Pharmaco-epidemiology--drugs, arthritis, and neoplasms: industry contribution to the data.

The real possibility that exogenous chemicals, including pharmaceuticals, increase risk of cancer is not a surprise. Epidemiologic systems to monitor exposure, as well as risk with or without such exposure, have been very hard to develop and very expensive to implement, and often have yielded ambiguous or questionably useful findings. The pharmaceutical industry is rising to the challenge of this situation, with attention to epidemiologic responsibilities and potential contributions deriving from them. This paper reviews the overall methodologic and public policy context surrounding pharmaco-epidemiology in the 1980s and then depicts an evolving program in pharmaco-epidemiology at one company, Burroughs Wellcome, as an example of one possible contributor. It discusses the company's epidemiologic approach to the question of association between treatment with disease-modifying antirheumatic drugs, one of which is manufactured by Burroughs Wellcome Company (Imuran brand azathioprine) and a possible increased risk of neoplasms in patients receiving these drugs as treatment for their rheumatoid arthritis. The paper will highlight specific epidemiologic applications of data bases derived from the published literature as well as voluntary reporting to industry. It will describe a proactive program in the development and conduct of epidemiologic studies to address these difficult problems.