Document Ozg4O0jZe6g6eEVNb44Zda3jw

Autografting Bone Marrow Transplantation (2001) 27, 821828 2001 Nature Publishing Group All rights reserved 02683369/01 $15.00 www.nature.com/bmt Thiotepa, busulfan, cyclophosphamide (TBC) and autologous hematopoietic transplantation: an intensive regimen for the treatment of multiple myeloma A Shimoni1, TL Smith2, A Aleman1, D Weber3, M Dimopoulos3, P Anderlini1, B Andersson1, D Claxton1, NT Ueno1, I Khouri1, M Donato1, M Korbling1, R Alexanian3, R Champlin1 and S Giralt1 1Department of Blood and Bone Marrow Transplantation, 3Department of Lymphoma and Myeloma, and 2Department of Biostatistics, The University of Texas, MD Anderson Cancer Center, Houston, TX, USA Summary: The study was designed to evaluate the efficacy and safety of an intensive, tri-alkylator conditioning regimen, consisting of thiotepa, busulfan and cyclophosphamide (TBC), prior to autologous hematopoietic cell transplantation in patients with multiple myeloma (MM) and to analyze factors associated with outcome. One hundred and twenty patients with MM received high-dose chemotherapy with TBC followed by autologous bone marrow (n = 24) or peripheral blood stem cell (PBSC) transplantation (n = 96). Fifty-four patients had chemosensitive disease and 66 had refractory disease at the time of transplantation. The overall response rate was 81% and the complete remission (CR) rate was 26%. Patients with chemosensitive disease had a CR rate of 52% vs 5% for patients with refractory disease. Multivariable analysis determined disease status at transplant as the factor most likely associated with long survival. Estimated median survival was 48, 35 and 9 months for patients with chemosensitive, primary refractory or disease in refractory relapse, respectively. Short interval from diagnosis to transplant among patients with primary refractory disease and younger age were also favorable prognostic factors for survival. Patients with refractory disease pre-transplant who achieved remission criteria rapidly after treatment had a worse outcome than the slow responders. Treatmentrelated mortality with the introduction of PBSC and better supportive care was 4.8%. In conclusion, TBC is an effective and relatively well-tolerated intensive conditioning regimen in patients with MM. A more favorable outcome was observed in patients with chemosensitive disease and with early treatment for primary refractory disease. TBC merits further study in these subgroups and comparison with alternative regimens in Correspondence: Dr S Giralt, Department of Blood and Bone Marrow Transplantation, University of Texas MD Anderson Cancer Center, 1515 Holcombe Blvd, Houston, TX 77030, Box 65, USA Received 12 October 2000; accepted 7 February 2001 prospective studies is warranted. Bone Marrow Transplantation (2001) 27, 821828. Keywords: multiple myeloma; autologous transplantation; response kinetics High-dose therapy followed by autologous bone marrow or peripheral blood stem cell transplantation has emerged as an effective treatment for many patients with multiple myeloma. Standard chemotherapy results in a complete remission (CR) rate of 5%, a median survival of 3 years, and 10% of patients surviving longer than 10 years.1,2 Studies have shown that survival is improved with high-dose chemotherapy.317 Disease status prior to transplant and time to transplant have been identified as very important prognostic factors for survival.1215 Age, 2-microglobulin level, and abnormalities of chromosome 13 have also been shown to be predictive.319 The kinetics of response after conventional chemotherapy (eg the time to achieve partial or complete remission) has been reported to be an important prognostic factor for treatment outcome by some, but not all investigators.2022 The kinetics of response after high-dose chemotherapy has not been thoroughly investigated and its prognostic significance remains unknown. In this study we report our long-term experience with a high-dose trialkylator regimen consisting of thiotepa, busulfan and cyclophosphamide for the treatment of multiple myeloma. The aim of the analysis was to define prognostic factors for survival as well as to determine if kinetics of response were important in prognosis. Patients and methods Eligibility criteria Patients were eligible for the study if they had symptomatic myeloma, stage II or III by Durie criteria.23 Patients were required to be younger than 70 years with a Zubrod performance status 2, left ventricular ejection fraction 50% High-dose chemotherapy in myeloma A Shimoni et al 822 with no uncontrolled arrhythmia, adequate pulmonary func- ing dose of 1 106 units administered subcutaneously three tion, serum creatinine 2 mg/dl, serum bilirubin 2 mg/dl times per week and started when platelet count was and no evidence of chronic liver disease. The study was 50 109/l. In addition, dexamethasone 20 mg/m2 was approved by the University of Texas MD Anderson Cancer administered orally for 4 days each month. Ninety patients Center Institutional Review Board and all patients provided received maintenance therapy. Twenty-three were not written informed consent according to institutional guide- treated due to early death or progression and seven for other lines. reasons. Ten of these patients were treated with interferon alone and three with dexamethasone alone for various Peripheral blood stem cell/Bone marrow collection reasons. This maintenance regimen was usually continued until there was evidence of disease progression or patient High-dose chemotherapy was supported by autologous per- intolerance. ipheral blood stem cells in 96 patients and by autologous bone marrow in 24 patients. Mobilization chemotherapy consisted of cyclophosphamide alone, cyclophosphamide Criteria for response and etoposide combination24 or fractionated cyclophos- Response to induction therapy and to transplant was phamide, vincristine, doxorubicin and dexamethasone assessed by the criteria developed by the EBMT/ (hyperCVAD)25 followed by growth factor stimulation. ABMTR/IBMTR for evaluation of response.27 Complete Apheresis was performed with the Cobe Spectra continuous response (CR) was defined as the absence of monoclonal flow blood cell separator (Cobe, Lakewood, CO, USA) protein in serum and urine by both immunoelectrophoresis using standard techniques. The target dose was at least and immunofixation, a normal bone marrow and no 2 106 CD34+ cells/kg and patients were allowed to have increase in size or number of bone lesions. Partial response bone marrow harvest to supplement peripheral blood stem (PR) was defined as reduction of more than 50% in the cell collection if this target was not reached. level of serum monoclonal protein or 90% of urinary light Twenty-four patients had autologous bone marrow har- chain excretion (or to less than 200 mg per day). All vests using conventional techniques, of whom 11 patients responses had to be maintained for at least 6 weeks. For underwent B cell purging with anti-CD19 monoclonal anti- patients with non-secretory disease responses were evalu- body and immuno-magnetic beads using previously ated by serial bone marrow aspirations. CR required less described techniques.14 than 5% plasma cells in the marrow and PR required 50% reduction of marrow plasmacytosis. No response was Preparative regimen and supportive care defined when the above criteria were not fulfilled. Relapse or progression required 25% increase in the level of para- The preparative regimen consisted of thiotepa 250 mg/m2 protein on two consecutive tests or reappearance of over 4 h given intravenously on days -9, -8, -7; busulfan paraprotein after CR. For assessment of response in patients 1 mg/kg given orally every 6 h on days -6, -5, -4 for a with refractory disease, post-transplant paraprotein levels total of 10 doses; cyclophosphamide 60 mg/kg intra- were compared with the immediate pre-transplant level. venously over 1 h on days -3, -2. Mesna 10 mg/kg was When transplantation was performed as consolidation of administered prior to the first dose of cyclophosphamide prior response post-transplant paraprotein levels were com- and then at 60 mg/kg/day by continuous infusion for 48 h. pared with the level prior to the start of initial chemo- Since November 1994 patients were enrolled on a study therapy. The post-transplant response in patients with investigating cyclosporine for induction of autologous chemosensitive disease was by definition at least PR, unless graft-versus-host disease. Forty-nine patients received they progressed or died within 6 weeks. Patients were cyclosporine starting on the day of stem cell infusion at a evaluated for response at approximately 1, 2, 3, 6, 9 and dose of 12 mg/kg per day initially by continuous intra- 12 months after transplantation and at 6 month intervals venous infusion and then orally for 28 days as previously thereafter. described.26 All patients were treated in private rooms. The prophylactic antibiotic regimen included oral norfloxacin (400 mg Statistical analysis twice per day), penicillin V (500 mg four times per day), The primary outcome variable for this study was overall fluconazole (200 mg per day), and valacyclovir (500 mg per survival (OS), defined from the day of autologous trans- day) or acyclovir. All were administered orally until hema- plantation until death from any cause. Event-free survival topoietic recovery. Trimethoprin sulfamethoxazole was (EFS) was defined as the time from the day of transplan- given orally for prevention of Pneumocystis carinii twice tation until either disease progression or death from any per day until day -2 and was restarted after hematopoietic cause. Time to engraftment and treatment-related mortality recovery for 6 months. Broad-spectrum intravenous anti- were secondary end-points. Survival time distributions biotics were initiated for temperature greater then 38.3C were estimated using the KaplanMeier method.28 Distri- or for any sign of infection. Filgastrim 5 g/kg/day was butions were compared between subsets of patients using administered subcutaneously beginning 1 day after periph- the log-rank test. A tree-structured survival analysis was eral blood stem cell/ bone marrow infusion and was con- employed in order to construct a decision tree relating mul- tinued until engraftment. Blood products were transfused tiple pre-treatment patient characteristics to survival out- according to standard criteria. come.29 The method is computer intensive and is a non- Maintenance therapy included alpha interferon at a start- parametric exploratory tool since it does not depend on any Bone Marrow Transplantation underlying distributional results. Continuous measures and categorical data were considered simultaneously; factors to be included were decided on the basis of a preliminary screening of individual factors. Initially, the program determines the optimal cutpoint of each possible predictor variable and then compares among variables to select the one which optimally splits the population into two subgroups most different in survival outcomes. The process is repeated in the resulting subgroups until no further partitioning is warranted. A sample re-use method termed cross-validation was used to confirm the final tree structure; this method deletes a random patient subset from the data, constructs a tree on the remaining data, and uses the group left out as new data on which to test the tree. As further confirmation of the multivariable results, a multiple regression analysis based on a proportional hazards model30 was carried out with continuous-measure variables dichotomized either using cutpoints suggested by conventional use or by the tree analysis. Results Patient and disease characteristics Between June 1991 and June 1998, 120 consecutive patients with multiple myeloma were included in the study. Patient and disease characteristics are shown in Table 1. Median age was 48 years (range 3267). Fifty-two were Durie stage II and 68 stage III on presentation. The abnormal protein was IgG in 75 patients and IgA in 19. Median time from initial treatment to transplant was 330 days (range 883350) and patients received a median of 3 (range High-dose chemotherapy in myeloma A Shimoni et al 15) prior treatments including the chemotherapy used for stem cell mobilization. Fifty-four patients were treated during first remission (43 patients) or chemosensitive relapse (11 patients). Forty-five patients were treated during primary refractory disease and 21 during chemo-refractory relapse. Patients with refractory disease had more prior treatments, longer interval from diagnosis, and a higher 2microglobulin level. Patients with refractory relapse were more likely to receive bone marrow that was stored in prior remission rather than peripheral blood stem cells for cellular support. Engraftment and toxicity The median time to ANC 0.5 109/l was 10 days (range 721) in those receiving peripheral blood stem cells vs 15 days (range 1026) in those receiving bone marrow (P 0.001). The median time to platelet count above 20 109/l was 11 days (range 667) and 29 days (range 1246), respectively (P = 0.002). None of the other pretransplant patient characteristics outlined in Table 1 predicted for a more rapid engraftment. Sixteen patients (13%, 95% CI 821%) died of treatment-related causes. Six patients died of multi-organ failure with or without superimposed infection, one died of intracranial bleeding, nine patients died of infections including three patients who died of respiratory syncytial virus pneumonia. Univariate analysis indicated that treatment-related mortality (TRM) was associated with age above 55 years, with the use of bone marrow as stem cell source and with an earlier transplant year (before 1994; Table 2). Seven additional patients (6%) experienced non-fatal grade III 823 Table 1 Summary of patient characteristics by disease status at transplant Total Remission No. Age 55 Sex (% male) Stage III Heavy chain IgG IgA BJPb IgD/NSc Light chain Prior tx 3 Time to tx (days) 1 year Source of stem cell (% BM)d Cyclosporinee LDHf 625 2 microglobulin 3 mg/dl Creatinine 120 48 (3267)a 18% 64% 57% 66% 17% 9% 9% 29% 50% 330 (883350) 58% 20% 42% 480 (1762023) 23% 3.0 (1.02.7) 50% 0.9 (0.51.8) 54 47 (3267) 15% 68% 53% 55% 20% 14% 10% 36% 30% 275 (883350) 72% 19% 43% 487 (1761624) 24% 2.3 (1.15.5) 23% 0.9 (0.51.8) aEntries in table are median (range) unless otherwise noted. bBence Jones proteinuria only. cIgD or non-secretory myeloma. dBone marrow as opposed to peripheral blood as the source of stem cells. eReceived cyclosporine for induction of autologous GVHD. fUpper normal limit in the institution is 618 U/l. Primary refractory 45 48 (3965) 20% 58% 57% 75% 11% 5% 9% 25% 62% 342 (882330) 53% 9% 44% 425 (2322023) 16% 3.5 (1.612.7) 73% 0.9 (0.51.7) Refractory relapse 21 48 (3459) 24% 67% 67% 71% 19% 5% 5% 18% 76% 520 (1612920) 33% 48% 33% 467 (2021450) 37% 3.5 (1.011.6) 58% 0.9 (0.51.5) Bone Marrow Transplantation High-dose chemotherapy in myeloma A Shimoni et al 824 Table 2 Treatment-related mortality by selected patient characteristics Prognostic factors for survival Characteristic Total Age 55 years 55 years Disease status Remission Primary refractory Refractory relapse Source of stem cells Bone marrow Peripheral blood Time to transplant 1 year 1 year Year of transplant 1994 1994 2 microglobulin 3 mg/dl 3 mg/dl Total No. Early deaths P value No. (%) 120 16 (13) 98 9 (9) 0.02 22 7 (32) 54 4 (7) 45 7 (16) 0.06 21 5 (24) 24 10 (42) 0.001 96 6 (6) 70 5 (8) 0.04 50 11 (22) 59 13 (22) 0.04 61 3 (5) 71 10 (14) NS 49 6 (12) toxicities by the Bearman criteria.22 Three patients had hemorrhagic cystitis and required invasive treatment. This toxicity is unique to this regimen and related to high-dose cyclophosphamide. Two patients developed grade III GI toxicity, one patient hepatic toxicity and one lung toxicity. All grade III toxicities were reversible. Table 3 summarizes the estimated survival by patient and disease characteristics measured prior to the start of highdose chemotherapy. Patients for whom therapy was initiated while they were in first remission or responsive relapse had the longest survival, on average, while those with disease in refractory relapse had the shortest. Patients with primary refractory disease had survival that was intermediate between these groups. The estimated median survival of patients with chemosensitive disease was 48 months (95% CI 44 months to not available). The survival rates of patients in first remission and sensitive relapse were similar (data not shown). The estimated median survival was 35 months (95% CI 2451 months) for patients with primary refractory disease and 9 months (95% CI 421 months) for patients with refractory relapse. The estimated median EFS was 23 months (95% CI 1837 months), 10 months (95% CI 519 months), and 5 months (95% CI 39 months), respectively for each of these groups. The estimated 5 year survival is 46% for responsive disease, 26% for primary refractory, and 0% for refractory relapse. The estimates for 5 year EFS are 22%, 9% and 0%, respectively. A tree-structured survival analysis was employed in order to consider the association of multiple patient characteristics with survival. With this approach it was possible to consider interaction among factors, to include continuous variables without specification of cutpoints, and to include factors which were unknown for some cases. Factors considered in the analysis were: age, stage of disease, serum Response to high-dose chemotherapy Of 54 patients with chemosensitive disease, 28 patients (52%) achieved CR including six patients in CR prior to transplant. Twenty-three patients (43%) remained in PR after the transplant with a median further reduction of myeloma protein of 70% (range 099%) in comparison with the pretransplant level. Three patients in this group (5%) were considered non-responders because of death or progression within 6 weeks of transplant. The overall response rate of patients with primary refractory disease was 67% and that of patients with refractory relapse was 76%. The CR rate in these two groups of patients with refractory disease was 5%. Survival rates after high-dose chemotherapy The median follow-up of surviving patients was 29 months (range 7 months to 7.5 years). The estimated 5-year survival for the total group was 29% (95% CI 2042%). The estimated 5-year EFS for the total group was 12% (95% CI 822%). Survival was highest among patients achieving CR. These patients had a median survival of 59 months. Actuarial survival and progression-free survival were 56% and 32% at 5 years. Partial remission was associated with a median survival of 30 months, 25% of patients achieving a partial remission are alive and 10% progression-free at 5 years. The median survival of non-responders was only 11 months (P = 0.001). Table 3 Summary of survival by selected pretransplant characteristics Percentiles of survival time (months) 75% 50% 25% P value Total Disease status remission primary refractory refractory relapse Stem cell source peripheral blood bone marrow Age 55 55 LDH low high Prior treatments 3 3 2 microglobulin 3.0 mg/dl 3.0 mg/dl Time to transplant 1 year 1 year Cyclosporine treated non-treated 9 38 80 30 48 NA 0.01 7 35 60 3 9 20 20 38 NA 0.01 3 8 45 14 38 NA 0.03 3 20 45 20 38 80 0.04 4 12 29 20 45 80 0.06 6 27 54 11 44 80 0.06 2 26 51 20 39 80 0.22 4 27 60 12 28 47 0.91 8 38 NA Bone Marrow Transplantation protein type, disease status, interval from diagnosis to transplant, source of stem cells, pretransplant LDH and 2microglobulin serum levels. The initial split in the tree was to divide patients into the three disease status categories. The only division within these subsets that could be crossvalidated was the interval between diagnosis and transplant among patients with primary refractory disease (Figure 1). The tree analysis determined 134 days as the cutpoint. To be clinically relevant, we determined 6 months to be the cutpoint for improved survival. This added only three patients to the group and did not change the statistical significance. Although there were only 10 patients with primary refractory disease who were treated within 6 months of diagnosis they had superior survival (lasting 2+ to 7+ years). There was no step-wise shortening of survival for those transplanted beyond 6 months, suggesting that the advantage of early transplantation was limited to those patients with primary refractory disease transplanted within 6 months of starting treatment. Survival curves for the four groups of patients selected by the multivariable tree analysis are plotted in Figure 2. There was a wide separation among the survival curves. As the second approach to considering the association of multiple patient characteristics with survival, a proportional hazards model, was fitted using a stepwise procedure. The same potential factors were included in the analysis. Disease status was determined as the factor most associated with survival. The only additional factor for which there was evidence of association with survival after considering disease status was age. The tree-structured analysis did not identify any age cutoff as a prognostic factor. Age 55 years was used as the cut-off for the proportional hazards model and was found to be a significant adverse prognostic factor for survival by this model. Total n = 120 Disease response: remission, primary refractory n = 99 Disease response: refractory relapse n = 21 Disease response: n = 54 Disease response: primary refractory n = 45 Primary refractory time to transplant: < 6 months n = 10 Primary refractory time to transplant: >6 months n = 35 Figure 1 Tree-structured survival analysis of 120 patients with multiple myeloma after TBC and autologous hematopoietic transplantation. Four groups were identified: chemosensitive disease (n = 54), primary refractory transplanted within 6 months (n = 10) or later than 6 months (n = 35) from diagnosis, and patients in refractory relapse. High-dose chemotherapy in myeloma A Shimoni et al Proportion surviving 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 Refractory relapse 0 12 24 36 Primary refractory early transplant Chemosensitive 48 60 Months Primary refractory late transplant 72 84 96 Figure 2 KaplanMeier curves of survival from the day of transplant by the risk groups determined by a tree-structured analysis. The four risk groups are as outlined in Figure 1. Kinetics of response Increasing response rates were observed over time after transplant while patients were on maintenance treatment alone. Twenty-four patients not at CR prior to transplant ultimately achieved CR. CR was achieved in a median of 146 days (range 15536). Interestingly, four patients (17%) achieved CR between 6 and 12 months after transplant and two patients (8%) achieved CR more than 1 year after transplant while on maintenance treatment alone (Figure 3). Forty-six of the 66 patients with refractory disease, either primary or refractory relapse, achieved a response (PR or CR) after transplant (70%). Criteria for response were achieved in a median of 30 days (range 1501). The median time to response was 56 days (range 1501) for patients with primary refractory disease and 14 days (range 4197) for refractory relapse. Forty patients achieved their response within 6 months of transplant (87%), and only three patients (6%) achieved their response after more than 12 months from the day of transplant. The kinetics of response in patients with refractory dis- Cumulative response (%) Refractory Chemosensitive 100 90 PR 80 CR 70 60 50 40 30 20 10 0 pre 1 3 6 12 >12 pre 1 3 6 12 >12 Months after BMT Figure 3 Cumulative response over time after hematopoietic transplantation. PR and CR rates are depicted in a cumulative fashion and show gradual increase in time. Responses are shown separately for patients with chemosensitive (n = 54) and patients with refractory disease (n = 66). Patients with chemosensitive disease were in at least PR after the transplant except for three patients who died or progressed within the first 6 weeks. 825 Bone Marrow Transplantation High-dose chemotherapy in myeloma A Shimoni et al 826 ease was found to correlate with survival. In Figure 4, sur- istics determined the disease status at transplant as the fac- vival subsequent to 1 month on treatment (landmark time) tor most highly associated with longer survival. Estimated is plotted for those achieving response at 1 month and those median survival was 48 months for patients with chemosen- still considered non-responders at that time. Estimated sitive disease and only 9 months for patients in refractory median subsequent survival for the 25 patients in the relapse. These results are comparable to other studies.319 responding group was 6 months compared to 37 months The outcomes in different studies are difficult to compare for the 37 patients who had not responded at that time because of different patient populations and specifically dif- (P 0.001). Twenty-one of the 36 patients who had not ferent percentages of patients with refractory disease. Fifty- achieved a response by 1 month eventually went on to five percent of the patients in this study had refractory dis- respond at a later time. Estimates of median subsequent ease at the time of transplant. Harousseau et al5 reported progression-free survival were 20 months for the late median survival of 54 months in patients with chemosensi- responders vs 3 months for refractory patients achieving tive disease and 30 months in chemo-refractory disease response criteria within a month of transplant. This corre- with melphalan/TBI-based regimens. Bensinger et al4 lation was also found when only patients with primary reported median survival of 36 months in a group of refractory were considered and even in a homogenous patients similar to our study and with a similar multi- group of primary refractory patients transplanted within the alkylator regimen. Vesole et al9 reported median survival first year (data not shown). of 41 months and a CR rate of 36% with the `Total therapy' or tandem transplant program in previously treated patients, 38% with refractory disease. Longer median survival rates Discussion up to 57 to 68 months10,11 have been reported in studies of newly diagnosed patients. However, these studies calculate High-dose chemotherapy and autologous hematopoietic survival from the day of diagnosis and include a lower per- transplantation has emerged as an effective treatment for centage of patients with refractory disease. Although the patients with multiple myeloma. This treatment modality outcomes seem comparable, comparison of outcomes results in high response rates, however, almost all patients following different regimens requires prospective are destined to relapse. High-dose melphalan with or with- comparative studies. out total body irradiation has been the most frequently used The only additional factor in this study for which there conditioning regimen. We have investigated an alternative was evidence of association with survival after considering regimen trying to optimize cytoreduction prior to transplant disease status was age. Other groups have not found age in anticipation of better disease control if this goal is achi- to be a prognostic variable.11,31 eved. Dimopoulos et al14 first reported on the combination Unresponsiveness to conventional chemotherapy seems of thiotepa, busulfan, and cyclophosphamide for the treat- to have a different biological significance in patients early ment of multiple myeloma. Each of the drugs used in this in the course of the disease. In the current study, 45 patients combination has proven activity against myeloma and has with primary refractory disease transplanted within 6 a different spectrum of extra-medullary toxicity, allowing months of initiation of therapy had survival rates that were escalation of the treatment doses and optimizing dose similar to those experienced by patients with chemosensi- intensity. tive disease despite lower CR rate. In contrast, patients with We have treated 120 patients with the TBC combination. refractory relapse had extremely poor outcome with a Overall, 81% of the patients responded and 26% achieving median survival of less than 1 year despite a high overall stringently defined CR. The median survival and EFS for response rate. Thus despite intensive therapy and a high the entire group was 38 months and 17 months, respect- remission rate, early disease recurrence and short survival ively. The multivariate analysis of pretransplant character- indicated a need for other investigational approaches in this subgroup. Early mortality for all patients in our series was 13%. 1.0 This is similar to the report of Barlogie et al6 for patients 0.9 with advanced myeloma and during the same years. How- Proportion surviving 0.8 ever, more recently with the introduction of peripheral 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 CR/PR at 1 month No response at 1 month blood stem cell transplants and better supportive care, the early mortality with this regimen was 4.8% (95% CI 1 13%). The toxicity profile was similar to that reported in other studies with the exception of severe hemorrhagic cystitis that occurred in 3% of all patients. The group most likely to benefit from TBC and autologous transplant were younger patients with chemosensitive disease. TRM was 0 12 24 36 48 60 72 84 96 higher among older patients and patients with refractory Months disease. Caution should be taken when administering intensive regimens in these groups. These patients should be Figure 4 KaplanMeier curves of OS subsequent to 1 month after trans- plant (landmark analysis) for 62 patients with refractory disease at trans- plant, alive at that stage. Patients achieving response criteria at 1 month (n = 25) have a worse outcome than non-responders (n = 37) as determined at that time. considered for alternative therapies. The kinetics of paraprotein clearance after high-dose chemotherapy has not been well studied. Comparison of the response kinetics in different studies is difficult due to Bone Marrow Transplantation the use of different response criteria. Singhal et al32 described the response kinetics in 33 patients in plateau phase with detectable protein. CR occurred in 96% of those ultimately achieving CR within 6 months with a median of 47 days. The time to CR was longer and the probability of achieving CR was lower with higher pretransplant paraprotein level. In a study of 50 previously untreated patients Gore et al33 found that the median time to CR was 112 days (range 16335) with seven of 25 patients achieving CR later than 6 months after transplant. Time to maximal response was 99 days in a similar series of previously untreated myeloma by Cunningham.34 The prognostic significance of response kinetics in patients undergoing stem cell transplant for multiple myeloma remain unknown. In our series the median time to CR was 146 days (range 15 536). Six of 31 CRs were achieved after more than 6 months and two after more than 1 year from transplant. For patients with refractory disease, remission criteria were achieved in a median of 30 days (range 1501); 12% of all responses were reached more than 6 months, and 6% more than 1 year after transplant. Maintenance therapy could have contributed to continuous and delayed responses. The study was not designed to assess the role of maintenance therapy since most responding patients were given longterm maintenance, however, delayed responses have also been noted in its absence. Interestingly, refractory patients that achieved response criteria early after transplant had a worse outcome. Landmark analysis showed that survival subsequent to 1 month after transplant was superior in those not achieving remission by that time. The delayed clearance of paraprotein after high-dose chemotherapy, sometimes for more than 1 year, cannot be explained only on the basis of long half-life. The continuous slow reduction of paraprotein level suggests that intensive treatment may have a more pronounced effect on an early precursor or colony-forming cell. The more mature protein secreting cells gradually disappear when there is no ongoing self-renewal. It has been shown35 that patients with refractory relapse have higher growth fractions and greater number of colony-forming cells in in vitro cultures. This explains the high response rate and rapid response but also rapid tumor regrowth and poor outcome. However, if the fraction of this more aggressive sub-clone is low, the response will be slow but durable. The pattern of response may be a clinical surrogate for the capability of the tumor for self-renewal and emergence of drug resistance and may have a significant prognostic value. Thus, patients with refractory disease who responded rapidly (within 1 month of transplant) probably had a more aggressive tumor with a high-dividing fraction, and despite a good initial response rapidly recurred with a resistant and rapidly fatal course. The same has been suggested by Hansen et al36 for rapid response to conventional chemotherapy. Boccadoro et al21 reported the same phenomenon in early responders to conventional chemotherapy when associated with high-labeling index. The normal kinetics of response should be considered when evaluating post-high-dose chemotherapy treatments such as tandem transplants. The occurrence of CR after 6 months from high-dose chemotherapy suggests that at least part of the response to any post-transplant intervention, High-dose chemotherapy in myeloma A Shimoni et al such as early administered second transplant, can still be related to the previous treatment. TBC might be an effective intensive regimen, given with a single transplant. With the improvement of supportive care it is as safe as the other regimens for autologous transplantation. TBC merits further study and comparison with alternative conditioning regimens as initial remission consolidation or early intensification of primary refractory disease. References 1 Alexanian R, Dimopoulos M. The treatment of multiple myeloma. New Engl J Med 1994; 330: 484489. 2 Kyle RA. Prognostic factors in multiple myeloma. Stem Cells 1995; 13 (Suppl. 2): 5663. 3 Attal M, Harousseau J, Stoppa A et al. A prospective randomized trial of autologous bone marrow transplantation and chemotherapy in multiple myeloma. New Engl J Med 1996; 335: 9197. 4 Bensinger WI, Rowley SD, Demirer T et al. High dose therapy followed by autologous hematopoietic stem cell infusion for patients with multiple myeloma. J Clin Oncol 1996; 14: 14471456. 5 Harousseau J, Attal M, Divine M et al. Autologous stem cell transplantation after first remission induction treatment in multiple myeloma: a report of the French registry on autologous transplantation in multiple myeloma. Blood 1995; 85: 3077 3085. 6 Barlogie B, Jagannath S, Naucke S et al. Long term followup after therapy for high risk multiple myeloma. Bone Marrow Transplant 1998; 21: 11011107. 7 Vesole DH, Crowly JJ, Catchatourian R. High-dose melphalan with autotransplantation for refractory multiple myeloma: results of a Southwest Oncology Group phase II trial. J Clin Oncol 1999; 17: 21732179. 8 Barlogie B, Desikar R, Munshi M et al. Durable CR in multiple myeloma in the absence of chromosome 13 deletion and with tandem, melphalan-based, high-dose therapy. The Arkansas experience with 1000 consecutive patients. Blood 1999; 94 (Suppl. 1): 714 (Abstr.). 9 Vesole DH, Tricot G, Jagannath S et al. Autotransplantation in multiple myeloma: what have we learned. Blood 1996; 88: 838847. 10 Attal M, Harousseau JL, Stoppa AM et al. High dose therapy in multiple myeloma: an update analysis of the IFM 90 protocol. Blood 1997; 90 (Suppl. 1): 418 (Abstr.). 11 Barlogie B, Jagannath S, Desikan KR et al. Total therapy with tandem transplants for newly diagnosed multiple myeloma. Blood 1999; 93: 5565. 12 Alexanian R, Dimopoulos MA, Delasalle KB et al. Myeloablative therapy for primary resistant myeloma. Stem Cells 1995; 13 (Suppl. 2): 118121. 13 Alexanian R, Dimopoulos M, Smith T et al. Limited value of myeloablative therapy in late multiple myeloma. Blood 1994; 83: 512516. 14 Dimopoulos MA, Alexanian R, Przepiorka D et al. Thiotepa, busulfan and cyclophosphamide: a new preparative regimen for autologous marrow or blood stem cell transplantation in high risk multiple myeloma. Blood 1993; 82: 23242328. 15 Vesole DH, Barlogie B, Jagannath S et al. High dose therapy for refractory myeloma: improved prognosis with better supportive care and double transplants. Blood 1994; 84: 950956. 16 Powles R, Raje N, Millar B et al. Outcome assessment of a population-based group of 195 unselected myeloma patients 827 Bone Marrow Transplantation High-dose chemotherapy in myeloma A Shimoni et al 828 under 70 years of age offered intensive treatment. Bone Mar- rine-induce graft versus host disease in patients with multiple row Transplant 1997; 20: 435443. myeloma undergoing high dose chemotherapy with autolog- 17 Jagannath S, Barlogie B, Dicke K et al. Autologous bone mar- ous stem cell rescue. J Clin Oncol 1997; 15: 667673. row transplantation in multiple myeloma identification of 27 Blade J, Samson D, Reece D et al. Criteria for evaluating prognostic factors. Blood 1990; 76: 18601866. disease response and progression in patients with multiple 18 Rajkumar SV, Fonseca R, Lacy MQ et al. Autotologous stem myeloma treated by high-dose therapy and hematopoietic stem cell transplantation for relapsed and primary refractory mye- cell transplantation. Br J Haematol 1998; 102: 11151123. loma. Bone Marrow Transplant 1999; 23: 12671272. 28 Kaplan EL, Meier P. Nonparametric estimation from incom- 19 Rajkumar SV, Fonseca R, Lacy MQ et al. Plasmablastic mor- plete observation. J Am Stat Assoc 1958; 53: 457481. phology is an independent predictor of poor survival after 29 Steinberg D, Colla P. CART: Tree-structured Non-parametric autologous stem-cell transplantation for multiple myeloma. J Data Analysis. Salford Systems: San Diego, CA, 1995. Clin Oncol 1999; 17: 15511557. 30 Cox DR. Regression models and life tables. J R Stat Soc 1972; 20 McLaughlin P, Alexanian R. Myeloma protein kinetics fol- 34: 187220. lowing chemotherapy. Blood 1982; 60: 851855. 31 Siegel DS, Desikan KR, Mehta J et al. Age is not a prognostic 21 Boccadoro M, Marmont F, Tribalto M et al. Early responder variable with autotransplants for multiple myeloma. Blood myeloma: kinetic studies identify a patient subgroup charac- 1999; 93: 5154. terized by very poor prognosis. J Clin Oncol 1989; 7: 119 32 Singhal S, Powles R, Milan S et al. Kinetics of paraprotein 125. clearance after autografting for multiple myeloma. Bone Mar- 22 Bearman SI, Appelbaum FR, Buckner CD et al. Regimen row Transplant 1995; 16: 537540. related toxicities in patients undergoing bone marrow trans- 33 Gore ME, Selby PJ, Clark PI et al. Intensive treatment of mul- plantation. J Clin Oncol 1988; 6: 15621568. tiple myeloma and criteria for complete response. Lancet 23 Durie BG, Salmon S. A clinical staging system for multiple 1989; 2: 879882. myeloma. Correlation of measured myeloma cell mass with 34 Cunningham D, Paz-Ares L, Milan S et al. High dose mel- presenting clinical features, response to treatment, and sur- phalan and autologous bone marrow transplantation as con- vival. Cancer 1975; 36: 842854. solidation in previously untreated myeloma. J Clin Oncol 24 Dimopoulos MA, Delasalle KB, Champlin R et al. Cyclophos- 1994; 12: 759763. phamide and etoposide therapy with GM-CSF for VAD-resist- 35 Takahashi T, Lim B, Jamal N et al. Colony growth and self- ant multiple myeloma. Br J Haematol 1993; 83: 240244. renewal of plasma cell precursors in multiple myeloma. J Clin 25 Dimopoulos MA, Weber D, Kantarjian H et al. HyperCVAD Oncol 1985; 3: 16131623. for VAD-resistant multiple myeloma. Am J Hematol 1996; 52: 36 Hansen OP, Jessen B, Videbaek A. Prognosis of myelomatosis 7781. on treatment with prednisone and cytostatics. Scand J Haema- 26 Giralt S, Weber D, Colome M et al. Phase I study of cyclospo- tol 1973; 10: 282290. Bone Marrow Transplantation