Evaluating drugs from cradle to grave--evolving systems for a complex activity.

Are medicines dangerous chemicals or life enhancing agents? The answer is ‘Both of those things, often at the same time, and worse in the wrong hands’. In this issue of the Annals, several authors address the problems besetting specialists who assess drug efficacy and safety. The last two articles describe activities (pharmacogenetics, and engineered drug delivery) which may produce unforeseen problems for drug evaluation. Some factors permeate the history of drug evaluation by physicians, scientists and government regulatory agencies. The first is the hypothetico-deductive method in measuring drug action. Most medicines were not truly effective before the highly selective drugs of the modern drug era, and there were few standards governing quality, safety, or efficacy. Only a drug-contamination mass poisoning in the 1920s provoked extensive quality and safety tests before the first administration to humans, and led to the US Food and Drugs Administration. The second feature is the unpredictability of some serious adverse effects. Systematic tests for drug toxicity merged only after the thalidomide catastrophe of the early 1960s. The third factor is proving drug efficacy in both animals and humans, to high scientific standards. Proof of principle in animal models cannot replace drug testing in humans. The need for rigorous proof produced elaborate guidelines and standards of investigation, e.g. the International Committee on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Use in Humans (ICH) Guidelines.