Is watchful waiting an option in breast cancer?

Some very interesting research was published this week in the Archives of Internal Medicine. Briefly, investigators in Norway looked at rates of breast cancer before and after mammography became a recommended screening test in 1996, and found higher rates in the regular mammography group. When you superimpose the rates of detection curves on each other (as was done in the New York Times graphic, from the article above) one can see that the rates are consistently lower in the earlier group. The authors argue that this is because there are indolent breast tumors which will never progress or may have even regressed during the period of observation. They go on to state that maybe watchful waiting, as we do for prostate cancer, may be a viable option for some women with breast cancer.

However, as I read the article, I believe there are several confounders which have not been adequately accounted for. For example, there was no way to account for the improvement in mammography technique and (importantly) interpretation by radiologists once mammography was done in a more regular fashion after 1996. In addition, women who are likely at higher risk for breast cancer (for whatever reason — genetics, past behavior, prior breast disease) were more likely to have obtained voluntary mammographies before 1996, meaning that that population was enriched with a group with a higher pretest probability of breast cancer. So, I don't buy their explanation, at least not yet. It's a very interesting finding, but I think there is more at play than just indolent or regressing breast tumors.