Background: Gastrointestinal cancers (GICs) continue to dominatein terms of both incidence and mortality worldwide. Dueto the absence of efficient and accurate prognostic biomarkers,the prognosis and treatment outcomes of many GICs are poor.Identifying biomarkers to predict individual clinical outcomesefficiently is a fundamental challenge in clinical oncology. Althoughseveral biomarkers have been continually discovered,their predictive accuracy is relatively modest, and their therapeuticuse is restricted. In light of this, the discovery of reliablebiomarkers for predicting prognosis and outcome in GIC is urgentlyrequired.Materials and Methods: We evaluated the HPA dataset andidentified
NPC۲ and
ITGAV as probable poor predictive genesfor these cancers. In addition, we used the GEPIA۲, cBio-Portal, UALCAN, LinkedOmics, STRING, Enrichr, TISDB,TIMER۲.۰, hTFTarget, miRTarBase, circBank, and DGIdb databasesto conduct a comprehensive and systematic analysis ofthe
NPC۲ and
ITGAV genes.Results: Our results found high expression levels of
NPC۲ and
ITGAV in most GICs. The aforementioned gene expressionswere linked to several clinicopathological characteristicsof GICs as well as poorer prognosis in LIHC and STAD. Themost common alteration type of
NPC۲ was amplification, andfor
ITGAV was deep deletion. Significant promotor hypermethylationwas also seen in
NPC۲ and
ITGAV in PAAD andCOAD, respectively. For the immunologic significance, NPC۲and
ITGAV were positively correlated with the abundance oftumor-infiltrating lymphocytes and macrophages. Furthermore,various immuno-modulators showed strong correlations withthe expression of these genes. There were currently ten smallmolecule drugs targeting ITGAV.Conclusion: Consequently, our bioinformatics analysis showedthat
NPC۲ and
ITGAV might be used as potential biomarkers todetermine the prognosis of various GICs and are also related toimmune infiltration