This study will use genomics-based technology, such as DNA microarrays, to more precisely diagnose subsets of lymphoma, leukemia and multiple myeloma patients. There have been many attempts to classify lymphoid cancers in ways that will be useful for clinical diagnosis and treatment. Although broad diagnostic categories have been reliably defined, patients within each category have distinct clinical courses, suggesting that these classifications could be further divided into molecular (genetic) subtypes. For example, 40 percent of patients with diffuse large B-cell lymphoma achieve long-term disease remissions following combination chemotherapy and are apparently cured, whereas the remaining 60 percent die from the disease. Similarly, some patients with follicular lymphoma develop aggressive disease within a few years of diagnosis, while others have stable disease over 10 to 20 years. Although the distinctions in clinical course of these diseases are recognized, there are no studies to determine the molecular (genetic) basis for this variability. This study will try to define new molecular diagnostic categories in these diseases and correlate them with clinical features, including treatment response, disease remission and overall survival following chemotherapy.

This retrospective study will use clinical data and tissue samples from participating centers in the Lymphoma/Leukemia Molecular Profiling Project LLMPP). New patients will not be recruited for this study.

Biopsy materials, including fresh frozen or OTC-embedded lymphoma biopsy material, viably frozen samples of peripheral blood cells from leukemia patients, and viably frozen samples of bone marrow aspirates from multiple myeloma patients will be collected from pathologists participating in the LLMPP. RNA and genomic DNA will be extracted from the tumor samples. A variety of technologies will be used to characterize the genome of the cancer cells, including lymphochip microarrays for array-based comparative genomic hybridization; Southern blotting and PCR for translocation of genes previously implicated in these malignancies; and PCR and DNA sequencing methods for analyzing base changes in the genome of the cancer cells. Clinical information from the initial diagnosis to disease relapse will be taken from existing databases and/or patient charts. Gene expression will be correlated with the clinical data. If a small number of genes is found to strongly predict clinical outcome, quantitative RT-PCR assays using the Taqman technology may be developed as an alternative to DNA microarray analysis.