(2014) 28, 1001–1007 & 2014 Macmillan Publishers Limited All rights reserved 0887-6924/14 www.nature.com/leu

ORIGINAL ARTICLE Chemomodulation of sequential high-dose by fludarabine in relapsed or refractory : a randomized trial of the AMLCG

M Fiegl1,14, M Unterhalt1,14, W Kern2, J Braess3, K Spiekermann1, P Staib4, A Gru¨ neisen5,BWo¨ rmann6, D Scho¨ ndube7, H Serve8, A Reichle9, M Hentrich10, X Schiel10, C Sauerland11, A Heinecke11, C Rieger1, D Beelen12, WE Berdel13,TBu¨ chner13 and W Hiddemann1 for the German AML Cooperative Group (AMLCG)

Chemomodulation of cytarabine by fludarabine has been attributed with a higher antileukemic efficacy, but randomized trials to address this question are rare. We therefore conducted a multicenter, randomized phase III study to evaluate the antileukemic efficacy of adding fludarabine to sequential high-dose cytarabine þ (SHAI) re-induction in relapsed or refractory acute myeloid leukemia (AML). Patients (n ¼ 326, of which 281 were evaluable) were randomly assigned to SHAI (cytarabine, 1 g/m2 bid, days 1–2 and 8–9 (3 g/m2 for patients p60 years with refractory AML or X2nd relapse); idarubicin 10 mg/m2 daily, days 3–4 and 10–11) or F-SHAI (SHAI with fludarabine, 15 mg/m2, 4 h before cytarabine). Although complete remission (CR) rates (35% SHAI and 44% F-SHAI) and overall survival did not differ between both regimens, fludarabine prolonged time to treatment failure from 2.04 to 3.38 months (median, Po0.05). Twenty-seven percent of patients proceeded to allogeneic stem cell transplantation, with a significantly higher number of patients in CR or incomplete remission in the F-SHAI group (22 vs 10%, Po0.01). In conclusion, fludarabine has a beneficial, although moderate, impact on the antileukemic efficacy of high-dose cytarabine-based salvage therapy for relapsed and refractory AML.

Leukemia (2014) 28, 1001–1007; doi:10.1038/leu.2013.297 Keywords: acute myeloid leukemia; relapse; chemotherapy; cytarabine; fludarabine

INTRODUCTION Besides the exploration of new drugs, additional approaches try In recent years, substantial progress has been achieved in the to optimize the antileukemic efficacy of cytarabine and/or 5 treatment of acute myeloid leukemia (AML), particularly for . A recent study by the HOVON investigated patients below 60 years of age. This progress is predominantly different doses of for remission induction and based on an intensification of induction therapy and a better revealed a clear dose dependency of response rates. The German understanding of AML biology, allowing the characterization of AML Cooperative Group explored a dose-dense schedule of 6 biologic subpopulations with different prognoses.1 These double induction therapy, suggesting a beneficial effect. Other achievements have resulted in the definition of AML risk groups investigated the addition of other drugs to the ‘7 þ 3’ 7 8 groups,2 which guides appropriate treatment strategies, regimen such as or gemtuzumab ozogamicin, and particularly for post-remission therapies. observed promising results. Initial therapy for remission induction, however, is still based on Another attempt to improve the efficacy of cytarabine-based cytarabine and anthracyclines, that is, two ‘old’ drugs, which were AML induction therapy is the modulation of intracellular introduced into treatment of AML more than 40 years ago,3,4 cytarabine metabolism. This approach aims at enhancing the 9 regardless of the respective genetic risk group (with the exclusion formation of cytarabine triphosphate (AraCTP), the active moiety. of acute promyelocytic leukemia). Attempts to improve these Cytarabine phosphorylation depends on the activity of the results by new agents that target specific molecular aberrations deoxycytidine kinase, the rate-limiting step of AraCTP 10 and pathways have so far been disappointing. Hence, the so- formation. Besides phosphorylation, AraCTP levels are called ‘7 þ 3’ regimen still remains the standard of AML induction maintained by dephosphorylation, regulated, for example, by treatment. . Fludarabine interacts with these

1Department of Internal Medicine III, University Hospital of Munich, Munich, Germany; 2MLL Munich Leukemia Laboratory, Munich, Germany; 3Department of Hematology and Oncology, Klinikum Barmherzige Bru¨der, Regensburg, Germany; 4Department of Hematology and Oncology, St-Antonius-Hospital, Eschweiler, Germany; 5Department of Hematology and Oncology, Vivantes Klinikum Berlin-Neuko¨lln, Berlin, Germany; 6German Society for Hematology and Oncology, Berlin, Germany; 7Department of Hematology and Oncology, Helios Klinikum Berlin Buch, Berlin, Germany; 8Department of Hematology and Oncology, J.W. Goethe-University, Frankfurt, Germany; 9Department of Internal Medicine II, University Hospital of Regensburg, Regensburg, Germany; 10Department of Hematology and Oncology, Krankenhaus Harlaching, Munich, Germany; 11Institute of Biostatistics and Clinical Research, University of Muenster, Muenster, Germany; 12Department of Hematopoietic Stem Cell Transplantation, University of Essen, Essen, Germany and 13Department of Medicine A, Hematology and Oncology, University of Muenster, Muenster, Germany. Correspondence: Dr M Fiegl, Department of Internal Medicine III, University Hospital of Munich, Marchioninistrasse 15, Munich 81377, Germany. E-mail: michael.fi[email protected] 14These authors contributed equally to this work. Received 14 August 2013; revised 27 September 2013; accepted 3 October 2013; accepted article preview online 22 October 2013; advance online publication, 8 November 2013 SHAI versus F-SHAI in relapsed or refractory AML M Fiegl et al 1002 pathways serving mainly as a ribonucleotide reductase inhibitor. allogeneic transplantation but without a donor, autologous SCT after a Hence, exposure of AML blasts to fludarabine several hours before conditioning regimen comprising busulfane and was cytarabine enhances intracellular AraCTP formation and DNA the recommended option, using cyclophosphamide mobilization and incorporation, resulting in an increased cell kill in vitro.11,12 G-CSF support for stem cell aphaeresis. On the basis of these in vitro data, fludarabine þ cytarabine combinations were investigated in clinical phase II and III studies. Evaluation and response criteria These studies lead to the establishment of the fludarabine þ A CR was defined as o5% blasts in the bone marrow, granulocytes cytarabine þ granulocyte colony-stimulating factor (G-CSF) (FLAG) 41500/mm3 and thrombocytes 4100 000/mm3, and the absence of and FLAG þ idarubicin (FLAG-Ida) regimens, which are widely used extramedullary leukemia. A CRi was defined as o5% blasts in the bone for the treatment of relapsed and refractory AML.13–17 Promising marrow and the absence of extramedullary leukemia without full recovery data were also reported for FLAG-Ida for first-line treatment of of the peripheral blood counts. Overall response rate (ORR) was defined as AML,18,19 most recently by the MRC15, where FLAG-Ida induced the combination of CR and CRi. Adequate or inadequate blast clearance high complete remission (CR) rates and reduced the rate of from the bone marrow was defined as p5or45% blasts in the post- relapse.20 treatment bone marrow evaluation (in aplasia, 1 week after the end of The German AML Cooperative Group therefore initiated a re-induction treatment). Patients dying within the first 60 days were considered as early deaths (EDs). Time to treatment failure (TTF) was multicenter, open-label, prospective randomized phase III trial in defined as the time from initiation of re-induction chemotherapy and relapsed or refractory AML to evaluate the antileukemic potential (I) death before assessment of early blast clearance, (II) inadequate blast of the pharmaco-modulation of sequential high-dose cytarabine clearance, (III) 45% blasts in the bone marrow after achievement of by fludarabine. This study was based on a preceding phase III trial remission or (IV) death in CR or CRi. Relapse-free survival was defined as in relapsed and refractory AML using the sequential application of time from achievement of CR/CRi and documentation of relapse or death. cytarabine þ mitoxantron (S-HAM),21 which revealed a high Overall survival (OS) was defined as the time from induction of efficacy in these patients. chemotherapy until death of any cause. Treatment-related side effects were assessed by the WHO (World Health Organization) criteria.

MATERIALS AND METHODS Statistical considerations Patients and entry criteria The primary objective of the present study was TTF (in a per protocol Adult patients (age X16 years) with relapsed or refractory (defined as analysis), because the best treatment option for patients with relapsed or primary resistance to induction chemotherapy (not specified) or CR refractory AML is allogeneic SCT. Therefore, it was the aim to establish a duration of o6 months) were eligible for the trial. There was no upper age salvage regimen that allows (I) more patients to undergo allogeneic SCT limit. Patients relapsing after prior allogeneic stem cell transplantation due to prolongation of time for organization of the procedure and (II) a (SCT) were excluded. In addition, patients were ineligible in case of severe high rate of allogeneic transplantation in remission (second CR or CRi), organ dysfunction (severe coronary heart disease or myocardial infarction which was preferred over transplantation in active disease. Patients within 6 months, cardiomyopathy, chronic pulmonal disease with undergoing allogeneic SCT were hence censored at the time of hypoxemia, uncontrolled hypertension, creatinine 41.5 mg/dl, bilirubin transplantation for analysis of TTF. With a predicted improvement of 42.0 mg/dl and/or transaminases 43 Â upper normal limit, uncontrolled 15%, a number of 127 evaluable patients per arm were needed for an infection and cerebral dysfunction), unless explained by AML. a-error of 0.05, using a sequential log rank test. Thus, with an estimate of No restrictions were made according to prior exposure. 10% of non-evaluable patients, a maximum total number of 140 patients per arm were to be included for a per protocol analysis. Cytogenetic subgroups Secondary objectives were non-hematologic toxicities, OS, relapse-free survival, rates of ED, CR and CRi, early blast clearance from the bone For detailed characterization of AML, bone marrow samples at relapse marrow and the rate of allogeneic SCT in CR or CRi. For additional analyses, underwent standardized diagnostics, including central sample registration, w2-tests and Student’s t-tests were used. The TTF and survival parameters preparation and evaluation by cytomorphology, cytochemistry and were calculated according to Kaplan–Meier analyses using log rank cytogenetics. Favorable cytogenetics were defined as t(8;21) and inv(16). statistics and/or Cox regression models. All analyses were performed with Unfavorable cytogenetics were defined as complex chromosomal aberra- SPSS Statistics Version 21 (IBM, New York, NY, USA). tions (X3 numerical or structural chromosomal aberrations without involvement of t(15;17), t(8;21) or inv(16)), aberrations of chromosomes 5 or 7, inv(3) and involvement of 11q23. Intermediate cytogenetics were Conduct of the study defined as normal karyotype or aberrations not qualifying for favorable or The study was carried out in accordance with the updated Declaration of unfavorable cytogenetics. At the time this trial was initiated, molecular risk Helsinki. All patients gave their written consent after being informed about groups (for example, FLT3-ITD and NPM1 mutations) were not used the purpose and the investigational nature of the trial. Before initiation, the routinely and, hence, were not required for this trial. study received approval of the responsible institutional review boards and the ethics committees of the participating institutions. Treatment protocol After inclusion, patients were randomized to receive either sequential high-dose cytarabine þ idarubicin (SHAI) or SHAI with fludarabine (F-SHAI). RESULTS 2 The SHAI regimen consisted of cytarabine 1000 mg/m as a 3-h infusion Study population and patient characteristics twice daily every 12 h on days 1–2 and 8–9, followed by idarubicin 10 mg/m2 daily on days 3–4 and 10–11. High-risk patients (refractory leukemia or Three hundred and twenty-six patients from 45 centers in X2nd relapse) o60 years of age received 3000 mg/m2 of cytarabine. The Germany, with relapsed or refractory AML, were included in the F-SHAI regimen comprised the same schedule with the addition of study. Patient disposition is shown in Figure 1. Patient details are fludarabine 15 mg/m2 4 h before each cytarabine dose. One week after given in Table 1 and show comparable pre-therapeutic character- completion of chemotherapy, a bone marrow aspiration was performed to istics. Median age was 57 years in the SHAI group and 52 years in assess initial blast clearance from the bone marrow. the F-SHAI group. Most patients suffered from relapse of de novo G-CSF at a dose of 5 mg/kg was administered subcutaneously once daily AML and intermediate cytogenetics was most frequently from day 14 onwards in every patient until peripheral granulocytes observed. In approximately one third of all patients cytogenetics exceeded 1500/mm3. In patients with 45% bone marrow blasts in the post-treatment bone marrow aspiration, G-CSF was discontinued. were not obtained at the time of relapse, as it was not a Patients who did not achieve a CR or incomplete remission (CRi) after mandatory requirement for participation. In both treatment arms, the treatment went off trial. Patients who achieved at least CRi stayed in B42% of all patients displayed a high-risk profile as defined by the trial; for these patients, post-remission therapy comprised preferably remission duration of o6 months before relapse, or second or allogeneic SCT from an HLA-identical donor. In patients suitable for higher relapse.

Leukemia (2014) 1001 – 1007 & 2014 Macmillan Publishers Limited SHAI versus F-SHAI in relapsed or refractory AML M Fiegl et al 1003 Patients randomly assigned (n=326)

F-SHAI (n=165) SHAI (n=161)

Received allocated treatment (n=162) Received allocated treatment (n=158) Did not receive allocated treatment (n=3) Did not receive allocated treatment (n=3)

Analyzed (n=142) Analyzed (n=139) Excluded from analysis (n=20) Excluded from analysis (n=19) (insufficient documentation) (insufficient documentation)

Treatment failures Treatment failures Early death (n=33) Early death (n=32) Persistent leukemia (n=61) Persistent leukemia (n=42)

Stem cell transplantations performed Stem cell transplantations performed Autologous (n=8) Autologous (n=3) Allogeneneic (n=44) Allogeneneic (n=33)

Deaths (n=112) Deaths (n=114) Censored, follow-up ended (n=30) Censored, follow-up ended (n=25)

Figure 1. CONSORT diagram: patient disposition. From 326 patients included, 281 were available for analysis for the primary end point. Forty-five patients had to be excluded from analysis, either because they received the wrong treatment (n ¼ 6) or were not documented, or insufficiently documented (n ¼ 39).

Table 1. Patient characteristics Table 2. Response to treatment of SHAI and F-SHAI in the whole study population and according to cytogenetic risk groups Characteristics SHAI (n ¼ 139) F-SHAI (n ¼ 142) Response rates SHAI F-SHAI Age Median age in years (range) 57 (17–83) 52 (19–76) Response in the whole population n ¼ 139 n ¼ 141 Patients X60 years n ¼ 54 (38.1%) n ¼ 48 (33.8%) CR (%) 35 44 ORR (CR þ CRi) (%) 42 54 Sex Non-responders (%) 37 26 Female 60 (43.2%) 62 (43.7%) Early death (%) 20 20 Male 79 (56.8%) 80 (56.3%) Response according to karyotype de novo AML 89.2% 89.4% Favorable (%) n ¼ 6 n ¼ 11 CR (%) 83.3 63.6 Karyotype (at relapse) CRi (%) 0.0 9.1 Not available 29.0% 32.0% Intermediate n ¼ 69 n ¼ 65 Favorable 4.0% 8.0% CR (%) 42.0 52.3 Intermediate 50.0% 46.0% CRi (%) 4.3 6.2 Unfavorable 17.0% 14.0% Unfavorable n ¼ 23 n ¼ 20 CR (%) 21.7 25.0 Duration of first remission CRi (%) 13.0 15.0 o6 months or 41 relapse 41.7% 43.0% Unknown n ¼ 41 n ¼ 46 46 months o18 months 41.8% 41.5% CR (%) 24.4 37.0 418 months 16.5% 15.5% CRi (%) 9.8 13.0

Clinical features Abbreviations: CR, complete remission; CRi, incomplete remission; F-SHAI, Median leukocyte count (/ml) 3820 3200 SHAI with fludarabine; ORR, overall response rate; SHAI, sequential Median thrombocytes (/ml) 52 000 49 500 high-dose cytarabine þ idarubicin. Median hemoglobin (g/dl) 10.3 10.6 ECOG 41 19.0% 23.0% Abbreviations: AML, acute myeloid leukemia; F-SHAI, SHAI with (CR CRi) of 42% in the SHAI group and 54% in the F-SHAI group fludarabine; SHAI, sequential high-dose cytarabine þ idarubicin. þ (P ¼ 0.056). Early blast clearance in the bone marrow was assessed in 153 patients (67 in SHAI and 86 in F-SHAI) and revealed a significantly higher adequate blast clearance after F-SHAI (90.7% Response to treatment after F-SHAI and 74.6% after SHAI, Po0.05). Response to treatment is shown in Table 2. Thirty-five percent of The rate of non-responders tended to be higher after SHAI with patients receiving SHAI and 44% of the F-SHAI treatment group 37% versus 26% after F-SHAI (P ¼ 0.054). The ED rate was 20% in achieved a CR. This difference was not statistically significant. CRi both arms and was higher in patients X60 years. Here, 26% of was seen in 8% and 10% of cases, respectively, resulting in an ORR elderly patients died within 60 days in the SHAI arm and 22% in

& 2014 Macmillan Publishers Limited Leukemia (2014) 1001 – 1007 SHAI versus F-SHAI in relapsed or refractory AML M Fiegl et al 1004 the F-SHAI arm. In younger patients, 16% treated with SHAI and 19% treated with F-SHAI succumbed to ED (not significant). Subgroup analyses revealed a significantly higher rate of non- responders in younger patients (o60 years of age) in the SHAI arm (39.5% vs 24.2%, Po0.05). However, ORR was not influenced by age as revealed by a multivariate analysis. Both regimens yielded high remission rates in the small subgroup (n ¼ 17) of patients with favorable cytogenetics (83.4% for SHAI and 72.7% for F-SHAI, not significant). For intermediate cytogenetics (n ¼ 134), 46.4% of patients treated with SHAI and 58.5% of patients treated with F-SHAI reached a CR or CRi. In the group with unfavorable cytogenetics (n ¼ 43), ORR was 34.8% with SHAI and 40.0% with F-SHAI. None of these differences was statistically significant. However, when combining the subgroups with intermediate and unfavorable cytogenetics (n ¼ 177) in a post-hoc analysis, F-SHAI was found to induce a significantly higher ORR of 54% as compared with SHAI (43%, Po0.05). Response rates according to cytogenetics (obtained at relapse) are given in Table 2. Owing to their nature as subgroup analyses, results should be interpreted with caution.

Survival parameters Re-induction chemotherapy with F-SHAI resulted in a significantly longer TTF (median 2.04 months vs 3.38 months, Po0.05) in the whole patient collective (Figure 2a). The OS was not affected by the type of re-induction chemotherapy. Median OS was 5.8 months for SHAI versus 6.7 months for F-SHAI, P ¼ 0.48 (Figure 2b). Relapse-free survival did not differ significantly between both groups (median 3.9 months for SHAI and 5.8 months for F-SHAI, P ¼ 0.31; Figure 2c). An overview of survival parameters according to cytogenetics is given in Table 3. In a multivariate analysis, age did not have an impact on TTF but on survival (data not shown).

Treatment-related toxicities Significantly higher rates of III/IV1 toxicities were found after F-SHAI for the occurrence of (non-lethal) bleeding (6% vs 9%, Po0.05), nausea and vomiting (11% vs 21%, Po0.01) and pulmonary side effects, that is, dyspnea, pulmonary edema and pneumonitis (8% vs 17%, Po0.05). Non-hematological side effects are listed in Table 4. Severe granulocytopenia was significantly longer after F-SHAI, requiring longer administration of G-CSF (median 24 days for F-SHAI vs 20 days for SHAI, Po0.01). However, this did not translate into significantly more frequent or more severe infections (243 infectious episodes for SHAI and 256 for F-SHAI, median duration of 5 days each).

Post-remission therapy Thirty-three patients (24%) in the SHAI group and 44 patients (31%) in the F-SHAI arm received SCT. In the SHAI group, 14 patients (10%) were transplanted in CR or CRi as compared with Figure 2. Outcome of re-induction chemotherapy. (a) TTF after initiation of therapy. F-SHAI significantly prolonged TTF by 1.3 31 patients (22%, Po0.01) in the F-SHAI treatment arm. months (SHAI, n ¼ 139; F-SHAI, n ¼ 141). (b) Overall survival. No Patients who received allogeneic SCT survived significantly significant difference was found between SHAI and F-SHAI (SHAI, longer than patients who did not receive SCT (median 11.8 vs 4 n ¼ 139; F-SHAI, n ¼ 141). (c) Relapse-free survival: the choice of months, Po0.01 (Cox regression and Kaplan–Meier); Figure 3a), re-induction chemotherapy with SHAI or F-SHAI had no impact on with seven patients (10%) being alive after 4 years as compared relapse-free survival (SHAI, n ¼ 49; F-SHAI, n ¼ 63). with four patients without allogeneic transplantation (2%). Remission status before allogeneic SCT did not significantly influence the outcome (Figure 3b): median OS for patients Additional prognostic factors transplanted in second CR was 12 vs 10.8 months for patients Overall outcome for the whole patient collective (regardless of re- transplanted in a non-remission status (P ¼ 0.57). induction chemotherapy) according to karyotype is shown in Autologous transplantation was performed in a small group of Figure 3c. Median survival for favorable karyotypes was 13.2 patients only (5% in SHAI and 2% in F-SHAI). Owing to the small months, for intermediate karyotypes was 7.2 months, whereas number, these patients were not further analyzed. unfavorable karyotypes had the shortest OS with a median

Leukemia (2014) 1001 – 1007 & 2014 Macmillan Publishers Limited SHAI versus F-SHAI in relapsed or refractory AML M Fiegl et al 1005 studies.14–17 The rationale for combining fludarabine and Table 3. Survival parameter according to cytogenetic risk groups cytarabine is the observation that pretreatment with fludarabine Karyotype SHAI F-SHAI increases intracellular AraCTP levels and consequently in vitro Median Median antileukemic efficacy. However, the results of clinical trials (95% CI) (95% CI) exploring fludarabine–cytarabine combinations were inconsistent. One trial by the MRC22 found no difference between a high-dose Favorable n ¼ 6n¼ 11 cytarabine þ fludarabine combination (without anthracyclines) Time to treatment failure 4.2 (0.0–9.8) 4.5 (3.6–5.5) and a ‘standard’ regimen consisting of intermediate dose (months) cytarabine, daunorubicin and in high-risk AML in Overall survival (months) 9.3 (4.4–14.1) 8.3 (1.8–14.7) terms of CR, relapse rate and relapse-free survival. Instead, a trend toward a decrease in OS after 4 years was observed for Intermediate n ¼ 69 n ¼ 65 22 23 Time to treatment failure 0.6 (0.6–0.6) 3.4 (0.0–7.4) cytarabine þ fludarabine. Russo et al. compared fludarabine, (months) cytarabine and idarubicin with cytarabine, idarubicin and Overall survival (months) 5.3 (4.1–6.5) 5.4 (2.9–7.9) etoposide as first-line therapy in a newly diagnosed AML case and reported a higher CR rate and a more favorable toxicity profile Unfavorable n ¼ 23 n ¼ 20 for fludarabine, cytarabine and idarubicin. Most recently, the Time to treatment failure 0.6 (0.6–0.6) 0.6 (0.5–0.7) MRC15 trial reported high remission rates for FLAG-Ida after one (months) course as compared with cytarabine, idarubicin and etoposide at Overall survival (months) 4.3 (0.2–8.4) 2.4 (0.7–4.2) the price of increased myelosuppression and reduced participation in consolidation therapy. Interestingly, relapse rates Unknown n ¼ 41 n ¼ 46 20 Time to treatment failure 0.6 (0.6–0.6) 0.6 (0.0–1.6) were still significantly reduced by FLAG-Ida. In all published (months) trials, treatment arms differed not only for the addition of Overall survival (months) 3.4 (2.4–4.4) 4.9 (3.4–6.5) fludarabine but also for the cytarabine dose and additional antileukemic agents. Hence, the question of the effect of Abbreviations: CI, confidence interval; F-SHAI, SHAI with fludarabine; SHAI, modulating cytarabine by fludarabine was not directly sequential high-dose cytarabine idarubicin. þ addressed. Judging the efficacy of FLAG-Ida-like regimens on the basis of published trials is even more hampered by the fact that various regimen are termed FLAG-Ida, yet they differ in Table 4. Non-hematologic toxicities (grades 3 and 4) of SHAI and dosage (for example, even for fludarabine) and dosing schedule. F-SHAI Therefore, the present study by the AML Cooperative Group Toxicities XIII1 SHAI F-SHAI aimed at evaluating the impact of fludarabine more precisely by a prospective randomized trial based on a SHAI or F-SHAI Liver combination in relapsed or refractory AML patients. Elevated bilirubin (%) 12 13 In our study we found only a moderate impact of chemo- Elevated AST/ALT (%) 5 8 modulation with fludarabin on the antileukemic efficacy of high-dose Elevated AP (%) 5 4 cytarabine. First, treatment with F-SHAI resulted in a statistically significant but clinically less relevant increase in TTF, which was Renal 2% 0% the primary end point of our trial. Second, the rate of early blast Cardiac 9% 6% Cerebral 8% 9% clearance from the bone marrow, obtained during aplasia (as an Pulmonary 8% 17%* early read-out for chemosensitivity), was significantly higher (by Allergic/exanthema 2% 4% 16%) in the F-SHAI arm. Third, although the rates of CR or CRi did Diarrhea 21% 24% not differ significantly in the whole study population, an 11% Alopecia 78% 68% increase in ORR was seen for F-SHAI in the subgroup of patients Bleeding 6% 9%* with non-favorable cytogenetics. This potential benefit of Nausea/vomiting 11% 21%** fludarabine for patients with non-favorable cytogenetics was also Abbreviations: F-SHAI, SHAI with fludarabine; SHAI, sequential high-dose demonstrated recently in a randomized trial comparing the cytarabine þ idarubicin. No significant differences were found except for combination of either fludarabine (DAF) or (DAC) to a (non-lethal) bleeding, pulmonary side effects and nausea/vomiting. 7 þ 3-like regimen (DA). Although DAC (other than DAF) improved *Po0.05, **Po0.01. CR rates and 3-year OS, DAF was superior to DA in terms of OS in the adverse cytogenetic risk group.24 In terms of pharmacodynamics, the schedule of fludarabine administration survival of 4.8 months (Po0.01). Patients achieving a CR or CRi was however suboptimal in the latter trial. show a significantly longer median OS (12 months, Po0.01) than TTF was chosen as the primary end point of the study, because patients who had persistent leukemia or progressive disease the best treatment option for patients with relapsed or refractory (median OS 2.4 months, Figure 3d). There was also a significant AML remains allogeneic SCT. Thus, the aim was to create a salvage difference between the responders, with patients achieving a CR regimen that allows more patients to undergo this treatment having a significantly longer OS of 13.2 months than patients option due to prolonged time for organization of the procedure achieving a CRi (7.2 months, Po0.05). (for example, search for donor and so on). Although this goal was As expected, remission duration before relapse had a significant achieved (more patients underwent allogeneic SCT in the F-SHAI prognostic impact with remission durations 418 months, arm and a significantly higher proportion received their transplant resulting in a median OS of 8.4 months as compared with in CR or CRi), the role of aggressive re-induction chemotherapy patients presenting with 41st relapse or remission duration o6 before allogeneic SCT has been questioned recently, and contrary months, who achieved an OS of 4.8 months only (Po0.01). to our initial hypothesis (that allogeneic SCT in second CR/CRi would be beneficial) this trial showed no advantage for patients that underwent allogeneic SCT in CR/CRi as compared with those DISCUSSION with active AML. One reason might be effective conditioning Fludarabine–cytarabine combinations are frequently used in regimens for SCT, for example, fludarabine–cytarabine combina- relapsed or refractory patients with AML based on several phase II tions as cytoreductive re-induction chemotherapy followed by a

& 2014 Macmillan Publishers Limited Leukemia (2014) 1001 – 1007 SHAI versus F-SHAI in relapsed or refractory AML M Fiegl et al 1006

Figure 3. Survival parameters independent of re-induction chemotherapy. (a) Patients receiving allogeneic SCT after SHAI or F-SHAI showed a significantly better OS than patients without SCT (11.8 vs 4.0 months, Po0.01). (b) OS after allogeneic SCT was not significantly better for SCT performed in second CR. (c) Cytogenetics remain a prognostic factor also in relapse. Patients with unfavorable karyotypes showed the worst OS. (d) Achievement of CR or CRi after induction therapy was associated with a significantly better OS (12.0 vs 2.4 months, Po0.01).

reduced-intensity conditioning regimen in high-risk AML.25,26 CONFLICT OF INTEREST Hence, a ‘bridge-to-transplantation approach’ with low-intensity The authors declare no conflict of interest. chemotherapy might also be a reasonable option, as chemotherapy-induced toxicities could be avoided as such. However, in our trial only 27% of patients proceeded to REFERENCES allogeneic SCT, displaying the fact that the only curative option 1 Vardiman JW, Thiele J, Arber DA, Brunning RD, Borowitz MJ, Porwit A et al. The in this setting is only available to a minority of patients. For 2008 revision of the World Health Organization (WHO) classification of myeloid all other patients, the question remains whether intensive neoplasms and acute leukemia: rationale and important changes. Blood 2009; re-induction chemotherapy, despite being able to induce 114: 937–951. objective response rates, is of a clinical benefit at all or if the 2 Dohner H, Estey EH, Amadori S, Appelbaum FR, Buchner T, Burnett AK et al. survival times can be achieved by less aggressive measures as well. Diagnosis and management of acute myeloid leukemia in adults: recommenda- In terms of toxicity, our data show that chemomodulation of the tions from an international expert panel, on behalf of the European LeukemiaNet. SHAI regimen with fludarabine resulted in acceptable toxicity. Blood 2010; 115: 453–474. 3 Ellison RR, Holland JF, Weil M, Jacquillat C, Boiron M, Bernard J et al. Arabinosyl After F-SHAI, G-CSF had to be administered 4 days longer, cytosine: a useful agent in the treatment of acute leukemia in adults. Blood 1968; indicating an increased myelosuppression. In addition, there was a 32: 507–523. significantly higher rate of nausea and vomiting, pulmonary side 4 Ohno R, Hirano M, Imai K, Koie K, Kamiya T, Nishiwaki H et al. Daunorubicin, effects and (non-lethal) bleeding. However, this increase in toxicity cytosine arabinoside, 6- riboside, and prednisolone (DCMP) did not translate into a higher rate of EDs or severe infections. In combination chemotherapy for acute myelogenous leukemia in adults. Cancer general, toxicities XIII were observed in the majority of patients in 1975; 36: 1945–1949. both treatment arms. Both arms comprised high-dose cytarabine 5 Lowenberg B, Ossenkoppele GJ, van Putten W, Schouten HC, Graux C, Ferrant A et al. with up to 3 g/m2 per single infusion, and in the light of one High-dose daunorubicin in older patients with acute myeloid leukemia. NEnglJMed recent study that implied a ceiling effect for high-dose cytarabine, 2009; 361: 1235–1248. 27 6 Braess J, Spiekermann K, Staib P, Gruneisen A, Wormann B, Ludwig WD et al. after which only side effects increase, dose modifications of the Dose-dense induction with sequential high-dose cytarabine and mitoxantone (F-)SHAI regimen might be considered for future studies. However, (S-HAM) and pegfilgrastim results in a high efficacy and a short duration of critical the toxicities observed in our trial, although severe, did not exceed in de novo acute myeloid leukemia: a pilot study of the AMLCG. the expected rate of complications. ED rates were in the upper Blood 2009; 113: 3903–3910. limits of death rates that have been reported for FLAG-Ida. 7 Nazha A, Kantarjian H, Ravandi F, Huang X, Choi S, Garcia-Manero G et al. In conclusion, chemomodulation of a SHAI regimen by Clofarabine, idarubicin, and cytarabine (CIA) as frontline therapy for patients p60 fludarabine improves antileukemic efficacy, but only at a modest years with newly diagnosed acute myeloid leukemia (AML). Am J Hematol 2013; rate. Allogeneic SCT remains the therapy of choice, but can only 88: 961–966. be performed in the minority of patients. Given the efficacy of 8 Burnett AK, Hills RK, Milligan D, Kjeldsen L, Kell J, Russell NH et al. Identification of patients with acute myeloblastic leukemia who benefit from the addition of chemomodulation on cytarabine even in a relapsed setting, gemtuzumab ozogamicin: results of the MRC AML15 trial. J Clin Oncol 2011; 29: incorporation of fludarabine into first-line therapy of AML seems 369–377. warranted and may help to improve the outcome of patients 9 Kufe DW, Major PP, Egan EM, Beardsley GP. Correlation of cytotoxicity with with AML. incorporation of ara-C into DNA. J Biol Chem 1980; 255: 8990–8997.

Leukemia (2014) 1001 – 1007 & 2014 Macmillan Publishers Limited SHAI versus F-SHAI in relapsed or refractory AML M Fiegl et al 1007 10 Plunkett W, Liliemark JO, Adams TM, Nowak B, Estey E, Kantarjian H et al. 20 Burnett AK, Russell NH, Hills RK, Hunter AE, Kjeldsen L, Yin J et al. Optimization of Saturation of 1-beta-D-arabinofuranosylcytosine 5’-triphosphate accumulation in chemotherapy for younger patients with acute myeloid leukemia: results of the leukemia cells during high-dose 1-beta-D-arabinofuranosylcytosine therapy. medical research council AML15 trial. J Clin Oncol 2013; 31: 3360–3368. Cancer Res 1987; 47: 3005–3011. 21 Kern W, Aul C, Maschmeyer G, Schonrock-Nabulsi R, Ludwig WD, Bartholomaus A 11 Gandhi V, Plunkett W. -specific metabolism of arabinosyl nucleosides in et al. Superiority of high-dose over intermediate-dose cytosine arabinoside K562 human leukemia cells. Cancer Chemother Pharmacol 1992; 31: 11–17. in the treatment of patients with high-risk acute myeloid leukemia: results 12 Gandhi V, Estey E, Keating MJ, Plunkett W. Fludarabine potentiates metabolism of of an age-adjusted prospective randomized comparison. Leukemia 1998; 12: cytarabine in patients with acute myelogenous leukemia during therapy. J Clin 1049–1055. Oncol 1993; 11: 116–124. 22 Milligan DW, Wheatley K, Littlewood T, Craig JI, Burnett AK. Fludarabine and 13 Estey EH. Treatment of relapsed and refractory acute myelogenous leukemia. cytosine are less effective than standard ADE chemotherapy in high-risk acute Leukemia 2000; 14: 476–479. myeloid leukemia, and addition of G-CSF and ATRA are not beneficial: results of 14 Parker JE, Pagliuca A, Mijovic A, Cullis JO, Czepulkowski B, Rassam SM et al. the MRC AML-HR randomized trial. Blood 2006; 107: 4614–4622. Fludarabine, cytarabine, G-CSF and idarubicin (FLAG-IDA) for the treatment of 23 Russo D, Malagola M, de Vivo A, Fiacchini M, Martinelli G, Piccaluga PP et al. poor-risk myelodysplastic syndromes and acute myeloid leukaemia. Br J Haematol Multicentre phase III trial on fludarabine, cytarabine (Ara-C), and idarubicin 1997; 99: 939–944. versus idarubicin, Ara-C and etoposide for induction treatment of younger, 15 Pastore D, Specchia G, Carluccio P, Liso A, Mestice A, Rizzi R et al. FLAG-IDA in the newly diagnosed acute myeloid leukaemia patients. Br J Haematol 2005; 131: treatment of refractory/relapsed acute myeloid leukemia: single-center experi- 172–179. ence. Ann Hematol 2003; 82: 231–235. 24 Holowiecki J, Grosicki S, Giebel S, Robak T, Kyrcz-Krzemien S, Kuliczkowski K et al. 16 Steinmetz HT, Schulz A, Staib P, Scheid C, Glasmacher A, Neufang A et al. Phase-II Cladribine, but not fludarabine, added to daunorubicin and cytarabine during trial of idarubicin, fludarabine, cytosine arabinoside, and filgrastim (Ida-FLAG) for induction prolongs survival of patients with acute myeloid leukemia: a multi- treatment of refractory, relapsed, and secondary AML. Ann Hematol 1999; 78: center, randomized phase III study. J Clin Oncol 2012; 30: 2441–2448. 418–425. 25 Schmid C, Schleuning M, Ledderose G, Tischer J, Kolb HJ. Sequential regimen 17 Yavuz S, Paydas S, Disel U, Sahin B. IDA-FLAG regimen for the therapy of primary of chemotherapy, reduced-intensity conditioning for allogeneic stem-cell refractory and relapse acute leukemia: a single-center experience. Am J Ther 2006; transplantation, and prophylactic donor lymphocyte transfusion in high-risk 13: 389–393. acute myeloid leukemia and myelodysplastic syndrome. J Clin Oncol 2005; 23: 18 Clavio M, Gatto S, Beltrami G, Quintino S, Canepa L, Pierri I et al. Fludarabine, ARA- 5675–5687. C, idarubicin and G-CSF (FLAG-Ida), high dose ARA-C and early stem cell trans- 26 Schmid C, Schleuning M, Schwerdtfeger R, Hertenstein B, Mischak-Weissinger E, plant. A feasable and effective therapeutic strategy for de novo AML patients. Bunjes D et al. Long-term survival in refractory acute myeloid leukemia after J Exp Clin Cancer Res 2002; 21: 481–487. sequential treatment with chemotherapy and reduced-intensity conditioning for 19 Kim I, Koh Y, Yoon SS, Park S, Kim BK, Kim DY et al. Fludarabine, cytarabine, and allogeneic stem cell transplantation. Blood 2006; 108: 1092–1099. attenuated-dose idarubicin (m-FLAI) combination therapy for elderly acute 27 Lowenberg B, Pabst T, Vellenga E, van Putten W, Schouten HC, Graux C et al. myeloid leukemia patients. Am J Hematol 2013; 88: 10–15. Cytarabine dose for acute myeloid leukemia. N Engl J Med 2011; 364: 1027–1036.

& 2014 Macmillan Publishers Limited Leukemia (2014) 1001 – 1007