Risk assessment is a set of formal methods used to try to quantify risk associated with a defined activity. Generally it involves integrating all the information that can be found on the subject to gain some insight into the likely magnitude of the risks, and to identify areas of uncertainty where better information would help improve the assessment. After reports from the USA and Canada indicated that transmission via feed (either from ingredients or cross contamination) may be of concern in PEDV transmission, we employed a risk assessment approach to evaluate PEDV transmission in feed ingredients of porcine origin – namely rendered products; spray dried blood products; and peptone products derived by a hydrolysis process from pig intestines.  The project was coordinated by a multidisciplinary group at the University of Minnesota (UMN) together with a stakeholder working group that included a range of technical experts from the animal feed and swine industry.
The approach taken was to acquire data from multiple sources (industry, scientific literature, experimental studies and industry reports), to document likely limitations of the data sources, and to identify needs for further research. Flow charts were designed to describe the various processes used to manufacture the ingredients, and to identify points of likely virus inactivation. All the processes we examined included heating steps which play a role in virus inactivation, though other mechanisms are also involved. Most of the data used in the analyses were from very recent studies that are yet to be replicated.
To assess the assess the likelihood of PEDV survival in all three ingredient types (rendering, spray-dried plasma, peptones), recent experimental data on heat inactivation of PEDV in various feed samples and ingredients were used (from Pork Checkoff funded research by Dr. Sagar Goyal , University of Minnesota). Due to the prolonged high temperatures involved in the rendering and hydrolyzed protein processes, the risk of survival in these products was deemed to be negligible.  No other relevant data were available for these products. Spray drying involves shorter exposure times and lower temperatures than rendering or hydrolysis processes. Based solely on the experimental heat inactivation data it was estimated that spray drying represented a low risk of PEDV survival. However, it is known that other processes, in addition to heat, have a role in pathogen inactivation during spray drying. Also, the experimental model from which the heat inactivation data were derived did not directly mimic the spray drying process. Recently published information on inactivation of PEDV by spray drying was also used in assessing the likelihood of PEDV survival. These data, from experiments with laboratory scale spray dryers, indicated a substantial reduction of PEDV that resulted in no infection of either laboratory cell cultures or live pigs in two experiments.  However, laboratory scale processes may not exactly reflect the range of conditions in commercial scale spray dryers. In line with practices recently adopted in industry, the effect of post-processing storage of spray-dried plasma at room temperature (20-22°C) for two weeks was also estimated to achieve additional inactivation. Taking in consideration this post-processing storage step, the risk of PEDV survival was estimated to be negligible to low (PEDV survival predicted only if 100% of the virus in raw plasma were viable) based on heat inactivation alone, and negligible based on the experimental data from laboratory spray drying. To assess the likelihood for post-processing contamination of the finished ingredient with PEDV, site visits were performed at the processing plants for each of the ingredients.  A number of potential routes of cross contamination were identified, but most of these pathways were categorized as negligible to low risk.
Overall, these assessments were constrained by the limited availability of specific data. However, based on thermal inactivation alone, the risk of PEDV surviving the processes of rendering and hydrolysis (peptone production) are negligible. Because the time and temperature profiles used in spray-drying are less severe, the possibility of virus survival is inherently greater if non-thermal mechanisms are ignored. However, currently available data indicate that probability of PEDV surviving the spray-drying process and current post-processing storage periods is extremely small.  Remaining uncertainty could be addressed in the future with better knowledge of 1) viral inactivation by spray drying, 2) the infectious dose of PEDV, and 3) the relationship between measures of virus RNA and infective dose.