Classification systems for acute and chronic leukaemias

Modern classification systems for acute and chronic leukaemias are based on cytomorphology, cytochemistry, immunophenotyping, immunogenetics and molecular cytogenetics. Morphology forms the initial diagnosis of leukaemia, but generally is not sufficient to identify biologically and clinically relevant subsets within the main categories of leukaemia. Immunophenotyping precisely defines the lineage and stage of differentiation of malignantly transformed haematopoietic cells. This is usually sufficient for precise classification of mature lymphoid malignancies, although immunogenetic and (molecular) cytogenetic studies might be helpful to confirm the diagnosis of disseminated non-Hodgkin's lymphomas. However, certain categories of disease that are clearly defined by cytomorphology and immunophenotyping, particularly acute leukaemias, are still heterogeneous, mainly owing to different underlying leukaemogenic events. Immunophenotyping can reveal subgroups highly suggestive of certain chromosome aberrations but reliable identification of such aberrations requires cytogenetic or molecular studies. Such combined diagnostic information forms the basis of current WHO classification of tumours of haematopoietic and lymphoid tissues. This will be complemented in the near future with novel criteria revealed by microarray gene expression profiling. This chapter summarizes and comments on the currently used immunophenotypic classification systems of acute and chronic leukaemias and on the added value of molecular diagnostics.

Chromosome aberrations evaluated by comparative genomic hybridization in B-cell chronic lymphocytic leukemia: correlation with CD38 expression

BACKGROUND AND OBJECTIVES

B-cell chronic lymphocytic leukemia (B-CLL) results from the accumulation of monoclonal CD5+ B cells. Despite its homogeneity at cellular level, B-CLL is clinically heterogeneous. Clinical studies indicate that CD38+ B-CLL are characterized by a more aggressive clinical course than are CD38- B-CLL. On the basis of these studies and considering the established correlation between specific chromosome aberrations and the clinical course of B-CLL, it is possible that CD38+ B-CLL cases are also characterized by specific subsets of chromosomal alterations.

DESIGN AND METHODS

Comparative genomic hybridization (CGH) was performed on purified B-cells from peripheral blood of 52 patients with B-CLL in order to detect chromosome imbalance. The immunophenotype of the patients, including CD38 expression, was also determined by flow cytometry. The results of CGH experiments were then compared with CD38 expression.

RESULTS

We found a clear correlation between the presence of chromosomal imbalances and CD38 expression: 13/16 CD38+ cases had chromosome imbalances, most of them (12/13) correlated with a poor prognosis. Among the CD38- B-CLL patients, only 8/36 displayed chromosome imbalances; the only three cases with loss in 13q as a single aberration, considered a good prognostic marker, were in this group. Moreover, we found that cytogenetic alterations were also more complex in the CD38+ B-CLL subset, since 9/10 with two or more aberrations were in the CD38+ group.

INTERPRETATION AND CONCLUSIONS

Collectively, the data reinforce the value of CD38 as a prognostic factor and indicate that genotypic/phenotypic features distinguish B-CLL subsets.

Non-steroidal anti-inflammatory drug use and risk of adult leukemia

Epidemiological studies have suggested that regular use of aspirin and other non-steroidal anti-inflammatory drugs (NSAIDs) may be associated with reduced risk of some types of malignancy. Leukemia incidence and self-reported aspirin, as well as other NSAID use, was examined in a prospective cohort of >28,000 postmenopausal women. Eighty-one incident leukemia cases occurred during the period 1993-2000. The multivariate-adjusted relative risk of leukemia was 0.45 (95% confidence interval: 0.27-0.75) for women who reported using aspirin two or more times per week compared with women who reported no use. Similar inverse associations were observed for the two subtypes of leukemia analyzed. In contrast, for women who reported using nonaspirin NSAIDs, the multivariate-adjusted relative risk of leukemia was 1.31 (95% confidence interval: 0.77-2.22). Analyses that excluded cases diagnosed before 1995 did not notably alter results. To our knowledge, this is the first prospective study to examine the association between NSAID use and incident adult leukemia. Although preliminary, the notable differences observed in leukemia risk between aspirin and nonaspirin NSAID use warrant further investigation.

DNA microarray data and contextual analysis of correlation graphs

BACKGROUND

DNA microarrays are used to produce large sets of expression measurements from which specific biological information is sought. Their analysis requires efficient and reliable algorithms for dimensional reduction, classification and annotation.

RESULTS

We study networks of co-expressed genes obtained from DNA microarray experiments. The mathematical concept of curvature on graphs is used to group genes or samples into clusters to which relevant gene or sample annotations are automatically assigned. Application to publicly available yeast and human lymphoma data demonstrates the reliability of the method in spite of its simplicity, especially with respect to the small number of parameters involved.

CONCLUSIONS

We provide a method for automatically determining relevant gene clusters among the many genes monitored with microarrays. The automatic annotations and the graphical interface improve the readability of the data. A C++ implementation, called Trixy, is available from http://tagc.univ-mrs.fr/bioinformatics/trixy.html.

An analysis of the risk of B-lymphocyte malignancies in industrial cohorts

Among numerous studies of occupational groups with varied chemical exposures (e.g., farmers, petroleum workers, and rubber workers), some have reported excess risk for non-Hodgkin's lymphoma (NHL), multiple myeloma, and other cancers of the B-lymphocyte cell line. While not conclusive, these studies raise questions about the effects of chemical exposures on the lymphocytic versus myeloid cell lines. Almost 70 occupational cohort studies were identified that addressed B-cell cancer risks in 9 major industrial categories, in order to look for common patterns across industries. This effort was substantially limited by the inconsistent nature of lymphohematopoietic (LH) classification schemes across studies and over time, and the relative paucity of B-cell-specific results in studies for any given industry. Taking these limitations into consideration, a descriptive, graphical analysis suggested a pattern of B-cell cancer elevations in the rubber and "general chemical" industries, but no consistent patterns in petroleum production/distribution or petrochemical production. The limited data sources, which lack detail about differences in hazard and exposure for different types of products/chemicals, did not allow a comprehensive look at possible common exposures associated with B-cell cancer elevations across industries. This study suggests that evaluation of possible associations between specific chemical exposures and B-cell malignancies would require additional studies with clear and common definitions of B-cell outcomes. The article concludes by giving an example of a possible common framework for categorizing NHL, the diseases for which most classification issues arise.

Marginal zone and mantle cell lymphomas: assessment of cytomorphology in subtyping small B-cell lymphomas

The diagnosis of malignant lymphomas by fine-needle aspiration biopsy (FNAB) is increasing in utilization. The cytomorphologic distinction among the small B-cell lymphomas may be quite difficult. We are unaware of anyone who has compared directly mantle cell lymphoma (MCL) and marginal zone lymphoma (MZL) in FNAB. Our major goal was to examine the cytomorphologic attributes of MCL and MZL and to look for features distinctive of or suggestive of either neoplasm. Seven immunophenotypic MCL and seven immunophenotypic MZL aspirates were evaluated for a number of cytomorphologic features in direct smears. Features favoring MCL include a relatively monomorphic cellular population, prominent nuclear membrane contour irregularities, and mitotic figures. Conversely, a polymorphic cellular population suggested MZL. However, due to extensive overlap of specific cytologic features, the two lymphomas cannot be definitively distinguished based solely on cytomorphology. Although there are cytomorphologic attributes suggestive of either MCL or MZL, considerable overlap exists. Based on an individual case basis, the distinction cannot be made reliably by morphology alone; ancillary studies, e.g., immunophenotyping, are essential.

Biological and clinical heterogeneity of B-cell chronic lymphocytic leukemia

B-cell chronic lymphoproliferative disorders have been recognized as a heterogeneous group of neoplastic diseases affecting the lymphoid system. They are frequently characterized by leukemic manifestations with peripheral blood and/or bone marrow involvement. B-cell chronic lymphocytic leukemia (B-CLL), a disease which is now accepted as derived from immunologicallycompetent antigen-activated B-cells, is the most common in the western countries. Although for several decades it was considered as a unique entity, a body of evidences is now emerging indicating the biological heterogeneity of B-CLL. Molecular and immunophenotypic data induce to consider two main subgroups of B-CLL exist at the cellular level: in fact, B-CLL cases can be divided into two categories according to IgV gene mutation status. One group develops from "memory" B-cells after antigenic stimulation in the germinal center of secondary follicles, displays mutated gene IgV and lacks CD38 expression; the second one derives from the so-called "naïve" antigen inexperienced B-cells, does not present IgV gene mutations and displays CD38 surface molecule. The two groups seem to be characterised by different clinical outcomes with the "mutated" group expected to have a more benign course than that arising from a "naïve" B-cell. Thus, the definition of the biological features of these different groups could be of help for clinicians in order to recognize patients needing more intensive therapeutic approaches.

[Chronic lymphocytic leukemia: high time for a risk-adapted approach]

B cell chronic lymphocytic leukaemia (CLL) is the most common leukaemia in the Western world. It is wrongly considered to be an indolent disease: even patients with Binet stage A can have disease-related morbidity, necessitating treatment in about 50% of them, and over 25% of these patients will die of CLL-related causes. It was recently discovered that there are in fact 2 subtypes of CLL: a pre-germinal centre type. Characterised by unmutated genes which code for the variable parts of immunoglobulin chains (IgV), and a post-germinal centre variant, characterised by IgV somatic mutations. This IgV-gene mutational status as well as surface expression of CD38 and characteristic cytogenic abnormalities has recently been shown to be powerful prognostic factors. Furthermore, over the past few years new treatment modalities have been developed including purine analogues, immunotherapy with monoclonal antibodies and stem cell transplantation. Therefore it is time to substitute the present policy used for the vast majority of patients (i.e. an expectant approach, if needs be followed by initial treatment with chlorambucil) with an approach based on the patient's individual risk profile. This should preferably be done within the framework of clinical trials.

Diagnostic role and prognostic significance of a simplified immunophenotypic classification of mature B cell chronic lymphoid leukemias

We verified the diagnostic and prognostic role of a simplified immunophenotypic classification (IC) in a series of 258 patients (M/F: 1.4; median age: 64 years; median follow-up: 64 months; 75 deaths) with mature B cell lymphoid leukemias (MBC-LL) for whom no histopathological diagnosis was available because of minimal or no lymph node involvement. The IC was based on the reactivity of three pivotal immunophenotypic markers: CD5, CD23 and SIg intensity. On the basis of different expression patterns, we identified four diagnostic clusters (C) characterized by distinct clinico-biological features and different prognoses: C1 (149 patients) identified most classical B cell chronic lymphocytic leukemias (CLL-type cluster; SIg(dim)/CD5+/CD23+); C2, 38 patients whose clinico-hematological characteristics were intermediate between C1 and C3 (CLL-variant cluster; SIg(bright)/CD5+/CD23+/-or SIg(dim)/CD5-/-/CD23 indifferent); C3 (16 patients) most situations consistent with mantle cell lymphoma in leukemic phase (MCL-type cluster; SIg(bright)/CD5+/CD23-); and C4, 55 cases, most of whom were consistent with leukemic phase lymphoplasmacytic/splenic marginal zone lymphomas (LP/S-type cluster; SIg(bright)/CD5-/+/CD23 indifferent). At univariate survival analysis, prognosis worsened from C1 to C4, C2 and C3 (P = 0.0001), and this was maintained at multivariate analysis (P = 0.006), together with CD11c expression (P = 0.0043), age at diagnosis (cut-off 70 years; P = 0.0008) and platelet count (cut-off 140 x 10(9)/l; P = 0.0034). Besides recognising the two well-known situations of classic B-CLL and MCL, our IC identified situations with distinct prognostic and/or clinical behaviors.

CD5-negative, CD10-negative small B-cell leukemia: variant of chronic lymphocytic leukemia or a distinct entity?

CD5- and CD10-negative chronic lymphocytic leukemias are quite uncommon as compared to the CD5-positive CLL. We reviewed 250 sequential cases of peripheral blood lymphocytosis to characterize cases of small B-cell lymphoproliferative disorders, submitted with a clinical diagnosis of chronic lymphocytic leukemia exhibiting a non-classic immunophenotypic profile. Six cases of CD5-, CD10-negative chronic lymphocytic leukemias and no tissue involvement were identified that revealed high-density surface-membrane immunoglobulin and CD20 expression, with variable expression of CD11c, CD23, and CD25. Most had a profound leukocytosis (mean WBC 180 x 10(9)/L) with proliferation of mature-appearing lymphocytes. Subsequent bone marrow biopsies showed diffuse infiltration by neoplastic cells in all evaluated patients. The clinical course appeared indolent, with follow-up revealing three patients alive (survival time 38-68 months), while two died of unrelated causes and one was lost to follow-up soon after diagnosis. These cases may represent somewhat unusual chronic lymphoproliferative disorders, with morphologic features and immunophenotypic profile not readily classifiable, but which are certainly atypical for classic chronic lymphocytic leukemia. Some of these features are reminiscent of those seen in marginal-zone lymphoma. However, it is most unusual for this known to be tissue-based disease to present primarily as leukemia rather than lymphoma.

Clustering of the SOM easily reveals distinct gene expression patterns: results of a reanalysis of lymphoma study

BACKGROUND

A method to evaluate and analyze the massive data generated by series of microarray experiments is of utmost importance to reveal the hidden patterns of gene expression. Because of the complexity and the high dimensionality of microarray gene expression profiles, the dimensional reduction of raw expression data and the feature selections necessary for, for example, classification of disease samples remains a challenge. To solve the problem we propose a two-level analysis. First self-organizing map (SOM) is used. SOM is a vector quantization method that simplifies and reduces the dimensionality of original measurements and visualizes individual tumor sample in a SOM component plane. Next, hierarchical clustering and K-means clustering is used to identify patterns of gene expression useful for classification of samples.

RESULTS

We tested the two-level analysis on public data from diffuse large B-cell lymphomas. The analysis easily distinguished major gene expression patterns without the need for supervision: a germinal center-related, a proliferation, an inflammatory and a plasma cell differentiation-related gene expression pattern. The first three patterns matched the patterns described in the original publication using supervised clustering analysis, whereas the fourth one was novel.

CONCLUSIONS

Our study shows that by using SOM as an intermediate step to analyze genome-wide gene expression data, the gene expression patterns can more easily be revealed. The "expression display" by the SOM component plane summarises the complicated data in a way that allows the clinician to evaluate the classification options rather than giving a fixed diagnosis.

Comparison of immunophenotypes of small B-cell neoplasms in primary lymph node and concurrent blood or marrow samples

Immunophenotyping of small B-cell neoplasms (SBCNs) may have a critical role in diagnosis. However, there are few data addressing whether the immunophenotypes of SBCNs in bone marrow (BM) and peripheral blood (PB) are representative of those in other tissue sites. We compared the immunophenotypic features of concurrently analyzed lymph node (LN) and BM/PB specimens using multiparameter flow cytometry. Fifty-five SBCNs were identified: 27 follicular lymphomas (FLs), 16 chronic lymphocytic leukemia/small lymphocytic lymphomas (CLL/SLLs), and 12 mantle cell lymphomas (MCLs). Major (presence vs absence) or minor (alteration of intensity) variations in expression of individual antigens between LN and BM/PB were observed in up to 25% of cases within a particular SBCN category. All FLs and CLL/SLLs maintained characteristic immunophenotypes in BM/PB. Potentially misleading variations included 1 case of MCL that failed to express CD5 in BM and likely would have been immunophenotypically misclassified as a marginal zone lymphoma and another MCL that expressed moderate CD23 in PB and would have required additional studies for precise classification. The remaining major and minor variations would not have affected interpretation.

Chronic lymphocytic leukaemia: only treat when symptomatic

(1) Chronic lymphocytic leukaemia usually occurs after the age of 60. It is more common in men than in women. (2) Life expectancy of asymptomatic patients differs little from that of the general population, and simple monitoring is generally recommended. (3) Chlorambucil is the standard drug used to slow progression of symptomatic disease. Fludarabine is used if chlorambucil fails. (4) There is no evidence that combination cytotoxic therapy is any better than chlorambucil or fludarabine alone. (5) Alemtuzumab is not yet adequately evaluated. It may prolong survival by a few months, but only at the cost of sometimes serious side effects.

[Epidemiologic, clinical and therapeutic aspects of chronic lymphoid leukemia: apropos of 120 cases]

We retrospectively studied 120 cases of chronic lymphocytic leukemia diagnosed between January 1988 and December 1998. The median age of our patients was of 66 years, 75% among them were male. The discovery of the illness was fortuitous in 20% of the cases, the peripheral adenopathy and the splenomegaly were noted respectively in 72 and 48% of the cases. The blood lymphocytosis was on average 51.109/1 with extremes of 5 and 818.109/1. Anemia was noted in 71% of the cases and a thrombopenia in 42%. Fifty patients were classified C stage of BINET and sixty elevated risk according to RAI. The therapeutic attitude was according to patient's age and the CLL stage. Thus, 94 patients received a chemotherapy and a complete or partial response was observed in 58 of the cases. The overall survival at 5 years were 47%. The retained prognostic factors were the stage according to the classifications of BINET and RAI, the thrombopenia and the lymphocytosis blood overhead 100.109/1.

Apoptosis in lymphocytic leukemias and lymphomas

Most current classifications of lymphoid neoplasms define the tumors based on the cell of origin, phenotype, genetic abnormalities, and clinical features. Here it is proposed that human lymphocytic tumors can be categorized based on the propensity and capacity of the tumor cells to undergo apoptosis. The first category is defined by malignant cells that are resistant to apoptosis due to expression of anti-apoptotic factors such as bcl-2 and cellular inhibitors of apoptosis (IAPs). These tumors would include CLL and follicular lymphomas, as well as some malignancies in which the tumor cells are infected by viruses that co-opt cell survival pathways, such as human T-cell leukemia/lymphoma virus (HTLV)-1. The second category, in which the malignant cells are apoptosis-prone, would include tumors arising in the context of impaired cytotoxic T-cell function. These neoplasms would include some human immunodeficiency virus (HIV)-related lymphomas such as Burkitt's lymphoma, and post-transplantation lymphomas. The third category would include neoplasms of intermediate sensitivity to apoptosis, some of which are associated with infection such as mucosa-associated lymphoid tissue (MALT) lymphomas of the stomach. Although this classification is tentative, it should evolve in parallel with our understanding of pathogenic mechanisms in lymphoid neoplasia, and provides a novel framework with which to consider the appropriateness of specific therapeutic strategies. Distinctions among lymphocytic tumors in terms of the likelihood of response to therapies such as antisense to bcl-2 related proteins, inhibitors of NF-kappa B activity, and new approaches aimed at bolstering the host's immune response, would cross standard classifications based on the T or B-cell origin of the tumor cells.

Telomere length and heavy-chain mutation status in familial chronic lymphocytic leukemia

We examined whether telomere lengths of peripheral blood mononuclear cells are associated with immunoglobulin gene usage in 21 familial chronic lymphocytic leukemia (CLL) patients. Subjects with unmutated V genes tended to have shorter telomeres than those with somatic mutations, especially after adjusting for age. Unlike V(H) mutation status, telomere length was not predictive for survival. Our results suggest that telomere length is associated with V(H) gene mutation status and provides further evidence that the biological basis of familial B-CLL is similar to that of sporadic patients.

Prognostic relevance of a scoring system based on clinical and biological parameters in early chronic lymphocytic leukemia

INTRODUCTION

Among patients with indolent form of B-cell chronic lymphocytic leukemia, some of them will progress into more advanced stages. To better define this subpopulation of patients, we attempted to define some parameters capable of predicting a pejorative clinical outcome.

MATERIALS AND METHODS

Eighty-eight previously untreated patients with B-cell chronic lymphocytic leukemia in Binet stage A were analysed to study the prognostic value of simple serological variables: soluble CD23 (sCD23), beta2 microglobulin (beta2m), lactate-dehydrogenase activities and albumin level. Results were compared to other conventional clinical and biological parameters by univariate and multivariate statistical analysis.

RESULTS

Our data show that: (1) among those studied, sCD23 >50 u/ml was the only serological significant parameter clearly correlated with disease progression and (2) stage A" patients (hemoglobin level between 100 and 120 g/l and/or lymphocytosis >30.10(9)/l), axillary lymph nodes and hypogammaglobulinemia were found to be other variables associated with a pejorative outcome. These four variables enabled the establishment of a scoring system, capable of predicting disease progression since 66% of the patients with a score < or =2 are going to evolve into advanced stages vs 12% with a score <2. Furthermore, the time to progression is shortened when the score is increasing.

CONCLUSION

Our findings show the prognostic relevance of a scoring system including sCD23 level. This score could be taken into account in the treatment strategy of B-cell chronic lymphocytic leukemia.