Objectives: Perihilar cholangiocarcinoma (Klatskin tumor) is a rare tumor entity, which is diagnosed late due to uncharacteristic symptoms. The therapeutic strategy for cure is still liver resection. Liver transplant in cases of locally irresectable tumors represents an alternative potential curative therapy for a select group of patients.
Materials and Methods: We present our data of 22 patients with irresectable Klatskin tumors who received transplants between 1996 and 2015. We analyzed relevant prognostic factors for the selection of patients to be transplanted to ensure an acceptable overall survival and reviewed known and established selection criteria.
Results: Four factors (age, tumor size, serum level of carbohydrate antigen 19-9, percutaneous transhepatic cholangiodrainage) could be detected for possible patient selection. Positive lymph node status and advanced tumor stage according to the Union for International Cancer Control were confirmed as negative prognostic factors for survival after transplant.
Conclusions: Liver transplant is a curative therapy for selected patients with irresectable Klatskin tumors, but further prospective studies are urgently needed.
Key words : Liver resection, Rare tumors
Perihilar cholangiocarcinoma, also known as Klatskin tumor, is a rare tumor entity, which is often diagnosed too late because of unspecific clinical symptoms.1 In Europe, Klatskin tumor has an incidence of about 1.2/100 000 per year.2,3 Worldwide, the highest prevalence can be found in Southeast Asia. So far, the cause has not yet been conclusively determined. There are several risk factors that play a role in the development of Klatskin tumor, but chronic infection of the bile ducts is of particular importance. In Asian countries, parasitic diseases with ophistorchiida (trematodes), including Chlonorchis sinensis and Opisthorchis viverrini, are worth mentioning. Other risk factors associated with the development of cholangiocarcinoma are high alcohol consumption, primary sclerosing cholangitis, chronic pancreatitis, hepato-/cholelithiasis, caroli syndrome, and choledochus cysts.2,4
Men and women are affected at the same proportions, although several studies have shown a slightly higher incidence in men.4 Histopathologically, 90% to 95% are adenocarcinomas of moderate to poor differentiation and highly desmoplastic cell stroma.2 Sclerosing tumor growth occurs frequently, with incidence of up to 70%.4
Surgical treatment of Klatskin tumor is the only curative option. It has been found that R0 resection, including liver resection, extrahepatic bile duct resection, and lymphadenectomy, is associated with increased survival.5,6 If R0 resection is attained, the 5-year survival rate can be 39% to 50%.7-11 Complications include perioperative mortality and tumor recurrence. The high recurrence rates depend on the radicality of the resection. Most recurrences are local and are followed by liver, lymph node, and peritoneal metastases.2,7 Neoadjuvant and adjuvant therapy concepts are presently under investigation.9,12-14
Treatment with liver transplant has been tried in younger patients in which R0 resection could not be attained after exploration or were not resectable because of poor hepatic parenchyma quality.4,15 Compared with liver resection, 2 major advantages can be achieved with liver transplant: (1) complications regarding the functional reserve from the residual liver with chronic cholestasis and chronic bile duct infection need not be considered and (2) because of discontinuous growth of the tumor intrahepatically, hepatectomy can avoid the risk of leaving undetected residual tumor.15,16 In times of worldwide organ shortage, this concept is seen with great restraint by some transplant surgeons, since results of earlier described procedures have been poor.12,17 Because of donor organ shortage, living-donor liver transplant is a good alternative for patients because it is predictable and can be integrated into multimodal treatment concepts.15
In the United States, studies in which patients had liver transplant after neoadjuvant radiotherapy and chemotherapy have shown marked improvements in survival.8,18 Survival rates of 65% after 5 years and 59% after 10 years with recurrence rates of 20% have been recently shown after neoadjuvant treatment.18,19 These good results of liver transplant are mainly attributable to the Mayo protocol, which was developed by Heimbach and associates.18 In the Mayo protocol, neoadjuvant treatment plays an important role and includes (1) external beam radiation therapy, (2) intravenous fluorouracil chemotherapy, (3) transluminal boost of radiation, (4) and oral capecitabine while waiting for transplant. In addition, (5) prior to transplant, an exploratory laparotomy is recommended to exclude tumor-affected locoregional lymph nodes.18 Inclusion and exclusion criteria for patients who can receive the neoadjuvant protocol are shown in Table 1.
An additional advantage of this neoadjuvant protocol is the “test of time,” as some patients may no longer be eligible for transplant because of tumor progression. These patients often have a poor prognosis after liver transplant due to an aggressive tumor biology.2
Poor results at the beginning of treatment of Klatskin tumor by liver transplant (30% 5-year survival) have been attributed to 3 factors: (1) poor patient selection, (2) no preoperative exploratory laparotomy, and (3) none of the patients received neoadjuvant therapy.8 In this work, we show our data of Klatskin tumor patients who received liver transplants (n = 22) at the Jena University Hospital (Jena, Germany) from 1996 to 2015. Ouraim was to detect further prognostic factors for the selection of patients to be transplanted to ensure recurrence-free and overall survival and to review known and established selection criteria.
Materials and Methods
The 22 patients with irresectable Klatskin tumors who had been transplanted at the Jena University Hospital had been listed in the Eurotransplant database according to Model for End-Stage Liver Disease (MELD)-based allocation after evaluative examinations.
According to the German guidelines for organ transplant, the following exceptional MELD criteria must be fulfilled: (1) biliary structure in cholangiography and biopsy/cytology with detection of a neoplasia, (2) tumor unresectable due to technical considerations or underlying liver disease, (3) lesion diameter < 3 cm as shown by computed tomography (CT) and/or magnetic resonance imaging, (4) no intrahepatic/extrahepatic metastases per CT and/or magnetic resonance imaging and no involvement of regional lymph nodes (exclusion via laparotomy), and (5) transplant within the framework of a prospective study.20 These criteria had been adapted several times during our study period; one of the most relevant changes was staging laparotomy for the exclusion of regional lymph nodes, which was introduced (in our center) in 2010.
Patients received deceased-donor or living-donor liver transplants. Immunosuppression was initiated with basiliximab, calcineurin inhibitors, glucocorticoids, and mycophenolate mofetil, with recent changes during the past years that included maintenance with dual therapy or monotherapy with mammalian target of rapamycin inhibitors. Follow-up examinations, which occurred every 3 months, were conducted by our transplant outpatient department and included laboratory tests and CT examinations (every 6 months).
Statistical analyses were performed with SPSS software (SPSS: An IBM Company, version 21, IBM Corporation, Armonk, NY, USA). Survival was determined with Kaplan-Meier analysis curves. Survival times were compared with the log-rank-test. Comparisons between subgroups were made with the chi-square-test or the Mann-Whitney U test, with comparisons of medians made with the Mood median test. P < .05 was considered significant.
Our evaluation showed a total of 32 patients listed at Eurotransplant. However, in 10 patients, transplant could not be achieved: 3 patients died while on the wait list, 5 patients started the liver transplant procedure but had to be canceled because of tumor spread (median time between listing to transplant and exploratory laparotomy was 47 days), and 2 patients were dropped from the wait list due to tumor progression on CT. Thus, the dropout rate in our population was over 31%.
Some patients received bridge therapy before transplant, including 1 patient who received 2 courses of photodynamic therapy. Another patient received 6 courses of intraluminal brachytherapy (total dose of 24 Gy), but this patient had a complicative course (recurrent relapses of cholangitis with development of a pronounced liver abscess). Explorative laparotomy was performed to determine irresectability and, since 2010, nodal negativity.
Of the 22 patients treated with transplant, 16 were men (72.7%) and 6 were women (27.3%), with mean age of 52.5 years (range, 30-71 y). One male patient had ulcerative colitis, and another male patient had a primary sclerosing cholangitis. Two other male patients had both ulcerative colitis and primary sclerosing cholangitis.
Before transplant, laboratory evaluation showed MELD scores of between 6 and 18 (average of 10). Regarding transplant type, 13 patients (59.1%) received whole organs from deceased donors with cava interposition (in one case as a domino transplant) and 9 patients (40.9%) received right lobes from living donors. All patients were treated with biliodigestive anastomosis. Immunosuppressive induction and therapy were performed according to the center’s guideline. Of the 22 transplanted patients, 9 were adjusted to mammalian target of rapamycin inhibitor-based immunosuppression in the course of observation.
Histologic workup revealed an average tumor diameter of 42.5 mm. It was striking that no correlation could be established with the preoperatively measured tumor size (see Table 2).
Perioperative mortality (30-day mortality) was 13.6% (3/22 patients). Two of the three deceased patients had hemorrhagic shock due to an erosion bleeding (owing to biliodigestive anastomosis insufficiency). The other patient died after aspiration with asphyxia during an epileptic attack (with known hydrocephalus in anamnesis). All complications (higher than grade IIIa) were divided according to the Clavien-Dindo classification and are shown in Table 3.
After exclusion of 30-day mortality, the median survival of the remaining 19 patients was 861 days (range, 145-4641 d), resulting in overall survival rates of 89.2%, 36.0%, and 28.8% at 1, 3, and 5 years. Recurrence-free survival was 691 days (range, 24-1402 d), with a 1- , 3-, and 5-year survival rates of 78.2%, 32.1%, and 24.1%.
The median overall survival, divided by Union for International Cancer Control (UICC) stage, was 3366 days for stage II patients, 1371 days for stage IIIa, 631 days for stage IIIb, and 381 days for stage IV. Median recurrence-free survival was 1402, 1089, 428, and 130 days for patients according to these mentioned tumor stages.
When patients were divided according to UICC stage I/II/IIIa versus UICC stage IIIb/IVa/IVb, survival times were significantly longer in the first group (3366 vs 564 d; P = .003). In addition, for these same groups, recurrence-free survival for patients with UICC stages II and IIIa was a median of 1402 days versus 213 days for patients with UICC stages IIIb or IV (P < .001). Divisions according to individual TMN tumor stages are shown in Table 4. Distribution of patients according to UICC classification (7th edition) is shown in Table 5.
When considering the results of Heimbach and associates and other modified risk factors for less favorable survival after transplant as described by Darwish Murad and associates,21-23 the confirmed major risk factors are as follows: age > 45 years, tumor size > 3 cm, carbohydrate antigen (CA) 19-9 concentration > 500 U/mL pretransplant, and percutaneous transhepatic cholangiodrainage before transplant.
The median overall survival for patients with 2 or fewer of these risk factors was significantly longer (P = .001) at 3366 days than that shown in patients who had 3 or 4 risk factors (564 days). The 3- and 5-year survival rates for patients with 2 or fewer risk factors were 66.7% and 53.3%, and the median recurrence-free survival was significantly longer for patients with fewer risk factors (1402 versus 428 days; P < .001), as shown in Figure 1. Table 6 compares median overall survival, recurrence-free survival, P values, and survival/recurrence after 1, 3, and 5 years for each individual factor.
Liver transplant is a suitable curative therapy for patients with Klatskin tumor who cannot be treated with R0 resection by partial hepatectomy due to intrahepatic tumor extension or an insufficient liver parenchyma reserve. As shown in our results, we conclude that transplant of Klatskin tumor needs further intense examination. Although, in the United States, proper selection and neoadjuvant treatment are presently nearly obligatory before transplant, we suggest that proper patient selection alone may be key to good long-term results, as similarly shown in data from partial liver resection regarding tumor characteristics. In addition to laboratory and radiologic selection criteria, staging laparotomy plays an important role.
In 2007, Robles and colleagues showed that the systematic use of laparotomy before liver transplant improved selection of patients to be transplanted.24 This also led to a paradigm shift in our center, with the integration of staging laparotomy into the standard procedure.
According to Croome and associates, a regional lymph node staging near the time of liver transplant is necessary.25 Regarding the lymphonodular metastasis pathways of the Klatskin tumor, regional (along common bile duct, cystic duct, hepatic artery, or portal vein) and distended metastases (periaortic, pericaval, superior mesenteric artery, or celiac lymph nodes) can be distinguished.26 During explorative laparotomy pretransplant, lymph node sampling should be carried out in the upper edge of the pancreas and left or right of the hepatoduodenal ligament. A “no-touch” technique of the tumor should be considered, and the ligament should not be altered. A minimum number of >2 lymph nodes should be removed. However, lymph node sampling and the no-touch technique (in direct tumor environment) can appear to be problematic in combination.27 Although lymph node sampling is required to assess the extent of the tumor, “extended” sampling can contact the no-touch technique, so that, ultimately, only positive lymph nodes (in the immediate vicinity of the tumor in the ligament) can be diagnosed on the histologic specimen after hepatectomy.
Patient selection for transplant plays a decisive role.28 In our population, the dropout rate was 31%. We were able to find similar results in the literature. Gores et al. as well as Darwish Murad et al. found that there is a high dropout rate while patients await liver tarnsplantation of approximately 30% at 12 months, due to tumor progression19.
Other risk factors for dropout from transplant include CA 19-9 level greater than 500 U/mL a calculated MELD score higher than 20, and a mass lesion greater than 3 cm in radial diameter.5,19,21 These factors were also confirmed in our analyses. The criterion of wait time of > 1 year proposed by Darwish Murad and associates21 was not taken into account because it was only fulfilled by 2 of our patients. In view of the tumor biology of Klatskin tumors, we believe that a waiting period of > 1 year seems unethical, placing both an exceptional MELD and living donations in the foreground of therapeutic concepts.21
The selection factors (CA 19-9 level, lymph node status, and tumor diameter) can give an indication of patients to be transplanted. In our patient population, it was evident that younger patients (< 45 y) had better outcomes, as also shown by Heimbach and associates, who showed that older patients and those with larger, more aggressive tumors are at a higher risk of disease recurrence.22,23 Our patient group also supported these results.
Regarding tumor size, as shown here, the preoperative tumor diameter cannot be correlated with the histopathologic preparation. The exclusion criteria of the Heimbach group of those with tumor sizes greater than 3 cm in preoperative sectional imaging must therefore be reconsidered, even if better survival (although not significant) for patients with a tumor size of less than 3 cm (in the histologic preparation) was shown in our patient group.18 Because most patients do not have mass-forming tumors, consideration of tumor size plays a major role in this process.4 As already described by Poruk and associates, sclerosing tumors are the most common type, comprising up to 70% of all cholangiocarcinomas; these tumors are firm and cause annular thickening of the bile duct, often with longitudinal and radial tumor infiltration.4,29,30 The two subtypes (nodular, papillary) tend to show mass formation but represent only a small number of tumors.
The data published by the Mayo Clinic showing an excellent survival rate can be confirmed here only for subgroups of patients. Because most patients at the Mayo Clinic were at UICC stage 0 or 1, as previously discussed, this effect could have been the neoadjuvant therapy or an incorrect interpretation of a stenosis of unclear dignity as a malignant tumor.16,31-33 Ravaioli and associates posed the question of whether tumor absence was the result of an initial false diagnosis or because of the efficacy of the chemoradiation protocol.34
Living-donor liver transplant is a plannable treatment. Thus, wait time for transplant can be shortened to ensure a better outcome. This was shown by Darwish Murad and associates, who observed that a wait period of > 1 year was a negative predictor for outcome.21 Another point that supports a short wait time for liver transplant is that patients with preexisting liver parenchyma damage run the risk of decompensation with increasing MELD score. Darwish Murad and associates also concluded that a higher MELD score was associated with a 3.5-fold increased risk of dropout.21 If the tumor is diagnosed at an early stage with stable hepatic function (and thus a low MELD score), the dropout rate could be circumvented and the outcome improved. However, living donations are rarely made for patients with Klatskin tumor in Germany.15 Even if not significant, the data presented here may show that perioperative mortality in living-donor liver transplant is equal to deceased-donor liver transplant in patients with Klatskin tumor.
With reference to our selection factors (age, tumor size, serum concentration of the tumor marker CA 19-9, percutaneous transhepatic cholangiodrainage), it can be stated that in our patient group comparisons, that is, in those with 2 or less of the criteria versus 3 or more (Table 6), we showed a significant advantage in both overall and recurrence-free survival in those with less of the selection criteria.
For overall survival, the 1-, 3-, and 5-year survival rates were 100%, 66.7%, and 53.3% for patients with < 2 risk factors versus 77.8% after 1 year and already 0% survival after 3 years for those with 3 or more risk factors (P < .001). Despite our low patient numbers, 4 prognostic factors (age, tumor size, serum concentration of the tumor marker CA 19-9, percutaneous transhepatic cholangiodrainage) could be detected for possible patient selection. The positive lymph node status and advanced tumor stage after UICC were confirmed as negative prognostic factors for survival after transplant.
Liver transplant is a curative therapy for selected patients with irresectable Klatskin tumors. Prospective studies are urgently needed to ensure appropriate patient selection, organ allocation, living donation policies, and administration of antineoplastic immunosuppression.35 Further (controlled) studies are recommended, which could shed light on neoadjuvant versus the “selection alone” concept.
Volume : 17
Issue : 3
Pages : 363 - 369
DOI : 10.6002/ect.2018.0024
From the Department of General, Visceral and Vascular Surgery, Jena University
Hospital, Jena, Germany
Acknowledgements: The authors have no sources of funding for this study and have no conflicts of interest to declare.
Corresponding author: Felix Dondorf, Department of General, Visceral and Vascular Surgery, Jena University Hospital, Am Klinikum 1, 07747 Jena, Germany
Phone: +49 3641 9322601
Table 1. Inclusion and Exclusion Criteria of the Mayo Protocol According to Heimbach and Associates18
Table 2. Comparison of Tumor Size in Preoperative Imaging and Pathologic Findings
Table 3. Perioperative Morbidity by Clavien-Dindo Classification
Table 4. Distribution of the 22 Transplanted Patients According to TNM Classification
Table 5. Division of the 22 Transplanted Patients According to Union for International Cancer Control Classification (7th edition)
Table 6. Comparisons of Median Overall Survival, Recurrence-Free Survival, P Results, and Survival/Recurrence After 1, 3, and 5 Years for Each Factor
Figure 1. Kaplan-Meier Curve of Patients Divided According to Number of Risk Factors