ADRIS – The Adverse Drug Reactions Information Scheme Hubert Hug, Dario Bagatto, Robert Dannecker, Richard Schindler, Oliver Horlacher, and Joseph Gut Under the Adverse Drug Reactions Information Scheme (ADRIS) data and knowledge relevant to the etiology of adverse drug reactions (ADRs) such as chemical structure of parent compounds, metabolites, covalent adducts, nucleic acid and protein sequences, protein structures, pharmaco-, toxico- and enzyme kinetics, pharmaco- and toxicodynamics, protein interactions, molecular pathways and complexes, as well as toxicological and clinical outcomes, are collected and logically and semantically related. ADRIS reflects the ontological prerequisite for the creation of databases and knowledge discovery systems for the abstraction and visualization of theragenomic concepts. A final outcome is the prediction of ADRs based on a profound knowledge of drug function and the molecular basics for personalized drug safety and eventually, personalized medicine. Pharmacogenetics 13:767–772 & 2003 Lippincott Williams & Wilkins Pharmacogenetics 2003, 13:767–772 Keywords: adverse drug reactions (ADRs), ADRIS, database, knowledge management, personalized drug safety, theragenomics TheraSTrat AG, Gewerbestrasse 25, 4123 Allschwil, Switzerland. Present address of Oliver Horlacher: Proteome Works Limited, 7/10 Tagalad Road, Mission Bay, Auckland, New Zealand. Corresponding author: Joseph Gut, TheraSTrat AG, Gewerbestrasse 25, 4123 Allschwil, Switzerland. Tel: +41 61 48550 12; fax: +41 61 48550 29; e-mail: joseph.gut@therastrat.com Received 7 May 2003 Accepted 20 September 2003 Introduction The person-to-person variability of a drug response is a major problem in clinical practice and in drug develop- ment [1]. It can lead to both adverse drug reactions (ADRs) of drugs or to therapeutic failure in individual patients or in sub-populations of patients. The occur- rence of serious or fatal ADRs has been extensively analyzed in hospital inpatients [2]. A meta-analysis of about 40 prospective studies from hospitals in the USA suggests that 6 to 7% of inpatients suffer from serious ADRs and 0.32% of patients have fatal ADRs. This results in about 100 000 deaths per year in the USA. This figure makes fatal ADRs between the fourth and sixth leading cause of death in hospital inpatients. ADRs can be classified in terms of their clinical, pharmacological, and chemical characteristics. From a clinical point of view, ADRs may comprise type A and type B reactions. Type A or augmented reactions can be predicted from the known pharmacology and often represent an exaggeration of the pharmacological ef- fects of the drug. These reactions are usually dose dependent and may be reversed by dose reduction. Examples of type A reactions include hypotension with antihypertensives and hemorrhage with anticoagulants. Type B or bizarre (idiosyncratic) reactions cannot be predicted from a knowledge of the basic pharmacology of the drug and have no simple dose response relation- ship, that is, there is a lack of correlation between dose and risk of toxicity. Host-dependent factors seem to be important in predisposition to these reactions, although for most forms of idiosyncratic drug reactions, these factors have not been elucidated. Such reactions are thought to have both metabolic and immunological components that may determine individual susceptibil- ity. These reactions tend to be serious and account for many drug-induced deaths. Examples of type B reac- tions include halothane-induced hepatitis [3] and anti- convulsant hypersensitivity [4]. From a chemical (i.e., structural) point of view, ADRs may comprise type C and type D reactions. Type C or chemical reactions are those reactions whose biological characteristics can be either rationalized or even pre- dicted based on the chemical structure of the parent drug, or of reactive intermediates and metabolites thereof. A classic example of a type C reaction is the hepatotoxicity caused by high doses of acetaminophen. Type D or delayed reactions may occur many years after treatment such as, for example, secondary tumors which appear years after treatment with chemothera- peutic agents. Also included in this category of ADRs are teratogenic effects seen in children after drug intake by the mother during pregnancy, such as, for example, in the fetal hydantoin syndrome with pheny- toin [5]. The collection of data related to ADRs is tedious. Relevant data and knowledge are dispersed in a variety of databases in non-compatible formats. Furthermore, a huge amount of new ADR data is created in the post- genomic area, for example single nucleotide poly- morphisms (SNPs), splice variants, gene expression, etc. Therefore, we developed the Adverse Drug Reac- Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited. Original article 767 0960-314X & 2003 Lippincott Williams & Wilkins DOI: 10.1097/01.fpc.0000054144.92680.d8