The pharmacovigilance symposium at the 37th annual meeting of the Japanese Society of Toxicology was a collaborative effort of the Pharmacovigilance Working Group (PVWG) and the Japanese Association of Pharmaceutical Medicine (JAPhMed) with key advice and contributions from expert colleagues from the PMDA and academia in addition to scientists and physicians from industry. The symposium follows a tradition established by the PVWG of mechanismbased examination of real examples of safety issues encountered by medical products in order to improve our knowledge of the predictability and preventability of adverse reactions with the ultimate goal of improving patient safety and the drug development process. The principles we try to follow are first to have clinical and non-clinical safety experts jointly evaluate safety data, increasing the communication and knowledge sharing between physicians and preclinical toxicologists. Second, we hope to increase the publication and open discussion of concrete examples of adverse reactions and safety issues. I believe that everyone involved in the process of drug development benefits when we can learn from examples of the investigation of adverse reactions from across the industry and that we have some ethical responsibilities to share safety data where possible because of the heavy responsibility we bear for engaging animals and people as experimental subjects in the process of drug discovery and development. Finally, we hope to encourage this sharing of expertise and concrete examples internationally because in today’s pharmaceutical industry a new chemical entity discovered in Japan could go on to pivotal clinical trials in the US and EU and ultimately be marketed in all countries from China to Brazil; drug safety issues can be newly discovered at any point in this international process.

　A major assumption when evaluating the toxicity of drugs or other chemicals is that the findings in animals predict what will happen in humans. This assumption has been in place for many years and more specifically the philosophy for evaluating the safety of drugs in animals before giving them to humans’ dates prior to the 1930s although this was voluntary as there were no laws in place to mandate this review. Following the death of many patients taking elixir of sulfanilamide contaminated with ethylene glycol, Leake (1929) suggested that competitive pressures and a general excitement for discovering new remedies resulted in an urgency where a “reliable scientific” evaluation of new drugs was often overlooked¹⁾. His proposal was to standardize the evaluation prior to initiating dosing in humans to include the following assessments:

• Minimal lethal dose

• Influence on isolated organs systems

• Mechanism of general systemic effects

• Fate in body and rate of destruction or elimination

• Acute and prolonged dosing

• Define therapeutic index

　From “general systemic studies” Leake proposed to evaluate “blood constituents, circulatory and respiratory activity, central nervous system and muscular response, glandular and alimentary action, kidney and sexual function and general metabolism”.

　The field of in silico toxicity prediction encompasses a broad range of techniques. Here, consideration is given to systems which are able to make predictions based on the structure of a chemical and, in particular, to the prediction of the mammalian toxicity of low- to medium-molecular weight organic chemicals in a pharmaceutical context.