Thursday, December 14, 2006

Jane Zones, Board Member

Discontinuation of Postmenopausal Hormone Therapy and the Decline in New Cases of Breast Cancer

Dr. PM Ravdin (of the M.D. Anderson Cancer Center in Houston) and colleagues gave an interesting talk speculating on possible reasons for the 7% decline in breast cancer incidence (new cases) in SEER data between 2002 and 2003. Incidence had been relatively stable in the several years prior to this.

The downward trend was seen only in women over 50, particularly in estrogen receptor (ER) positive breast cancers. There was no change in incidence rates for cases that were ER negative.

Ravdin and his group posited that the best explanation for the decline was the large number of postmenopausal women who stopped taking hormone therapy (HT) after the Women’s Health Initiative reported their first results.  WHI found that use of Prempro (Premarin plus progestin) not only was not beneficial in preventing a number of age-related chronic conditions, but also was correlated with an increase in breast cancer incidence.

At the time of the announcement, in July 2002, an estimated 30% of older women were taking HT. Half of these women stopped the regimen in subsequent months. About half of breast cancers are hormone related. Voila! Thirty per cent halved is 15%, and halved again is 7%, an appealing explanation for the 7 % decline in incidence.

Clearly the effect would not be on the initiation of breast cancer, since the proposed effect happened quickly, and we know that breast cancer develops over a long period before detection. Rather, HT discontinuation would affect progression of already existing breast cancer.

Using a variety of sources of evidence, Ravdin, et al., discounted a number of possible explanations for declining breast cancer rates. Decreases in mammographic screening were too small (3% in women ages 50 to 65) to account for the large decrease in incidence. Likewise, few healthy postmenopausal women have decided to take tamoxifen (1%) or raloxifene (for osteoporosis prevention)—both of which have been shown to lower, but not eliminate, risk for developing breast cancer.

The Ravdin research group continues to monitor other alternative explanations for decreasing breast cancer incidence, which could be a product of a number of factors, including questionable reliability of SEER data. Meanwhile they eagerly await publication of SEER data for 2004 this coming April to see if breast cancer incidence continues to decline.

Two recent publications have supported the HT-breast cancer incidence association, one by SEER researchers (Cancer 107, October 2006, pp. 1711-42) and a letter by researchers using Kaiser hospital data from Northern California (Journal of Clinical Oncology 24 (33), November 20, 2006, pp. e49-e50).