To review recent advances in research on per- and polyfluoroalkyl substances (PFAS) and outline priority steps for risk assessment in consumer health protection, the German Federal Institute for Risk Assessment (BfR) organized the 'International PFAS Conference' in Berlin in October 2025. Building on the European Food Safety Authority's (EFSA) opinion in 2020, global research activities on PFAS have intensified. The conference was attended by 200 participants from 18 countries and covered topics such as analytical methods, human exposure, toxicokinetics, toxicity, and future perspectives. Given that there are more than 21,000 different PFAS in use, discussions highlighted the need for further data collection and a basis for prioritizing substances. Robust exposure assessment requires improved analytical methods combined with newly developed predictive tools to quantify, identify, and make better use of non-target data. Hazard characterization may benefit from the combined use of classic experimental data sets, epidemiological data, and new approach methodologies (NAMs) data. The participants emphasized the continuous need for refined data and proposed a systematic consolidation of global data on shared data platforms, accompanied by corresponding guidelines for data usage. In the final panel discussion, effective risk communication was identified as a critical challenge, necessitating clear and consistent messaging for the public and policymakers. The conference concluded with five key recommendations for future health risk assessments: prioritizing PFAS; improving analytical methods; promoting generation of robust data and data gap filling; promoting open science, including the development of shared data infrastructure and cross-institutional knowledge exchange; and strengthening risk communication. These recommendations aim to support transparent, evidence-based and effective PFAS consumer health risk management in the decades ahead.

Aflatoxin B1 (AFB1) is a hepatocarcinogenic mycotoxin found in animal feed contaminated by mold from the genus Aspergillus. It can be transferred into food of animal origin, such as the milk of dairy cows, as the metabolite aflatoxin M1 (AFM1), which has 10

Polychlorinated dibenzo-p-dioxins (PCDDs) and dibenzofurans (PCDFs), commonly known as “dioxins”, and polychlorinated biphenyls (PCBs) are persistent, environmentally ubiquitous lipophilic contaminants. Most human exposure comes from animal products, such as milk and meat. In this study, a whole-body physiologically-based toxicokinetic (wbPBTK) model is presented for the quantitative prediction of the transfer of these contaminants to milk and multiple tissues of high-yielding dairy cows. The model is parametrized using a comprehensive in vivo study in which cows were orally exposed to 35 dioxin and PCB congeners. To describe the transfer to cow tissues, the wbPBTK model incorporates dynamic changes in the physiological characteristics across lactation and comprises 11 compartments, including blood, adipose tissues, liver and udder. Congener-specific parameters were derived from the data based on a novel modular Bayesian inference approach designed to handle the numerical challenges of this experimental dataset, such as the imbalance between numerous milk measurements and comparatively few tissue measurements. A final model was derived for 29 congeners, showing at least reasonable performance in describing milk data for 28, adipose tissue data during lactation for 27, blood data for 21, and organ tissue data at slaughter for 21 congeners. It includes an estimation of the prediction uncertainties using credible intervals. A code implementation for German Holstein cows is included, which may be modified for other breeds, ages or milk yields. The model is also available for interactive calculation in the ConTrans.bfr.bund.de platform tool to support exposure assessment and risk management for dioxins and PCBs in dairy production.

Per- and polyfluoroalkyl substances (PFAS) constitute a group of anthropogenic compounds that are ubiquitously distributed in the environment due to their widespread use in various industrial processes and consumer products. Among foodstuffs of animal origin, wild boar (Sus scrofa) livers may contain particularly high levels of PFAS. Additionally, wild boars have been identified as suitable bioindicators and are used by German national authorities for monitoring purposes. In the context of consumer health risk assessment, information on influencing environmental factors is crucial for the risk analysis process. The objective of this study was to systematically analyze reports from the national monitoring program on PFAS concentrations in wild boar livers from 3 German federal states (n = 217). The analysis was based on publicly available data on land use characteristics, including urbanization, industry, agriculture, and forest areas in the sampled regions. Overall, data on sample characteristics (e.g., age, sex, body weight) varied between and within Federal States and counties. Despite these inconsistencies, statistical analyses revealed significant differences in the average PFAS concentrations between counties within the Federal States. Additionally, significant correlations were observed between land use features and PFAS concentrations, varying by Federal State and substance. In conclusion, land-use-based evaluations of food monitoring data may prove valuable for characterizing PFAS levels in wild boar liver. Harmonizing sampling protocols, data collection, and analytical methods among state laboratories will improve the comparability and interpretation of official monitoring data in future assessments.

Understanding the transfer of polychlorinated dibenzo-p-dioxins and dibenzofurans (PCDD/Fs) in farm animals is essential for ensuring food safety, but such information for suckler ewes (Ovis aries) has been lacking. This work quantifies the accumulation, tissue distribution, and depuration kinetics of PCDD/Fs in these animals. Six suckler ewes (EXP group) were exposed to PCDD/Fs through contaminated hay (2.3-12.7 ng toxic-equivalent kg-1 dry matter) and then allowed to depurate by switching to noncontaminated hay from 29 days of lactation. Four control ewes were fed continuously with noncontaminated hay. At different time points covering depuration, weaning and slaughter, PCDD/F analysis of milk (three time points), blood and sternal adipose tissue (five time points), Longissimus thoracis muscle, liver, and empty body homogenate at slaughter (188 days of depuration) was performed. A relevant PCDD/F bioaccumulation was observed from oral intake in milk and adipose tissue (biotransfer factors of 1.24 and 1.06 day kg-1 lipids for the sum toxic-equivalent, respectively) in the EXP ewes, especially for penta- and hexa-chlorinated congeners. The EXP ewes' adipose tissue started at 10-fold the EU maximum level (ML) and showed depuration below the ML after 130 days. Specific PCDD/F accumulation in the ewe liver was observed, especially for dibenzofurans. These toxicokinetic data can inform recommendations to ensure the chemical safety of sheep food products.

Polychlorinated biphenyls (PCBs) are persistent environmental pollutants that accumulate in tissues of exposed animals and humans. This case report refers ton=3 dairy cows accidentally exposed to non-dioxin-like PCBs (ndl-PCBs) of unknown origin on a German farm. At study start they had a cumulative total of 122-643 ng/g fat in milk and 105-591 ng/g fat in blood, consisting mainly of PCBs 138, 153, and 180. Two cows calved during the study and their calves were raised on their mothers' milk, resulting in cumulative exposure until slaughter. A physiologically based toxicokinetic model was developed to describe the fate of ndl-PCBs in the animals. The toxicokinetic behavior of ndl-PCBs was simulated in individual animals, including transfer of contaminants into calves via milk and placenta. Both the simulations and experimental data indicate that contamination via both routes is significant. In addition, the model was used to estimate kinetic parameters for risk assessment.

With increasing global focus on planetary boundaries, food safety and quality, the presence of per-and poly-fluoroalkyl substances (PFAS) in the food chain presents a challenge for the sustainable production and supply of quality assured food. Consumption of food is the primary PFAS exposure route for the general population. At contaminated sites, PFAS have been reported in a range of agricultural commodities including cattle. Consumer exposure assessments are complicated by the lack of validated modelling approaches to estimate PFAS bio-accumulation in cattle. Previous studies have shown that PFAS bioaccumulation in livestock is influenced by environmental, spatial and temporal factors that necessitate a dynamic modelling approach.This work presents an integrated exposure and population toxicokinetic (PopTK) model for cattle that esti-mates serum and tissue concentrations of PFAS over time. Daily exposures were estimated from intakes of water, pasture, and soil, and considered animal growth, seasonal variability (pasture moisture content and temperature) and variable PFAS concentrations across paddocks. Modelled serum and tissue estimates were validated against monitoring data from Australian and Swedish cattle farms. The models were also used to develop and test practical management options for reducing PFAS exposure and to prioritise remediation for farms. Model outputs for exposure management scenarios (testing cattle rotation and targeted supplementation of feed and water) showed potential for marked reductions in consumer exposures from cattle produce.

Understanding the transfer of polychlorinated dibenzo-p-dioxins (PCDDs) and dibenzofurans (PCDFs) as well as polychlorinated biphenyls (PCBs) from oral exposure into cow's milk is not purely an experimental endeavour, as it has produced a large corpus of theoretical work. This work consists of a variety of predictive toxicokinetic models in the realms of health and environmental risk assessment and risk management. Their purpose is to provide mathematical predictive tools to organise and integrate knowledge on the absorption, distribution, metabolism and excretion processes. Toxicokinetic models are based on more than 50 years of transfer studies summarised in part I of this review series. Here in part II, several of these models are described and systematically classified with a focus on their applicability to risk analysis as well as their limitations. This part of the review highlights the opportunities and challenges along the way towards accurate, congener-specific predictive models applicable to changing animal breeds and husbandry conditions.

The use of lead-based rifle bullets in hunting poses a risk to human and animal health when bullet fragments remain in the game meat. The objective of this study was to assess, for the first time, the number, size and spatial distribution of bullet fragments in game animals collectively and in three dimensions. Four roe deer (Capreolus capreolus) hunted with lead bullets were radiographed using computed tomography (CT) at each step of meat processing from animal bodies to edible parts. The animal bodies contained 43–199 fragments with a mean volume of 3.71 ± 5.49 mm 3 and a mean distance of 6.5 ± 4.6 cm perpendicular to the wound channel. About 40% of the fragments were in the lower size range of < 1.00 mm 3 . Individual fragments were located up to 22.2 cm from the wound channel and were located in the edible parts (ribs, n = 2; haunch, n = 1; shoulder, n = 3) with up to 0.29 g (0.16 ± 0.15 g, mean ± SD) of estimated bullet fragment mass (BFM) deposited in the edible parts of the roe deer. Tissues in a radius of at least 16 cm around the wound channel should have been discarded in order to remove 95% of the BFM from the investigated roe deer. Additionally, around 85.1% of the initial bullet mass corresponding to 9.948 ± 1.040 g BFM were estimated to be introduced into the environment. This study highlights the challenges of ensuring sufficient removal of lead-based rifle bullet fragments in game meat processing for either human consumption purposes or for use as pet food.

Dioxin-like polychlorinated biphenyls (dl-PCBs) as well as polychlorinated dibenzo-p-dioxins (PCDDs) and dibenzofurans (PCDFs) are a major concern for food safety, especially in fat-containing foods of animal origin, such as milk. Due to the lipophilic character of PCDD/Fs and PCBs, it is of special interest to explore whether the metabolic state of high-yielding cows influences the transfer rates into milk. Five German Holstein cows were orally exposed to a mixture of 17 PCDD/Fs, 12 dl-PCBs, and 6 non-dioxin-like PCBs (ndl-PCBs) for two dosing periods of 28 days each. The first period covered the negative energy balance (NEB) after calving, while the second period addressed the positive energy balance (PEB) in late lactation. Each dosing period was followed by a depuration period of around 100 days. During the NEB phase, the transfer rates of 14 PCDD/Fs and 7 dl-PCBs quantified were significantly (p ≤ 0.1) higher compared to the PEB phase, indicating an influence of the metabolic state on the transfer. Furthermore, the congener-specific transfer rates (0.3-39%) were in the range of the results from former studies. This indicates that the milk yield of the exposed cows is not the only determining factor for the transfer of these congeners into milk.

Non-dioxin-like polychlorinated biphenyls (ndl-PCBs) are persistent environmental pollutants that accumulate in the tissues of exposed animals and humans. Contaminated feed can lead to ndl-PCB contaminated food of animal origin; such foods are the main route of human exposure. Therefore, predicting ndl-PCB transfer from feed into animal products is important for human health risk assessment. Here, we developed a physiologically based toxicokinetic model describing the transfer of PCBs-28, 52, 101, 138, 153 and 180 from contaminated feed into the liver and fat of fattening pigs. The model is based on a feeding study with fattening pigs (PIC hybrids) that were temporarily fed contaminated feed containing known concentrations of ndl-PCBs. Animals were slaughtered at different ages, and ndl-PCB concentrations in muscle fat and liver were determined. The model accounts for animal growth and excretion via the liver. Based on their elimination speed and half-lives, they can be categorized into fast (PCB-28), intermediate (PCBs 52 and 101) and slow (PCBs 138, 153 and 180). Using a simulation with realistic growth and feeding patterns, the following transfer rates were found: 10 % (for fast), 35-39 % (intermediate) and 71-77 % (slow eliminated congeners). Using the models, the highest level of 3.8 μg/kg dry matter (DM) was calculated for any sum of ndl-PCBs in pig feed to ensure that the current maximum levels in pork meat and liver (40 ng/g fat) are not be exceeded. The model is included in the Supplementary Material.

Polychlorinated dibenzo-para-dioxins (PCDDs) and dibenzofurans (PCDFs) (collectively and colloquially referred to as 'dioxins') as well as polychlorinated biphenyls (PCBs) are persistent and ubiquitous environmental contaminants that may unintentionally enter and accumulate along the food chain. Owing to their chronic toxic effects in humans and bioaccumulative properties, their presence in feed and food requires particular attention. One important exposure pathway for consumers is consumption of milk and dairy products. Their transfer from feed to milk has been studied for the past 50 years to quantify the uptake and elimination kinetics. We extracted transfer parameters (transfer rate, transfer factor, biotransfer factor and elimination half-lives) in a machine-readable format from seventy-six primary and twenty-nine secondary literature items. Kinetic data for some toxicologically relevant dioxin congeners and the elimination half-lives of dioxin-like PCBs are still not available. A well-defined selection of transfer parameters from literature was statistically analysed and shown to display high variability. To understand this variability, we discuss the data with an emphasis on influencing factors, such as experimental conditions, cow performance parameters and metabolic state. While no universal interpretation could be derived, a tendency for increased transfer into milk is apparently connected to an increase in milk yield and milk fat yield as well as during times of body fat mobilisation, for example during the negative energy balance after calving. Over the past decades, milk yield has increased to over 40 kg/d during high lactation, so more research is needed on how this impacts feed to food transfer for PCDD/Fs and PCBs.

The industrial hemp sector is growing and, in recent years, has launched many novel hemp-derived products, including animal feed. It is, however, unclear to what extent individual cannabinoids from industrial hemp transfer from the feed into products of animal origin and whether they pose a risk for the consumer. Here we present the results of a feeding experiment with industrial hemp silage in dairy cows. Hemp feeding included changes in feed intake, milk yield, respiratory and heart rates, and behaviour. We combined liquid chromatography–tandem mass spectrometry-based analyses and toxicokinetic computer modelling to estimate the transfer of several cannabinoids (Δ 9 -tetrahydrocannabinol (Δ 9 -THC), Δ 8 -THC, Δ 9 -tetrahydrocannabinolic acid, Δ 9 -tetrahydrocannabivarin, 11-OH-Δ 9 -THC, 11-nor-9-carboxy-Δ 9 -THC, cannabidiol, cannabinol and cannabidivarin) from animal feed to milk. For Δ 9 -THC, which has a feed-to-milk transfer rate of 0.20% ± 0.03%, the acute reference dose for humans was exceeded in several consumer groups in exposure scenarios for milk and dairy product consumption when using industrial hemp to feed dairy cows.

Non-dioxin-like polychlorinated biphenyls (ndl-PCBs) are a subclass of persistent bioaccumulative pollutants able to enter the food chain. Toxicokinetic models for the transfer of the six ndl-PCB congeners (PCBs 28, 52, 101, 138, 153, and 180) from contaminated feed and soil into chicken eggs and meat are presented. Three independent controlled feeding study datasets were used to estimate the model parameters and four studies for evaluating the model performance. The yolk deposition of ndl-PCBs is modeled in a novel way that mimics the physiology of yolk growth and development, resulting in improved prediction of the experimental data without introducing an ad hoc time delay between ovulation and oviposition. Using the models, the highest level of 2.4 μg/kg dry matter (DM) was calculated for the sum of ndl-PCBs in laying hen feed to ensure that the current maximum levels in meat and eggs (40 ng/g fat) will not be exceeded. It is also shown how this highest level in feed should be adapted in case soil, in addition to feed, is also a source of ndl-PCBs for free-range chickens.

Lupin varieties with a low content of quinolizidine alkaloids (QAs) like blue sweet lupin (BSL) have long been used as a protein source for dairy cows. A health concern for humans may arise from the transfer of acute toxic QAs from feed into cow's milk. This study is the first to quantify the transfer of QAs from BSL into cow's milk with experimental and modeling methods. Four lactating dairy cows were subjected to two 7 day feeding periods with 1 and 2 kg/d BSL, respectively, each followed by a depuration period. BSL contained 1774 mg/kg dry matter total QAs. Individual milk samples were taken twice daily and QA contents in feed and milk determined with liquid chromatography-tandem mass spectrometry. Transfer of QAs into the milk was already seen with the administration of 1 kg/d BSL, with differences in transfer rates (TRs) between individual QAs. A toxicokinetic model was derived to quantify and predict QA feed-to-food transfer. For the four most prominent QAs, our model shows an alpha-half-life of around 0.27 d. TRs were obtained for six QAs and were between 0.13 (sparteine) and 3.74% (multiflorine). A toxicological assessment of milk containing QAs as measured in this study indicated a potential health concern.

Undesirable substances in feed can transfer into foods of animal origin after ingestion by livestock animals. These contaminants in food may threaten consumer health. Commonly, feeding trials are conducted with animals to assess the transfer of undesirable substances into animal tissues or milk. Such feeding trials explore the effects of the various physiological systems (e.g., ruminant and non-ruminant gastro-intestinal tracts) as well as different livestock production intensities on transfer. Using alternative methods to mimic the complex physiological processes of several organs is highly challenging. This review proposes a potential cascade of in vitro and ex vivo models to investigate the transfer of contam­inants from feed into foods of animal origin. One distinct challenge regarding the models for ruminants is the simulation of the forestomach system, with the rumen as the anaerobic fermentation chamber and its epithelial surfaces for absorption. Therefore, emphasis is placed on in vitro systems simulating the rumen with its microbial ecosystem as well as on ex vivo systems to replicate epithelial absorption. Further, the transfer from blood into milk must be evaluated by employing a suitable model. Finally, in silico approaches are introduced that can fill knowledge gaps or substitute in vitro and ex vivo models. Physiologically-based toxicokinetics combines the information gained from all alternative methods to simulate the transfer of ingested undesirable substances into foods of animal origin.

Background During food or feed contamination events, it is of utmost importance to ensure their rapid resolution to minimize impact on human health, animal health and finances. The existing Rapid Alert System for Food and Feed (RASFF) is used by the European Commission, national competent authorities of member countries and the European Food Safety Authority to report information on any direct or indirect human health risk arising from food or feed, or serious risks to animal health or the environment in relation to feed. Nevertheless, no methods exist to to collectively evaluate this vast source of supply chain information. Methods To aid in the extraction, evaluation and visualization of the data in RASFF notifications, we present the Rapid Alert Supply Network Extractor (RASNEX) open-source tool available from https://doi.org/10.5281/zenodo.4322555 freely. Among RASNEX’s functions is the graphical mapping of food and feed supply chain operators implicated in contamination events. RASNEX can be used during ongoing events as a support tool for risk analysis using RASFF notifications as input. Results In a first use case, we showcase the functionality of RASNEX with the RASFF notification on a 2017/2018 contamination event in eggs caused by the illegal use of fipronil. The information in this RASFF notification is used to visualize nine different flows of main and related food products. In a second use case, we combine RASFF notifications from different types of food safety hazards (Salmonella spp., fipronil and others) to obtain wider coverage of the visualized egg supply network compared to the first use case. Actors in the egg supply chain were identified mainly for Italy, Poland and Benelux. Other countries (although involved in the egg supply chain) were underrepresented. Conclusions We hypothesize that biases may be caused by inconsistent RASFF reporting behaviors by its members. These inconsistencies may be counteracted by implementing standardized decision-making tools to harmonize decisions whether to launch a RASFF notification, in turn resulting in a more uniform future coverage across European food and feed supply chains with RASNEX.

Non-dioxin-like polychlorinated biphenyls (ndl-PCBs) are a subclass of persistent bioaccumulative pollutants able to enter the food chain. We investigated the transfer of ndl-PCBs from contaminated feed into meat and liver of fattening chickens. A total of 48 chicks were divided into five treatment and one control groups. Treated animals were fed with contaminated diets (11.7 ± 0.4 μg/kg sum of indicator ndl-PCBs; 88% dry matter (DM)) before slaughter for different subperiods of time: 16, 23, 28, 32, and 36 days for groups 1-5, respectively. One day after the end of each subperiod, three animals per group were slaughtered to determine the congener-specific ndl-PCB content. All remaining animals were fed the control feed until slaughter on day 37 to probe depuration. We used these data to generate congener-specific physiologically based toxicokinetic (PBTK) models for indicator ndl-PCBs. The models show that PCBs 28, 138, 153, and 180 form a more slowly eliminated cluster (with an observed transfer rate into meat over 74% and observed half-lives over 8.7 days) than PCBs 52 and 101 (with a transfer rate under 13% and half-lives under 2.6 days). Our simulations show that ndl-PCB levels in feed lower than 3.9 (long 56-day) or 4.4 μg/kg (short 37-day fattening period) would be necessary to ensure the current maximum level in muscle meat (fat basis), according to EU Regulations 1881/2006 and 1259/2011. The PBTK models are made available in the Python and Food Safety Knowledge Exchange formats.