Search

Molecular routine diagnostics for BCR-ABL1-negative myeloproliferative neoplasms (MPN) currently focusses on mutations in JAK2, CALR and MPL. In recent years, recurrent mutations in MPNs have been identified in several other genes. We here present the validation of a next generation sequencing (NGS)-based 28-gene panel and its use in MPN. We analysed the mutation status of 28 genes in 100 MPN patients [40 essential thrombocythaemia (ET), 30 primary myelofibrosis (PMF), 30 polycythaemia vera (PV)] and found two or more mutated genes in 53 patients. Moreover, significantly more mutated splicing genes (SF3B1, SRSF2 and U2AF1) were present in PMF (0·60 mutated genes/patient) compared to ET (0·15) while no mutations in splicing genes were found in PV. Additionally, chromatin modification genes (ASXL1 and EZH2) were frequently mutated in PMF patients (0·50) and, to a significantly lesser extent, in ET (0·13) and PV (0·07). Contrarily, DNA methylation genes (DNMT3A, IDH1, IDH2 and TET2) were mutated most often in PV (0·5) and less frequently in ET (0·23) and PMF (0·20), but without reaching statistical significance. Our results demonstrate the feasibility and utility of NGS-based panel diagnostics for MPN. With 53% of the patients bearing two or more mutated genes, their prognostic relevance needs further studies.

Chronic Lymphocytic Leukaemia (CLL) is a heterogeneous disease with a clinical course dependent on cytogenetic features. However, in 15-20% of cases both chromosome banding and fluorescence in situ hybridisation analyses do not show any kind of abnormality. With the aim to identify dependable molecular prognostic factors in this subgroup, we performed a comprehensive analysis on 171 patients including genomic arrays (comparative genomic hybridisation and single nucleotide polymorphism), immunoglobulin heavy chain variable region genes (IGHV) status, flow cytometry and targeted sequencing. Genomic arrays detected 73 aberrations in 39 patients (23%). Most frequently, patients had 1 aberration (25/171; 15%), while 14 patients (8%) had at least 2 aberrations. IGHV status was unmutated in 53/171 (31%) patients. SF3B1 was the most frequently mutated gene (26/171 patients; 15%), followed by NOTCH1 (15/171; 9%). At univariate analysis, an adverse impact on time to treatment (TTT) was evident for SF3B1 mutations, higher white blood cell count, higher CLL cells percentage by flow cytometry, CD38 positivity, IGHV unmutated status and at least 2 genomic array abnormalities. Of these, SF3B1 mutations, CLL cells percentage, IGHV unmutated status and number of genomic array aberrations maintained their impact in multivariate analysis. In conclusion, by integrating genomic and molecular data, we identified patients at higher risk for treatment need.

Correct action upon re-emergence of minimal residual disease in acute myeloid leukaemia (AML) patients has not yet been established. The applicability of demethylating agents and use of allogeneic stem cell transplantation will be dependent on pre-relapse AML growth rates. We here delineate molecular growth kinetics of AML harbouring MLL partial tandem duplication (MLL-PTD; 37 cases) compared to those harbouring MLL translocations (43 cases). The kinetics of MLL-PTD relapses was both significantly slower than those of MLL translocation positive ones (median doubling time: MLL-PTD: 24 d, MLL-translocations: 12 d, P = 0·015, Wilcoxon rank sum test), and displayed greater variation depending on additional mutations. Thus, MLL-PTD+ cases with additional RUNX1 mutations or FLT3-internal tandem duplication relapsed significantly faster than cases without one of those two mutations (Wilcoxon rank sum test, P = 0·042). As rapid relapses occurred in all MLL subgroups, frequent sampling are necessary to obtain acceptable relapse detection rates and times from molecular relapse to haematological relapse (blood sampling every second month: MLL-PTD: 75%/50 d; MLL translocations: 85%/25 d). In conclusion, in this cohort relapse kinetics is heavily dependent on AML subtype as well as additional genetic aberrations, with possibly great consequences for the rational choice of pre-emptive therapies.

In patients with myelodysplastic syndromes (MDS), sole 20q deletion [del(20q)] is a recurrent favourable abnormality. We studied additional molecular and cytogenetic lesions and their prognostic impact in 305 MDS patients with del(20q) (229 males/76 females; 29-90 years). All patients were investigated by cytomorphology and chromosome banding analysis (CBA), subsets by fluorescence in situ hybridization, molecular mutation screening, and array comparative genomic hybridization (aCGH). By aCGH (n = 30), the minimal common deleted region (CDR) was flanked by PTPRT (20q13·11) and EYA2 (20q13·12). 210 (68·9%) patients had 'early MDS' without blast increase, 95 (31·1%) 'advanced' MDS with blast increase (5-19%). Additional chromosomal abnormalities (ACAs) were detected in 88/305 (28·9%) patients. Patients with advanced MDS more frequently had ACAs (P = 0·003) and had a higher mean number of ACAs (P = 0·020) and of molecular mutations (P = 0·060). Spliceosome mutations were frequent (U2AF1: n = 31/155; 20·0%; SRSF2: n = 31/159; 19·5%; SF3B1mut: n = 8/159; 5·0%). ASXL1mut (25/153; 16·3%) were associated with advanced MDS (P = 0·001). Presence of ≥3 ACAs (P = 0·003) and ASXL1mut (P = 0·002) were associated with worse 2-year survival. In conclusion, the cytogenetic subgroup of MDS with del(20q) has a good prognosis but may be further subclassified by additional cytogenetic and molecular lesions. U2AF1mut is overrepresented in MDS with del(20q), and ASXL1mut is prognostically adverse.

Multiparameter flow cytometry (MFC) identifies rare cases of biclonal disease in chronic lymphocytic leukaemia (CLL). By MFC, we identified 76 patients with biclonal disease in a cohort of 5523 CLL patients (1·4%). Fluorescence in situ hybridization and chromosome banding analysis revealed five and six cases, respectively, with two different cytogenetic aberrations due to clonal evolution. Two different B-cell receptor rearrangements and IGHV subtypes were more frequent in biclonal than in monoclonal CLL by MFC (37·1% vs. 2·7%; P < 0·001). Patients with biclonal CLL by MFC showed a trend to a shorter time to treatment than monoclonal CLL (P = 0·080).

Summary Acute myeloid leukaemia (AML) with CEBPA mutations is listed as a provisional entity in the current World Health Organization classification. A difference in clinical outcome between single- (sm) and double-mutated (dm) cases has been reported, whereupon CEBPAdm cases were shown to be associated with better overall survival (OS). The occurrence and prognostic impact of concomitant molecular mutations in addition to CEBPAdm has not been assessed until now with exception of GATA2 mutations. Here, we investigated a cohort of 95 AML CEBPAdm cases for concomitant mutations. TET2 was found to be most frequently mutated (34·0%) gene, followed by GATA2 (21·0%), WT1 (13·7%), DNMT3A (9·6%), ASXL1 (9·5%), NRAS (8·4%), KRAS (3·2%), IDH1/2 (6·3%), FLT3-internal tandem duplication (6·3%), FLT3-tyrosine kinase domain (2·1%), NPM1 (2·1%), and RUNX1 (1/94). Patients harbouring additional mutations in the TET2 gene showed significantly worse OS than TET2 wild-type cases (P = 0·035), whereas GATA2-mutated patients showed improved OS (P = 0·032). Serial analyses were performed for 39 CEBPAdm cases with concomitant mutations. Here, we observed that CEBPA mutations present the primary pathogenetic event in the majority of cases (76·9%). Further, a distinct gene expression profile (GEP) was confirmed for CEBPAdm versus CEBPAsm or CEBPA wild-type cases while no significant changes in GEP were observed related to additional mutations within the CEBPAdm AML.

High BAALC gene expression has been associated with poor prognosis in cytogenetically normal acute myeloid leukaemia (CN-AML) and has been suggested as a suitable marker for assessing minimal residual disease (MRD). The purpose of this study was to substantiate these findings by the analysis of a large data set of 632 diagnostic and follow-up samples in 142 intensively treated CN-AML patients. Paired diagnostic/relapse samples of 35 patients revealed stable high BAALC expression in 89%, irrespective of a high proportion of clonal evolution found in 49% of these cases. High BAALC expression, both directly after induction chemotherapy and within 3-6 months after induction chemotherapy, correlated significantly with shorter event-free survival and overall survival. Moreover, 8 of 10 patients displaying high BAALC expression levels after completion of induction therapy as well as 5 of 5 patients exhibiting high BAALC expression levels within 3-6 months after induction chemotherapy experienced relapse with a median of 197 and 101 days, respectively, from sampling to relapse. Thus, BAALC expression-based MRD detection during therapy may be considered a strategy to identify patients at high risk of relapse.

Summary The World Health Organization classification uses a cut-off point of 5·0 × 109/l cells with a chronic lymphocytic leukaemia (CLL)-phenotype in peripheral blood to discriminate between monoclonal B-lymphocytosis (MBL) and B-CLL. This study analysed 298 MBL patients by multi-parameter flow cytometry, chromosome banding analysis (CBA)/fluorescence in situ hybridization (FISH), and IGHV mutation status and compared them with 356 CLL patients. In MBL, CBA more frequently revealed a normal karyotype and FISH identified less frequently del(6q), del(13q) (as sole alterations), and del(17)(p13). Within the MBL cohort, a shorter time to treatment (TTT) was found for ZAP-70-positivity, 14q32/IGH-translocations (CBA), del(11)(q22·3) (FISH) and unmutated IGHV status. Higher CD38 and ZAP-70 expression, del(11)(q22·3) (FISH), trisomy 12 (FISH), and 14q32/IGH-translocations (CBA) were correlated with a shorter TTT in the combined cohort (MBL + CLL); a sole del(13)(q14) (FISH) correlated with longer TTT. Regarding overall survival, unmutated IGHV status and ‘other’ alterations (CBA) had an adverse impact. There was no correlation between the concentration of CLL-cells and TTT or overall survival. Multivariate analysis confirmed a negative impact on TTT for del(11)(q22·3)/ATM, trisomy 12 (both by FISH), and 14q32/IGH-translocations by CBA. These data emphasize a close relationship between MBL and CLL regarding clinically relevant parameters and provide no evidence to strictly separate these entities by a distinct threshold of clonal B-cells.

TET2 mutations have been identified in 10-25% of patients with myeloid malignancies, but TET2 gene copy number alterations remain to be evaluated. We performed interphase and metaphase fluorescence in situ hybridization (FISH), using newly designed probes that span TET2, in 893 patients with acute myeloid leukaemia (AML) and chronic myeloid malignancies and validated the new assay by single nucleotide polymorphism arrays. Next-generation sequencing for molecular TET2 mutations was performed in 37/50 del(TET2) patients. We identified TET2 deletions in 50/893 patients (26 males, 24 females; 44-87 years) resulting in a 5.6% frequency [22/425 AML (5.2%), 15/217 chronic myelomonocytic leukaemia (CMML; 6.9%), 9/188 myelodysplastic syndromes (4.8%), 4/63 myeloproliferative neoplasms (6.3%)]. Cytogenetic alterations (mostly complex) were detected in 35/50 (70.0%) of del(TET2) cases. TET2 deletions were cytogenetically cryptic in 25/50 (50.0%). Sequencing detected TET2mut in 19/37 (51.4%) del(TET2) cases investigated, predominantly CMML (10/14; 71.4%). In de novo AML with intermediate cytogenetics, presence of any TET2 alteration was related to worse median overall survival (347 d vs. not reached; P = 0.052) and event-free survival (164 vs. 457 d; P = 0.024). JAK2(V617F) was found in 6/18 (33.3%) del(TET2) cases analysed. Thus, TET2 haploinsufficiency recurrently occurs in myeloid malignancies, frequently combined with TET2 mutations on the remaining allele. The prognostic impact of del(TET2) in myeloid malignancies should be further evaluated.

In AML, cooperation of mutations suppressing differentiation ('class-II-mutations') with 'class-I-mutations' increasing cell proliferation is frequent. In rare cases of myeloid malignancies, the BCR-ABL1 fusion was reported to cooperate as class-I-mutation with class-II-mutations, but most cases had to be classified as blast phase of chronic myeloid leukaemia (CML). We identified five cases of Philadelphia positive subclones in AML occurring in coincidence with other genetic lesions: 1:220 patients with inv(16)/CBFB-MYH11 (0·5%), 2:272 AML cases with t(8;21)/RUNX1-RUNX1T1 (0·7%), 1:1029 NPM1-mutated AML (0·1%), and one patient with s-AML following MDS with a 5q-deletion. Four patients had m-BCR (e1a2) BCR-ABL1 transcripts; one case only had an M-BCR (b3a2) breakpoint. These cases allow some interesting conclusions: The BCR-ABL1 rearrangement apparently can cooperate with the NPM1 mutation similar to other class-I-mutations. The identification of Philadelphia positive subclones in

14q-deletions have been repeatedly described in mature B-cell neoplasms, but not yet characterized in a larger cohort. Based on chromosome banding analysis, the present study identified 47 del(14q) cases in 3054 mature B-cell neoplasms (1·5%) (chronic lymphocytic leukaemia [CLL]: 1·9%; CLL/prolymphocytic leukaemia [PL]: 9·0%; others: 0·2%). Interphase fluorescence in situ hybridization was performed with probes for 14q22.1, 14q24.1, 14q32.33, and IGH@ (14q32.3). The del(14q) had heterogeneous size but showed a breakpoint cluster at the centromeric site in 14q24.1 (62% of cases). At the telomeric side, the most frequent breakpoint was within the IGH@ locus (14q32.3) between IGH@ 3'-flanking and IGHV (IgVH) probes (45%). In 16 cases (34%), breakpoints occurred within 14q24.1 and 14q32.3. Eighty-one percent of del(14q) cases showed 1-3 additional cytogenetic alterations (in 45%, +12), and 56% were IGHV-unmutated. In all cases (16/16) with breakpoints in 14q24.1 and 14q32.3, a B-CLL immunophenotype was found. Clinical follow-up in 32 del(14q) patients was compared to 383 CLL and CLL/PL patients without del(14q). While 3-year-overall survival did not differ significantly, time to treatment was significantly shorter in the del(14q) cohort (21·0 months vs. 80·1 months, P = 0·015). In conclusion, the del(14q) is a rare recurrent alteration in diverse mature B-cell neoplasms, shows variable size but distinct clustering of breakpoints, and is associated with short time to treatment.

Summary. The French–American–British (FAB) classification and the new World Health Organization (WHO) classification distinguish acute monoblastic leukaemia (AML M5a) from acute monocytic leukaemia (AML M5b). Not much is known about the underlying genetic differences leading to these clearly different phenotypes. We analysed 58 patients with de novo AML M5a and 66 patients with denovo AML M5b in comparison with a whole group of 1603 de novo AML. An aberrant karyotype was found in 75·9% of AML M5a but in only 28·8% of M5b (P

The biological mechanisms responsible for the association of specific karyotypes with prognosis in acute myeloid leukaemia (AML) remain largely unclear. A prospective study was performed to evaluate how far cytogenetically defined prognostic subgroups of AML differ in their proliferative activity as a potential mechanism for differential sensitivities to S-phase-specific induction chemotherapy comprising cytosine arabinoside (AraC). One hundred and eighty-seven patients with de novo AML were included in the study; 25 patients with a favourable [inv(16), t(8;21), t(15;17)] karyotype, 99 with a normal karyotype, 29 with an unfavourable karyotype (-5, 5q-, -7, 7q-, complex aberrations) and 34 with cytogenetic aberrations of unknown prognostic significance (all others). The favourable group demonstrated the highest ex vivo proliferative activity (PA) (3.41 pmol/105 cells), significantly (P = 0.02) exceeding the unfavourable group with the lowest PA (0.72) and the group with a normal karyotype (1.06) or with karyotype of unknown significance (1.05) that both demonstrated an intermediate PA. Samples with a high PA (> median of the whole group) were more likely to produce interleukin 3, granulocyte macrophage colony-stimulating factor (GM-CSF), granulocyte CSF (G-CSF) (56%, 43% and 50%) than cells with a low PA (33%, 36% and 36%; n.s.). The effect of priming by exogenous GM-CSF or G-CSF was significantly more pronounced in samples with a low PA than in rapidly proliferating samples (P < 0.01). For the whole group, a high PA was closely associated with an increased incorporation of AraC triphosphate (AraCTP) into the DNA (P < 0.0001). Clinically, a high PA was associated with a better complete remission (CR) rate in the normal (95% versus 62%) and the unfavourable group (75% versus 33%). The significant differences in proliferative activity between cytogenetic subgroups of AML are associated with increased cytosine arabinoside pharmacodynamics and constitute one potential mechanism for the different response of cytogenetic subgroups to AraC-based induction therapy.

The current study was initiated to explore the mechanisms underlying the previously demonstrated association between the proliferative activity of leukaemic blasts and the response to cytosine arabinoside (AraC)-based therapy in de novo acute myeloid leukaemia (AML). The activity of key enzymes of AraC metabolism–deoxycytidine kinase (DCK), cytidine deaminase (DCD) and polymerase α (PolyA) were determined in blast cells from 33 patients. In addition, formation and retention of intracellular levels of AraC triphosphate (AraCTP) and DNA incorporation of AraC were measured, as was the proliferative activity of leukaemic blasts by [3H]-TdR incorporation before and after stimulation with granulocyte–macrophage colony-stimulating factor (GM-CSF) or granulocyte CSF (G-CSF) for 48 h. AraC incorporation into the DNA (median 0·60 pmol/105 cells) was significantly related to the proliferative activity of AML blasts (r = 0·74, P

High-dose cytosine arabinoside (AraC)-containing regimens have shown the highest antileukaemic efficacy of all currently used regimens in the treatment of acute myeloid leukaemia (AML). This study aimed at increasing the antileukaemic potential of high-dose AraC by raising intracellular levels of AraC triphosphate (AraCTP), which is the mediator of cytotoxicity, via biochemical modulation by inhibitors of ribonucleotide reductase (RR) or haematopoietic growth factors (HGFs). Blasts from patients with de novo AML were analysed for their formation of AraCTP under high-dose AraC conditions (20 µm over 3 h) without prior modulation (n = 47) after a 2-h pre-exposure with fludarabine (50 µg/ml) (n = 40) or gemcitabine (30 ng/ml) (n = 40) and after a 48-h pre-exposure to granulocyte colony-stimulating-factor (G-CSF; 100 ng/ml) (n = 27) or granulocyte–macrophage colony-stimulating-factor (GM-CSF; 100 U/ml) (n = 28). Unmodulated formation of AraCTP (median 239·8 ng/107 cells) could not be increased via modulation by gemcitabine (232·4 ng/107 cells) or fludarabine (247·8 ng/107 cells). The lack of effect of RR inhibitors was also observed for all other known metabolites of AraC [Ara-cytosine monophosphate (CMP), Ara-cytosine diphosphate (CDP), AraCDP-choline, Ara-uridine monophosphate (UMP), Ara-uridine diphosphate (UDP) and Ara-uridine triphosphate (UTP)]. In contrast, pre-exposure to HGFs led to significant increases in AraCTP formation (G-CSF 556·0 ng/107 cells, 2·31-fold increase, P

Different metabolites of cytosine arabinoside (AraC) contribute to its cytotoxicity including incorporation of AraCTP into DNA, the incorporation of AraUMP into RNA, inhibition of polymerase alpha and beta (AraCMP/CTP), an impairment of repair mechanisms (AraCTP), alterations of phospholipid metabolism (AraCDP-choline), a direct membrane interaction (AraC), the alteration of signal transduction pathways (AraCDP-choline, AraCTP) and the induction of apoptosis. Since little is known about the potential differences in AraC metabolism between leukaemic blasts and normal haemopoietic progenitor cells, the formation of all known AraC metabolites was determined in bone marrow samples from patients with acute myeloid leukaemia (AML), healthy volunteers and specimens of cellsorted CD34+ haemopoietic stem cells. Highly significant differences were found for phosphorylated AraC metabolites (AraCMP, -CDP, -CTP, AraUMP) between AML and normal mononuclear bone marrow (ng/107 cells respectively 1.30 v 2.66; 2.65 v 7.50; 33.68 v 99.0; 1.18 v 5.70). The highest differences were found for formation of AraCDP-choline (3.75 v 12.86) which might be relevant for the high efficacy of high-dose AraC regimens. In contrast, no differences were found in the deamination product AraU (2.01 v 2.91). Only minute amounts of phosphorylated AraU derivatives were detected, providing an explanation for the lacking contribution of AraU to cytosine arabinoside cytotoxicity. Results in normal CD34+ haemopoietic stem cells did not differ significantly from normal bone marrow mononuclear cells and therefore justify their use as a surrogate in determining AraC-induced haematotoxicity. These data suggest a metabolic basis for the relative selectivity of AraC cytotoxicity for AML blasts and provide a means to determine the role of different metabolites and their related mechanism of action for overall AraC cytotoxicity.

Acute myeloid leukaemia (AML) with CEBPA mutations is listed as a provisional entity in the current World Health Organization classification. A difference in clinical outcome between single- (sm) and double-mutated (dm) cases has been reported, whereupon CEBPAdm cases were shown to be associated with better overall survival (OS). The occurrence and prognostic impact of concomitant molecular mutations in addition to CEBPAdm has not been assessed until now with exception of GATA2 mutations. Here, we investigated a cohort of 95 AML CEBPAdm cases for concomitant mutations. TET2 was found to be most frequently mutated (34·0%) gene, followed by GATA2 (21·0%), WT1 (13·7%), DNMT3A (9·6%), ASXL1 (9·5%), NRAS (8·4%), KRAS (3·2%), IDH1/2 (6·3%), FLT3-internal tandem duplication (6·3%), FLT3-tyrosine kinase domain (2·1%), NPM1 (2·1%), and RUNX1 (1/94). Patients harbouring additional mutations in the TET2 gene showed significantly worse OS than TET2 wild-type cases (P = 0·035), whereas GATA2-mutated patients showed improved OS (P = 0·032). Serial analyses were performed for 39 CEBPAdm cases with concomitant mutations. Here, we observed that CEBPA mutations present the primary pathogenetic event in the majority of cases (76·9%). Further, a distinct gene expression profile (GEP) was confirmed for CEBPAdm versus CEBPAsm or CEBPA wild-type cases while no significant changes in GEP were observed related to additional mutations within the CEBPAdm AML.

The World Health Organization classification uses a cut-off point of 5·0 × 10(9)/l cells with a chronic lymphocytic leukaemia (CLL)-phenotype in peripheral blood to discriminate between monoclonal B-lymphocytosis (MBL) and B-CLL. This study analysed 298 MBL patients by multi-parameter flow cytometry, chromosome banding analysis (CBA)/fluorescence in situ hybridization (FISH), and IGHV mutation status and compared them with 356 CLL patients. In MBL, CBA more frequently revealed a normal karyotype and FISH identified less frequently del(6q), del(13q) (as sole alterations), and del(17)(p13). Within the MBL cohort, a shorter time to treatment (TTT) was found for ZAP-70-positivity, 14q32/IGH-translocations (CBA), del(11)(q22·3) (FISH) and unmutated IGHV status. Higher CD38 and ZAP-70 expression, del(11)(q22·3) (FISH), trisomy 12 (FISH), and 14q32/IGH-translocations (CBA) were correlated with a shorter TTT in the combined cohort (MBL + CLL); a sole del(13)(q14) (FISH) correlated with longer TTT. Regarding overall survival, unmutated IGHV status and 'other' alterations (CBA) had an adverse impact. There was no correlation between the concentration of CLL-cells and TTT or overall survival. Multivariate analysis confirmed a negative impact on TTT for del(11)(q22·3)/ATM, trisomy 12 (both by FISH), and 14q32/IGH-translocations by CBA. These data emphasize a close relationship between MBL and CLL regarding clinically relevant parameters and provide no evidence to strictly separate these entities by a distinct threshold of clonal B-cells.

In AML, cooperation of mutations suppressing differentiation ('class-II-mutations') with 'class-I-mutations' increasing cell proliferation is frequent. In rare cases of myeloid malignancies, the BCR-ABL1 fusion was reported to cooperate as class-I-mutation with class-II-mutations, but most cases had to be classified as blast phase of chronic myeloid leukaemia (CML). We identified five cases of Philadelphia positive subclones in AML occurring in coincidence with other genetic lesions: 1:220 patients with inv(16)/CBFB-MYH11 (0·5%), 2:272 AML cases with t(8;21)/RUNX1-RUNX1T1 (0·7%), 1:1029 NPM1-mutated AML (0·1%), and one patient with s-AML following MDS with a 5q-deletion. Four patients had m-BCR (e1a2) BCR-ABL1 transcripts; one case only had an M-BCR (b3a2) breakpoint. These cases allow some interesting conclusions: The BCR-ABL1 rearrangement apparently can cooperate with the NPM1 mutation similar to other class-I-mutations. The identification of Philadelphia positive subclones in1% of patients with CBF-leukaemias fits well with previous observations that most CBF-AML are accompanied by activating mutations in genes enhancing proliferation. Since we observed the occurrence of the Philadelphia positive subclones at diagnosis, at relapse, or throughout the disease, the time point of the emergence of Philadelphia subclones seems variable in AML. Clinical research should further concentrate on Philadelphia positive subclones in AML to assess the clinical impact.

14q-deletions have been repeatedly described in mature B-cell neoplasms, but not yet characterized in a larger cohort. Based on chromosome banding analysis, the present study identified 47 del(14q) cases in 3054 mature B-cell neoplasms (1·5%) (chronic lymphocytic leukaemia [CLL]: 1·9%; CLL/prolymphocytic leukaemia [PL]: 9·0%; others: 0·2%). Interphase fluorescence in situ hybridization was performed with probes for 14q22.1, 14q24.1, 14q32.33, and IGH@ (14q32.3). The del(14q) had heterogeneous size but showed a breakpoint cluster at the centromeric site in 14q24.1 (62% of cases). At the telomeric side, the most frequent breakpoint was within the IGH@ locus (14q32.3) between IGH@ 3'-flanking and IGHV (IgVH) probes (45%). In 16 cases (34%), breakpoints occurred within 14q24.1 and 14q32.3. Eighty-one percent of del(14q) cases showed 1-3 additional cytogenetic alterations (in 45%, +12), and 56% were IGHV-unmutated. In all cases (16/16) with breakpoints in 14q24.1 and 14q32.3, a B-CLL immunophenotype was found. Clinical follow-up in 32 del(14q) patients was compared to 383 CLL and CLL/PL patients without del(14q). While 3-year-overall survival did not differ significantly, time to treatment was significantly shorter in the del(14q) cohort (21·0 months vs. 80·1 months, P = 0·015). In conclusion, the del(14q) is a rare recurrent alteration in diverse mature B-cell neoplasms, shows variable size but distinct clustering of breakpoints, and is associated with short time to treatment.

A retrospective comparison of WT1 and BCR-ABL1 expression was performed in 40 imatinib-treated chronic myeloid leukaemia patients. The overall correlation of WT1 and BCR-ABL1 was low. In two patients WT1 expression was increasing despite very low BCR-ABL1 levels. As both revealed Ph-negative aberrant clones, a second independent cohort of 20 cases, all with Ph-negative clonal evolution, was analysed. High WT1 expression (5.0-177.0%) was detected in a case with +11 and in four of eight cases with +8, but not in cases with del(20q) or -Y. Thus, increasing WT1 levels in molecular responders may indicate Ph-negative clonal cytogenetic evolution during imatinib treatment.

The French-American-British (FAB) classification and the new World Health Organization (WHO) classification distinguish acute monoblastic leukaemia (AML M5a) from acute monocytic leukaemia (AML M5b). Not much is known about the underlying genetic differences leading to these clearly different phenotypes. We analysed 58 patients with de novo AML M5a and 66 patients with de novo AML M5b in comparison with a whole group of 1603 de novo AML. An aberrant karyotype was found in 75.9% of AML M5a but in only 28.8% of M5b (P0.0001) and in 54.7% of all other AML subtypes (P = 0.0015). 11q23/MLL aberrations were detected in 31% of M5a, 12.1% of M5b (P = 0.01) but only 1.3% of all other AML subtypes (P0.0001). Trisomy 8 as the sole cytogenetic aberration was found in 22.4% of M5a, but in only 3% of M5b and in 2.5% of all other AML subcategories (P0.0001). Although the frequency of the MLL-partial tandem duplication (MLL-PTD) did not differ between the three cohorts (1.7%, 4.5% and 6.1% respectively, NS), the detection of FLT3 length mutations (FLT3-LM) differed significantly. AML M5a showed a low frequency of only 6.9%, but 28.8% of M5b (P = 0.0014) and 23.5% of all other AML revealed a FLT3-LM. In conclusion, we demonstrated genetic, i.e. biological, differences between AML M5a and AML M5b and all other AML. Therefore, AML M5 should further be categorized as two different groups, as proposed by the WHO classification.