In situ immunoradiometric assay of fibrinogen adsorbed to artificial surfaces

Acute pulmonary embolism caused by highly aggregated intravenous immunoglobulin

BACKGROUND AND OBJECTIVES

Six patients died and one patient survived following infusion of a specific lot of intravenous immunoglobulin (IVIG) within half an hour in May 2008. This study elucidated the underlying pathogenesis.

MATERIALS AND METHODS

A variety of protein fractionation and identification approaches were employed to determine the abnormal components in IVIG products obtained from the hospital where the patients were treated. Animal studies using mice and monkeys were conducted to elucidate the pathophysiological mechanisms. In animal experiments, the effect and distribution of immunoglobulin was investigated using HE staining and immunohistochemistry (IHC) separately, while platelets and fibrinogen depletion were utilized to determine a possible link between thromboembolism formation in animals and the lethal effect of the IVIG. The size and distribution of the protein aggregates were determined with Coulter Counter Multisizer-3 after the dilution of the IVIG with plasma, and the lethal effect of the protein aggregates was simulated with artificial microparticles.

RESULTS

The IVIG retrieved from the hospital was found to have striking similarities to the heat-treated IVIG in terms of protein aggregation profiles and lethal effects. Post-mortem examination indicated that immunoglobulin aggregates were mainly found in the lung of the animals, while depletion of platelets and fibrinogen from the IVIG preparations failed to prevent the death of the animals. Similar amount of artificial microparticles caused animal death in similar fashion.

CONCLUSIONS

Our findings indicate that the retrieved IVIG exerted its lethal effects by blocking the pulmonary circulation without markedly altering the coagulation cascade or immunological events.

Adsorption-desorption of proteins on phospholipid polymer surfaces evaluated by dynamic contact angle measurement

Adsorption-desorption of plasma protein on various polymer membranes was evaluated by a dynamic contact angle (DCA) measurement using the Wilhelmy plate method. Poly(ethylene terephthalate) (PET) was used as a substrate membrane; we examined this membrane coated with hydrophilic polymers such as poly[2-methacryloyloxethyl phosphorylcholine (MPC)-co-n-butyl methacrylate (BMA)] or poly[2-hydroxyethyl methacrylate (HEMA)]. Although the zeta-potential of the PET membrane was negative, the coating with poly(MPC-co-BMA) induced increase of value to nearly zero. The DCA loops observed on the polymer membranes after protein adsorption were unity and hysteresis of the loop was reduced. In the cases of protein adsorbed on both the PET and the poly(HEMA) membranes, the shape and hysteresis of the loops were almost same during the rinsing process with a phosphate-buffered solution (PBS). However, the hysteresis of the DCA loops that appeared on the protein-adsorbed poly(MPC-co-BMA) membrane became large during the rinsing process with the PBS, and the shape of the DCA loop returned to its non-protein-adsorbed state. Therefore, proteins adsorbed on poly(MPC-co-BMA) could desorb more readily than those on PET and poly(HEMA) because of the weak interaction between poly(MPC-co-BMA) and the proteins.

Identification of plasma proteins adsorbed to hemodialyzers during clinical use

An investigation has been made of the protein layers formed on hemodialysis membranes during clinical use. Dialyzers having membranes of polymethylmethacrylate (PMMA), Cuprophane, cellulose acetate (CA), and saponified cellulose ester (SCE) were examined. Immediately following dialysis the dialyzers were washed free of blood and the membranes eluted with 2% SDS. The eluates were examined by SDS-PAGE followed by protein immunoblotting. Antisera to 16 common plasma proteins were used to probe for the presence of these proteins in the eluates. Most of the proteins tested for were found in the different eluates, suggesting that the protein layers are extremely complex. The protein compositions were qualitatively different on the different membranes. Except for HMWK the contact phase clotting factors were present in very small amounts and were largely activated. The clear presence of HMWK and the relatively small amounts of fibrinogen provide support for the occurrence of the Vroman effect. Fibrinogen was found to be degraded and this may be related to the observation that plasminogen was activated to plasmin. Complement C3 was an abundant component of all eluates. It was degraded to small fragments in a way which could not be related to complement activation. Many of the other proteins, particularly those of high molecular weight, were extensively degraded. It is speculated that this heretofore unremarked phenomenon may be due to the action of enzymes released by cell damage.

Immunochemical detection by specific antibody to thrombin of prothrombin conformational changes upon adsorption to artificial surfaces

Polyclonal antihuman alpha-thrombin antibodies produced in rabbits reacted minimally (less than 0.05%) in solution with human prothrombin. However, when prothrombin was adsorbed to artificial surfaces such as polyvinyl chloride (PVC), the cross-reactivity of surface-bound prothrombin with antibody IgG to thrombin (greater than 95% purity) was shown to be significantly enhanced. On PVC, the molar ratios of antibody IgG to thrombin/prothrombin approached the same level as that of antibody IgG to thrombin/thrombin when thrombin was adsorbed to the same material. The analyses of antigen-antibodies interaction, in solution with a direct binding assay by immune precipitation at high-speed centrifugation (160,000 g, 30 min), and on solid-phase PVC, were accomplished by use of double-labeling technique, i.e., 131I-thrombin (or 131I-prothrombin) and 125I-antibody IgG to thrombin. The results appear to suggest that prothrombin adsorption to PVC has resulted in some molecular conformational changes so that immunologically the adsorbed prothrombin resembles that of adsorbed thrombin on the same PVC surface.