Perihilar cholangiocarcinoma: Current therapy

Abstract

Perihilar cholangiocarcinoma, which is a rare primary malignancy, originates from the epithelial cells of the bile duct. Usually invading the periductal tissues and the lymph nodes, perihilar cholangiocarcinoma is commonly diagnosed in the advanced stage of the disease and has a dismal prognosis. Currently, complete hepatectomy is the primary therapy for curing this disease. Perioperative assessment and available surgical procedures can be considered for achieving a negative margin resection, which is associated with long-term survival and better quality of life. For patients with unresectable cholangiocarcinoma, several palliative treatments have been demonstrated to produce a better outcome; and liver transplantation for selected patients with perihilar cholangiocarcinoma is promising and desirable. However, the role of palliative treatments and liver transplantation was controversial and requires more evidence and substantial validity from multiple institutions. In this article, we summarize the data from multiple institutions and discuss the resectability, mortality, morbidity and outcome with different approaches.

Key Words: Cholangiocarcinoma; Klatskin tumor; Surgery; Liver transplantation; Therapy

Core tip: Perihilar cholangiocarcinoma is a type of malignant tumor with vague and insidious symptoms, and is often diagnosed at an advanced stage. Currently, negative margin resection (R0) is the only way to cure patients with perihilar cholangiocarcinoma. In this article, we describe the surgical procedure and the criteria for operation and illustrate the palliative therapy and liver transplantation options for unresectable perihilar cholangiocarcinoma.

INTRODUCTION

Cholangiocarcinoma, which is a rare malignant tumor, constitutes less than 1% of all human malignancies[1]. The spectrum of cholangiocarcinoma is divided into three types, according to the anatomical location. Perihilar cholangiocarcinoma (PHC) is the most common type of the malignant tumor accounting for 50%-67% of all cases, followed by distal cholangiocarcinoma (DCCA) and intrahepatic cholangiocarcinoma (ICCA), which account for 27%-42% and 6%-8%, respectively[2,3]. When first described by Klatskin, PHC was commonly called Klatskin tumor[4]. Ben-Menachem summarized that the most common risk factors of PHC were liver flukes, primary sclerosing cholangitis, choledochal cysts, hepatolithiasis and cirrhosis, which account for 10% of the cases[5]. Patients with PHC are usually admitted to the hospital with severe painless jaundice and are diagnosed at an advanced stage, which means a poor prognosis and a shortened life span.

Complete resection is recognized as an effective therapy for many carcinomas. Similarly, resection has long been demonstrated to be the best option for patients with PHC, and is associated with long-term survival and better quality of life[6]. PHC surgery was previously considered to be a challenge for hepatobiliary surgeons, because of the complex, intimate and variable anatomical relationship of the bile duct and vascular structures[7]. Because of the anatomical characters and the slow progression of the tumor, palliative procedures have been used to treat cancers involving the hepatic hilus, whereas definitive surgery can only be applied to a minority of patients with well-localized lesions[8]. From 1955 to 1973, Longmire collected 63 patients with extrahepatic cholangiocarcinoma (ECCA), and 34 of those patients had lesions that originated near the confluence of the hepatic duct. However, only six patients (18%) were likely candidates for hepatic resection. Guthrie et al[9] gathered 107 patients with ECCA divided into two periods, 1980-1985 and 1986-1991. They found that the overall resectablity rate (17%) was similar to that reported in other studies, while the use of percutaneous transhepatic cholangiography decreased and the use of endoscopic retrograde cholangiography increased in the second period. However, palliative treatments had unsatisfactory results and were associated with a high incidence of recurrent cholangitis and jaundice. Furthermore, the palliative approaches did not provide a method for curing the tumors; the techniques only served to relieve the symptoms of biliary obstruction.

With the development of radiology, oncology, liver transplantation and a better understanding of the pathways of tumor spread, surgical methods have recently improved significantly. Radical resection with a microscopically negative margin is believed to be the only way to cure patients with PHC. During recent decades, various surgical innovations and strategies have been introduced to achieve this goal. Currently, left or right hepatic resection, routine caudate lobe resection, lymphadenectomy, vascular resection and portal vein arterialization were promoted to improve outcome in patients with PHC. Nevertheless, for those patients who were not candidates for curative resection, several palliative treatments, such as chemotherapy, radiotherapy and photodynamic therapy, could be used to improve the quality of their life.

SURGERY

Staging and assessment of resectability

For various types of cancers, the American Joint Committee on Cancer (AJCC) TNM staging system is the most useful classification. The latest AJCC edition (7th edition) separates the ECCA into PHC and DCCA, which shows that the two subtypes have their own characteristics in pathology, treatment and prognosis. Based on the primary tumor (T), regional lymph nodes (N) and metastasis (M), the stage group is divided into 0-IV. Except for the “basic stage”, the TNM classification has additional descriptions for residual tumor and histological grade. This classification is usually associated with the histological classification, also known as pathological staging, which is mostly used to stage tumors after surgical resection[10]. However, the majority of experts thought that the classification failed to indicate local respectability of the tumor and to distinguish between various surgical options, which limited the use of the staging system in the preoperative setting[11].

Proposed in the 1970s, the Bismuth-Corlette classification is the most useful stage system for predicting the resectability and for assessing the longitudinal intraductal extension of resection. Four types are classified according to the location and the longitudinal extension of the tumor in the biliary tree. Type I lesions involve the common hepatic duct immediately below the confluence; Type II lesions involve the hepatic bile duct confluence, that is beyond the confluence; Type IIIa and IIIb lesions occlude the common hepatic duct and either the right or the left hepatic duct, respectively; and Type IV lesions involve the confluence and both right and left hepatic ducts[12,13]. In Bismuth’s opinion, Types I and II lesions would require only a local resection of the bile duct with a hepaticojejunostomy reconstruction, whereas the right or left hepatectomy for Type IIIa or IIIb lesions and hepatectomy plus liver transplantation for Type IV lesions, could be a contraindication for resection[13]. However, the Bismuth classification fails to describe the radical extension of the cancerous lesion and cannot provide complete information concerning vascular involvement and lymph node involvement, distant metastasis and liver atrophy. Thus, the staging system is primarily used as a convenient guideline for a surgical approach.

Combining the radial and longitudinal extensions of PHC, a preoperative clinical staging system was introduced by Jarnagin and Blumgart at Memorial Sloan-Kettering Cancer Center (MSKCC). This system, which was formally summarized and published in 2001, is also known as the T-staging system , and consists of local tumor extent, biliary duct, portal vein and hepatic lobar atrophy (Table 1)[14,15]. This system could be used to stratify patients preoperatively for the likelihood of respectability and to counsel patients on the potential for an R0 resection. In 2007, Chen et al[16] used this staging system to assess 85 patients with PHC. The 1-year survival rates of T1, T2 and T3 patients were 71.8%, 50.8% and 12.9%, respectively; whereas the 3-year survival rates were 34.4%, 18.2% and 0%, respectively[16]. The patients with PHC in the T1 and T2 stages were likely candidates for curative resection, whereas those in the T3 stage could not achieve R0 resection even if they had undergone resection[16]. Another retrospective test in 380 patients showed that the R0 resection rates for T1, T2 and T3 patients were 44.1%, 36.1% and 1.3%, respectively; whereas the median survival was 22.8, 23 and 10.8 mo, respectively[17]. Both surveys demonstrated that the T stage was associated with resectability and long-term survival. Moreover, the MSKCC provided the criteria for unresectable PHC, which included the following: locally advanced tumor extending bilaterally to the secondary biliary radicles, unilateral sectional bile ducts with contralateral portal vein branch involvement, encasement or occlusion of the primary portal vein proximal to its bifurcation, and atrophy of one hepatic lobe with contralateral tumor extension to sectional bile ducts (Table 2)[15].

Table 1 Memorial sloan-kettering cancer center classification.

Stage	Criteria
T1	Tumor involving biliary confluence ± unilateral extension to second-order biliary radicles.
T2	Tumor involving biliary confluence ± unilateral extension to second-order biliary radicles and ipsilateral portal vein involvement ± ipsilateral hepatic lobar atrophy
T3	Tumor involving biliary confluence + bilateral extension to second-order biliary radicles; or unilateral extension to second-order biliary radicles with contralateral portal vein involvement; or unilateral extension to second-order biliary radicles with contralateral hepatic lobar atrophy; or main or bilateral portal venous involvement

Table 2 Criteria for unresectability[15].

Patient factors

Medically unfit or otherwise unable to tolerate a major operation

Hepatic cirrhosis

Local tumor-related factors

Tumor extension to secondary biliary radicles bilaterally

Encasement or occlusion of the main portal vein proximal to its bifurcation

Atrophy of one hepatic lobe with contralateral portal vein branch encasement or occlusion

Atrophy of one hepatic lobe with contralateral tumor extension to secondary biliary radicles

Unilateral tumor extension to secondary biliary radicles with contralateral portal vein branch

encasement or occlusion

Metastatic disease

Histologically proven metastases to N2 lymph nodes1

Lung, liver, or peritoneal metastases

¹Metastatic disease to peripancreatic, periduodenal, celiac, superior mesenteric, or posterior pancreaticoduodenal lymph nodes was considered to represent disease not amenable to a potentially curative resection. By contrast, metastatic disease to cystic duct, pericholedochal, hilar, or portal lymph nodes (i.e., within the hepatoduodenal ligament) did not necessarily constitute unresectability.

A recent report indicated that a new system was designed by the international cholangiocarcinoma group, which incorporated the size of the tumor, the extent of the disease in the biliary system, the involvement of the hepatic artery and portal vein, the involvement of lymph nodes, distant metastases, and the volume of the putative remnant liver after resection[10]. Despite its comprehensiveness, this new classification must be validated and accepted.

We searched the key words “hilar cholangiocarcinoma”, “Klatskin tumor” and “resection” using Pubmed and Medline, and we summarized the respectability and the outcomes from different institutions in different periods. The results of the surgical treatment are shown in Table 3[2,3,8,15,17-51]. Although the data were not fully calculated and were derived from tertiary referral centers, the number of patients with PHC who had undergone the resection was small, and only few large institutions contained more than 300 cases[42,50,51]. These findings attested to the rarity of this disease; additionally, these results indicated that the majority of patients lost the opportunity to undergo a curative operation when diagnosed, and therefore, these patients were not counted in the total number of study participants. Table 3 shows that the resectability rate was significantly variable, ranging from 28% to 95%, and that the curative resection rate ranged from 14% to 95%. This wide variability may be attributed to the differences in the sample content, the broad range of dates for inclusion, the characteristics of patients in different geographical areas, the methods of patient selection and the preoperative techniques in these studies.

Table 3 Results of surgical resection for perihilar cholangiocarcinoma.

Ref.	Published year	Resections	Resectablity(%)	Negative margin(%)	Liver resection(%)	Morbidity	Mortality	5-yr survival rate(%)
Hadjis et al[18]	1990	27	NA	56	60	NA	7	22
Nakeeb et al[2]	1996	109	56	26	14	47	4	11
Su et al[19]	1996	49	28	49	57	47	10	15
Klempnauer et al[20]	1997	151	45	77	77	NA	10	28
Miyazaki et al[21]	1998	76	NA	71	86	34	13	26
Neuhaus et al[22]	1999	80	NA	61	85	55	8	22
Kosuge et al[8]	1999	65	73	52	80	37	9	33
Gerhards et al[23]	2000	112	NA	14	29	65	18	NA
Nimura et al[24]	2000	142	80	61	90	49	9	26
Todoroki et al[25]	2000	101	89	14	58	14	4	28
Jarnagin et al[15]	2001	80	50	78	78	64	10	26
Kawarada et al[27]	2002	65	89	64	75	28	2.3	26
Capussotti et al[26]	2002	36	NA	89	83	47	3	27
Kawasaki et al[28]	2003	79	75	68	87	14	1.3	22
Seyama et al[29]	2003	87	94	64	67	43	0	40
Rea et al[32]	2004	46	NA	80	100	52	9	26
Kondo et al[31]	2004	40	95	95	65	48	0	NA
I.Jitsma et al[30]	2004	42	NA	65	100	76	12	19
Hemming et al[33]	2005	53	50	80	98	40	9	35
Jarnagin et al[34]	2005	106	70	77	82	62	8	NA
Dinant et al[35]	2006	99	NA	31	38	66	15	27
DeOliveira et al[3]	2007	173	62	19	20	61	5	10
Ito et al[36]	2008	38	55	63	53	32	0	33
Konstadoulakis et al[37]	2008	59	81	68.6	86.4	25.5	6.8	34.9
Igami et al[41]	2010	298	70	74	98	43	2	42
Hirano et al[40]	2010	146	NA	87	87	44	3.4	35.5
Lee et al[42]	2010	302	86	70.9	89	43	1.7	32.5
Unno et al[44]	2010	125	NA	63.2	100	48.7	8	34.7
Ercolani et al[38]	2010	51	49.6	72.5	98	51	10	34.1
Shimizu et al[43]	2010	224	NA	69.1	78	47.6	10.7	30.3
Giuliante et al[39]	2010	43	29	77	93	52.5	6.9	36.1
Regimbeau et al[45]	2011	56	NA	76.9	100	72	8	NA
Young et al[48]	2012	83	92	42.2	93	62.7	7	20
Saxena et al[47]	2012	54	64	64.3	42	45.2	2.4	24
Ribero et al[46]	2012	82	NA	81.7	91.5	64.6	9.7	28
De Jong et al[50]	2012	305	NA	64.2	73	NA	10.6	20.2
Matsuo et al[17]	2012	157	78	76	90	59.2	7.6	37.5
Cheng et al[49]	2012	176	34	78.4	97	26.3	2.9	13.5
Nagino et al[51]	2013	574	76.1	76.5	96.7	57.3	4.7	32.5

NA: Not applicable.

Surgical procedures and strategies

In several reports, the surgical procedures were as follows: (1) preoperative biliary drainage was conducted to reduce the serum bilirubin concentration below 2 mg/dL; (2) preoperative percutaneous transhepatic portal embolization was performed when the volume of the liver remnant was estimated to be less than 40%; (3) the operative procedures for hilar resection were determined and planned using multidetector row computed tomography (MRCT); (4) the skeletonization of the portal vein and hepatic artery was performed using nodal clearance around the head of the pancreas; (5) portal vein resection and reconstruction were conducted before hepatic dissection if necessary; (6) frozen sections of the resected margins of the bile duct were investigated; and (7) lymph nodes in the hepatoduodenal ligament, around the head of pancreas and around the common hepatic artery were completely removed, whereas lymph nodes in the para-aortic region were removed, if possible, with a curative resection[44]. In other institutions, the surgical procedure included the hepatic artery resection, reconstruction and arterioportal shunt.

Obstructive jaundice, which is the most common symptom in patients with PHC, may increase the in-hospital mortality by 10% and is associated with many complications, such as bacterial translocation, malnutrition, renal insufficiency and postoperative liver dysfunction[52,53]. To avoid the risk of hepatic resection, preoperative biliary drainage (PBD) is recommended by many surgical teams. Percutaneous transhepatic biliary drainage (PTBD) had previously been widely used; however, several prospective randomized studies showed that PTBD had no benefit in postoperative morbidity and mortality but increased potential risks, such as vascular injury, infectious complications and tumor seeding metastasis[54-56]. Currently, endoscopic nasobiliary drainage (ENBD) is performed instead of PTBD because of fewer complications and better outcomes. More recently, the Nagoya Institute demonstrated that unilateral ENBD of the future remnant lobe(s) exhibited a high success rate as an effective and suitable PBD method even in BC type III to IV lesions[57]. To avoid the postoperative liver dysfunction resulting from extended hepatic resection, many institutions have promoted portal vein embolization (PVE) to increase volume of the future liver remnant (FLR). In several cautious surgical centers, when the FLR was 40% or less of the total liver volume, PVE was performed because the serum bilirubin level had decreased to less than 10 mg/dL[41,46]. Subsequently, surgery was performed after 2-4 wk of liver hypertrophy due to clonal expansion and cellular response[58].

When determining the surgical approach, the local excision, hepatectomy, and extended hepatectomy with or without caudate resection should be considered. In the Bismuth’s opinion, Bismuth Type I and II would require only a local resection. Recently, bile duct resection alone without hepatectomy has been largely abandoned in favor of a more aggressive approach. Capussotti et al[59] conducted a systematic review of the effect of local resection compared with hepatectomy. In the pathologic aspect, the isolated bile duct cannot be adequately resected, because of the following: the necessity for wide surgical margins; neoplastic extension along the perineural sheaths and segment 1 neoplastic invasion. From another perspective, the R0 resection rate was higher after combined liver resection, although, in the earlier years of its application, local resection could be associated with fewer complications and shorter lengths of hospital stay[15,21,35]. In conclusion, according to this systematic review, local resection should only be scheduled for small papillary Klatskin tumors without bile duct confluence involvement confined to the bile duct wall[59]. Because of the rarity and the advanced stage of the disease at the time of diagnosis, a local resection was rarely performed.

Despite the incomplete accuracy, the Bismuth classification initiated the idea of wider resection for PHC[13]. Table 3 shows that liver resection rates increased from 14% to 100% with an increased R0 resection rate. The common liver resection strategies are as follows: right or left hepatectomy (resection of hepatic segments 5, 6, 7, 8 or 2, 3, 4 ± 1), right or left hepatic trisectionectomy, also called extended right or left hepatectomy (resection of hepatic segments 4, 5, 6, 7, 8 or 2, 3, 4, 5, 8 ± 1), and central hepatectomy. Bisectionectomy or more was defined as a major hepatectomy; sectionectomy or less was defined as a minor hepatectomy[38]. Currently, for those patients with Bismuth type I and II, the right hepatectomy with caudate lobectomy was recommended, which has been demonstrated to decrease the rate of recurrence[29]. However, for those patients with Bismuth types III and IV lesions, the approaches varied in different institutions. Recently, Cheng et al[49] reported 171 patients with PHC of Bismuth types III and IV lesions. For Bismuth Type III lesions, right, left or central hepatectomy with caudate lobectomy was performed. For Bismuth IV lesions, the right or left hepatectomy or extended right or left hepatectomy with caudate lobectomy was conducted to increase the negative margin rates. The choice of surgical side may depend on the predominance of the tumor; however, the right trisectionectomy is indicated for centrally located tumors because of the length of each hepatic duct, the location of the hilar common bile duct in the hepatoduodenal ligament, the ease of complete caudate lobectomy and portal vein reconstruction, and the frequent involvement of the right hepatic artery[7,28]. The left hepatectomy is considered to be a more complicated procedure than the right hepatectomy and requires greater skill, especially in cases involving portal vein resection and reconstruction. Moreover, preserving the right hepatic artery and the right portal vein could be an oncological problem with left or extended left resection, which could increase the tumor cell dissemination. Therefore, the rate of left hepatectomies is approximately 25%-30% of all resections[60]. In the study by Shimizu et al[43], the R0 resection was achieved in all 7 patients who underwent right trisectionectomy, but in only 8 (61.5%) of 13 patients who underwent left trisectionectomy. This finding suggests that a more extended resection from the right side, but not from the left side, may provide greater potential for curability. However, several authors believed that the left extended hepatectomy could achieve the same result. Nagino et al[51] analyzed the patients with PHC who underwent surgery and compared the surgical strategies in different periods (Table 4). From their experience, the incidence of left hepatic trisectionectomies gradually increased while the incidence of central hepatectomies decreased. Totally, the left or extended left hepatectomy represented nearly 55% of all of the resections performed on patients with PHC.

Table 4 Surgery performed according to the time period[51] n (%).

		Time period
	Total	Earlier period		Later period		P
		1997-1990	1991-2000	2001-2005	2006-2010
Number of patients resected	574	72	116	168	218
Resectability	574/754 (76.1)	72/93 (77.4)	116/148 (78.4)	168/216 (77.8)	218/297 (73.4)	0.406
Type of hepatectomy1						< 0.001
S1,4,5,6,7,8	43 (7.5)	5 (6.9)	11 (9.5)	4 (2.4)	23 (10.6)
S1,5,6,7,8	177 (30.8)	17 (23.6)	40 (34.5)	53 (31.5)	67 (30.7)
S1,2,3,4,5,8,	110 (19.2)	4 (5.6)	12 (10.3)	29 (17.3)	65 (29.8)
S1,2,3,4	187 (32.6)	27 (37.5)	35 (30.2)	68 (40.5)	57 (26.1)
S1,4,5,8/S1,5,8/S1,4/S1	38 (6.6)	13 (18.1)	10 (8.6)	11 (6.5)	4 (1.8)
Without hepatectomy	19 (3.3)	6 (8.3)	8 (6.9)	3 (1.8)	2 (0.9)
Combined resection
Pancreatoduodenectomy	74 (12.9)	9 (12.5)	13 (11.2)	20 (11.9)	32 (14.7)	0.553
Portal vein resection	206 (35.9)	23 (31.9)	36 (31.0)	58 (34.5)	89 (40.8)	0.116
Wedge resection	36	15	6	10	5
Segmental resection	170	8	30	48	84
Hepatic artery resection	76 (13.2)	0	5 (4.3)	25 (14.9)	46 (21.1)	< 0.001
Operative time, min2	668 ± 134	664 ± 162	787 ± 170	675 ± 145	605 ± 134	< 0.001
Blood loss, mL2	2491 ± 2156	4414 ± 2791	3773 ± 3024	1898 ± 1268	1768 ± 1130	< 0.001
Homologous blood transfusion	271 (47.2)	68 (94.4)	93 (80.2)	46 (27.4)	64 (29.4)	< 0.001

Homologous blood includes packed red blood cell and fresh-frozen plasma. Note that P indicates the statistical difference between the earlier period (1977-2000) and the later period (2001-2010).

¹Expressed as Couinaud’s hepatic segments resected;

²Excluding 19 patients who did not undergo hepatectomy.

Nimura et al[61] introduced the concept of routine caudate lobectomy (CL). Bilateral biliary branches of the caudate lobe are confluent with the right hepatic duct, the left hepatic duct, the confluence of these and the right posterior hepatic duct. Therefore, the caudate lobe is usually involved in PHC in 40% to 98% of patients, which indicates a need for CL[61-63]. Moreover, routine CL combined with resection had high curative resectablity rates and increased the likelihood of long-term survival for patients with advanced stage PHC[49]. Similarly, Kow et al[64] showed that the patients with CL had a significantly better overall survival rate of 64.0 mo compared to the survival rate of 34.6 mo in type III PHC patients in the group without CL. Although mechanisms for CL have not been established, the outcome remains optimistic while undertaking CL in PHC.

A major hepatectomy combined with pancreatoduodenectomy, for example, hepatopancreatoduodenectomy (HPD), was routinely used in the PHC surgery in several institutions. This procedure occupied 12.9% of the total surgery cases, and was indicated in the following cases: (1) diffusely infiltrating tumors of the entire extrahepatic bile duct; and (2) downward superficial spreading, or bulky nodal metastases of the pancreatoduodenal region (Table 4)[65]. Therefore, HPB provides an important method for treating spreading unresectable cholangiocarcinoma; thus, it is now the fourth standard procedure following hepatectomy, bile duct resection, and pancreatoduodenectomy[66].

In several high-volume samples, PHC was frequently reported to metastasize via the lymphatics in 24% to 75% of the patients[42,51]. Moreover, many authors had demonstrated that lymph node metastasis had a negative impact on survival in PHC[3,28,29,31,33,42,51]. Thereafter, lymphadenectomy played a crucial role in the outcome of patients with PHC. However, the 5-year survival rate is related to the location of the metastasis of the lymph node. Therefore, lymph node metastasis that is confined to the hepatic pedicle or the hepatoduodenal ligament is not a reason for abandoning resection. The tumor positive lymph nodes along the common hepatic artery or celiac axis are usually considered a contraindication for resection[7]. Kitagawa et al[67] showed that, in 110 patients after resection of PHC, there was a 5-year survival rate of 31%, if the lymph nodes were negative. However, in patients suffering from a local or a para-aortic lymph node infiltration, the 5-year survival rates were 15% and 12%, respectively. Interestingly, in the same report, 12% of the patients with positive para-aortic lymph nodes who lived more than 5 years were found to have macroscopically negative nodes in surgery[67]. Although the routine lymph node dissection beyond the hepatoduodenal ligament is not generally recommended, several authors still believe that lymph node dissection is beneficial.

Due to the intimate relationship between the bile duct and vessels, PHC could usually infiltrate the portal vein and hepatic artery. The indication for portal vein resection (PVR) and reconstruction for PHC is controversial. Previously, tumors involving the portal vein were considered unresectable. However, more recently, several surgeons have advocated this approach and its clinical benefit has been validated in many studies[22,28,29,31,33,42,50]. de Jong et al[50] reported the results of the analysis of an international, multicenter database from seven major hepatobiliary centers. They found that the PVR for PHC was associated with a greater risk for 30-d and 90-d perioperative mortality. Nevertheless, they thought that PVR should be undertaken, when necessary, to extirpate all of the disease because of its association with long-term survival in several patients with PHC[50]. Similarly, Nanigo recommended that PVR should be performed only when the vessel adhered to and could not be freed from the tumor during the skeletonization resection of the hepatoduodenal ligament and that PVR should not be performed as a routine procedure because it lacked scientific validation[68]. Because of the short distance between the tumor and the portal vein, Neuhaus et al[22] proposed a “no-touch” concept in 1999 and recommended routine PVR to achieve a wider distal radicality. Additionally, Neuhaus et al[69] proposed a survey to compare the effect of the “no-touch” resection with the traditional curative resection. The 5-year survival rate was significantly higher in the “no-touch” group at 58% compared to 29% in the traditional curative resection group (P = 0.021). However, this new technique has not been accepted by many institutions because it lacks scientific validation and more random studies are warranted for additional investigation.

In earlier reports, few institutions proposed the surgical strategy of hepatic resection combined with hepatic artery resection in patients with advanced PHC. In small samples, the outcome and survival rates were disappointing. Therefore, many authors did not recommend this surgical strategy[43,62,70]. Shimizu et al[43] showed that all of the nine patients undergoing left-sided hepatectomy combined with hepatic artery resection lived less than 3 years, and they considered that the hepatic artery resection was a primary prognostic factor (RR = 3.063; 95%CI: 1.289-7.282). However, in 2010, the Nagoya Institute reported their experiences with major hepatectomies with simultaneous resections and reconstructions of the portal vein and hepatic artery; the investigators showed that the challenging surgery could be performed with an acceptable mortality rate of 2% and offered a better likelihood of long-term survival with a 5-year survival rate of 30%[71]. Currently, the number of patients undergoing hepatic artery resection has been increasing (Table 4). In the institute’s published data of 107 patients, the majority of patients (95%) underwent left-sided hepatectomies, of which 59% were left trisectionectomies and 36% were left hepatectomies. The overall mortality rate was 2.8% and the 5-year survival rate was 34.1%. The resected hepatic arteries were reconstructed primarily by end-to-end anastomosis, with an arterioportal shunt or an interposition graft using the radial artery or great saphenous vein[68]. For those patients who are unable to undergo hepatic artery reconstruction after resection, portal vein arterialization (PVA) could be a new approach. Using this method, adequate oxygen delivery to hepatocytes and biliary ducts can be assured. Moreover, several animal experiments showed that PVA could promote hepatic cell proliferation and enhance liver regeneration after extended hepatic resection[72]. The clinical significance of hepatic artery resection is debatable, yet also promising and encouraging.

Morbidity and mortality

In Table 3, we summarize the morbidity and mortality which show significant variations, ranging from 14% to 76% and from 0% to 18%, respectively. Sano et al[73] defined the complications as major when they resulted in organ failure or required another surgery or interventional radiology, such as liver failure, lung failure and renal failure[51,73]. Complications that were classified as minor include pleural effusion necessitating thoracocentesis, wound infection, intra-abdominal infection with positive culture of the drainage fluid, delayed gastric emptying, anastomotic leakage, clinically silent pancreatic fistula with amylase-rich serous fluid or contaminated fluid with positive culture, and bile leakage from the raw surface of the liver healing spontaneously or responding to conservative management[73]. The most common complications observed in most institutions were infective complications, especially during earlier years of the use of these procedures, representing 50% or more of the observed complications[3,15,36]. Nagiono et al[51] compared the complications between the earlier years and the more recent years, and they demonstrated that the incidence of grade C liver failure, which is clinically serious, decreased markedly from 18.2% from 1977 to 1990 to 3.2% from 2006 to 2010. Wound sepsis was the second most common complication, followed by intra-abdominal abscess and bile leakage[51].

The operative mortality included all in-hospital deaths as defined by Sano. All postoperative complications that affected the outcome or lengthened the hospital stay were considered. Death may be associated with acute liver failure after extended right hepatectomy and combined portal vein resection, and sepsis with multi-organ failure[45]. Overall, these extended liver as well as vascular resections were found to be significant predictors of increased mortality[23]. In addition to liver function, operative time and blood loss may be associated with mortality[51]. Several reports have demonstrated that preoperative portal vein embolization may decrease mortality even with extended hepatectomy[73].

Outcomes and recurrence

The average 5-year survival rates after resection for PHC range from 11% to 42% (Table 3). Factors associated with favorable outcome include the following: R0 resection, no lymph node metastasis, absence of perineural and perivascular invasion, and well-differentiated histological grade. Complete resection with negative histologic margins is the only modifiable factor and, for that reason, the primary aim of surgical therapy. Recently, several reports demonstrated that patients undergoing R1 resection (microscopically positive margin) had a longer overall survival rate than patients with unresectable PHC[36,74]. Moreover, patients undergoing R0 resections with a margin less than 5 mm had the same survival rate as those patients undergoing R1 resections[29]. The surgeons were encouraged to perform more aggressive surgery to achieve a better outcome.

Few studies have analyzed recurrence patterns and time to recurrence in patients with PHC. In several reports, tumor recurrence rates can be as high as 50% to 76%, and the median time to recurrence rates has been reported to be 12 to 43 mo[36,47,75,76]. The most common site of recurrence is a local site, followed by the liver, lymph node, peritoneum and other organs. Only histologic grade was associated with recurrence-free survival[47]. Generally, the patients with recurrent disease are not candidates for curative therapy and can only receive adjuvant therapy to improve long-term outcome.

ORTHOTOPIC LIVER TRANSPLANTATION

Theoretically, orthotopic liver transplantation (OLT) offers the advantage of the resection of all of the structures that may be affected by tumor, for example, the portal vein, bilateral hepatic ducts and atrophic liver lobes. Compared to surgical resection, OLT has several advantages: (1) patients with Bismuth IV type lesions and peripheral vascular lesions cannot undergo resection; (2) patients with PHC arising from primary sclerosing cholangitis (PSC) will tolerate resection poorly because of the underlying liver impairment; (3) dissection in the hepatic hilum has the potential for causing spillage, which is an adverse prognostic factor; and (4) a clear circumferential margin is usually not achievable, which might increase the recurrence rates of PHC[77]. However, in the early years of the application of this procedure, the results were disappointing. The Cincinnati Transplant Tumor Registry collected global data between 1968 and 1997. The 1-, 2-, and 5-year survival rates were 72%, 48%, and 23%, respectively. Eighty four percent of the patients had a recurrence within 2 years of transplantation[78]. This undesirable result may have been associated with the unselected patients who had distant metastasis. Despite this finding, PHC was considered to be a relative contraindication to OLT due to the lack of organs. Interestingly, several investigations found that those patients with negative margins in transplantation and the absence of regional lymph node metastases had a better survival rate. Moreover, 22% of the patients receiving radiotherapy and chemotherapy alone had a 5-year survival, which inspired several surgeons to explore a new OLT approach for PHC.

From 1987 to 2000, Miyazaki et al[70] collected 17 patients who were treated with systemic chemotherapy and intraluminal bile duct irradiation as they awaited liver transplantation. Eleven patients underwent liver transplantation, and until 2000, five patients were alive without evidence of tumor recurrence with a median follow-up of 7.5 years (range, 2.8-14.5 years). In 1994, the Mayo Clinic developed a protocol employing preoperative chemoradiation therapy followed by liver transplantation, which showed encouraging results. Currently, according to the Mayo Clinic protocol, patients receive EBRT (a target dose of 4500 cGy) with protracted venous infusion of 5-FU (225 mg/m² per day). Following this treatment, transcatheter iridium-192 brachytherapy (a target dose of 2000 cGy) is administered. Subsequently, the patients receive oral capecitabine (1000 mg/m² per day in two divided doses) until the time of OLT. Importantly, a staging laparotomy is performed on all of the patients before OLT to rule out metastatic disease. Only the patients with negative staging operations are eligible for transplantation[79]. Although there is a high dropout rate as patients await liver transplantation, the 5-year survival rate could achieve approximately 65% to 70%. However, the majority of patients undertaking OLT were diagnosed with PSC, and only 58% patients had histologically proven cancer which limited the use of OLT[80].

In 1996, Pichlmayr et al[81] proposed the indications for OLT in patients with PHC as follows: (1) unresectablity in presumed UICC stage II confirmed by laparotomy; (2) status postresection with the intention for R0 with R or R2 positive resection margins due to advanced central tumor infiltration; and (3) local intrahepatic recurrence. After additional exploration and analysis of PHC, the Mayo Clinic proposed their criteria for neoadjuvant therapy and liver transplantation[82] (Table 5). These types of patients would be excluded if they had the following: (1) intrahepatic cholangiocarcinoma; (2) uncontrolled infection; (3) prior radiation or chemotherapy; (4) prior biliary resection or attempted resection; (5) intrahepatic metastases; (6) evidence of extrahepatic disease; (7) history of other malignancy within 5 years; and (8) transperitoneal biopsy[82]. Although the Mayo Clinic protocol has been accepted in the majority of institutions, the role of OLT requires additional substantial evidence and data confirmation from multiple institutions.

Table 5 Criteria for neoadjuvant therapy and liver transplantation[82].

Diagnosis of cholangiocarcinoma

Transcatheter biopsy or brush cytology

CA-19.9 > 100 mg/mL and/or a mass on cross-sectional imaging with a malignant appearing stricture on cholangiography

Biliary ploidy by FISH with a malignant appearing stricture on cholangiography

Unresectable tumor above cystic duct

Pancreatoduodenectomy for microscopic involvement of the common bile duct

Resectable cholangiocarcinoma arising in PSC

Radial tumor diameter ≤ 3 cm

Absence of intra- and extrahepatic metastases

Candidate for liver transplantation

PSC: Primary sclerosing cholangitis.