DNA Microarrays in Lymphoid Malignancies
DNA Microarrays in Lymphoid Malignancies
In this issue of ONCOLOGY,
Rosenwald describes in detail the
current state of gene expression
profiling and its role in the classification
and prediction of outcome in lymphoid
malignancies. Since the first
publication describing the use of gene
expression signatures to predict outcome
in diffuse large B-cell lymphoma
(DLBCL), a few subsequent papers
have appeared, detailing comprehensive
findings in other lymphoma subtypes.[
1-4] Together these works bring
clarity to the molecular taxonomy
of the non-Hodgkin's lymphomas
(NHL). Insights gleaned from these
studies not only provide an improved
understanding of the biology of these
tumors, but have also led to a significant
improvement in our ability to assign
risk for individual patients with
Microarray analysis of NHL has generated significant excitement in the areas of lymphoma classification and biomarker studies. This new technology has provided insights into the biology of NHL, as well as the tools for subclassification of diseases into unique categories; in addition, it can be used to develop models for risk stratification of patients. Combined with clinical prognostic factors such as the International Prognostic Index (IPI) scoring system, microarray analyses aid our ability to determine prognosis by revealing biomarkers that contribute to the prediction of outcome. For example, subclassification of DLBCL into germinal center B-cell (GCB) and activated B-cell (ABC) subtypes adds independent predictive power for determining outcome in this disease.[1,2] The distinction between these two subtypes based on cell of origin not only has prognostic significance, but also identifies potential new avenues of therapy. ABC-type DLBCLs rely on constitutive activation of the nuclear factor-kappaB pathway, suggesting that this subgroup may benefit from therapeutic strategies that target the proteosome (eg, bortezomib [Velcade]), reducing levels of downstream nuclear factor-kappaB target genes. Such clinical trials are under way at the National Cancer Institute. Similarly, targeting of protein kinase C-beta is being tested in a group of DLBCL patients with inferior survival characteristics. The 17-gene linear predictor developed for DLBCL has clearly been shown to contribute to the prognosis of the IPI score, as assessed by multivariate models. Moreover, the significant predictive ability of the proliferation signature in mantle cell lymphoma based on 20 proliferationassociated genes, far exceeds any previous biomarkers used to predict outcome in this disease. Thus, gene expression profiling has provided a new list of important biomarkers that, in addition to helping define the biology of NHL, also contribute significantly to the development of new predictive models. The single study of microarray analysis of Hodgkin's lymphoma clinical samples similarly was able to predict patients destined to treatment failure. Gene Expression Profiles
The comprehensive analyses of mRNA levels in tumor cells provide a genome-wide view of the expression profile on malignant cells. Because activation of specific signaling pathways and novel gene expression can be anticipated from these studies, an improved understanding of the biology of these tumors is an unexpected benefit of this work. For example, a recent study of gene expression profiling in primary mediastinal B-cell lymphoma patients revealed a gene expression pattern that is distinct from either the GCB or ABC subtypes of nodal DLBCL, but overlaps significantly with classic Hodgkin's lymphomas.[ 8] A complete list of the genes that allow a distinction between primary mediastinal B-cell lymphomas and classic Hodgkin's lymphomas may hold the key to understanding the unique biology of Hodgkin's lymphoma. Novel gene discovery in this context holds promise for identifying heretofore-undiscovered deregulated gene expression in NHL that has the potential to translate into novel therapies. For example, the relative overexpression of a previously unidentified gene unique to a subtype of NHL could become a new target for specific therapy. DLBCL Subtypes
The finding that unique oncogenic events are differentially associated with subclassification based on gene expression profiling supports the validity of distinguishing at least two DLBCL subtypes. For example, the t(14;18) translocation present in approximately 15% to 20% of de novo DLBCL cases, is only seen in the GCB subtype. So also are cases associated with amplification of the c-rel locus located at chromosome 2p. Ongoing somatic mutation of the immunoglobulin heavy chain locus (IgVH)-a physiologic event associated with the germinal center-is also exclusive to the GCB subtype. In contrast, bcl-2 expression, resulting predominantly from mechanisms other than translocation, is characteristic of the ABC subtype. These data lend credence to the molecular distinctions and suggest that gene expression profiling is, in fact, distinguishing morphologically similar lymphomas into biologically distinct subgroups- a distinction that has clinical impact. Thus, microarray analysis has revolutionized the taxonomy of NHL classification when analyzed in these retrospective studies. The role of these new technologies in prospective studies is untested, as is clinical decisionmaking when gene expression studies are used in "real time." It is entirely possible that discovery of new genes may be realized using microarray analyses, but the implementation of these tests in routine practice may be at the level of immunohistochemistry (IHC). The answers to these questions await future studies. However, ZAP70 expression in chronic lymphocytic leukemia is a perfect example of the clinical use of novel gene discovery being rapidly implemented in the clinic using IHC and flow cytometry. Conclusions
Lastly, a degree of caution is required before embracing microarray gene expression as the overarching final arbiter for the control of gene expression in human tumors. A steadystate measurement of mRNA is but one measure of gene expression-it is unlikely the sole mechanism for the control of gene expression. Proteomic approaches, particularly when combined with cDNA microarray, will undoubtedly lead to the discovery of other levels of gene expression control, including altered mRNA stability, translational control mechanisms, posttranslational modification, and intracellular targeting of nascent proteins. Surely these other levels of gene expresion control commonly used in lower organisms will also prove to be important in human tumors.
2. Rosenwald A, Wright G, Chan WC, et al: The use of molecular profiling to predict survival after chemotherapy for diffuse large-Bcell lymphoma. N Engl J Med 346:1937-1947, 2002.
3. Shipp MA, Ross KN, Tamayo P, et al: Diffuse large B-cell lymphoma outcome pre- diction by gene-expression profiling and supervised machine learning. Nat Med 8:68-74, 2002.
4. Staudt LM: Molecular diagnosis of the hematologic cancers. N Engl J Med 348:1777- 1785, 2003.
5. Davis RE, Brown KD, Siebenlist U, et al: Constitutive nuclear factor kappaB activity is required for survival of activated B cell-like diffuse large B cell lymphoma cells. J Exp Med 194:1861-1874, 2001.
6. Rosenwald A, Wright G, Wiestner A, et al: The proliferation gene expression signature is a quantitative integrator of oncogenic events that predicts survival in mantle cell lymphoma. Cancer Cell 3:185-197, 2003.
7. Devilard E, Bertucci F, Trempat P, et al: Gene expression profiling defines molecular subtypes of classical Hodgkin's disease. Oncogene 21:3095-3102, 2002.
8. Rosenwald A, Wright G, Leroy K, et al: Molecular diagnosis of primary mediastinal B cell lymphoma identifies a clinically favorable subgroup of diffuse large B cell lymphoma related to Hodgkin lymphoma. J Exp Med 198:851-862, 2003.
9. Rosenwald A, Alizadeh AA, Widhopf G, et al: Relation of gene expression phenotype to immunoglobulin mutation genotype in B cell chronic lymphocytic leukemia. J Exp Med 194:1639-1647, 2001.
10. Crespo M, Bosch F, Villamor N, et al: ZAP70 expression as a surrogate for immunoglobulin- variable-region mutations in chronic lymphocytic leukemia. N Engl J Med 348:1764-1775, 2003.