Search

To characterise the genetics of splenic marginal zone lymphoma (SMZL), we performed whole exome sequencing of 16 cases and identified novel recurrent inactivating mutations in Kruppel-like factor 2 (KLF2), a gene whose deficiency was previously shown to cause splenic marginal zone hyperplasia in mice. KLF2 mutation was found in 40 (42%) of 96 SMZLs, but rarely in other B-cell lymphomas. The majority of KLF2 mutations were frameshift indels or nonsense changes, with missense mutations clustered in the C-terminal zinc finger domains. Functional assays showed that these mutations inactivated the ability of KLF2 to suppress NF-κB activation by TLR, BCR, BAFFR and TNFR signalling. Further extensive investigations revealed common and distinct genetic changes between SMZL with and without KLF2 mutation. IGHV1-2 rearrangement and 7q deletion were primarily seen in SMZL with KLF2 mutation, while MYD88 and TP53 mutations were nearly exclusively found in those without KLF2 mutation. NOTCH2, TRAF3, TNFAIP3 and CARD11 mutations were observed in SMZL both with and without KLF2 mutation. Taken together, KLF2 mutation is the most common genetic change in SMZL and identifies a subset with a distinct genotype characterised by multi-genetic changes. These different genetic changes may deregulate various signalling pathways and generate cooperative oncogenic properties, thereby contributing to lymphomagenesis.

Histone deacetylases (HDACs) have been identified as therapeutic targets due to their regulatory function in chromatin structure and organization. Here, we analyzed the therapeutic effect of LBH589, a class I-II HDAC inhibitor, in acute lymphoblastic leukemia (ALL). In vitro, LBH589 induced dose-dependent antiproliferative and apoptotic effects, which were associated with increased H3 and H4 histone acetylation. Intravenous administration of LBH589 in immunodeficient BALB/c-RAG2(-/-)γc(-/-) mice in which human-derived T and B-ALL cell lines were injected induced a significant reduction in tumor growth. Using primary ALL cells, a xenograft model of human leukemia in BALB/c-RAG2(-/-)γc(-/-) mice was established, allowing continuous passages of transplanted cells to several mouse generations. Treatment of mice engrafted with T or B-ALL cells with LBH589 induced an in vivo increase in the acetylation of H3 and H4, which was accompanied with prolonged survival of LBH589-treated mice in comparison with those receiving vincristine and dexamethasone. Notably, the therapeutic efficacy of LBH589 was significantly enhanced in combination with vincristine and dexamethasone. Our results show the therapeutic activity of LBH589 in combination with standard chemotherapy in pre-clinical models of ALL and suggest that this combination may be of clinical value in the treatment of patients with ALL.

In B-cell chronic lymphocytic leukemia (B-CLL), somatic mutation of IgVH genes defines a subgroup with favorable prognosis, whereas the absence of IgVH mutations is correlated with a worse outcome. Mutations of the BCL-6 gene are also observed in a subset of B-CLL, but the clinical significance of this molecular alteration remains uncertain. We examined the distribution of IgVH and BCL-6 gene mutations in 95 well-characterized patients with Binet stage A B-CLL, and correlated them with clinical, laboratory, cytogenetic findings and disease progression. Mutations of the BCL-6 gene were observed only in cases harboring mutated IgVH. Unexpectedly, coexistence of IgVH and BCL-6 mutations was correlated with shorter treatment-free interval (TFI) compared to cases harboring only IgVH mutation (median, 55 months vs not reached; P=0.01), resembling the clinical course of unmutated IgVH cases (median TFI, 44 months). As expected, deletions of 17p13 (P53 locus) and 11q22 (ATM locus) were observed in cases with unmutated IgVH, except one patient who showed mutations of both IgVH and BCL-6. No other statistically significant differences were observed among the genetic subgroups. Our data indicate that BCL-6 mutations identify a subgroup of Binet stage A B-CLL patients with a high risk of progression despite the presence of mutated IgVH gene.

The BCL6 gene, which functions as a transcription repressor, is the target of multiple chromosomal translocations in non-Hodgkin's lymphomas (NHL). These translocations occur in the nontranslated region of the BCL6 gene, juxtaposing regulatory sequences of the diverse partner genes to the open reading frame of the BCL6 gene and thus are thought to deregulate BCL6 gene expression. The levels of expression of the BCL6 gene and protein have been demonstrated to predict the clinical outcome of diffuse large B-cell lymphomas. By contrast, the prognostic significance of BCL6 gene translocations is unclear. In this study we have sought an explanation for this apparent discrepancy. We examined tumors with a variety of different BCL6 translocations and therefore with a variety of potentially substituted promoters. We found no increase in total BCL6 mRNA levels in the NHL specimens harboring BCL6 gene translocation. Indeed, some of these tumors expressed relatively low quantities of the BCL6 mRNA. We also sought to determine whether BCL6 transcription occurs from the rearranged or from the normal untranslocated allele in these tumors. We demonstrate that lymphoma cell lines and majority of NHL tumor specimens expressed BCL6 mRNA predominantly from the rearranged allele that may come under the control of various partner gene promoters. However, few NHL tumors with BCL6 gene translocations expressed BCL6 mRNA equally from the rearranged and the nonrearranged alleles. Neither the nature of the substituted promoters nor the presence of activating mutations in the BCL6 regulatory sequences correlated with the allelic expression of the BCL6 gene in these tumors.

Cytogenetic abnormalities of band 11q23 have been found in more than 50% of infant leukemias regardless of the phenotype. Using probes for the MLL gene at 11q23, MLL rearrangements have been identified in 70-80% of all infant leukemias including virtually all of the cases with 11q23 translocations, as well as cases with apparently normal karyotypes. We reviewed the chromosomal pattern of 26 cases of infant leukemias (12 ALL, 12 AML, two AUL). Eleven had 11q23 translocations, five had other abnormalities, and 10 had a normal karyotype. To determine whether 11q23/MLL rearrangements were present in the leukemia cells of patients with a normal karyotype, we performed FISH and molecular studies of eight of these patients who had adequate material. Three were found to have 11q23/MLL abnormalities, two of them detected by FISH; one ALL case had a t(11;19) (q23;p13.3), and one AML case had a t(11;19) (q23;p13.1). Retrospective review confirmed the presence of the t(11;19) in a small percentage of poor quality metaphase cells in both cases. A rearrangement of the MLL gene was detected by Southern blot analysis of leukemic cells from a third patient with ALL; one cell with a deletion of 11q23 was found on karyotypic review. Therefore, in our series the actual incidence of 11q23 abnormalities in infant leukemias was 54% (14/26): 67% in ALL (8/12) and 50% in AML (6/12). Our findings suggest that most infant leukemias with apparently normal karyotypes that have a molecular rearrangement of the MLL gene are undetected subtle translocations.(ABSTRACT TRUNCATED AT 250 WORDS)

We report on the chromosomal pattern of 120 patients with childhood AML de novo. One hundred and fifteen patients (96%) had adequate samples for analysis; 98 (85%) of these showed clonal karyotypic abnormalities. They were classified into cytogenetic subgroups which were closely correlated with FAB subtypes: t(8;21) and M2 (n = 9); t(15;17) and M3 (n = 12); inv(16) and M4Eo (n = 9); t(9;11) and M5a (n = 10); t(11q23) other than t(9;11) and M4-M5 (n = 11); and t(1;22) and M7 (n = 4). In patients with -7/del(7q) (n = 6), leukemia was preceded by MDS in half of the cases, although they had diverse FAB subtypes. Thirty-seven patients had miscellaneous abnormalities. Despite a high CR rate, patients with t(8;21) had a very poor survival: only one child was event-free at 3 years from diagnosis. One third of patients with t(15;17) died during induction. Those eight who achieved CR fared well: only two relapsed, and six were event-free survivors. Patients with inv(16) had a high remission rate and a long survival: five children were in CR 20 to 136 months. Both groups with t(9;11) and t(11q23) had a high remission rate: however, outcome was superior for the t(9;11) group when compared to either the t(11q23) group (EFS at 3 years +/- SE, 56 +/- 17% vs. 11 +/- 10%, p = 0.07) or to the remaining patients (p = 0.06). Both -7/del(7q) and t(1;22) groups had low CR rates (50%) and poor survival. Cytogenetic analysis identifies clinically distinct subsets of childhood AML and is useful in tailoring treatment for these patients. Favorable cytogenetic groups (t(15;17), inv(16), and t(9;11)) may do well with current therapy protocols, whereas unfavorable groups (t(11q23), t(8;21), -7/del(7q), and t(1;22)) require more effective therapies.