Hematopoietic cell destruction by immune mechanisms in acquired aplastic anemia

D ESPITE THE SIMPLEST of pathologies-the empty bone marrow-the pathophysiology of aplastic anemia is complicated. Complexity can be inferred from the diversity of putative etiologic associations, from drugs and chemicals to viruses and pregnancy. Clinically, the disease may run a variety of courses, from acutely explosive and rapidly catastrophic, as in El-h molecular lesions that now define mouse models of constitutional marrow failure are not present in acquired human aplastic anemia. A diversity of mechanisms might be responsible for idiosyncratic drug-associated marrow failure, the largest category of aplastic anemia other than idiopathic disease, but such drug-induced aplastic anemia is indistinguishable from idiopathic disease on clinical grounds.* Drugs used in cancer chemotherapy are selected for their cytotoxicity and their regular, dosedependent induction of transient marrow aplasia is expected, but there is a paucity of evidence that the typically rare drug reactions associated with marrow failure are by direct toxicity Idiosyncratic reactions are extremely diI?icult to study in the laboratory’* and simple serum assays are unhelpful because antibodies to either drugs or cells have only occasionally been identified in aplastic anemia.3* Finally, it should be recalled that even massive doses of drugs that were designed or selected for their toxicity for mammalian cells may be not be fully myeloablative. (For a recent review of chemical and drugs as agents in aplastic anemia, see Young.86) In contrast, the immune model provides a powerful prism through which to focus unanswered questions in the etiology and consequences of marrow failure. Currently, the characterization of aplastic anemia as an autoimmune disease helps explain the response of marrow failure to immunosuppressive treatments and many laboratory abnormalities in immune system function. The model must be able to explain known features of the disease, such as the severity of marrow damage and the rarity of disease, and it should have heuristic power to guide future laboratory investigations and to design new therapeutic approaches. These challenges are shared with other autoimmune processes characterized by organ-specific damage. While in some cases possible initiating antigens have been inferred from animal models or serologic studies in humans (myelin basic protein for multiple sclerosis, islet cell proteins for type I diabetes mellitus, keratin in uveitis), hematologists have the great advantage of facile experiiental manipulation of target and effector cells, and clinicians have proven successful in the application of immunologic therapies to patients with marrow failure syndromes.