Characteristics and outcome of patients with therapy-related acute promyelocytic leukemia front-line treated with or without arsenic trioxide.

Therapy-related acute promyelocytic leukemia (t-APL) is relatively rare, with limited data on outcome after treatment with arsenic trioxide (ATO) compared to standard intensive chemotherapy (CTX). We evaluated 103 adult t-APL patients undergoing treatment with all-trans retinoic acid (ATRA) alone (n=7) or in combination with ATO (n=24), CTX (n=53), or both (n=19). Complete remissions were achieved after induction therapy in 57% with ATRA, 100% with ATO/ATRA, 78% with CTX/ATRA, and 95% with CTX/ATO/ATRA. Early death rates were 43% for ATRA, 0% for ATO/ATRA, 12% for CTX/ATRA and 5% for CTX/ATO/ATRA. Three patients relapsed, two developed therapy-related acute myeloid leukemia and 13 died in remission including seven patients with recurrence of the prior malignancy. Median follow-up for survival was 3.7 years. None of the patients treated with ATRA alone survived beyond one year. Event-free survival was significantly higher after ATO-based therapy (95%, 95% CI, 82-99%) as compared to CTX/ATRA (78%, 95% CI, 64-87%; P=0.042), if deaths due to recurrence of the prior malignancy were censored. The estimated 2-year overall survival in intensively treated patients was 88% (95% CI, 80-93%) without difference according to treatment (P=0.47). ATO when added to ATRA or CTX/ATRA is feasible and leads to better outcomes as compared to CTX/ATRA in t-APL.

Expression of CD33 is a predictive factor for effect of gemtuzumab ozogamicin at different doses in adult acute myeloid leukaemia.

It remains unclear in adult acute myeloid leukaemia (AML) whether leukaemic expression of CD33, the target antigen for gemtuzumab ozogamicin (GO), adds prognostic information on GO effectiveness at different doses. CD33 expression quantified in 1583 patients recruited to UK-NCRI-AML17 (younger adults) and UK-NCRI-AML16 (older adults) trials was correlated with clinical outcomes and benefit from GO including a dose randomisation. CD33 expression associated with genetic subgroups, including lower levels in both adverse karyotype and core-binding factor (CBF)-AML, but was not independently prognostic. When comparing GO versus no GO (n=393, CBF-AMLs excluded) by stratified subgroup-adjusted analysis, patients with lowest quartile (Q1) %CD33-positivity had no benefit from GO (relapse risk, HR 2.41 (1.27-4.56), P=0.009 for trend; overall survival, HR 1.52 (0.92-2.52)). However, from the dose randomisation (NCRI-AML17, n=464, CBF-AMLs included), 6 mg/m2 GO only had a relapse benefit without increased early mortality in CD33-low (Q1) patients (relapse risk HR 0.64 (0.36-1.12) versus 1.70 (0.99-2.92) for CD33-high, P=0.007 for trend). Thus CD33 expression is a predictive factor for GO effect in adult AML; although GO does not appear to benefit the non-CBF AML patients with lowest CD33 expression a higher GO dose may be more effective for CD33-low but not CD33-high younger adults.

Arsenic trioxide-based therapy of relapsed acute promyelocytic leukemia: registry results from the European LeukemiaNet.

In 2008, a European registry of relapsed acute promyelocytic leukemia was established by the European LeukemiaNet. Outcome data were available for 155 patients treated with arsenic trioxide in first relapse. In hematological relapse (n=104), 91% of the patients entered complete hematological remission (CR), 7% had induction death and 2% resistance, 27% developed differentiation syndrome and 39% leukocytosis, whereas no death or side effects occurred in patients treated in molecular relapse (n=40). The rate of molecular (m)CR was 74% in hematological and 62% in molecular relapse (P=0.3). All patients with extramedullary relapse (n=11) entered clinical and mCR. After 3.2 years median follow-up, the 3-year overall survival (OS) and cumulative incidence of second relapse were 68% and 41% in hematological relapse, 66% and 48% in molecular relapse and 90 and 11% in extramedullary relapse, respectively. After allogeneic or autologous transplantation in second CR (n=93), the 3-year OS was 80% compared with 59% without transplantation (n=55) (P=0.03). Multivariable analysis demonstrated the favorable prognostic impact of first remission duration ⩾1.5 years, achievement of mCR and allogeneic or autologous transplantation on OS of patients alive after induction (P=0.03, P=0.01, P=0.01) and on leukemia-free survival (P=0.006, P<0.0001, P=0.003), respectively.

Hyperdiploidy with 49-65 chromosomes represents a heterogeneous cytogenetic subgroup of acute myeloid leukemia with differential outcome.

Chromosome gain is frequent in acute myeloid leukemia (AML) and is counted alongside structural abnormalities when determining karyotype complexity. However, there are few studies investigating the cytogenetic profile and outcome of patients with a hyperdiploid karyotype (49-65 chromosomes, HK). We identified 221 (14%) patients with HK out of 1563 patients with three or more chromosomal abnormalities. HK was not associated with sex, white cell count and secondary disease status, but was more prevalent among children (22% vs 13%). The pattern of chromosomal gain and loss was non-random and chromosomes 8, 13 and 21 were the most frequently gained. Three distinct subgroups (numerical, structural and adverse) were identified with differential outcome: 5-year cumulative incidence of relapse of 52%, 68% and 76%, respectively (P=0.008). Patients in the adverse subgroup had poorer survival compared with patients with only numerical abnormalities (adjusted hazard ratio: 2.01 (95% confidence interval: 1.43-2.83), P=0.0002). This outcome heterogeneity was similar among children and adults. In conclusion, AML patients with a HK should not automatically be assigned to the adverse cytogenetic risk group on the basis of complexity. Instead they should be assessed for the presence of specific chromosomal abnormalities, which are known to harbour an adverse effect.

Establishing optimal quantitative-polymerase chain reaction assays for routine diagnosis and tracking of minimal residual disease in JAK2-V617F-associated myeloproliferative neoplasms: a joint European LeukemiaNet/MPN&MPNr-EuroNet (COST action BM0902) study.

Reliable detection of JAK2-V617F is critical for accurate diagnosis of myeloproliferative neoplasms (MPNs); in addition, sensitive mutation-specific assays can be applied to monitor disease response. However, there has been no consistent approach to JAK2-V617F detection, with assays varying markedly in performance, affecting clinical utility. Therefore, we established a network of 12 laboratories from seven countries to systematically evaluate nine different DNA-based quantitative PCR (qPCR) assays, including those in widespread clinical use. Seven quality control rounds involving over 21,500 qPCR reactions were undertaken using centrally distributed cell line dilutions and plasmid controls. The two best-performing assays were tested on normal blood samples (n=100) to evaluate assay specificity, followed by analysis of serial samples from 28 patients transplanted for JAK2-V617F-positive disease. The most sensitive assay, which performed consistently across a range of qPCR platforms, predicted outcome following transplant, with the mutant allele detected a median of 22 weeks (range 6-85 weeks) before relapse. Four of seven patients achieved molecular remission following donor lymphocyte infusion, indicative of a graft vs MPN effect. This study has established a robust, reliable assay for sensitive JAK2-V617F detection, suitable for assessing response in clinical trials, predicting outcome and guiding management of patients undergoing allogeneic transplant.

Inclusion of chemotherapy in addition to anthracycline in the treatment of acute promyelocytic leukaemia does not improve outcomes: results of the MRC AML15 trial.

Two hundred eighty-five patients, median age 42, with PML-RARα-positive acute promyelocytic leukaemia were randomised to Ara-C-containing 'Medical Research Council (MRC) Chemotherapy'+ATRA (All-trans-retinoic acid) or anthracycline+ATRA (modified 'Spanish') therapy. MRC treatment comprised four courses with ATRA in courses 1-2. Spanish treatment comprised four anthracycline-based courses with ATRA in courses 1-3. In course 3 patients were randomised to gemtuzumab ozogamicin (GO) or not. The Spanish arm received 24-month maintenance. Patients were sequentially molecularly monitored. Quality of life was assessed at baseline, 3, 6, 9, 12, 24 months. Remission rates were similar in both arms (93%): cumulative incidence of haematological relapse (CIHR) was 6% at 5 years; 5 patients relapsed molecularly. Survival post relapse was 80%. There were more deaths in remission in the MRC arm (4% vs 10%: P=0.2). The overall 5-year relapse-free and overall survival was similar between arms (81% vs 82% and 84% vs 83%, respectively). More supportive care and hospitalisation (81.8 vs 63 days, P<0.0001) was required in the MRC arm. GO did not provide benefit. High white blood cell count (>10 × 10(9)/l) was not prognostic overall, or within treatment arms. Both approaches deliver similar results with minor differences in quality of life. MRC treatment required more hospitalisation. This suggests that additional chemotherapy, Ara-C in particular, is not required.

Azacitidine fails to eradicate leukemic stem/progenitor cell populations in patients with acute myeloid leukemia and myelodysplasia.

Epigenetic therapies demonstrate significant clinical activity in acute myeloid leukemia (AML) and myelodysplasia (MDS) and constitute an important new class of therapeutic agents. However hematological responses are not durable and disease relapse appears inevitable. Experimentally, leukemic stem/progenitor cells (LSC) propagate disease in animal models of AML and it has been postulated that their relative chemo-resistance contributes to disease relapse. We serially measured LSC numbers in patients with high-risk AML and MDS treated with 5'-azacitidine and sodium valproate (VAL-AZA). Fifteen out of seventy-nine patients achieved a complete remission (CR) or complete remission with incomplete blood count recovery (CRi) with VAL-AZA therapy. There was no significant reduction in the size of the LSC-containing population in non-responders. While the LSC-containing population was substantially reduced in all patients achieving a CR/CRi it was never eradicated and expansion of this population antedated morphological relapse. Similar studies were performed in seven patients with newly diagnosed AML treated with induction chemotherapy. Eradication of the LSC-containing population was observed in three patients all of whom achieved a durable CR in contrast to patients with resistant disease where LSC persistence was observed. LSC quantitation provides a novel biomarker of disease response and relapse in patients with AML treated with epigenetic therapies. New drugs that target this cellular population in vivo are required.

Molecular pathogenesis of secondary acute promyelocytic leukemia.

Balanced chromosomal translocations that generate chimeric oncoproteins are considered to be initiating lesions in the pathogenesis of acute myeloid leukemia. The most frequent is the t(15;17)(q22;q21), which fuses the PML and RARA genes, giving rise to acute promyelocytic leukemia (APL). An increasing proportion of APL cases are therapy-related (t-APL), which develop following exposure to radiotherapy and/or chemotherapeutic agents that target DNA topoisomerase II (topoII), particularly mitoxantrone and epirubicin. To gain insights into molecular mechanisms underlying the formation of the t(15;17) we mapped the translocation breakpoints in a series of t-APLs, which revealed significant clustering according to the nature of the drug exposure. Remarkably, in approximately half of t-APL cases arising following mitoxantrone treatment for breast cancer or multiple sclerosis, the chromosome 15 breakpoint fell within an 8-bp "hotspot" region in PML intron 6, which was confirmed to be a preferential site of topoII-mediated DNA cleavage induced by mitoxantrone. Chromosome 15 breakpoints falling outside the "hotspot", and the corresponding RARA breakpoints were also shown to be functional topoII cleavage sites. The observation that particular regions of the PML and RARA loci are susceptible to topoII-mediated DNA damage induced by epirubicin and mitoxantrone may underlie the propensity of these agents to cause APL.

Development of standardized approaches to reporting of minimal residual disease data using a reporting software package designed within the European LeukemiaNet.

Quantitative PCR (qPCR) for detection of fusion transcripts and overexpressed genes is a promising tool for following minimal residual disease (MRD) in patients with hematological malignancies. Its widespread clinical use has to some extent been hampered by differences in data analysis and presentation that complicate multicenter clinical trials. To address these issues, we designed a highly flexible MRD-reporting software program, in which data from various qPCR platforms can be imported, processed, and presented in a uniform manner to generate intuitively understandable reports. The software was tested in a two-step quality control (QC) study; the first step involved eight centers, whose previous experience with the software ranged from none to extensive. The participants received cDNA from consecutive samples from a BCR-ABL+ chronic myeloid leukemia (CML) patient and an acute myeloid leukemia (AML) patient with both CBFß-MYH11 and WT1 target genes, they conducted qPCR on their respective hardware platforms and generated a series of reports with pre-defined features. In step two, five centers used the software to report BCR-ABL+ MRD in a harmonized manner, applying their recently obtained CML international scale conversion factors. The QC study demonstrated that this MRD-reporting software is suitable for efficient handling of qPCR data, generation of MRD reports and harmonization of MRD data.

Detection and molecular monitoring of FIP1L1-PDGFRA-positive disease by analysis of patient-specific genomic DNA fusion junctions.

To evaluate current detection methods for FIP1L1-PDGFRA in hypereosinophilic syndrome (HES), we developed a means to rapidly amplify genomic break points. We screened 202 cases and detected genomic junctions in all samples previously identified as RT-PCR positive (n=43). Genomic fusions were amplified by single step PCR in all cases whereas only 22 (51%) were single step RT-PCR positive. Importantly, FIP1L1-PDGFRA was detected in two cases that initially tested negative by RT-PCR or fluorescence in situ hybridization. Absolute quantitation of the fusion by real-time PCR from genomic DNA (gDNA) using patient-specific primer/probe combinations at presentation (n=13) revealed a 40-fold variation between patients (range, 0.027-1.1 FIP1L1-PDGFRA copies/haploid genome). In follow up samples, quantitative analysis of gDNA gave 1-2 log greater sensitivity than RQ-PCR of cDNA. Minimal residual disease assessment using gDNA showed that 11 of 13 patients achieved complete molecular response to imatinib within a median of 9 months (range, 3-17) of starting treatment, with a sensitivity of detection of up to 1 in 10(5). One case relapsed with an acquired D842V mutation. We conclude that detection of FIP1L1-PDGFRA from gDNA is a useful adjunct to standard diagnostic procedures and enables more sensitive follow up of positive cases after treatment.

The molecular anatomy of the FIP1L1-PDGFRA fusion gene.

The FIP1L1-PDGFRA fusion gene is a recurrent molecular abnormality in patients with eosinophilia-associated myeloproliferative neoplasms. We characterized FIP1L1-PDGFRA junction sequences from 113 patients at the mRNA (n=113) and genomic DNA (n=85) levels. Transcript types could be assigned in 109 patients as type A (n=50, 46%) or B (n=47, 43%), which were created by cryptic acceptor splice sites in different introns of FIP1L1 (type A) or within PDGFRA exon 12 (type B). We also characterized a new transcript type C (n=12, 11%) in which both genomic breakpoints fell within coding sequences creating a hybrid exon without use of a cryptic acceptor splice site. The location of genomic breakpoints within PDGFRA and the availability of AG splice sites determine the transcript type and restrict the FIP1L1 exons used for the creation of the fusion. Stretches of overlapping sequences were identified at the genomic junction site, suggesting that the FIP1L1-PDGFRA fusion is created by illegitimate non-homologous end-joining. Statistical analyses provided evidence for clustering of breakpoints within FIP1L1 that may be related to DNA- or chromatin-related structural features. The variability in the anatomy of the FIP1L1-PDGFRA fusion has important implications for strategies to detect the fusion at diagnosis or for monitoring response to treatment.

Characterization of a reference material for BCR-ABL (M-BCR) mRNA quantitation by real-time amplification assays: towards new standards for gene expression measurements.

Monitoring of BCR-ABL transcripts has become established practice in the management of chronic myeloid leukemia. However, nucleic acid amplification techniques are prone to variations which limit the reliability of real-time quantitative PCR (RQ-PCR) for clinical decision making, highlighting the need for standardization of assays and reporting of minimal residual disease (MRD) data. We evaluated a lyophilized preparation of a leukemic cell line (K562) as a potential quality control reagent. This was found to be relatively stable, yielding comparable respective levels of ABL, GUS and BCR-ABL transcripts as determined by RQ-PCR before and after accelerated degradation experiments as well as following 5 years storage at -20 degrees C. Vials of freeze-dried cells were sent at ambient temperature to 22 laboratories on four continents, with RQ-PCR analyses detecting BCR-ABL transcripts at levels comparable to those observed in primary patient samples. Our results suggest that freeze-dried cells can be used as quality control reagents with a range of analytical instrumentations and could enable the development of urgently needed international standards simulating clinically relevant levels of MRD.

Impact of karyotype on treatment outcome in acute myeloid leukemia.

Over the course of the last three decades it has become apparent that the majority of cases of acute myeloid leukemia (AML) are characterized by at least one of a variety of recurrent chromosomal abnormalities. Whilst in many instances it remains uncertain as to which abnormalities represent primary events in the pathogenesis of AML, those which provide critical second hits that are required for progression to full blown leukemia or those that are merely markers of the leukemic process, it is nevertheless clear that diagnostic karyotype is a key determinant of outcome in this disease. Indeed there is mounting evidence to support the notion that cytogenetic analysis can serve to identify biologically distinct subsets of AML that demand tailored therapeutic approaches. This underpins the trend towards more widespread adoption of routine cytogenetic and molecular analysis in the characterization of patients with a diagnosis of acute leukemia. A key challenge for the future is to use this information to achieve greater consensus in risk group assignment of AML which will provide a more reliable framework for determining the most appropriate treatment approach for individual patients with this disease.