Gastrointestinal cancers in India: Treatment perspective

Metastatic gastrointestinal stromal tumor: A regional cancer center experience of 44 cases

BACKGROUND

Gastrointestinal stromal tumors (GISTs) are the most common mesenchymal neoplasms of the gastrointestinal tract. Historically, a poor prognosis for metastatic disease has been reported with systemic chemotherapy. Significant advances have been made in the last decade, since the introduction of different tyrosine kinase inhibitors (TKIs). Unfortunately, even though the TKIs have been used for a long time, there are very few published data of the experience of TKI therapy in metastatic GIST from India.

MATERIALS AND METHODS

Patients diagnosed with metastatic GIST from January 2005 to October 2016 at our center, who received first-line therapy with imatinib 400 mg/day, were reviewed retrospectively. Patients' profile, response to treatment, toxicity of TKI therapy, time to progression, and survival were evaluated.

RESULTS

Of the 44 metastatic GIST patients, 23 (52.2%) were males. Median age at diagnosis was 48 years. The most common presenting symptom was an abdominal pain (52%), followed by weight loss (23%). Most frequently affected metastatic site was liver (57%), followed by peritoneum (16%), and lungs (4.5%). Metastases to both liver and peritoneum were found in 10 patients (22.5%). All patients were initially treated with imatinib at a dose of 400 mg/day. Disease stabilization was documented in 21 cases (48%), and 13 patients (29%) achieved a partial response. TKI therapy was well-tolerated in most cases. Median progression-free survival (PFS) was 26 months, and estimated median survival was 48 months. Patients with lung metastases have a significantly inferior median PFS and overall survival, in comparison to patients with other metastatic sites (P < 0.05).

CONCLUSIONS

Imatinib therapy was well tolerated and induced a sustained clinical benefit in more than half of the patients with metastatic GIST. Lung metastases seemed to be a poor prognostic factor in this patient population.

GIDB: a knowledge database for the automated curation and multidimensional analysis of molecular signatures in gastrointestinal cancer

Gastrointestinal (GI) cancer is common, characterized by high mortality, and includes oesophagus, gastric, liver, bile duct, pancreas, rectal and colon cancers. The insufficient specificity and sensitivity of biomarkers is still a key clinical hindrance for GI cancer diagnosis and successful treatment. The emergence of `precision medicine', `basket trial' and `field cancerization' concepts calls for an urgent need and importance for the understanding of how organ system cancers occur at the molecular levels. Knowledge from both the literature and data available in public databases is informative in elucidating the molecular alterations underlying GI cancer. Currently, most available cancer databases have not offered a comprehensive discovery of gene-disease associations, molecular alterations and clinical information by integrated text mining and data mining in GI cancer. We develop GIDB, a panoptic knowledge database that attempts to automate the curation of molecular signatures using natural language processing approaches and multidimensional analyses. GIDB covers information on 8730 genes with both literature and data supporting evidence, 248 miRNAs, 58 lncRNAs, 320 copy number variations, 49 fusion genes and 2381 semantic networks. It presents a comprehensive database, not only in parallelizing supporting evidence and data integration for signatures associated with GI cancer but also in providing the timeline feature of major molecular discoveries. It highlights the most comprehensive overview, research hotspots and the development of historical knowledge of genes in GI cancer. Furthermore, GIDB characterizes genomic abnormalities in multilevel analysis, including simple somatic mutations, gene expression, DNA methylation and prognosis. GIDB offers a user-friendly interface and two customizable online tools (Heatmap and Network) for experimental researchers and clinicians to explore data and help them shorten the learning curve and broaden the scope of knowledge. More importantly, GIDB is an ongoing research project that will continue to be updated and improve the automated method for reducing manual work.

Identification of significant biomarkers and pathways associated with gastric carcinogenesis by whole genome-wide expression profiling analysis

The incidence of gastric cancer (GC) is extremely high in East Asia. GC is also one of the most common and lethal forms of cancer from a global perspective. However, to date, we have not been able to determine one or several genes as biomarkers in the diagnosis of GC and have also been unable to identify the genes which are important in the therapy of GC. In this study, we analyzed all genome-wide expression profiling arrays uploaded onto the Gene Expression Omnibus (GEO) database to filtrate the differentially expressed genes (DEGs) between normal stomach tissues and GC tissues. GSE13911, GSE19826 and GSE79973 were based on the GPL570 platform, and GSE29272 was based on the GPL96 platform. We screened out the DEGs from the two platforms and by selecting the intersection of these two platforms, we identified the common DEGs in the sequencing data from different laboratories. Finally, we obtained 3 upregulated and 34 downregulated DEGs in GC from 384 samples. As the number of downregulated DEGs was greater than that of the upregulated DEGs, functional analysis and pathway enrichment analysis were performed on the downregulated DEGs. Through our analysis, we identified the most significant genes associated with GC, such as secreted phosphoprotein 1 (SPP1), sulfatase 1 (SULF1), thrombospondin 2 (THBS2), ATPase H+/K+ transporting beta subunit (ATP4B), gastric intrinsic factor (GIF) and gastrokine 1 (GKN1). The prognostic power of these genes was corroborated in the Oncomine database and by Kaplan-Meier plotter (KM-plotter) analysis. Moreover, gastric acid secretion, collecting duct acid secretion, nitrogen metabolism and drug metabolism were significantly related to GC. Thus, these genes and pathways may be potential targets for improving the diagnosis and clinical effects in patients with GC.