|Home | About | Journals | Submit | Contact Us | Français|
Diffuse large B cell lymphoma (DLBCL) is the most common subtype of non-Hodgkin lymphoma (NHL) comprising approximately one third of all cases of NHL.1 Although the addition of rituximab to conventional chemotherapy has improved outcomes for patients with DLBCL, a significant number still relapse and eventually die from lymphoma.2 For patients with chemosensitive relapsed DLBCL autologous stem cell transplantation (auto-SCT) provides better outcomes than chemotherapy alone, with approximately 50% of patients able to achieve long term disease free survival.3, 4 However, outcomes are worse for patients with chemoresistant or high risk disease. This includes patients with early relapse, defined as relapse less than one year after the date of diagnosis, primary induction failure (PIF), defined as less than complete response to first line chemotherapy, and/or relapse after rituximab.
High-risk patients may benefit from allogeneic stem cell transplantation (allo-SCT) as it provides a tumor free source of stem cells and the potential benefit of a graft versus lymphoma effect (GVL). Two larger retrospective studies comparing allo-SCT and auto-SCT for relapsed DLBCL showed higher non relapsed mortality (NRM) and similar overall survival (OS) and disease free survival (DFS) in allo-SCT recipients, but these studies did not focus on patients with refractory disease and early relapses 5, 6. Here for the first time, we compare the clinical outcomes of patients with early relapsed and primary induction failure DLBCL undergoing allo-SCT versus auto-SCT.
The Washington University School of Medicine transplant database was searched to identify patients with DLBCL who had undergone auto-SCT or allo-SCT for early relapse or primary induction failure between January 1, 1997 and December 30, 2010. The final study cohort was 121 patients (42 allo-SCT, 79 auto-SCT). Patient, disease and transplant characteristics are summarized in Table 1. The allo-SCT group was younger, more likely to have a transplant before 2003, included a higher proportion of patients with PIF and chemoresistant disease at the time of transplant, and received higher numbers of chemotherapy regimens before transplant. Notably, there were no differences in stage at diagnosis, B symptoms, co-morbidity index, rituximab exposure, or time from diagnosis to transplant.
The maximal response to SCT was not significantly different between auto (63% complete response, 6% partial response, 4% stable disease, 17% progressive disease, 10% unknown) and allo-SCT (52% complete response, 3% partial response, 7% stable disease, 24% progressive disease, 14% unknown) (P=0.53). Overall survival at 1, 3, and 5 years after transplant was 40%, 26%, and 23%, respectively after allo-SCT and 67%, 54%, and 50% respectively after auto-SCT (P=0.002). In multivariate analysis, allo-SCT, age > 50 years, and chemoresistance at the time of transplant were associated with lower survival (Table 2). Progression free survival (PFS) at 1, 3, and 5 years after transplant was 36%, 23%, and 20% respectively in allo-SCT and 62%, 53%, and 49%, respectively after auto-SCT (P=0.02). In multivariate analysis, PFS was adversely influenced by allo-SCT, age > 50 years and 3 or more prior chemotherapy regimens (Table 2). Cumulative incidence of relapse/progression was not different in allo-SCT group compared with auto-SCT groups. At 1, 3, and 5 years after transplant, cumulative incidence of relapse/progression was 37%, 41%, and 41%, respectively in allo-SCT group, and 33%, 37%, and 39%, respectively in auto-SCT group (P=0.59). In multivariate analysis only refractory disease at transplant and longer interval between date of diagnosis and transplant were associated with greater risk of relapse/progression (Table 2). Cumulative incidence of non-relapse mortality (NRM) at 1, 3, and 5 years after transplant was 40, 56, and 60% respectively in allo-SCT and 9%, 17%, 20% respectively in auto-SCT (P=0.0002). In multivariate analysis, only type of transplant independently predicted NRM (Table 2).
Multiple studies have found that early relapse predicts worse outcomes for DLBCL.3, 7, 8 In the phase III CORAL trial, patients with early relapsed DLBCL had half the response rate to salvage chemotherapy, were much less likely to proceed to auto-SCT, and had lower PFS than those who relapsed more than 12 months after the date of diagnosis.3 For patients who do undergo auto-SCT, early relapse is a negative predictor of both OS and PFS.3, 9 Allogeneic SCT for early relapse is largely understudied, but it is associated with a high risk of relapse post transplant for patients with NHL.10 Our study is the first to focus on auto and allo-SCT for early relapsed/PIF DLBCL. Overall, our study shows that for a combined group of DLBCL patients who do not achieve CR with first line chemotherapy and patients who relapse early after achieving CR, auto-SCT compared to allo-SCT is associated with better OS, PFS, and NRM and a similar relapse rate. These results are similar to the retrospective registry study performed by Lazarus et al 6 despite variation in the two study populations. Compared to our study, the Lazarus study included patients with late relapses, had lower percentage of patients with PIF, and fewer patients with progressive disease during first line chemotherapy. Despite greater chemoresistance among patients receiving allo-SCT (15% auto vs 33% allo), the incidence of relapse/progression was not different and the auto-SCT group was more likely to die of lymphoma (61% vs 35%). While this difference might be reflective of a GVL effect, it is offset by significantly higher NRM in the early post-transplant period. Of note, nearly all of the allo-SCT patients in our study were treated with myeloablative conditioning regimens. Reduced intensity conditioning (RIC) or non-myeloablative conditioning regimens have been shown to reduce NRM but at a greater risk of relapse. Comparison of RIC allo-SCT and auto-SCT for DLBCL with PIF or early relapse is best addressed by a prospective randomized trial.
In summary, our study suggests that while allo-SCT could be considered in patients with DLBCL with early relapse or refractory disease, auto-SCT is associated with better survival and generally more preferable. It is possible that some subsets of early relapse patients benefit from the allo-SCT. Comparison of auto and allo-SCT for early relapsed chemosensitive DLBCL with particular attention to various patient and disease characteristics deserves further analysis in a randomized clinical trial.
Conflict of interest: The authors declare no conflict of interest.