A retrospective study was performed on a total of 106 patients diagnosed with ICC from January 2015 to December 2019 at the Mengchao Hepatobiliary Hospital of Fujian Medical University (Fuzhou, China). Inclusion criteria were as follows: (1) patients with radical resection of intrahepatic bile duct tumor in our hospital (R0), and complete resection of liver tumors, (2) patients with full records of clinicopathological data, (3) postoperatively histopathologically proven ICC, (3) no history of other anti-tumor therapies before surgery, (4) no history of other malignancies. Exclusion criteria included the following: (1) Patients died within 30 days after surgical operation. (2) Patients with mixed intrahepatocellular carcinoma diagnosed by pathology after operation. This study was approved by the Ethics Committee at Mengchao Hepatobiliary Hospital of Fujian Medical University (approval number: 2021-044-01), and all methods were performed in accordance with the relevant guidelines and regulations. The clinical information and characteristics were recorded and analyzed after written consent was obtained from the patients and their families.
Diagnosis and Treatment
After a baseline history and detailed physical examination, blood was obtained from patients in order to detect the serological indicators, including hepatitis B surface antigen, hepatitis B virus DNA level, anti-hepatitis C virus (HCV) antibody, serum albumin, total bilirubin, ALT, CA125, CA 19-9, and carcinoembryonic antigen (CEA). All patients were assessed with the contrast-enhanced CT or magnetic resonance, and positron emission tomography was chosen to determine whether there existed intrahepatic or extrahepatic metastases. According to tumor features and anatomy of the liver, partial liver resection combined with regional lymph node dissection in the hepatoduodenal ligament and retropancreatic and/or para-aortic lymph nodes was performed.
Postoperative pathological diagnosis and classification of ICC pathological subtypes
A serial slide was cut from representative formalin-fixed paraffin-embedded (FFPE) samples, stained with hematoxylin and eosin (H&E) and then observed under microscope by two independent pathologist. Tumor pathological indicators were recorded, including tumor size, number, tumor envelope, gross classification (mass type, peritubular infiltration type, and intratubular growth type), tumor differentiation (poorly differentiated, moderately differentiated, and well-differentiated), extrahepatic metastasis, bile duct invasion, lymph node metastasis and liver cirrhosis. According to histologic appearance and immunohistochemical index, ICCs was sub-classified into large duct and small duct types [12, 13].
KRAS and IDH1/2 gene mutation detection
Since IDH1/2 and KRAS mutations were reported to be the most frequent genetic alterations according to several large-scale genomic analyses, tumor tissues were macro-dissected from FFPE tissue blocks. Then, the total DNA of ICC tissues was extracted by a commercial kit (DNB400-50RXN, Merck, Germany) according to manufacturer’s instructions. This study focus on the mutation hot spot of exon 4 at codon 132 of IDH1, exon 4 at codon 172 of IDH2, and exon 2 of KRAS by polymerase chain reaction (PCR). The primer pairs were showed as follows: IDH1-R132: F: 5’-GATGAGAAGAGGGTTGAGGAGTT-3' and R: 5’-TACCTTGCTTAATGGGTGTAGATAC-3’. IDH2-R172: F: 5’-AGCTGAAGAAGATGTGGAAAAGTC-3’ and R: 5’-TTTGGGGTGAAGACCATTTTG-3’. KRAS: F: 5’-ACGTCTGCAGTCAACTGGAATT-3’ and R: 5’-TCTGTATCAAAGAATGGTCCTGC-3’. The PCR productions were analyzed by 3730xl DNA Analyzer (Applied Biosystems; USA).
ICC patients were followed up every 3 months after surgery. At each post-surgery visit， physical examinations were carried out. Blood was collected to detect serum level of CA19-9, CA125, CEA and contrast-enhanced CT or magnetic resonance imaging was also performed. When the patients were suspected of tumor recurrence or metastasis, the contrast-enhanced CT or magnetic resonance could be performed earlier. In this study, the endpoint of follow-up was defined as tumor recurrence confirmed by two radiologic images or death caused by ICC. Overall survival (OS) was defined as the interval between the date of surgery and death or the last date of follow-up. Relapse-free survival (RFS) was defined as the time from the date of surgery to the time of recurrence or death.
Categorical variables were displayed as mean±SD, and categorical variables were expressed as frequency. All recorded variables associated with prognosis were firstly analyzed by univariate Cox regression analysis. Subsequently, the potential variables those are potentially associated with prognosis at a significant level by univariate Cox regression analysis were further enrolled in multivariate Cox regression analysis to verify the independent risk factors. Nomogram was plotted based on these independent differential factors by using the package of rms in R version 4.0.3 (http://www.r-project.org/). The performance of the nomogram was measured by concordance index (C-index) and assessed by 1 000 bootstrap samples to compare nomogram-predicted versus observed Kaplan-Meier estimates of survival probability. Then, Receiver operating characteristic (ROC) curve analysis was used for comparison between our nomogram and other models on the basis of C-index. Survival curves were calculated using the Kaplan-Meier method and compared using the log-rank test. P＜ 0.05 was considered statistically significant.