Search

INTRODUCTION: Allogeneic stem cell transplantation (allo-SCT) is an important curative strategy in adults with high risk acute myeloid leukemia (AML) and myelodysplasia (MDS). Disease relapse represents the major cause of treatment failure and whilst retrospective analyses have identified that pre-transplant measurable residual disease (MRD) is an important predictor of transplant outcome this has never been examined prospectively. The advent of reduced intensity conditioning (RIC) regimens has substantially increased the number of older adults eligible for allo-SCT but the optimal RIC regimen in high risk AML remains unknown. Registry data have demonstrated improved outcomes using a sequential transplant regimen utilizing cytosine arabinoside (araC)/amsacrine (AMSA) cytoreduction followed by a fludarabine (Flu)/busulfan (Bu) based RIC regimen (FLAMSA-Bu). However, although the FLAMSA-Bu regimen is now widely used in adults with high risk AML and MDS its benefit has not been evaluated in a randomized trial. We report the results of a randomized trial evaluating the FLAMSA-Bu regimen compared with standard RIC regimens which also represents the first prospective evaluation of the impact of pre-transplant MRD levels on transplant outcome. PATIENTS AND METHODS: 244 patients (median age 59 yrs) with high risk AML (n=164) or high risk myelodysplasia (n=80) were randomized 1:1 to a control arm determined by investigator's choice of either Flu/B2/ATG (Flu, Bu 3.2 mg/kg x 2 days, ATG 2.5 mg/kg x 2 days); Flu/Mel/Alemtuzumab (A) (Flu, Mel 140 mg/m2, A 50 mg) or Flu/Bu2/A (Flu, Bu 3.2 mg/kg x 2 days, A 50 mg) versus an experimental arm of FLAMSA-Bu (Flu, araC 2g/m2 x 4 days, AMSA 100mg/m2 x 4 days, Bu -total dose 11.2 mg/kg). Patients over the age of 60 received an adjusted FLAMSA-Bu regimen utilising a reduced dose of araC (1mg/m2 x 4 days) and a total Bu dose of 6.4 mg/kg. Patients were transplanted using either an HLA identical sibling (n=49) or matched (10/10 or 9/10) unrelated donor (n=195). All patients received cyclosporine GVHD prophylaxis. 155 patients with AML were in CR1 or CR2 at the time of transplant and 9 had primary refractory disease. MRD was monitored by flow cytometry (applying different-from-normal analysis when no diagnostic/relapse leukemic aberrant immunophenotype was available). Pre-transplant MRD levels were measured up to four weeks prior to transplantation in 201 patients (MRD positive = 80 (40%), MRD negative = 94 (47%), inadequate sample = 27 (13%)). MRD results were not reported to clinicians. The primary outcome was overall survival. RESULTS: Baseline characteristics including CR1/CR2 status, adverse cytogenetics and MRD levels were similar between regimens. Median follow up was 35 months. Transplant outcomes were comparable between patients allografted in the control and FLAMSA-Bu arms. 2 yr overall survival (OS) and cumulative incidence of relapse (CIR) were 61% and 30% respectively in the control arm vs 62% and 26% for the FLAMSA-Bu arm. Transplant related mortality at 100 days was 3.0% in patients allografted using the control regimen vs 14% in patients allografted using the FLAMSA-Bu regimen and 17% vs 21% at 1 year. In the study cohort pre-transplant MRD positivity was associated with both an increased CIR compared to patients testing MRD negative (2 yr CIR 42% vs 19%, p=0.009) and decreased OS (2 yr OS 52% vs 71%, p=0.048). The FLAMSA-Bu regimen failed to improve OS or reduce CIR in either MRD positive or MRD negative patients. CONCLUSIONS: This trial, the largest randomized trial of RIC regimens in AML to date, did not detect any benefit of intensification of the conditioning regimen in adults with high risk AML or MDS. Specifically, the FLAMSA-Bu regimen was not associated with improved transplant outcome in patients who were MRD positive pre-transplant. These data include the first demonstration in a prospective analysis that the presence of pre-transplant MRD measured in real time is associated with reduced OS consequent upon an increased risk of disease relapse. Further randomized studies of novel conditioning regimens in adult AML, crucially with integrated MRD studies, are now required but these results support exploration of alternative strategies, such as pre or post-transplant pharmacological intervention, as the most promising strategy to reduce the risk of disease relapse post allograft. Disclosures Russell: Jazz: Consultancy, Honoraria, Speakers Bureau; DSI: Consultancy, Honoraria, Speakers Bureau; Pfizer Inc: Consultancy, Honoraria, Speakers Bureau; Astellas: Consultancy, Honoraria, Speakers Bureau. Freeman:Jazz Pharmaceuticals: Speakers Bureau. OffLabel Disclosure: We report data using the combination of fludarabine, busulphan, amsacrine and cytosine arabinsoide as a conditioning regimen in patients allografted for high risk acute myeloid leukemia

Introduction Most children with acute myeloid leukaemia (AML) harbour fusion genes which are ideal targets for molecular minimal residual disease (MRD) monitoring. However, evidence of prognostic significance is currently lacking and consequently most current paediatric AML treatment protocols rely on flow cytometric (FCM) evaluation to allocate treatment. Molecular MRD techniques provide significantly greater sensitivity and specificity and could allow more accurate outcome prediction, and consequently more personalised therapy, which is highly relevant in a disease where treatment related mortality, morbidity and relapse remain significant. Methods Between June 2016 and February 2019, MyeChild01 enrolled 170 children aged 0-18y with newly diagnosed AML who were randomly assigned to induction therapy with liposomal daunorubicin or mitoxantrone with cytarabine with or without gemtuzumab ozogamicin. Consolidation treatment was determined by karyotype, mutational profile and MRD status. Comprehensive centralised diagnostic assessment consisted of: Karyotype and fluorescence in-situ hybridisation (FISH) using a custom panel of probes to detect paediatric AML-associated gene fusions.PCR based screening for mutations in FLT3, NPM1 and CEPBA.Targeted capture of known fusion loci and paired end sequencing.RNA-seq using the Illumina TruSight fusion panel. Where a fusion gene was identified, RT-qPCR assays were designed and optimised for each patient. NPM1 mutation was also used as an MRD target if present. Paired PB and BM samples were requested after each cycle of treatment. Patients could have sequential monitoring after completion of therapy although this was not mandated. For this analysis, patients with core-binding factor (CBF) leukaemias i.e. inv(16)(p13q22) or t(8;21)(q22q22) with transcript levels above previously defined thresholds (Yin et al, 2012) were considered MRD positive. For all other targets, amplification in at least 2/3 replicates at For patients with CBF or NPM1 mutation, both molecular and FCM MRD status contributed to treatment allocation. Otherwise, FCM MRD was used unless there was no FCM target. Allogeneic stem cell transplantation (HSCT) was recommended for all patients with poor risk cytogenetics, those with intermediate risk cytogenetics who were MRD positive after cycle 2, and those with favourable risk cytogenetics who were MRD positive at the end of cycle 3. After completion of treatment, no specific advice was given regarding management of positive MRD results. Results We identified fusion genes in 126/170 patients (74%). We designed 55 unique RT-qPCR assays to specifically amplify 27 fusions with calculated sensitivity between 1:104 and 1:106. 62/126 children provided a complete set of samples: 31 (60%) with favourable, 18 (29%) with intermediate and 13 (21%) with high risk cytogenetic / molecular profiles. We analysed the effect of molecular MRD status at the end of protocol specified treatment, which included SCT in 17/62 (27%). One year event-free survival from diagnosis was 70% (95% confidence interval 52-82%) in patients who tested MRD negative compared to 11% (1-39%) in patients who tested MRD positive at the end of treatment (p Four patients received non-protocol specified SCT due to physician / family preference based on persistently positive or serially rising transcript levels after completion of treatment. At last follow up 4/4 were alive in complete remission. Rising transcript levels were observed in a further 8 patients to whom no pre-emptive treatment was given due to individual preference or lack of appropriate therapy. Haematological relapse subsequently occurred in 8/8. Conclusion Molecular MRD monitoring is feasible in the majority of children with AML, permitting refinement of response-adapted therapy. Molecular MRD status at the end of treatment is highly predictive of outcome, identifying molecular complete remission as a treatment goal for these children. Serially rising MRD levels reliably predict relapse in the absence of therapeutic intervention. Disclosures Dillon: Abbvie: Consultancy, Honoraria; Novartis: Consultancy, Honoraria; Pfizer: Consultancy, Honoraria; TEVA: Consultancy, Honoraria. Vyas:Astellas: Speakers Bureau; Abbvie: Speakers Bureau; Forty Seven, Inc.: Research Funding; Celgene: Research Funding, Speakers Bureau; Novartis: Research Funding, Speakers Bureau; Pfizer: Speakers Bureau; Daiichi Sankyo: Speakers Bureau. Amrolia:UCLB: Patents & Royalties. Baruchel:Bellicum: Consultancy; Servier: Consultancy, Honoraria; Celgene: Consultancy, Honoraria; Novartis: Consultancy, Honoraria.

BACKGROUND Although the overall survival for children with AML has serially improved due to better supportive care, intelligent use of transplant in 1st complete remission (CR1), and better salvage in CR2, the relapse rate (RR) in CR1 remains unacceptably high at 30-35%. Increasing the intensity of induction chemotherapy may improve outcome. Intensifying anthracycline dosage in induction has been reported to improve outcome in adult AML1, but this remains unproven in children in whom anthracycline-related cardiotoxicity is a major concern. Gemtuzumab ozogamicin (GO), the anti-CD33-calicheamicin conjugated antibody is the most promising new drug being explored in AML treatment presently. Cardiac myocytes do not express CD33 and therefore GO may enable intensification of induction chemotherapy without increasing cardiotoxicity. In children, GO is given as a single 3 mg/m2 dose, however the optimum dose, schedule and timing when combined with chemotherapy remains uncertain. METHODS A dose finding study (DFS) to establish the optimum tolerated number of 3 mg/m2 GO doses that can be combined with induction treatment: cytarabine plus mitoxantrone or liposomal daunorubicin (LD) was embedded within MyeChild 01; an international randomised phase III study for children with AML and high risk MDS. This major DFS recruited pts aged ≥12 months and High-risk patients (defined by poor risk cytogenetics and/or a failure to achieve remission) received fludarabine, cytarabine and idarubicin for course 2. All other patients received a second course of the randomised therapy. Progression to next cohort was determined by an independent data monitoring committee (DMC) taking into account known toxicities of the induction chemotherapy. Dose limiting toxicities (DLTs) were defined as neutrophils ≤1.0 x109/l by d45 post course 1 or 2, non-transfusion dependent platelets ≤80 x109/l by d45 post course 1 or 2 due to documented bone marrow aplasia/hypoplasia and non-haematological events: death from any cause other than AML, veno-occlusive disease (VOD) and any grade ≥3 non-haematological toxicity persisting for >48 hours without resolution to grade ≤2 with specific protocol defined exceptions. RESULTS At data cut-off (27 July 2018), 55 pts were recruited, of which 51 were AML, 1 MDS, 2 myeloid sarcoma and 1 APL (not dosed as ineligible) (Table 1). The chemotherapy randomisation was not balanced in cohort 3 due to suspension of the randomisation when LD became unavailable due to a manufacturing failure; 39 pts received cytarabine with mitoxantrone and 15 cytarabine with LD. Post-course 1, 22 (40%) were stratified as high risk. 54 patients were dosed with GO. In cohort 1; 15 pts received GO 3 mg/m2 on d4; 5 were added as an expansion cohort on the recommendation of the DMC due to 2 patients having unevaluable haematological toxicities. Cohort 2; 20 pts received GO 3 mg/m2 on d 4 and 7. Both cohorts have completed the DLT evaluation period. In cohort 3; 19 pts received GO 3 mg/m2 on days 4, 7 and 10. The DLT evaluation period is near completion for all cohort 3 pts and the data will be presented. Table 2 presents grade ≥3 AEs at data cut off. 29 DLTs were reported, 6 in cohort 1 and 14 in cohort 2. In cohort 3, 9 DLTs are reported but 13 pts are still in the DLT evaluation period. The commonest DLTs were failure to recover either neutrophil or platelet count by d45 post course 1 or 2. DMC review confirmed the reported DLTs for cohorts 1 and 2 were consistent with expected toxicity from the combined chemotherapy alone. Six cases of VOD were reported, all outside the DLT evaluation period, none were fatal. CONCLUSIONS The preliminary findings indicate that 2 doses of 3 mg/m2 GO combined with induction chemotherapy is well tolerated in this patient population, consistent with data from adult AML studies. The DMC concluded for cohorts 1 and 2 that the reported toxicities, including myelosuppression and non-haematological toxicities did not exceed the expected events associated with intensive induction chemotherapy in the absence of GO and there was a notable absence of significant DLTs. There was no apparent difference in AEs between mitoxantrone or LD arms. Cohort 3 DLT reporting is ongoing and the data will be presented.Lowenberg, NEJM 2009; Ohtake, Blood 2011. Disclosures Gibson: Pfizer: Research Funding; Galen: Research Funding. Houlton:Pfizer: Research Funding. Baruchel:Novartis: Membership on an entity's Board of Directors or advisory committees; Shire: Research Funding; Celgene: Consultancy; Amgen: Consultancy; Jazz Pharmaceuticals: Consultancy, Honoraria, Other: Travel, accommodations or expenses; Servier: Consultancy; Roche: Consultancy. Wheatley:Pfizer: Research Funding. Kearns:Galen: Research Funding; Pfizer: Consultancy, Research Funding.

Diagnostic karyotype provides the framework for risk-stratification schemes in acute myeloid leukemia (AML); however, the prognostic significance of many rare recurring cytogenetic abnormalities remains uncertain. We studied the outcomes of 5876 patients (16-59 years of age) who were classified into 54 cytogenetic subgroups and treated in the Medical Research Council trials. In multivariable analysis, t(15;17)(q22;q21), t(8;21)(q22;q22), and inv(16)(p13q22)/t(16;16)(p13;q22) were the only abnormalities found to predict a relatively favorable prognosis (P < .001). In patients with t(15;17) treated with extended all-trans retinoic acid and anthracycline-based chemotherapy, additional cytogenetic changes did not have an impact on prognosis. Similarly, additional abnormalities did not have a significant adverse effect in t(8;21) AML; whereas in patients with inv(16), the presence of additional changes, particularly +22, predicted a better outcome (P = .004). In multivariable analyses, various abnormalities predicted a significantly poorer outcome, namely abn(3q) (excluding t(3;5)(q25;q34)), inv(3)(q21q26)/t(3;3)(q21;q26), add(5q)/del(5q), −5, −7, add(7q)/del(7q), t(6;11)(q27;q23), t(10;11)(p11∼13;q23), other t(11q23) (excluding t(9;11)(p21∼22;q23) and t(11;19)(q23;p13)), t(9;22)(q34;q11), −17, and abn(17p). Patients lacking the aforementioned favorable or adverse aberrations but with 4 or more unrelated abnormalities also exhibited a significantly poorer prognosis (designated “complex” karyotype group). These data allow more reliable prediction of outcome for patients with rarer abnormalities and may facilitate the development of consensus in reporting of karyotypic information in clinical trials involving younger adults with AML. This study is registered at http://www.isrctn.org as ISRCTN55678797 and ISRCTN17161961.

We investigated the benefit of adding all-trans retinoic acid (ATRA) to chemotherapy for younger patients with nonacute promyelocytic acute myeloid leukemia and high-risk myelodysplastic syndrome, and considered interactions between treatment and molecular markers. Overall, 1075 patients less than 60 years of age were randomized to receive or not receive ATRA in addition to daunorubicin/Ara-C/thioguanine chemotherapy with Ara-C at standard or double standard dose. There were data on FLT3 internal tandem duplications and NPM1 mutations (n = 592), CEBPA mutations (n = 423), and MN1 expression (n = 195). The complete remission rate was 68% with complete remission with incomplete count recovery in an additional 16%; 8-year overall survival was 32%. There was no significant treatment effect for any outcome, with no significant interactions between treatment and demographics, or cytarabine randomization. Importantly, there were no interactions by FLT3/internal tandem duplications, NPM1, or CEBPA mutation. There was a suggestion that ATRA reduced relapse in patients with lower MN1 levels, but no significant effect on overall survival. Results were consistent when restricted to patients with normal karyotype. ATRA has no overall effect on treatment outcomes in this group of patients. The study did not identify any subgroup of patients likely to derive a significant survival benefit from the addition of ATRA to chemotherapy. This study is registered at http://www.controlled-trials.com under ISRCTN17833622.

The prognostic impact of tyrosine kinase domain (TKD) mutations of the fms-like tyrosine kinase-3 (FLT3) gene in acute myeloid leukemia (AML) is currently uncertain. To resolve this issue we screened 1107 young adult nonacute promyelocytic leukemia AML patients with known FLT3 internal tandem duplication (ITD) status for FLT3/TKDs; they were detected in 127 (11%) cases. Mutations were associated with a high white cell count (P =.006) and patients with inv(16) (P = .005) but were infrequent in patients with adverse cytogenetics and secondary AML. Overall survival (OS) at 5 years was 53% and 37% for FLT3/TKD mutant and wild-type patients respectively (odds ratio, 0.72; 95% confidence interval, 0.58 to 0.89; P = .002). For both the cumulative incidence of relapse and OS the difference in outcome between FLT3/ITDs and FLT3/TKDs was highly significant (P < .001). In multivariate analysis, impact of FLT3/TKDs on OS when including all mutant-positive patients was not significant, but patients with high-level mutations (more than 25% mutant) had a significantly improved outcome (P = .004). The novel finding that biologically distinct activating mutations of the same gene can be associated with markedly different clinical outcomes has implications for risk stratification and therapy and is significant to the understanding of chemoresistance in AML.

The pathogenesis of acute myeloid leukemia (AML) involves the cooperation of mutations promoting proliferation/survival and those impairing differentiation. The RAS pathway has been implicated as a key component of the proliferative drive in AML. We have screened AML patients, predominantly younger than 60 years and treated within 2 clinical trials, for NRAS (n = 1106), KRAS (n = 739), and HRAS (n = 200) hot-spot mutations using denaturing high-performance liquid chromatography or restriction fragment length polymorphism (RFLP) analysis. NRAS mutations were confirmed in 11% of patients (126/1106) and KRAS mutations in 5% (39/739). No HRAS mutations were detected in 200 randomly selected samples. Codons most frequently mutated were N12 (43%), N13 (21%), and K12 (21%). KRAS mutations were relatively overrepresented in French-American-British (FAB) type M4 (P < .001). NRAS mutation was over-represented in the t(3;5)(q21 approximately 25;q31 approximately q35) subgroup (P < .001) and underrepresented in t(15;17)(q22;q21) (P < .001). KRAS mutation was overrepresented in inv(16)(p13q22) (P = .004). Twenty-three percent of KRAS mutations were within the inv(16) subgroup. RAS mutation and FLT3 ITD were rarely coexistent (14/768; P < .001). Median percentage of RAS mutant allele assayed by quantitative RFLP analysis was 28% (N12), 19% (N13), 25% (N61), and 21% (K12). RAS mutation did not influence clinical outcome (overall/disease-free survival, complete remission, relapse rate) either for the entire cohort or within cytogenetic risk groups.

In an attempt to improve induction chemotherapy for older patients with acute myeloid leukemia (AML),1314 patients were randomized to 1 of 3 induction treatments for 2 courses of DAT (daunorubicin, cytarabine, and thioguanine) 3 + 10, ADE (daunorubicin, cytarabine, and etoposide) 10 + 3 + 5, or MAC (mitoxantrone-cytarabine). The remission rate in the DAT arm was significantly better than ADE (62% vs 50%; P = .002) or MAC (62% vs 55%;P = .04). This benefit was seen in patients younger and older than 70 years. There were no differences between the induction schedules with respect to overall survival at 5 years (12% vs 8% vs 10%). A total of 226 patients were randomized to receive granulocyte colony-stimulating factor (G-CSF) or placebo as supportive care from day 8 after the end of treatment course 1. The remission rate or survival were not improved by G-CSF, although the median number of days to recover neutrophils to 1.0 × 109/L was reduced by 5 days. Patients who entered remission (n = 371) were randomized to stop after a third course (DAT 2 + 7) or after 6 courses, ie, a subsequent COAP (cyclophosphamide, vincristine, cytarabine, and prednisolone), DAT 2 + 5, and COAP. The relapse risk (81% vs 73%), disease-free survival (16% vs 23%), and overall survival at 5 years (23% vs 22%) did not differ between the 3-course or 6-course arms. In addition to a treatment duration randomization, 362 patients were randomized to receive 12-month maintenance treatment with low-dose interferon, but no benefit was seen with respect to relapse risk, disease-free survival, or overall survival.

Introduction: Azacitidine (AZA) represents a significant advance in the treatment of adults with acute myeloid leukaemia (AML) (Dombret et al Blood 2016 126:291-9) and myelodysplasia (MDS) (Fenaux et al Lancet Oncol 2009) 10:223-32) ineligible for intensive chemotherapy. However complete response (CR) rates remain modest and disease progression is almost inevitable. Consequently, additional agents which improve the clinical activity of AZA and other hypomethylating agents are urgently required. Pre-clinical and Phase II studies have identified the potential of co-administration of histone deacteylase inhibitors, notably vorinostat (VOR), with AZA as a strategy to improve its clinical activity. We have therefore compared outcomes after AZA monotherapy or combined AZA and VOR in a prospective randomized trial in adults with high risk AML and MDS. Patients and Methods: 259adults with AML or high risk MDS deemed ineligible for intensive chemotherapy were randomized to receive either AZA (75 mg/m2) x 7 days every 28 days or AZA in combination with VOR 300 mg bd days 3-9 po for a minimum of 6 cycles. Patients were assessed for response, using modified Cheson criteria, after 3 and 6 cycles of therapy. Patients who achieved a major clinical response (CR, CRi or PR) were eligible to receive continued treatment until loss of response, toxicity or death. Major response after 6 cycles and overall survival (OS) were co-primary endpoints. 219 patients had AML (newly diagnosed n= 108) and 44 had MDS (newly diagnosed n= 30). Results: Co-administration of VOR did not increase either the major response rate at 6 months compared with AZA monotherapy (33% AZA v 37% AZA+VOR, OR (95% CI)=1.2 (0.7,2.0), p=0.55) or improve OS (1 year OS AZA 43% versus AZA/VOR 44%, HR (95% CI)=1.06 (0.8, 1.4), p=0.68). No benefit of addition of VOR was observed in any patient subgroup. In multivariate analysis of the study population, a low presentation white blood count (p=0.029) and a diagnosis of MDS vs AML (p=0.0048) was associated with significantly improved survival, whilst disease status (CR v advanced disease) (p=0.023) and low presentation white blood count (p=0.034) significantly improved major response rates. Presentation karyotype was not predictive of either survival or response. Both adverse events (AE) and serious adverse events (SAE) were balanced between treatment arms with 64 and 89 patients in the AZA arm experiencing AE and SAEs respectively versus 75 and 97 in the combination arm respectively. Discussion: To our knowledge this is the largest randomized study of epigenetic therapy in AML and MDS. We conclude that the addition ofVOR to AZA therapy does not increase either major response rates or overall survival in patients with AML and high risk MDS. Alternative therapeutic strategies with the potential to increase response to AZA are required in this patient population. Disclosures Ferguson: Celgene: Consultancy. Raghavan:Celgene: Consultancy. Quek:Celgene: Research Funding. Cavenagh:Celgene: Consultancy, Speakers Bureau; Janssen: Consultancy, Speakers Bureau; Novartis: Consultancy, Speakers Bureau; Amgen: Consultancy, Speakers Bureau. McMullin:Novartis: Honoraria, Speakers Bureau. Pillai:Celgene: Honoraria. Vyas:Celgene: Honoraria, Research Funding.

Cytogenetics is considered one of the most valuable prognostic determinants in acute myeloid leukemia (AML). However, many studies on which this assertion is based were limited by relatively small sample sizes or varying treatment approach, leading to conflicting data regarding the prognostic implications of specific cytogenetic abnormalities. The Medical Research Council (MRC) AML 10 trial, which included children and adults up to 55 years of age, not only affords the opportunity to determine the independent prognostic significance of pretreatment cytogenetics in the context of large patient groups receiving comparable therapy, but also to address their impact on the outcome of subsequent transplantation procedures performed in first complete remission (CR). On the basis of response to induction treatment, relapse risk, and overall survival, three prognostic groups could be defined by cytogenetic abnormalities detected at presentation in comparison with the outcome of patients with normal karyotype. AML associated with t(8;21), t(15;17) or inv(16) predicted a relatively favorable outcome. Whereas in patients lacking these favorable changes, the presence of a complex karyotype, −5, del(5q), −7, or abnormalities of 3q defined a group with relatively poor prognosis. The remaining group of patients including those with 11q23 abnormalities, +8, +21, +22, del(9q), del(7q) or other miscellaneous structural or numerical defects not encompassed by the favorable or adverse risk groups were found to have an intermediate prognosis. The presence of additional cytogenetic abnormalities did not modify the outcome of patients with favorable cytogenetics. Subgroup analysis demonstrated that the three cytogenetically defined prognostic groups retained their predictive value in the context of secondary as well as de novo AML, within the pediatric age group and furthermore were found to be a key determinant of outcome from autologous or allogeneic bone marrow transplantation (BMT) in first CR. This study highlights the importance of diagnostic cytogenetics as an independent prognostic factor in AML, providing the framework for a stratified treatment approach of this disease, which has been adopted in the current MRC AML 12 trial.

The relative efficacy and toxicity of the chemotherapeutic agents thioguanine (6TG) and etoposide (VP16) were assessed by a randomized comparison of the DAT (daunorubicin, cytarabine, thioguanine) versus ADE (daunorubicin, cytarabine, etoposide) regimens in the Medical Research Council's 10th acute myeloid leukaemia trial (MRC AML 10), which was open to patient entry from May 1988 to April 1995. In this, the largest reported trial of AML therapy to date, 1,857 eligible patients, mostly less than 56 years old, were randomized: 929 (including 143 children under 15 years old) were allocated to DAT and 928 (143 children) to ADE. The two groups were well matched for presentation features. The complete remission (CR) rate was 81% with DAT and 83% with ADE (P = .3). The percentages of remitters achieving remission after 1, 2, or more than 2 courses were 70%, 22%, and 8% for DAT and 74%, 21%, and 5% for ADE. The percentages failing to achieve a CR due to resistant disease were 11% with DAT versus 9% with ADE (P = .07). There was a slightly higher death rate in CR during consolidation chemotherapy with ADE (9%) than with DAT (6%) (P = .06). Patients receiving DAT took slightly but significantly longer to recover from neutropenia and thrombocytopenia but the median number of days in hospital were similar in each group. ADE patients experienced slightly more severe nonhematologic toxicity. There was also no significant difference between the groups in the longer-term measures of efficacy: disease-free survival at 6 years from CR was 42% (±4) for DAT and 43% (±4) for ADE (P = .8); relapse rate at 6 years was 50% (±4) for DAT and 49% (±5) for ADE (P = .6); survival at 6 years was 40% (±4) for both DAT and ADE (P = .9). Subgroup analysis failed to show any benefit for etoposide in patients with monocytic or myelomonocytic disease, or in any other diagnostic subgroup. In conclusion, DAT and ADE both achieve high remission rates and good long-term survival, and are equally effective chemotherapy regimens for the treatment of AML patients aged up to 55 years.

During the immunodiagnosis of 517 cases of acute myelogenous leukemia (AML) entered into the Medical Research Council (MRC) AML10 trials, we have observed the CD34 precursor cell antigen more frequently in AML of M2 morphology, especially in the 84% of cases with the t(8;21) chromosomal translocation, than in any other French-American-British classification group. CD34 expression was then quantified (using QIFI and Quantum Simply Cellular beads [Flow Cytometry Standards, Research Triangle Park, NC] and CD34(+) standard cells). When CD34 antibody-binding capacity (ABC) of normal bone marrow (BM) precursors and leukemic blasts was compared, it was shown that AML M2 cases with t(8;21) not only had the highest percentages of CD34(+) blasts, but in >80% of CD34(+) cases the individual blasts expressed higher than normal levels of CD34 antigen (>60 x 10(3) ABC per cell). In addition, in 73% of this group CD34 antigen was overexpressed in an asynchronous combination with cytoplasmic myeloperoxidase (MPO). Other signs of asynchrony included high CD34 expression with CD15 and/or CD56, as well as aberrant combinations of CD13 with terminal deoxynucleotidyl transferase (TdT) and CD19. These findings demonstrate that asynchrony is identifiable in virtually every case of AML with t(8;21), although it does not always involve the same antigens, M2 cases with t(8;21), mostly CD34(+), had a 100% remission rate and 71% 5-year survival rate; other patients with CD34(+) or CD34(-) AML showed 69% and 84% remission rates and 31% and 36% 5-year survival rates, respectively. Consequently, individual markers such as CD34 should be interpreted in relation to other features such as chromosomal changes. These simple methods, which are well suited to quantify the expression of ligands, are a useful contribution to diagnosis: 60% to 65% of M2 cases with t(8;21) are rapidly identified by CD34 overexpression alone. This aberration, together with the other signs of asynchrony seen at presentation, can be used to search for residual leukemia after therapy. (C) 1996 by The American Society of Hematology.

The clinical value of serial minimal residual disease (MRD) monitoring in core binding factor (CBF) acute myeloid leukemia (AML) by quantitative RT-PCR was prospectively assessed in 278 patients [163 with t(8;21) and 115 with inv(16)] entered in the United Kingdom MRC AML 15 trial. CBF transcripts were normalized to 105ABL copies. At remission, after course 1 induction chemotherapy, a > 3 log reduction in RUNX1-RUNX1T1 transcripts in BM in t(8;21) patients and a > 10 CBFB-MYH11 copy number in peripheral blood (PB) in inv(16) patients were the most useful prognostic variables for relapse risk on multivariate analysis. MRD levels after consolidation (course 3) were also informative. During follow-up, cut-off MRD thresholds in BM and PB associated with a 100% relapse rate were identified: for t(8;21) patients BM > 500 copies, PB > 100 copies; for inv(16) patients, BM > 50 copies and PB > 10 copies. Rising MRD levels on serial monitoring accurately predicted hematologic relapse. During follow-up, PB sampling was equally informative as BM for MRD detection. We conclude that MRD monitoring by quantitative RT-PCR at specific time points in CBF AML allows identification of patients at high risk of relapse and could now be incorporated in clinical trials to evaluate the role of risk directed/preemptive therapy.

Activating mutations of MPL exon 10 have been described in a minority of patients with idiopathic myelofibrosis (IMF) or essential thrombocythemia (ET), but their prevalence and clinical significance are unclear. Here we demonstrate that MPL mutations outside exon 10 are uncommon in platelet cDNA and identify 4 different exon 10 mutations in granulocyte DNA from a retrospective cohort of 200 patients with ET or IMF. Allele-specific polymerase chain reaction was then used to genotype 776 samples from patients with ET entered into the PT-1 studies. MPL mutations were identified in 8.5% of JAK2 V617F− patients and a single V617F+ patient. Patients carrying the W515K allele had a significantly higher allele burden than did those with the W515L allele, suggesting a functional difference between the 2 variants. Compared with V617F+ ET patients, those with MPL mutations displayed lower hemoglobin and higher platelet levels at diagnosis, higher serum erythropoietin levels, endogenous megakaryocytic but not erythroid colony growth, and reduced bone marrow erythroid and overall cellularity. Compared with V617F− patients, those with MPL mutations were older with reduced bone marrow cellularity but could not be identified as a discrete clinicopathologic subgroup. MPL mutations lacked prognostic significance with respect to thrombosis, major hemorrhage, myelofibrotic transformation or survival.

The role of histopathology in the diagnosis of essential thrombocythemia (ET) is controversial, and there has been little attempt to quantitate interobserver variability. Diagnostic bone marrow trephine biopsy specimens from 370 patients with ET by Polycythemia Vera Study Group (PVSG) criteria were assessed by 3 experienced hematopathologists for 16 different morphologic features and overall diagnosis according to the World Health Organization (WHO) classification. Our results show substantial interobserver variability, particularly for overall diagnosis and individual cellular characteristics such as megakaryocyte morphology. Reticulin grade was the dominant independent predictor of WHO diagnostic category for all 3 hematopathologists. Factor analysis identified 3 independent factors likely to reflect underlying biologic processes. One factor related to overall and lineage-specific cellularity and was significantly associated with JAK2 V617F status (P < .001), a second factor related to megakaryocyte clustering, and a third was associated with the fibrotic process. No differences could be discerned between patients labeled as having “prefibrotic myelofibrosis” or “true ET” in clinical and laboratory features at presentation, JAK2 status, survival, thrombosis, major hemorrhage, or myelofibrotic transformation. These results show that histologic criteria described in the WHO classification are difficult to apply reproducibly and question the validity of distinguishing true ET from prefibrotic myelofibrosis on the basis of subjective morphologic criteria. This study was registered at http://isrctn.org as #72251782 and at http://eudract.emea.europa.eu/ as #2004-000245-38.

The optimum chemotherapy schedule for reinduction of patients with high-risk acute myeloid leukemia (relapsed, resistant/refractory, or adverse genetic disease) is uncertain. The MRC AML (Medical Research Council Acute Myeloid Leukemia) Working Group designed a trial comparing fludarabine and high-dose cytosine (FLA) with standard chemotherapy comprising cytosine arabinoside, daunorubicin, and etoposide (ADE). Patients were also randomly assigned to receive filgrastim (G-CSF) from day 0 until neutrophil count was greater than 0.5 × 109/L (or for a maximum of 28 days) and all-trans retinoic acid (ATRA) for 90 days. Between 1998 and 2003, 405 patients were entered: 250 were randomly assigned between FLA and ADE; 356 to G-CSF versus no G-CSF; 362 to ATRA versus no ATRA. The complete remission rate was 61% with 4-year disease-free survival of 29%. There were no significant differences in the CR rate, deaths in CR, relapse rate, or DFS between ADE and FLA, although survival at 4 years was worse with FLA (16% versus 27%, P = .05). Neither the addition of ATRA nor G-CSF demonstrated any differences in the CR rate, relapse rate, DFS, or overall survival between the groups. In conclusion these findings indicate that FLA may be inferior to standard chemotherapy in high-risk AML and that the outcome is not improved with the addition of either G-CSF or ATRA.

The prognostic significance of FLT3 mutations in acute promyelocytic leukemia (APL) is not firmly established and is of particular interest given the opportunities for targeted therapies using FLT3 inhibitors. We studied 203 patients with PML-RARA–positive APL; 43% of the patients had an FLT3 mutation (65 internal tandem duplications [ITDs], 19 D835/I836, 4 ITD+D835/I836). Both mutations were associated with higher white blood cell (WBC) count at presentation; 75% of the patients with WBC counts of 10 × 109/L or greater had mutant FLT3. FLT3/ITDs were correlated with M3v subtype (P < .001), bcr3 PML breakpoint (P < .001), and expression of reciprocal RARA-PML transcripts (P = .01). Microarray analysis revealed differences in expression profiles among patients with FLT3/ITD, D835/I836, and wild-type FLT3. Patients with mutant FLT3 had a higher rate of induction death (19% vs 9%; P = .04, but no significant difference in relapse risk (28% vs 23%; P = .5) or overall survival (59% vs 67%; P = .2) at 5 years. In in vitro differentiation assays using primary APL blasts (n = 6), the FLT3 inhibitor CEP-701 had a greater effect on cell survival/proliferation in FLT3/ITD+ cells, but this inhibition was reduced in the presence of ATRA. Furthermore, in the presence of CEP-701, ATRA-induced differentiation was reduced in FLT3/ITD+ cells. These data carry implications for the use of FLT3 inhibitors as frontline therapy for APL.

Fetal liver tyrosine kinase 3 (FLT3) internal tandem duplications (ITDs) are powerful adverse prognostic indicators for relapse in acute myeloid leukemia (AML) but the most efficacious therapy for FLT3/ITD+ patients is currently unknown. We evaluated outcome according to FLT3/ITD status in 1135 adult patients treated according to United Kingdom Medical Research Council (UK MRC) AML protocols: 141 received an autograft, and 170 received a matched sibling allograft in first complete remission (CR). An FLT3/ITD was detected in 25% of patients and was an independent predictor for relapse (P < .001). It remained prognostic for increased relapse in patients who received a transplant (odds ratio [OR] = 1.91; 95% confidence intervals [CIs] = 1.13-3.21; P = .02), with no evidence of a difference in effect between patients who received an autograft (OR = 2.39; CIs = 1.24-4.62) and patients who received an allograft (OR = 1.31; CIs = 0.56-3.06) (test for interaction, P = .3) or between patients who did or did not receive a transplant (P = .4). These results were confirmed in an analysis of 186 patients randomized to receive or not receive an autograft in first CR and in a donor-versus–no donor analysis of 683 patients to assess the role of allograft (for latter, FLT3/ITD- OR = 0.70, CIs = 0.53-0.92; FLT3/ITD+ OR = 0.59, CIs = 0.40-0.87; test for interaction, P = .5). These results suggest that at present there is no strong evidence that FLT3 status should influence the decision to proceed to transplantation.

Essential thrombocythemia (ET) is associated with arterial and venous thrombosis, hemorrhage and transformation to acute leukemia and myelofibrosis. Cytoreduction with hydroxyurea (Hu) reduces thrombotic risk, but concerns about potential toxicity have led to the widespread use of a newer agent, anagrelide (Ag) as first-line therapy. However, no randomized trials comparing the efficacy and safety of Ag with Hu have been performed. The MRC PT1 trial is an international, multi-centre, randomized controlled trial comparing Ag with Hu in patients with ET at high-risk of vascular events (based on age, platelet count or cardiovascular risk factors). The trial was independently conceived, conducted and analysed by the investigators on behalf of the UK Myeloproliferative Disorders Study Group and the MRC Adult Leukaemia Working Party. Patients (newly diagnosed or previously treated) were randomized to receive either Ag plus aspirin (Ag+asp) or Hu+asp. Clinical end-points together with blood and bone marrow morphology were assessed centrally by experts blinded to treatment allocation. With 809 patients randomized and median follow-up of 39 months, it is the largest and most comprehensive randomized study of ET performed to date. Patients in the Ag+asp arm were significantly more likely to reach the primary end-point and several secondary end-points. Intention-to-treat log-rank analyses of time to event indicate that Ag+asp patients were significantly more likely than Hu+asp patients to reach the composite primary end-point of arterial thrombosis, venous thrombosis or major hemorrhage (55 v 36 events; odds ratio 1.57; 95% CI 1.04–2.37; p=0.03). Ag+asp was also associated with increased rates of arterial thrombosis (37 v 17 events; OR 2.16; 95% CI 1.27–3.69; p=0.004) and major hemorrhage (22 v 8 events; OR 2.61; CI 1.27–5.33; p=0.008) but, interestingly, decreased venous thrombosis compared to Hu+asp (3 v 14 events; OR 0.27; CI 0.11–0.71; p=0.006). Myelofibrotic transformation, the diagnosis of which required new clinical and/or laboratory features in addition to grade III/IV reticulin fibrosis, was significantly increased in the Ag+asp arm (16 v 5 events; OR 2.92, CI 1.24–6.86, p=0.01). Transformation to MDS/AML was comparable between the two arms (4 Ag+asp v 6 Hu+asp), although the small number of transformations and short follow-up prevent firm conclusions about leukemogenicity. Overall survival was not statistically different. Ag+asp was more poorly tolerated than Hu+asp. In the Ag+asp arm significantly fewer patients remained on their allocated treatment (p These results demonstrate that, compared to Hu+asp, Ag+asp is associated with an excess rate of arterial thrombosis, major hemorrhage and myelofibrotic transformation but decreased venous thrombosis, and suggest that Hu should remain first-line therapy in patients with ET at high risk for vascular events.

The xeroderma pigmentosum group D (XPD) gene encodes a DNA helicase that functions in nucleotide excision repair of chemotherapy-induced DNA damage, the efficiency of which is predicted to be affected by a lysine to glutamine variant at codon 751. We hypothesized that this constitutive genetic variant may modify clinical response to chemotherapy, and we have examined its association with outcome following chemotherapy for acute myeloid leukemia (AML) in 341 elderly patients entered into the United Kingdom Medical Research Council AML 11 trial, and with the risk of developing chemotherapy-related AML. Among subjects treated for AML, disease-free survival at one year was 44% for lysine homozygotes, compared with 36% for heterozygotes and 16% for glutamine homozygotes (hazard ratio [HR], 1.30; 95% confidence interval [CI], 1.01-1.70; P � .04). Similarly, overall survival at one year was 38% for lysine homozygotes, 35% for heterozygotes, and 23% for glutamine homozygotes (HR, 1.18; 95% CI, 0.99-1.41; P � .07). Furthermore, homozygosity for the XPD codon 751 glutamine variant was associated with a significantly increased risk of developing AML after chemotherapy (odds ratio, 2.22 for Gln/Gln vs Lys/Lys; 95% CI, 1.04-4.74). These data suggest that the XPD codon 751 glutamine variant protects against myeloid cell death after chemotherapy

In acute myeloid leukemia (AML), further prognostic determinants are required in addition to cytogenetics to predict patients at increased risk of relapse. Recent studies have indicated that an internal tandem duplication (ITD) in the FLT3 gene may adversely affect clinical outcome. This study evaluated the impact of a FLT3/ITD mutation on outcome in 854 patients, mostly 60 years of age or younger, treated in the United Kingdom Medical Research Council (MRC) AML trials. An FLT3/ITD mutation was present in 27% of the patients and was associated with leukocytosis and a high percentage of bone marrow blast cells (P

Between May 1988 and June 2000, 698 children were treated in the Medical Research Council acute myeloid leukemia 10 and 12 trials. The presenting features and outcomes of therapy in these children were compared by age. Although there was no single cutoff in age, younger children were more likely to have intermediate risk and less likely to have favorable cytogenetics (P

Abstract 236 Background: There is little consensus on which maintenance therapy clinicians should choose for their patients. Since 1999, the three novel agents of bortezomib, lenalidomide, and thalidomide have been approved for use among patients with MM. These agents have been increasingly used as maintenance therapy. To date, only two randomized controlled trials of maintenance therapy have examined the efficacy of these novel agents head-to-head. Here, we conduct a network meta-analysis of bortezomib, lenalidomide, and thalidomide to determine which of these novel agents could potentially increase overall survival (OS) and progression-free survival (PFS). Methods: A comprehensive literature search of MEDLINE (PubMed), the Cochrane Central Register of Controlled Trials (CENTRAL), and meetings abstracts from American Society of Hematology, American Society of Clinical Oncology, European Society for Medical Oncology and European Hematology Association was undertaken to identify all phase III randomized controlled trials (RCTs) of maintenance therapy published until July 2012. We applied the Bayesian mixed treatment comparison (MTC) method under the random-effects model. The indirect comparisons were constructed from trials that have one treatment in common. For each included RCT, we calculated the hazard ratio (HR) and its corresponding standard error and used this to calculate the indirect estimates of HR and corresponding credible intervals (CrI). We also ranked the treatments according to the probability of best treatment and calculated the surface underneath the cumulative ranking curve (SUCRA). All analyses were conducted in WinBUGS 1.4.3 and Stata 11.2. Results: The network, number of trials for each comparison, and number of patients enrolled is shown in Figure 1. The network for OS was based on 12 RCTs enrolling 5542 patients and the network for PFS was constructed from 13 RCTs and 5784 patients. The MTC networks were consistent for both OS and PFS. For both OS and PFS, two comparisons were produced (Figure 2). For OS, the analysis showed that none of the treatments were superior. For PFS, lenalidomide was superior to thalidomide (HR = 0.58, 95% CrI [0.37, 0.94]). The estimates of SUCRA and rank probabilities (Figure 3) suggested that for OS bortezomib was best followed by lenalidomide and thalidomide. For PFS, lenalidomide was best followed by bortezomib and thalidomide. Conclusion: Using the MTC method, we found no evidence that any of the novel agents are superior to one another in terms of OS. Lenalidomide was the only novel agent which was superior to another active therapy (thalidomide). While these results provide preliminary evidence to which novel agent may be more beneficial as maintenance therapy, definitive conclusions cannot be reached until large, well designed RCTs evaluating these therapies head-to-head are conducted. Disclosures: No relevant conflicts of interest to declare.

Abstract 4271 Background: The role of various maintenance therapies in the management of MM is unclear and evidence on the efficacy of these regimens is conflicting. In order to provide the totality of available randomized evidence on the role of maintenance therapy in MM, we conduct a comprehensive systematic review and meta-analysis of all RCTs studying maintenance therapy. Here, we report the pooled results of trials which directly examined the novel agents of bortezomib, lenalidomide, or thalidomide and reported the outcomes of overall survival (OS) and/or progression-free survival (PFS). Methods: A comprehensive literature search of MEDLINE (PubMed), the Cochrane Central Register of Controlled Trials (CENTRAL), and meetings abstracts from American Society of Hematology, American Society of Clinical Oncology, European Society for Medical Oncology and European Hematology Association was undertaken to identify all phase III randomized controlled trials (RCTs) of maintenance therapy published until July 2012. We extracted data on OS and PFS. Time to event data were pooled under the random effects model as hazard ratios (HR) and its corresponding 95% confidence interval (CI). Heterogeneity was assessed using the chi square test and I2statistic. All analyses were done in Review Manager 5.1. Results: Twenty-two RCTs met the inclusion criteria. (Figure 1) However, only data from the following RCTs were able to be pooled for the direct head-to-head comparison: 2 RCTs of bortezomib maintenance therapy enrolling 792 patients, 5 RCTs of lenalidomide maintenance therapy enrolling 1776 patients, 11 RCTs of thalidomide maintenance therapy enrolling 3952 patients. The pooled HR and 95% CI, number of RCTs, and number of patients for each comparison are presented in Figure 2. Only two trials compared the novel agents of bortezomib and thalidomide head-to-head. There was no significant different in terms of PFS. For the novel agent of lenalidomide, there was no significant difference is OS compared to placebo. The pooled PFS was in favor of lenalidomide maintenance compared to placebo. For thalidomide, OS was significantly in favor of the intervention when compared to placebo or prednisone/dexamethasone. There was no significant difference in OS between thalidomide maintenance when compared to interferon control. For the outcome of PFS, the pooled results favored thalidomide when compared to prednisone/dexamethasone or interferon control. There was no significant difference between thalidomide and placebo. Conclusion: To date, the largest number of trials has been among thalidomide as maintenance therapy. In our meta-analysis, thalidomide is the only agent which improves survival compared to no treatment. Other novel agents have been evaluated in a smaller number of trials and current data does not allow for firm conclusions that any agent is superior to the other. An indirect, network meta-analysis is called for to provide additional insights regarding comparative efficacy of the novel agents as the maintenance treatment for MM. Disclosures: No relevant conflicts of interest to declare.

Abstract 2975 Background: Meta-analyses (MAs) with few participants (i.e. small number of primary studies) are at risk of producing random errors and consequently overestimating treatment effects. With insufficient information the risk of obtaining a false positive result (type I error) increases, which may lead to false conclusions. Trial sequential analysis (TSA) has been proposed as a method to ascertain whether results of MAs are conclusive (true vs. false positive, true vs. false negative). It adjusts for the risk of random error by constructing monitoring boundaries under the sample size necessary to conclude significant treatment effect. In TSA, the information size is calculated based on a pre-specified event rate in the control group, a minimum intervention effect (risk ratio reduction), and a desired maximum risk of type I error α and type II error β. Here we apply TSA on MAs of randomized controlled trials of maintenance therapies in the management of multiple myeloma. Methods: A comprehensive literature search of MEDLINE (PubMed), the Cochrane Central Register of Controlled Trials (CENTRAL), and meetings abstracts from American Society of Hematology, American Society of Clinical Oncology, European Society for Medical Oncology and European Hematology Association was undertaken to identify all phase III randomized controlled trials (RCTs) of maintenance therapy published until July 2012. We extracted data on overall survival and progression-free survival comparing treatments that could be pooled in random effects meta-analysis. We performed TSA for the apriori diversity-adjusted information size (APDIS) under risk ratio reduction of 20% and 25%. The information size was adjusted for between-study trial diversity, which is defined as the total relative variance expansion changing from a fixed effect into a random effects meta-analysis. We used two-sided α = 5% and 1 – β = 80% power. All analyses were done in Stata 11.2 using metacumbounds command. Results: Nine separate meta-analyses (18 randomized controlled trials) met the inclusion criteria (Table 1). The median number of patients was 1193 (range 351–2824) and median diversity 0% (range 0%-91%). Under both risk ratio reductions of 20% and 25%, 4/9 MAs were false negative and 1/9 false positive. The observed power based on the accrued sample size and observed risk ratio reduction was greater than 80% in 5/9 MAs. Conclusion: TSA detected one false positive MA of two trials comparing thalidomide with prednisone/dexamethosone for the outcome of overall survival. Future MAs need to consistently undertake TSA to avoid misleading conclusions. Disclosures: No relevant conflicts of interest to declare.

Abstract 582 We previously showed in the MRC AML15 Trial (Burnett et al JCO 2011:29(4):369-77 that 70% of younger patients with AML derived a 10% survival benefit by the addition of the immuno-conjugate, Gemtuzumab Ozogamicin (GO) (Mylotarg™), to induction chemotherapy. In a second trial (AML16 non-intensive) its addition to Low dose Ara-C doubled the CR rate but did not improve OS (Burnett et al. Blood 2010:116 (21): Abstract 18). We now report the result of the NCRI AML16 (Intensive) Trial in which older patients were randomised to receive, or not, GO 3mg/m2 on day 1 of course 1 of induction chemotherapy. Patients were randomised to two courses of DA (daunorubicin/ara-C) or DClo (daunorubicin/clofarabine), followed, or not, by a 3rd course (DA) with or without Azacytidine maintenance. Between December 2006 and July 2010, 1115 patients were randomised to the GO vs no GO comparison from 149 centres in the UK & Denmark. The median age was 67 years (range 51–84). 806 had de novo AML, 194 secondary disease and 115 high risk MDS (>10% marrow blasts). Of the 806 patients with cytogenetic data, 33 were favourable/629 intermediate/204 adverse risk (Grimwade et al, Blood 1998: 92:2322-33). 96% of those allocated GO received it. The overall response rate was 69% (CR 60%; CRi 9%) with 52% of patients achieving response after course 1. Overall survival at 4 years was 17% with a median follow up of 29.5 months (range 0.5–54.6 months).CR%CRi%ORR%Res Dis %Ind Death %30d mortality %60d mortality %GO629711712915No GO5810682111814p-value0.180.30.070.40.80.8 Induction Results: There was no significant difference in blood count recovery kinetics; other toxicities and resource usage were not significantly increased with the exception that for course 1 nausea and oral toxicity were marginally higher with GO, more platelet transfusions were given (13.7 vs 9.6 mean units; p Relapse/RFS/OS: The rate of relapse was significantly reduced (GO vs no GO: CIR at 2 years 61% vs 70%; p=0.004), leading to significantly better RFS (28% vs 23% at 2 years; p=0.03), and survival from CR and OS were significantly better on the GO arm (2 year survival from CR 47% vs 39%; p=0.02; 2 year OS 35% vs 29%; p=0.04). While the benefit appeared lower in patients with secondary disease or with adverse cytogenetics (GO vs no GO 2 year OS 19% vs 21% and 13% vs 6%) compared to those with de Novo disease or intermediate cytogenetics (2 year OS 38% vs 32% and 41% vs 36%), as was seen in our previous AML15 trial, in this trial there was no significant evidence of interaction between treatment and any demographics or underlying chemotherapy. Conclusion: The addition of GO to induction chemotherapy did not lead to unacceptable increases in toxicity. It improved disease control and delivered a significant OS benefit for older patients overall. While there is no significant interaction, the greater benefit in patients with de novo disease or intermediate risk cytogenetics is consistent over the 2228 patients randomised in the AML15 and AML16 trials. Disclosures: Burnett: Pfizer: Consultancy. Off Label Use: Gemtuzumab Ozogamicin in AML.

Abstract 18 In a significant proportion of older patients with AML intensive chemotherapy is not considered a viable option[1]. Such patients may receive low-dose Ara-C (LDAC), best supportive care (BSC) with hydroxyurea or an experimental agent but outcomes are poor. We have shown that LDAC is superior to BSC [2], but only in patients who enter CR (fewer than 20%).[2]. There is therefore scope to improve outcomes in these patients, and a number of possible treatments have been evaluated in the randomised UK NCRI AML16 trial, one of which is gemtuzumab ozogamicin (GO), which we have shown to benefit the majority of younger patients when given in conjunction with standard chemotherapy[3]. In AML16 novel agents or combinations are tested in untreated older patients with AML or high risk MDS (marrow blasts >10%) using a “pick a winner” design. The design allows unpromising treatments to be identified early (typically after 50 or 100 patients per arm): only those arms which show promise will continue to a trial with OS and DFS as endpoints. The aim is to at least double the remission rate from 15% to 30%, and thus improve overall survival. We now report the results of LD Ara-C (20mg bd days 1–10 for 4 courses) versus LDAC combined with GO (at a fixed dose of 5mg on day 1 of each course for 4 courses at 6–8 week intervals). Patient Details: The comparison opened as part of the UK LRF AML14 trial and was carried forward to AML16 unchanged. Between June 2004 and June 2010, 495 patients were randomised, 249 to LDAC plus ATO, 246 to LDAC. The median age was 75 years (range 54–90); 83% of patients were aged over 70 years, and 61% were male. To be eligible patients’ LFTs had to be less than twice ULN. Treatment arms were balanced for age, gender, performance status, de novo/secondary AML/high risk MDS, presenting WBC and cytogenetic risk group. Follow-up is complete to 1st January 2010, with median follow up of 21 months. Survival and remission data is available on 412, 404 patients respectively. Treatment Results: The trial passed through both stopping hurdles based on CR/CRi rates and therefore continued to full accrual, with overall survival as primary outcome measure. The table shows the distribution of outcomes: there was no heterogeneity by recruitment period (AML14 vs AML16). The causes of death (316) were:- There were no significant interactions between treatment and any of the baseline variables on either remission or survival outcomes. Likewise there were no major toxicity implications, although resource usage tended to be higher in patients given GO. Discussion: While the addition of GO significantly improves CR rate, achieving the 30% response originally sought, this does not translate into survival. This is predominantly due to an increase in relapse, indicating that if GO is to have a role in this setting, it will require effective treatment to maintain remission. [1] Juliusson G et al. Blood 2009; 113: 4179 – 4187 [2] Burnett et al. Cancer 2007 109: 1114–1124 [3] Burnett et al. JCO 2010 to appear. Disclosures: Off Label Use: Mylotarg (gemtuzumab ozogamicin).

Abstract 3028 Background: Bisphosphonates are used as supportive therapy for patients with multiple myeloma (MM) to prevent bone destruction. There are multiple types of bisphosphonates. Currently, which bisphosphonate is superior to others in the treatment of patients with MM is not known. There are only three randomized controlled trials (RCTs) which employed head to head comparison of bisphsophonates in patients with MM. We performed indirect meta analysis of various bisphohsphonates and did not find the superiority of any particular type of bisphosphonate over others (Mhaskar et al Cochrane Database Syst Rev. 2010 Mar 17;3). In June 2010, Morgan et al published largest RCT to date involving 1960 MM patients employing head to head comparison of zoledronate and clodronate (Morgan et al J Clin Oncol, 2010. 28(15_suppl): p. 8021). Hence, the aim of this analysis is to assess if any specific type of bisphosphonate is superior to others incorporating data from this RCT. Methods: We extracted data regarding overall survival (OS), progression free survival (PFS), skeletal related events (SREs), and grade III-IV treatment related harms. Superiority of bisphosphonates was assessed by the mixed treatment comparison (MTC) as described by Lu and Ades (Lu et al. Stat Med, 2004. 23(20): p.3105-24). The fixed and random effects MTC models were compared using the deviance information criterion. Network consistency was checked using the back-calculation method. Since there was no evidence suggesting the superiority of random effects model we have reported fixed effects model estimates. Results: Eighteen RCTs were included enrolling 4,970 patients. For the outcome of OS the pooled analysis demonstrated a beneficial effect of zoledronate in comparison with clodronate [HR = 0.83 (95% CI: 0.73 to 0.94)] and etidronate [HR = 0.48 (95% CI: 0.31 to 0.71)]. However, there was no evidence that zoledronate was superior to pamidronate [HR = 0.84 (95% CI: 0.61 to 1.13)], or ibandronate [HR = 0.70 (95% CI: 0.42 to 1.11)] for the outcome of OS. Zoledronate was superior to clodronate [HR = 0.88 (95% CI: 0.78 to 0.99)] for the outcome PFS. Zoledronate was also superior to clodronate [HR = 0.75 (95% CI: 0.64 to 0.88)], pamidronate [HR = 0.65 (95% CI: 0.42 to 0.95)], and ibandronate [HR = 0.44 (95% CI: 0.26 to 0.72)] for the outcome of SREs. There were no significant adverse effects associated with the administration of bisphosphonates. Only three RCTs reported the incidence of osteonecrosis of jaw (ONJ). The incidence of ONJ was higher in zoledronate [OR: 12.03 (95% CI: 3.68 to 39.26)] compared to clodronate. We also identified 7 observational studies evaluating 1068 patients for ONJ. These studies suggest that ONJ may be a common event (range: 0%− 51%). Conclusion: The results demonstrate a beneficial effect of zoledronate on OS compared to clodronate and etidronate. Similarly, zoledronate was superior to clodronate in improving PFS. zoledronate is superior to clodronate, pamidronate and ibandronate in prevention of SREs in patients with MM. In the context of MTC uncertainty analysis, zoledronate ranked as the best treatment followed by clodronate and pamidronate. These finding underscore the need for RCT with head to head comparison of zoledronate and pamidronate for their efficacy and safety in MM patients. Disclosures: No relevant conflicts of interest to declare.

Abstract 523 Introduction: The advent of reduced intensity conditioning (RIC) regimens permits the delivery of allogeneic stem cell transplant (alloSCT) with a potentially curative graft-versus-leukaemia effect in patients with acute myeloid leukaemia (AML) whose outcome with conventional chemotherapy or conventional myeloablative SCT would be poor. Although widely utilised in the management of older patients with AML in 1st complete remission (CR) there are no randomised trials or “donor v. no donor” analyses supporting a role for RIC alloSCT in this setting. Methods: We have examined the impact of RIC allograft in 1st CR on the outcome of patients over the age of 45 treated within the UK MRC AML15 trial. Patients with favourable cytogenetics (t8;21, inv16, t15;17) were not eligible for transplant and were thus excluded from the analysis, which was restricted to patients with intermediate and adverse cytogenetics (defined as: -5, -7, 5q-, abn3q, or 5 or more abnormalities). Patients who were not considered suitable for transplant at the tissue typing stage were excluded. The data was initially analysed by treatment actually received, using the method of Mantel-Byar to allow for time to transplant. The analysis was stratified by presence of a sibling donor, and by cytogenetic risk group. End points were death in 1st CR (DCR), relapse risk (RR), relapse free survival (RFS) and overall survival (OS). A hazard ratio (HR) less than 1 indicates benefit for alloSCT/donor. Results: A total of 700 patients were eligible, of whom 185 received RIC alloSCT in 1st CR (120 sibling, 65 MUD), 109 patients had adverse cytogenetics, and 247 patients had a matched sibling donor. Median follow up from 1st CR was 2.5 years. Mantel-Byar analysis showed increased DCR for RIC alloSCT (HR 6.73; CI 3.49-12.97; p Discussion: In the absence of randomised trials, evaluation of the benefit of SCT is difficult. Mantel-Byar analyses may well be subject to selection biases; sibling donor versus no sibling donor analyses are difficult to interpret in this context, since non-receipt of SCT in the donor group and use of MUD SCT (in the no donor group) both tend to reduce any effect. Hence, these results need to be interpreted cautiously. While our data are consistent with the possibility that RIC alloSCT in 1st CR may be effective in some patients aged 45+ (those with both a sibling donor and intermediate cytogenetics being the most likely candidates based on the current evidence), further data are clearly required, ideally from large scale prospective randomised controlled trials. Disclosures: No relevant conflicts of interest to declare.

486 A significant proportion of older patients with AML are not treated with conventional intensive chemotherapy [1] because they are unfit or not considered likely to benefit from an intensive treatment approach. Their outcomes are poor. Such patients have typically been treated with low-dose Ara-C (LDAC) or best supportive care (BSC) with hydroxyurea, and unrandomised studies of new agents have been used in this population. A recent randomised trial has shown that LDAC is superior to BSC in these patients [2]. Randomised trials are underway to assess the value of other novel treatments compared to LDAC. In an unrandomised phase 2 trial in 64 patients, Roboz et al found encouraging results using the combination of Arsenic Trioxide and LDAC. [3]. The UK NCRI AML16 trial is a programme of development which aims to test novel agents or combinations in untreated older patients with AML or high risk MDS (marrow blasts >10%) following a “pick a winner” design. The intention is to identify a “winner” which will produce a remission rate in excess of 30%, compared with 15-20% with LDAC. Using LDAC as the standard arm,the design allows unpromising treatments to be identified early (typically after 50 patients per arm), so that only those arms which show promise will continue to a trial with OS and DFS as endpoints. We report our experience of LD Ara-C (20mg bd days 1-10 for 4 courses) versus LDAC combined with Arsenic Trioxide (ATO, 0.25 mg/kg d1-5, d9, d11 for 4 courses at 6-8 week intervals). Patient Details: Between December 2006 and until its conclusion in May 2009, 166 patients were randomised, 84 to LDAC plus ATO, 82 to LDAC. The median age was 74 years; 80% of patients were aged over 70 years, 62% were male. There were no differences between the treatment arms with respect to age distribution, gender, performance status, de novo/secondary AML, high risk MDS, presenting WBC or cytogenetic risk group. Follow-up is complete to 1st January 2009, with median follow up for survivors of 8 months (range 0.1-17), at which point 122 patients had been recruited, and there were a total of 60 deaths (LDAC n=25; LDAC+ ATO, n=35). CR status is known on 113 patients. Overall, 24 patients have entered CR/CRi (LDAC n=13, LDAC+ATO n=11) with 8 relapses (2 vs 6; 1 vs 3 patients have died following relapse). | | CR | CRi | 30-day mortality | 8 week mortality | 6 month survival from CR | 12 month OS | |:--------------------- | -------------------- | ------------------------ | ---------------- | ---------------- | ------------------------ | ---------------- | | All patients | 15/113 (13%) | 9 (8%) | 13% | 22% | 91% | 33% | | LD Ara-C (n=60) | 7/54 (13%) | 6 (11%) | 12% | 22% | 83% | 41% | | LD Ara-C + ATO (n=62) | 8/59 (14%) | 3 (5%) | 13% | 22% | 100% | 27% | | OR/HR & 95% CI | CR: 0.95 (0.32-2.81) | CR/CRi: 1.28 (0.56-3.39) | | | 2.31 (0.31-17.0) | 1.46 (0.88-2.44) | | P-value | 0.9 | 0.6 | | | 0.4 | 0.03 | Treatment Results: The causes of death (60) were:- | | Infection | Haemorrhage/CVA | Resistant disease | Cardiac | Relapse | Other | |:--------------------- | --------- | --------------- | ----------------- | ------- | ------- | ----- | | LD Ara-C (n=25) | 11 | | 12 | | | 2 | | LD Ara-C + ATO (n=35) | 13 | 1 | 15 | 1 | 2 | 3 | The DMEC recommended closure of the LDAC + ATO arm of the trial because follow-up data on the first 50 patients per arm showed that ATO had failed to provide the 2.5% improvement in CR/CRi required for continued recruitment and that the required improvement in remission was unlikely with LDAC + ATO. Conclusions: While ATO has a definite role in treating patients with APL, and may be of benefit in combination with other drugs in AML, the combination of LDAC + ATO in this patient population was not beneficial. [1] Juliusson G et al. Blood 2009; 113: 4179—4187 [2] Burnett et al. Cancer 2007 109: 1114—1124 [3] Roboz Gail J et al. Cancer 2008;113(9):2504—11. Disclosures: Off Label Use: Arsenic Trioxide is not licensed in this indication.

Abstract 1419 Poster Board I-442 For more than 20 years high dose chemotherapy followed by allogeneic stem cell transplantation (SCT) has been considered as a reasonable approach for the treatment of patients with AML. Moreover, during the last decade new scientific and technical developments results in major changes of clinical practice of transplantation. Enhanced donor availabilities and new strategies, e.g. dose-reduced conditioning, now make allogeneic stem cell transplantation available to patients who do not have a related donor or would not tolerate high-dose chemotherapy due to age or comorbidities. Usually, the decision to start the work-up process for allogeneic transplantation in AML patients is based on the availability of a donor, the assignment to the cytogenetic risk group, and the response to induction therapy, as well as patient factors. However, there would be greater confidence in defining who should, or should not, receive an allograft if the available recommendations given in guidelines are consistent and similar. In this analysis, a comprehensive systematic literature search for best available evidence from controlled clinical trials was performed in the bibliographic databases MEDLINE, EMBASE and Cochrane Central. In addition, the websites of major organizations in Europe and the US (European Group for Blood and Marrow Transplantation, EBMT; European Society for Medical Oncology, ESMO; British Committee for Standards in Hematology, BCSH; American Society for Blood and Marrow Transplantation, ASBMT; National Comprehensive Cancer Network, NCCN) were screened and the specific databases of the National Guideline Clearinghouse and the Guideline International Network Database were also searched to identify the latest recommendations and guidelines. The following points were selected for systematic comparison of the best available evidence: Factors for risk assessment and categorization of AML, donor categories for allogeneic SCT (sibling donors / matched unrelated donors), allogeneic transplantation in first CR, allogeneic transplantation in relapse/progressive disease or second CR, and allogeneic transplants with reduced intensity conditioning regimen. Several interesting findings emerge from this analysis: 1) For patients with relapse or refractory disease donor availability should be explored and discussed, though this is not based on reliable evidence from randomized studies; 2) Patients in CR1 with intermediate or high risk disease who have a matched related donor available should receive allogeneic stem cell transplantation (intermediate risk; ASBMT: reasonable, NCCN: option); 3) For patients who lack a family donor the recommendations are not consistent; 4) Allogeneic transplantation with reduced conditioning in AML patients is feasible, but the superiority over standard therapeutic regimens has not been proven yet. In summary, current guidelines differ in critical points in the recommendation for allogeneic stem cell transplantation. Furthermore, it is likely that only well-defined subgroups of AML patients will benefit from stem cell transplantation. Disclosures: No relevant conflicts of interest to declare.

Abstract 3858 Poster Board III-794 Background Bisphosphonates are currently used as supportive therapy for multiple myeloma (MM). In 2001 we conducted a Cochrane systematic review (SR) showing that Bisphosphonates reduce vertebral fractures and pain but have no effect on other important outcomes. Here we report an update of that SR. Methods A comprehensive literature search of MEDLINE, EMBASE, LILACS, Cochrane database of randomized controlled trials (RCTs) and www.clinicaltrials.gov and meetings abstracts from American Society of Clinical Oncology, American Society of Hematology and European Hematology Association was undertaken to identify all phase III RCTs published until January 2009. We extracted data regarding overall survival (OS), progression free survival (PFS), vertebral and non vertebral fractures, skeletal related events (SREs), pain, hypercalcemia, grade III-IV treatment related harms. The time to event data and dichotomous data were pooled under the random effects model as hazard ratios (HR) and risk ratios (RR) respectively. Heterogeneity was assessed using the chi square test and I2 statistic. Indirect comparison of various bisphosphonates was conducted according to the methods developed by Bucher and Glenny et al and were extended to calculate HR/RR. Results Seventeen RCTs were included enrolling 3,010 patients. In comparison with placebo / no treatment, the pooled analysis demonstrated a beneficial effect of bisphosphonates on prevention of pathological vertebral fractures (7 RCTs, 1116 patients) [RR= 0.74 (95% CI: 0.62 to 0.89), P = 0.001], SREs (6 RCTs, 1334 patients) [RR= 0.81 (95% CI: 0.72 to 0.92), P = 0.001] and on amelioration of pain (8 RCTs, 1281 patients) [RR = 0.75 (95% CI: 0.60 to 0.95), P = 0.01]. We found no significant effect of bisphosphonates on OS, PFS, hypercalcemia or on the reduction of non-vertebral fractures. There was statistically significant heterogeneity for OS and pain endpoints. The heterogeneity for the outcome of pain could be explained by the variation in the pain scales used to measure pain. However, we also found that the beneficial effect of bisphosphonates on pain reduction was greater in patients who were asymptomatic at the start of treatment [RR= 0.28 (95% CI: 0.12 to 0.67)] compared to symptomatic patients [RR= 0.83 (95% CI: 0.69 to 1.00)] (Test of interaction: p = 0.005). The heterogeneity for OS was attributed to one RCT with unrealistic treatment effects (“an outlier effect”). Results of indirect meta analysis were consistent with the results from direct comparisons for the outcomes of vertebral fractures, SREs and pain. The indirect meta analyses did not find the superiority of any particular type of bisphosphonate over others. There were no significant adverse effects associated with the administration of bisphosphonates. In fact, only two RCTs reported osteonecrosis of jaw (ONJ). We also identified 7 observational trials evaluating 1068 patients for ONJ. These studies suggest that ONJ may be a common event (range: 0%- 51%). Since ONJ was only sporadically reported in RCTs the results from observational studies may be an overestimate due to their non-controlled design. Conclusion Adding bisphosphonates to the treatment of MM reduces pathological vertebral fractures, SREs and pain but - from the published evidence - not mortality. Assuming the baseline risk of 20%-50% for vertebral fracture without treatment, we estimate that between 8 - 20 MM patients should be treated to prevent vertebral fracture(s) in one patient. Similarly, assuming the baseline risk of 31%-76% for pain amelioration without treatment, we estimate that between 5 - 13 MM patients should be treated to reduce pain in one patient. Also, with the baseline risk of 35%-86% for SREs without treatment, we estimate that between 6 - 15 MM patients should be treated to prevent SRE(s) in one patient. No bisphosphonate appears to be superior to others. Disclosures: Glasmacher: Celgene: Employment, Equity Ownership.

Abstract 484 On behalf of the NCRI AML Working Group, United Kingdom There is a general acceptance that survival in younger patients (75% relapse remains a problem. Optimisation of induction and defining the schedule and duration of post-induction chemotherapy remains important even as novel approaches such as antibody directed therapy or small molecule therapy are added. The MRC AML15 compared 3 induction chemotherapy schedules and 4 post induction options. In induction patients were randomised to two course of DA or ADE or FLAG-Ida. In each case patients were also eligible to be randomised to receive Gemtuzumab Ozogamicin (GO) or not, which has been separately reported. In consolidation patients were randomised to MRC schedule (MACE/MidAC) or two courses of High Dose Ara-C and for those randomised to Ara-C to a dose of 3g/m2 or 1.5g/m2. Finally patients could be randomised to receive, or not, a 5th course comprising Ara-C. Between October 2002 and January 2009, 1982 patients were randomised between DA (n=994) and ADE (n=988); up to May 2007, patients could also be randomised to FLAG-Ida; the randomised comparison with ADE recruited 1266 patients (FLAG-Ida n=635, ADE n=631). For part of the trial patients could be randomised to GO or not (306 patients in the DA v ADE and 297 in the FLAG-Ida v ADE comparisons were randomised). The median age of patients in the randomised comparisons was 49 years. Overall the ORR was 85% (CR 80%, CRi 5%) and the OS at 5 years is 43%. The randomisations were balanced for age, sex, performance status, de Novo/secondary disease, cytogenetics, white cell count, and GO allocation. With the exception of a suggestion of greater benefit on ORR with FLAG-Ida among patients given GO, overall outcomes were not affected by GO treatment. In consolidation 1445 patients were randomised between MRC (n=723) and HD Ara-C (n=722). Of the 723 Ara-C patients, 328 were randomised to 3g/m2 and 329 to 1.5g/m2: 227 of the 1619 patients who received 4 courses were randomised for a 5th course with 112 allocated 5 courses. Conclusions: There were no significant differences in the response rate between induction treatments. FLAG-Ida resulted in a reduced relapse rate but this was balanced by increased myelosuppression and deaths in CR resulting in no overall benefit in survival. No survival differences were seen in any of the post induction comparisons, but the MACE/MidAc arm produced more myelosuppression and required more supportive care, and led to increased death in remission. Disclosures: No relevant conflicts of interest to declare.

Relapse remains the main cause of treatment failure in Acute Myeloid Leukaemia (AML). Some 65% of AML patients lack a specific fusion transcript which can serve as a molecular target for detecting residual disease. However, in over 75% of AML cases, the Wilm’s Tumor (WT1) gene is over-expressed and can therefore be used for MRD monitoring by RQ-PCR. The clinical value of serial MRD monitoring in WT1 positive patients was prospectively assessed in the AML-15 Trial which opened in 2002. The trial compared 3 induction regimens (DA v/s ADE v/s FLAG Ida), followed by randomisation in consolidation (courses 3 and 4) to either MACE + MidAC or 2 doses of Ara-C (3g/m2 or 1.5g/m2) and to stop or have a 5th course (Ara-C 1.5g/m2). Patients were also randomised to receive Gemtuzumab Ozogamicin (3mg/m2) at induction and/or consolidation. Adults in the standard and poor risk groups were scheduled to receive an allograft if a matched donor was available. An “in house” WT1 assay used primers/probes covering exons 7 and 8 in the 3′ region of the WT1 gene. WT1 transcripts from BM and PB were measured by real-time quantitative PCR (RQ-PCR) on the 7900 HT ABI machine, at diagnosis, after each course of chemotherapy, stem cell transplant, and during remission for up to 3 years. WT1 copies were normalised to ABL gene and expressed per 105ABL copies. WT1 was over-expressed in >80% of AML cases screened, when compared to normal BM copies (median 544, range 31–1173), PB (median 10, r 0–58), PBSC (median 71, r 0–483). In this study, 700 patients (excluding Core Binding Factor positive AML) with diagnostic BM WT1 levels >104 copies have been recruited and data on 416 are presented. Follow up is complete to 1/1/08 with 244 relapses after a median follow-up of 40 months. At complete remission (CR) following induction, a greater log reduction in WT1 transcripts was associated with a significantly reduced relapse risk even after adjustment for age, WBC, cytogenetics, performance status and de Novo/Secondary AML (hazard ratio (HR) 0.57 (0.46–0.71) per log reduction p500 copies. In all cases PB transcript levels were more prognostic than levels from BM. We conclude that MRD monitoring in WT1 positive AML patients allows risk stratification based on treatment response, and in this group of patients, MRD monitoring adds to already known risk factors. Whether patients identified as being at high risk of relapse benefit from further treatment, including allogeneic stem cell transplantation, remains to be determined in future studies.

Background: The prognostic value of mutations in NPM1 and FLT3-ITD is well known. Previously, Schlenk et al. (ASH 2007, Abstract 297) have reported that survival was significantly improved in a group of older patients who were NPM1 mutant/FLT3-ITD Wild Type when treated with ATRA; there was no significant improvement in overall survival in patients who were either NPM1 WT or FLT3-ITD mutant. We report here on a subset of the UK MRC AML12 trial who have been characterised for FLT3-ITD and NPM1 and who were randomised between ATRA in induction and no ATRA. Patients and Methods: A total of 393 patients were identified and characterised. The median age was 46 years (range 16–68); NPM1 and FLT3-ITD status were determined using methods previously reported (Gale et al. Blood 2008). Median follow-up for survival is 7.1 years. The overall results of the ATRA randomisation have previously been reported (Burnett et al. ASH 2002 Abstract 529) and show no benefit for ATRA treatment. All patients were treated with Daunorubicin, Ara-C and Thioguanine (DAT) with a randomisation between two doses of Ara-C, and were randomised to receive, or not, ATRA 45mg/m2/d during courses 1 and 2 of chemotherapy. ATRA was given for a median of 56 days. Results: A total of 143 (36%) patients had an NPM1 mutation and 93 (24%) had a FLT3-ITD mutation. No significant interactions were seen between either NPM1 status, or FLT3-ITD status and ATRA treatment with respect to complete remission, overall survival or relapse free survival (see Table). Estimates of the hazard ratios (HR) for the interaction between FLT3-ITD and ATRA, and NPM1 and ATRA for overall survival were 0.75 (95% CI 0.42–1.32 p=0.3) and 0.66 (95% CI 0.38–1.12 p=0.13), where an HR Conclusions: In this randomised comparison of ATRA therapy in younger patients with AML there were no significant interactions. Any impact of NPM1 or FLT3-ITD status on treatment with ATRA is likely to be relatively small or non-existent. CR OS at 5years RFS at 5 years ATRA No ATRA OR, 95% CI ATRA No ATRA HR, 95% CI ATRA No ATRA HR, 95% CI NPM1 WT 80% 86% 1.59 (0.82–3.09) 36% 35% 1.07 (0.79–1.45) 33% 29% 1.05 (0.76–1.46) NPM1 Mutant 91% 89% 0.81 (0.27–2.43) 57% 44% 0.70 (0.45–1.11) 49% 39% 0.83 (0.63–1.31) Interaction with ATRA p=0.3 0.1 0.4 FLT3 WT 83% 87% 1.46 (0.77–2.74) 45% 42% 0.97 (0.73–1.30) 42% 35% 0.87 (0.64–1.18) FLT3 ITD 89% 88% 0.89 (0.25–3.17) 40% 25% 0.73 (0.44–1.21) 32% 26% 0.89 (0.53–1.50) Interaction with ATRA p=0.5 0.4 0.9

Novel treatments are needed for older patients with AML who are not thought likely to benefit from standard chemotherapy. A recent randomised trial demonstrated that Low Dose Ara-C (LD Ara-C) was superior to best supportive care (BSC) (Burnett et al Cancer2007: 109: 1114–1124). The Farnesyl Transferase inhibitor, Tipifarnib, has produced encouraging responses in this patient group (Lancet 109; 1387–1394) although in a subsequent randomised comparison vs BSC it was not superior (Haroussea et al Blood1997:110,135a) The UK NCRI AML16 trial aims to test novel agents or combinations in untreated older patients with AML or high risk MDS (marrow blasts >10%) following a “pick a winner” design. The intention is to randomise up to 50 patients per arm with the expectation that a “winner” will achieve a remission rate in excess of 30%, compared with 15–20% with LD Ara-C. If that is achieved randomisation will continue with OS and DFS as endpoints. We report our experience of LD Ara-C (20mg bd days 1–10 for 4 courses versus LD Ara-C combined with Tipifarnib (300mg bd days 1–21) for 4 courses at 6–8 week intervals. Patient Details: Sixty-five patients were randomised between December 2006 and October 2007. The median age was 74.4 years with 82% of patients over 70 years. There were no differences between the treatment arms with respect to age distribution, gender, performance status, de novo/secondary AML, high risk MDS presenting WBC or cytogenetic risk group. Because of concerns about excess deaths in the Tipifarnib arm and therefore the low probability that LD Ara-C + Tipifarnib would be superior to LD Ara-C alone, the DMEC recommended premature closure of the Tipifarnib arm. Conclusions: While other agents combined with Tipifarnib may continue to show promise, the combination with LD-Ara-C in this patient population was not beneficial.

The UK MRC AML12 Trial in Acute Promyelocytic Leukaemia was conducted between 1994 and 2002 and involved the simultaneous administration of ATRA with standard chemotherapy (Daunorubicin/Etoposide/Ara-C) until complete remission. This approach resulted in an overall survival at 5 years of 80% for patients given long ATRA; however it was associated with prolonged periods of neutropenia, days on antibiotics and hospitalisation. In the successor trial, AML15, which recruited 291 patients between May 2002 and June 2007 from 90 treatment centres, the same treatment was compared with the Idarubicin/ATRA approach developed by the Spanish PETHEMA Group. Patients under 60 years of age were randomised to receive either the MRC Arm (four treatment courses ADE/ADE/MACE and MidAC with ATRA given for the first 60 days of treatment) or the Spanish Arm (two courses of Idarubicin/ATRA followed by two courses of Mitoxantrone/ATRA, and then 18 months of maintenance. In course 3 of each arm patients are randomised to receive, or not, Gemtuzumab Ozogamicin (GO) on day 1. Since it was considered unlikely that sufficient patients would be available to demonstrate any effectiveness difference the primary endpoints were toxicity, neutropenic days, days on antibiotics, hospitalisation, supportive care requirements and Quality of Life which was measured at baseline, and 3, 6, 12 and 24 months from entry using the EORTC QLQ30, plus leukaemia module, and Hospital Anxiety and Depression Score. Results: No difference in CR was seen (91% MRC vs 93% Spanish; OR 0.73, 95%CI 0.30–1.77, p=0.5); 10 vs 8 patients died during induction and 1 patient in each arm had resistant disease. Five (MRC) vs 6 (Spanish) patients have relapsed. More patients (n=11) in the MRC arm died in CR than in the Spanish arm (n=2) (p=0.009). Overall survival, with a median follow up of 24 months, was not significantly different (85% Spanish vs 81% MRC at 4 years; HR 0.57 95% CI 0.29–1.11, p=0.10). The MRC treatment was significantly more myelosuppressive which resulted in significantly greater need for blood product support, days on antibiotics and hospitalisation, particularly in the second course (p Conclusion: We confirm that the Spanish combination of Idarubicin and ATRA is as effective as the more intensive MRC chemotherapy based approach. The extra chemotherapy involved in the MRC approach increased supportive care requirements and resulted in an excess of deaths in CR and an adverse impact on quality of life. We will now compare the Idarubicin/ ATRA treatment with the chemotherapy free combination of ATRA/ Arsenic Trioxide.

A key challenge in the management of APL is to improve upon survival rates now achieved with ATRA and anthracycline-based chemotherapy. One approach involves minimal residual disease (MRD) monitoring to identify patients otherwise destined to relapse, whose outcome may be improved by pre-emptive therapy. Patients with PML-RARA+ APL (n=261) were prospectively monitored by RQ-PCR in the MRC AML15 trial, comparing a MRC treatment schedule comprising 4 courses of combination chemotherapy (ADE+ATRA,ADE,MACE,MidAC) with a less intensive PETHEMA schedule involving anthracycline monotherapy+ATRA followed by maintenance (ATRA/6MP/MTX). Paired peripheral blood (PB) and bone marrow (BM) samples were analyzed after each course, then 3-monthly for 2 years. Patients with persistent PCR positivity or molecular relapse (defined by successive PCR positive results with rising PML-RARA transcript level) received ATO and proceeded to autologous/allogeneic transplant according to molecular response; patients unfit for transplant received 3 ATO courses. 4813 samples were analyzed (median 16/pt), including 1507 paired samples. No difference in kinetics of disease response was observed between schedules (median time to molecular CR (CRm): 70d MRC v 67d PETHEMA,p=0.3). Furthermore, neither response kinetics nor MRD status during or immediately after consolidation identified those most at risk of relapse. 6/182 (3%) patients tested PCR positive after course 4. While this has been considered an indication for allogeneic transplant, serial monitoring showed progression in only 2 (given ATO pre-emptively); the other 4 remain in CRm 3–37mo later. Most patients subject to relapse tested PCR negative post-consolidation (10/12,83%), with fusion transcripts first detected at a median of 8mo (3–33mo) later. While concordance between BM and PB MRD results was high in early phases of treatment (r>.46,p

Activating mutations in MPL, the thrombopoietin receptor, were originally described in idiopathic myelofibrosis (IMF) and subsequently in a small minority of patients with essential thrombocythemia (ET). The prevalence and clinical significance of MPL mutations in ET has not been fully characterised. Using prospectively acquired diagnostic and follow-up data from the large MRC PT-1 study, we describe the clinical characteristics, complications and laboratory features of these mutations. Sequencing the entire MPL cDNA from 19 JAK2 wild-type ET patients did not reveal any mutations outside exon 10. Sequencing MPL exon 10 DNA in 88 ET and 112 IMF patients revealed MPLW515L (n=11), MPLW515K (n=3) and MPLS505N (n=2) mutations. The MPLS505N mutation, previously associated with inherited thrombocythemia, was clearly an acquired mutation in both patients (1 ET and 1 IMF). Sensitive PCR-based methods were then developed for all three mutant MPL alleles and used to genotype 776 patients with ET from the PT-1 trial. MPL mutations were detected in 4% of all patients, comprising 9% of JAK2 wild-type patients. MPL mutant patients were older at diagnosis than both JAK2 wild-type (p=0.0001) and JAK2V617F patients (p=0.09). Compared to JAK2V617F patients, MPL mutant patients had lower hemoglobin and higher platelet levels at diagnosis (p=0.0001 and p=0.006 respectively). There were no differences in diagnostic blood counts between MPL mutant and JAK2 wild-type patients. Diagnostic trephine histology for JAK2V617F (n=168), JAK2 wild-type (n=130) and MPL mutant patients (n=13) was reviewed by three independent histopathologists blinded to mutation status. MPL mutant samples were less cellular than JAK2V617F and JAK2 wild-type samples (p=0.0001 and p=0.003 respectively). Compared to both JAK2V617F and JAK2 wild-type samples, MPL mutant samples had lower erythroid cellularity (p=0.0006 and p=0.005 respectively) and granulocyte cellularity (p=0.006 and p=0.05 respectively). There were no differences in megakaryocyte cellularity or reticulin grade between the three groups. MPL mutant patients had an increased risk of venous thrombosis after trial entry when compared to JAK2 wild-type patients (p=0.02); the rate of venous thrombosis in the MPL mutant group was comparable to the JAK2V617F group. The mutant allele burden (assessed quantitatively by pyrosequencing) was significantly higher in patients with MPLW515K compared to patients with MPLW515L mutations (p=0.0001), suggesting biological differences in the transforming abilities of the two mutations. In colony assays using patient mononuclear cells the MPLW515L mutation was associated with cytokine-independent growth of megakaryocyte colonies but not with erythropoietin-independent erythroid colony growth (4/4 patients). Of 48 MPL mutant patients, only one also carried the JAK2V617F mutation by allele-specific PCR. Taken together, our results describe the prevalence, clinical features and laboratory manifestations of MPL mutations in the context of a large prospective study of ET.

We welcome the contribution from Bornhauser and colleagues on the impact of a FLT3/ITD in acute myeloid leukemia (AML) patients treated with different types of consolidation therapy in the AML 96 study of the DSIL. There is general agreement that the presence of a FLT3/ITD is a poor prognostic

The MRC AML15 Trial is primarily for patients with any form of AML who are under 60 years. One of the questions addressed was whether the addition of the immunoconjugate, Gemtuzumab Ozogamicin (GO) to induction (course 1) and/or consolidation (course 3) is beneficial. In induction patients are randomised to receive either DA (Daunorubicin/Ara-C) or ADE (Ara-C/Daunorubicin/Etoposide) or FLAG-Ida (Fludarabine/Ara-C/Idarubicin/G-CSF) and in consolidation either MACE (Amsacrine/Etoposide) or HD Ara-C (3.0g/m2 or 1.5g/m2 per dose). Our prior pilot trial had shown that GO 3mgs/m2 could be safely added to day 1 of each of these treatments (Kell et al Blood102, 4277–4283). Here we report the preliminary results of the effect of combining GO with induction chemotherapy. This randomisation achieved its recruitment target and was closed on 30 June 2006. All other comparisons in the trial, including GO in consolidation, remain open. Patients: A total of 1115 patients were randomised between July 2002 and June 2006. The median age was 49 (range 0–71) years: 53% of patients were male: 92% (n=1027) had de novo disease: 95% had WHO performance score of 30 x 109/l and LFT’s > normal were initially excluded but admitted from March 2004. APL patients were not eligible for entry. 15% of patients with data had favourable 71% intermediate, and 14% adverse cytogenetics. Over 83% were CD33 positive. Results: The overall remission rate was 85% with no differences between the arms for GO vs no GO in CR (85% vs 85%) induction death (8% vs 7%) or resistant disease (7% vs 8%). There was a modest increase in mucositis on the GO arm in course 1 only (p=0.04) and increased AST and Alt toxicity in C1 (p=.002; p=.03) but no difference in bilirubin grades. GO patients used more platelets (19 vs 14; p

Treatment strategies for younger patients (2.667 (poor risk) which gave excellent discrimination for survival from CR at 5 years of 63%, 47% and 24% (p MRC Risk Group Total Good Standard Poor Crosstabulation of old and new risk group classification New Risk Classification Good 309 28 0 337 (14%) Standard 51 1288 42 1381 (59%) Poor 2 274 353 629 (27%) Total 362 (15%) 1590 (68%) 395 (17%) 2347

VTE is a common clinical problem in cancer patients. Uncertainty remains as to whether prophylactic treatment of oral anticoagulants (mainly warfarin) or LMWHs, given by subcutaneous injection, is beneficial and practice varies widely. It has been proposed that anticoagulant therapy may also improve survival in cancer patients. The largest trial of VTE prophylaxis in cancer patients with central venous catheters (CVCs), WARP, has just been completed and showed no apparent benefit for low dose warfarin - among 811 patients randomised to low dose warfarin versus not, the VTE rate was 5% in both arms (odds ratio [OR]=1.04, 95% confidence interval [CI]=0.56–1.92, p=1.0) and mortality was not reduced (OR=1.02, CI=0.74–1.4, p=0.8) (Young et al., ASCO, 2005, LBA8004). This result needs to be put in the context of other RCTs of prophylaxis with anticoagulants, so we performed a meta-analysis to evaluate the effect of warfarin and LMWH on VTE rates and cancer mortality. Computerized searches for trials were performed, including MedLine, Embase, NCI trials register, ASCO and ASH meeting abstracts. Standard meta-analysis methods were used and the results presented as ORs and CIs, with heterogeneity of treatment effect between trials examined using tests for interaction. In 5 trials (n=1355 patients) of warfarin v. not, there was no clear evidence of a decrease in VTEs (OR=0.68, CI=0.46–1.01, p=0.06). An additional 3 trials (n=2034 for all 8 trials) had survival data and there was no clear evidence of a decrease in mortality (OR=0.88, CI=0.77–1.01, p=0.06). In 5 trials (n=793) of LMWH v. not, VTE rate was reduced (OR=0.53, CI=0.32–0.87, p=0.01). 6 trials (n=1295) contained survival data and mortality was reduced with LMWH (OR=0.79, CI=0.67–0.92, p=0.003). In 3 trials (n=985) of LMWH v. oral anti-coagulation (mainly with warfarin), VTE rates were lower with LMWH (OR=0.46, CI=0.32–0.67, p

The NCRI AML14 Trial was devised for patients aged over 60 years with de novo or secondary AML or high risk MDS (defined as >10% blasts). There was a non-intensive option which compared Low Dose Ara-C vs Hydroxyurea each with or without All-Trans Retinoic Acid which has been reported previously (Burnett et al Blood2004: 249a), and an intensive approach which we now report. This compared two dose levels of Daunorubicin (50mg/m2 vs 35mg/m2) and Ara-C (200mg/m2 vs 400mg/m2) within a Daunorubicin/Ara-C (3+10) followed by (3+8) schedule. After a 3rd course (MidAC: Mitoxantrone/Ara-C), patients were randomised to receive or not a 4th course (ICE: Idarubicin/Ara-C/Etoposide). In addition some patients (n=200) were randomised to receive PSC-833 in addition to Dauno 35. The results of this randomisation have been reported previously (Burnett et al Blood 2003614a). A total of 1273 patients entered the trial between December 1998 and closure in May 2005, from 136 centres. Follow-up is complete to 1st April 2005, with median follow-up of 33 months. The median age was 67 (range 44-88). Cytogenetics were known for 67% of patients: of these, 3% had favourable, 74% intermediate and 23% adverse cytogenetics. 72% had de novo AML, 17% had secondary disease and 11% had high risk MDS. The overall CR rate was 62% and the Relapse Risk (RR), Disease Free Survival (DFS), and Overall Survival (OS) were 84%, 13%, and 13% at 5 years. 896 patients were randomised to D50 vs D35. No significant differences were found in CR (63% vs 64% OR 1.04 (95% CI 0.78–1.37)), RR (84% vs 85% OR 0.85 (95%CI 0.65–1.06)), DFS 14% vs 13% OR 0.86 (95% CI 0.70–1.06)), or OS (15% vs 13% OR 0.92, 95% CI (0.72–1.08)). Likewise, there were no significant differences in outcome for the 1264 patients randomised to Ara-C 200 vs 400: CR (62% vs 62% OR 1.00 (95% CI 0.80–1.27)), RR (84% vs 85% OR 1.14 (95% CI 0.95–1.37)), DFS (13% vs13% OR 1.14 (95% CI 0.96–1.14)) or OS (13% vs 12% OR 1.00 (95%CI 0.87–1.14)). In the 255 patients randomised to a total of 4 vs 3 courses the RR (83% vs 76% OR 0.90 95%CI 0.66–1.24), DFS (16% vs 18% OR 0.94 95% CI 0.69–1.28) and OS (23% vs 22% OR 1.05, 95% CI 0.75–1.47) were not significantly different. From this preliminary analysis we conclude that there is no difference between a reduced dose of Daunorubicin (35mg/m2) compared with standard dose, or between enhanced Ara-C dose (400mg/m2) or standard dose and the confidence intervals are consistent with at most a moderate difference in treatment effect. We found no benefit for giving more than 3 courses of total treatment to patients in this age group, although the confidence intervals here do not rule out moderate, but potentially meaningful differences.

AML is a heterogeneous disease and several prognostic factors are well established in younger patients. Less work has been done on parameters affecting outcome in older patients, aged 60 years or over, especially in the majority who receive non-intensive therapy. We used data from the Medical Research Council AML11 trial (n=842) to develop a prognostic index and validated this using the subsequent Leukaemia Research Fund AML14 trial (n=943). Patients in AML11 received intensive standard induction chemotherapy (daunorubicin/Ara-C based); those in AML14 received either similar intensive therapy (AML14I, n=790) or non-intensive therapy with either hydroxyurea or low-dose Ara-C (AML14NI, n=153). Patients with APL were excluded from AML11 analysis since they were not eligible for AML14. Overall survival (OS) was analysed using multivariate Cox regression modelling to identify factors independently related to outcome, and a prognostic index was created for OS using the regression coefficients. From this patients were assigned to 3 equal sized risk groups (good, standard, poor). In AML11, adverse cytogenetics (−5, del(5q), −7, abn(3q), complex), increasing white blood count (WBC), secondary AML, poor performance status (PS) and increasing age were related to poor OS (all p Overall survival by risk group Survival at 1 year (%) Survival at 3 years (%) Risk Group Good Standard Poor Good Standard Poor AML11 55 40 15 25 13 7 AML14I 55 30 20 25 8 8 AML14NI 23 12 2 2 0 0

It is widely accepted that internal tandem duplications (ITDs) of the juxtamembrane domain of FLT3 occur in about one quarter of cases of acute myeloid leukaemia (AML) in young adults and predict for early relapse from complete remission (CR). Constitutive activation of FLT3 can also arise from mutations in the tyrosine kinase domain (TKD) but there is controversy as to the clinical significance of this class of mutation. This is partly due to the small sample size of some studies and the inclusion of elderly patients who have a poor prognosis regardless of their FLT3 status. To definitively resolve this issue we have screened for TKD mutations in AML blast cells from 1339 young adult patients included in the UK MRC AML 10 and 12 trials using a sensitive denaturing HPLC technique (Transgenomic WAVE®). Mutant samples were confirmed by sequencing or specific restriction digest. 161 of 1339 (12%) patients had a TKD mutation which is a higher frequency than previously reported both because of the sensitivity of the technique and the detection of mutations outside the EcoRV digest site. 91 patients (6.8%) were deemed to have high level mutant arbitrarily defined as ≥ 20% of all FLT3 alleles and 70 (5.2%) had a low level mutant. 79 of the 161 (49%) mutants were Asp835Tyr. 8 mutants occurred outside the EcoRV digest site and 3 novel mutations were characterised. 372 (28%) patients in this cohort had an ITD. There was a negative correlation between the presence of an ITD and a TKD mutation with only 2.5% having evidence of both mutations. Furthermore, high levels of both class of mutation in the same patient were not seen so that it is possible that both mutations never arise in the same cell. The demographics of patients with a TKD mutation differed from those with a FLT3 ITD in that the presence of TKD mutations were not correlated with FAB type and were infrequent in patients with secondary AML. Both types of mutation were more frequent in patients with a high white count but were infrequent in patients with adverse cytogenetics. The presence of TKD mutations did not impact on CR rate, the incidence of resistant disease or the induction death rate. In contrast to FLT3 ITDs, TKD mutations were associated with a reduced relapse rate (odds ratio [OR] 0.77, 95% confidence intervals [CI] 0.59–0.99, P.04), improved disease free survival (OR 0.75, CI 0.60–0.93, P.008) and increased overall survival (OR 0.79, CI 0.64–0.97, P.02). In patients with wild type FLT3 the actuarial relapse rate at 5 years was 46%, compared to 57% with an ITD (excluding rare double mutants) and 34% in those with a TKD mutation. The overall survival at 5 years was 44%, 35% and 55% respectively. In multivariate analysis, the presence of a TKD mutation still had an effect on the relapse rate (OR 0.82, CI 0.67–1.01, P.05) and overall survival (OR 0.83, CI 0.70–0.98, P.03). These data suggest that different classes of activating mutation of the same tyrosine kinase receptor can be associated with markedly different clinical outcomes. FLT3 ITDs are associated with a poor prognosis and FLT3 TKD mutations with a relatively good prognosis. This unexpected genotype-phenotype relationship has not previously been described with oncogenic mutations and is significant to the understanding of the pathophysiology of chemoresistance as well as prognostic stratification.

Between 1988 and 2002, 868 children (0–15 years) were entered into MRC AML 10 (1988–95, n=341) and AML 12 (1995–2002, n=527) trials. Children were allocated to one of three MRC risk groups: good risk - patients with t(8,21),inv(16),t(15,17) irrespective of bone marrow status after course 1 or the presence of other genetic abnormalities; standard risk - patients with neither favourable nor adverse cytogenetics and not more than 15% blasts in the bone marrow after course 1; poor risk - patients with more than 15% blasts in the bone marrow after course 1 or with adverse abnormalities of -5,-7, del(5q), abn(3q), complex (>/-5 abnormalities) and without favourable genetic abnormalities. Outcome from CR - death in CR (DCR), relapse risk (RR), disease-free survival (DFS) and survival from CR (OSCR) - was analysed by MRC risk group. RISK GROUP AML 10 AML 12 Good Standard Poor Good Standard Poor DCR (8yr %) 9 13 11 7 5 10 RR (8yr %) 35 40 66 21 37 53 DFS (8yr %) 59 52 31 74 60 42 OSCR (4yr %) 81 60 39 88 75 49 OSCR (8yr %) 78 57 37 84 72 49 In AML 10 all patients were eligible for SCT with a histocompatible sibling donor, but unrelated donor transplantation was not part of the protocol. Because of their favourable outcome in AML 10, good risk children were not eligible for SCT in AML 12 and part way through the trial a similar approach was adopted for standard risk patients, whilst SCT continued to be recommended for poor risk patients with their inferior outcome. Both sibling and unrelated donor transplantation were permitted. In AML 10 and AML 12, 38 of 139 (27%) poor risk children underwent SCT - 17 sibling allografts, 11 unrelated donor allografts and 10 autografts. The procedural mortality was: 6%, 55% and 0% respectively. Mantel-Byar analysis (to account for time to SCT) comparing transplanted with non-transplanted poor risk children showed no evidence of reduction in relapse risk (HR 1.02, 95% CI 0.58–1.79, p=0.9), disease-free survival (HR 1.47, 95% CI 0.87–2.50, p=0.16) or survival benefit (HR 1.64, CI 0.94–2.85, p=0.08 against SCT), both overall or for any type of SCT. The survival at 8 years from SCT was 41% for sibling allografts, 18% for unrelated donor allografts and 60% for autografts. OUTCOME BY TREATMENT - AML 10 & AML 12 Poor Risk (all) Poor Risk (censored at SCT) DCR (8yr %) 11 5 RR (8yr %) 58 55 DFS (8yr %) 37 43 OSCR (4yr %) 45 50 OSCR (8yr %) 44 50 Outcome is relative and, whilst poor risk children still do worse than good and standard risk patients, the outcome for poor risk children has improved. A survival at 8 years from CR of about 50% for poor risk children in AML12 (and no deaths beyond 4 years suggesting that most of those who survive are cured), raises questions as to whether any children with AML should be transplanted in 1st CR given the mortality and morbidity of the procedure. The high mortality associated with unrelated donor transplantation requires further investigation.

An acquired V617F mutation in JAK2 occurs in most patients with polycythemia vera (PV) but only half of those with essential thrombocythemia (ET) and idiopathic myelofibrosis. It is not known whether mutation-bearing patients are biologically distinct from those lacking the mutation, or why the same mutation is associated with different phenotypes. Two sensitive PCR-based methods were used to assess the JAK2 mutation status of 806 patients with ET, including 776 from the MRC PT-1 trial and two other prospective studies. The combined cohort provides a unique resource for studying ET, particularly in view of its large size, centralized review of end-points and comprehensive follow-up. The involvement of a large number of secondary and tertiary centers, together with the inclusion of patients in all risk categories, suggest the results are of general relevance. JAK2 mutation status divided the cohort into two distinct subgroups. Mutation-positive patients (53.4%) displayed multiple features resembling PV, with significantly increased hemoglobin levels (p JAK2 mutation status defines two biologically distinct subgroups of ET with differences in presentation, outcome and response to therapy. Our results suggest a model in which V617F-positive ET and PV form a continuum, with the degree of erythrocytosis determined by physiological or genetic modifiers, including iron stores, epo homeostasis, gender and V617F-homozygosity. This concept has major implications for the classification, diagnosis and management of MPDs.

There is no adequate treatment for older patients with AML who are not considered fit for intensive chemotherapy who comprise a significant proportion of the AML population. As part of the ongoing NCRI (formerly MRC) AML14 Trial in patients over 60 years patients were randomised to Low Dose Ara-C (20mg bd for 10 days every 4–6 weeks) versus Hydroxycarbamide (Hydroxyurea). Because of preclinical studies showing sensitization to Ara-C, patients were in addition randomised to receive All-transretinoic acid or not (45mgs/m2 for 60 days). Two hundred and four patients entered; 129 had a WHO score of 65 yrs; 108 had do novo, 53 secondary disease and 28 had high risk MDS (blasts >10%). One hundred and ninety-nine patients entered the HU vs LD-Ara-C randomisation and 204 entered the ATRA randomisation. The arms were balanced with respect to age; sex; disease type; WBC and performance score. Complete remission was seen in 1 of 92 (1%) patients in the HU arm and 15 of 94 (17%) in the LD Ara-C arm (P=0003). Overall survival was considerably improved (hazard ratio 0.61, 95% Cl 0.45 to 0.82, p=0.001). This improvement in outcome was not obtained at the expense of less well tolerated treatment: toxicity and supportive care requirements were similar between the two groups. However, there were no significant differences in outcome between patients given ATRA or not overall or within the treatment arms, although numbers were too small to rule out moderate benefits or disbenefits of treatment (hazard ratio for OS 0.97, 95% Cl 0.73 to 1.28, p=0.8). We conclude that LD-Ara-C in the schedule chosen could be adopted as a standard of care for older patients not fit for intensive chemotherapy but the outcome remains poor. Future strategies could combine Low Dose Ara-C with novel agents. This trial received a research grant from the UK Leukaemia Research Fund.

ATRA has a key role in the treatment of APL where the presence of the PML-RARα protein is predictive of response. Pre-clinical data and some non-randomised clinical studies suggested that ATRA may enhance the effect of chemotherapy particularly based on its potential to reduce the Bcl2 expression which may thereby improve apoptotic responses to chemotherapy. We have evaluated this possibility in 3 randomised trials in remission induction in patients under 60 years (AML12) n=1097; in relapse (AML-HR) n=362; and as a non-intensive approach in older patients (AML14-non-intensive) n=207. In AML12 patients were randomised to receive standard induction chemotherapy options (ADE vs DAT or MAE vs DAT with Ara-C at either Ara-C 200mg/m2/day or 400mgs/m2/day). There was no significant difference in CR rate (83% vs 85%) or survival at 5 years (40% vs 36%) on ATRA or no ATRA overall or in any demographic, risk or treatment subgroup. Patients with high risk disease (adverse cytogenetics or less than PR after course 1 of induction) or relapse were randomised to receive Fludarabine/Ara-C or ADE with or without G-CSF, with or without ATRA for 60 days. The addition of ATRA made no overall impact on remission rate (61% vs 61%) or survival at 3 years (23% vs 27%). There was no subgroup where an advantage was seen. In AML14 patients not considered fit for intensive treatment were randomised to receive Hydroxyurea or Low Dose Ara-C with or without ATRA for 90 days. There was no evidence that patients who received ATRA derived any benefit either overall or within the treatment arms. In this extensive assessment, no advantage of adding ATRA to chemotherapy - as a continuous daily dose during induction - could be detected in any trial or in any subgroup within each trial.