Familial Aggregation: Sorting Susceptibility From Shared Environment

Extract

The contribution of family-based studies to elucidating the etiology of cancer has not been as widely appreciated as deserved. During the past decade, headlines have repeatedly trumpeted the discoveries of many cancer genes. In a number of cases, the discoveries have been translated to the clinic to improve cancer risk assessment and more precisely to devise potential interventions (1–5). In the laboratory, moreover, new avenues have been opened to characterizing the functional role of these genes (6). However, none of this would have been possible without the painstaking research over many years by hundreds of epidemiologists, clinicians, and genetic epidemiologists. They have labored to piece together the puzzle, study by study, on dozens of different cancers involving thousands of patients and their families. When viewed as a whole, this foundation of evidence provides the interpretive framework for the publicized breakthrough studies that capture media attention.

Looking back, it perhaps is not surprising that genetic studies of cancer patients were not a major research goal prior to 1990. While clinical geneticists had long known of and reported rare cancer family syndromes (7), the more common cancers appeared to be largely sporadic (nonfamilial). Epidemiologists searching for susceptibility factors would often choose the achievable and intuitive approach to test for familial aggregation by comparing family histories of case subjects with those of control subjects. Familial aggregation of cancer occurs when cancers cluster in families and one can statistically demonstrate that the risk of the disease in families is greater than the risk in the general population (8). The literature reports scores of studies observing familial aggregation of different cancers by this approach. Such evidence does not automatically mean that these cancers are due to genetic causes. Familial aggregation is an observation that has several interpretations: The cancers can be due to shared environmental exposure(s), shared susceptibility, or a combination of the two. Transmission of environmental factors within families can also mimic or confound true genetic susceptibilities (9). The challenge is to sort through the various hypotheses.