Plasma cell myeloma is a more complex neoplasm than suggested by the relative uniformity of its dominant plasma cells, which represent the terminal stage of normal B-cell differentiation. Phenotypic, molecular, and cellular genetic data favor the presence of a myeloma stem cell early in hematopoietic development so that, as in chronic myelogenous leukemia (CML), a far distance exists between the primordial malignant cell that was the target of malignant transformation and the dominant clinical phenotype. Traces of pre-B, myeloid, and T cells are coexpressed with the mature B-cell phenotype, an occurrence unknown in normal B-cell differentiation. Analogous to CML, disease progression is marked by disease dedifferentiation, occasionally with cessation of myeloma protein production and development instead of extramedullary lymphomalike features with high LDH or myelodysplasia/acute myelogenous leukemia (AML) syndromes. The prognostic importance of serum LDH levels even in newly diagnosed myeloma suggests the early presence of tumor cells with “LDH phenotype,” which, as a result of drug resistance and proliferative advantage, expand preferentially during disease progression. Further characterization of these cells may provide important clues about the ontogeny of multiple myeloma. Myeloma cells express many receptors for different biological signals that might be exploitable for therapy with immunotoxins or radioisotopes. Plasma cells and their precursors also produce a variety of cytokines, some of which have putatively autostimulatory functions (eg, IL-1, IL-5, IL-6) and/or are related to disease manifestations (eg, IL-1 and TNF-beta as OAF). The wealth of cellular expression by plasma cells provides clues for understanding the mechanisms of gene activation and the nature of abnormal growth and differentiation. The accuracy of prognostically relevant staging systems has been refined with the use of new quantitative parameters that reflect tumor mass (ie, serum B2M levels) and biology. Further studies of cellular and molecular biology (ie, CAL- LA, H-ras) may reveal those tumor cell features that define clinical entities, response to therapy, and long-term prognosis. The lack of a major advance in prognosis despite the use of more drugs and more intensive regimens justifies the continued use of standard melphalan- prednisone for patients with a highly favorable prognosis, for the very aged, and for those with a short life expectancy due to other major medical problems. However, a radical departure from standard practice is required to improve the prognosis for younger patients with poor risk features.(ABSTRACT TRUNCATED AT 400 WORDS)