Response Re: Markers of DNA Repair and Susceptibility to Cancer in Humans: an Epidemiologic Review

Extract

We are pleased that Hemminki et al. have taken seriously the challenge in our recent review (1) to develop definitive molecular assays for DNA repair. Their conclusions echo ours: neither they nor we really know what the studies in lymphocytes measure. Clearly, their research publications in 1999 and 2000 will be valuable in developing the techniques for studying DNA repair.

We find it interesting that the two studies using kinetic tests on skin biopsies, which Hemminki et al. believe fulfill the criterion of biologic relevance, had positive and statistically significant results (although carried out in only a few subjects). Roth et al. (2) showed that melanoma subjects had lower DNA repair rates than control subjects (P < .001), and Alcalay et al. (3) showed that dimers in basal cell carcinoma subjects were eliminated less rapidly than in control subjects (P < .05).

The overall message that we derive from our analysis of the literature is that a considerable number of studies consistently found an association between biologic tests, whose significance is still obscure, and cancer at several sites. Looking at it this way, we agree with Hemminki and colleagues that further clarification of the biologic background is crucial. However, at this time, we think that focusing our attention on the design of the studies, reproducibility, sample size, and selection of control subjects was necessary to draw valid conclusions at the population level. The issue of biologic relevance of studies of DNA repair in lymphocytes is critical for epidemiologic studies where associations are derived from numerous subjects to define small reproducible alterations that may be important. Most of the work Hemminki et al. cite has been conducted in very small groups and has not yet been reproduced.