Can we Interpret Analyses of Etiological Heterogeneity when Molecular Markers are Used to Classify Tumors?

Abstract Background: Analyses of etiological heterogeneity use molecular markers to sub-type cancers (e.g. breast cancer) and test whether an exposure is more strongly associated with one tumor sub-type as opposed to other sub-types. However, these molecular markers (e.g. a mutated oncogene) may best be conceptualized as mediators of exposure-disease relationships. Methods: Seven model causal scenarios, based on data from the literature on post-menopausal breast cancer, were created with varying relationships between an exposure (E), a mediating or non-mediating molecular marker (M) and a disease status (D), with M used to sub-type the disease. Using numerical examples, we assessed how well analyses of etiological heterogeneity identify the distinct etiological pathways in these seven scenarios. Results: etiological heterogeneity analyses of the data from the four causal mediation models produced similar results despite the mediating role of M differing across scenarios, with heterogeneity effect sizes ranging from 1.47 to 1.60. Analyses of data from the two scenarios in which M did not affect the risk of D, also produced evidence of etiological heterogeneity, with heterogeneity effect sizes of 1.53 and 2.66. The smallest heterogeneity effect size (1.33) was observed in the scenario where M was a sufficient cause of D. Conclusions: This methodological research indicates that analyses of etiological heterogeneity cannot distinguish between the modeled causal scenarios when a mediating or non-mediating molecular marker is used to sub-type tumors. Impact: We discuss these results in the context of the literature on obesity and etiological heterogeneity in breast tumors sub-typed by sex hormone receptor status.