. The disease recurrences and metastases in the cohort were predominantly abdominal (75%) 282.5 days (median; range = 120-741 days) after LT. Overall, the rates of disease recurrence were 27% in the first year and 4.5% in the second and third years after LT.

DISCUSSION

First described by Durand-Fardel24 in 1840, CC is an aggressive bile duct tumor that still remains a diagnostic challenge; this challenge is compounded by poor therapeutic options, and the result is a high mortality rate. Currently available chemotherapeutic regimens neither are curative nor have a survival benefit. There is no role for tumor-debulking surgical procedures. Successful surgical resection with a chance of cure requires R0 margins (ie, negative microscopic margins), but this technique is limited by the tumor's infiltrative nature and longitudinal spread pattern and is complicated by the tumor's location and close proximity to vital structures, underlying liver disease and dysfunction. In such complex situations, LT hypothetically should prove to be superior to and more beneficial than surgical resection, but historically LT for CC has invariably resulted in prohibitive recurrence rates and poor survival outcomes.

In 1987, the University of Nebraska initiated a neoadjuvant radiation therapy protocol for LT for CC patients without extrahepatic metastases: brachytherapy was followed by daily intravenous 5FU until the time of LT. As of 2002, their LT protocol had resulted in a tumor-free survival rate of approximately 45% with a median follow-up of 7.5 years.18 In 1993, Mayo Clinic Rochester developed a combination protocol consisting of external-beam radiation therapy and 5FU chemotherapy followed by brachytherapy before LT for carefully screened and selected patients. The protocol was based on Alden and Mohiuddin's experience (Thomas Jefferson University Hospital, Philadelphia, PA, 1984-1990) and Foo et al.'s experience (Mayo Clinic Florida, 1980-1991) with external-beam radiation therapy and iridium brachytherapy for bile duct cancer.25, 26 The Mayo Clinic Rochester LT protocol resulted in 1-, 3-, and 5-year survival rates of 90%, 80%, and 71%, respectively, as of 2008.19

Our LT protocol for CC is an adaptation of the Mayo Clinic Rochester protocol with some modifications during the study period of 2001-2008. The main difference between our protocol and the original Mayo Clinic Rochester protocol is the practice of staging laparotomy. We did not use staging laparotomy after the completion of neoadjuvant chemoradiation to determine the eligibility for the LT wait list. Rather, we explored each patient at the time of LT, and if an obvious metastatic cancer was found outside the porta hepatis, LT was aborted. We aborted 1 case. If we also had aborted LT for the 6 patients with lymph node and adjacent visceral metastases (who would have been excluded by staging laparotomy), the equivalent dropout rate would have been 30% [(6 + 1)/(22 + 1)]. This rate is similar to the dropout rate reported by Mayo Clinic Rochester.27 Our results thus corroborate the necessity of staging laparotomy before a patient with CC is listed for LT. On November 17, 2009, UNOS adopted the Mayo Clinic Rochester protocol (policy 3.6.4.5.2) and began to allow priority Model for End-Stage Liver Disease (MELD) exception scores for CC patients who have completed the neoadjuvant chemoradiation protocol and for whom staging laparotomy is negative. Since March 2009, we have strictly incorporated staging laparotomy into our protocol as the ultimate criterion for determining the candidacy for the LT wait list. We perform staging laparotomy after the completion of neoadjuvant chemoradiation therapy, and the findings determine the final eligibility for LT listing. Our LT protocol for CC has resulted in a median survival time of 3.3 years and 1-, 2-, and 3-year survival probabilities of 90%, 70%, and 63%, respectively, for a cohort that incidentally included patients with intrahepatic and metastatic CC that was undetected before LT. Because staging laparotomy before LT can potentially identify cases with lymph node and adjacent visceral metastases (6/22 in our cohort) and can aid in preventing these patients from undergoing LT, we anticipate better outcomes in the future.

In our series, we have noted that disease recurrence is significantly associated with a larger residual tumor [6.3 versus 2.0 cm (mean values), P = 0.008] after the neoadjuvant chemoradiation therapy, and this emphasizes the importance of selecting candidates with an initially low tumor burden and an effective neoadjuvant chemoradiation regimen before LT.28 The tumor measurements reviewed here are based on explant findings. In patients with infiltrating and longitudinal CC, the explants revealed tumor dimensions beyond the selection criteria, and this emphasizes the diagnostic difficulties for this unique tumor with a periductal infiltrating and intraductal growth pattern. With technological advances, we may be able to accurately determine the tumor location and volumetrics in the near future, and along with stringent LT recipient selection criteria, LT outcomes ought to improve. In the last few years, major changes have been seen in the field of nonsurgical CC management. Recent data about the outcomes of gemcitabine/cisplatin combination therapy for inoperable CC are encouraging in comparison with the outcomes of traditional regimens based on gemcitabine, 5FU, or oxaliplatin. In the future, the incorporation of a more effective chemotherapy regimen into our protocol may further decrease the tumor burden and improve outcomes. In our cohort, brachytherapy (n = 12) had no impact in comparison with a radiation boost (n = 10) on the residual tumor size (P = 0.6) or disease recurrence (P > 0.99), but the outcomes may have been influenced by the small sample size. Currently, at our institution, a radiation boost is more commonly employed. Interestingly, a shorter waiting time for LT after the completion of the neoadjuvant chemoradiation protocol was associated with a higher disease recurrence rate [18 versus 56 days (mean values), P = 0.04]. On the basis of these data, one could infer that a longer waiting period after the neoadjuvant chemoradiation therapy and a tumor reassessment before LT perhaps provide a better assessment of the tumor biology and behavior and thus enable ideal candidate selection for LT. However, a definite waiting period for the tumor biology assessment (before LT and after the neoadjuvant chemoradiation therapy) needs to be defined with future studies and larger cohorts. Historically, the wait time for LT in UNOS region 3 (our region) is very short. As of June 2011, the median waiting time for LT from the time of listing is 157 days for patients with a MELD score of 11 to 18 and 49 days for patients with a MELD score of 19 to 24; this accounts for the short wait time of our cohort before LT.

Our series has demonstrated that a neoadjuvant chemoradiation protocol for LT for CC is promising for patients with truly extrahepatic CC (as defined in our protocol in the introduction) and for patients within the early stages (I-IIB) of the American Joint Committee on Cancer Staging (AJCCS) system. Better survival was observed for patients within extrahepatic CC stages IA, IB, and IIB (100% survival at a median follow-up of 2.2 years).29 The survival was notably poor when patients underwent transplantation for stage III extrahepatic CC (median survival = 1.2 years, log-rank test P = 0.02) and was uniformly poor for patients with intrahepatic CC (stage IIIC and even stage I; median survival = 1 year). This emphasizes the need for the accurate preoperative diagnosis and staging of CC for ideal LT recipient selection and favorable outcomes. At our institution, preoperative staging is performed with a combination of multiplanar, multisequence MRI or triple-phase liver protocol CT, ERCP, EUS, and aspiration of the regional lymph nodes. Preoperatively, the cohort staging was T1N0M0; all patients with intrahepatic CC were at stage I, and patients with extrahepatic CC were at stage IA according to the AJCCS system. However, according to the explant findings, the TNM staging varied and affected the AJCCS results: T1N0M0 (7 patients), T0N0M0 (5), T2N0M0 (3), T4N0M0 (3), T3N1M0 (2), T3N0M0 (1), and T1N1M0 (1). Using these data, we noted that the accuracy of the preoperative AJCCS staging of CC was only 53% (receiver operating characteristic analysis: area under the curve = 0.6, P = 0.1).29 Additionally, in a recent study conducted at our institute and presented at the 2010 American Society of Clinical Oncology Gastrointestinal Cancers Symposium, no specific EUS morphological criteria or echo features were found to correlate with nodal malignancies in patients with CC, and this finding is in agreement with published data.30 Until prospective studies establish specific EUS morphological criteria or echo features that correlate with nodal malignancies in patients with CC (which are crucial for staging and management), staging laparotomy will remain the gold standard for excluding metastatic CC for better LT outcomes. However, staging laparotomy involves logistical challenges in cadaveric donor orthotopic LT because the waiting time on the transplant list and the actual time to LT are unpredictable; this is compounded by the individual's tumor biology and the rate of growth between staging and the actual time of LT. Hence, because of the potential for detecting disease extension at the time of transplantation that was undiagnosed by conventional testing, another LT recipient should be readily available to receive the organ for cadaveric donor orthotopic LT.

In our study, when patients with recurrent CC and metastases after LT were compared with patients in the recurrence-free group, certain favorable characteristics were determined to be statistically significant (t test, Fisher's exact test, and log-rank test; Table 3). An older recipient age, PSC, hepatic fibrosis on the explant, an indeterminate biliary cytology (determined by ERCP) but suspicion for CC (according to the aforementioned criteria), an extrahepatic (ie, hilar) tumor, a smaller tumor size, a longer waiting time for LT after neoadjuvant therapy, a smaller residual tumor after neoadjuvant therapy, and an absence of adjacent visceral involvement were identified as favorable features. Interestingly, in contrast to previously published data,27 explant findings of perineural invasion did not appear to affect the survival outcomes. Also, in comparison with a radiation boost, brachytherapy failed to show a significant correlation with the outcomes.

Table 3. Better Survival Predictors
Better Survival PredictorsPatients With RecurrencePatients With No RecurrenceP Value*
PSC (%)50930.03
Hepatic fibrosis on the explant (%)25930.002
Cytology (%)   
 Definite88360.03
 Indeterminate13640.03
Age at LT (years)37510.002
Time from neoadjuvant therapy to LT (days)18560.04
Residual tumor (cm)6.32.00.008
Visceral invasion (%)3800.03
Better Survival FeaturesP Value
  • NOTE: This table is based on comparisons of patients with post-LT disease recurrence and disease-free patients as well as Kaplan-Meier survival curves for various subgroups.

  • *

    Fisher's exact test and t test.

  • Log-rank test.

Hepatic fibrosis on the explant0.007
Indeterminate biliary cytology for CC0.009
Extrahepatic CC0.05
Absence of adjacent visceral invasion0.001

In our cohort, we observed that patients with evident CC according to biliary brushings (via ERCP) during the pre-LT evaluation had poorer outcomes and a higher tumor recurrence rate than patients with biliary brushings that were indeterminate for CC (87% versus 13%, P = 0.03). However, 60% of the same patients with indeterminate CC had clear histological evidence of CC on their corresponding explants (Table 4). It is conceivable that the remaining CC cases (40%) responded well to the neoadjuvant therapy and hence had no residual tumors.

Table 4. Outcomes Based on the Definite and Indeterminate Diagnostic Criteria and the Residual Tumor Evidence on Explants
CC PresentationPSCDiagnostic CriteriaResidual CC on ExplantLymphovascular and Perineural InvasionVisceral MetastasisDisease Recurrence
MassNoDefiniteYesNoNoYes
MassYesDefiniteYesYesYesYes
MassNoDefiniteYesYesYesYes
StrictureYesDefiniteYesYesNoYes
MassYesDefiniteYesNoNoNo
MassYesDefiniteYesYesNoNo
MassNoDefiniteYesNoYesYes
StrictureYesDefiniteNoNoNoNo
MassYesDefiniteYesYesNoYes
MassYesDefiniteYesNoNoNo
MassYesDefiniteYesNoNoNo
MassYesDefiniteYesNoNoYes
StrictureYesIndeterminateNoNoNoNo
MassNoIndeterminateYesYesNoYes
StrictureYesIndeterminateYesYesNoNo
StrictureYesIndeterminateYesYesNoNo
StrictureYesIndeterminateNoNoNoNo
MassNoIndeterminateYesYesNoNo
StrictureYesIndeterminateNoNoNoNo
MassYesIndeterminateYesYesNoNo
StrictureYesIndeterminateYesNoNoNo
StrictureYesIndeterminateNoNoNoNo

The results of our LT protocol for CC and its outcomes are limited by the small cohort size and the retrospective status. Nevertheless, most of our observations are in agreement with published data for similar neoadjuvant chemoradiation protocols, which emphasize encouraging outcomes when these protocols are undertaken.

In conclusion, the results of this study emphasize the role of neoadjuvant chemoradiation as an important factor for improved survival after LT for CC. More effective chemoradiation regimens may further decrease the tumor burden and improve outcomes. Moreover, our data underscore the important role of staging laparotomy as a part of the UNOS-approved LT protocol for CC. Rigorous adherence to each step in the protocol is essential for optimal outcomes after LT for CC. Further investigations of the exact timing of LT with deceased donor allografts after the completion of neoadjuvant chemoradiation therapy are needed.

Ancillary