Comparison of clinicopathological characteristics between cirrhotic and non ‐ cirrhotic patients with intrahepatic cholangiocarcinoma : A large ‐ scale retrospective study

The effect of cirrhosis on the characteristics of intrahepatic cholangiocarcinoma (ICC) has not been fully elucidated. The purpose of this study was to investigate how cirrhosis affects the clinicopathological characteristics and survival of surgically treated ICC patients. A total of 1,312 ICC patients surgically treated between January 2007 and December 2011 at a single institution were retrospectively reviewed and the clinicopathological data were compared between cirrhotic and non-cirrhotic patients. Univariate and multivariate analyses were performed to identify significant and independent prognostic factors in this cohort. A total of 302 patients (23.0%) were cirrhotic. Compared with cirrhotic patients, the tumors in non-cirrhotic patients were usually larger, less differentiated, and more likely to have lymphatic metastasis, vascular and perineural invasion. Following resection, cirrhotic patients achieved a longer survival compared with non-cirrhotic patients (16.0 vs. 13.0 months, respectively; P<0.038). Multivariate analysis demonstrated that hepatitis B virus infection and cirrhosis were independent favorable prognostic factors, while the presence of cholelithiasis, elevated carbohydrate antigen 19-9 and carcinoembryonic antigen levels, multiple tumors, lymphatic metastasis, vascular invasion and positive surgical margin status were independent unfavorable prognostic factors. Overall, the clinicopathological characteristics of ICC patients with and without cirrhosis differed significantly. Compared with cirrhotic patients, in whom the biological behavior of ICC was similar to that of HCC, non-cirrhotic patients exhibited higher-risk pathological characteristics, lower curative resection rate and worse survival.


Introduction
Intrahepatic cholangiocarcinoma (ICC) is the second most common primary liver cancer after hepatocellular carcinoma (HCC) (1)(2)(3).ICC has been categorized as peripheral and perihilar types based on location (1,4), and as mass-forming, periductular infiltrating and intraductal growth types, based on the growth pattern classification of ICC by the Liver Cancer Study Group of Japan (5).An increasing number of studies suggest that surgical resection usually offers the possibility of long-term survival to patients with this disease (2,6,7).Although there has been a worldwide increase in the incidence and mortality of ICC in recent years (2,3), ICC has not been investigated as extensively as HCC (1).
Previous studies suggested that hepatitis B virus (HBV) infection and cirrhosis, which are well-documented pathogenic factors in the development of HCC (8)(9)(10)(11), may also be associated with an increased risk of ICC (12)(13)(14)(15).Cirrhosis is common among HCC patients (8)(9)(10), and has been proven to be a poor prognostic factor following surgical treatment of HCC (10,16,17).A significant proportion of ICC patients are also cirrhotic; however, the prognostic role of this finding has not been extensively investigated.Although HBV infection has been reported to be a favorable prognostic factor for ICC patients and the clinicopathological characteristics differ between patients with and those without HBV infection (13,18,19), the role of cirrhosis in the prognosis of ICC patients has not been fully elucidated due to the limited number of related studies.Cirrhosis has been found to be a favorable prognostic factor for ICC patients in our former study (20); however, the opposite result was reported by another previous study (21).The aim of the present study was to determine the effect of cirrhosis on the prognosis of ICC patients and the

Comparison of clinicopathological characteristics between cirrhotic and non-cirrhotic patients with intrahepatic cholangiocarcinoma: A large-scale retrospective study
mechanism underlying this effect through comparing clinicopathological characteristics and survival data in large series of ICC patients with and without cirrhosis.Pathological and immunohistochemical methods.All the resected and bioptic specimens were pathologically examined, including tumor size and number, capsule formation, LN metastasis, vascular invasion, perineural invasion and tumor cell differentiation.Each tumor was staged according to the 7th edition of the American Joint Committee on Cancer (AJCC) staging system for ICC (24).The surgical margins were examined for the presence of residual tumor and were classified according to the R classification as R0 (no residual tumor and resection margin >0 mm), R1 (microscopic residual tumor or null-margin resection) or R2 (macroscopic residual tumor) (25).Curative resection was defined a negative resection margin on histopathological examination.

Patients recruiting and grouping
Follow-up.All the patients were followed up postoperatively by X-ray of the chest, ultrasound scan of the liver, liver function tests and serum levels of carbohydrate antigen (CA) 19-9, carcinoembryonic antigen (CEA) and α-fetoprotein (AFP) at an interval of 1-3 months.When recurrence or metastasis were suspected, a CT or MRI scan was performed to confirm the diagnosis.Treatments for recurrent disease included surgery, transarterial chemoembolization, radiotherapy and supportive therapy.Survival was evaluated from the date of surgery; the patients were followed up for survival until death or until the study deadline date of September 30, 2014.
Statistical analysis.Continuous variables are presented as the mean ± standard deviation or as median values and range.Categorical variables are presented as total and percentage.Comparisons were performed using the unpaired t-test for continuous variables and the Chi-squared or Wilcoxon test for categorical variables.Overall survival (OS) rates were calculated using the Kaplan-Meier method.The statistically significant prognostic factors were analyzed by univariate analysis, evaluated using the Kaplan-Meier method and compared with the log-rank test.The multivariate analysis was performed using the Cox proportional hazards model to identify the independent prognostic factors for survival.Statistical analysis was performed using the SPSS 19.0 software for Windows (SPSS Inc., Chicago, IL).Differences with P-values of <0.05 were considered statistically significant.III).

Clinical
Survival of the entire cohort.The duration of survival was defined as the time from surgery to the date of death or the last follow-up, and the median follow-up period was 47 months (range, 1-93 months).The 1-, 3-and 5-year OS rates for the entire cohort were 57.0, 19.9 and 13.3%, respectively, with a median survival time (MST) of 14.0 months.
It is generally hypothesized that the prognosis of ICC is worse compared with that of HCC following surgical treatment (2,5,27).Complete surgical resection is the only curative treatment for ICC; however, similar to previous reports (2,6,7,25), the OS of this entire cohort indicated that the prognosis of ICC is dismal following surgical management, with an MST of only 14.0 months, as the disease is usually advanced at the time of diagnosis.There are several known factors affecting the prognosis of ICC after surgery, including surgical margin status, multiple tumors, LN metastasis and vascular invasion (2,6,7,25,28).In earlier studies on ICC, cirrhosis (21), unlike HBV infection (13,18,19), was shown to be an independent unfavorable prognostic factor for survival, but little is known on the underlying mechanism.However, the multivariate analysis in this series and our former study (20) revealed that cirrhosis was an independent favorable prognostic factor for survival of ICC patients following surgery.Different sample sizes may be the reason for the different results reported by these studies regarding the role of cirrhosis in ICC prognosis.
In the present study, all the patients with underlying liver disease had well-compensated liver function (Child A), which, in non-cirrhotic cases, should not significantly affect the extent of hepatectomy or postoperative morbidity and mortality.However, as reported by an earlier study on ICC (21), cirrhosis exerts a negative effect on major hepatectomy, and cirrhotic patients in our study tended to have a smaller resection range (Table III).Non-cirrhotic patients may be more amenable to resection due to a relatively better preserved liver function, as in HCC tumors (16,17).However, the R0 resection rate was significantly higher among cirrhotic patients (P<0.001), while the rates of non-curative resection [R1 (P=0.002) and R2 (P=0.027)] and exploratory laparotomy (P= 0.045) were significantly higher among non-cirrhotic patients, which may be one of the reasons for the superior survival of ICC patients with cirrhosis.ICCs in non-cirrhotic patients tended to be larger, with a lower incidence of capsule formation, and at a more advanced stage at diagnosis compared with those in cirrhotic patients (Table II), which may be the reason for the better surgical margin status and survival in ICC patients with cirrhosis.ICC patients with cirrhosis exhibited a significantly better survival compared with those without cirrhosis (Fig. 1), which may not be attributed to early tumor detection.Although our data demonstrated that patients with cirrhosis had significantly smaller tumors compared with those without cirrhosis (Table I), tumor size was not found to be an independent prognostic factor for ICC patients.In the  present study, the presence of cholelithiasis, HBV infection, elevated CA19-9 and CEA levels, surgical margin status and certain pathological characteristics, such as multiple tumors, capsule formation, lymphatic metastasis and vascular invasion, were significantly associated with the presence of cirrhosis, but the associations were not causative, and multivariate analysis demonstrated that these factors together with cirrhosis were all independent prognostic factors for ICC (Table IV).The differences in the clinicopathological characteristics between ICC patients with and those without cirrhosis may be due to different underlying pathogenic mechanisms in the two groups of patients.
According to previous studies, the development of HCC in cirrhotic and non-cirrhotic livers may be underlined by distinct mechanisms (8,9), which has not been proven in ICC patients.In the present study, in the clinical setting, ICC patients with cirrhosis exhibited different and unique characteristics compared with patients without cirrhosis.The findings of this study suggested that ICC associated with cirrhosis may display a biological behavior similar to that of HCC and, thus, have a better prognosis.Although the etiology of ICC remains unclear, there is growing evidence suggesting that ICC associated with cirrhosis may be derived from the same hepatic progenitor cells as HCC (13,14,18,19) and, thus, behaves more like HCC, which is generally considered to have a more favorable prognosis compared with ICC (2,5,27,29).The observations of the present study indicate that ICC occurring in patients with cirrhosis may share a common carcinogenic process with HCC.Compared with non-cirrhotic patients, cirrhotic ICC patients were more likely younger and male, a profile resembling that of HCC patients (18,19).The formation of vascular tumor thrombi, one of pathological characteristics of HCC, was observed more often among ICC patients with cirrhosis compared with those without cirrhosis.In contrast to a previous study (21), LN metastasis and perineural invasion, which are typical pathological characteristics of adenocarcinoma, were less often found in ICC patients with cirrhosis compared with those without cirrhosis.AFP is often used as a tumor marker for HCC and, in the present study, a significantly higher number of cirrhotic ICC patients exhibited elevated serum AFP levels compared with non-cirrhotic patients, suggesting that the ICC cells may exhibit hepatocellular differentiation.These findings also suggest that ICC with cirrhosis and HCC may share a common carcinogenic process.
The present study had several limitations.First, a small number of patients with mild fibrosis or steatosis were included, which may have affected the findings; however, none of these patients had true cirrhosis and, therefore, were considered eligible for inclusion in the cohort of non-cirrhotic patients.Patients with HCV infection, a known inciting factor of hepatocarcinogenesis (12), were not included in the present study due to the small case series of HCV infection.A number of patients received non-radical resection and a considerable percentage of non-anatomical hepatectomies were included in this study, due to the advanced tumor stage at the time of diagnosis and the high incidence of chronic liver disease, such as HBV infection and cirrhosis, prevalent in China.Furthermore, although it included the largest case series of ICC patients, this study was retrospective in nature, which may be associated with certain limitations with regards to data selection.
In conclusion, cirrhosis is an independent favorable prognostic factor for survival of ICC patients, due to the distinct biological characteristics as well as the different pathogenic mechanism in this subgroup of patients.More emphasis should be placed on aggressive surgical treatment for ICC patients with cirrhosis, considering safety and better survival in this group.Non-cirrhotic patients may lack the typical ʻfield-defectʼ of a cirrhotic liver; however, these patients may harbor a molecular field defect that differs from that of a cirrhotic liver, leading to higher-risk pathological characteristics, lower resection rates and worse survival.Further investigation should be focused on the genomic profile of livers with and without cirrhosis in order to elucidate the different pathogenic mechanisms underlying the development of ICC, in order to design novel targeted treatments to improve the survival of ICC patients.

Figure 1 .
Figure1.Overall survival (OS) in intrahepatic cholangiocarcinoma patients with and without cirrhosis: The 1-, 3-and 5-year OS rated for patients with cirrhosis were 62.3, 24.1 and 14.7%, respectively, which were significantly higher compared with the corresponding rates in patients without cirrhosis (55.4,18.7 and 13.0%, respectively), with median survival times of 16.0 vs. 13.0 months, respectively (P<0.038).

Table I .
Comparison of clinical characteristics between intrahepatic cholangiocarcinoma patients with and without cirrhosis.
(6),13,18,19)in status, were statistically significant prognostic factors affecting the survival of ICC patients (TableIV).Cox's regression multivariate analysis identified HBV infection and cirrhosis as independent favorable prognostic factors, while the presence of cholelithiasis, elevated CA19-9 and CEA levels, multiple tumors, lymphatic metastasis, vascular invasion and positive surgical margin status were independent unfavorable prognostic factors, with hazard ratios of 1.330, 1.726, 1.380, 1.297, 1.193 and 1.788, respectively (TableIV).infectionmayalsobeassociatedwiththe occurrence of ICC(12,13,18,19); however, the association between cirrhosis and the pathogenesis/prognosis of ICC remains unknown.The present study confirmed the earlier observation that cirrhosis is prevalent among patients with ICC in highly endemic areas(6), as it was observed in 23.0% of our patients, which is a markedly higher percentage compared with Western TableIII.Distribution of different types of liver resection in intrahepatic cholangiocarcinoma patients with and without cirrhosis.

Table II .
Comparison of pathological characteristics between intrahepatic cholangiocarcinoma patients with and without cirrhosis.

Table IV .
Univariate and multivariate analyses of variables associated with overall survival after surgery in 1,312 patients with intrahepatic cholangiocarcinoma.

Table IV .
Continued.