© 1982 Oxford University Press
research-article |
Drugs in the Aetiology of Cancer: A Retrospective Study


* Dr H M Anthony, University Department of Radiotherapy, University of Leeds, Cookridge Hospital, Leeds LS16 6QB.
St Helens
**G R Addleston, M J Berger, D E Bowden, E Breen (dec), I G C Brown, J M Brown, D B Cathcart, B Colville, A Cooke, T W Cookson, E Coupland, J I Currie, J H de Graeve, Y M Dias, M Dunsby, D G Eastham, J G Edgar, D Elliott, D A Furniss, J F Goodall, S C C Hamilton, P H Holmes, N Hurwitz, D M D Lambert, J S Lawson, L J Letty, J W Lewis, H Lindsay, D MacAdam, J M MacKinnon, J M McArdle, J S Merrell. R G Morey, L H Moss, D B Oughtibridge, J M Perkin, C W D Philips, R A Robinson, T Rose, A Rushton (dec), J Sagar, S H Samuel, J F Scarlett, K Scott, R O Selby, S Sheffrin, M W Sweet-Escott, G P Ward, S Watson.
Leeds
Anthony H M (University Department of Radiotherapy, Cookridge Hospital, Leeds LS 16 6QB, England), Kenny T E,et al. Drugs in the aetiology of cancer: a retrospective study. International Journal of Epidemiology 1982, 11; 336344.
A retrospective study of the history of drug use in cancer patients, conducted in general practice and involving 972patients, has suggested that concurrent treatment with barbiturates and benzodiazepines may involve a cancerrisk. Concurrent usage of other drugs with barbiturates was investigated because barbiturates have been reported to alter the handling of other drugs by liver microsomes and to interact with chemical carcinogens in experimental animal systems. If the risk has a basis in fact, it may lie behind the recent report of a small cancer risk associated with barbiturate treatment, in a prospective study based on computerized prescription information in a defined population, raising the hope that our finding can be tested in a larger study. A possible risk associated with prolonged treatment with cortisone related drugs could not be excluded.
Four per cent of the cancer patients had tumours at sites consistent with reported occupational hazards and afurther 4% suspected site-hazard associations. Occupation and smoking did not contribute to the drug associations noted.
Entry of cancer patients from general practice gave a true sample of cancer patients in the area, judged by thedistributions of age, sex, marital status and cancer site. Choice of controls from consultations for a new complaint resulted in a group showing deviation in the types of patient similar to those reported for consultations in general practice.
Received 22 March 1982