ABSTRACT
'Double-hit lymphomas' (DHL), defined by concurrent MYC and BCL2 (or, alternatively, BCL6) rearrangements, have a very poor outcome compared to standard-risk, diffuse large B-cell lymphomas (DLBCL). Consequently, dose-intensive (DI) therapies and/or consolidation with high-dose therapy and transplant have been explored in DHL, although benefit has been debated. This meta-analysis compared survival outcomes in DHL patients receiving dose-escalated regimens [DI: R-Hyper-CVAD (rituximab, cyclophosphamide, vincristine, doxorubicin, dexamethasone) or R-CODOX-M/IVAC (rituximab, cyclophosphamide, doxorubicin, vincristine, methotrexate/ifosfamide, etoposide, high dose cytarabine); or intermediate-dose: R-EPOCH (rituximab, etoposide, doxorubicin, cyclophosphamide, vincristine, prednisone)] versus standard-dose regimens (R-CHOP; rituximab, cyclophosphamide, doxorubicin, vincristine, prednisone) in the first-line setting. Data were synthesized to estimate hazard ratios of dose-escalated treatments versus R-CHOP using a Weibull proportional hazards model within a Bayesian meta-analysis framework. Eleven studies examining 394 patients were included. Patients were treated with either front-line R-CHOP (n = 180), R-EPOCH (n = 91), or R-Hyper-CVAD/rituximab, methotrexate, cytarabine (R-M/C), R-CODOX-M/R-IVAC (DI) (n = 123). Our meta-analysis revealed that median progression-free survival (n = 350) for the R-CHOP, R-EPOCH and DI groups was 12·1, 22·2, and 18·9 months, respectively. First-line treatment with R-EPOCH significantly reduced the risk of a progression compared with R-CHOP (relative risk reduction of 34%; P = 0·032); however, overall survival (n = 374) was not significantly different across treatment approaches. A subset of patients might benefit from intensive induction with/without transplant. Further investigation into the role of transplant and novel therapy combinations is necessary.