Accurately distinguishing tumor cells from normal cells is a key issue in tumor diagnosis, evaluation, and treatment. Fluorescence-based immunohistochemistry as the standard method faces the inherent challenges of the heterogeneity of tumor cells and the lack of big data analysis of probing images. Here, we have demonstrated a machine learning-driven imaging method for rapid pathological diagnosis of five types of cancers (breast, colon, liver, lung, and stomach) using a perovskite nanocrystal probe. After conducting the bioanalysis of survivin expression in five different cancers, high-efficiency perovskite nanocrystal probes modified with the survivin antibody can recognize the cancer tissue section at the single cell level. The tumor to normal (T/N) ratio is 10.3-fold higher than that of a conventional fluorescent probe, which can successfully differentiate between tumors and adjacent normal tissues within 10 min. The features of the fluorescence intensity and pathological texture morphology have been extracted and analyzed from 1000 fluorescence images by machine learning. The final integrated decision model makes the area under the receiver operating characteristic curve (area under the curve) value of machine learning classification of breast, colon, liver, lung, and stomach above 90% while predicting the tumor organ of 92% of positive patients. This method demonstrates a high T/N ratio probe in the precise diagnosis of multiple cancers, which will be good for improving the accuracy of surgical resection and reducing cancer mortality.

Hepatic tuberculosis (HTB) is rare extrapulmonary tuberculosis that is clinically similar to liver malignancy, making it difficult for correct diagnoses. Pathology is the gold standard for tuberculosis diagnosis. However, there are few reports on the pathological features of HTB. A total of 32 HTB cases were considered and the differences in pathological features and drug resistance were analyzed and compared with those for pulmonary tuberculosis (PTB). Enhanced CT scans showed ring-shaped delayed enhancement during the arterial, venous, and delayed phases. Most HTB cases were single lesions, with the highest incidence in the right lobe, and the average lesion volume was smaller than that of PTB. The frequency of granuloma in pathological changes, the overall share of the lesion area in the HTB group, and the number of foxp3+ cells were significantly higher than in the PTB group. However, no statistically significant differences were observed between the two groups’ other pathological features and immune cell numbers. The immune microenvironment of the normal tissues surrounding the lesion was further analyzed. The findings showed that the number of macrophages and foxp3+ cells in the HTB group was significantly higher than in the PTB group. No significant difference in drug resistance was detected between the HTB and PTB groups. In conclusion, there are substantial differences in the characterization of pathological feature and immune microenvironment between HTB and PTB. The frequency of granuloma and subsequent overall share of the lesion area was significantly higher in HTB compared to PTB.

As a regulatory subunit of cyclin kinase, CKS1B promotes cancer development and is associated with poor prognosis in multiple cancer patients. However, the intrinsic role of CKS1B in pancreatic cancer remains elusive. In our research, CKS1B expression in pancreatic tumor tissue was higher than that in normal tissue by TCGA, Oncomine and CPTAC databases analysis. Similar result was verified in our center tissues by qRT-PCR. CKS1B expression was closely relevant to histologic grading, prognosis, and TMB. GSEA showed that CKS1B mainly participated in the regulation of autophagy and T cell receptor signaling pathway. Furthermore, CIBERSORT analysis showed that there was a strong correlation between CKS1B expression and tumor immune cells infiltration. Drug sensitivity analysis showed that patients with high CKS1B expression appeared to be more sensitive to gemcitabine, 5-fluorouracil, and paclitaxel. We then investigated cell viability and migratory ability by CCK8 and transwell assay, respectively. Results indicated that CKS1B knockdown by short hairpin RNA significantly reduced pancreatic cancer cell viability and invasion via regulating PD-L1 expression. In conclusion, our research further demonstrates the role of CKS1B in pancreatic cancer and the signaling pathways involved. The association of CKS1B with immune infiltration and immune checkpoint may provide a new direction for immunotherapy of pancreatic cancer.

Background Intrahepatic cholangiocarcinoma (ICC) is a fatal primary liver cancer, and its long-term survival rate remains poor. RNA-binding proteins (RBPs) play an important role in critical cellular processes, failure of any one or more processes can lead to the development of multiple cancers. This study aimed to explore pivotal biomarkers and corresponding mechanisms to predict the prognosis of patients with ICC. Methods The transcriptomic and clinical information of patients were collected from The Cancer Genome Atlas and Gene Expression Omnibus databases. Bioinformatic methods were used to identify survival-related and differentially-expressed biomarkers. Quantitative real-time PCR (qRT-PCR) and immunohistochemistry were used to detect the expression levels of key biomarkers in independent real-world cohorts. Subsequently, a prognostic signature was constructed that effectively distinguished patients in the high- and low-risk groups. Independent prognosis analysis was used to verify the signature’s independent predictive capabilities, and two nomograms were developed to predict survival. Results PIWIL4 and SUPT5H were identified and considered as pivotal biomarkers, and the same expression trends of upregulation in ICC were also validated via qRT-PCR and immunohistochemistry in the separate real-world sample cohorts. The prognostic signature showed good predictive capabilities according to the area under the curve. The correlation of the biomarkers with the tumour microenvironment suggested that the high riskScore was positively related to the enrichment of resting natural killer cells and activated memory CD4 + T cells. Conclusion In the present study, we demonstrated that PIWIL4 and SUPT5H could be used as novel prognostic biomarkers to develop a prognostic signature. This study provides potential biomarkers of prognostic value for patients with intrahepatic cholangiocarcinoma.