Gene expression profiling reveals clear differences between EBV-positive and EBV-negative posttransplant lymphoproliferative disorders.

Posttransplant patients are at risk of developing a potentially life-threatening posttransplantation lymphoproliferative disorder (PTLD), most often of diffuse large B cell lymphoma (DLBCL) morphology and associated with Epstein-Barr Virus (EBV) infection. The aim of this study was to characterize the clinicopathological and molecular-genetic characteristics of posttransplant DLBCL and to elucidate whether EBV(+) and EBV(-) posttransplant DLBCL are biologically different. We performed gene expression profiling studies on 48 DLBCL of which 33 arose posttransplantation (PT-DLBCL; 72% EBV+) and 15 in immunocompetent hosts (IC-DLBCL; none EBV+). Unsupervised hierarchical analysis showed clustering of samples related to EBV-status rather than immune status. Except for decreased T cell signaling these cases were inseparable from EBV(-) IC-DLBCL. In contrast, a viral response signature clearly segregated EBV(+) PT-DLBCL from EBV(-) PT-DLBCL and IC-DLBCL cases that were intermixed. The broad EBV latency profile (LMP1+/EBNA2+) was expressed in 59% of EBV(+) PT-DLBCL and associated with a more elaborate inflammatory response compared to intermediate latency (LMP1+/EBNA2-). Inference analysis revealed a role for innate and tolerogenic immune responses (including VSIG4 and IDO1) in EBV(+) PT-DLBCL. In conclusion we can state that the EBV signature is the most determining factor in the pathogenesis of EBV(+) PT-DLBCL.

Diffuse large B-cell lymphoma: a heterogeneous group of non-Hodgkin lymphomas comprising several distinct clinicopathological entities.

Diffuse large B-cell lymphoma (DLBCL) as defined by the World Health Organization (WHO) classification is clinically, morphologically and genetically a heterogeneous group of malignant proliferations of large lymphoid B cells. Over the last 6 years, several studies have been published improving our understanding of these lymphomas. These studies analyzed DLBCL by their gene expression profile, provided further information on some of the variants of DLBCL listed in the WHO classification and stressed the impact of the site of origin of these tumors. This review summarizes these recent data and explores their impact on the recognition of new clinicopathological lymphoma entities.

The dynamics of the B follicle: understanding the normal counterpart of B-cell-derived malignancies.

The repertoire of B cells secreting antibodies with unique antigen-binding specificities is produced at two stages: a primary B-cell repertoire is formed in the bone marrow through immunoglobulin gene rearrangements, whereas a secondary B-cell repertoire is generated in the peripheral lymphoid organs (spleen, lymph nodes and mucosa-associated lymphoid tissue) through somatic hypermutation and class-switch recombination upon antigen encounter. The latter events take place within highly specialized histological structures, designated B follicles, which are composed of distinct microanatomical compartments namely the follicle centre, lymphocytic corona and marginal zone. Each compartment comprises a particular subset of B cells, characterized by unique properties, thereby reflecting the complexity and variability in the spectrum of defence mechanisms against invading pathogens. The past years have spawned an avalanche of new data and information that encompasses both the structure and function of each compartment and its B cells. This review incorporates up-to-date information on peripheral B-cell differentiation into a challenging working model, thereby pointing to the structural and functional imprint of both the T-cell-dependent and T-cell-independent immune response on the B follicle. As such, this article aims to form an excellent base for a better understanding of the normal counterpart of B-cell-derived haematological malignancies (leukemias and lymphomas).

The pathogenesis of MALT lymphomas: where do we stand?

Mucosa-associated lymphoid tissue (MALT) lymphoma is a heterogeneous form of a B-cell non-Hodgkin's lymphoma with extranodal location. The gastrointestinal tract is the most common site of disease, but involvement of multiple other organ systems has been documented. Four translocations, t(11;18)(q21;q21), t(1;14)(p22;q32), t(14;18)(q32;q21) and t(3;14)(p13;q32), are specifically associated with MALT lymphoma. Remarkably, the genes targeted by at least three of these translocations are involved in one and the same pathway, leading to the activation of nuclear factor-kappaB (NF-kappaB). This review presents MALT lymphoma as a model of how sustained inflammation increases the risk of genotoxic insults and how these genetic events initiate oncogenesis.

CTLA-4 blockade in murine bone marrow chimeras induces a host-derived antileukemic effect without graft-versus-host disease.

We studied the effect of CTLA-4 blockade on graft-versus-leukemia and graft-versus-host responses in a mouse model of minor histocompatibility-mismatched bone marrow transplantation. Early CTLA-4 blockade induced acute graft-versus-host disease. Delayed CTLA-4 blockade resulted in a lethal condition with lymphosplenomegaly, but with stable mixed T-cell chimerism, unchanged alloreactive T-cell frequencies and absent anti-host reactivity in vitro. In contrast, multiorgan lymphoproliferative disease with autoimmune hepatitis and circulating anti-DNA auto-antibodies were documented. Splenic lymphocytes exhibited ex vivo spontaneous proliferation and a marked proliferative response against host-type dendritic cells pulsed with syngeneic (host-type) tissue-peptides. Both phenomena were exclusively mediated by host and not donor T cells, supporting an autoimmune pathogenesis. Selectively host-derived T-cell immune reactivity was equally documented against leukemia-peptide-pulsed dendritic cells, and this was paralleled by a strong in vivo antileukemic effect in anti-CTLA-4-treated and subsequently leukemia-challenged chimeras. In conclusion, delayed CTLA-4 blockade induced a host-derived antileukemic effect, occurring in the context of an autoimmune syndrome and strictly separated from graft-versus-host disease. Both antileukemic and autoimmune responses depended on the allogeneic component, as neither effect was seen after syngeneic bone marrow transplantation. Our findings reveal the potential of using CTLA-4 blockade to establish antileukemic effects after allogeneic hematopoietic stem cell transplantation, provided autoimmunity can be controlled.

A comprehensive genetic and histopathologic analysis identifies two subgroups of B-cell malignancies carrying a t(14;19)(q32;q13) or variant BCL3-translocation.

The biologic and pathologic features of B-cell malignancies bearing a translocation t(14;19)(q32;q13) leading to a fusion of IGH and BCL3 are still poorly described. Herein we report the results of a comprehensive cytogenetic, fluorescence in situ hybridization (FISH), molecular and histopathological survey of a large series of B-cell malignancies with t(14;19) or variant translocations. A total of 56 B-cell malignancies with a FISH-proven BCL3 involvement were identified with the translocation partners being IGH (n=51), IGL (n=2), IGK (n=2) and a non-IG locus (n=1). Hierarchical clustering of chromosomal changes associated with the t(14;19) indicated the presence of two different groups of IG/BCL3-positive lymphatic neoplasias. The first group included 26 B-cell malignancies of various histologic subtypes containing a relatively high number of chromosomal changes and mostly mutated IgVH genes. This cluster displayed three cytogenetic branches, one with rearrangements in 7q, another with deletions in 17p and a third one with rearrangements in 1q and deletions in 6q and 13q. The second group included 19 cases, mostly diagnosed as B-cell chronic lymphocytic leukemia (B-CLL), and characterized by few additional chromosomal changes (e.g. trisomy 12) and unmutated IgVH genes. In conclusion, our study indicates that BCL3 translocations are not restricted to B-CLL but present in a heterogeneous group of B-cell malignancies.

Aggressive systemic mastocytosis complicated by protein-losing enteropathy.

A 46-year-old woman was admitted to the hospital with complaints of chronic diarrhoea, vomiting and severe muscle weakness. Clinical examination showed a lethargic, malnourished, dehydrated patient with ascites and bilateral leg oedema. Laboratory evaluation revealed mild normochromic normocytic anaemia and severe hypoproteinaemia with hypoalbuminaemia. Upper gastrointestinal endoscopy showed a thickened, friable duodenal mucosa with multiple erosions. Colonoscopy revealed nodular, pseudopolypoid lesions with patchy erosions in the left hemicolon. Haematoxylin-eosin stained sections from biopsies of endoscopically abnormal bowel segments showed multi-focal aggregates of large, histiocyte-like cells with abundant pale cytoplasm in the lamina propria. These cells were negative on PAS, Ziehl-Neelsen, Giemsa and toluidine blue stains. Their immunophenotype was CD68 (+), c-kit/CD117 (+) and mast cell tryptase (+), which is consistent with mast cells. A trephine biopsy showed diffuse replacement of the bone marrow by atypical, monomorphic, frequently spindle-shaped mast cells. No associated haematopoietic malignancy was detected. The final diagnosis was aggressive systemic mastocytosis with involvement of the gastrointestinal tract complicated by protein-losing enteropathy. This association has not been reported previously. The patient has been treated with prednisolone and interferon-alpha and has since recovered.

FOXP1, a gene highly expressed in a subset of diffuse large B-cell lymphoma, is recurrently targeted by genomic aberrations.

The transcription factor Forkhead box protein P1 (FOXP1) is highly expressed in a proportion of diffuse large B-cell lymphoma (DLBCL). In this report, we provide cytogenetic and fluorescence in situ hybridization (FISH) data showing that FOXP1 (3p13) is recurrently targeted by chromosome translocations. The genomic rearrangement of FOXP1 was identified by FISH in three cases with a t(3;14)(p13;q32) involving the immunoglobulin heavy chain (IGH) locus, and in one case with a variant t(2;3) affecting sequences at 2q36. These aberrations were associated with strong expression of FOXP1 protein in tumor cells, as demonstrated by immunohistochemistry (IHC). The cases with t(3p13) were diagnosed as DLBCL ( x 1), gastric MALT lymphoma ( x 1) and B-cell non-Hodgkin's lymphoma, not otherwise specified ( x 2). Further IHC and FISH studies performed on 98 cases of DLBCL and 93 cases of extranodal marginal zone lymphoma showed a high expression of FOXP1 in approximately 13 and 12% of cases, respectively. None of these cases showed, however, FOXP1 rearrangements by FISH. However, over-representation of the FOXP1 locus found in one additional case of DLBCL may represent another potential mechanism underlying an increased expression of this gene.

A novel N-ras mutation in malignant melanoma is associated with excellent prognosis.

Mutations in the ras gene are key events in the process of carcinogenesis; in particular, point mutations in codon 61 of exon 2 of the N-ras gene occur frequently in malignant melanoma (MM). We searched for point mutations in the N-ras gene in a large series of primary and metastatic MM from 81 different retrospectively selected patients using the very sensitive denaturing gradient gel electrophoresis technique, followed by sequencing. The classical codon 12 and codon 61 mutations were found in 21 and 17% of the cases, respectively. No codon 13 mutation was found. A novel mutation at codon 18 of exon 1, consisting of a substitution of alanine (GCA) by threonine (ACA), was found in 15% of the primary MMs but in none of the metastatic MMs. All of the other cases were free of mutations. Using microdissected cells from distinctive MM growth phases as source of DNA for mutation analysis, this particular N-ras exon 1 mutation at codon 18 was already present in the radial growth phase and preserved throughout the successive growth phases; it was also found in a dysplastic nevi in continuity with a MM, indicating a clonal relationship between both lesions. Our findings also illustrate the clonal relationship between the distinctive growth phases in MM and suggest the codon 18 mutation to occur early in MM development. The MM in patients with this mutation were significantly thinner than those without a codon 18 mutation (P = 0.0257). Statistical analysis, comparing the group of codon 18 patients with the group of patients with the classical mutations and without mutations, revealed a highly significant difference in overall outcome. The cumulative probability of developing metastasis was significantly lower for the group patients with a codon 18 mutation (P = 0.0130). We can thus conclude that this codon 18 mutation identifies a group of patients with better prognosis than patients with melanoma that harbor wild-type sequence or classical activating point mutations in codon 12 or 61. Preliminary nucleotide binding measurements could not detect a difference between wild-type Ras protein and the mutant Ras(A18T) protein. However, for a precise elucidation of the role of the N-Ras(A18T) mutant in melanoma, additional studies aimed to measure the affinity to guanine nucleotide exchange factors and GTPase-activating proteins are needed.

T-cell/histiocyte-rich large B-cell lymphoma: a distinct clinicopathologic entity.

PURPOSE: Although it has proven difficult to delineate diagnostically reproducible and clinically relevant subgroups, the heterogeneity of diffuse large B-cell lymphomas (DLBCL) is widely acknowledged. In 1992, we reported on six cases that suggested that large B-cell lymphoma rich in stromal histiocytes and T cells may be identified as a distinct clinicopathologic entity within DLBCL. PATIENTS AND METHODS: An integrated clinicopathologic study of 40 cases of this DLBCL subtype is presented. RESULTS: Distinguishing a DLBCL rich in histiocytes and reactive T cells, designated T-cell/histiocyte--rich large B-cell lymphoma (THR-BCL), may be justified from a clinical point of view. The disease typically affects middle-aged male patients who usually present with advanced-stage disease that is not adequately managed with current therapeutic strategies. Whereas proliferation fraction and p53 overexpression, in addition to the clinical variables incorporated in the International Prognostic Index (IPI), significantly correlate with response to treatment and survival in a univariate analysis, only the IPI score identifies relevant prognostic THR-BCL subpopulations in a multivariate model. The morphologic and immunophenotypic profile of the neoplastic B cells in THR-BCL suggests that they may originate from a germinal center ancestor. CONCLUSION: THR-BCL constitutes a distinct clinicopathologic entity that is characterized by an aggressive behavior. Experimental therapeutic strategies may be indicated to obtain a more favorable response to treatment in this disease.

Additional prognostic value of bone marrow histology in patients subclassified according to the International Prognostic Scoring System for myelodysplastic syndromes.

PURPOSE: The most recent and powerful prognostic instrument established for myelodysplastic syndromes (MDS) is the International Prognostic Scoring System (IPSS), which is primarily based on medullary blast cell count, number of cytopenias, and cytogenetics. Although this prognostic system has substantial predictive power in MDS, further refinement is necessary, especially as far as lower-risk patients are concerned. Histologic parameters, which have long proved to be associated with outcome, are promising candidates to improve the prognostic accuracy of the IPSS. Therefore, we assessed the additional predictive power of the presence of abnormally localized immature precursors (ALIPs) and CD34 immunoreactivity in bone marrow (BM) biopsies of MDS patients. PATIENTS AND METHODS: Cytogenetic, morphologic, and clinical data of 184 MDS patients, all from a single institution, were collected, with special emphasis on the determinants of the IPSS score. BM biopsies of 173 patients were analyzed for the presence of ALIP, and CD34 immunoreactivity was assessable in 119 patients. Forty-nine patients received intensive therapy. RESULTS: The presence of ALIP and CD34 immunoreactivity significantly improved the prognostic value of the IPSS, with respect to overall as well as leukemia-free survival, in particular within the lower-risk categories. In contrast to the IPSS, both histologic parameters also were predictive of outcome within the group of intensively treated MDS patients. CONCLUSION: Our data confirm the importance of histopathologic evaluation in MDS and indicate that determining the presence of ALIP and an increase in CD34 immunostaining in addition to the IPSS score could lead to an improved prognostic subcategorization of MDS patients.

Prognostic value of positron emission tomography (PET) with fluorine-18 fluorodeoxyglucose ([18F]FDG) after first-line chemotherapy in non-Hodgkin's lymphoma: is [18F]FDG-PET a valid alternative to conventional diagnostic methods?

PURPOSE: A complete remission (CR) after first-line therapy is associated with longer progression-free survival (PFS). However, defining CR is not always easy because of the presence of residual masses. Metabolic imaging with fluorine-18 fluorodeoxyglucose ([18F]FDG) positron emission tomography (PET) offers the ability to differentiate between viable and fibrotic inactive tissue. In this study, we evaluated the value of PET in detecting residual disease and, hence, predicting relapse after first-line treatment in patients with non-Hodgkin's lymphoma (NHL). PATIENTS AND METHODS: Ninety-three patients with histologically proven NHL, who underwent a whole-body [18F]FDG-PET study after completion of first-line chemotherapy and who had follow-up of at least 1 year, were included. Persistence or absence of residual disease on PET was related to PFS using Kaplan-Meier survival analysis. RESULTS: Sixty-seven patients showed a normal PET scan after first-line chemotherapy; 56 of 67 remained in CR, with a median follow-up of 653 days. Nine of these patients with a residual mass considered as unconfirmed CR received additional radiotherapy. Only 11 of 67 patients relapsed (median PFS, 404 days). Persistent abnormal [18F]FDG uptake was seen in 26 patients, and all of them relapsed (median PFS, 73 days). Because standard restaging also suggested residual disease, 12 patients received immediate secondary treatment. In 14 of 26 patients, only PET predicted persistent disease. From these patients, relapse was proven either by biopsy (n = 8) or by progressive disease on computed tomography or magnetic resonance imaging (n = 6). CONCLUSION: Persistent abnormal [18F]FDG uptake after first-line chemotherapy in NHL is highly predictive for residual or recurrent disease. In relapsing patients, PFS was significantly shorter after a positive scan than after a negative scan.

Maintenance of remission with human recombinant interferon alfa-2a in patients with stages III and IV low-grade malignant non-Hodgkin's lymphoma. European Organization for Research and Treatment of Cancer Lymphoma Cooperative Group.

PURPOSE: Interferon alfa has shown significant activity in patients with low-grade malignant non-Hodgkin's lymphoma (NHL). In 1985, we initiated a prospective randomized study in which the potential benefit of interferon alfa given as maintenance treatment was investigated after tumor load reduction was achieved with chemoradiotherapy in patients with advanced low-grade malignant non-Hodgkin's lymphoma. PATIENTS AND METHODS: The study involved 347 patients with stage III or IV disease, 315 satisfying the eligibility criteria. All were treated with a regimen of cyclophosphamide, vincristine, and prednisone (CVP) given every 3 weeks for eight cycles. Thereafter, patients were eligible for iceberg irradiation. Finally, all patients were completely restaged, and responding and stable-disease patients were then randomized, 122 to interferon alfa-2a maintenance, 3 million U three times weekly for 1 year; and 120 to no further treatment. RESULTS: Seventy-nine percent of the patients response to CVP, ie, 45% complete remissions (CR) and 34% partial remissions (PR). In the group of randomized patients, the response rate after CVP plus or minus radiotherapy was 90%. As compared with control patients, patients in the interferon (IFN) maintenance group had a tendency toward a prolonged time to progression (TTP) (median, 132 v 87 weeks; P = .054, adjusted for response to CVP). However, overall survival was similar in both groups. Interferon was well tolerated. The median dose of IFN actually received corresponded to 90% of the planned cumulative dose. The treatment had to be stopped because of toxicity in 16 patients (15% of the patients in whom IFN was started). CONCLUSION: Interferon maintenance treatment in the phase of minimal residual disease of patients with advanced low-grade malignant NHL increased TTP at the borderline of statistical significance, without remarkable toxicity. However, overall survival was not influenced.

A new t(2;5) translocation in a null cell type CD30 positive anaplastic large cell lymphoma case.

Anaplastic large cell lymphoma (ALCL) expressing the CD30 antigen is an uncommon subtype of non-Hodgkin's lymphoma characterized by distinct morphological and clinical features. The recurrent chromosomal abnormality found in these tumours is a t(2;5)(p23;q35) which has been detected in a minority of these cases, predominantly with a T cell immunophenotype. We report here a CD30 positive null cell type ALCL case cytogenetically characterized by a new type of t(2;5) translocation with distinct breakpoints at 2q37 and 5q31. FISH with a panel of 5q specific DNA probes applied in this case allowed for a mapping of a 5q31 breakpoint region between the locus for IL-3 (proximally) and CI5-56 probe (distally). These results point to a localization of unknown gene(s) on the long arm of chromosome 5 that, in addition to the NPM gene at 5q35, may be involved in the pathogenesis of some CD30+ ALCL.

A single course of remission reinduction chemotherapy for acute myelogenous leukemia relapsing after allogeneic bone marrow transplantation is complicated by graft-versus-host disease and followed by sustained complete remission.

Graft-versus-host disease (GVHD) remains a major immunological complication after allogeneic bone marrow transplantation (allo-BMT), but also favors development of the beneficial graft-versus-leukemia (GVL) effect. A patient with AML-M4 (inv (16)) is described, who was given non-myeloablative remission reinduction therapy for leukemic relapse (inv (16), trisomy 8) diagnosed on day 184 after HLA-compatible sibling BMT. On day 236, ie about 6 weeks after completion of this course, a clinical syndrome suggestive of acute GVHD grade 3 had developed. Skin biopsy confirmed the clinical diagnosis of GVHD, with a compatible liver biopsy. Transfusion-associated GVHD was ruled out by analysis of short tandem repeat (STR) alleles in the skin biopsy, revealing alleles from donor and recipient but not from third party origin. Cyclosporin A (CsA) therapy, which had been tapered between days 150 and 175, was resumed, resulting in a favorable response and gradual transition to limited chronic GVHD. The patient has since remained in complete remission with an excellent performance status for more than 40 months, without further chemotherapy. Thus this biopsy proven case of GVHD was induced by marrow donor lymphocytes more than 200 days after transplantation and apparently triggered by remission reinduction chemotherapy. The case indicates that intensive non-myeloablative chemotherapy can cure AML relapsing after allo-BMT. The therapeutic effect in this case probably involved a direct pharmacological suppression of the leukemic clone followed by a GVL effect initiated by donor-derived alloreactive T lymphocytes.

Expression of B7-2 (CD86) molecules by Reed-Sternberg cells of Hodgkin's disease.

Ligation of CD28 on T cells with its natural ligands B7-1 (CD80) or B7-2 (CD86) provides a major costimulatory signal for T cells and is of potential importance for tumor rejection. We previously reported a strong expression of B7-1 on Reed-Sternberg cells and anaplastic large cell lymphoma cells. We report here our findings on B7-2 expression by malignant lymphomas (n = 70). B7-2 was present on the neoplastic cells of anaplastic large cell lymphoma in two of three cases studied, and on a subpopulation of the malignant cells in one out of four cases of follicular lymphoma. B7-2 was not expressed by the neoplastic cells of the other non-Hodgkin's lymphomas (n = 32), including T cell-rich B cell lymphoma. In contrast, Reed-Sternberg cells in lymph nodes affected by Hodgkin's disease are strongly positive for B7-2 (n = 31). Evidence for a functional correlate of this expression was obtained by our findings that the combination of anti-B7-1 and anti-B7-2 monoclonal antibodies was more effective than each separately in blocking allogeneic T cell activation (proliferation and cytokine secretion) by Hodgkin's disease-derived cell lines as stimulators. The possible role of B7-1 and B7-2 expression for the course and symptomatology of Hodgkin's disease is discussed.

Small B cell NHL and their leukemic counterpart: differences in subtyping and assessment of leukemic spread.

Three subtypes of small lymphocytic lymphoma were studied, namely B cell chronic lymphocytic leukemia (B-CLL), mantle cell lymphoma (MCL) and follicle center lymphoma (FCL). Agreement between tissue diagnosis, based on the proposal for a revised European-American classification of lymphoid neoplasms from the International Lymphoma Study Group, and the cytomorphological diagnosis on peripheral blood and/or bone marrow smears, using the proposals for the classification of chronic (mature) B and T lymphoid leukemias of the French-American-British Cooperative Group, was studied. Full agreement was found in 90% of the CLL and 82% of the FCL cases. In MCL cases, agreement was 65% including all cases classified as intermediate/mantle zone lymphoma according to FAB criteria. The incidence of bone marrow involvement detection in trephines compared to smears was equal in CLL (both 100%) and slightly higher in MCL (56 vs 48.5%); in FCL, however, trephine biopsies provided more reliable material (71 vs 35%).

Primary diffuse large B cell lymphoma of the stomach: analysis of somatic mutations in the rearranged immunoglobulin heavy chain variable genes indicates antigen selection.

Gastric low grade MALT lymphomas show a pattern of somatic mutations in their rearranged immunoglobulin genes, indicative of antigen selection. This provides evidence for antigen stimulation in the lymphomagenesis. Gastric diffuse large B cell lymphomas develop secondary to low grade MALT lymphoma or de novo. To study whether antigen-selection is also a feature of primary diffuse large B cell lymphomas, we analysed somatic mutations in the rearranged immunoglobulin heavy chain (IgH) variable genes (VH). The rearranged VH genes of six cases of gastric primary diffuse large B cell lymphoma were amplified from genomic or complementary DNA by a VH gene family-specific polymerase chain reaction method. The PCR products were directly sequenced and were compared to published germline sequences to analyse somatic mutations. Similarly to low grade MALT lymphomas 5/6 primary diffuse large B cell lymphomas show a pattern of somatic mutation in their rearranged VH genes, indicative of antigen selection and suggesting a role for antigens in lymphomagenesis. One case showed bi-allelic VH gene rearrangements, which were non-functional due to extensive deletions. Antigen selection could not be demonstrated or excluded. Antigen selection is a common feature in most analysed primary diffuse large B cell lymphomas, although some heterogeneity in the mechanisms involved in the lymphomagenesis of gastric primary diffuse large B cell lymphomas has not been excluded entirely (case 4).

Characteristic pattern of chromosomal gains and losses in marginal zone B cell lymphoma detected by comparative genomic hybridization.

Marginal zone B cell lymphoma (MZBCL) represents a distinct subtype of B cell non-Hodgkin's lymphoma, which has been recently recognized and defined as a disease entity. We investigated 25 cases (18 at primary diagnosis and seven during the course of disease) of MZBCL by comparative genomic hybridization (CGH) and compared these results with cytogenetic, fluorescence in situ hybridization (FISH), and Southern blot data. Twenty of the 25 cases (80%) showed gains (total 49) or losses (total 15) of genetic material. In extranodal, nodal, and splenic MZBCL, material of chromosomes 3 (52% of cases), 18 (32%), X (24%), and 1q (16%) was most frequently gained, whereas losses predominantly involved chromosomes 17 (16%) and 9 (12%). High-level amplifications involving the regions 18q21-23 and 18q21-22, respectively, were detected in two cases. Gains of chromosomes 1q and 8q and losses of chromosome 17 or 17p occurred more frequently in relapsed or progressive lymphomas. For all of the frequently affected chromosomes, CGH allowed narrowing of the relevant subregions including 3q21-23, 3q25-29 and 18q21-23. By Southern blot analysis, the BCL2, BCL6, and CMYC proto-oncogenes were found to be a part of the over-represented regions in two cases, one case, and two cases, respectively, with gains involving 18q, 3q or 8q. In 13 cases, CGH revealed chromosomal imbalances which were not detected by cytogenetic analysis but could be confirmed by FISH or Southern blot analysis in all cases investigated. On the other hand, CGH failed to detect trisomy 3, trisomy 18, and deletion 7q in three cases with a low proportion of tumor cells bearing these abnormalities, as shown by interphase FISH. The characteristic pattern of chromosomal gains and losses detected in this study confirms the distinct nature of MZBCL and may point to chromosomal regions involved in the pathogenesis of these neoplasms.

Analysis of N- and K-ras mutations in the distinctive tumor progression phases of melanoma.

Mutations in the ras genes are key events in the process of carcinogenesis; in particular, point mutations in codon 61 of exon 2 of the N-ras gene occur frequently in cutaneous melanoma. To investigate whether these mutations occur in early or late tumor progression phases, we searched for point mutations in the N- and K-ras genes in 69 primary cutaneous melanoma, 35 metastases, and seven nevocellular nevi in association with cutaneous melanoma. Lesions were microdissected in order to procure pure tumor samples from the distinctive growth phases of the cutaneous melanoma; the very sensitive denaturing gradient gel electrophoresis technique was used to visualize the mutations, and was followed by sequencing. Point mutations in the N-ras gene but not in the K-ras gene were detected on denaturing gradient gel electrophoresis. Twenty-three primary (33%) and nine metastatic (26%) melanomas showed bandshifts for N-ras. In the majority of cases, mutations occurring in early growth phases (i.e., the "intraepidermal" radial growth phase), were preserved in later growth phases (i.e., the invasive radial growth phase, vertical growth phase, and metastatic phase), which proves the clonal relationship between the successive growth phases. In three cases, however, the mutations differed between the distinctive growth phases within the same cutaneous melanoma, due to the occurrence of an additional mutation (especially in codon 61) in a later tumor progression phase. Our approach also permitted us to analyze the mutational status of nevi, associated with cutaneous melanoma. Six out of seven associated nevi carried the same sequence (mutated or wild-type) as the primary cutaneous melanoma, whereas in one case the sequence for N-ras differed between the primary melanoma and the associated nevus. In conclusion, this approach allowed us to demonstrate the clonal relationship between subsequent growth phases of melanoma and associated nevi; our results suggest that N-ras exon 1 mutations preferentially occur during early stages of tumor progression and hence may be involved in melanoma initiation, whereas those in N-ras exon 2 are found preferentially during later stages and hence are more probably involved in metastatic spread of cutaneous melanoma.