A statistical approach to an individualized prognostic index (IPI) for breast cancer survivability

The authors present 611 and 262 case histories of patients with breast cancer, studied 5 and 10 years after mastectomy, respectively; 27 clinical and 10 histologic parameters were considered for the statistical evaluation, in order to define an Individualized Prognostic Index (IPI) for breast cancer survivability. The probability of survival was estimated by a Bayesian formula using selected prognostic parameters, these parameters were placed in order of discriminant resolution and, for the calculation of the IPI, were selected according to their importance, as it follows: 5 years after surgery: percent affected nodules, dermal infiltration, TNM phase, Scarff-Bloom index and evolutive outbreak (PEV); 10 years after surgery: TNM phase, dermal infiltration, percent affected nodules and Scarff-Bloom index. The current information considers that out of several parameters, the selected prognostic parameters used for the IPI are sufficient to establish probability tests and a reliable estimation of life expectancy following breast cancer surgery.