Palmer, JR, EE Hatch, CL Rosenberg, P Hartge, RH Kaufman, L Titus-Ernsto, KL Noller, AL Herbst, RS Rao, R Troisi, T Colton and RN Hoover. 2002. Risk of breast cancer in women exposed to diethylstilbestrol in utero: preliminary results (United States). Cancer Causes and Control 13: 753–758.

Palmer et al. open up the next saga in the continuing story of diethylstilbestrol (DES): increased risk of breast cancer in adult women following their exposure in the womb.

The trends reported by Palmer et al. point to yet another health burden to be bourne by victims of the a drug industry eagerly pushing its products in the face of mounting evidence that not only did DES not achieve its intended effects, it caused harm. These findings also highlight the need for DES daughters to be especially vigorous in monitoring for breast cancer. for more on DES, see Chapter 4 in Our Stolen Future, or visit DES Stories, a book by Margaret Braun

Many women were exposed in the womb to DES during the 1950's and 1960's because the drug was recommended for control of difficult pregnancies. Other health effects have long since been established, most notably clear cell carcinoma. That cancer is a disease normally seen in older women, but rare even in that age group. Its appearance in teenage girls gave the first indications of DES's impacts.

No early-age association between breast cancer risk and DES exposure comparable to the link between clear cell carcinoma and DES emerged from studies of young adult DES daughters. With DES daughters now reaching the age at which breast cancer incidence rises in women in general, another question becomes possible: does DES increase the risk of breast cancer as DES daughters enter into the decades of life when breast cancer is most common?

What did they do? Palmer et al. assembled data on breast cancer incidence in a cohort of women exposed in utero to DES and now participating in a long-term health study. Their original enlistment in the study was unrelated to presence or absence of breast cancer. The exposed women (a total of 4281) were matched with unexposed women (a total of 2095) of the same age. Among them, 78 women developed breast cancer. One case was excluded because the medical records could not confirm the diagnosis. For most cases, data allowed classification of tumor type (estrogen receptor-positive or -negative) as well as histology, tumor size and nodal status.

Palmer et al. based their calculation of risk on person-years, or the number of years each woman was in the cohort through to one of four possible ending dates: (1) breast cancer diagnosis, (2) the last date of last-known check-up, (3) date of death (not from breast cancer) or (4) date of response to a questionnaire distributed in 1997.

In their statistical analysis, Palmer et al. considered a number of known factors affecting breast cancer risk, including age at menarche, age at first birth and number of births. They also examined the impact of several other potentially confounding variables, including family history of breast cancer, menopausal stauts, and use of hormone replacement therapy, but did not employ them in the final statistical analysis because they did not affect the outcome.

What did they find? Looking at all women in the study group combined, irrespective of age, there is no significant elevation in breast cancer risk associated with DES exposure. The risk ratio is 1.4 but the 95% confidence interval extends far beneath 1.0 (table below).

Separating the entire cohort by age, however, reveals a substantially different picture. For exposed women over 40, breast cancer risk is elevated compared to unexposed women, with a rate ratio of 2.5. The number of cancer cases in women above the age of 40 still remains small, and the trends hover at the edge of statistical significance.

DES-exposed women under the age of 40 do not appear to have increased risk of breast cancer compared to unexposed women of the same age. Hence the lack of significant risk elevation in all women combined appears to result from a dilution of the real effect in women over 40-yrs old by pooling them with younger women.

An analysis of the cumulative incidence of breast cancer in relation to age suggests what is happening: Beginning in their fifth decade of life (ages 40+), the data suggest that breast cancer risk rises faster in DES daughters than in unexposed women. adapted from Palmer et al. 2002

The analysis of tumor characteristics revealed suggestive but inconclusive patterns. For the entire cohort, risk was stronger for estrogen-positive tumors compared to estrogen-negative tumors (risk ratio of 1.9 vs. 0.4). The data also suggested a greater elevation in risk for tumors with no nodal involvement vs. metastatic disease (risk ratio of 3.6 vs. 0.8).

Information about the importance of timing of exposure during pregnancy was also inconclusive, but trends suggested that exposure during the first 8 weeks did not elevate risk whereas exposure subsequently did.

What does it mean? The overall implication of these results is that in utero exposure to a potent synthetic estrogen, diethystilbestrol, alters the developmental pathway of exposed women in ways that increase their vulnerability to breast cancer decades later in life. This adds to a body of emerging evidence indicating that to understand breast cancer, research must focus on exposures during critical windows of vulnerability in development, including in the womb. Studies that only ask about exposure levels at the time of diagnosis will provide misleading results, as for example, the 2002 Long Island Breast Cancer Study Project.

If the current trends in these data holds as the study cohort ages and moves further into the decades in life of higher breast cancer risk, the DES effect should become clearer and the significance level of the results should strengthen. This already appears to be happening: Palmer et al. compare their results with an earlier study of breast cancer risk in the same study cohort, published in 1998, which found a rate ratio of 1.2 (compared to 1.4. in the current study) based on roughly 50% fewer cases. They also note that the median age of women at the end of the current study was 43, and hence the incidence of breast cancer is still quite low within the study population.

On the basis of these results, the authors make two recommendations