ALBERT

Description: Introduction MRD is an important predictor of outcome in childhood ALL. Since 2000, MRD detected by quantitative PCR (qPCR) for immunoglobulin and T-cell receptor gene rearrangements with a minimal sensitivity of 1E-04 has been used for risk group stratification in pediatric BFM trials. Next generation sequencing (NGS) permits rapid parallel sequencing of large numbers of DNA segments. It can overcome most of the limitations of qPCR: it allows highly specific molecular detection of MRD without laborious optimization of patient-specific assays. Moreover it enables not only monitoring of malignant clone but also shows the picture of entire immune background. Aims To develop an assay for immunoglobulin heavy chain (IgH) rearrangements detection on Ion Torrent PGM/Ion Proton platforms and compare the MRD levels with qPCR at BFM stratification timepoints. Methods Two round PCR was used for library preparation. Libraries were created from 450ng (equivalent of 70,000 DNA copies) of bone marrow DNA and 50ng of Human Genomic DNA (Roche). In the first round of PCR rearranged IgH genes were amplified using IGH FR3 BIOMED-2 primers (van Dongen, Leukemia 2003). In the second round the sequencing adapters and multiplex identifiers were added. Sequencing was performed on Ion Torrent PGM/Ion Proton sequencers using a 200bp chemistry. We developed a bioinformatics algorithm for detection of reads with known clonal V-D-J rearrangements within the resulting fastq files. For validation of the assay we sequenced 1E-1 to 1E-5 dilutions of diagnostic samples from 2 patients in multiplicates. The results show that the assay gives reproducible quantitative results up to 1E-4 dilution. Results We sequenced 183 samples from 67 patients (52×day 15, 65×day 33, 66×day 78) with childhood ALL treated according to AIEOP-BFM ALL 2000 protocol with the median coverage 587,406 reads per sample. Eighty-three (45.4%) samples were negative by both methods. Fifteen (8.2%) samples were positive by NGS and negative by qPCR and 14 (7.7%) samples were positive by qPCR and negative by NGS. All the discordant samples had MRD levels below the sensitivity of both methods. The overall correlation of all double positive and double negative samples was very good (R2=0.93). If risk group stratification based on NGS results would be performed, 8 patients would be classified as intermediate risk (IR) instead of standard risk (SR) (one of whom relapsed) and 8 patients as SR instead of IR. One patient would be relocated from IR to slow early responders (SER) group, and two patients from SER to IR (one of them relapsed). One patient who relapsed would be classified as high risk (HR) instead of SER. All 5 patients who were MRD negative at d15 by NGS remained MRD negative in later timepoints and none of them relapsed. Discussion We present a cost-effective and widely adoptable NGS-based method that provides clinically relevant results in childhood ALL. NGS has a great potential to reduce the laboriousness associated with patient-specific qPCR analysis and to speed up the process of MRD detection. The sensitivity of both methods is comparable when ~500ng of DNA is used. The majority of the differences were in the samples with MRD levels below 1E-4 and most of treatment stratification changes occurred between SR/IR. However, the different stratification mostly concerned patients who did not relapse. The sensitivity of NGS could be improved if more DNA was analyzed. However, the benefit of increased MRD sensitivity is questionable due to possible overtreatment of patients with very low MRD loads after induction treatment. “Online” identification of d15 MRD negative patients reported as having an excellent prognosis previously is possible only by NGS, because optimization of patient-specific qPCR takes several weeks. The addition of 10% polyclonal DNA seems to solve the problem of MRD overestimation by NGS in samples with B-cell aplasia. At present, the main drawback of the Ig/TCR-exploring NGS methods is lack of standardization both in the experimental setting and in data analysis. Therefore, recently a European network, the EuroClonality NGS Consortium, has been formed to standardize the whole workflow of analytics, pre-analytics and bioinformatics not only for MRD quantification but also for clonality assessment in lymphoid neoplasms and for repertoire analysis. Supported by IGA NT14343, IGA NT12397, IGA NT13462-4 and GAUK 394214. Disclosures No relevant conflicts of interest to declare.

Description: In acute myeloid leukemia (AML), aberrant expression and mutations of transcription factors have been correlated with disease outcome. In the present study, we performed expression and mutation screening of GATA2, which is an essential transcription factor for regulation of myeloid lineage determination, in de novo pediatric AML patients. GATA2 mutations were detected in 5 of 230 patients, representing a frequency of 2.2% overall and 9.8% in cytogenetically normal AML. GATA2 expression analysis demonstrated that in 155 of 237 diagnostic samples (65%), GATA2 expression was higher than in normal BM. In complete remission, normalization of GATA2 expression was observed, whereas GATA2 expression levels stayed high in patients with resistant disease. High GATA2 expression at diagnosis was an independent poor prognostic factor for overall survival (hazard ratio [HR] = 1.7, P = .045), event-free survival (HR = 2.1, P = .002), and disease-free survival (HR = 2.3, P = .004). The prognostic impact of GATA2 was particularly evident in specific AML subgroups. In patients with French-American-British M5 morphology, inv(16), or high WT1 expression, significant differences in survival were observed between patients with high versus normal GATA2 expression. We conclude that high GATA2 expression is a novel poor prognostic marker in pediatric AML, which may contribute to better risk-group stratification and risk-adapted therapy in the future.

Description: Abstract 2738 Poster Board II-714 L-Asparaginase (L-Asp) is an important component in the combined chemotherapy for childhood acute lymphoblastic leukemia (ALL). Administration of L-Asp leads to depletion of plasmatic asparagine and consequently causes loss of intracellular asparagine. As a non-essential amino acid, asparagine is synthesized from aspartate and glutamine by asparagine synthetase (ASNS). Primary ALL cells are believed to have low ASNS expression and therefore to be sensitive to asparagine depletion. Although increased ASNS level was shown to be connected with L-Asp resistance the exact relationship between ASNS expression and L-Asp sensitivity is not clear. We and others have previously shown TEL/AML1[+] ALL blasts express more ASNS mRNA than TEL/AML[-] do although primary TEL/AML[+] cell are in vitro more sensitive to treatment with L-Asp. Hutson et al (1997) showed that amino acid deprivation led to increased expression of ASNS on mRNA and protein level as well as to increased biological activity. On the other hand, Nan Su et al described negative correlation between L-Asp sensitivity and ASNS protein rather than mRNA levels. Therefore, in our studies we concentrated on protein expression of ASNS in patients' samples. So far, there has been no reproducible published data on ASNS protein detection by Western blot in primary patients' samples. Despite using 3 different antibodies and precise optimization we were not able to detect ASNS protein in patients' samples in contrast to cell lines. Transcripts' levels confirmed significantly lower (2 log) expression of ASNS in patients' leukemic cells compared to leukemic cell lines. Therefore, for further studies on gene and protein relation we had to rely on cell lines as a model. We detected ASNS gene expression and ASNS protein content in four ALL cell lines: REH (TEL/AML1[+]), UOCB6 (TEL/AML1[+]), NALM6 (TEL/PDGFRB[+]) and RS4;11 (MLL/AF4[+]). ASNS mRNA levels were in accord with sensitivity to L-Asp. UOCB6 as the most resistant cell line (IC50=0.04U/ml) had the highest expression of ASNS (normalized ASNS, nASNS=4.946), then NALM6 (IC50=0.01U/ml; nASNS=1.8), REH (IC50=0.6.10−4; nASNS=1.176) and RS4;11 (IC50

Description: Leukaemias with MLL gene rearrangement are usually considered prognostically unfavourable and the clinical symptoms typically follow the translocation formation rapidly. MLL rearrangement is thus thought to be a major hit in leukaemogenesis that is either sufficient to cause the disease or it is a very strong and rapid inducer of the subsequent hit(s) required for the malignant transformation. We report an unusual presentation of secondary acute lymphoblastic leukaemia (sALL) with MLL rearrangement. Our patient was diagnosed originally with acute myeloid leukaemia (AML-M3) characterised by PML/RARα fusion and an internal tandem duplication of FLT3 (FLT3/ITD). After 30 months of complete remission of AML, she developed sALL with MLL/FOXO3A fusion gene. Bone marrow (BM) samples taken during AML therapy were analysed for the presence of these aberrations. Both the PML/RARα fusion and FLT3/ITD disappeared shortly after AML onset and did not reappear. However, FISH and quantitative RT-PCR showed the presence of the MLL/FOXO3A fusion 20 months before the diagnosis of sALL, present in 10–90% of BM cells. Morphological examination showed no blast infiltration of the BM at this time. Experiments combining FISH and morphology confirmed the presence of an MLL rearrangement in myeloid as well as lymphoid cells, indicating that the fusion arose in a multipotent progenitor. In order to identify potential secondary genetic events precipitating sALL in this patient, we used Affymetrix 50K single nucleotide polymorphism (SNP) array analysis on DNA from the diagnostic sALL sample versus the "preleukaemic" (remission AML) sample taken 16 months before. This analysis revealed a 10 Mb amplification on 19q13.32 in the sALL sample, not present in the preleukaemic sample: this was confirmed by FISH with a BAC from the amplified region. A difference between the pre-leukaemic and leukaemic cells is also demonstrated by the incomplete rearrangement of IgH gene (DH1/JH) present only at the diagnosis of sALL. There are about 450 genes in the amplified region on 19q and several of them might be involved in deregulation of the preleukaemic cell if overrepresented (e.g. FLT3 ligand, interleukin 11, Ras interacting protein 1, Stem cell growth factor, Aurora C). The long latency period prior to the onset of the secondary leukaemia in our case resembles the mouse model of MLL/FOXO3A. However, in contrast to the animal model and also to the previous reports of MLL/FOXO3A patients (2 cases described so far, both secondary AMLs after Hodgkin’s disease), our child developed leukaemia from the lymphoid lineage. Taken together, these results indicate that the MLL/FOXO3A fusion alone is not sufficient to cause leukaemia and that second hit is required to the onset of the disease. A responsible gene is possibly located on the telomeric part of the 19q.

Description: Pediatric acute myeloid leukemia (AML) is a heterogeneous disease and 30-40% of the patients still die. Prognosis is dependent on relevant genetic aberrations. Although many driving genetic alterations causing AML have been defined, in ~20% of the pediatric AML patients the oncogenic events remain unidentified. The ETS-Variant gene 6 (ETV6) encodes a transcription factor that functions as a tumor suppressor gene and is required for proper hematopoiesis in the bone marrow niche. Point mutations, deletions and translocations can lead to silencing of the gene, resulting in loss of transcriptional repression activity. ETV6 aberrations strongly associate with leukemia. In pediatric B-cell precursor acute lymphoblastic leukemia, translocation ETV6/RUNX1 occurs in ~25% of cases. Mutations in ETV6 are identified in ~25% of early immature T-cell ALL and also reported as event in adult AML (Van Vlierberghe et al, J Exp Med 2011; Barjesteh van Waalwijk van Doorn-Khosrovani et al, Oncogene 2005). We previously reported that pediatric AML patients can be divided in three clusters based on HOX-expression; (1) low HOXA/B expression, (2) high HOXA and low HOXB expression, and (3) high HOXA/B expression, and identified new repetitive genetic abnormalities in the third cluster, especially in NUP98. Cluster 1 is mainly represented by core-binding factor (CBF) AML, but in ~20% of these cases we did not find specific genetic abnormalities. Helton et al presented ETV6 aberrations in pediatric CBF-AML at ASH 2011, identified with whole genome sequencing, and with poor clinical outcome. We hypothesized that ETV6 aberrations might reduce the number of patients without known driving abnormality, especially in the low HOXA/B cluster. We screened a large representative de novo pediatric AML cohort for ETV6 mutations in exons 2-8 with direct sequencing, for ETV6 deletions by multiplex ligation-dependent probe amplification and for ETV6 translocations using split signal FISH, and analyzed outcome. In a well-characterized de novo pediatric AML cases with available gene-expression data, 6/275 (2.2%) patients had mutations affecting the predicted amino acid sequence of ETV6 and one had a silent mutation, 4/259 (1.5%) had an ETV6 deletion and 6/65 (9.2%) patients an MNX1/ETV6 translocation. Additionally, we identified 3 cases with a positive split signal FISH suggestive of a break in which ETV6 is involved, and a similar gene expression profile was found in these three cases. The aberrations of ETV6 were seen in patients of all three HOX-groups; n=9, n=6 and n=4 for cluster 1, 2 and 3 respectively. In patients with an ETV6 mutation (n=6) or deletion (n=4) 13 and 38 genes, respectively, were significantly up-regulated, including CLDN5,DPEP1 and BIRC7. This is consistent with the up-regulated genes in functional studies silencing ETV6 in LOUCY cells (Van Vlierberghe et al, J Exp Med 2011). High expression of BIRC7 has been associated with poor prognosis in adult acute leukemia (El-Mesallamy et al, Leuk Res 2011). The median age of patients with an ETV6-mutation or deletion (n=10) was 11.3 years (range 4.0-15.3) and 40% were female. Median WBC was significantly lower (15.1x109/L vs 47.0x109/L, p

Description: Key Points NUP98/KDM5A, CBFA2T3/GLIS2, KMT2A-rearrangements, and monosomy 7 are associated with poor outcome; RBM15/MKL1 and others fare better. Screening for NUP98/KDM5A, RBM15/MKL1, CBFA2T3/GLIS2, and KMT2A rearrangements combined with conventional karyotyping is advisable.

Description: Abstract 756FN2 Bone marrow (BM) aspiration at the end of induction therapy plays a crucial role for the evaluation of remission and the minimal residual disease (MRD), both critical for treatment stratification in modern treatment protocols for paediatric acute lymphoblastic leukaemia (ALL). However, the aspiration is repeated in 15–20% of patients, either due to non-representative morphology or to insufficient material needed for MRD analysis. We prospectively analysed 320 paediatric ALL patients treated according to ALL-BFM 2000 (n=301) or ALL IC-BFM 2002 (n=19) protocols with repeated BM aspiration at the end of induction therapy, on treatment day 33. Fourteen patients had more than one re-puncture. The median follow-up was 69 months, 45 (14%) patients had an event (relapse/death). The cause for the repeated BM aspiration was non-representative morphology (32%), insufficient material for MRD analysis (33%) or both (35% cases). In order to evaluate prognostic significance of the re-punctures and to determine which of the repeated samples should be used for the final treatment stratification we analysed MRD levels and MRD stratification, morphology, leukocyte count (WBC) and the length of treatment delay caused by waiting for the repeated aspiration. MRD data were collected and interpreted according to the EuroMRD guidelines in one central reference laboratory per each participating country. Morphology was evaluated centrally using an own scoring system (with a max value of 26 points). Treatment delay between the original and the last aspiration was one-third longer in patients with subsequent event compared to patients remaining in complete remission (CR) (median 8 (range 2 – 21) vs. 6 (1 - 28) days, respectively; p=0.020). Patients with a subsequent event had significantly higher WBC at the time of the last repeated BM aspiration, compared to patients without event (p=0.019), while there was no difference relative to the original aspiration (p=0.9). Analysis of the BM morphology at the original aspiration showed no significant difference between patients with an event vs. those in CR. However, the repeated aspiration of patients with a subsequent event had significantly better morphology (median 18.5/26 vs. 15/26 points, p=0.0012) mainly due to higher cellularity (p=0.003) and number of megakaryocytes (p=0.048). MRD levels were identical or decreased in 88% and increased in 12% of cases comparing the original aspiration to the repeated aspiration. In 63 patients (20%) the different MRD levels would lead to different treatment stratification. Higher MRD was associated with treatment failure; the best predictive values for subsequent event were obtained using the MRD results of the original aspiration (p=3.1e-07) or the highest of the detected MRD levels (p=6.0e-07). The last aspiration before proceeding with treatment had the lowest, though still a highly significant predictive value (p=8.6e-06). Corresponding results are obtained when MRD levels are substituted by final MRD risk stratification into standard, medium or high risk (p

Description: FC is still not employed in MRD based treatment protocols. One problem is lack of standardization suitable for prospective trials involving multiple clinical centers and FC laboratories. Therefore, we established a MiniMini Project, an international collateral study within the ALL IC BFM 2002 treatment protocol for childhood acute lymphoblastic leukemia (ALL). The MiniMini Project provides a mainframe of minimal panel of monoclonal antibody combinations to evaluate MRD by FC. Patients (pts) are stratified according non-MRD criteria (prednisone response at day 8 in peripheral blood (PB), percentage of blasts at day 15 and day 33 in bone marrow (BM), leukocytosis, and age at diagnosis and presence of BCR/ABL or MLL/AF4 fusions). Identical immunophenotypic populations are reported in all pts regardless presenting phenotype. Each laboratory investigates at least 2 pts with B lineage ALL by the T ALL combinations and vice versa. These “cross-lineage controls” together with data on subpopulations that are negative at diagnosis were used to set the specificity cutoff values at each time point (diagnosis, day 8 BM and PB, day 15, day 33 day 52 BM). MRD levels obtained by Ig/TCR rearrangements RQ-PCR in 32 pts (24 pts BCP ALL, 8 pts T ALL) were used to define specificity thresholds. 185 pts were investigated in the participating laboratories. We used data from first Czech cohort of pts (92 pts in total, 16 pts T lineage, 74 pts B lineage, in standard risk group (SRG), n=36, IRG, n=40 and HRG, n=16) in whom clinical data as well as standard FC analysis results were available. We compared morphological percentage of blasts (used for stratification) to a level of residual disease by FC. There was high concordance in SRG of both methods, except 1 patient redirected into IRG group (M3 BM vs. only 14% of blasts by FC). In IRG, concordance was in 92.5% of pts, 3 pts should be placed in HRG group according FC. 98.9% of pts morphologically in complete remission at day 33 were confirmed by FC. Although FC data confirm a significant difference between PGR and PPR in PB specimens at day 8 (p=0.0014), there is an overlap in percentage of leukemic cells between these categories. In total, MRD level above 0.1% was observed in BM of 100, 99, 84, 32 and 3.5 % pts in days 0, 8, 15, 33 and 52, respectively and in PB of 95% pts at day 8. Our first results show feasibility of FC standardization. The choice of subpopulations and the cutoff points will be validated in an independent cohort within the same Project.

Description: Minimal residual disease (MRD) testing based on a unique Ig/TCR gene rearrangement pattern of each patient’s leukaemia turned out to be an independent tool to determine treatment response and the risk of relapse in paediatric acute lymphoblastic leukaemia (ALL). Since 07/2000, MRD information at week 5 and 12 of therapy has been used for stratification in ALL-BFM 2000 trial. In parallel, ALL IC-BFM 2002 has been designed by the International-BFM Group to test the morphological assessment of the early treatment response. Patients are stratified according to the blast proportion in peripheral blood (PB) at day 8 and in bone marrow (BM) at day 15 and 33 of therapy, age, initial WBC and the presence of BCR/ABL and MLL/AF4 fusion. One of the goals of the study is the comparison of this risk group assessment to the MRD-based criteria used in ALL-BFM 2000. In the Czech Republic, 73 patients were treated according to ALL IC-BFM 2002 protocol from 11/2002 to 12/2003, 29 in the standard-risk (SR), 35 in the intermediate-risk (IR) and 9 in the high-risk (HR) group. The SR, HR and all T-cell ALL patients were examined for clonal Ig/TCR rearrangements. RQ-PCR patient-specific systems were designed for each of these patients according to the ESG-MRD-ALL criteria. For 39 of the 40 patients tested (97.5%) at least one target with minimal sensitivity of 10(−4) was identified. MRD was evaluated in BM samples from 34 patients at several time points inclusive of the mandatory 5 and 12 week ones. Simultaneously the PB specimens of the T-ALL patients were tested. In total, 205 BM and 64 PB specimens were included. In 7 patients of 24 in the SR group, MRD positivity at week 5 and/or at week 12 was observed (ranging between 9.7x10(−4) and 1.5x10(−2)), thus identifying patients who would not qualify to the MRD-based SR group in ALL-BFM 2000 despite the identical induction regimen. In T-ALL patients, PB-MRD levels paralleled those in BM. MRD results showed no separation of MRD levels between IR- and HR-stratified T-ALL patients. These preliminary findings reveal a significant difference between the stratification results of ALL IC-BFM 2002 and ALL-BFM 2000. A fast response as measured by the morphology criterion (M1 or M2 bone marrow at day 15) together with other low-risk features does not necessarily correspond with rapid MRD clearance. The complete analysis of MRD is planned for the international consortium participating in the ALL IC-BFM 2002 protocol.