Estimation of prognosis of the individual with primary breast cancer and its applications.

Abstract

Three patients with 2c mb reast lumps which are diagnosed as carcinomas may have hugely differing prognoses: one may die of breast cancer within two years of diagnosis with liver metastases; one may live 10 years without recurrence and then be diagnosed with bone metastases and one may be cured of breast cancer by th ep rimary surger ya lone .C learly al lt hree require different management programmes. Only with accurate prediction of the clinical course are the clinicians able to apply the optimal treatments for each. Ag reat many prognostic factors in breast cancer have been described in the medical literature but very few when placed in multivariate analysis return independent significance. Tumour size is well established as is lymph node (LN) stage; that tumour grade is an equally strong factor is often overlooked (1). Size and LN Stage are Time Dependent factors and are closely related. Grade is af actor of Tumour Biology determining the natural aggressiveness of the tumour, and this factor does not change with time during the growth and metastatic spread of ab reast cancer. The majority of the factors which have been described are biological factors and grade has its powerful effect because it is the common endpoint of these. COMBINING PROGNOSTIC FACTORS The best estimate of prognosis is obtained by combining the factors of the two different sets. The Nottingham Prognostic Index (NPI) was derived by a multivariate Cox analysis of 9p otential prognostic factors (2). The three factors returning independent significance were Size and LN Stage and Grade. The analysis allocated values for the relative contribution of each factor to th ee stimatio no fs urviva la nd thereby allowed af ormula to be calculated which has re