Abstract
Abstract
Background: Follicular lymphoma (FL) is the second most common type of lymphoma in the United States. The clinical course is variable--some patients have an indolent course while others experience relatively aggressive disease, transformation to higher grade, and short survival. Despite a wide range of treatment options, no consensus exists concerning optimal therapy. Clinical predictors of outcome such as the FL International Prognostic Index (FLIPI) exist; however, biologic risk factors are needed to assist in providing accurate risk stratification. Recent data have suggested that the anti-tumor immune response may affect the clinical course and survival. In this study we sought to identify biologic indicators of prognosis in FL patients using a tissue microarray.
Methods: Biopsies from 94 newly diagnosed FL patients who presented between 1985 and 2002 were reviewed for cytologic grade (CG)(Mann-Berard criteria, WHO classification) and presence of diffuse areas (DAs) prior to placement in a tissue microarray (two 1.5 mm cores/case). Cases with diffuse large B-cell components were not included. Immune response was assessed with immunostains for CD3 and CD25 as indices of T-cell infiltration and activation, respectively, and CD68 and CD163 to evaluate macrophage infiltration. Lymphoma cells were evaluated for bcl-2 expression, as well as CD10 and MUM1 expression to assess germinal center phenotype. Initial and subsequent patient management was individualized and varied considerably. Clinical data were obtained for FLIPI and overall survival (OS). Statistical analysis was performed using P ≤ 0.10 as significant for univariable and P ≤ 0.05 for multivariable analysis.
Results: The patients consisted of 50 males (53.2%) and 44 females (46.8%), with a median age at diagnosis of 58 (range 24–89) years. The FLIPI distribution was 34 low risk (0–1), 26 intermediate risk (2), and 34 high risk (>2). There were 31 deaths. The median survival was 174.1 months and the median follow-up among surviving patients was 68.8 months (range 19.5–196.3). Cox proportional hazards analysis for risk of death showed no association with CG, DAs, or expression of bcl-2, CD3 (intrafollicular), CD25, CD68, or CD163. A high FLIPI score and lack of CD10 expression were both of borderline significance for higher risk of death (HR 2.06, 95% CI 0.86–4.92; P = 0.10 and HR 2.8, 95% CI 0.81–9.65; P = 0.10, respectively). MUM1 was expressed in 23 of 87 evaluable cases and was also associated with higher risk of death (HR 2.14, 95% CI 0.97–4.71; P = 0.059). Multivariable analysis showed that only MUM1 expression was associated with a higher risk of death (HR 2.30, 95% CI 1.04–5.12; P = 0.04).
Conclusions: These data demonstrate that expression of MUM1 in FL cells, consistent with a late germinal center/post germinal center phenotype, identifies a group of FL patients with poor prognosis. We found no correlation between survival and host immune response, using individual analysis of immunohistochemical markers of T-cell infiltration/activation (CD3 and CD25) or macrophage infiltration (CD68 and CD163). More detailed analysis of molecules involved in the transition of centrocyte to post-germinal center B-cells in FL is warranted.