Prognostic factors in breast cancer.

A variety of tumor characteristics can provide prognostic information useful in managing the patient with primary breast cancer. Some of these characteristics are firmly established, whereas others are observer dependent or require prospective validation. No single characteristic, however, is likely to fully define which patient with primary breast cancer is destined to relapse. This clinical dilemma--recognition of the high-risk patient--is particularly important in the management of women with node-negative breast cancer. Because most women with this early stage of disease will be cured by surgery alone, the use of adjuvant chemotherapy must be limited to high-risk subsets. Tumor size and ER status are established prognostic factors. Histologic and nuclear grade may be important, but problems of interobserver variability remain. Some studies have shown that aneuploidy or a high S-phase fraction may be independent, high-risk characteristics. Flow cytometric DNA content analysis must be applied with caution, however, because the calculation of S-phase fraction has not been standardized and because the prognostic utility of this approach has not been prospectively confirmed. For now, a prudent approach might be to gather as much prognostic information about each patient's tumor as possible. Those with several of the high-risk characteristics listed in Table 2 should receive strongest consideration for adjuvant treatment. Some of these same prognostic factors, along with several others, can be used to characterize the high-risk node-positive patient. The number of involved axillary nodes is the most important established predictor. Progesterone-receptor status is associated with both disease-free and overall survival, whereas ER status is independently related only to overall survival. Histologic grade, DNA ploidy, and S-phase fraction can also be used to help define the high-risk patient. Finally, tumors that amplify or overexpress the HER-2 gene may have a higher risk of relapse, although this finding has been questioned. Management of patients with breast cancer requires an individualized approach that is based on a careful weighing of a variety of prognostic considerations. The relative importance of these factors will require further large-scale, prospective, multiparameter studies. Although results from such studies are awaited, an understanding of the clinical heterogeneity of breast cancer must be based on a multiplicity of observations, each of which characterizes, in a limited way, the biology of this disease.