Six patients with thrombotic microangiopathy associated with drug therapy had serial analyses of von Willebrand factor (vWF) multimeric patterns in their EDTA-plasma samples by sodium dodecyl sulfate-1% agarose gel electrophoresis and autoradiography. In the plasma of five patients (one with chronic myelogenous leukemia, two with prostatic cancer, and two with lymphoma), vWF abnormalities were observed during evolution of the thrombotic microangiopathy. These abnormalities were either the presence of unusually large (UL)vWF multimers of the type similar to those found within, and released or secreted by, endothelial cells (three patients) or a relative decrease in the largest plasma vWF multimers of the type that can be induced to attach to platelets (one patient) or both vWF abnormalities in different serial samples (one patient). In the one cardiac transplant patient who did not develop vWF multimeric abnormalities associated with thrombotic microangiopathy, vWF antigen levels were elevated more than threefold. This later individual received therapy with cyclosporin A alone. The other five thrombotic microangiopathy patients received cyclosporin A in combination with other chemotherapeutic agents (two patients); mitomycin-C, along with other chemotherapy (two patients); or multiple chemotherapeutic drugs, but not cyclosporin A or mitomycin C (one patient). The finding of vWF multimeric abnormalities during serial analysis of plasma samples from five of six patients with drug-associated thrombotic microangiopathy suggests the possibility that ULvWF forms derived from damaged or stimulated endothelial cells, along with the largest plasma vWF multimers, may be involved in the intravascular platelet clumping that is an essential part of the pathophysiology of this disorder.

Assessment of plasma vWF abnormalities by clinical coagulation laboratories is difficult because the available test systems for vWF antigen quantification and multimer analysis are expensive, laborious, and require days, radioactive anti-vWF antibodies and autoradiographic methods. We have devised simple, rapid, sensitive alternative techniques for vWF quantification and multimer analysis that can be readily installed in clinical laboratories. Plasma vWF antigen quantification is by a 2 hour enzyme immunoassay that accurately detects levels as low as 0.23% of normal. Plasma vWF to be quantified is bound to polyclonal monospecific antihuman vWF attached to small glass beads, and anti-human vWF conjugated with alkaline phosphatase is added to make an insoluble "sandwich." A substrate solution consisting of phenylphosphate and 4-amino-antipurine is added, followed by potassium ferricyanide. Optical density (at 490-510 nm) of the red color that develops is directly proportional to the plasma concentration of vWF antigen. Plasma vWF multimeric analysis is by a one-day electrophoretic immunobiot procedure. Plasma vWF multimer forms are solubilized in SDS-urea-Tris-EDTA, separated by horizontal 1% agarose gel electrophoresis, and transferred to a cationic membrane. Other protein binding sites on the membrane are blocked with milk proteins, and the membrane is overlaid with anti-vWF IgG linked to alkaline phosphatase. vWF multimers are then displayed as blue bands by soaking the membrane in an alkaline solution of the histochemical stain, fast blue RR (commonly used for leukocyte alkaline phosphatase scoring) dissolved in naphtol AS-MX phosphate. These simple, non-radioactive procedures performed together permit the rapid distinction of classical (Type I) von Willebrand's disease (vWD), characterized by low vWF antigen and normal multimers, from the Type II vWD syndromes, characterized by a relative deficiency of the largest plasma vWF forms. Unusually large vWF multimers, present in remission plasma of patients with chronic relapsing thrombotic thrombocytopenic purpura (TTP), are also easily detected using this rapid system of multimer analysis.