Should we replace dexamethasone, cytarabine, and cisplatin for relapsed lymphoma?

Abstract

With a 5-year overall survival of 53%, high-dose chemotherapy with autologous stem-cell transplantation (ASCT) served as the standard of care for relapsed aggressive diffuse large B-cell lymphoma (DLBCL) before the rituximab (R) era. The most important factor governing the success of this procedure is the response to salvage chemotherapy with DHAP (dexamethasone, cytarabine, and cisplatin) before ASCT. Adding R to second-line chemotherapy was expected to increase the efficacy of salvage followed by ASCT. Early results of several phase II studies of R-ICE (R, ifosfamide, etoposide, and carboplatin) suggested such an improvement. Then, a wellconducted randomized study assessing the addition of R to salvage combination chemotherapy demonstrated a significant improvement in progression-free survival (PFS) compared with regimens lacking R. However, only 5% of the patients received R as part of their firstline treatment. In addition to standard DHAP, several combinations of salvage chemotherapy have been used to increase the efficacy and/or reduce the substantial treatment-related toxicity of DHAP. Moreover, the identification of a salvage regimen that is less toxic and more accessible to elderly patients remains an unmet need. Curiously, no comparative studies were performed before the R era, and institutional tradition has served as the main guide for choosing a salvage regimen. Thanks to a worldwide academic collaborative effort, two randomized comparative studies for patients with relapsed aggressive lymphoma are now available. In the CORAL study (Collaborative Trial in Relapsed Aggressive Lymphoma), only patients with CD20 DLBCL experiencing their first relapse and/or refractory episode after upfront therapy were randomized between two regimens that included drugs with distinct modes of action: R-ICE (243 patients) and R-DHAP (234 patients). The patients who responded to these treatments proceeded to high-dose chemotherapy and ASCT using the same conditioning regimen, BEAM (carmustine, 1,3-bis(2-chloroethyl)-1-nitrosourea, etoposide, cytarabine, and melphalan). No difference was observed in the response rate after salvage chemotherapy and before transplantation; the response rate was 63% in the R-ICE group and 64% in the R-DHAP group, with 142 (58%) of responding patients presenting complete remission or uncertain and 92 patients (38%) presenting partial remission before ASCT. However, only 53% of the patients underwent transplantation, and the expected benefit of adding R has remained unclear. In the article that accompanies this editorial, Crump et al reported a similar design (NCIC Clinical Trials Group LY.12) based on a noninferiority comparison between R-GDP (R, gemcitabine, dexamethasone, and cisplatin) and R-DHAP. B-cell lymphoma patients were also administered R. Compared with DHAP treatment, treatment with GDP administered on an outpatient basis before high-dose chemotherapy and ASCT was as effective and associated with less morbidity. In total, 619 patients with relapsed/refractory aggressive lymphoma were randomly assigned to a treatment group. Of these patients, 71% presented with DLBCL; 15% presented with a lymphoma that had transformed from an indolent B-cell histology; and 8% presented with T-cell or anaplastic large-cell lymphoma. Among the 554 patients with B-cell lymphoma, 74% were administered R as part of a prior therapy. In the intention-to-treat population, the response rate after two cycles of treatment was 45.2% in the GDP group and 44.0% in the DHAP group (P .84). No differences were observed between the two regimens according to the histological subtype, revised IPI score, relapse less than 12 months or prior use of R (Appendix Table A6). However, a formal statistical comparison of the response rates in these patient subgroups was not performed. The transplantation rate was 52.1% in the GDP group and 49.3% in the DHAP group. The overall response rate observed following GDP was not inferior to that of DHAP, and GDP was associated with fewer grade 3 and 4 adverse events. The authors achieved their major end point. Considering that both studies revealed similar response rates, the results are disappointing. Interestingly, the T-cell lymphoma subset treated in the LY.12 study display an even poorer response rate of 30% to 38%. No differences were observed with respect to stem-cell collection and the percentage of patients receiving ASCT. The event-free survival and overall survival outcomes were also similar among the three regimens (R-DHAP, R-ICE and R-GDP), but these results were disappointing and revealed only a minimal, nonsignificant advantage for R-DHAP in the CORAL study. Toxicity was the main difference among the three regimens, with GDP displaying the least toxicity. In the CORAL study, secondary International Prognostic Index (IPI), a relapse/refractory episode less than 12 months from diagnosis and prior R exposure were significantly associated with poor response and survival in multivariable analyses. In the LY.12 study, poor Eastern Cooperative Oncology Group performance status, advanced stage at the start of the study and two or more disease sites were associated with a worse overall survival. Secondary IPI was not directly examined, JOURNAL OF CLINICAL ONCOLOGY E D I T O R I A L VOLUME 32 NUMBER 31 NOVEMBER 1 2014