Key points

Introduction

Despite significant progress in the treatment of multiple myeloma, most patients will have recurrent disease and die. New drugs and autologous stem cell transplantation (ASCT) have increased rates of remission, improved remission durations, and prolonged overall survival rates by 3 to 5 years. Relapses, however, continue for most patients, emphasizing the need for new drugs and treatment strategies. Allogeneic stem cell transplantation (AlloSCT) offers the potential of harnessing an immunologic graft-versus-myeloma (GVM) effect capable of controlling residual disease and offering the potential for cure. Unfortunately, the complexity, complications, and relatively poor outcomes of AlloSCT for myeloma have limited its application and relegated this therapy to investigational studies. As survival of patients with myeloma has improved with new drugs and ASCT, the role of AlloSCT, if any, in the treatment of myeloma has been further questioned. This article reviews the current knowledge about AlloSCT for myeloma and discusses current research activities.

Introduction

Despite significant progress in the treatment of multiple myeloma, most patients will have recurrent disease and die. New drugs and autologous stem cell transplantation (ASCT) have increased rates of remission, improved remission durations, and prolonged overall survival rates by 3 to 5 years. Relapses, however, continue for most patients, emphasizing the need for new drugs and treatment strategies. Allogeneic stem cell transplantation (AlloSCT) offers the potential of harnessing an immunologic graft-versus-myeloma (GVM) effect capable of controlling residual disease and offering the potential for cure. Unfortunately, the complexity, complications, and relatively poor outcomes of AlloSCT for myeloma have limited its application and relegated this therapy to investigational studies. As survival of patients with myeloma has improved with new drugs and ASCT, the role of AlloSCT, if any, in the treatment of myeloma has been further questioned. This article reviews the current knowledge about AlloSCT for myeloma and discusses current research activities.

History

An initial attempt at treating refractory myeloma with total body irradiation (TBI) and bone marrow from an unmatched cadaver, reported in 1957, was unsuccessful. Success with AlloSCT for the treatment of multiple myeloma was first reported in 1982 and again in 1984. One patient with progressive myeloma received cyclophosphamide and 800 cGy TBI followed by HLA identical sibling marrow and achieved engraftment and a remission. Two patients with advanced myeloma received cyclophosphamide and TBI followed by sibling bone marrow. Both patients engrafted with subsequent resolution of their myeloma; however, one died of disseminated zoster at 6 months, while the other relapsed at 3 years. These reports were followed by small series from Seattle and the European Bone Marrow Transplant Cooperative Group (EBMT) showing the feasibility of this procedure in myeloma. As experience accumulated in the 1980s and 1990s, it became clear that AlloSCT using ablative techniques was associated with high morbidity and mortality, as a consequence, many transplant centers abandoned myeloablative allograft conditioning for myeloma. A US trial of early versus late ASCT had an option of AlloSCT for patients with matched sibling donors. Patients received myeloablative conditioning with melphalan plus TBI. After 36 patients were enrolled in the AlloSCT arm, this group was closed to further accrual because of transplant-related mortality (TRM) exceeding 50%. Interestingly, after more than 7 years of follow-up, the overall survival (OS) rate of the AlloSCT group was 40%, better than either arm of the ASCT groups. Because of the GVM effect of AlloSCT, deeper and more durable remissions were observed, resulting in a reduced rate of relapse in the AlloSCT group compared with either ASCT group. Studies using patient-specific primers to look for minimal residual disease in myeloma patients achieving a complete response (CR) reported higher rates of minimal residual disease negativity among recipients of AlloSCT compared with ASCT. Furthermore, patients with minimal residual disease–negative remissions have much longer disease-free intervals than patients whose marrow is minimal residual disease positive at any time after transplant.

By the mid-1990s the EBMT reported a retrospective comparison of 189 patients with myeloma who received myeloablative AlloSCT and were matched for gender and prior courses of chemotherapy with 189 patients with myeloma who received ASCT. The groups were comparable except the median age of the AlloSCT group was 6 years younger, with 16-month longer median follow-up. OS was superior for the ASCT group (median, 34 vs 18 months; P = .001) because of a much higher TRM in the AlloSCT group (41% vs 13%; P = 0001). This higher TRM was not offset by the lower rates of relapse seen in the AlloSCT group (at 4 years, 50% vs 70%; P = .04).

Thus, the perception of an increase in morbidity and mortality after AlloSCT for patients with myeloma draws directly from outcomes of patients reported in the literature. Recently, this idea has been challenged by pooled analyses of 56,000 transplants reported to EBMT. Patients were given a risk score of 0 to 7 based on age, disease stage, time from diagnosis to transplant, donor type, and donor-recipient gender. Patients with myeloma had a greater mortality risk after AlloSCT but were also observed to have a higher risk score because of more advanced age, longer interval from diagnosis to transplant, and more advanced disease stage. When similar risk score patients with myeloma or acute leukemia underwent transplant, the mortality was similar. Although interesting, as a practical matter, these observations are of limited benefit, because the diagnosis of myeloma is frequent at advanced age, and most patients do not achieve complete remission before transplant. Furthermore, the EBMT observed a marked reduction in nonrelapse mortality after 2 years from 46% for patients undergoing transplant between 1983 and 1993 to 30% for patients undergoing transplant between 1994 and 1998, likely because of improvements in supportive care and more careful patient selection. Nevertheless, with a mortality rate of even 30%, most transplant centers began to explore alternatives to myeloablative AlloSCT in patients with myeloma and other hematologic malignancies.

Reduced-intensity conditioning regimens

In the late 1990s, interest in exploiting the GVM effects of AlloSCT was adopted by several groups using a reduced-intensity conditioning designed to minimize early toxicity and mortality associated with high intensity, myeloablative treatments. A reduced-intensity regimen delivers therapy that is immunosuppressive and designed mainly to facilitate donor engraftment. This reduced-intensity approach relies heavily on the immunologic effect of the graft for disease control, as prior demonstrations of the GVM effect in recipients of donor lymphocyte infusions showed. Several conditioning regimens have been reported ranging from minimal conditioning using 2 Gy TBI to ablative yet reduced doses of melphalan or busulfan. A summary of the principal reduced intensity regimens evaluated in clinical trials of 40 or more patients is shown in Table 1 . Almost all of these series used peripheral blood stem cells rather than bone marrow from matched related or unrelated donors because of higher rates of graft failure or rejection observed with marrow. Trials using a single dose of 2 Gy TBI, mycophenolic acid, and cyclosporine for GVHD prophylaxis with matched related donors have reported 8% grades 3 to 4 GVHD, 11% to 22% nonrelapse mortality rate, and 50% to 60% CR rates. The addition of fludarabine to 2 Gy TBI in a mix of related and unrelated donors was associated with more severe GVHD (24%) but similar nonrelapse mortality. Fludarabine melphalan combinations with cyclosporine, methotrexate, and antithymocyte globulin (ATG, unrelated donors only) reported 21%, severe GVHD, less than 20% nonrelapse mortality, and CR rates of approximately 50%. Intermediate dose (100 mg/m ² ) melphalan, with or without 2.5 Gy TBI and cyclosporine and prednisone for GVHD prophylaxis, was associated with high rates of severe GVHD (42%) and nonrelapse mortality (38%). A regimen of fludarabine and ablative but reduced doses of busulfan, 8 mg/kg, and ATG resulted in low (7%) severe GVHD, moderate (17%) nonrelapse mortality, and low (27%) rates of CR. It is likely that the relatively high doses of ATG used in this trial interfered with the GVM effects of the allograft. A novel 3-drug regimen of thiotepa, 5 mg/kg, fludarabine, and melphalan reported low rates (5%) of severe GVHD, 13% nonrelapse mortality rate, and a relatively high (62%) CR rate. None of the regimens in Table 1 have been directly compared with one another, and differences in the mix of patient characteristics, donor types, and the GVHD prophylaxis may account for many of the outcome disparities.

GVHD and relapse

Multiple trials show a strong association between the development of acute or chronic GVHD and improved progression-free survival (PFS). An analysis of registry data after reduced-intensity AlloSCT for myeloma found that patients who had chronic extensive or chronic limited GVHD had superior OS and PFS compared with patients who had no chronic GVHD (84% and 46% vs 29% and 12%, respectively). A subsequent multivariate analysis found a lower risk of relapse and better event free-survival with limited chronic GVHD, but this failed to translate to improved OS.

Because of the increased morbidity and mortality associated with AlloSCT for myeloma, it is recommended by some experts that this treatment be deferred until ASCT fails. Relapse after ASCT, however, is an important negative risk factor for outcome after AlloSCT. Several studies report that patients whose disease has progressed after ASCT and who are then offered AlloSCT, have increased rates of TRM and relapse. A retrospective study has compared outcomes in a multicenter analysis of patients with myeloma who progressed after ASCT. The study compared 68 patients who had relapse after ASCT and were treated with AlloSCT from matched related or unrelated donors with 94 patients who had relapse after ASCT but did not have suitable donors and received salvage chemotherapy with novel agents. The groups were comparable, except for a 4-year median greater age in the no-donor group and slightly more patients with extramedullary disease in the donor group. Two-year TRM was 22% versus 1% in the donor versus no-donor groups ( P <.0001), and PFS at 2 years was 43% versus 18% ( P <.001), but 2-year OS was not different at 54%. Of the 68 patients, 19 (28%) were reported to be in CR at 29 months, indicating that a proportion of patients can achieve long-term control with AlloSCT, even after ASCT relapse. The above studies would indicate that waiting until failure of ASCT may not be the best strategy for considering AlloSCT.

HI-risk myeloma

Another recommendation is that AlloSCT be reserved for patients with myeloma who have the highest risk of disease relapse with conventional therapy, with or without ASCT, based on the hypothesis that the GVM effect of the donor graft is independent of chemosensitivity. Indeed, high β-2 microglobulin, conventional cytogenetic abnormalities such as deletion of chromosome 13, and certain chromosomal abnormalities detected by fluorescence in situ hybridization (FISH) including translocation 4;14 or deletion 17p have been associated with PFS intervals as short as 9 to 20 months after ASCT. Data using newer drugs such as bortezomib for both induction before ASCT and maintenance or consolidation after ASCT have indicated improved PFS and OS of certain high-risk groups compared with patients not receiving bortezomib. Unfortunately, outcomes with bortezomib were still not as good as in the group of patients without these high-risk features.

Retrospective analyses of AlloSCT suggest that this approach may overcome high-risk cytogenetics such as del17 or 4;14, and very small prospective trials have validated this theory. The German myeloma study group reported the preliminary results of a prospective trial, DSMM V, for patients with high-risk myeloma based on del13 by FISH who received one ASCT followed by either a reduced-intensity AlloSCT or a second ASCT based on the availability of a related or unrelated donor. Of 199 patients enrolled, 126 received AlloSCT using fludarabine and melphalan 140 mg/m ² , whereas 73 received ASCT after a second course of melphalan. At 1 year, the CR rate was significantly higher for ASCT-AlloSCT (59% vs 32%; P = .003). TRM for the ASCT-AlloSCT group was 12.6%. With short follow-up of 25 to 34 months, the OS was 72% for the tandem ASCT group versus 60% for the ASCT-AlloSCT group ( P = .22). A longer follow-up and analysis of this trial is currently underway.