Identification of Combined Clinical and Biomarker Prognostic Factors in Follicular Lymphoma.

Follicular lymphoma (FL) is generally an indolent disease, progressing slowly over a period of many years. However, some patients experience a rapid progression or histological transformation to an aggressive lymphoma requiring more aggressive treatment. Identification of these patients at the time of the initial diagnosis would allow more informed decisions to be made regarding clinical management. We investigated whether the expression of specific proteins in FL cells in combination with baseline clinical data could predict for relapse, transformation and survival of patients with follicular lymphoma. A tissue microarray (TMA) was generated of 67 cases of FL diagnosed between 1974 and 2003. Clinical baseline and follow-up data were obtained (median of 58 months). Forty-one patients experienced relapse within the follow-up period and thirty-one individuals transformed to aggressive lymphoma. Median overall survival was 13.6 years. Immunohistochemistry (IHC) staining was obtained for markers routinely used in lymphoma diagnosis, as well as oncogenesis-relevant proteins including p53, bcl-2, bcl-6, mum1, p16 and p65. Evaluations of the presence of benign T-cells, infiltrating macrophages, and the follicular dendritic cell network were also made. Univariate analysis was undertaken using the Kaplan-Meier method in order to assess the prognostic power of various clinical, histological and IHC variables with respect to overall survival, time to transformation and time to relapse. Variables that had prognostic significance (p<0.05) in univariate analysis were included in multivariate models of these end-points. High tumour grade, expression in tumour cells of p16 and lack of bcl-2 predicted shorter transformation-free survival in multivariate analysis. Advanced tumour stage, presence of B symptoms, expression of p53 or p16 in tumour cells, presence of benign T-cells in the tumour environment and integrity of the benign follicular dendritic cell network had independent prognostic power for shorter relapse-free survival. Greater than 5 nodal sites of disease, presence of B symptoms at diagnosis and expression of p53 in lymphoma cells each emerged as significant and independent predictors of overall survival using a Cox regression model. These results indicate that protein-based expression profiling using TMAs in combination with clinical data is a potentially productive means of identifying a prognostic index that can be implemented in the clinical management of follicular lymphoma patients. Further study is required to compare these results to existing clinical prognostic scores.