Improved Outcome of Elderly Poor-Prognosis DLBCL Patients with 6xCHOP-14 and 8 Applications of Rituximab (R) Given Over An Extended Period: Results of the SMARTE-R-CHOP-14 Trial of the German High-Grade Non-Hodgkin Lymphoma Study Group (DSHNHL)

Abstract 592

Even though CHOP-14 significantly improved the outcome of elderly patients with DLBCL compared to CHOP-21 (Pfreundschuh et al., Blood 2004;104: 634–641), no superiority of dose-dense R-CHOP-14 compared to R-CHOP-21 was observed in two randomized trials (Delarue R et al., Blood 2009;114:169s; Cunningham et al., J Clin Oncol 2011;29:504s). Pharmacokinetics of R-CHOP-14 showed a plateau of R serum levels not until cycle 5 (Reiser et al., Blood 108, 778a, 2006). In order to achieve high R levels early and maintain R serum levels over a prolonged period of time with 8 R applications, in the SMARTE-R-CHOP-14 phase-II study 6 cycles of CHOP-14 chemotherapy were combined with 8 applications of rituximab (375 mg/m²), of which the first 3 applications were given in a dose-dense fashion followed by increasing intervals with the last 3 R administered after the completion of chemotherapy. 200 elderly pts (61 to 80 years) with DLBCL received 6 cycles of 2-weekly CHOP-14 combined with 8 × R on days −4, −1, 10, 29, 57, 99, 155, and 239. Prophylaxis with levofloxacin, acyclovir and cotrimoxazole was mandatory. The primary endpoint was event-free survival (EFS). 306 pts treated within the RICOVER-60 trial with 6xR-CHOP-14 + 2 R (with R given on days 1, 15, 29, 43, 57, 71, 85 and 99) served as control. 99/200 pts were evaluable for this planned interim analysis. Baseline characteristics were well balanced except more IPI high-risk patients in the SMARTE-R-CHOP-14 population (p=0.015) than in the RICOVER-60 population. Despite the worse prognostic profile of the SMARTE-R-CHOP-14 population, grade 3&4 infections decreased from 6.6% to 3.5% per cycle (p=0.009), and therapy-associated deaths from 5.6% to 3% (p=0.482) with intensified infection prophylaxis. Despite a less favorable study population, SMARTE-R-CHOP-14 resulted in a higher CR (84% vs. 78%) and a lower progression (5% vs. 7%) rate. More importantly, despite the worse prognostic profile of the SMART-E-R-CHOP-14 population 3-year outcome was similar to the RICOVER-60 population (EFS: 67.5% [57.9; 77.1] vs. 66.5% [60.9; 72.0]; PFS: 73.2% [64.2; 82.2] vs. 73.2% [64.2; 82.2]; OS: 81.4% [73.6; 89.2] vs. 78.1% [73.2; 83.0]). When the analysis was restricted to poor-prognosis patients (IPI=3-5), 3-year EFS of SMART-E-R-CHOP-14 patients (n=50) was 15.3% better (68.9% vs. 53.6%), PFS was 13.1% (73.0% vs. 59.9%) and OS was 10.8% (77.9% vs. 67.1%) better compared to 123 poor-prognosis RICOVER-60 patients. A pharmacokinetic analysis showed that with the 2-weekly application of rituximab maximal through levels were reached not before the 6^th application, while the SMARTE-R schedule was associated with an early maximal rituximab through level (after the 2^nd application) and a prolonged period of measurable R serum levels (420 vs. 320 days). We conclude that while kinetics of 8 applications of 2-weekly rituximab appears to be sufficient for optimal results in elderly patients with low tumor burden (and low IPI), the SMART-E-R schedule achieves superior results in patients with high tumor burden (and high IPI). This is probably due to the longer exposure of tumor cells to relevant rituximab serum levels with the SMART-E-R schedule. The results also indicate that a potential advantage of CHOP-14 over CHOP-21 chemotherapy might be compromised by the shorter exposure to rituximab when 8 applications are given every 2 weeks (R-CHOP-14: last application day 99) compared to 3 weeks (R-CHOP-21: last application day 148). Results of the completed SMART-E-R CHOP-14 trial with 200 patients and a detailed statistical analysis will be presented and discussed. Supported by Deutsche Krebshilfe and Roche