Prognostic ability of five clinical risk scores in follicular lymphoma: A single?center evaluation

Abstract

With the intention of identifying follicular lymphoma (FL) patients at higher risk of progression, early relapse (POD24), histological transformation (HT) or death, multiple risk scores (RS) have been proposed. However, it has not yet been established whether any of them globally outperforms the others. We evaluated the clinical utility and statistical performance of the five most widely used clinical scores (IPI, ILI, FLIPI, FLIPI2, PRIMA-PI) in a single-center series of 414 grade 1-3A FL patients diagnosed in the rituximab era. Overall concordance (proportion of patients allocated to the same risk category by all five RS) was 24%. FLIPI and FLIPI2 were predictive of time to first treatment. All five scores were predictive of response, POD24, progression-free, and OS, while only FLIPI predicted HT. IPI identified a small subset (7%) of truly high-risk patients (10-year OS of 16%). In subgroup analyses, we showed that ILI is useful in the prognostication of limited-disease patients, and PRIMA-PI is an age-independent score that can identify a high-risk subset of older patients. Performance metrics were slightly better for IPI in terms of calibration (Harrell's c-index 0.73), without major differences among RS regarding other parameters. Although the incorporation of molecular and imaging data will continue to refine the stratification of FL patients, FLIPI remains the most powerful clinical prognostic index in the rituximab era, predicting the greatest number of endpoints.