Comparison of nanofiltration efficacy in reducing infectivity of centrifuged versus ultracentrifuged 263K scrapie-infected brain homogenates in "spiked" albumin solutions.

ABSTRACT

BACKGROUND: The safety of plasma-derived products is of concern for possible transmission of variant Creutzfeldt-Jakob disease. The absence of validated screening tests requires the use of procedures to remove or inactivate prions during the manufacture of plasma-derived products to minimize the risk of transmission. These procedures need proper validation studies based on spiking human plasma or intermediate fractions of plasma fractionation with prions in a form as close as possible to that present in blood. STUDY DESIGN AND METHODS: Human albumin was spiked with low-speed or high-speed supernatants of 263K scrapie-infected hamster brain homogenates. Spiked albumin was then passed through a cascade of filters from 100 nm down to 20 to 15 nm. Residual infectivity was measured by bioassay. RESULTS: The overall removal of infectivity spiked into albumin through serial nanofiltration steps was 4 to 5 logs using low-speed supernatant and 2 to 3 logs with high-speed supernatant. CONCLUSION: These findings confirm the utility of nanofiltration in removing infectivity from plasma (or other products) spiked with scrapie brain homogenate supernatants. However, efficiency is diminished using supernatants that have been ultracentrifuged to reduce aggregated forms of the infectious agent. Thus, filtration removal data based on experiments using "standard" low-speed centrifugation supernatants might overestimate the amount of prion removal in plasma or urine-derived therapeutic products.