Maintenance Rituximab or Observation after Frontline Treatment with Bendamustine-Rituximab (BR) for Follicular Lymphoma: A Real World Analysis Across 13 US Cancer Centers

INTRODUCTION: For follicular lymphoma (FL) patients (Pts) in need of therapy, frontline treatment with bendamustine-rituximab (BR) has become a standard treatment option. The prospective randomized PRIMA trial demonstrated that rituximab maintenance (RM) after response to treatment with R-CHOP is safe and improves progression-free survival (PFS) vs observation (Obs). However, there remains lack of consensus regarding the application of RM after front-line BR; there are no randomized clinical trials demonstrating its safety/efficacy. Recently-presented data from the GALLIUM study comparing outcomes of FL Pts treated with either BR or B in combination with obinutuzumab (G) followed by RM or G maintenance showed a fatal adverse event (AE) rate of 4.7% and 5.9%, respectively with a median follow-up of 41 months (Hiddemann, EHA 2017). The aim of this study was to assess the efficacy and safety of RM vs. Obs after frontline BR for FL.

METHODS: We conducted a comprehensive multi-institutional retrospective analysis of outcomes of FL Pts (grade I/II & IIIA) initiated on therapy with BR between 2011-2016 followed by either RM or Obs. The decision to apply RM was based on preferences of individual treating physician and/or patient. Individual patient records were queried for baseline demographic, disease characteristics, and treatment history. Outcomes included adverse events during or after BR induction, response rates, PFS, overall survival (OS) and cause of death (COD).

RESULTS: We collected data on 640 Pts across 13 US academic medical centers. Median age was 60 years (range 21-88). 52% were male and 87% were Caucasian (N = 397 with available data). FLIPI was low (20%), intermediate (38%) and high risk (41%) (N=545), 21% had B symptoms (N = 422) and 23%, had bulky disease (N=408). 91% had grade I/II FL and 9% had grade IIIA (N=632). The median time from diagnosis to treatment was 1 month (range 0 - 135 months). With a median follow-up of 36 months, 63 Pts died (9.8%): COD was lymphoma (n=26), infection or multi-organ system failure (MOSF) (n=12), unknown (n=12) solid tumor (n=10), MDS (N=1) cardiovascular (n=1) and progressive multifocal leukoencephalopathy (n=1), representing a fatal AE rate of 2.3% or 4.1% if cases of unknown COD are excluded or included in determining fatal AE rate, respectively. Median time to death was 22 months (range 2-74 months). Among the 584 Pts who received ≥4 cycles of BR, responses to induction therapy were available in 394 cases. RM was more commonly applied in responding Pts than Obs: in the 260 Pts selected for RM, the complete and partial remission (CR, PR) rates were 71% and 28%, respectively and in the 134 Pts selected for Obs, the CR and PR rates were 58% and 30%, respectively, (P <0.001). Median number of cycles was 6 and median B dose was 90 mg/m² and did not differ between RM and Obs Pts (P=0.14 for cycle, P=0.40 for dose). RM was administered q2 months (66%), q3 months (30%) or 4 doses q6 months (4%) for a mean of 18 months (range 3-24). The 3-year PFS for PR Pts was 80% and 44% for RM vs Obs, respectively (P=0.002). The 3-year PFS for CR Pts was 86% and 80% for RM vs. Obs, respectively (P=0.54). For Pts in PR/CR, multivariable analysis suggested that high FLIPI vs. low/intermediate was associated with worse PFS [hazard ratio (HR) 2.47, 95% confidence interval (95% C.I.) 1.43-4.28), P=0.001] and OS [HR 5.47 (95% CI 1.81-16.53), P=0.003]. RM was associated with improvement in PFS for Pts in PR [HR 0.36 (95% CI 0.18-0.71), P=0.003] but no benefit for Pts in CR [HR 0.80 (95% CI 0.38-1.66), P=0.55]. There was no association between use of RM and OS. Kaplan Meier estimates of PFS and OS (Figure) are depicted based on response (CR or PR) to BR induction.

CONCLUSIONS: In this large outcomes analysis of FL Pts treated with BR followed by RM, the fatal AE rate was comparable to that reported in GALLIUM when accounting for the real world nature of this unselected Pt cohort. RM was associated with a significant improvement in PFS vs. Obs in Pts who achieved PR after induction but no observed benefit for Pts in CR. In addition, we identified that the FLIPI was strongly prognostic for Pt survival for responding patients. Within the limitations inherent to retrospective analysis, these data suggest that FL Pts with partial response to BR induction therapy experience prolongation of PFS with RM, but consistent with the findings in the PRIMA study, there was no obvious impact on OS.

View largeDownload slide

View largeDownload slide

Close modal