Seven animal drugs [penicillin G (PENG), sulfadimethoxine (SDMX), oxytetracycline (OTET), erythromycin (ERY), ketoprofen (KETO), thiabendazole (THIA), and ivermectin (IVR)] were used to evaluate the drug distribution between milk fat and skim milk fractions of cow milk. More than 90% of the radioactivity was distributed into the skim milk fraction for ERY, KETO, OTET, PENG, and SDMX, approximately 80% for THIA, and 13% for IVR. The distribution of drug between milk fat and skim milk fractions was significantly correlated to the drug's lipophilicity (partition coefficient, log P, or distribution coefficient, log D, which includes ionization). Data were fit with linear mixed effects models; the best fit was obtained within this data set with log D versus observed drug distribution ratios. These candidate empirical models serve for assisting to predict the distribution and concentration of these drugs in a variety of milk and milk products.

It is important to understand the partitioning of drugs in processed milk and milk products, when drugs are present in raw milk, in order to estimate the potential consumer exposure. Radioisotopically labeled erythromycin, ivermectin, ketoprofen, oxytetracycline, penicillin G, sulfadimethoxine, and thiabendazole were used to evaluate the distribution of animal drugs among rennet curd, whey, and protein fractions from skim cow milk. Our previous work reported the distribution of these same drugs between skim and fat fractions of milk. Drug distribution between curd and whey was significantly correlated (R = 0.70) to the drug's lipophilicity (log P), with improved correlation using log D (R = 0.95). Distribution of drugs was concentration independent over the range tested (20-2000 nM). With the exception of thiabendazole and ivermectin, more drug was associated with whey protein than casein on a nmol/g protein basis (oxytetracycline experiment not performed). These results provide insights into the distribution of animal drug residues, if present in cow milk, among milk fractions, with possible extrapolation to milk products.

Norovirus (NoV) infections are the leading cause of foodborne illness in the United States. Effective disinfection is important for controlling outbreaks caused by this highly infectious virus but can be difficult to achieve because NoV is very resistant to many common disinfection protocols. The inability of human NoV to replicate in tissue culture complicates NoV research, generally necessitating genome copy quantification, the use of surrogate viruses, or the use of other substitutes such as virus-like particles. To date, comprehensive comparisons among NoV surrogates and between surrogates and human NoV are missing, and it is not clear how best to extrapolate information from surrogate data. We conducted a systematic review and meta-analysis of comparisons of NoV surrogates with regard to their susceptibility to disinfection on hard surfaces or in suspension. Restricting our analysis to those studies in which two or more virus surrogates were compared allowed us to improve the signal-to-noise ratio in our analysis, similar to the epidemiological concept of matching. Using meta-analysis methods, our results indicate that hepatitis A virus, murine norovirus 1, and phage MS2 are significantly more resistant to disinfection than is feline calicivirus, but average differences in viral titer reduction appeared to be modest, 1.5 log PFU or less in all cases. None of the studies that compared surrogates and human NoV met our inclusion criteria, precluding a direct comparison between human NoV and NoV surrogates in this study. For all surrogates with sufficient data available to permit subgroup analyses, we detected strong evidence that the type of disinfectant impacted the relative susceptibility of the surrogates. Therefore, extrapolation of results between surrogates or from surrogates to human NoV must consider the type of disinfectant studied.