A probabilistic model for analyzing viral risks of plasma‐derived medicinal products

Abstract

BACKGROUND: The prevention of transmission of viral infections by plasma-derived medicinal products is of concern to manufacturers, legislators, and patient representative groups. Recent European legislation requires a viral risk assessment for all new marketing applications of such products. STUDY DESIGN AND METHODS: A discrete event Monte Carlo model was developed to determine the viral transmission risks of the plasma-derived medicinal products. The model incorporates donor epidemiology, donation intervals, efficiency of screening tests for viral markers, inventory hold period, size and composition of the manufacturing pool, production time, process virus reduction capacity, and product yield. With the model, the human immunodeficiency virus (HIV) and hepatitis C virus (HCV) contamination risks of a typical hypothetical plasma product were calculated, and the sensitivity of the risk to various model variables was analyzed. RESULTS: The residual HIV and HCV risks of the finished products are linear in change with viral incidence rate and inversely linear with product yield and process virus reduction capacity. For the product analyzed in this article, the residual risk is less sensitive to changes in screening test pool size, donation frequency, and inventory hold period. There is only a limited dependency on the donation type (apheresis or whole-blood donations) and a negligible dependency on the manufacturing pool size. CONCLUSIONS: The use of probabilistic model simulation techniques is indispensable when estimating realistic residual viral risks of plasma-derived medicinal products. In contrast to conventional deterministic residual risk estimations, the probabilistic approach allows incorporation of specific manufacturing decisions and therefore provides the only feasible alternative for a correct assessment of residual risks.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 19 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.