The natural history of breast cancer indicates that it develops over years-even decades-and may progress through recognizable stages of proliferative breast disease. Although breast cancers are slow growing, they metastasize early. The ability to recognize and treat high-risk, precursor breast lesions is therefore desirable.

Many human carcinomas evolve via a sequence of changes from benign hyperplasia through atypical hyperplasia to carcinoma in situ and eventually to fully malignant invasive tumors with the potential to metastasize. In the case of colorectal neoplasia, the well recognized polyp/carcinoma sequence is associated with a series of specific genetic alterations (Fearon and Vogelstein, 1990). Although the colon cancer model has become a paradigm for soUd tumor development, definition of critical genetic events in breast cancer has been hampered by its often lengthy natural history and architectural complexity. In the human breast, a spectrum of microscopic changes has been termed proliferative breast disease (PBD).

The progression of histopathological features of PBD has been correlated with increased risk for the development of invasive carcinoma (Page and Dupont, 1990), but the focal and microscopic lesions of PBD provide scant tissue for genetic or other biological analyses. Although hyperplastic lesions are observed in human breast (Wellingsetal., 1975; Dupont and Page, 1985; Dupont etal., 1993), their role in disease progression is not understood.

Patients with the most severe form of PBD, atypical hyperplasia, do have a four-to fivefold increased relative risk of developing breast cancer (Page and Dupont, 1990; Palh etal., 1991; London etal., 1992; Dupont etal., 1993), but it has been unclear whether these lesions are precursors of cancer or simply markers of breasts likely to give rise to independent neoplastic lesions. Carcinoma in situ (CIS), with an associated tenfold increased risk for breast cancer, is accepted as a precursor lesion because subsequent development of invasive cancer is frequently in the same breast in which CIS was detected earlier. In addition, risk factors for CIS and invasive cancer are nearly identical (Kerlikowske et al., 1997).

On the other hand, the risk for cancer in patients with biopsies diagnosed as atypical hyperplasia is nearly equal in the contralateral as in the ipsilateral breast. Thus, it has been suggested that atypical hyperplasia may reflect a field effect or condition rather than being a precursor lesion. However, occurrence of cancer in the ipsilateral breast is slightly higher, and the mean time to occurrence is lower in the breast in which a proliferative lesion is detected compared to the contralateral breast. For example, a study of 116 benign breast disease biopsy patients who subsequently developed invasive breast cancer during a 25-year follow-up indicated that 56% of breast cancers occurred in the ipsilateral breast and 44% occurred in the contralateral breast and the mean time till occurrence of cancer following biopsy was 11 years for the ipsilateral breast and 14 years for the contralateral breast cancers (Krieger and Hiatt, 1992).

These data suggest that some proliferative breast disease (PBD) lesions are precursors because if all hyperplastic lesions were simply identifying a field condition, incidence and time till occurrence should be equal in contralateral and ipsilateral breasts.

Other benign lesions, not part of the proliferative, hyperplastic sequence, have also been found to be associated with increased risk for later development of breast cancer. Fibroadenomas, particularly ones associated with cysts, sclerosing adenosis, papillary apocrine changes, or epithelial calcification are associated with elevated risk for developing breast cancer (Dupont et al., 1994).

Remarkably, the increased risk for cancer in fibroadenoma patients remains raised for decades, whereas the risk of developing invasive cancer declines with time following diagnosis of atypical hyperplasia such that after 10 years there is no increased risk (Dupont and Page, 1989). The apparent transient increased risk may also be due to the existence of multiple types of hyperplasia, only some of which are precursor lesions and progress within the 10-year period.

It is possible that the true precursor hyperplastic lesions can be identified. For example, 17% of patients having atypical hyperplasia with sclerosing adenosis subsequently develop cancer versus the 4% of cases without sclerosing adenosis that do so (Tavassoli and Norris, 1990); moreover, invasive cancer that develops in women previously having atypical hyperplasia within papillomas is nearly always ipsilateral to the biopsy site (Page et al., 1996).

Fred Raymond Miller
Advances in Oncobiology