Myelofibrosis—a disease (or diseases) with variable degrees of anemia, splenomegaly, fibrotic bone marrow, and myeloid metaplasia—eludes precise definition. Theories on pathogcnesis range from necrobiosis (due to the toxic action of benzene, radioactive phosphorus, radium, and other agents) and primary bone-marrow failure (in which myeloid metaplasia is viewed as a compensatory process), to the current, widely held theory that it is but one of a group of myeloproliferative disorders. Proponents of this latter view regard intramedullary and extramedullary changes as stemming from a common "myelostimulatory factor." Depending on the phase of the disease, or the form it takes in a particular patient, the myeloproliferative disorder might be manifested as polycythemia vera, myelogenous leukemia, erythroleukemia of di Guglielmo, or myelofibrosis.
The so-called myelostimulatory factor, having also eluded precise definition, has remained, to a large extent, an abstraction, a convenient rack which can accommodate many hats. However, Anderson, Hoshino, and Yamamoto,1 in