Deep learning has been applied to many areas in health care, including imaging diagnosis, digital pathology, prediction of hospital admission, drug design, classification of cancer and stromal cells, doctor assistance, etc. Cancer prognosis is to estimate the fate of cancer, probabilities of cancer recurrence and progression, and to provide survival estimation to the patients. The accuracy of cancer prognosis prediction will greatly benefit clinical management of cancer patients. The improvement of biomedical translational research and the application of advanced statistical analysis and machine learning methods are the driving forces to improve cancer prognosis prediction. Recent years, there is a significant increase of computational power and rapid advancement in the technology of artificial intelligence, particularly in deep learning. In addition, the cost reduction in large scale next-generation sequencing, and the availability of such data through open source databases (e.g., TCGA and GEO databases) offer us opportunities to possibly build more powerful and accurate models to predict cancer prognosis more accurately. In this review, we reviewed the most recent published works that used deep learning to build models for cancer prognosis prediction. Deep learning has been suggested to be a more generic model, requires less data engineering, and achieves more accurate prediction when working with large amounts of data. The application of deep learning in cancer prognosis has been shown to be equivalent or better than current approaches, such as Cox-PH. With the burst of multi-omics data, including genomics data, transcriptomics data and clinical information in cancer studies, we believe that deep learning would potentially improve cancer prognosis.

Esophageal carcinoma (EC), consists of two histological types, esophageal squamous carcinoma (ESCC) and esophageal adenocarcinoma (EAC). EAC accounted for 10% of EC for centuries; however, the prevalence of EAC has alarmingly risen 6 times and increased to about 50% of EC in recent 30 years in the western countries, while treatment options for EAC patients are still limited. Stratification of molecular subtypes by gene expression profiling methods had offered opportunities for targeted therapies. However, the molecular subtype in EAC has not been defined. Hence, Identification of EAC molecular subtypes is needed and will provide important insights for future new therapies. We performed meta-analysis of gene expression profiling data on three independent EAC cohorts and showed that there are two common molecular subtypes in EAC. Each of the two EAC molecular subtypes has subtype specific expression patterns and mutation signatures. Genes which were over-expressed in subtype I EACs rather than subtype II EAC cases, were enriched in biological processes including epithelial cell differentiation, keratinocyte differentiation, and KEGG pathways including basal cell carcinoma. TP53 and CDKN2A are significantly mutated in both EAC subtypes. 24 genes including SMAD4 were found to be only significantly mutated in subtype I EAC cases, while 30 genes including ARID1A are only significantly mutated in subtype II EACs. Two EAC molecular subtypes were defined and validated. This finding may offer new opportunities for targeted therapies.