Double Intensive Consolidation Chemotherapy in Adult Acute Myeloid Leukemia By J.L. Harousseau, N. Milpied, J. Briere, B. Desablens, P.Y. Leprise, N. Ifrah, B. Gandhour, and P. Casassus Of 115 adult patients with de novo acute myeloid leukemia (AML), 87 (75.5%) achieved complete remission (CR) after induction treatment with zorubicin and conventional doses of cytarabine (Ara-C). Patients under age 45 years with histocompatibility locus antigen-identical sibling underwent bone marrow transplantation (BMT). The others were treated with two courses of intensive consolidation chemotherapy (ICC): course 1 with 4 days of high-dose Ara-C and 3 days of amsacrine (m-AMSA); course 2 with carmustine (BCNU), Ara-C, cyclophosphamide, and etoposide. Forty-two patients received both planned courses, 15 received only the first, and 13 patients could only support conventional maintenance therapy. Four patients died during consolidation. With a median follow-up of 60 months, the disease-free survival (DFS) after ICC at 5 years is 40.3% (± 6.5%), with no statistically significant difference between patients receiving one or two courses. The DFS for the 17 trans-
planted patients is comparable (P = .72) and is lower for the 13 excluded patients (23% -- 11.5%, P = .046). Age did not influence the probability of remaining in CR. In univariate analysis, three parameters had a negative impact on the 5-year DFS: a high initial WBC count (52%for patients with < 30 x 10' WBC/L v 12% for patients with > 30 x 10' WBC/L, P = .01), a long delay between induction treatment and course 1 (± 60 days; 63% v 29%, P = .01), and a long delay between course 1 and course 2 (± 60 days, 61.5% v 28.5%, P = .05). In multivariate analysis (Cox model), only the WBC count remained significant. This study confirms the value of intensive postremission chemotherapy, which can be compared in AML with allogeneic or autologous BMT. It also demonstrates the prognostic value of the initial WBC count. The optimal modalities of ICC remain to be defined by further studies. J Clin Oncol 9:1432-1437. c 1991 by American Society of Clinical Oncology.
THE
VALUE OF some form of postremission
patients with refractory or relapsed AML,9 0 high-
therapy in acute myeloid leukemia (AML) has been demonstrated,' but the nature of the postremission therapy remains a controversial issue. Three different approaches are currently debated: allogeneic bone marrow transplantation (BMT), autologous hematopoietic support, and
dose cytarabine (HD-Ara-C) was the major compo-
chemotherapy. Several modalities of chemotherapy have been proposed and classified: maintenance, consolidation, and intensification therapies.2 Some recently published series have shown that in nonrandomized studies, a disease-free survival (DFS) rate of 30% to 50% could be achieved after a short-term intensified chemotherapy without maintenance treatment." Because of remissions achieved in
nent of the intensive consolidation chemotherapy (ICC) in these studies. Autologous BMT after myeloablative therapy has also been shown to be an
attractive alternative
in the
treatment
of
AML."1 2"However, this method risks reinfusion of an unknown quantity of leukemic clonogenic cells. We present the results of a multicenter pilot study of a double ICC in AML. The first course contained HD-Ara-C. The second course contained drugs used in the conditioning regimen for autologous BMT but at lower doses to avoid the need of an autologous hematopoietic support.
PATIENTS AND METHODS Study Group From the Departmentsof Hematology of Centre Ilospitalier Regional, Nantes; Centre Hospitalier Regional, Brest; Centre Hospitalier Rggional, Amiens; Centre Ilospitalier Rdgional, Rennes: Centre Hospitalier Regional, Angersn and Il6pital Bobigny, France. Submitted June 8, 1990; accepted February 11, 1991. Address reprint requests to J.L. Harousseau, MD, HIotel Dieu, 44035 Nantes Ceder, France. (.'•1991 by American Society of Clinical Oncology. 0732-183X/91/0908-000553.00/0
1432
Adult patients under 65 years of age with de novo AML were included in a multicenter study involving six centers. Patients with a diagnosis of preleukemic myelodysplastic syndrome or with blastic transformation of a chronic myeloproliferative disorder were excluded. Secondary (chemoinduced and/or radioinduced) leukemias were also excluded. From September 1984 to March 1987, 115 consecutive patients (53 male, 62 female) were enrolled after providing informed consent.
Journal of Clinical Oncology, Vol 9, No 8 (August), 1991: pp 1432-1437
Downloaded from ascopubs.org by EAST CAROLINA UNIVERSITY on January 25, 2020 from 150.216.132.043 Copyright © 2020 American Society of Clinical Oncology. All rights reserved.
INTENSIVE CONSOLIDATION CHEMOTHERAPY IN AML
1433
Treatment Regimen
RESULTS 2
The induction treatment consisted of Ara-C 200 mg/m /d (continuous infusion) for 7 days and zorubicin (Rubidazone; Rhone-Poulenc, Antony, France) 200 mg/m2/d bolus day 1 to day 4. Patients in complete- remission (CR) were allografted if they were under the age of 45 years and had a histocompatibility locus antigen-identical sibling. Patients over 45 years or without a suitable donor received two courses of ICC. The first course was Ara-C (3 g/m 2 in a 2-hour infusion every 12 hours for eight doses, day 1 to day 4) and amsacrine (m-AMSA) 100 mg/m2 /d intravenously (IV), day 5 to day 7. This course had to be performed as soon as possible after CR achievement. However, while waiting for a laminar air-flow room in one center, 14 patients received one course of maintenance therapy with Ara-C 100 mg/m2 subcutaneously every 12 hours for 10 doses and m-AMSA 150 mg/m 2 day 1. These patients did not differ from the others in hematologic or extrahematologic toxicity of the consolidation courses or in CR duration. The second course was carmustine (BCNU) 200 mg/m2 (IV bolus) day 1, Ara-C 300 mg/m2 /d (continuous infusion) day 1 to day 4, cyclophosphamide 1.5 g/m 2/d (1-hour infusion) day 2 and day 3, etoposide 300 mg/m'/d (IV bolus) day 2 and day 3. No further treatment was administered.
Supportive Care According to available facilities, patients were housed in standard single-bed rooms with reverse isolation or in laminar air-flow rooms (48% of the patients during the first course, 52% of the patients during the second course). Supportive care included broad-spectrum antibiotics, prophylactic platelet transfusions, and parenteral nutrition according to the methods used in each center.
StatisticalAnalysis Differences in patient distributions were assessed by X2 test for qualitative variables and by Wilcoxon test for nonparametric analysis."3 CR was defined by the criteria of Ellison et al.' 4 DFS was calculated from achievement of CR to relapse or death in CR. Survival curves were calculated by the method of Kaplan and Meier,"5 and differences between distributions were based on the log-rank test?6 The significance of prognostic factors for DFS was determined using a stepwise linear regression method based on Cox's proportional hazards model.'7
Patient Characteristics The initial characteristics of the 115 patients are listed in Table 1. Their median age was 44 years (range, 13 to 65 years). The French-AmericanBritish (FAB) classification was available for 111 patients, as follows: M1, 18 patients; M2, 44; M3, nine; M4, 16; M5, 20; M6, three; M7, one. The initial hemoglobin level ranged from 4 to 15 g/dL (median, 8.4 g/dL), the initial WBC count ranged from 0.8 to 314 x 10'/L (median, 9.8 x 109/L), and the initial platelet count ranged from 6 to 305 x 10'/L (median, 49.5 x 109/L). Eighty-seven (75.5%) patients achieved CR after induction treatment (one course, 80 patients; two courses, seven). The achievement of CR was not influenced by age, sex, initial hemoglobin level, WBC and platelet counts, FAB classification, or fever at presentation. Of these 87 patients, 17 received a supralethal therapy followed by BMT. Of the 70 other patients, 13 (18.5%) refused the ICC regimen or were considered unable to tolerate high-dose chemotherapy because of infectious (aspergillosis) or visceral complications of the induction treatment; they received conventional monthly maintenance chemotherapy. Thus, 57 patients are assessable for the ICC program, 42 of them having received both courses. Fifteen patients could only receive the first planned course. The causes for not receiving the second course were death during the first course (three patients), relapse between the first and the second courses (three patients), refusal or medical contraindication (nine patients). The median interval between induction treatment and course 1 was 69 days (range, 36 to 124 days), and the median interval between course 1 and course 2 was 64 days (range, 40 to 152 days). The clinical characteristics of the 57 patients treated with ICC, the 13 excluded
Table 1. Initial Patient Characteristics Characteristic
All Patients (N = 115)
ICC Patients (n = 57)
Sex (M/F) Median age (years) Range FAB M4 + M5 subclasses (%) Fever at presentation (%) Median hemoglobin (g/dL) Median WBC count (x 1 0 //L) Median-platelet count (x 10'/L)
53/62 44 13-65 30 43 8.4 9.8 49
26/31 44 13-65 17.5 43 8 7.7 52
P
planted patients is comparable (P = .72) and is lower for the 13 excluded patients (23% -- 11.5%, P = .046). Age did not influence the probability of remaining in CR. In univariate analysis, three parameters had a negative impact on the 5-year DFS: a high initial WBC count (52%for patients with < 30 x 10' WBC/L v 12% for patients with > 30 x 10' WBC/L, P = .01), a long delay between induction treatment and course 1 (± 60 days; 63% v 29%, P = .01), and a long delay between course 1 and course 2 (± 60 days, 61.5% v 28.5%, P = .05). In multivariate analysis (Cox model), only the WBC count remained significant. This study confirms the value of intensive postremission chemotherapy, which can be compared in AML with allogeneic or autologous BMT. It also demonstrates the prognostic value of the initial WBC count. The optimal modalities of ICC remain to be defined by further studies. J Clin Oncol 9:1432-1437. c 1991 by American Society of Clinical Oncology.
THE
VALUE OF some form of postremission
patients with refractory or relapsed AML,9 0 high-
therapy in acute myeloid leukemia (AML) has been demonstrated,' but the nature of the postremission therapy remains a controversial issue. Three different approaches are currently debated: allogeneic bone marrow transplantation (BMT), autologous hematopoietic support, and
dose cytarabine (HD-Ara-C) was the major compo-
chemotherapy. Several modalities of chemotherapy have been proposed and classified: maintenance, consolidation, and intensification therapies.2 Some recently published series have shown that in nonrandomized studies, a disease-free survival (DFS) rate of 30% to 50% could be achieved after a short-term intensified chemotherapy without maintenance treatment." Because of remissions achieved in
nent of the intensive consolidation chemotherapy (ICC) in these studies. Autologous BMT after myeloablative therapy has also been shown to be an
attractive alternative
in the
treatment
of
AML."1 2"However, this method risks reinfusion of an unknown quantity of leukemic clonogenic cells. We present the results of a multicenter pilot study of a double ICC in AML. The first course contained HD-Ara-C. The second course contained drugs used in the conditioning regimen for autologous BMT but at lower doses to avoid the need of an autologous hematopoietic support.
PATIENTS AND METHODS Study Group From the Departmentsof Hematology of Centre Ilospitalier Regional, Nantes; Centre Hospitalier Regional, Brest; Centre Hospitalier Rggional, Amiens; Centre Ilospitalier Rdgional, Rennes: Centre Hospitalier Regional, Angersn and Il6pital Bobigny, France. Submitted June 8, 1990; accepted February 11, 1991. Address reprint requests to J.L. Harousseau, MD, HIotel Dieu, 44035 Nantes Ceder, France. (.'•1991 by American Society of Clinical Oncology. 0732-183X/91/0908-000553.00/0
1432
Adult patients under 65 years of age with de novo AML were included in a multicenter study involving six centers. Patients with a diagnosis of preleukemic myelodysplastic syndrome or with blastic transformation of a chronic myeloproliferative disorder were excluded. Secondary (chemoinduced and/or radioinduced) leukemias were also excluded. From September 1984 to March 1987, 115 consecutive patients (53 male, 62 female) were enrolled after providing informed consent.
Journal of Clinical Oncology, Vol 9, No 8 (August), 1991: pp 1432-1437
Downloaded from ascopubs.org by EAST CAROLINA UNIVERSITY on January 25, 2020 from 150.216.132.043 Copyright © 2020 American Society of Clinical Oncology. All rights reserved.
INTENSIVE CONSOLIDATION CHEMOTHERAPY IN AML
1433
Treatment Regimen
RESULTS 2
The induction treatment consisted of Ara-C 200 mg/m /d (continuous infusion) for 7 days and zorubicin (Rubidazone; Rhone-Poulenc, Antony, France) 200 mg/m2/d bolus day 1 to day 4. Patients in complete- remission (CR) were allografted if they were under the age of 45 years and had a histocompatibility locus antigen-identical sibling. Patients over 45 years or without a suitable donor received two courses of ICC. The first course was Ara-C (3 g/m 2 in a 2-hour infusion every 12 hours for eight doses, day 1 to day 4) and amsacrine (m-AMSA) 100 mg/m2 /d intravenously (IV), day 5 to day 7. This course had to be performed as soon as possible after CR achievement. However, while waiting for a laminar air-flow room in one center, 14 patients received one course of maintenance therapy with Ara-C 100 mg/m2 subcutaneously every 12 hours for 10 doses and m-AMSA 150 mg/m 2 day 1. These patients did not differ from the others in hematologic or extrahematologic toxicity of the consolidation courses or in CR duration. The second course was carmustine (BCNU) 200 mg/m2 (IV bolus) day 1, Ara-C 300 mg/m2 /d (continuous infusion) day 1 to day 4, cyclophosphamide 1.5 g/m 2/d (1-hour infusion) day 2 and day 3, etoposide 300 mg/m'/d (IV bolus) day 2 and day 3. No further treatment was administered.
Supportive Care According to available facilities, patients were housed in standard single-bed rooms with reverse isolation or in laminar air-flow rooms (48% of the patients during the first course, 52% of the patients during the second course). Supportive care included broad-spectrum antibiotics, prophylactic platelet transfusions, and parenteral nutrition according to the methods used in each center.
StatisticalAnalysis Differences in patient distributions were assessed by X2 test for qualitative variables and by Wilcoxon test for nonparametric analysis."3 CR was defined by the criteria of Ellison et al.' 4 DFS was calculated from achievement of CR to relapse or death in CR. Survival curves were calculated by the method of Kaplan and Meier,"5 and differences between distributions were based on the log-rank test?6 The significance of prognostic factors for DFS was determined using a stepwise linear regression method based on Cox's proportional hazards model.'7
Patient Characteristics The initial characteristics of the 115 patients are listed in Table 1. Their median age was 44 years (range, 13 to 65 years). The French-AmericanBritish (FAB) classification was available for 111 patients, as follows: M1, 18 patients; M2, 44; M3, nine; M4, 16; M5, 20; M6, three; M7, one. The initial hemoglobin level ranged from 4 to 15 g/dL (median, 8.4 g/dL), the initial WBC count ranged from 0.8 to 314 x 10'/L (median, 9.8 x 109/L), and the initial platelet count ranged from 6 to 305 x 10'/L (median, 49.5 x 109/L). Eighty-seven (75.5%) patients achieved CR after induction treatment (one course, 80 patients; two courses, seven). The achievement of CR was not influenced by age, sex, initial hemoglobin level, WBC and platelet counts, FAB classification, or fever at presentation. Of these 87 patients, 17 received a supralethal therapy followed by BMT. Of the 70 other patients, 13 (18.5%) refused the ICC regimen or were considered unable to tolerate high-dose chemotherapy because of infectious (aspergillosis) or visceral complications of the induction treatment; they received conventional monthly maintenance chemotherapy. Thus, 57 patients are assessable for the ICC program, 42 of them having received both courses. Fifteen patients could only receive the first planned course. The causes for not receiving the second course were death during the first course (three patients), relapse between the first and the second courses (three patients), refusal or medical contraindication (nine patients). The median interval between induction treatment and course 1 was 69 days (range, 36 to 124 days), and the median interval between course 1 and course 2 was 64 days (range, 40 to 152 days). The clinical characteristics of the 57 patients treated with ICC, the 13 excluded
Table 1. Initial Patient Characteristics Characteristic
All Patients (N = 115)
ICC Patients (n = 57)
Sex (M/F) Median age (years) Range FAB M4 + M5 subclasses (%) Fever at presentation (%) Median hemoglobin (g/dL) Median WBC count (x 1 0 //L) Median-platelet count (x 10'/L)
53/62 44 13-65 30 43 8.4 9.8 49
26/31 44 13-65 17.5 43 8 7.7 52
P
