Résumé
Introduction: high dose chemotherapy (HDT) and autologous stem cell transplantation (ASCT) is an effective treatment in relapsed follicular lymphoma (FL). Alternative regimens including bendamustine based conditioning (BeEAM) had been developed to improve lymphoma outcome. Visani et al. showed in a phase I/II for relapsed aggressive lymphomas and relapsed Hodgkin lymphomas that bendamustine used at a dose of 200mg/m2, 2 days in combination with cytarabine (400 mg/m2, day-5 to day-2), etoposide (200 mg/m2, day-5 to day-2) and melphalan (140 mg/m2, day-1) was well tolerated with promising efficacy results. We designed a single arm, multicenter, phase II study for first and second chemosensitive relapses in FL patients (pts) evaluating this conditioning regimen.
Methods: The primary endpoint is the efficacy of BeEAM by measuring the 2-year event-free survival (2-year EFS) rate defines by relapse, progression, death from any cause and initiation of a new therapy. The hypothesis was the improvement of 2-year EFS rate from 70% (H0) to 85% (H1). Based on a Fleming-A'Hern single-stage design, 50 pts are needed. Secondary endpoints are: overall response rate (ORR) at day 100 according to 2007 Cheson criteria; progression-free survival (PFS); overall survival (OS) and safety profile of BeEAM. Main eligible criteria are: histologically confirmed FL relapsed (WHO grade 1, 2, 3a); aged 18-65 years; first or second chemosensitive relapses after salvage immunochemotherapy based on 2007 Cheson criteria with a complete (CR) or a partial response (PR) before the BeEAM; eligibility for ASCT; signed informed consent. This trial was funded by the French government PHRC program 2012.
Results: Twenty-one pts were included between July 2014 and November 2016 in 7 LYSA centers. One patient did not receive the treatment. The final analysis was based on 20 pts. Inclusion were performed for 16 (80%) and 3 (15%) pts at 1st and 2nd relapses, respectively. One patient (5%) was included at 3rd relapse and is considered as protocol minor deviation. The median time between the initial diagnostic and the 1st relapse was 3.6 years (0.7-12.9). In first-line therapy, 17/20 pts (85%) were treated with immunochemotherapy. At the relapse allowing inclusion, the FLIPI score were 0-1, 2, ≥3 for 6 (32%), 7 (36%), 6 (32%) pts, respectively. Salvage treatments were rituximab (R) associated with, dexamethasone, cytarabine, and platin in 19 pts and R-CHOP in one patient. The median age at inclusion was 57 years (range, 39-63); 16 (80%) and 4 (20%) pts were in CR and PR before BeEAM. After 18 included pts, the protocol was amended due to unexpected toxicities especially two hepatic veino-occlusive diseases (VOD), leading to the reduction of bendamustine dosage. Finally, 17 and 3 pts were respectively treated at 200mg/m2, 2 days and 160mg/m2, 2 days. The median time for neutrophil>0.5 G/L and platelet recoveries>20 G/L were 7.5 (0-11) and 7 (2-18) days. Patients received a median of 4 (0-12) packed red blood cell units and 6 (3-21) platelet units until J100. The median duration of hospital stay duration was 27 days (20-50). Response evaluation at J100 showed 19 CR pts and one patient in progressive disease. The ORR was 95.0% (95%CI, 75.1-99.9). Six pts progressed with two deaths of the FL (Figure 1A). With a median follow-up of 31 months (16-42), the 2-year EFS rate was 59.6% (95%CI, 35.1-77.4) but two events correspond to unplanned rituximab maintenance in two CR pts (Figure 1B). The PFS and OS rates were 69.6% (95%CI, 44.5-85.1) and 90% (95%CI, 65.6-97.4), respectively (Figure 1C). Fifteen pts (75%) had infectious complications until J100 that needed treatments for more than 7 days. Grade ≥3 toxicities were infections (N=18, 90%), gastrointestinal disorders (N=14, 70%), mucositis (N=12, 60%); atrial flutter (N=1, 5%); acute renal failure (N=2, 10%); HHV6 reactivation (N=3, 15%). In total, 29 serious adverse events (SAE) were declared for 16 pts. Two pts experienced VOD classified as unexpected SAE; one of them needed a hepatic transplantation. No patient presented toxic death.
Conclusions: the prospective evaluation of BeEAM conditioning in FL pts with a chemosensitive relapses showed excessive toxicities especially infectious complications, cases of VOD leading to premature termination of the trial. Based on the 20 treated pts, the outcome seems not improve with the use of BeEAM. Our study does not encourage the use of BeEAM conditioning in relapsed FL pts.