Autologous Stem-Cell Transplantation (AutoSCT)
The initial policy on myeloablative chemotherapy with autologous stem-cell transplantation (AutoSCT) to treat multiple myeloma was based on a 1996 Blue Cross Blue Shield Association (BCBSA) Technology Assessment Center (TEC) Assessment that summarized evidence on outcomes for patients with newly diagnosed or responsive multiple myeloma separately from those with refractory or resistant myeloma. Patients with responsive myeloma were defined as patients who achieved a complete or partial (at least 50% tumor reduction) response to chemotherapy, while patients with resistant or refractory multiple myeloma were defined as those who achieved < 50% reduction in tumor burden. The following conclusions were offered.
Newly Diagnosed or Responsive Multiple Myeloma:
- The available data support the conclusion that AutoSCT is at least as effective and may be more effective than conventional-dose chemotherapy for improving the health outcomes of these patients. In a key randomized trial, event-free and overall survival (OS) after myeloablative therapy was clearly better than after conventional chemotherapy.
Multiple Myeloma in a Refractory or Resistant Relapse:
- In contrast, insufficient data are available to support a conclusion regarding the outcomes of AutoSCT for myeloma patients in a refractory relapse. Most of the data consisted of uncontrolled clinical series of patients, and outcomes were poor, since treatment-related toxicity was frequent and severe in heavily pretreated patients. No randomized controlled trials were reported.
Reviews in 2003 summarized data from trials on newly diagnosed patients completed after the initial 1996 TEC Assessment. These reviews confirmed that a larger proportion of patients survive after AutoSCT than after conventional dose induction therapy for newly diagnosed disease. A randomized controlled trial (RCT) by Child et al., published after these reviews, also reported longer survival after AutoSCT, and included a meta-analysis of results from three RCTs that favored AutoSCT (odds ratio for survival, 0.70; 95% confidence interval, 0.53 to 0.93). In contrast, another RCT that compared intensified induction chemotherapy alone with intensified induction followed by myeloablative AutoSCT reported no survival benefit from the myeloablative therapy.
An evidence-based review and clinical guideline of the American Society for Blood and Marrow Transplantation (ASBMT) emphasize the lack of data from RCTs on outcomes of AutoSCT for salvage therapy of MM. The review and guideline conclude that AutoSCT is preferred as de novo rather than salvage therapy, since this minimizes risks of myelodysplasia and other toxicities from extended treatment with alkylating agents. In addition, data show that newly diagnosed patients given AutoSCT as part of de novo therapy spend more time without symptoms, treatment, or treatment toxicity than those given conventional dose induction and AutoSCT for salvage at relapse or progression.
Note that myeloma in a refractory relapse differs from primary refractory disease (newly diagnosed myeloma that does not enter a partial or complete remission after conventional-dose induction). The RCTs on newly diagnosed patients showed that AutoSCT induces durable remissions and extends the duration of OS and disease-free survival in a substantial proportion of patients with primary refractory myeloma.
The policy regarding tandem transplantation initially was based on two 1998 BCBSA TEC Assessments that focused on tandem transplants for newly diagnosed or responsive multiple myeloma or for resistant multiple myeloma. The following conclusions were offered.
Newly Diagnosed or Responsive Multiple Myeloma
- Only five published studies were found that reported data on the outcomes of treatment in patients receiving tandem transplants. Although one of the studies was a randomized study comparing single and tandem transplants for newly diagnosed patients, the available data were preliminary, published as abstracts only, and did not permit conclusions on survival. The University of Arkansas reported a large, nonrandomized, single-institution case series of multiple myeloma treated with tandem transplants. However, the data compared results to historical controls treated with conventional-dose regimens and not to a single course of AutoSCT, the more appropriate comparator.
- Using indirect comparisons of previously published outcomes for patients given a single course of AutoSCT as historical controls, results reported for tandem and single transplants overlapped substantially for nearly all outcomes of interest. These data were considered inadequate to demonstrate superiority of tandem transplants over single transplants. Data on the duration of survival after tandem transplants were scant and nearly all other outcomes data were only available from single-arm studies with highly selected patients. Thus, the comparison of outcomes was subject to a high degree of patient selection bias.
Refractory or Resistant Multiple Myeloma
- Two reports with a total of 69 patients treated at one institution and a third report with 30 patients provided the only data on outcomes of tandem transplants to treat resistant multiple myeloma. There was no control group for direct comparison of outcomes in the most updated report on the larger series of highly selected patients. The earlier report from this institution and the third paper included nonrandomized controls, but aggregated outcomes for patients with resistant myeloma and those transplanted as part of first-line therapy. In addition, insufficient detail was provided to determine if the patients given tandem transplants for resistant myeloma were sufficiently comparable to those given either single Autologous SCT or conventional-dose salvage therapy to permit conclusions based on indirect comparison of outcomes from separate studies. Thus the available data were insufficient to permit conclusions on outcomes of tandem transplants.
Randomized Trials of Tandem Autotransplants
IFM-94 Results. The 2003 study by Attal et al. randomized newly diagnosed myeloma patients to single or tandem Autologous transplants. Outcomes were analyzed by intent-to-treat at 75 months’ median follow-up. Among those randomized to single transplants (n = 199), 148 relapsed: 33 were salvaged with a second autotransplant, 13 received no salvage, and the remainder received conventional chemotherapy plus thalidomide. Among those randomized to tandem autotransplants (n = 200), 129 relapsed: 34 were salvaged with another (3rd) transplant, 12 received no salvage, and the remainder received conventional chemotherapy plus thalidomide. Seven years after diagnosis, patients randomized to tandem transplants had higher probabilities than those randomized to single transplants for event-free (EFS; 20% vs. 10%, p = 0.03), relapse-free (RFS; 23% vs. 13%; p < 0.01), and OS (42% vs. 21%, p = 0.010). Treatment-related mortality was 6% and 4% after tandem and single transplants, respectively (p = 0.40). Second transplants apparently extended survival only for those who failed to achieve a complete response (CR) or very good partial response (VGPR) after one transplant (OS at seven years: 43% vs. 11%, p < 0.001).
Generalizability Issue. An accompanying editorial by Stadtmauer raised concerns that these results might be specific to the regimens used for myeloablative therapy in IFM-94. Patients in the single transplant arm received 140 mg/m2 melphalan plus total body irradiation (TBI), while those in the tandem arm received the same dose without TBI for the initial transplant and with TBI for the second transplant. The editorial cites the IFM-95 study as evidence, suggesting 140 mg/m2 melphalan plus TBI may be less effective and more toxic as myeloablative therapy than 200 mg/m2 melphalan and no TBI. Based on this, the author hypothesizes increased survival in IFM-94’s tandem arm may have resulted from greater cumulative exposure to melphalan (280 vs. 140 mg/m2).
The IFM-95 randomized trial directly compared 200 mg/m2 melphalan without TBI (n = 142) versus 140 mg/m2 plus TBI (n = 140) as conditioning regimens for single autotransplants. The trial reported OS at 45 months of 66% in the arm given 200 mg/m2 and 46% in the arm given 140 mg/m2 plus TBI. However, the CR rates were identical in the two arms, and there was no difference between arms in event-free survival or relapse rates. The authors hypothesized the difference in OS might have resulted from more frequent use of a second transplant on relapse in the arm originally given 200 mg/m2 melphalan (50% of 51 relapsed patients) than in the arm originally given 140 mg/m2 plus TBI (25% of 44 relapsed patients). The published report on this trial does not explain the reasons for this difference in salvage therapy.
Additional evidence supporting the superiority of 200 mg/m2 melphalan as the pretransplant conditioning regimen comes from a comparison of two separate single-arm studies, a registry analysis from the European Group for Blood and Marrow Transplantation, and a U.K. retrospective analysis. On the other hand, a Spanish registry analysis and a German retrospective study reported equivalent OS and event-free survival after either regimen.
Taken together, the evidence from IFM-95 and the series data comparing different conditioning regimens does not appear sufficiently compelling to justify considering tandem autotransplants investigational for myeloma. The concerns regarding generalizability of IFM-94’s results appear relevant more to the clinical research agenda to identify the best pre-transplant conditioning regimen for myeloma than to the principal finding of IFM-94 that a second transplant extends survival for myeloma patients who fail to achieve a complete or very good partial remission.
Other RCTs. Three other randomized trials compare single with tandem autotransplants using different conditioning regimens than IFM-94. Each has completed accrual and treatment, but requires another 1-2-years’ follow-up to mature. Barlogie et al. summarized recent interim results in Hematology 2003. One trial (HOVON) reports improved EFS in the tandem arm but no difference in OS, while the other two trials (Bologna 96 and MAG95) report no differences between arms for either outcome. However, survival curves in IFM-94 did not begin to separate until after 4–5 years of median follow-up.
Autotransplant Followed by Reduced-Intensity Conditioning Allotransplant
Rationale. The Stadtmauer editorial also stresses that all but 23% of IFM-94 patients randomized to tandem autotransplants relapsed or died by seven years. Recent reviews and commentaries agree that tandem autotransplants rarely cure myeloma. There is evidence of a graft-versus-myeloma effect after allotransplants, but conventional (myeloablative) allotransplants for myeloma are associated with excess early mortality. Thus, investigators hypothesize that an autologous transplant followed by a non-myeloablative, or reduced-intensity conditioning (RIC) allotransplant might be optimal myeloma therapy for those with an HLA-matched sibling donor. This approach uses the autotransplant to maximally reduce the malignant plasma cell burden, and the subsequent RIC allotransplant to (potentially) eradicate remaining malignant clones through a graft-versus-malignancy effect.
Uncontrolled Data. A recent study included 54 patients (median age, 52 years; range, 29–71) with previously treated myeloma (52% refractory or relapsed disease) given an initial autotransplant conditioned with 200 mg/m2 melphalan. Of these, 52 received a subsequent RIC allotransplant. Investigators reported 78% OS at a median 552 days after allografting. Treatment achieved a CR in 57% and an overall response rate of 83%. Acute graft-versus-host disease (GVHD) developed in 38% of patients (grades III/IV in four cases; grade II in all others), and chronic GVHD requiring therapy occurred in 46%. Twelve patients died: one from viral infection after the initial autotransplant, two from myeloma progression (three and 23 months post-RIC allotransplant), seven from GVHD, and one each from lung cancer and encephalopathy.
Ongoing RCT. At the 2003 American Society of Hematology (ASH) meeting, the IFM group reported preliminary results from a multi-arm trial (IFM99-03 and IFM99-04) comparing autotransplant followed by RIC allotransplant with tandem autotransplants for newly diagnosed high-risk myeloma (defined by elevated beta-2 microglobulin and/or chromosome 13 deletions). Each autotransplant cycle was conditioned with > 200 mg/m2 melphalan. They used "biological randomization" to select those with an HLA-identical sibling donor for the auto/min-allo arm (IFM99-03 protocol; n = 47). Those lacking a donor were selected for tandem autotransplants (IFM99-04 protocol; n = 128), and were further randomized to treatment with or without an antibody to IL-6 after the second transplant. A preliminary intent-to-treat analysis compared the IFM99-03 patients with all those in the IFM99-04 protocol. At median follow-up of 18 months (range, 9-39) for those still alive, estimated EFS at three years (52% vs. 54%; p = 0.37) and median EFS (21 vs. 25 months; p = 0.08) were not significantly different between groups. Relapse rate was 45% in the auto/RIC-allo arm, without a plateau on the EFS Kaplan-Meier curve. Also, 50% of those given an RIC allotransplant experienced chronic GVHD, which was extensive in only two cases.
Other Data. Two additional ASH presentations reported outcomes of the auto/RIC-allo approach for managing myeloma. One was an ongoing uncontrolled multicenter Italian trial in which 57 newly diagnosed patients have completed the protocol. At a median 287 days (range 6-1,177) after allotransplants, OS is 84% with 58% in CR and in 25% in PR. A registry analysis compared outcomes of RIC allotransplants used alone (n = 183) with those of the auto/min-allo strategy (n = 53) for newly diagnosed myeloma patients. OS at two years was 50% (95% CI: 42%-59%) for those given RIC allotransplants alone and 40% (95% CI: 23%-58%) for those given an autotransplant followed by a RIC allotransplant.
Conclusions on which of the strategies yields optimal outcomes require mature data from the IFM99-03 and IFM99-04 comparison and from the NIH (National Institutes of Health)-sponsored trial.
Repeat Autotransplants for Relapse After Initial Autotransplant
Literature searches found no prospective controlled studies comparing autotransplants with alternative strategies to manage myeloma that has relapsed after a complete or partial remission in response to a first autotransplant. This lack of controlled studies is not surprising, as patients rarely are considered eligible for a second autotransplant, either because their relapsed myeloma progresses too rapidly or because their advanced physiologic age, poor health status, or extensive prior treatment increases the likelihood of adverse outcomes. The ASBMT-sponsored evidence-based systematic review summarized data from four relevant clinical series. Investigators reported that some myeloma patients who relapsed after a first autotransplant achieved durable complete or partial remissions after a second autotransplant as salvage therapy. Factors that apparently increased the likelihood of durable remissions and extended survival included a chemosensitive relapse, younger age, a long disease-free- or progression-free-interval since the initial autotransplant, and fewer chemotherapy regimens prior to the initial autotransplant. Thus, clinical judgment, confirmed by external review, can play an important role to select patients for this treatment with a reasonable likelihood that potential benefits may exceed harms.
The policy on allogeneic SCT (AlloSCT) initially was based on a 1996 BCBSA TEC Assessment that offered the following conclusions.
- No studies directly compared the outcomes of allogeneic stem-cell transplantation (AlloSCT) with either conventional chemotherapy or AutoSCT. One retrospective study directly compared the outcomes of allogeneic transplantation with those of autologous transplantation. However this report only provided outcomes that were combined for all myeloma patients, regardless of whether their disease was responsive or refractory to treatment.
- Indirect comparisons suggest that AlloSCT is associated with a 39%–55% 5-year survival, while the comparable figure for AutoSCT is 36%–52%.
In a 1999 review on AlloSCT for myeloma, Kyle reported a mortality rate of 25% within 100 days and overall transplant-related mortality of approximately 40%. In addition, relapse of multiple myeloma was common such that few patients were cured. Candidates for AlloSCT tend to be younger than the average age of patients with multiple myeloma and in better overall condition and thus may have a better prognosis no matter what the treatment. Therefore, randomized trials were required to determine whether reported outcomes of AlloSCT were a true reflection of the treatment’s effectiveness or were due to patient selection bias.
In a 2003 update, recent reviews summarized data comparing AutoSCT with AlloSCT for newly diagnosed or responsive multiple myeloma. These included two prospective but nonrandomized controlled trials (total n = 135) and four retrospective comparisons (total n = 544), all using myeloablative conditioning regimens for those given AlloSCT. The data summarized in these reviews do not alter the conclusions of the 1996 BCBSA TEC Assessment that treatment-related toxicity after AlloSCT exceeds potential benefits from a graft-versus-myeloma effect. The recent ASBMT guideline also concludes that AutoSCT is preferred over AlloSCT based on presently available evidence. However, for the rare patient with an available syngeneic donor, data from a retrospective case-matched analysis suggest outcomes of syngeneic SCT are superior to those of AutoSCT.
Allogeneic SCT After a Failed Prior Autologous SCT
A 2000 BCBSA TEC Assessment focused on AlloSCT after a prior failed course of AutoSCT to treat various malignancies, including multiple myeloma. The TEC Assessment found that data were inadequate to permit conclusions about this treatment strategy.
In a 2003 update, a few uncontrolled series published since the 2000 BCBSA TEC Assessment reporting outcomes of AlloSCT to treat relapse or failure after an autologous transplant were described in a review article; another study was published after the review was completed. Most of the newer studies used reduced intensity or non-myeloablative conditioning regimens. These were small pilot/feasibility studies that lacked control groups for comparison of outcomes.
Additional Review of Literature Through 2008
Barlogie and colleagues published final results of the S9231 trial, initiated in 1993, which randomized 516 patients with multiple myeloma to receive either standard therapy or myeloablative conditioning with melphalan 140 mg/m2 plus total body irradiation followed by AutoSCT. The authors reported virtually no difference in outcomes, including response rates, progression-free survival, and OS. Fermand and colleagues reported a randomized study on newly diagnosed patients aged 55 to 65 years, and also reported no significant difference in OS between those randomized to myeloablative chemotherapy (melphalan 200 mg/m2 or melphalan 140 mg/m2 plus busulfan 16 mg/kg) and AutoSCT versus conventional dose chemotherapy. However, the myeloablative chemotherapy group reported a significant improvement in a composite outcome (TwiSTT) that measures time without symptoms, treatment, or treatment toxicity.
The reasons for the discrepant results compared to the positive findings of other randomized trials are uncertain, but may be related to the conditioning regimen or patient age. For example, conditioning with total body irradiation and a smaller dose of melphalan, typical of trials such as S9231 begun in the early 1990s, was later shown to be less effective than a larger dose of melphalan alone (e.g., the IFM-95 trial). A search of the clinical trials database identified a large number of ongoing trials exploring different conditioning regimens and maintenance therapies.
Research continues on many aspects of treatment for multiple myeloma, such as chemotherapy regimens and SCT. The Web site, www.clinicaltrials.gov, lists 305 studies relating to multiple myeloma currently recruiting patients. The current policy may change as additional results, especially with longer follow-up, are released. During this review, the policy statement regarding tandem SCT with an autograft followed by an allograft in younger, newly diagnosed patients was changed. This decision was based primarily on a recent study using “genetic randomization,” i.e., 80 patients with an HLA-identical sibling were allowed to chose allografts or autografts for the second transplant (58 completed an autograft/allograft sequence) and 82 without an HLA-identical sibling who were assigned to tandem autografts (46 completed the double autograft sequence). The transplantation was preceded by chemotherapy with VAD. The validity of the “genetic randomization” is open to question, but the results among those completing tandem transplantation showed a higher complete response rate at the completion of the second transplant for the autograft/allograft group (55%) than for the autograft/autograft group (26%; p = 0.004). Analyzing the group with HLA-identical siblings versus those without, in a pseudo intention-to-treat analysis, event-free survival and OS were significantly longer in the group with HLA-identical siblings. However, it is difficult to gauge the impact of the different percentages in each group who actually complete treatment. The treatment-related mortality rate at two years was 2% in the double autograft group and 10% in the autograft/allograft group; 32% of the latter group has extensive, chronic GVHD.
A meta-analysis of randomized controlled trials compared chemotherapy versus myeloablative chemotherapy with single AutoSCT. The nine trials that met the selection criteria (N = 2,411) started enrolling patients in the 1990s and include the two studies mentioned above that did not detect a survival benefit from myeloablative chemotherapy. The authors of the meta-analysis concluded that myeloablative therapy with AutoSCT increased the likelihood of progression-free survival (hazard of progression = 0.75; 95% CI: 0.59-0.96) but not OS (hazard of death = 0.92; 95% CI: 0.74-1.13); the odds ratio for treatment-related mortality was 3.01 (95% CI: 1.64-5.50) in the group with AutoSCT. However, the effects of myeloablative chemotherapy and AutoSCT may have been diluted by the fact that up to 55% of patients in the standard chemotherapy group received myeloablative chemotherapy with AutoSCT as salvage therapy when the multiple myeloma progressed. This could account for the lack of a significant difference in OS between the two groups in the study.
Finally, results of the Bologna 96 clinical study have been released, comparing single with double autologous SCT (n = 321). Patients undergoing tandem AutoSCT were more likely than those with a single SCT to attain at least a near complete response (47% vs. 33%; p = 0.008), to prolong relapse-free survival (median, 42 vs. 24 months; p < 0.001), and extend event-free survival (median, 35 vs. 23 months; p = 0.001). There was a difference between single and double AutoSCT only for patients who initially responded to VAD. There was no significant difference between the groups in treatment-related mortality (3%–4%). This study supports the use of tandem autografts; it does not address the costs and benefits of novel chemotherapy regimens versus AutoSCT, which is currently being studied, or of autografts versus allografts for the second round of SCT.
Treatment of multiple myeloma has been evolving over the past few years with the use of newer agents including bortezomib, thalidomide, and lenalidomide. A recent review noted that, “the need for early ASCT [autologous SCT] in an era of new drugs is the most important clinical question in myeloma today.”
Abdelkefi reported on a European study of 195 patients with de novo symptomatic myeloma younger than age 60 who were randomly assigned to receive either tandem autologous SCT transplantation up front (arm A, n = 97) or one AutoSCT followed by a maintenance therapy with thalidomide (day + 90, 100 mg/day during six months) (arm B, n = 98). In both arms, autologous SCT was preceded by first-line therapy with thalidomide-dexamethasone and subsequent collection of peripheral blood stem cells with myeloablative cyclophosphamide (4 g/m2) and granulocyte colony stimulating factor. With a median follow-up of 33 months (range, 6-46 months), the three-year OS was 65% in arm A and 85% in arm B (P = .04). Additional trials with longer follow-up are needed to understand the role of tandem transplants in myeloma with the use of newer agents.
For those with primary progressive disease (disease progression during induction therapy), a risk-adapted approach to treatment is recommended. This risk-adapted approach is based on several large studies that have shown that patients defined as high risk genetically do not derive durable responses to AutoSCT strategies and may relapse within one year of treatment. However, patients with primary progressive disease who do not have these characteristics are considered candidates for AutoSCT. Emerging studies suggest that early introduction of bortezomib results in better quality of response and prolonged survival for high-risk patients.
Stewart suggests an approach to separate the 25% of patients at high risk from those at standard risk. This involves the detection of t(4:14), t(14:16), or 17p deletion by FISH assay, chromosome 13 deletion or hypodiploidy by karyotyping, or plasma cell labeling index greater than 3%. Finding one abnormality identifies a patient at high risk in this approach. In addition, patients with beta-2-microglobulin levels greater than 5.5 mg per liter are often considered high risk.
Interest continues in use of allogeneic SCT in multiple myeloma. As noted in a recent review by Bensinger, despite improvements in survival for multiple myeloma patients, the disease remains incurable for most. Allogeneic SCT is potentially curative, due in part to a graft-versus-myeloma effect. However, high transplant-related mortality with AlloSCT is a major limitation to wider use of this modality. While mortality can be reduced through the use of RIC regimens, this comes at a cost of higher rates of disease progression and relapse. Studies are ongoing in an attempt to improve overall outcomes of allogeneic SCT. In addition, the risk of chronic GVHD is significant with RIC-AlloSCT.
The policy statements are in concordance with current National Comprehensive Cancer Network guidelines, which indicate that AlloSCT be performed in myeloma patients on clinical trial (except as already noted in the policy statement).
Tandem and Triple SCT, and Donor Leukocyte Infusion
Search of published literature did not locate any random, controlled studies that indicate improved patient outcomes for tandem or triple SCT, or donor leukocyte infusion.
Disclaimer for coding information on Medical Policies
Procedure and diagnosis codes on Medical Policy documents are included only as a general reference tool for each policy. They may not be all-inclusive.
The presence or absence of procedure, service, supply, device or diagnosis codes in a Medical Policy document has no relevance for determination of benefit coverage for members or reimbursement for providers. Only the written coverage position in a medical policy should be used for such determinations.
Benefit coverage determinations based on written Medical Policy coverage positions must include review of the member’s benefit contract or Summary Plan Description (SPD) for defined coverage vs. non-coverage, benefit exclusions, and benefit limitations such as dollar or duration caps.