Search

Background CNS recurrence of aggressive lymphoma remains a distressing and usually incurable event. In an analysis of patients on protocols of the German High-Grade Non-Hodgkin’s Lymphoma Study Group (DSHNHL) between 1990 and 2000, the CNS relapse rate was 2.2% in 1693 patients treated with modern chemotherapy but without rituximab (Boehme et al., Ann. Oncol. 2007). We now analyzed the incidence and prognostic factors for CNS recurrence in the RICOVER-60 study (Pfreundschuh et al., Blood abstract 2006) where patients had been randomized to 6 or 8 courses of CHOP-14 +/- rituximab. Patients and Methods From 2000–2005, 1242 patients (pts.) between 61 and 80 years with CD20+ aggressive lymphoma were randomized to receive 6 or 8 cycles of CHOP-14 chemotherapy with or without 8 courses of rituximab. Twenty patients were excluded because of missing or retracted informed consent, leaving 1222 pts. for the intention-to-treat analysis. For patients with suspected or confirmed lymphoma manifestation in testes, bone marrow, upper cervical nodes, sinuses or other extranodal involvements in the cranial region the protocol asked for intrathecal (i.th.) prophylaxis with 15 mg methotrexate twice in the first two cycles of treatment. Results 1217 patients were evaluable, of whom 58 patients (4.8%) after a median time of 8 months (1–39) developed relapse or progression to the CNS with a median time of survival of 3 months (0.1–38). Multivariate Cox regresssion analysis identified involvement of more than one extranodal site (RR = 3.4; p< 0.001), the presence of B-symptoms (RR = 1.9; p = 0.025) and increased LDH values (RR = 1.5; p = 0.146) as relevant predictors of CNS recurrence. Patients with these three characteristics had a CNS relapse rate of approximately 24% at 2 years, about 6-fold the incidence rate observed in all other patients. The addition of rituximab to chemotherapy reduced the risk of CNS recurrence (RR = 0.5; p = 0.025), whereas the number of treatment cycles (6 vs. 8) did have no influence on the CNS-specific outcome. 273 of 1217 patients (22%) received i.th. prophylaxis at least during one cycle and 202 of 273 prophylaxis patients (74%) were treated intrathecally in full compliance with the protocol, thus reducing the rate of CNS complications (RR = 0.3; p = 0.023). The majority of these patients had involvement of testes, bone marrow, sinuses or upper skeletal and/or nodal sites. Conclusion The incidence of CNS relapse in 1217 patients treated for aggressive lymphoma with CHOP-14 with or without rituximab was low (4.8%) but slightly higher than reported in other recent series. The prognostic factors for CNS-disease and the poor prognosis of CNS relapse remain largely unchanged although rituximab significantly reduced the incidence of CNS disease (3.6 vs. 5.9%) and the risk of CNS recurrence (RR 0.5; p = 0.025). Prophylactic i.th. therapy given to patients with distinct extranodal involvement reduces the risk of CNS disease in these individuals. Due to the low overall incidence general prophylactic strategies cannot be recommended.

Interval reduction from 3 (CHOP-21) to 2 weeks (CHOP-14; Pfreundschuh et al., Blood, 2004) and the addition of rituximab to CHOP-21 (R-CHOP-21; Coiffier et al., NEJM, 2002) improved outcome in elderly patients with DLBCL to a similar extent without increasing toxicity compared to CHOP-21. In the RICOVER-60 trial, elderly patients (61–80 years, stages I–IV) were randomized to receive 6 or 8 cycles of CHOP-14 with or without 8 applications of rituximab given on days 1, 15, 29, 43, 57, 71, 85, and 99. Radiotherapy was planned to sites of initial bulk and/or extranodal involvement. The primary endpoint was freedom from treatment failure (FFTF) with events defined as additional therapy, failure to achieve complete remission, progressive disease, relapse, or death. The trial was powered to show a 9% difference in FFTF rate after 3 years. Between 07/2000 and 06/2005, 1330 patients were recruited. A planned interim analysis was performed on 828 evaluable patients with CD20+ DLBCL (median age 68 years; IPI=1: 32%, IPI=2: 29%; IPI=3: 23%; IPI=4,5: 16%). As by intention to treat, there was no difference in FFTF between 6 (n=414) and 8 (n=413) cycles (p=0.23), but FFTF after R-CHOP-14 (n=414) was significantly better than after CHOP-14 (n=413) alone (p=0.000025). As the empirical p-value of the log rank test statistics for FFTF was considerably lower than the critical value for the interim analysis (pcrit=0.031), the formal criterion for stopping the trial according to the alpha spending function (O’ Brien and Fleming boundary) was met and the trial was stopped on June 17, 2005 with 50/1330 patients still under therapy. After a median observation time of 26 months, there was a trend for a better FFTF after 8 cycles of CHOP-14 (n=210) compared to 6 cycles (n=203; 58% vs. 53%; p=0.13), but this trend was neutralized after the addition of rituximab: 70% FFTF for both 6 (n=211) and 8 cycles R-CHOP-14 (n=203). The advantage of R-CHOP-14 over CHOP-14 with respect to overall survival after 26 months is not yet significant (74% vs. 78%; p=0.13). Excluding patients with stage I, the RICOVER-60 population is very similar to the one included in the GELA 98.5 trial; however, the projected 2.5-year survival rate for elderly stage II–IV patients after 6 x R-CHOP-14 (74%) compares favorably with 8 x R-CHOP-21 in the GELA trial (64%; Feugier et al. JCO 2005). The superiority of 6 x R-CHOP-14 over 8 x R-CHOP-21 is mostly due to the better 2.5-year survival of poor-prognosis patients (IPI=3,4,5: 64% in the RICOVER trial vs. 54% in the GELA trial). In conclusion, the results observed with 6 cycles of R-CHOP-14 in this largest randomized trial of DLBLC performed to date are the best ever reported for elderly patients with DLBCL. 6 x R-CHOP-14 should be considered as reference standard in future trials for elderly patients with DLBCL.