Progression of disease within 2 years (POD24) is a clinically significant endpoint to identify follicular lymphoma patients with high risk of death

Background: Follicular lymphoma (FL) is an indolent non-Hodgkin's lymphoma with heterogeneous outcomes among patients. Casulo et al (JCO 2015) showed that progression of disease within 2 years (POD24) after diagnosis for FL patients treated by R-CHOP was associated with poor outcomes, needing further validations before using it as a standard endpoint to evaluate treatment efficacy. We investigated the POD24 predictive value for all patients treated or not with R-CHOP in our institution (Nantes Medical University, France). Methods: Patients with grade 1, 2 or 3a FL treated from 2007 were registered in our local database (approved by French authorities, CNIL) and included in the present retrospective monocentric study, with up-dating of patient's outcomes. FL diagnosis was performed by local pathologist experts (members of the national LYSA-pathologist group, France). Results: Between 2007 and 2016, 317 patients with confirmed FL were included. At diagnosis: 24 did not received any treatment (Wait and watch), 259 were treated with Rituximab (R) (including R alone in 98 cases), 143 received an anthracycline-containing regimen (mainly R-CHOP like), 5 received bendamustine-containing regimen, radiotherapy alone in 11 cases and another chemotherapy regimen in 36 cases (mainly R-COP). Second line treatment (N = 151) consisted of chemotherapy in 91 cases, R alone in 37 cases. After first line therapy, 61 patients relapsed or died within 2y (POD24+), 99 patients after 2y, including 21 transformations, and 154 patients did not progress or die (missing = 3). At the time of the present analysis, the median follow-up is 5y. Median PFS is 58.2 months. OS at 1y, 3y, and 5y are 98.4% [97.0-99.8], 95.1% [92.6-97.6] and 92.5% [89.3-95.9] respectively. The 5y OS was statistically worst for POD24+ patients (82 % [71.9-93.5]) than for POD24- patients (93.3% [88.98-97.8]) (p = 0.00001). Age at diagnosis (≥60), performance status (PS ≥ 1), FLIPI, FLIPI2 scores (high) and transformation are predictive of OS in univariate analysis. PS (≥1) at diagnostic is predictive of POD24+. Conclusions: POD24 is predictive of a worse OS regardless of first line treatment nature and can be recommended as a relevant endpoint for clinical trials. Legal entity responsible for the study: CHU de Nantes. Funding: Has not received any funding. Disclosure: All authors have declared no conflicts of interest.