Intensive consolidation versus oral maintenance therapy in patients 61 years or older with acute myeloid leukemia in first remission: results of second randomization of the AML HD98-B treatment Trial

In patients over the age of 60 years with acute myeloid leukemia (AML), treatment results are disappointing with only modest improvement over the last four decades.1 In addition to the limited success rates after induction therapy, one major problem is that older adults have a high rate of relapse, and no clearly effective postremission therapy has been established.2, 3 We recently reported the results of our randomized phase III AML HD98-B treatment trial.4 In this study, we showed that the addition of all-trans retinoic acid (ATRA) to induction and first consolidation therapy significantly improved complete remission (CR) rates and survival in patients with AML (excluding acute promyelocytic leukemia (APL)) over the age of 60 years. The present report focuses on the second randomization of this trial that was performed to assess the role of a second intensive, intravenously (i.v.) given consolidation with idarubicin and etoposide versus a 1-year oral maintenance therapy with the same drugs. Patients 61 years or older with de novo AML or refractory anemia with excess of blasts in transformation (RAEB-t) defined by the French–American–British (FAB) classification system, secondary AML (s-AML) with a preceding history of myelodysplasia, or therapy-related AML following treatment of primary malignant disease (t-AML) were eligible for entry into the trial. Patients with APL, patients with concomitant liver, renal or cardiac disease, and patients with a performance status WHO >2 were not included. The treatment schedule was as follows: after a first up-front randomization for the addition of ATRA to induction therapy with two cycles of ICE (idarubicin 12 mg/m2 i.v. days 1 and 3, cytarabine 100 mg/m2 cont. i.v. days 1–5, etoposide 100 mg i.v. days 1 and 3) and first consolidation therapy with HAM (cytarabine 0.5 g/m2/12 h i.v., days 1–3; mitoxantrone 10 mg/m2 i.v., days 2 and 3), patients in CR were randomized to either second intensive consolidation therapy IEiv (idarubicin 12 mg/m2 i.v. days 1 and 3, etoposide 100 mg/m2 i.v. days 1–5) or a 1-year oral maintenance therapy IEpo (idarubicin 5 mg p.o. days 1, 4, 7, 10, 13; etoposide 100 mg p.o. days 1 and 13; repeated on day 29 for 12 courses). Between August 1997 and April 2003, 329 patients were recruited. After two cycles of induction therapy 151 of 329 eligible patients achieved CR (46%). First consolidation therapy with HAM or A-HAM was given in 119 patients. After first consolidation therapy, randomization between IEiv versus IEpo was intended. Causes for not proceeding to randomization were allogeneic transplantation from an HLA-identical sibling donor (n=7), no informed consent (n=14), and relapse in two patients after first consolidation therapy. Finally, 96 patients were randomized between IEiv (n=47) and IEpo (n=49). Table 1 shows the clinical parameters according to second randomization. Of 47 patients randomized to IEiv, three withdrew informed consent and received no further therapy, and two patients relapsed before the start of IEiv. Forty-two patients received IEiv in median 55 days (range: 40–89 days) after first consolidation therapy with HAM or A-HAM. All patients were treated in-hospital, and the median duration of hospital stay was 29 days (range, 7–42 days). All patients suffered from hematological toxicity with median times of neutropenia (neutrophils <0.5 109/l) and thrombocytopenia (platelets <20 109/l) of 13 (range, 1–27 days) and 7.5 days (range, 0–26 days), respectively. In 26 of 42 patients (62%), i.v. antibiotics were administered due to neutropenic fever (n=15), septicemia (n=4), pneumonia (n=4), and soft tissue infection (n=3). There was one death 12 days after start of IEiv due to pneumonia and subsequent septicemia. Of 49 patients randomized to IEpo, three received intensive late consolidation (IEiv, n=2; HAM, n=1), and two patients relapsed before the start of IEpo. Forty-four patients received the first cycle of IEpo in median 55 days (range 25–268 days) after first consolidation therapy. A median of four cycles (range, one to 12) of IEpo were administered; eight patients completed 12 cycles of IEpo. Causes of premature termination of IEpo were relapse in 34 cases, persistent thrombocytopenia below 50 109/l after the forth cycle in one case, and infectious spondylitis after the second cycle in one case. A total of 251 cycles IEpo were administered. No hematological toxicity WHO grade III/IV was observed in patients who were in continuous CR. Four patients were admitted to the hospital due to fever (>38.0°C) and one due to infectious spondylitis, and all received i.v. antibiotics (11%). The median duration of their hospital stays was 6 days (range 3–9 days). The median estimated follow-up time after second randomization was 52 months. Seventy-one of the 96 patients have died resulting in an estimated survival probability after 48 months of 24% (95 confidence interval, 17–36%). Of the 96 randomized patients, 80 relapsed, and one patient died during IEiv-induced cytopenia leading to a cumulative incidence of relapse (CIR) and death after 48 months of 82% (standard error, 0.04) and 1% (standard error, 0.01), respectively. Of the 80 relapsed patients, 39 received either no or palliative treatment. Following an intensive reinduction therapy, 13 of 41 (32%) patients achieved a second CR two of whom received an autologous and one an allogeneic stem cell transplantation. Analysis according to second randomization on an intention-to-treat basis revealed a statistically significant lower CIR (P=0.005) in patients assigned to IEiv compared to those assigned to IEpo. This translated into a statistically significant better survival (P=0.04) with estimated median survival times from second randomization of 22.3 and 14.3 months in the IEiv and the IEpo arm, respectively (Figure 1). The principle conclusion of the second randomization of the AML HD98-B trial is that a second intensive consolidation with idarubicin and etoposide resulted in lower CIR and superior survival compared to a mild p.o. maintenance therapy with the same drugs. Myelosuppression during the p.o. maintenance therapy with IEpo was mild or absent in the majority of patients. Therefore, one could argue that there was only a minor or no therapeutic effectiveness of IEpo. Even in this scenario, our study in elderly patients adds to previous knowledge, that is, that four cycles of intensive chemotherapy may be superior to only three cycles.5 Additionally, our study highlights two problems of performing a randomization late during treatment course in elderly AML patients. On the one hand, only 29% of the initial patient cohort could be randomized after first consolidation therapy. Although this proportion is comparable to that of other recently published randomized studies addressing this issue,6, 7 the value of these results for the whole population is very limited. On the other hand, selection bias towards favorable risk factors is enormous. Only two of 96 randomized patients where classified as high-risk according the Medical Research Council classification system.8 This raises the question whether a stratification of elderly patients according to cytogenetics might be useful in the future to identify a population of patients who will benefit from longer treatment1, 8 (Schlenk et al. Blood 2005; 106: 544, abstract). We are grateful to all members of the German-Austrian Acute Myeloid Leukemia Study Group (AMLSG) for providing leukemia specimens.