At the end of the 1990s, 2 prospective trials compared autologous stem cell transplantation (auto-SCT) with myeloablative allo-SCT in MM.2,3 In both of these, extremely high rates of transplant-related mortality (TRM) were observed in the allogeneic arms, ranging from 30% despite partial T-cell depletion in the study conducted by the HOVON group,2 up to 53% in the United States Intergroup trial (S9321), in which T-cell depletion was not carried out.3 As a result, myeloablative allo-SCT was abandoned and the procedure is not recommended as part of frontline therapy in patients with symptomatic de novo MM.4
Although the introduction of RIC regimens has resulted in a decrease in the high incidence of TRM associated with myeloablative conditioning, an intense debate is currently ongoing regarding the question of whether patients should be subjected to the substantial morbidity, especially graft-versus-host disease (GVHD), and the risk of mortality associated with RIC allo-SCT as part of first-line therapy compared with the relatively safe procedure of auto-SCT.5,,,,,–11 Lokhorst et al performed a donor versus no-donor analysis of patients included in the phase 3 HOVON- 50 MM trial.1 After a single auto-SCT, eligible patients were randomized according to the availability of an HLA-identical sibling to receive either maintenance therapy consisting of thalidomide or α-interferon or RIC allo-SCT after low-dose 2 Gy total body irradiation. Importantly, the trial included a large number of patients, and the median follow-up, which exceeds 6 years, is long enough to draw reliable conclusions. The complete response, the overall survival (OS), and progression-free survival (PFS) rates were not significantly different between the donor and no-donor groups. As expected, the cumulative incidence of nonrelapse mortality was significantly higher, while the incidence of relapse was significantly decreased in the RIC allo-SCT group. The PFS of patients who actually received RIC allo-SCT was significantly prolonged compared with the patients who continued maintenance, but this difference did not translate into a survival benefit.
These results differ from those previously reported by the Italian5,6 and the European Blood and Marrow Transplantation (EBMT) groups7 regarding the comparison of tandem auto-SCT with a sequential application of single auto-SCT followed by RIC allo-SCT. In both of these trials PFS and OS were improved for patients with a sibling donor. Interestingly, in all 3 studies, the Lokhorst trial as well as the Italian and EBMT studies, which used the same conditioning regimen before RIC allo-SCT, the PFS and OS of patients who had a donor and underwent the auto-SCT/RIC allo-SCT procedure appeared similar. The absence of a difference between the donor and the no-donor groups in the Lokhorst study, which is the major finding of this trial, is related to a much better outcome of the no-donor group compared with the EBMT and the Italian series. The HOVON study was the last of the 3 trials to be initiated, and was conducted when lenalidomide and/or bortezomib could be routinely given at the time of relapse in the no-donor group of patients. The results of the HOVON trial are in line with those of a recent donor versus no-donor comparison reported by the Blood and Marrow Transplant Clinical Trials Network (BMT-CTN) group, in which after a shorter follow-up no difference was seen in outcome between patients in the tandem auto-SCT arm and those undergoing auto-SCT followed by RIC allo-SCT.11 Overall, 6 donor versus no-donor studies are now available, 2 supporting5,–7 and 4 opposing1,8,,–11 the sequential approach of auto-SCT followed by RIC allo-SCT.
The impact of the HOVON study is substantial, because it is the only one to be performed at a time of widespread use of novel agents at relapse. The systematic use of combinations of novel agents upfront as induction therapy before auto-SCT, as well as their incorporation into the consolidation and/or maintenance settings after auto-SCT, which can be considered an optimal strategy in the 2010s, will further improve the results of procedures not including RIC allo-SCT. Moreover, recent attempts aimed at bettering outcomes with RIC allo-SCT, such as the use of maintenance treatment with low-dose lenalidomide in the HOVON-76 study, failed because of the rapid induction of acute GVHD.12 It can be anticipated that the gap between the results achieved with auto-SCT versus those obtained with RIC allo-SCT, also taking into account toxicity, will widen in the future to the detriment of the allogeneic procedure.
In their important report and analysis, Lokhorst et al note that the study was not powered to address the significant issue of the outcome of poor-risk patients, such as those harboring the 17p deletion. This subgroup, which constitutes 7% to 10% of all patients, might be the only one to derive a benefit from an allogeneic effect, which, however, will have to be proven in clinical trials.
In conclusion, front-line myeloablative allo-SCT in MM died 10 years ago, and the latest results from the HOVON-50 study further corroborate that RIC allo-SCT should not be recommended as part of front-line therapy in patients with symptomatic MM.
Conflict-of-interest disclosure: The author declares no competing financial interests. ■