Abstract
Background/Aim: Intrahepatic cholangiocarcinoma (ICC) is the second most common primary liver cancer after hepatocellular carcinoma, and has a poor prognosis. Surgical resection is the only option for a cure of ICC. Here we attempted to define the cure rate after hepatic resection for ICC and to identify the predictors for a cure. Patients and Methods: Among the 96 patients who underwent R0 resections for primary ICC between 1990 and 2011 at the Kumamoto University Hospital and Kyushu University Hospital, those who were followed for ≥5 years after surgery were enrolled. “Cure” was defined as recurrence-free survival (RFS) of ≥5 years after surgery. Results: A total of 81 patients were eligible. A cure was achieved in 37 patients (45.7%). The 5-year overall survival and RFS rates were 55.0% and 41.7%, respectively. A multivariate logistic regression analysis identified the absence of lymph node metastasis (relative risk (RR) 7.5, p=0.011) and the absence of microvascular invasion (RR 5.5, p=0.0137) as the independent predictors of achieving a cure. Conclusion: R0 resections achieved a cure in 45.7% of this series of ICC patients. The predictors of a cure identified here, i.e., absence of lymph node metastasis and absence of microvascular invasion, could contribute to the selection of patients who are not candidates for adjuvant chemotherapy.
Intrahepatic cholangiocarcinoma (ICC) is a malignant tumor arising from the epithelial cells of the intrahepatic bile duct, and ICCs are located within the hepatic parenchyma. ICC is the second most common primary liver cancer after hepatocellular carcinoma, and ICCs account for 5-15% of all primary liver cancers in Japan (1-3). As ICC cases are always asymptomatic, only 20-40% of ICC patients are operable at the time of their presentation (4-6); ICC is thus one of the poor-prognosis cancers. Surgical resection is still the only option for long-term survival and a potential cure.
However, the surgical outcomes for ICC remain unsatisfactory with a 5-year survival rate of 23.6-35% (1, 4, 7, 8) and 58% even after R0 resection (1, 9). The rate of recurrence was also reported to be high, at 53-71% (4, 10, 11). The reported poor prognostic factors for ICC are lymph node metastasis, microvascular invasion, tumor size ≥5 cm, and multiple tumors (1, 7, 10). Data regarding the prediction of ICC recurrence could contribute to the decisions regarding treatment strategies. That is, knowledge of reliable predictors of recurrence could suggest indications for selecting patients who need adjuvant chemotherapy.
In this study, we investigated the cure rate after R0 resection for ICC, based on the 5-year recurrence-free survival (RFS), and we identified the predictors of achieving a cure.
Patients and Methods
Patients. From 1990 to 2011, we treated 121 patients for primary ICC at Kumamoto University Hospital and Kyushu University Hospital. After the exclusion of patients with R1/R2 resection (n=23) or operative mortality (n=2), 96 patients were enrolled in this study. After the exclusion of patients who are not completed 5-year follow-up (n=11) and those who deceased due to diseases other than ICC (n=4), 81 patients were followed for ≥5 years after surgery. “Cure” was defined as RFS of ≥5 years after surgery. The median follow-up time of our series was 45 months (range=0.43-197 months).
Surgical techniques. The details of our surgical techniques and the follow-up methods have been reported (1, 9, 12, 13). Major hepatic resection (HR) with bile duct resection was performed when bile duct invasion of ICC was suspected to have affected the first hepatic duct. Partial HR was performed in cases of peripheral ICC without bile duct invasion. When we considered that it would be better to confirm the negative surgical margin, the surgical stump was preserved for frozen pathology (1). As the right and left lobes of the liver have different routes of lymphatic drainage, the style of lymph node dissection was different according to the tumor location on the right or left lobe (12).
Survival curves related to OS and RFS. OS (a) and) RFS (b after surgery.
Ethical approval: All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. Informed consent was obtained from all individual participants included in the study.
Statistical analysis. Continuous variables are expressed as means±standard deviation (SD) and were compared using Student's t-test. Categorical variables were compared using either the χ2 test or Fisher's exact test, as appropriate. Grade II or higher complications according to the Clavien-Dindo classification were recorded as morbidity (14). The overall survival (OS) and RFS curves were generated by the Kaplan–Meier method and compared by the log-rank test. We subjected the variables that exhibited a probability (p) value <0.05 in a univariate analysis to a multivariate analysis using a multivariate logistic regression model. All analyses were performed using JMP® Pro 9.0.2 software (SAS, Cary, NC). p-Values <0.05 were considered significant.
Results
Patients' prognoses. A potential cure was achieved in 37 of the 81 patients (45.7%). The survival curves related to OS and RFS are illustrated in Figure 1. The 5-year OS rate for the complete series was 55.0%, and the median OS time was 5.5 years. For the complete series, the 5-year RFS rate was 41.7%, and the median RFS time was 2.8 years.
Patients' clinicopathological characteristics. The clinical characteristics of the cured (n=37) and non-cured (n=44) groups in our series are summarized in Table I. There were no significant differences in any variables between the two groups. The surgical factors of the two groups are summarized in Table II. Hepatic resection of two or more segments was performed in 54% of the cured group (20 patients) and 75% of the non-cured group (33 patients) (p=0.048). The median operation time was 402 (range=200-785 min) min in the cured group and significantly longer, i.e., 483 (range=142-755 min) min in the non-cured group (p=0.032).
Our comparison of tumor-related factors in the two groups is summarized in Table III. The median tumor size was 3.3 (range=1.1-12 cm) cm in the cured group and significantly higher at 4.5 (range=1.6-11 cm) cm in the non-cured group (p=0.014). Invasion of the serosa was recognized at a significantly higher rate in the non-cured group (63%; 27 patients) compared to the cure group (38%; 14 patients) (p=0.026). Microvascular invasion was recognized in 12 cure-group patients (34%) and 31 non-cured patients (70%) (p=0.0013). Pathological intrahepatic metastasis was identified in six patients (21%) of the cure group and 16 patients (46%) of the non-cured group (p=0.045). Lymph node metastasis was recognized in 11% (4 patients) of the cure group and 50% (17 patients) of the non-cured group (p=0.0093).
Univariate and multivariate analysis of predictors for a cure. The univariate analysis revealed the following as predictors of a cure: operation time <500 min. (p=0.0073), hepatic resection of <2 segments (p=0.048), tumor diameter <3 cm (p=0.0032), absence of invasion of the serosa (p=0.026), absence of microvascular invasion (p=0.0013), absence of pathologic intrahepatic metastasis (p=0.045) and absence of lymph node metastasis (p=0.0093) (Table IV). The multivariate logistic regression analysis identified the absence of lymph node metastasis (relative risk (RR) 7.5, p=0.0110) and the absence of microvascular invasion (RR 5.5, p=0.0137) as the independent predictors for achieving a cure (Table V).
Discussion
A cure was achieved in 45.7% of the 81 patients with ICC after R0 resection in the present series. This is higher than the previously reported cure rates of 27%-37% in ICC patients (10, 15, 16). This may be because surgical procedures and multidisciplinary therapy for ICC have progressed over the past decade, with the use of agents such as gemcitabine plus cisplatin or S-1. However, a standard therapy for recurrence or adjuvant chemotherapy against ICC has not been established (17-22).
Comparisons of clinical characteristics between the cured and non-cured groups.
Comparisons of surgical factors between the cured and non-cured groups.
ICC is a relatively rare disease, and R0 resections for ICC are also relatively rare. The limitations of this study include its retrospective nature and relatively small sample size, despite its collection of data from two high-volume liver cancer centers in Japan. Multi-institutional studies of larger numbers of patients are warranted to test our present findings.
We identified two independent predictors for a cure of ICC; the absence of lymph node metastasis and the absence of microvascular invasion. These factors were previously reported to be poor-prognosis factors in ICC patients who have undergone resections; that is, several studies reported lymph node metastasis (1, 7, 10, 15, 23), microvascular invasion (7, 10), tumor size ≥5 cm (10, 15, 23) and multiple tumors (7, 23) as poor-prognosis factors. In our present analysis too, only tumor-related factors and no clinical or surgical factors were identified as predictors of a cure.
Comparisons of tumor-related factors between the cured and non-cured groups.
Unfortunately, the absence of lymph node metastasis and the absence of microvascular invasion can be determined only after surgery, but these factors are still valuable when considering indications for adjuvant chemotherapy for ICC patients after R0 resections. For example, it was reported that the patients who showed the greatest benefit from adjuvant therapy were those with lymph node metastasis (24). That is, patients without factors such as lymph node metastasis or microvascular invasion need not receive adjuvant chemotherapy after R0 resections.
We also performed a receiver operating curve (ROC) analysis to determine the diagnostic accuracy of the identified risk factors for cure. Using the predictor of absence of lymph node metastasis, the sensitivity was 38.6% and the specificity was 89.2%. Using the predictor of absence of microvascular invasion, the sensitivity was 70.5% and the specificity was 65.7%. Thus, high specificity (89.2%) was observed for absence of lymph node metastasis as a predictor, whereas high sensitivity (70.5%) was observed for absence of microvascular invasion. In the future, when ICC recurrence is being treated by using drugs with a risk of severe adverse events, lymph node metastasis rather than microvascular invasion should be adopted as a factor when considering indications for adjuvant chemotherapy.
Moreover, with the use of both absence of lymph node metastasis and absence of microvascular invasion as predictors, the sensitivity was 60.0% and the specificity was 90.9%. We propose that for selecting contra-indications for adjuvant chemotherapy, high specificity is critical rather than high sensitivity. The indications for adjuvant chemotherapy can be accurately identified by using our two predictors of a cure. A model to predict a cure with higher accuracy along with advances in molecular targeted therapies and radiation techniques are needed (25).
In conclusion, we achieved a potential cure in 45.7% of 81 patients with ICC who underwent R0 resections, and identified the absence of lymph node metastasis and the absence of microvascular invasion as the independent predictors of a cure.
Univariate analysis of factors that predicted a cure.
Multivariate analysis of predictors for a cure.
Footnotes
Conflicts of Interest
Masayo Tsukamoto and other co-authors have no conflicts of interest concerning of this work.
- Received September 12, 2017.
- Revision received October 7, 2017.
- Accepted October 11, 2017.
- Copyright© 2017, International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved
