Plu-Bureau G, Bossard N, Thalabard J C
Médecine de la Reproduction, Hôpital Necker, 149 rue de Sèvres, 75743 Paris Cedex 15.
Rev Epidemiol Sante Publique. 2000 Jun;48(3):294-303.
The analysis of the interaction between environmental and genetic factors is a matter of increasing interest in cancerology. More particularly the discovery of the BRCAx family and the high cumulated incidence of familial breast cancers related to mutations of these proteins raised the issue of the differential effect of long term and/or early exposure to oral contraceptives in the presence of these mutations. The classical case-control design assumes the presence of a control group, which can be sometimes difficult to obtain from both the technical and ethical points of view. Case-only or case-case studies, which are based only on series of cases, making them apparently attractive, have been proposed to analyze more specifically the interaction term. The aim of the present paper is to review and discuss the methodological basis and main assumptions of the case-only design, and their applicability to breast cancer studies. The measure of the interaction between an environmental factor and a susceptibility genetic factor differs in an important aspect from the measure of the association between an environmental factor and a acquired tumoral genetic factor; this aspect is reminded.
