Purpose: The Nottingham Prognostic Index (NPI) is used to predict survival in patients with breast cancer. This index is based on tumor size, lymph node stage, and histological grade and allows the stratification of patients into three different prognostic groups. Our aim was to verify the effect of some prognostic variables on survival and to establish the independent influence of each of these variables by a survival regression analysis. We applied the NPI to the same group of patients to assess its predictive power and reproducibility.
Methods: We evaluated 311 women with breast cancer treated between January 1993 and December 1998.
Results: In a multivariate analysis (Cox proportional hazard model), only size, lymph node involvement, and histological grade were independent prognostic factors. The survival curves obtained after applying the NPI were similar to those for the factors with independent prognostic significance derived from our multivariate analysis.
Conclusion: The NPI allows us to accurately predict prognosis, and we advocate its standardized use.