Investigation of cancer case clusters: possibilities and limitations

Discovery of meaningful biologic events through investigation of individual cancer case clusters is limited by the greater likelihood that any given cluster, otherwise unspecified, will be a random event and by restrictions arising from the necessarily retrospective nature of any studies undertaken. In the face of long and variable latencies, it is difficult to document potential oncogenic exposures and to estimate dosage. Since case numbers are usually small, statistical assessments may have only limited value. This, however, encourages greater attention to be given to case features and clinical details from which patterns of case associations within clusters can occasionally be seen. When similar epidemiologic case patterns are repeatedly encountered in different cluster situations (as was apparent with parish and school associations, reviewed here, in several communities in the United States), some biologic basis for such case clustering may be suspected. Similar suspicions are aroused when additional cases of rare cancer are observed in community settings to which case clustering has already drawn attention.