Inactivation of hepatitis A variants during heat treatment (pasteurization) of human serum albumin

BACKGROUND: Pasteurization of human serum albumin (HSA) is detailed in the US and European Pharmacopoeial monographs and therefore a process that allows for little variation in physiochemical variables. Nevertheless, differences of up to 3.9 log in hepatitis A virus (HAV) inactivation by pasteurization have been reported. Here, the hypothesis that the choice of HAV variant used in the pasteurization might contribute to this inactivation variability is evaluated experimentally. STUDY DESIGN AND METHODS: The identity of four widely used cytopathic variants of the original HAV HM175 strain was determined by partial sequencing. These variants were used in pasteurization studies conducted under the principles of good laboratory practice, for which HAV-spiked HSA of 5 or 25% protein content was kept at 58 +/- 1 degrees C for 600 +/- 10 minutes, and the virus inactivation was assessed. In addition, data from previous pasteurization studies were included in the analysis. RESULTS: The four HAV variants could be divided into two subgroups, with significantly different (p <= 0.0001) virus inactivation by pasteurization (4.7 and 4.8 log vs. 2.3 and 2.6 log, respectively). Also, the protein concentration of the HSA solution used for pasteurization had a significant effect on the achieved HAV inactivation, with reduction factors obtained in 5% HSA significantly lower than in 25% HSA (p < 0.002). CONCLUSION: HAV variant and protein concentration of the HSA solution affect the overall HAV inactivation that is achieved during pasteurization. As the HAV inactivation capacity should not be overestimated, an HAV variant more resistant to heat inactivation should be used for studies investigating the viral safety profiles of plasma derivatives.