Fax: (404) 727-4455
AAML03P1, a pilot study of the safety of gemtuzumab ozogamicin in combination with chemotherapy for newly diagnosed childhood acute myeloid leukemia
A report from the Children's Oncology Group
Article first published online: 15 JUL 2011
Copyright © 2011 American Cancer Society
Volume 118, Issue 3, pages 761–769, 1 February 2012
How to Cite
Cooper, T. M., Franklin, J., Gerbing, R. B., Alonzo, T. A., Hurwitz, C., Raimondi, S. C., Hirsch, B., Smith, F. O., Mathew, P., Arceci, R. J., Feusner, J., Iannone, R., Lavey, R. S., Meshinchi, S. and Gamis, A. (2012), AAML03P1, a pilot study of the safety of gemtuzumab ozogamicin in combination with chemotherapy for newly diagnosed childhood acute myeloid leukemia. Cancer, 118: 761–769. doi: 10.1002/cncr.26190
- Issue published online: 20 JAN 2012
- Article first published online: 15 JUL 2011
- Manuscript Accepted: 15 MAR 2011
- Manuscript Revised: 3 MAR 2011
- Manuscript Received: 27 JAN 2011
- gemtuzumab ozogamicin;
- newly diagnosed;
- childhood acute myeloid leukemia;
The development of antigen-targeted therapies may provide additional options to improve outcomes in children with acute myeloid leukemia (AML). The Children's Oncology Group AAML03P1 trial sought to determine the safety of adding 2 doses of gemtuzumab ozogamicin, a humanized anti-CD33 antibody-targeted agent, to intensive chemotherapy during remission induction and postremission intensification for children with de novo AML.
AAML03P1 enrolled 350 children with previously untreated AML. Patients with a matched family donor received 3 courses of chemotherapy followed by hematopoietic stem cell transplantation; those without a matched family donor received 5 courses of chemotherapy. Gemtuzumab ozogamicin 3 mg/m2/dose was administered on Day 6 of Course 1 and Day 7 of Course 4.
Toxicities observed in all courses of therapy were typical of AML chemotherapy regimens, with infection being most common. Patients achieved a complete remission rate of 83% after 1 course and 87% after 2 courses. The mortality rate was 1.5% after the first gemtuzumab ozogamicin-containing induction course and 2.6% after 2 induction courses. The 3-year event-free survival and overall survival rates were 53 ± 6% and 66 ± 5%, respectively.
This trial determined that it is safe and feasible to include gemtuzumab ozogamicin in combination with intensive chemotherapy. The survival rates compare favorably with the recently published results of clinical trials worldwide. Cancer 2012;. © 2011 American Cancer Society.
The prognosis for children with acute myeloid leukemia (AML) has improved over the past 3 decades.1-10 Complete remission (CR) rates as high as 93% and overall survival (OS) rates of up to 65% are now reported worldwide. Dose intensification, improved supportive care, and effective salvage regimens have transformed a disease that was once uniformly fatal into one that is potentially curable for most patients. Despite these advances, the cure rate for children with AML lags behind that for children with acute lymphoblastic leukemia due foremost to relapse.
Specific treatment strategies for children with AML have focused on intensifying induction and postremission therapy. Children's Cancer Group (CCG) study 2891 successfully tested intensively timed induction chemotherapy consisting of dexamethasone, cytarabine (ara-C), thioguanine, etoposide, and rubidomycin (daunomycin).11 The successor trial, CCG-2961, incorporated idarubicin into this induction regimen, achieving 3-year event-free survival (EFS) and OS rates of 48 ± 5% and 63 ± 5%, respectively, after much more aggressive supportive care measures were instituted.12 The Medical Research Council (MRC) clinical trials have attempted to intensify the dosing, duration, and total number of courses of treatment for patients with AML. By using an extended induction regimen and postinduction intensification therapy based on high-dose ara-C and anthracyclines, the MRC 10 and 12 clinical trials achieved excellent 5-year EFS and OS rates (MRC 10: 48% and 57%; MRC 12: 55% and 67%).5
As our understanding of the biology of AML improves, the development of antigen-targeted therapies provides us with additional options to improve outcomes. Gemtuzumab ozogamicin is a recombinant humanized anti-CD33 monoclonal antibody linked to calicheamicin, a potent antitumor antibiotic.13 Phase 1 and 2 studies have shown the safety and efficacy of gemtuzumab ozogamicin both as a single agent and in combination with chemotherapy in adults and children.14-19 There was concern about liver toxicity in adult trials using higher doses (6 and 9 mg/m2) in combination with cytotoxic agents.20 Pediatric and younger adult clinical trials have successfully addressed these concerns by reducing the dose of gemtuzumab ozogamicin, eliminating repeated or consecutive course doses, and avoiding administration in tandem with thioguanine.16,19,21 In July of 2010, gemtuzumab ozogamicin was voluntarily withdrawn from the US market based in part on the results of Southwest Oncology Group (SWOG) study 106. This study demonstrated no benefit when gemtuzumab ozogamicin was added to induction or postconsolidation therapy, and reported an increase in induction deaths compared with those receiving chemotherapy alone (5.4 % vs 1.4 %, respectively). However, these results may be confounded because of a lower dose of daunorubicin in those patients receiving gemtuzumab ozogamicin as well as an induction death rate lower than that usually seen in chemotherapy trials for patients of that age.22 In fact, MRC 15 recently reported no significant increase in toxicity in those randomized to a gemtuzumab ozogamicin-containing arm. Importantly, this study showed a significant survival benefit for patients with favorable cytogenetics who received gemtuzumab ozogamicin during induction. This effect mimicked that seen in the SWOG 106 study.23 Definitive randomized trials evaluating the effect of gemtuzumab ozogamicin continue in the Children's Oncology Group and the Berlin, Frankfurt, Munster Cooperative Group.
Here we describe the findings of the Children's Oncology Group pilot study AAML03P1. On the basis of the excellent EFS and OS and the low treatment-related mortality (TRM) rates (9%) observed in MRC 12, the Children's Oncology Group opted to test the feasibility of a modified MRC regimen in combination with gemtuzumab ozogamicin. The primary aim of this study was to determine the safety and CR rate of the well-established, 2-course intensive induction regimen with the addition of a single dose of gemtuzumab ozogamicin. In addition, the study evaluated the safety of adding a subsequent dose of gemtuzumab ozogamicin to the mitoxantrone and cytarabine intensification course administered to children not undergoing a matched family donor hematopoietic stem cell transplantation (HSCT). Initially, the study was designed to accrue 150 patients. A subsequent amendment expanded AAML03P1 enrollment to accrue 15 additional patients with primary induction failure in whom a suitable unrelated donor could be identified. As a result, AAML03P1 enrolled 350 patients, making it the largest pediatric de novo AML clinical trial since the closure of CCG-2961, and among the first to implement MRC-based therapy in the United States.
MATERIALS AND METHODS
AAML03P1 opened to accrual in December 2003 and completed accrual in November 2005. Data analyses are current as of August 21, 2009. Eligible patients were ≥1 month and ≤21 years old with primary, untreated AML who met diagnostic criteria as set out by the French-American-British (FAB) classification system. Patients with acute promyelocytic leukemia, juvenile myelomonocytic leukemia, documented bone marrow (BM) failure syndromes, Down syndrome, or secondary/treatment-related leukemia were not eligible. Patients with myelodysplastic syndrome were not eligible unless they presented with karyotypic abnormalities characteristic of de novo AML. In May of 2005, an amendment expanded AAML03P1 enrollment to accrue 15 additional patients with primary induction failure in whom a suitable unrelated donor could be identified (or 180 additional patients from whom these additional 15 could be identified). This was necessary to achieve a secondary objective of the study, which sought to determine the feasibility of finding a suitable unrelated stem cell donor within 21 days of study entry. Institutional review boards at participating centers approved the study, and participating patients or their parents signed written informed consent. The original clinical trial was registered at www.clinicaltrials.gov as NCT00070174.
Risk classifications were not assigned prospectively, as the primary aim was limited to safety across all risk groups. Morphologic, cytogenetic, and molecular analyses were performed according to study guidelines. FAB data were limited to institutional review only. The cytogenetic data presented here are a combination of institutional and central review. Because this was a retrospective subanalysis, response to induction therapy was not included in this risk assignment. Low-risk patients were defined for analysis purposes as those having inv(16)/t(16;16) or t(8;21), and high-risk patients were defined as those having monosomy 7 or del(5q). Patients with an accepted cytogenetic sample who did not meet criteria for low or high risk were defined as intermediate risk. Patients without cytogenetic data were not included in the retrospective subanalysis.
Figure 1 shows the schema of the study, and the legend describes the details of drug dose and administration. Treatment consisted of a remission induction phase (Induction I and Induction II) followed by an intensification phase (Intensification I, II, and III). Patients with <20% blasts after Induction I as determined by BM examination went on to receive a second course of Ara-C, daunorubicin, and etoposide, and those with ≥20% blasts were removed from the study. Patients with ≥5% blasts after Induction II were removed from the study, and those with <5% blasts received systemic intensification therapy based on the MRC trials. Intensification I consisted of high-dose ara-C and etoposide and was based on the MRC amsacrine, cytarabine, and etoposide (MACE) regimen. Amsacrine, an intercalator and topoisomerase II inhibitor, was not available in the United States and was replaced by a higher dose of cytarabine than was given in the MRC trials, 10 g/m2/course versus 1 g/m2/course. The etoposide dose was increased from 500 mg given in the MRC trials to 750 mg. After Intensification I, patients with a 5 of 6 or 6 of 6 matched family donor were assigned to receive allogeneic HSCT. Patients without matched family donors received 2 additional intensification cycles (total of 3). Intensification II consisted of mitoxantrone and cytarabine with a slightly different administration schedule than that given in the MRC trials but identical to that used and piloted in prior CCG trials.24 Central nervous system (CNS) prophylaxis consisted of intrathecal cytarabine on Day 1 of Induction I and II, as well as Day 1 of Intensification I and II. For patients with overt CNS disease at diagnosis (any number of blasts in an atraumatic lumbar puncture), intrathecal cytarabine was given twice weekly until the cerebrospinal fluid was clear of blasts, followed by 2 additional intrathecal treatments.
It was suggested, but not required, that patients have an absolute neutrophil count >1000/μL and a platelet count >106/μL before beginning each course. CR was defined as <5% blasts with trilineage maturation and no evidence of extramedullary disease. Partial remission (PR) was defined as the presence of 5% to 20% blasts after 1 course of induction therapy seen on 2 separate BM examinations by morphology or the persistence of extramedullary disease. Supportive care guidelines were consistent with those implemented in postamendment CCG-2961.12
Statistical Plan and Analysis
The primary outcome measures were remission status after Induction I and II, OS, EFS, disease-free survival (DFS), relapse risk, and TRM. The significance of observed differences in proportions was tested using the chi-square test and Fisher exact test when data were sparse. The Mann-Whitney test was used to determine the significance of differences between medians. The Kaplan-Meier method was used to estimate OS, EFS, and DFS. OS was defined as the time from study entry (or from the end of Induction I or II for CR patients) to death. EFS was defined as time from study entry until death, Induction I failure (ie, refractory disease with BM blasts >20%, CNS relapse, persistent CNS disease, or a second malignancy), Induction II failure (ie, fail to achieve CR), or relapse. DFS was defined as either the time from the end of Induction I until death, Induction II failure, or relapse or the time from the end of Induction II until death or relapse for patients who achieved CR. Estimates of relapse risk and TRM were obtained using the method of cumulative incidence that accounts for competing events. Relapse risk was calculated as the risk of relapse from the end of Induction I to relapse, progressive disease (PD) death, or Induction II failure, and non-PD deaths were considered competing events. TRM was calculated as the risk of mortality from study entry or from the end of Induction I to non-PD death, where relapses, PD deaths, and induction failures were considered competing events. Relapse risk and TRM were also assessed similarly for patients in CR at the end of Induction II. The significance of predictor variables was tested with Gray's statistic for relapse risk and TRM. Cox proportional hazards models were used for univariate and multivariate analyses comparing study entry characteristics. Patients lost to follow-up were censored at their date of last known contact or at a cutoff 6 months before the date of analyses: August 21, 2009 for AAML03P1 and October 30, 2006 for CCG-2961. For AAML03P1 subjects in CR after Intensification I, outcomes were compared for those with and without an matched family donor.
Of 350 patients enrolled in the study, 340 were eligible. Ten patients were identified as ineligible because of starting treatment before enrollment (3 patients), starting treatment before consent was obtained (1 patient), having AML with an FAB classification of M3 by anatomic pathology report (2 patients), having acute promyelocytic leukemia by cytogenetic report (2 patients), omission of cardiac function evaluation before treatment (1 patient), or not meeting the study definition for AML at study entry (1 patient). Table 1 lists the characteristics of the 340 eligible patients enrolled in AAML03P1. The FAB and cytogenetic data are presented in Table 1. Cytogenetic data were not available for 7% of eligible patients.
|Cytogenetic risk group|
|WBC, ×103/μL, median (range)||19.6||(0.7-495)|
|BM blasts %, median (range)||67||(0-100)|
|Platelets, ×103/μL, median (range)||51||(3-578)|
|Hemoglobin, g/dL, median (range)||8.2||(2.5-64)|
The length in days of each cycle for patients completing all therapy and time to neutrophil and platelet count recovery are summarized in Table 2. Intensification II was the longest cycle of therapy (median, 49 days), only slightly longer than Intensification III (median, 48 days). This was because of prolonged absolute neutrophil count recovery (median, 52 days) and platelet recovery (median, 46 days).
|Cycle||N||Days of Period for Patients Completing All Treatment||Days Until ANC Recovery (≥1000/μL)||Days Until Platelet Recovery (≥100,000/μL)|
|Median (Days)||Range (Days)||Median (Days)||Range (Days)||Median (Days)||Range (Days)|
Table 3 shows the response of patients to Induction I and Induction II therapy. Of the treatment failures, 5.6% had >20% blasts after 1 induction cycle, and 3.2% had >5% blasts after 2 induction cycles. The death rate during Induction I was 1.5%. The causes of death during Induction I included hemorrhage (n = 2), acute respiratory distress syndrome (n = 1), postoperative abdominal bleeding (n = 1), and multiorgan failure (n = 1). For Induction I and II together, the overall death rate was 2.6%, with 3 deaths attributed to infection during Induction II. Overall, 34 patients were not evaluable by the end of 2 cycles. This included 9 patients who withdrew after Induction I with a CR (n = 4) or PR (n = 5), 9 patients who withdrew from study without a BM evaluation performed after Induction I, and 16 patients without a BM evaluation performed after Induction II. The subsequent TRM rate in later courses of therapy from the end of Induction I for patients in CR was 9 ± 3%, and the TRM rate from the end of Induction II for patients in CR was 7 ± 3%.
|Induction I Response|
|BM blasts >20%||17|
|Persistent CNS disease||2|
|Cumulative response by the end of Induction II|
The most common non–HSCT-related, nonhematologic adverse events (grade ≥3) are summarized in Table 4. Infection data were collected in a manner that may have included minor and nonsterile site infections along with bloodstream infections. As expected, documented infections were the most common toxicity during each treatment course. Transaminitis grade ≥3 did occur in a higher percentage of patients during Induction I (aspartate aminotransferase [AST], 9%; alanine aminotransferase [ALT], 11%) than in Induction II (AST, 3%; ALT, 4%), which could possibly reflect a gemtuzumab ozogamicin effect. Elevations in ALT/AST were common during Intensification II (11% and 13%, respectively), but were just as high during Intensification III (10% and 15%, respectively), which did not contain gemtuzumab ozogamicin. Cardiac toxicities grade ≥3 did not exceed 3% in any course. One exception was hypotension, which was probably related to sepsis. Hypotension was reported in 2% of patients during Induction I, 3% during Induction II, 6% during Intensification I, 7% during Intensification II, 7% during Intensification III, and 12% during HSCT. One patient became hypotensive during Induction I, and the attribution indicated probable relation to gemtuzumab ozogamicin. For other patients, there was no recorded attribution to gemtuzumab ozogamicin.
|Induction I, N (%)||Induction II, N (%)||Intensification I, N (%)||Intensification II, N (%)||Intensification III, N (%)|
|Anorexia||55 (16%)||25 (9%)||28 (10%)||39 (19%)||19 (11%)|
|Fever and Neutropenia||105 (31%)||57 (19%)||59 (21%)||27 (13%)||30 (18%)|
|Documented infectionsa||184 (54%)||129 (44%)||171 (61%)||167 (81%)||138 (82%)|
|Hypokalemia||66 (19%)||27 (9%)||25 (9%)||42 (20%)||31 (18%)|
Veno-occlusive disease (VOD) was defined as 1) weight gain of >10% of baseline, 2) right upper quadrant pain or tender hepatomegaly, 3) jaundice with total bilirubin of 2 mg/dL, and 4) edema or ascites. VOD developed in 19 patients (5.6%) during therapy (n = 11) or during follow-up (n = 8), including 7 patients who received HSCT after a relapse. Two patients were diagnosed during Intensification II, at 15 and 18 days after receiving gemtuzumab ozogamicin. Of 9 patients with VOD during transplant, the diagnosis was made a mean of 23 days from the beginning of that reporting period. One death was reported to be a result of VOD that developed Day +16 post-HSCT.
Patients with an matched family donor were to receive an HSCT after Intensification I. Table 5 lists intent-to-treat survival outcomes of AAML03P1, comparing those with matched family donors (73 patients) and those without (195 patients). Children with a matched family donor had a lower 3-year relapse risk than those without (18 ± 10% vs 36 ± 7%, P = .005). The 3-year DFS rate was 75 ± 12% versus 57 ± 7% (P = .008), but the difference in OS rates (80 ± 10% vs 73 ± 6%, P = .171) was not statistically significant. Three deaths were reported during HSCT, and the causes were listed as infection, VOD, and acute respiratory distress syndrome. A review of treatment outcomes for patients with a matched family donor by risk group did not reveal any data of statistical significance.
|N||HR||Estimate at 3 years ± 2 SE (%)||N||HR||Estimate at 3 years ± 2 SE (%)|
|DFS from end of Int I in CR||73||1||75 ± 12||195||1.99||57 ± 7||.008|
|OS from end of Int I in CR||73||1||80 ± 10||195||1.52||73 ± 6||.171|
|RR from end of Int I in CR||73||1||18 ± 10||195||2.35||36 ± 7||.005|
|TRM from end of Int I in CR||73||1||7 ± 6||195||1.13||7 ± 3||.873|
Table 6 lists multivariate Cox analysis results for commonly regarded prognostic factors in pediatric AML. Analysis was restricted to include only patients with all data available (n = 305). Black patients had significantly worse EFS and OS than patients of other racial groups, with hazard ratios (HRs) of 1.93 (P = .004) and 2.48 (P<.001), respectively. Consistent with findings from previous AML trials, univariate analysis of EFS showed HRs of 0.55 (P = .007) for those with low-risk karyotypes compared with standard risk and 2.69 (P = .002) for those with high-risk cytogenetic features compared with standard risk. Black race along with low-risk and high-risk cytogenetic features retained statistical significance in a multivariate analysis.
|N||EFS From Study Entry||OS From Study Entry|
|HR||95% CI||P||HR||95% CI||P|
|Cytogenetics (risk class)|
|White (not Hispanic)||178||1||1|
|Black (not Hispanic)||41||1.93||1.24–3.02||.004||2.48||1.50–4.10||<.001|
|Hispanic or other||86||0.97||0.66–1.44||.891||1.07||0.67–1.72||.767|
|CNS at diagnosis|
|Age at diagnosis (years)|
Table 7 shows the EFS and OS for the study and by remission status after Induction I and II. The median follow-up time for patients alive at last contact was 98.8 months, with an upper range of 154.4 months. The 3-year EFS rate from study entry was 53 ± 6%, and the OS rate was 66 ± 5% (Fig. 2). Patients in remission after Induction II achieved 3-year DFS and OS of 60 ± 6% and 72 ± 6%, respectively.
|N||3 yr % ± 2 SE (%)|
|Outcome for de novo patients|
|EFS from study entry||340||53 ± 6|
|OS from study entry||340||66 ± 5|
|TRM from study entry||340||9 ± 3|
|DFS from end of course 1 for patients in CR||270||58 ± 6|
|OS from end of course 1 for patients in CR||270||71 ± 6|
|RR from end of course 1 for patients in CR||270||33 ± 6|
|TRM from end of course 1 for patients in CR||270||9 ± 3|
|DFS from end of course 2 for patients in CR||266||60 ± 6|
|OS from end of course 2 for patients in CR||266||72 ± 6|
|RR from end of course 2 for patients in CR||266||33 ± 6|
|TRM from end of course 2 for patients in CR||266||7 ± 3|
This study showed that it is safe and feasible to add a dose of gemtuzumab ozogamicin to an intensive induction and a second dose to a postremission intensification regimen in children with newly diagnosed AML. As expected, infection was the most common complication. The most common infections were sepsis and catheter-related infections. In the 2 courses in which gemtuzumab ozogamicin was added, the rates of infection were 54% and 81%. These high numbers reflect that the study included information on minor infections as well as those grade ≥3. A comprehensive review of the AAML03P1 infection data are underway and will be reported in a separate article. However, the infection rates reported in the gemtuzumab ozogamicin-containing regimens did not differ substantially from the other chemotherapy courses within this trial. The addition of gemtuzumab ozogamicin did not appear to significantly prolong the time to neutrophil or platelet recovery. The median duration of Induction I was 36 days, consistent with comparable induction regimens.16,19,25 The median duration of Intensification II was 47 days, similar to the 42-day recovery period seen with the CCG-2951 regimen, which used mitoxantrone and cytarabine without gemtuzumab ozogamicin for children with AML in relapse.24
The addition of gemtuzumab ozogamicin did not appear to affect TRM. The observed toxic mortality rate in Induction I was 1.5% (5 patients), which compares favorably with the early death rates in postamendment CCG-2961 (4.3%) and MRC 12 (4%). The TRM rate for Intensification II was 1.8%, below the rate seen using comparable regimens containing mitoxantrone and cytarabine, as in CCG-2951 (3%) and AAML00P2 Arm A (5%).19,24 The primary safety objectives in this trial were met, demonstrating that the addition of gemtuzumab ozogamicin to established chemotherapy regimens is tolerable in pediatric patients.
Of particular concern before this study was the incidence of VOD with gemtuzumab ozogamicin administration. Known risk factors for VOD in those receiving gemtuzumab ozogamicin include higher doses (6-9 mg/m2), coadministration of hepatotoxic drugs such as 6-thioguanine, administration in sequential courses of therapy, and administration within 3 months of HSCT.16,18,26,27 The observed rate of VOD in this study for those who received HSCT (18%) did not differ significantly from those seen with other regimens that combine gemtuzumab ozogamicin with chemotherapy19,27 and was consistent with the general HSCT population, in whom VOD rates as high as 18% have been reported.21,28
This study was intended to determine the CR rate after 2 courses of Ara-C, daunorubicin, and etoposide plus a single dose of gemtuzumab ozogamicin in the initial course. This primary objective was met, with a cumulative response rate of 87% (95% confidence interval, 83%-90%) after 2 courses. The CR rates were similar to those seen with CCG-2961, which reported a CR/PR rate of 88%. The MRC 10 and 12 regimens, on which the backbone chemotherapy was based, both yielded CR rates of 93%. However, these CR rates were reported after 4 courses of induction chemotherapy. MRC 10 yielded a 63% response rate after Induction I and an 83% response rate after Induction II.29 MRC15 did not include enough patients on the Ara-C, daunorubicin, and etoposide + gemtuzumab ozogamicin arm to make a valid comparison, but the overall induction CR rate for all patients was 83% after 2 cycles.23 The 3-year EFS and OS of AAML03P1 were at least equivalent to postamendment CCG-2961 (Fig. 2), and comparable to the 5-year EFS and OS rates reported in MRC 12 (EFS rate, 55%; OS rate, 67%). Definitive comparisons between all of these studies are difficult because of potential biases encountered when using historical comparisons.30
AAML03P1 showed that it is safe and feasible to include gemtuzumab ozogamicin in combination with cytotoxic chemotherapy and that the EFS and OS rates were at least equivalent to those seen in CCG-2961. Future studies will build upon our refinement of risk stratification and molecular classification by moving novel agents and chemotherapy combinations into phase 3 clinical trials.
The work was supported by National Institutes of Health grants: Children's Oncology Group Chair's grant NIH U10 CA98543 and SDC U10 CA98413.
CONFLICT OF INTEREST DISCLOSURES
The authors made no disclosures.
- 4Treatment of childhood acute myeloblastic leukemia: dose intensification improves outcome and maintenance therapy is of no benefit—multicenter studies of the French LAME (Leucemie Aigue Myeloblastique Enfant) Cooperative Group. Leukemia. 2005; 19: 2082-2089., , , et al.
- 22Preliminary results of Southwest Oncology Group Study S0106: an international intergroup phase 3 randomized trial comparing the addition of gemtuzumab ozogamicin to standard induction therapy versus standard induction therapy followed by a second randomization to post-consolidation gemtuzumab ozogamicin versus no additional therapy for previously untreated acute myeloid leukemia [abstract]. Blood. 2009; 114: 790., , , et al.
- 25Comparison of ‘sequential’ versus ‘standard’ chemotherapy as re-induction treatment, with or without cyclosporine, in refractory/relapsed acute myeloid leukaemia (AML): results of the UK Medical Research Council AML-R trial. Br J Haematol. 2001; 113: 713-726., , , .
- 29A simple, robust, validated and highly predictive index for the determination of risk-directed therapy in acute myeloid leukaemia derived from the MRC AML 10 trial. United Kingdom Medical Research Council's Adult and Childhood Leukaemia Working Parties. Br J Haematol. 1999; 107: 69-79., , , et al.