A Gene-ADR Network Based Pharmacovigilance Study Of Small Molecular Targeted Anti-tumor Drugs

With the development of biotechnology and increase of cancer incidence,molecular targeted agents have been applied in clinical therapy.Molecular targeted agent is a kind of drug that has higher specificity in tumor cells with little damage to normal cells.In recent years,several molecular targeted drugs are available in China and widely used in clinical therapy.Molecular targeted agents specifically restrain the growth of cancer cells by blocking the pathway needed for tumor growth and carcinogenesis,whose advantages lie in a relatively clear efficacy and lack of myelotoxicity as traditional chemotherapy.But molecular targeted agents have different adverse drug reactions such as allergic reaction,coagulation dysfunction,cutaneous reactions,.etc.and severe ADR also occurs.Currently the ADR study of molecular targeted agents mostly focused on case report,review and clinical trial results while there are still few research on ADR mechanism.Research has indicated that there is association between gene and ADRs.But preclinical study,clinical trials and post-market ADR surveillance cannot effectively and efficiently detect the relationship between ADR and specific genes,and molecular biologic or pharmacoepidemiologic study is needed to determine the association,which spends much labor and material.Presently some bioinformatics research teams are working on gene related ADRs,but they usually focused on database integration and establishment.Their study object is a large variety of drugs,so they pay little attention to a specific category of therapeutic drugs or a certain drug.Therefore,our research aims at the association between genes and molecular targeted drugs' ADR by bioinformatics methods.Objective: Due to lack of efficient and effective supporting methods to study gene-related ADRs,literature research,such methods as database searching and integrating and conditional probability algorithm are applied to analysis and evaluate the relationship between gene and ADRs.We conduct the research to find a new way for screening and mining gene-ADR relationships,thus providing a theoretical basis for further epidemologic and biologic study.Methods: Gene-drug reaction data and ADR information are collected and consolidated to form a gene-drug-ADR network.Probability value are calculated based on the network and four drugs with high gene-ADR associations are detected.Three of the four drugs(Lapatinib,Bosutinib & Vismodegib)are deep dived.Results: Totally 19 molecular targeted drugs are involved in the study to establish a gene-drug-ADR network and association of each gene-ADR pair is calculated.The method of reporting odds ratio(ROR)are used in two random sample drugs and the comparison of results of the two methods shows an overlap ratio of 60%.By screening the result of probability value several gene-ADR pairs with high association are found while related 4 drugs are emerged(Lapatinib,Bosutinib,Crizotinib and Vismodegib).1)Literature research indicate that Lapatinib-related genes express low when using lapatinib,which means a potential relationship with organ abnormality.2)Detected genes of Vismodegib gastrointestinal reactions do show a relationship with gastric acid secretion and gastric carcinogenesis.Conclusion: The research established a gene-drug-ADR network of molecular targeted anti-tumor drugs and association of each gene-ADR pair is valued.The validation of algorithm indicates a preferable overlap.Four drugs with high gene-ADR relationships are filtered and further study on three of the four drugs suggests our findings match the result of genomics research.This study is a primary research and gives a basic result and conclusion,but further biologic and epidemiologic research are required to validate.