Bolaños-Meade et al provide a concise review of tandem transplantation for patients with multiple myeloma. High-dose chemotherapy with autologous stem cell support has been shown to improve response rates, event-free survival, and overall survival over conventional chemotherapy in a major randomized clinical trial-the Intergroupe Français du Myélome (IFM)-90 trial.[1] This procedure is now accepted as the standard of care for newly diagnosed myeloma patients younger than age 70 years. However, the same study demonstrated the need to improve upon single autografts, as the overall survival rate at 7 years in the transplant group was only 43%.
Conceptual Basis of Tandem Autografts
Using tandem transplants, investigators have aimed to improve outcomes by incrementally achieving higher complete response rates with repeated cycles of high-dose therapy requiring stem cell support. With their Total Therapy protocol-a series of non-cross-resistant chemotherapy regimens culminating in tandem transplantation-researchers from the Arkansas Cancer Research Center showed that the complete response rate increased from 26% to 41% following the first and second transplant, respectively.[2] On multivariate analysis, complete response was a significant prognostic factor for improved outcome.
Although a complete response, by either immunofixation electrophoresis or molecular technology, has an overall positive influence on outcome in most studies, it may not be the best surrogate marker of treatment efficacy. Consider, for example, the preliminary results of a trial conducted by the Spanish Program for the Study and Treatment of Malignant Hemopathies (PETHEMA),[3] a randomized series of 216 patients treated with either conventional or high-dose therapy. This study showed a significant increase in complete responses but no significant prolongation in event-free and overall survival with high-dose therapy intensification in patients responding to conventional chemotherapy with VBMCP (carmustine [BiCNU], vincristine, melphalan(Drug information on melphalan) [Alkeran], cyclophosphamide(Drug information on cyclophosphamide) [Cytoxan, Neosar], prednisone(Drug information on prednisone)) or VBAD (vincristine, carmustine(Drug information on carmustine), doxorubicin(Drug information on doxorubicin) [Adriamycin], dexamethasone(Drug information on dexamethasone)).
In contrast to their earlier report, the Arkansas group showed in a more recent analysis of tandem transplant data that disease biology (absence of abnormalities in chromosomes 11 or 13, low beta-2-microglobulin, and transplant within 12 months of diagnosis) rather than posttransplant response status are better predictors of long-term survival.[4]
Evidence for Tandem Autografts
The largest series of tandem autotransplants by Barlogie et al had impressive patient numbers, long follow- up, and a cumulative increase in complete response rates with each transplant.[2] In fact, a small subset of patients (n = 46) with good prognostic variables tended to plateau in event-free survival, suggesting an operational "cure" among these goodrisk patients.[4] The series, however, was a nonrandomized heterogeneous population of patients who received treatment at a single institution, making the risk of selection bias significant.
Currently, four major European multicenter randomized trials are addressing the issue of tandem vs single transplantation; to date, none have been published in peer-reviewed medical journals. The results of these trials are reviewed in the main article. The most mature study, IFM-94, showed that a tandem transplant was associated with improved event-free and overall survival.[5] The subsets of patients who achieved better outcomes with randomization to tandem transplants were those with low beta- 2-microglobulin levels and those who received tandem transplants and survived more than 3 years posttransplant.
Preliminary analysis of the Italian Bologna 96 study showed a significant prolongation in response duration for the tandem transplant arm.[6] However, there were no significant differences in the complete response rate or overall survival between groups.
The Dutch Hemato-Oncology Association (HOVON) trial[7] tested the sequential vs single high-dose question by comparing two cycles of intermediate- dose melphalan (70 mg/m2) without stem cell support vs the same therapy followed by cyclophosphamide and total-body irradiation-based autologous transplantation. Although the complete response rate increased significantly in the transplant group (13% vs 29%), the second intensification did not result in improvement in event-free or overall survival or quality of life.
Finally, the randomized French Myelome-Autograffe (MAG) study failed to show any differences between single and tandem transplantation.[ 8]
In summary, IFM-94 is the only tandem transplant trial to report an increased overall survival. On an intent- to-treat basis, excluding subgroup analyses, the other trials show that there are no statistically significant progression-free or overall survival benefits associated with tandem autografts. Unlike the single-arm Arkansas study,[4] which showed an increase in complete responses from 26% to 41% with the second transplant, the four other studies failed to demonstrate as marked an increase in the complete response rate after the second transplant (26% to 35%). This may explain the small (or absent) survival benefit from the second transplant in these studies.
Implications for Practice
In the nonrandomized Arkansas study, the inherent selection bias stemming from a referred patient population cannot be excluded. The multicenter randomized trials that have failed to show significant benefit have only been reported in abstract form with short follow-up. Clearly, more data are necessary before tandem transplants can be considered the standard of care. Tandem transplantation should not be recommended as an upfront strategy for newly diagnosed myeloma patients outside the setting of a well-designed clinical trial.
The identification of patients who might benefit from planned upfront tandem transplants remains an unresolved issue. Patients with good biologic risk factors may achieve good outcomes without upfront tandem transplants, whereas the poorest-risk patients clearly have unfavorable outcomes despite tandem autografts. Conversely, the best-risk patients with chemotherapy-sensitive disease may achieve plateau survival curves without repeated dose intensification. To this end, the Spanish group found that 73 patients with a complete response after a single transplant had event-free and overall survivals similar to the 26 patients who completed tandem transplant and achieved a complete response after the first transplant (personal communication, J. Blade, 2002). These data imply that there is no clinical advantage to additional therapy after achieving a complete response.
The role of the second transplant as a salvage procedure vs an "operatively curative" upfront procedure is unclear. Apart from historical comparisons reported in abstract form,[9] there are no randomized studies comparing early vs late second transplant nor any randomized studies comparing second transplant as salvage therapy to conventional salvage therapy. The prudent approach would be to harvest adequate stem cells for two autografts prior to stem cell exposure to alkylators and then complete the first course of high-dose therapy. Subsequently, a second autograft as salvage therapy could be considered based on host and disease characteristics at the time of disease progression.
Autografts Followed by Tandem Allografts
Allogeneic transplants provide a unique opportunity not observed with autologous transplants: a tumor-free stem cell source and a well-proven graft-vs-myeloma effect. Even with these advantages, only 10% to 20% of allografted patients remain in continued complete response 5 or more years posttransplant. However, a definite subset of myeloma patients remain disease-free at the molecular level beyond 5 years, suggesting that allogeneic transplantation can be curative in some patients.
Allogeneic transplantation is limited to a small cohort of younger myeloma patients who have human leukocyte antigen (HLA)-compatible donors. Furthermore, the high transplant- related mortality of conventional myeloablative allografts in myeloma has essentially precluded this option. In at least three case control comparisons of autologous vs allogeneic transplant, the autologous transplant group had a longer median survival.
In an attempt to decrease transplant- related mortality while optimizing the graft-vs-myeloma effect, the use of a reduced nonablative conditioning regimen (minitransplant) has recently been investigated in myeloma and other malignancies. Several small studies in myeloma with short follow-up have demonstrated the "proof of principle" that alloreactive donor T cells can induce impressive responses, even in heavily pretreated patients with resistant disease. The article by Bolaños-Meade et al summarizes the major series in nonablative transplants. The incidence of acute and chronic graft-vs-host disease associated with minitransplants has been reported to be as high as 67% and 47%, respectively, and the transplant-related mortality has ranged from 13% to 67%.[10]
One of the most intriguing concepts in myeloma management is the use of sequential autologous transplant to provide maximal cytoreduction followed by nonablative allogeneic transplant to maximize the graft-vs-myeloma effect of alloreactive T cells. The largest sequential tandem auto/allograft series (all HLAidentical siblings) was reported by the Seattle group,[11] who observed an overall response rate of 84% and a complete response rate of 53%. The incidence of acute (grade II-IV) and chronic graft-vs-host disease was 45% and 55%, respectively. Of the 31 patients who received nonablative allotransplants, 5 died at a median follow-up of 328 days-1 from progressive disease, 3 from graft-vs-host disease, and 1 from encephalopathy.
More recently, a German-Israeli multicenter group reported similar results with nonablative matched unrelated donor allografts in 21 patients with advanced myeloma[12] and prior autografts (9 as part of an auto/allo tandem protocol). They observed an overall response rate of 90% and a complete response rate of 40%. The incidence of acute graft-vs-host disease was 38% (grade III/IV graft-vshost disease in 19%) and chronic graft-vs-host disease, 37%. Transplant- related mortality was 10% at 100 days and 26% at 1 year. At a median follow-up of 13 months, the 2-year estimated overall and progression- free survival rates are 74% and 53%, respectively. Not surprisingly, patients who had already relapsed following an autograft had a shorter progression-free survival (26% vs 86%, P = .04).
Conclusions
These studies indicate that nonablative tandem transplants, although conceptually sound and promising, require larger studies and longer follow- up to determine efficacy. At this time, nonablative or autologous-miniallogeneic tandem transplantation should be considered investigational and only in the context of welldesigned clinical trials. The Eastern Cooperative Oncology Group (ECOG) E4A98 trial (now enrolling) is also using a tandem auto-mini allograft approach to evaluate its feasibility, efficacy, and immunoreconstitution. The Blood and Marrow Transplant Clinical Trials Network (BMTCTN) will soon initiate a trial comparing autologous transplant to autologous-mini-allogeneic transplants.
Further advances in myeloma may not come from dose intensification alone. A multipronged attack using targeted, cytoreductive, nonablative, and immunologic therapies coupled with increasing biologic risk stratification strategies will be necessary to ultimately produce operational or true cures in this disease. To this end, current clinical efforts should be focused on novel sequential treatment strategies to increase response rates and prolong disease control and overall survival while maintaining good quality of life.
