Microarrays in Diffuse Large B-Cell Lymphoma Chen Lossos Biochemistry 118Q Professor Brutlag December 3, 2009Diffuse large B-cell lymphoma (DLBCL) is an aggressive malignancy of mature B lymphocytes, with more than 25,000 new occurrences in the United States every year and accounting for nearly 40% of Non-Hodgkins Lymphoma cases. Typical presentation of DLBCL includes rapidly enlarging lymph nodes or tumors in extra-nodal sites. Night sweats, fever and weight loss are observed in approximately 30% of patients. Patients with DLBCL have highly variable clinical courses: although most patients initially respond to chemotherapy, fewer than half of the patients achieve a durable remission (Fauci). Clinical prognostic models, such as the International Prognostic Index (IPI) have been developed to identify DLBCL patients who are unlikely to be cured with standard therapy. The clinical factors which constitutes IPI (age, performance status, stage of disease, number of extranodal sites and serum lactate dehydrogenase levels), however, frequently fail to accurately predict patient survival. This, according to Schipp et al. suggests that there exist underlying biological mechanisms which are responsible for the observed differences of survival of patients with similar IPI scores (Shipp). Like many cancers, the biological mechanisms underlying DLBCL pathogenesis are markedly intricate, involving relationships between numerous genes, signaling pathways, and regulatory processes. Therefore, to study a single gene to identify disease cause and pathogenesis does not accurately reflect gene expression changes. Instead, a molecular technique capable of evaluating multiple components of the biological processes of tumorgenesis and pathogenesis is necessary to fully understand these processes in DLBCL. The advent of DNA microarray technology has done just this. Microarrays have not only allowed for the development of better diagnostic and treatment techniques of DLBCL, but have also illuminated the pathogenesis of the disease. Most importantly, microarrays have allowed for the subclassification of DLBCL intotwo distinct categories with unique clinical outcomes and survival: Germinal Center Diffuse Large B-Cell Lymphoma (GC-DLBCL) and Activated peripheral Blood B-Cell Diffuse Large B-Cell Lymphoma (ABC-DLBCL). The first microarray study of DLBCL, carried out at Stanford by Alizadeh et al. provided the strongest initial evidence for the subcategorization of DLBCL. This study utilized a specialized cDNA array called Lymphochip, which was constructed by selecting genes expressed in lymphoid cells or which were previously reported to be involved in cancer biology. Using an unsupervised hierarchical clustering method (i.e. microarray data are analyzed without using external information), 42 tumors from patients treated with anthracycline based chemotherapy revealed two distinct subgroups of DLBCL based on gene expression. The expression patterns were characteristic of normal germinal center (GC) B-cells and activated peripheral blood B-cells. In addition to these categories, expression signatures (genes of similar function which cluster together) were observed in genes associated with proliferation and lymph nodes. The strongest evidence for the categorization of DLBCL into the two subtypes, however, came from patients’ survival rate. Although the average five-year survival for all patients was 52%, 76% of GC B-like DLBCL patients alive after five years, as compared with only 16% of activated B-like DLBCL patients. Thus, this difference in patient survival based on the GC and ABC classification system provides a possible explanation for the observed differences in survival between patients with similar IPI scores. Thus, the molecular differences between the two patient groups were accompanied by a remarkable divergence in clinical behavior, suggesting that GC B-like DLBCL and activated B cell DLBCL should be regarded as distinct diseases. Although this data indicated for the need to subclassify DLBCL into GC-DLBCL and ABC-DLBCL, there remained a need to validate this classification in independent sets of patientsor to confirm the findings by other biological parameters. The support needed for these DLBCL subgroups was provided by a study performed by the Lymphoma/Leukemia Molecular Profiling Project Group, which analyzed tumors from 240 DLBCL patients treated with anthracycline-based chemotherapy. Using a similar methodology to Alizadeh et al. (a cDNA Lymphochip and unsupervised clustering), this study confirmed the clustering of patients into GC and ABC-like DLBCL categories. Moreover, the clinical impact of this molecular classification was confirmed, as a much better prognosis was seen in patients with GC-like DLBCL (5-year survival of 60%) compared with ABC (35%) (Rosenwald et al). The subdivision into GC-DLBCL and ABC-DLBCL were further confirmed by biological presentations of patients. The mutation of the variable heavy (VH) chain of the Ig gene, frequently rearranged and over-expressed in B cell lymphomas, provided the opportunity to do this. 14 DLBCL patient tumor samples were clustered by hierarchical clustering using the Lymphochip. 7 of the patients occupied a common branch and showed a gene expression profile similar to the GC-DLBCL seen in Alizadeh et al. All 7 of these patients displayed the mutation of the Ig gene. In contrast, none of the 7 patients who exhibited ABC-DLBCL characteristic expression displayed ongoing mutations in the Ig gene. This suggests that GC-DLBCL and ABC-DLBCL originate from different ontogeny lymphocyte precursors (Lossos et al PNAS). In addition to the mutation of the Ig gene, other findings substantiated the need to subdivide DLBCL into two entities. The t(14;18)(q32;q21) translocation, which involves the BCL-2 gene and amplification of the c-rel locus on chromosome 2p, was detected almost exclusively in Germinal Center-like DLBCL. Furthermore, Bea et al. found that trisomy 3, gains of chromosomal regions 3q and 18q21-q22, and losses of 6q21-22 were associated with ABC-like DLBCL. Further, GC-like DLBCL had frequent gains of 12q12. A parallel analysis yieldedthat these DNA amplifications were strongly correlated with impact on genes whose expression is involved chromosomal regions in a subgroup-specific fashion (Huang et al). These findings strengthened the need to subdivide DLBCL into GC-like and ABC-like subtypes. Despite the advances that
View Full Document
Unlocking...