Possible Cohort for Breast Cancer Prevention Trials Defined

NEW ORLEANS--Cytologic and biomarker assessment can be used to identifya cohort of women at extremely high risk for short-term breast cancer development,Carol Fabian, MD, said at the American Society of Preventive Oncology annualmeeting.

This, in turn, raises the possibility that such a cohort could be enrolledin breast cancer chemoprevention trials using biomarker changes as surrogateendpoints rather than development of cancer.

The researchers, from the University of Kansas Medical Center, KansasCity, previously reported that evidence obtained by random fine needleaspiration (FNA) of epithelial hyperplasia with atypia and biomarker abnormalitieswas more prevalent in women at increased epidemiologic risk for breastcancer than in low-risk controls.

The current study included 335 women (median age, 44) who were at highrisk for breast cancer because of a first-degree relative with breast cancer,prior precancerous biopsy, prior invasive cancer, or some multiple thereof.They underwent FNA and were followed for subsequent cancer developmentbetween 1989 and 1996. These results were compared with FNA findings andcancer development in a group of paid volunteers who were at low risk forbreast cancer.

The study found that cancer was more likely to develop in women withFNA evidence of hyperplasia with atypia. In turn, hyperplasia with atypiawas predicted by the presence of multiple biomarker abnormalities--DNAaneuploidy, overexpression of p53, epidermal growth factor receptor (EGFR),and Her-2/neu overexpression.

FNA evidence of hyperplasia with atypia was present in 18% of womenwith risk factors for breast cancer. An additional 13% had evidence ofhyperplasia without atypia but with multiple abnormal biomarkers. Thus,in a population at increased epidemiologic risk, largely because of positivefamily history (69% of study subjects), 31% had random FNA biomarkers associatedwith a marked short-term risk of breast cancer, Dr. Fabian said.

Significant differences were found in low-risk vs high-risk subjectsin the prevalence of normal, hyperplastic, or atypical pathology. And exceptfor ploidy, there was a stepwise increase in all biomarkers with the degreeof cytologic abnormality.

Twice the Expected Rate

At 34 months' follow-up, 10 women had developed ductal carcinoma insitu or invasive cancer; an additional four women developed lobular carcinomain situ. This is twice the expected rate, Dr. Fabian said.

Cancer developed in 5 of 40 subjects with both atypical hyperplasiaand multiple biomarker abnormalities; in 2 of 21 women with atypical hyperplasiaalone; in 2 of 70 with multiple biomarkers alone; and in only one of 203subjects with neither atypical hyperplasia nor multiple bio-marker abnormalities.

All of these cancer cases occurred in women in the high-risk group;however, Dr. Fabian pointed out that the low-risk volunteers were not followedwith the same intensity.

According to this analysis, 5% of high-risk women with atypical hyperplasiaor multiple biomarkers will develop breast cancer within three years, Dr.Fabian said. The strongest predictors of breast cancer were atypical hyperplasiaand multiple positive biomarkers, especially EGFR and p53. One third ofsubjects in the study had one or both of these findings.