Tuesday, November 25, 2008

Can some breast cancers just "go away"? Data mining says maybe, but it's complicated.

There's a paper this week in the Archives of Internal Medicine discussing the phenomena of some breast cancers possibly going away without treatment. As I do a lot of breast cancer related surgery, I know I'm going to get asked about this by a patient one of these days.

The paper is titled "The Natural History of Invasive Breast Cancers Detected by Screening Mammography" and can be read online here.

It opens with the observation that

...screening mammography has been associated with increased breast cancer incidence among women of screening age. If all of these newly detected cancers were destined to progress and become clinically evident as women age, a fall in incidence among older women should soon follow. The fact that this decrease is not evident raises the question: What is the natural history of these additional screen-detected cancers?

From autopsy studies of the elderly, we know we find many breast and prostate tumors which are clinically silent and that the patients died with rather then from. In an idealized world we could understand tumor biology enough that we could safely say some breast cancers could be watched, just as we already do with some prostate cancer.

This idea of "benign neglect" (no pun intended) for malignancies in regards to current standard treatments of surgery, chemotherapy, and radiation could potentially spare people significant morbidity and save the health system a great deal of money. One example of this would be the emerging idea that the drugs that block estrogen hormone metabolism (Arimidex) or estrogen receptors (Tamoxifen) may be just as effective as chemotherapy in post-menopausal women with estrogen receptor positive (ER+) tumors.

Now the study in question is taking some BIG leaps in logic making their conclusion. Much like financial analysts use "back casting" to test stock/bond buying strategies in the rear view mirror, these type of retrospective ideas can suffer from the fallacy of taking a result and looking back to make the data fit. This idea of watching these tumors would need to be done prospectively with very close followup. It would never be possible to do this trial in the United States due to internal review boards (IRB) and medical malpractice issues, but such an experiment might be possible in other countries (In the New York Times write up, Mexico is suggested for instance as a candidate. Gracias muchacho!)

Something to think about!



holt said...

from perusing the paper another angle that I have not heard discussed is that the patient population discussed in the paper is perimenopausal by age. They do not present break down in the differences in the pathology by hormone receptor status, or the patient's menopausal status. I have a suspicion that what they are seeing is a pseudotreatment effect of patients going through menopause with subclinical hormone sensitive tumors.

Dr. Rob Oliver Jr. said...

Leave it to my little brother to solve cancer! :)