Options
2019
Journal Article
Titel
Analyses of frequency and sensitivity of affected organs in preclinical studies from RepDose and eTOX database
Titel Supplements
Abstract
Abstract
Question To better understand the predictivity of preclinical study outcomes to adverse effects in humans we analyzed the most prevalent adverse outcomes and the frequency of affected organs in preclinical studies from eTOX (http://etoxsys.com) and Toxtool. Toxtool links the study data of ToxRef DB (US EPA), Hess DB (NEDO) and RepDose DB (www.fraunhofer-repdose.de). While, Toxtool contains mainly chemicals, eTOX focus on drugs. The chemical domain of both dataset differ as indicated by principle component analysis. Methods ETOX comprises about 8000 repeated dose toxicity studies, which differ with regard to study duration, species, route and dose regimen; whereas Toxtool comprises about 6900 studies. Frequency and sensitivity of targets and organs in preclinical studies were compared between eTOX, and ToxTool. The frequency of organs in RepDose were identified under the condition of subacute, subchronic or chronic duration. Targets/organs affected with a frequency of at least 10% in Toxtool studies were compared to targets/organs observed in eTOX studies. Significant differences are indicated (chi2 test; p < 0.05). Results In subacute studies with oral exposure, the most frequently affected targets/organs are similar in both data sets at any dose, namely liver, clinical chemistry, clinical symptoms, hematology and kidney. Body weight changes belong to the most affected targets in Toxtool, but are less often observed in eTOX. The most sensitive frequently affected targets/organs are liver, clinical symptoms, kidney, hematology and clinical chemistry. Significant differences are observed comparing targets/organs from inhalation and oral studies. Liver, clinical symptoms, kidney, hematology respiratory tract, lung, nose, body weight and clinical chemistry were the most frequently observed targets in subacute inhalation studies, whereas not significant differences were observed comparing bolus and continuous application from oral studies. Conclusion Using high quality datasets is a critical step in the drug development chain. Promoting adequate data quality control, standardisation and annotation were the important challenges in this study. Main target organs did not differ between drugs and chemicals in subacute studies. For next approaches, we will evaluate the predictivity of apical findings in short term studies compared to long-term studies in preclinical studies.