Maintenance therapy with thalidomide improves overall survival after autologous hematopoietic progenitor cell transplantation for multiple myeloma

Authors

  • Brett T. Brinker M.D.,

    1. Division of Hematology and Oncology, Winship Cancer Institute, Emory University School of Medicine, Atlanta, Georgia
    Search for more papers by this author
  • Edmund K. Waller M.D., Ph.D.,

    1. Division of Hematology and Oncology, Winship Cancer Institute, Emory University School of Medicine, Atlanta, Georgia
    Search for more papers by this author
    • Dr. Waller is a Scholar of the Leukemia and Lymphoma Society.

  • Traci Leong Ph.D.,

    1. Division of Hematology and Oncology, Winship Cancer Institute, Emory University School of Medicine, Atlanta, Georgia
    2. Department of Biostatistics, Rollins School of Public Health, Emory University, Atlanta, Georgia
    Search for more papers by this author
  • Leonard T. Heffner, Jr. M.D.,

    1. Division of Hematology and Oncology, Winship Cancer Institute, Emory University School of Medicine, Atlanta, Georgia
    Search for more papers by this author
    • Dr. Lonial and Dr. Heffner are members of the Speakers' Bureau for Celgene.

  • Istvan Redei M.D.,

    1. Division of Hematology and Oncology, Winship Cancer Institute, Emory University School of Medicine, Atlanta, Georgia
    Search for more papers by this author
  • Amelia A. Langston M.D.,

    1. Division of Hematology and Oncology, Winship Cancer Institute, Emory University School of Medicine, Atlanta, Georgia
    Search for more papers by this author
  • Sagar Lonial M.D.

    Corresponding author
    1. Division of Hematology and Oncology, Winship Cancer Institute, Emory University School of Medicine, Atlanta, Georgia
    • Winship Cancer Institute, Emory University School of Medicine, 1365 Clifton Road, Building C, Room 4004, Atlanta, GA 30322
    Search for more papers by this author
    • Dr. Lonial and Dr. Heffner are members of the Speakers' Bureau for Celgene.

    • Dr. Lonial is a Lymphoma Research Foundation Clinical Investigator.

    • Fax: (404) 778-5530


Abstract

BACKGROUND

High-dose chemotherapy with autologous hematopoietic progenitor cell (HPC) transplantation improves survival for patients with multiple myeloma (MM); however, most patients develop recurrent disease after undergoing transplantation, and new treatment approaches are needed. The objective of this retrospective review of autologous HPC transplantation for patients with MM was to evaluate the impact of conditioning regimens and posttransplantation therapy on survival.

METHODS

The authors reviewed 112 patients with MM who received autologous HPC grafts at their institution. Between June 1992 and August 2001, 54 patients received busulfan, cyclophosphamide, and etoposide (Bu/Cy/VP-16), and 58 patients received high-dose melphalan (MEL-200) followed by autologous HPC transplantation. After transplantation, 36 patients received thalidomide for maintenance or salvage therapy, and 76 patients received no posttransplantation thalidomide.

RESULTS

At a median follow-up of 58 months, the median survival was 54 months. There was no statistically significant difference noted with regard to response to conditioning regimen, progression-free survival, or overall survival between the Bu/Cy/VP-16 and MEL-200 cohorts. Patients who received thalidomide after transplantation had improved median survival (65.5 months) compared with patients who did not receive thalidomide (44.5 months; P = .09). When they were separated according to reasons for thalidomide use, patients who received thalidomide as maintenance had improved overall survival compared with patients who received thalidomide as salvage (65 months vs. 54 months; P = .05).

CONCLUSIONS

Combination chemotherapy provided no advantage over high-dose melphalan in patients with MM who underwent autologous HPC transplantation. The posttransplantation use of thalidomide seemed to improve the survival of patients compared with historical controls from the prethalidomide era. Further prospective trials are underway to confirm the benefit of thalidomide in the posttransplantation setting. Cancer 2006. © 2006 American Cancer Society.

Multiple myeloma is a clonal disorder of malignant plasma cells that infiltrate the bone marrow, suppress polyclonal immunoglobulin production, and cause lytic bone lesions as well as renal and immune dysfunction. Despite therapeutic advances over the last decade, multiple myeloma remains an incurable disease; however, long-term disease-free survival occasionally has been achieved with allogeneic transplantation.

Historically, standard therapy for myeloma has involved combinations of alkylating agents and corticosteroids. Randomized trials comparing the use of melphalan and prednisone with multiagent infusional chemotherapy resulted in similar overall survival (OS), although the use of infusional chemotherapy yielded higher response rates.1 Results from the Intergroupe Francophone du Myelome (IFM) 90 trial demonstrated that the use of high-dose chemotherapy (HDT) followed by autologous hematopoietic progenitor cell (HPC) transplantation improved response rates, OS, and event-free survival compared with standard therapy.2 Attempts to improve the survival advantage conferred by HDT and autologous HPC transplantation with autograft purging,3–5 the use of radiation or busulfan-based regimens,6–9 or the use of tandem transplantations10–12 have not demonstrated that strategy has consistent benefit over standard, single transplantations, with the exception of the IFM 94 trial, which demonstrated a benefit for tandem transplantation over single autologous transplantation.12 Based on single-agent activity in the newly diagnosed and recurrent setting,13–17 several groups are evaluating the benefit for maintenance thalidomide after HPC transplantation.18, 19

We evaluated the impact of 2 different conditioning regimens and the use of posttransplantation thalidomide therapy on OS after HDT with either autologous bone marrow transplantation or peripheral blood stem cell transplantation. In this retrospective analysis, we have demonstrated that, although conditioning regimens containing multiple chemotherapeutic agents offered no benefit over single-agent, high-dose melphalan (MEL-200), the use of thalidomide at any time after autologous transplantation appeared to confer a survival advantage compared with similarly matched patients who did not receive posttransplantation thalidomide, and patients who received maintenance therapy had improved OS compared with patients who received thalidomide at the time of salvage therapy.

MATERIALS AND METHODS

Patients

The study population comprised 112 patients with pathologically confirmed multiple myeloma who received HDT and underwent autologous HPC transplantation between June 1992 and August 2001. Patient outcomes were updated through October 2005. Two conditioning treatment regimens were used during that period, and their clinical outcomes were compared. The first group of patients received a combination of busulfan, cyclophosphamide, and etoposide (Bu/Cy/VP-16) between June 1992 and February 1998. From March 1998 through August 2001, a second group of patients received MEL-200. All patients underwent either autologous bone marrow transplantation or blood HPC transplantation. Two different strategies for posttransplantation therapy were compared: patients who received thalidomide at any point after autologous transplantation and patients who did not receive thalidomide after transplantation.

Treatment

All patients received standard induction chemotherapy prior to undergoing transplantation with regimens like combined vincristine, doxorubicin, and dexamethasone. Blood HPC grafts were collected by apheresis after chemomobilization, which consisted of cyclophosphamide at a dose of 4 gm/m2 (given over 2 hours for 1 day) and etoposide at a dose of 200 mg/m2 (100 mg/m2 each day for 2 days) or cyclophosphamide alone in combination with 10-15 μg/kg per day of hematopoietic growth factors (granulocyte–colony-stimulating factor [G-CSF] or G-CSF plus granulocyte-macrophage–colony-stimulating factor20). Patients who received the Bu/Cy/VP-16 conditioning regimen also received targeted busulfan21, 22 (area under the curve of <1500) at a dose of 16 mg/kg (1 mg/kg every 6 hours on Days − 8, − 7, − 6, and − 5), cyclophosphamide at a dose of 60 mg/kg per day × 2 (on Days − 3 and − 2), and etoposide at a dose of 10 mg/kg per day × 3 (on Days − 4, − 3, and − 2). Patients received conditioning with melphalan at a dose of 100 mg/m2 per day × 2 (on Days − 3 and − 2). Patients in the Bu/Cy/VP-16 cohort received either autologous bone marrow or blood HPC grafts, whereas patients in the MEL-200 cohort received blood HPC transplantation. CD34-positive cell selection of autografts was performed using the Isolex 2.0 System (Baxter Health Corporation; Irvine, CA).23 Patients who underwent transplantation were housed in private rooms outfitted with high-efficiency particulate air filtration during their hospital stay, and they received standard antibacterial, antiviral, and antifungal prophylaxis.24 Blood products were administered according to standard transfusion parameters for the unit.

Thalidomide was administered in the posttransplantation period as maintenance or salvage therapy. When it was administered as maintenance therapy, thalidomide was initiated at a dose of 200 mg per day starting between 3 and 6 months after transplantation and continuing either until toxicity precluded further therapy or until patients had disease progression. When it was administered as salvage therapy, thalidomide was initiated at a dose of 200 mg per day at the time of diagnosis of disease progression. Patients with progressive disease after thalidomide and patients in the nonthalidomide cohort who had disease progression received additional treatment, which included local radiation and chemotherapy. For the purpose of statistical analysis, patients who received posttransplantation thalidomide as maintenance therapy and patients with disease progression who received posttransplantation thalidomide as salvage therapy were evaluated as a single cohort. Subsequent survival analyses for maintenance or salvage therapy with thalidomide also were performed. Thirty-eight patients received maintenance treatment with interferon-α at a dose of 3 × 106 units subcutaneously 3 times per week starting from 3 to 6 months after HPC transplantation and continuing until there was evidence of either disease progression or patient intolerance.

Response Assessment

The objective evidence of response to pretransplantation induction chemotherapy, high-dose therapy, and autologous HPC transplantation was assessed by using serum and urine protein electrophoresis with immunofixation, skeletal survey, and bone marrow aspirate and biopsy using the European Group for Blood and Marrow Transplant/Autologous Bone Marrow Transplant Registry/International Bone Marrow Transplant Registry criteria.25 Pretreatment disease status was assessed between 30 and 60 days prior to transplantation. Response to HPC transplantation was assessed between 3 and 6 months after transplantation. Patients were reevaluated at yearly intervals to determine the extent of remission.

Statistical Analysis

Differences between treatment groups with respect to patient baseline characteristics, treatment modalities, and response to induction therapy were compared by chi-square analysis or Fisher exact test when appropriate. Continuous variables were compared by using the Wilcoxon rank-sum test. OS was evaluated for the 2 conditioning cohorts (Bu/Cy/VP-16 vs. MEL-200) and for the posttransplantation treatment cohorts (thalidomide vs. no thalidomide). Progression-free survival (PFS) was evaluated only for the high-dose conditioning cohorts. Complete restaging to assess response to posttransplantation thalidomide was variable because of patient follow-up and precluded an accurate analysis of posttransplantation therapy PFS. Disease status at follow-up was determined by the last documented complete restaging evaluation to confirm continued remission. Complete restaging was not required for patients who had evidence of confirmed disease recurrence. OS was calculated from the time of HPC transplantation to death from any cause. PFS was calculated from the date of HPC transplantation to either the first date of progressive disease documented on restaging studies or the date of death. Patients who were lost to follow-up or who were alive at the associated last clinical contact were censored. Survival curves were estimated by using the Kaplan–Meier method, and differences between the curves were compared by using the log-rank and Wilcoxon tests. All P values were 2-sided, and significance was declared at the 5% level.

Univariate analyses to determine factors that had an impact on OS were calculated by using the Cox proportional hazards regression model. The baseline patient characteristics that were assessed for impact on survival included age, gender, stage at presentation, and β2-microglobulin level immediately prior to transplantation. Additional factors that were evaluated included high-dose conditioning cohort (Bu/Cy/VP-16 or MEL-200), source of stem cells (bone marrow or peripheral blood), CD34-positive cell selection of autografts (yes or no); and complete responses to transplantation, posttransplantation thalidomide, and posttransplantation interferon. The effects of thalidomide use and prognostic factors on survival were assessed by using a proportional hazards model.

RESULTS

Patient Characteristics and Treatment Modalities

The median age for the 112 patients evaluated in this study was 53 years (range, 35-71 years). Patients who received conditioning with Bu/Cy/VP-16 (n = 54 patients) underwent grafts between 1992 and 1998, whereas patients who received conditioning with MEL-200 (n = 58 patients) underwent transplantation between 1998 and 2001. Patients in each HDT cohort were similar with respect to gender, age, disease stage, and response to pretransplantation chemotherapy, as shown in Table 1. Patients in the Bu/Cy/VP-16 cohort received either a peripheral blood stem cell graft (n = 31 patients) or a bone marrow graft (n = 23 patients) compared with patients in the MEL-200 cohort, all of whom received peripheral blood stem cell grafts (n = 58 patients). More patients (n = 38) in the MEL-200 cohort received grafts who had undergone autograft purging with CD34-positive cell selection as a means of reducing tumor cell contamination. Finally, 25 patients in the Bu/Cy/VP-16 cohort received posttransplantation interferon-α as maintenance therapy compared with 29 patients in the MEL-200 cohort who received posttransplantation thalidomide either as maintenance therapy or as salvage therapy.

Table 1. Frequency of Patient Baseline Characteristics According to Treatment Modalities by High-Dose Chemotherapy Conditioning Cohort
VariableNo. of Patients (%)P Value
Bu/Cy/VP-16 (n = 54)MEL-200 (n = 58)
  • Bu/Cy/VP-16: busulfan, cyclophosphamide, and etoposide; MEL-200: high-dose melphalan; CR: complete response; PR: partial response; BM: bone marrow; PBSC: peripheral blood stem cells.

  • *

    Response to induction therapy immediately prior to autologous hematopoietic progenitor cell transplantation.

Gender (male)31 (57)30 (52).55
Age ≥55 y20 (37)26 (45).40
 Disease stage at diagnosis   
 Stage I9 (18)9 (16).69
 Stage II8 (16)13 (22) 
 Stage III33 (66)36 (62) 
 Data not available40 
Pretransplantation response*   
 CR4 (7)5 (8.5).65
 PR47 (87)52 (90) 
 Stable1 (2)1 (1.5) 
 Progression2 (4)0 (0) 
Stem cell source   
 BM23 (43)0 (0)<.001
 PBSC31 (57)58 (100) 
Autograft CD34-negative selection  
 Yes1 (2)38 (66)<.001
 No53 (98)20 (44) 
Posttransplantation therapy   
 Thalidomide7 (13)29 (50)<.001
 Interferon25 (48)13 (22) 
 Neither22 (39)16 (28) 

After high-dose conditioning and HPC transplantation, 36 patients received single-agent thalidomide, and 76 patients did not. Among the individuals who received thalidomide after HPC transplantation, 20 patients received thalidomide as maintenance therapy, and 16 patients received thalidomide as salvage therapy. The median time to the initiation of thalidomide treatment after HPC transplantation was 8 months (range, 2-76 months). The thalidomide-treated cohort and the no thalidomide cohort were similar across a broad range of characteristics, including gender, age, stage, response to pretransplantation chemotherapy, and initial response to transplantation, as shown in Table 2. Among the 76 patients who did not receive thalidomide, 56 patients underwent transplantation prior to 2000, and a greater proportion of those patients received Bu/Cy/VP-16 (62%) than patients who received posttransplantation thalidomide (19%). There was a significant difference in the frequency of conditioning regimen between each cohort. Maintenance interferon-α was administered to 38 patients (10 patients with thalidomide and 28 patients without thalidomide).

Table 2. Frequency Characteristics According to the Receipt of Posttransplantation Thalidomide Therapy
VariableNo. of Patients (%)P Value
Thalidomide (n = 36)No Thalidomide (n = 76)
  • CR: complete response; PR: partial response; BM: bone marrow; PBSC: peripheral blood stem cells; Bu/Cy/VP-16: busulfan, cyclophosphamide, and etoposide; MEL-200: high-dose melphalan.

  • *

    Initial response to autologous hematopoietic progenitor cell transplantation at 3-6 months.

Gender (male)22 (61)39 (51).33
Age ≥55 y18 (50)28 (37).19
Disease stage at diagnosis   
 Stage I6 (17)12 (16).08
 Stage II11 (31)10 (14) 
 Stage III18 (52)51 (70) 
 Data not available13 
Response to induction   
 CR4 (11)5 (7).57
 PR32 (89)67 (88) 
 Stable0 (0)2 (2.5) 
 Progression0 (0)2 (2.5) 
Stem cell source   
 BM4 (11)19 (25).09
 PBSC32 (89)57 (75) 
Autograft CD34-negative selection  
 Yes16 (44)23 (30).14
 No20 (56)53 (70) 
Conditioning regimen   
 Bu/Cy/VP-167 (19)47 (62)<.001
 Mel-20029 (81)29 (38) 
Response to transplantation*   
 CR12 (35)18 (27).58
 PR11 (32)28 (42) 
 Stable5 (15)6 (9) 
 Progression6 (18)15 (22) 
 Data not available29 
Maintenance interferon   
 Yes10 (28)28 (37).52
 No26 (72)48 (63) 

Response to Autologous HPC Transplantation

Among 112 study participants, 101 patients had complete staging evaluations after autologous transplantation and were evaluable for disease response. Eleven patients had incomplete staging evaluations within the 6-month posttransplantation interval and could not be evaluated. The response at 3 to 6 months after transplantation for each high-dose conditioning cohort is shown in Table 3. For the evaluable HPC graft recipients, 30 patients (29.7%) achieved a complete response and 39 patients (38.6%) achieved a partial response to transplantation within 6 months. Progressive disease was observed in 21 patients (21%). There were no significant differences noted between the MEL-200 group and the Bu/Cy/VP-16 group with respect to response of myeloma to HPC transplantation.

Table 3. 3-Month to 6-Month Response to High-Dose Chemotherapy and Autologous Hematopoietic Progenitor Cell Transplantation
ResponseNo. of Patients (%)P Value
Bu/Cy/VP-16 (n = 54)MEL-200 (n = 58)
  1. Bu/Cy/VP-16: busulfan, cyclophosphamide, and etoposide; MEL-200: high-dose melphalan.

Complete response15 (30)15 (29).32
Partial response19 (38)20 (39) 
Stable3 (6)8 (16) 
Progression13 (26)8 (16) 
Data not available47 

Survival

The median duration of follow-up for the surviving patient population was 58.1 months (range, 35.2-137.9 months). Sixty-nine patients died during the study period. The Kaplan–Meier estimate for median OS is presented in Figure 1A. The median OS was 53.9 months (95% confidence interval [95% CI], 43.5-73.1 months) with an estimated 3-year survival rate of 0.64 (95% CI, 0.59-0.70). The median PFS is shown in Figure 1B and was 14.0 months (95% CI, 12.5-24.6 months), and 77% of patients had progressive disease during the study period.

Figure 1.

(A) The probability of overall survival for all patients (n = 112). (B) The probability of progression-free survival for all patients (n = 112).

The OS stratified by each high-dose conditioning regimen cohort is presented in Figure 2. The median follow-up for all patients who received conditioning with Bu/Cy/VP-16 was 44 months (range, 6.6-138.0 months), and the median follow-up for all patients who received conditioning with MEL-200 was 47 months (range, 4.7-80.0 months). The groups that received Bu/Cy/VP-16 or MEL-200 were balanced with respect to the number of prior therapies, response to induction, and response to HDT. The only statistically different strata between the 2 HDT cohorts were the stem cell source, CD34 selection, and the type of maintenance therapy (Table 1). There were 41 deaths in the Bu/Cy/VP-16 cohort and 28 deaths in the MEL-200 cohort. The 3-year probability of survival was 0.57 (95% CI, 0.50-0.66) for patients who received Bu/Cy/VP-16, and 0.71 (95% CI, 0.63-0.70) for patients who received MEL-200. The difference in OS was not significant (P = .44). There also was no difference in PFS between the 2 high-dose cohorts (P = .13).

Figure 2.

The probability of survival according to high-dose conditioning regimen: busulfan, cyclophosphamide, and etoposide (Bu/Cy/VP-16) versus high-dose melphalan (MEL-200). Vertical segments represent censored patients. NS: not significant.

Posttransplantation therapy with thalidomide or interferon was evaluated for impact on patient outcomes. Again, the 2 groups were balanced well for response assessment after transplantation, and for response to induction therapy, and for disease status at the time of transplantation. The only statistically difference between the maintenance cohorts were related to the type of conditioning used in each group, with more patients in the MEL-200 group receiving thalidomide posttransplantation (Table 2). The Kaplan–Meier estimate of OS based on the presence or absence of the use of posttransplantation thalidomide is shown in Figure 3. The median survival for patients who received thalidomide was 65 months (95% CI, 33.5-114.0 months). Patients who did not receive thalidomide had a median survival of 46 months (95% CI, 28.7-53.9 months). The probability of 3-year survival for the thalidomide group was 0.67 (95% CI, 0.46-0.88) versus 0.54 (95% CI, 0.42-0.67) among patients in the no thalidomide group. The difference in OS trended toward significance (P = .09; Wilcoxon test). Treatment with posttransplantation interferon-α did not improve OS compared with patients who did not receive interferon after transplantation (P = .31).

Figure 3.

The probability of survival according to posttransplantation therapy cohort (thalidomide vs. nonthalidomide). Vertical segments represent censored patients.

When the reason for thalidomide use was evaluated with respect to OS, the 36 patients were evaluated based on thalidomide either as salvage therapy (n = 16 patients) or as maintenance therapy (n = 20 patients). The groups were well balanced with respect to response to transplantation and other potential confounding variables between these 2 subsets. OS was improved for patients who received thalidomide as maintenance therapy compared with patients who received thalidomide as salvage therapy when patients were evaluated from the time of thalidomide initiation (P = .05) (Fig. 4).

Figure 4.

The probability of survival from the time of thalidomide initiation (maintenance vs. salvage therapy).

The patient baseline characteristics and treatment modalities that were assessed for impact on OS in the univariate model are shown in Table 4. In the univariate analysis, younger age (P = .04) was associated with improved OS. The use of thalidomide in the posttransplantation setting also appeared to enhance OS and trended toward significance (Wilcoxon test; P = .09) (Fig. 3). Because of the multiple confounding variables between these historic cohorts, a proportional hazards model and analyses were conducted to eliminate the historic and treatment-related imbalances between the 2 groups. Using this model with extended follow-up, only age < amedian of 53 years (P = .0012) was found to have a favorable effect on OS. Gender, disease stage at presentation, β2-microglobulin level <3.5 μg/mL, high-dose conditioning regimen, source of stem cells, CD34-positive cell selection of autografts, complete response to transplantation, and posttransplantation interferon α did not have an impact on survival.

Table 4. Median Overall Survival and Univariate Analysis for Association with Survival by Patient Baseline Characteristics
Characteristic*No. of PatientsMedian Survival (Months)P Value
  • HDT: high-dose chemotherapy; HPC: hematopoietic progenitor cell; Bu/Cy/VP-16: busulfan, cyclophosphamide, and etoposide; MEL-200: high-dose melphalan.

  • *

    Variables that were not included in the table because of statistical insignificance were CD34-negative stem cell selection of HPC grafts, source of stem cells (bone marrow vs. peripheral blood), and complete response to HPC transplantation.

  • There were 9 patients with incomplete data.

All patients11253.9
Gender   
 Male6153.9.52
 Female5151.6 
Age, y   
 <535562.1.036
 ≥535744.6 
Disease stage   
 Stage I-II3947.2.65
 Stage III6957.7 
β2 microglobulin, μg/mL   
 <3.59151.6.78
 ≥3.51265.5 
HDT and HPC transplantation   
 Bu/Cy/VP-165444.44
 MEL-2005846 
Posttransplantation interferon   
 Yes3851.6.31
 No7232.6 
Posttransplantation thalidomide   
 Yes3665.5.09
 No7644.6 

Toxicity

Treatment-related mortality was similar between the 2 HDT conditioning arms, with 3 deaths prior to Day 100 after transplantation in the Bu/Cy/VP-16 arm and no deaths in the MEL-200 cohort. The median time to hematopoietic engraftment also was similar between the 2 transplantation cohorts (data not shown).

The median dose of thalidomide after transplantation was 200 mg per day (range, 50-450 mg per day), and the median duration of treatment was 8.3 months (range, 1-40 months). For patients who received thalidomide as maintenance therapy, the median length of treatment was 10.1 months (range, 1-40 months), and it was 7.7 months (range, 2.1-39 months) for patients who received thalidomide as salvage therapy. The difference in duration of thalidomide treatment between the salvage and maintenance groups was not statistically significant. Toxicities related to thalidomide are shown in Table 5. Approximately 44% of thalidomide recipients did not have adverse effects. The most common complications associated with thalidomide were neuropathy (28%), fatigue (19%), and constipation (11%). Thalidomide therapy was discontinued because of patient intolerance in 6 patients (17%), with neuropathy the most common cause for discontinuation.

Table 5. Frequency of Toxicities from Thalidomide
SymptomFrequency (%)Discontinued for Toxicity
Neuropathy10 (28)4
Fatigue7 (19)1
Constipation4 (11)0
Dizziness2 (6)0
Depression1 (3)0
Pancytopenia1 (3)0
Rash1 (3)1
No adverse effects16 (44)0

DISCUSSION

To our knowledge, the current study provides the first data set that suggests a survival benefit for the use of posttransplantation thalidomide. Two key features render this a compelling finding. First, although this analysis is retrospective and potentially susceptible to time and treatment bias, the baseline characteristics, risk strata, and response to high-dose therapy were similar among patients in both groups. Second, patients in both cohorts received salvage therapy at the time of recurrence after transplantation with the exception of patients in the nonthalidomide group, who did not receive single-agent thalidomide at any time point. Therefore, it was not that patients in the nonthalidomide group did not receive salvage therapy at recurrence; they simply received salvage therapy that did not involve the use of thalidomide. Thalidomide was tolerated well among our patients when appropriate dose adjustments were made to address the well described toxicities of therapy. Only 6 patients had therapy discontinued because of intolerable side effects. In fact, many patients continued to receive thalidomide for >1 year with manageable toxicities. What is encouraging is the finding that patients who received maintenance thalidomide appeared to have an improved OS compared with patients who received salvage thalidomide. It is unlikely that this was related to the duration of therapy, because patients received thalidomide for a similar duration in both arms (10.0 months maintenance and 7.7 months salvage; P = .54).

The OS curves for the thalidomide group versus the nonthalidomide group (Fig. 3) show a clear improvement in OS for the thalidomide group that appears to disappear at 70 months. Because, currently, no therapy is curative for myeloma, it is not surprising that the curves eventually do cross, especially with such prolonged follow-up. These data, taken in concert, support the use of posttransplantation thalidomide and further suggest that the use of maintenance therapy is associated with improved OS.

The current analysis also confirmed that the use of combination intense chemotherapy as conditioning for autologous HPC transplantation does not confer an advantage over the use of single-agent, high-dose melphalan. In terms of initial response to HPC transplantation, PFS, OS, and toxicity, there were no statistically significant differences noted between patients who received conditioning with Bu/Cy/VP-16 or MEL-200. Because MEL-200 is tolerated well, has a low rate of transplantation-related complications (0 early deaths in the current analysis compared with 3 early deaths in the Bu/Cy/VP-16 group), and is relatively easy to administer, it is standard practice to use melphalan-based conditioning. Furthermore, other pretransplantation covariates, including CD34-positive cell selection of autografts, autologous graft source, initial stage, and complete response to induction therapy, were not associated with OS in univariate and multivariate modeling.

Although the current study was not randomized, the univariate and proportional hazards model analyses did account for differences between the high-dose cohorts and between the thalidomide and nonthalidomide groups. Although there is the potential for selection bias in a retrospective analysis, patients who received thalidomide in the current study were similar to patients who did not receive thalidomide, except that thalidomide was not available to patients who underwent transplantation in the earlier cohort. Another concern is that the observed survival benefit associated with thalidomide may be subject to bias because of the variation in timing of treatment initiation after transplantation. This certainly is possible; however, the current analysis suggests that maintenance therapy is superior to salvage therapy with respect to OS. When this approach is taken with the use of rituximab maintenance, OS between the planned maintenance therapy group versus the retreatment group suggests no real difference in duration of rituximab sensitivity or OS; however, these results suggest that the same is not the true for posttransplantation thalidomide use.

HDT followed by autologous HPC transplantation has become the standard of care for most patients age < 65 years with myeloma based on the observations of several large, randomized clinical trials.2, 26 Although the results of transplantation clearly are superior to standard-dose therapy for most patients, transplantation is not a curative strategy.27 Attempts to improve on the benefits observed with HDT and autologous HPC transplantation through the choice of conditioning regimen, stem cell source, and purging autografts of malignant cells have not produced significant improvements in event-free survival or OS.28 Several investigators have noted the correlation between initial response to transplantation and survival.29, 30 Strategies directed at targeting minimal residual disease after transplantation may be more successful with less toxicity than standard salvage approaches. Berenson et al. first demonstrated a survival advantage to alternate-day corticosteroids after induction therapy in a Southwest Oncology Group (SWOG) study.31 Posttransplantation therapy with maintenance interferon was examined in a meta-analysis and demonstrated a modest survival benefit.32 However, similar to corticosteroids, interferon was associated with significant toxicity and patient compliance issues. Furthermore, a recent SWOG randomized trial demonstrated no benefit for patients who were randomized to receive maintenance interferon after transplantation.33 The activity of thalidomide as a single agent in patients with recurrent or refractory myeloma makes it an attractive candidate for further analysis as maintenance therapy. Through antiangiogenic effects by inhibiting fibroblast growth factor and vascular endothelial growth factor,34, 35 immunomodulatory properties that alter the activity of tumor necrosis factor-α,36 enhanced cytotoxic T-cell activity,37 and alterations in cell adhesion molecule activity,38 thalidomide has now demonstrated favorable response rates in various settings for patients with myeloma.

Recently, Attal et al. reported results of the IFM 99-02 trial in which patients were randomize to receive maintenance thalidomide or placebo. The group that was randomized to receive thalidomide (100 mg per day) had an improved PFS and median progression; however, to our knowledge, no survival benefit has been reported to date.18 Potential improvements in OS because of maintenance thalidomide may have been negated by the finding that patients in the placebo group received thalidomide as salvage therapy. In this historic cohort analysis, we compared OS in the prethalidomide era and the postthalidomide era and demonstrated improved OS for patients who received thalidomide at any time in the posttransplantation period, and we observed a clear difference in OS for maintenance therapy over salvage therapy with thalidomide.

This single-center, retrospective analysis evaluated the potential benefit of mutliagent, busulfan-based chemotherapy conditioning and posttransplantation thalidomide on the survival of patients after transplantation. The data support the use of high-dose melphalan over combination chemotherapy as conditioning prior to HPC transplantation for patients with myeloma. The current results also extend emerging data demonstrating the activity of thalidomide in the posttransplantation setting by exhibiting a survival advantage for the use of thalidomide as either salvage or maintenance therapy posttransplantation, although the current results suggest that salvage therapy is inferior. However, several questions remain with respect to the use of thalidomide, including the timing of initiation, dosing schemes (continuous vs. intermittent), and the duration of therapy. Additional prospective, randomized studies will be needed to test the efficacy of maintenance thalidomide therapy and to confirm our data.

Ancillary