Intrahepatic cholangiocarcinoma (ICC) is an aggressive malignancy, and there is a need for effective systemic therapies. Gene expression profile-based analyses may allow for efficient screening of potential drug candidates to serve as novel therapeutics for patients with ICC. The RNA expression profile of ICC and normal biliary epithelial cells were downloaded from the Gene Expression Omnibus (GEO) and The Cancer Genome Atlas (TCGA) databases. Function annotation and enrichment pathway analyses of the differentially expressed genes (DEGs) were finished using the Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) databases. A weighted gene co-expression network (WGCN) was constructed by WGCN analysis (WGCNA). Key genes from the DEGs and co-expression gene modules were analyzed to generate a protein–protein interaction (PPI) network. The association between the top 10 screened hub genes and the overall and disease-free survival of ICC patients was examined. The Connectivity Map (cMap) analysis was performed to identify possible drugs for ICC using hub genes. A total of 151 key genes were selected from the overlapping genes of 1287 GSE-DEGs, 8183 TCGA-DEGs and 1226 genes in the mixed modules. A total of 10 hub genes of interest (CTNNB1, SPP1, COL1A2, COL3A1, SMAD3, SRC, VCAN, PKLR, GART, MRPS5) were found analyzing protein–protein interaction. Using the cMap, candidate drugs screened with potential efficacy for ICC included three tyrosine kinase inhibitors (dasatinib, NVP-BHG712, tivantinib), two cannabinoid receptor agonists (palmitoylethanolamide, arachidonamide), two antibiotics (moxifloxacin, amoxicillin), one estrogen receptor agonist (levonorgestrel), one serine/threonine protein kinase inhibitor (MK-2206) and other small molecules. Key genes from network and PPI analysis allowed us to identify potential drugs for ICC. The identification of novel gene expression profiles and related drug screening may accelerate the identification of potential novel drug therapies for ICC.

Pancreatic ductal adenocarcinoma (PDAC) and distal cholangiocarcinoma (dCCA) are very aggressive tumors with a high mortality rate. Pancreas and distal bile ducts share a common embryonic development. Hence, PDAC and dCCA exhibit similar histological features that make a differential diagnosis during routine diagnostic practice challenging. However, there are also significant differences, with potential clinical implications. Even if PDAC and dCCA are generally associated with poor survival, patients with dCCA seem to present a better prognosis. Moreover, although precision oncology-based approaches are still limited in both entities, their most important targets are different and include alterations affecting BRCA1/2 and related genes in PDAC, as well as HER2 amplification in dCCA. Along this line, microsatellite instability represents a potential contact point in terms of tailored treatments, but its prevalence is very low in both tumor types. This review aims at defining the most important similarities and differences in terms of clinicopathological and molecular features between these two entities, also discussing the main theranostic implications derived from this challenging differential diagnosis.

Gallbladder cancer is often associated with poor prognosis, especially when patients experience early recurrence after surgery. Machine learning may improve prediction accuracy by analysing complex non-linear relationships. The aim of this study was to develop and evaluate a machine learning model to predict early recurrence risk after resection of gallbladder cancer. In this cross-sectional study, patients who underwent resection of gallbladder cancer with curative intent between 2001 and 2022 were identified using an international database. Patients were assigned randomly to a development and an evaluation cohort. Four machine learning models were trained to predict early recurrence (within 12 months) and compared using the area under the receiver operating curve (AUC). Among 374 patients, 56 (15.0%) experienced early recurrence; most patients had T1 (51, 13.6%) or T2 (180, 48.1%) disease, and a subset had lymph node metastasis (120, 32.1%). In multivariable Cox analysis, resection margins (HR 2.34, 95% c.i. 1.55 to 3.80; P < 0.001), and greater AJCC T (HR 2.14, 1.41 to 3.25; P < 0.001) and N (HR 1.59, 1.05 to 2.42; P = 0.029) categories were independent predictors of early recurrence. The random forest model demonstrated the highest discrimination in the evaluation cohort (AUC 76.4, 95% c.i. 66.3 to 86.5), compared with XGBoost (AUC 74.4, 53.4 to 85.3), support vector machine (AUC 67.2, 54.4 to 80.0), and logistic regression (AUC 73.1, 60.6 to 85.7), as well as good accuracy after bootstrapping validation (AUC 75.3, 75.0 to 75.6). Patients classified as being at high versus low risk of early recurrence had much worse overall survival (36.1 versus 63.8% respectively; P < 0.001). An easy-to-use calculator was made available (https://catalano-giovanni.shinyapps.io/GallbladderER). Machine learning-based prediction of early recurrence after resection of gallbladder cancer may help stratify patients, as well as help inform postoperative adjuvant therapy and surveillance strategies.