Regulatory landscape of alternatives to animal testing in food safety evaluations with a focus on the western world.

Traditional animal models are increasingly being replaced by new approach methodologies (NAMs) which focus on predicting toxicity of chemicals based on mechanistic data rather than apical endpoint data usually obtained from animal models. Beyond in vitro genetic tests, however, only a handful of NAMs have been successfully implemented in regulatory decision-making processes, mostly in the cosmetics and chemicals sector. Regulatory guidance on food safety testing in many jurisdictions still relies on data obtained from animal studies. This is due to the lack of validated models to predict systemic toxicity, which is essential to develop health-based guidance values for food additives. Other factors limiting the adoption of NAMs into food safety assessment include sector legislation lagging behind scientific progress, and lack of training and expertise to use the new models. While regulatory and industry bodies are working to combat these challenges, more needs to be done before these models can be used as standalone tools for regulatory decision-making. This review summarizes the current state and challenges of regulatory acceptance of NAMs for decision-making, and the efforts by governing bodies and industry to transition from animal testing for food safety assessments.

Genotoxicity and neonatal subchronic toxicity assessment of a novel mixture of the human-identical milk oligosaccharides lacto-N-fucopentaose I and 2'-fucosyllactose.

Human milk oligosaccharides (HMOs) are a complex group of bioactive molecules largely observed in human breast milk but also occurring in limited amounts in other mammalian milks. Advances in biotechnology have enabled production of human-identical milk oligosaccharides (HiMOs), structurally identical molecules to HMOs found naturally in human milk, intended for addition to infant formula to more closely replicate breast milk. Biosynthesis of a novel mixture of two major HMOs, lacto-N-fucopentaose I and 2'-fucosyllactose (LNFP-I/2'-FL), recently became possible. To support the safety of LNFP-I/2'-FL for use in infant formula and other foods, it was subject to a safety assessment comprising a bacterial reverse mutation test, an in vitro mammalian cell micronucleus test, and a 90-day oral gavage study in neonatal rats. In the 90-day study (the first HiMO study to include the new endocrine-sensitive endpoints described in the 2018 version of OECD Test Guideline 408), LNFP-I/2'-FL was administered by oral gavage to neonatal rats once daily (from Day 7 of age) for 90 consecutive days, at doses up to 5000 mg/kg bw/day, followed by a 4-week recovery period. Concurrent reference controls received 5000 mg/kg bw/day of the approved infant formula ingredient oligofructose. LNFP-I/2'-FL was nongenotoxic in vitro. The highest dose tested (5000 mg/kg bw/day) was established as the no-observed-adverse-effect level in the 90-day study, as there were no test article-related adverse effects on clinical observations, body weight, food consumption, clinical pathology, and organ weights nor any noteworthy macroscopic or microscopic findings. This supports the safety of LNFP-I/2'-FL for its intended uses in food.

Mutagenicity and genotoxicity studies of aspartame.

Two studies were conducted to further assess its mutagenic and genotoxic potential. In a bacterial reverse mutation pre-incubation study, Salmonella typhimurium strains TA100, TA1535, TA98, and TA1537 and Escherichia coli WP2 uvrA were treated with aspartame at concentrations of up to 5000 µg/plate with or without metabolic activation and showed no mutagenic potential. Similarly, in vivo micronucleus testing of aspartame following gavage administration (500-2000 mg/kg body weight) to Crlj:CD1(ICR) strain SPF male mice showed no increase in the proportion of micronucleated polychromatic erythrocytes in bone marrow cells collected and evaluated 24 or 48 h post administration. Overall, aspartame had no potential for mutagenic or genotoxic activity.

Systematic review and evaluation of aspartame carcinogenicity bioassays using quality criteria.

The current review assessed cancer studies of aspartame based on a quality appraisal using the Klimisch grading system. Nine studies having complete histopathology were included: three 2-year studies by Searle; three transgenic mice studies by the NTP; three lifetime studies by the Ramazzini Institute. A tenth study limited to brain tumors was not rated. None were determined as Klimisch Code 1 (reliable without restrictions). The Searle studies predated GLP standards but their methodology was comparable; transgenic mouse models are not validated, but are accepted as supporting data. These studies were rated Klimisch Code 2 (reliable with restrictions). The Ramazzini Institute used a lifetime model of their own design that has been questioned due to high rates of spontaneous tumors, issues with tumor type diagnosis and concerns about the impact of chronic infections. As many of these problems could be attributed to using animals that died or were terminated near end of life, along with the other problems noted, these studies were rated Klimisch Code 3 (not reliable). As the Klimisch Code 2 studies demonstrated a lack of carcinogenic potential, and as aspartame is hydrolyzed to common components and lacks genotoxic activity, a conclusion that aspartame is not carcinogenic is supported.

Evaluation of aspartame cancer epidemiology studies based on quality appraisal criteria.

Given the widespread use of the low-calorie sweetener aspartame over the last 30 years, the current work was undertaken to evaluate aspartame epidemiology studies looking at cancer endpoints against quality appraisal criteria. The quality appraisal tool used was from the National Heart, Lung and Blood Institute (NHLBI) of the National Institute of Health. Studies identified included nine case-control studies and five prospective cohort studies. Most studies assessed low-calorie or diet beverages rather than aspartame intake specifically; however, common use of aspartame in diet sodas does allow for some general extrapolation of results. Following consideration of study quality, two case-control and five prospective studies were considered to meet the majority of the NHLBI criteria. The primary limitation of the other case-control studies was an inadequate sample size. Overall, the results of the studies do not support that exposures to low and no-calorie sweeteners and beverages, and by extension aspartame, are associated with an increased risk of cancer in humans.

Histological analyses of the Ishii (1981) rat carcinogenicity study of aspartame and comparison with the Ramazzini Institute studies.

Researchers from the Ramazzini Institute have reported that lifespan dosing of rats with aspartame treatment is associated with an increased overall incidence of malignant tumors, including leukemias/lymphomas, transitional cell carcinomas of the renal pelvis/ureter, and malignant schwannomas of the peripheral nerves. Other carcinogenicity studies conducted on aspartame have shown no such carcinogenic potential in any organ system. Additional data to assess the carcinogenic potential of aspartame, especially in relation to the publications of the Ramazzini Institute, were obtained from a third-party histological evaluation of tissues from a carcinogenicity study previously conducted to assess the potential for aspartame to induce tumors of the brain. The results of this histological evaluation provide no evidence of a tumorigenic effect of aspartame in any organ group, including those organs/tissues reportedly affected in the Ramazzini Institute's studies. The only effects identified were an increased incidence of renal pelvic mineralization and renal pelvic hyperplasia secondary to the irritant properties of the mineralization process. The toxicological significance of these particular findings is widely considered minimal. There is no evidence that aspartame is carcinogenic in rats, at least to doses of 4 g/kg body weight/day administered over a 2-year period.

Toxicological safety evaluation of the human-identical milk oligosaccharide 6'-sialyllactose sodium salt.

Human milk oligosaccharides (HMOs) are abundant in breastmilk, but their presence in infant formula is negligible. Sialylated HMOs, such as 6'-sialyllactose, constitute a significant portion of the HMO fraction of human milk and are linked to important biological functions. To produce infant formula that is more comparable with human milk, biosynthesized sialyllactoses known as human-identical milk oligosaccharides (structurally identical counterparts to their respective naturally occurring HMOs in breastmilk) are proposed for use in infant formula and other functional foods for the general population. To support the safety of 6'-sialyllactose sodium salt (6'-SL), a 90-day oral (gavage) toxicity study and in vitro genotoxicity tests were conducted. The 90-day study is the first to be conducted with 6'-SL using neonatal rats (day 7 of age at the start of dosing), thus addressing safety of 6'-SL for consumption by the most sensitive age group (infants). In the 90-day study, neonatal rats received 6'-SL at doses up to 5000 mg/kg body weight (BW)/day and reference controls received 5000 mg/kg BW/day of fructooligosaccharide (an ingredient approved for use in infant formula) for comparison with the high-dose 6'-SL group, followed by a 4-week recovery period. There was no evidence of genotoxicity in vitro. No test item-related adverse effects were observed on any parameter in the 90-day study, thus the high dose (5000 mg/kg BW/day) was established as the no-observed-adverse-effect level. These results confirm that 6'-SL is safe for use in formula milk for infants and in other functional foods for the general population.

Toxicological safety assessment of the human-identical milk oligosaccharide 3'-sialyllactose sodium salt.

Human breastmilk is a mixture of nutrients, hormones and bioactive molecules that are vital for infant growth and development. Infant formula (IF) lacks many of these compounds, most notably human milk oligosaccharides (HMOs), which are abundant in breastmilk but scarce in IF. Sialyllactoses, such as 3'-sialyllactose, constitute a large portion of the HMO fraction. To produce IF that matches breastmilk more closely, biosynthesized human-identical milk oligosaccharides (structurally identical to HMOs) such as 3'-sialyllactose sodium salt (3'-SL) are proposed for use in IF and foods for the general population. The safety assessment of 3'-SL comprised in vitro genotoxicity tests and a 90-day oral (gavage) toxicity study. This is the first 90-day study conducted with 3'-SL using neonatal rats (7 days old at the start of dosing-equivalent age to newborn human infants in terms of central nervous system and reproductive development), demonstrating the safety of 3'-SL for consumption by infants, the most sensitive age group. The neonatal rats received 3'-SL at doses up to 5,000 mg/kg body weight (BW)/day and reference controls received 5,000 mg/kg BW/day of fructooligosaccharide (an ingredient approved for use in IF) for comparison with the high-dose 3'-SL group, followed by a 4-week recovery period. There was no evidence of genotoxicity in vitro. In the absence of any test item-related adverse effects in the 90-day study, the high dose (5,000 mg/kg BW/day) was established as the no-observed-adverse-effect level. This confirms the safety of 3'-SL for use in IF for infants, as well as in functional foods for the general population.

Dietary Exposures to Common Emulsifiers and Their Impact on the Gut Microbiota: Is There a Cause for Concern?

Emulsifiers are commonly used in food processing for the technological purpose of altering the flavor or to improve the texture of foods. Due to their ubiquity, these substances are consumed daily at low levels in the human diet. Recently published in vitro and in vivo studies suggest dietary exposure to emulsifiers modulate the gut microbiota and contribute to the increasing prevalence of metabolic disease. A literature search was conducted which identified eight studies investigating the interaction of sodium carboxymethyl cellulose, polysorbate 80, gum arabic, carrageenan, and arabinogalactan with the gut microbiota in murine and in vitro models. Numerous inconsistent changes in various phyla and genera were identified. These studies were conducted at high doses that have no relevance to the current dietary levels consumed in the United States. Subtle changes in gut microbiota composition as a toxicological endpoint is not supported by established internationally recognized toxicology testing guidelines. Therefore, the results of these studies are difficult to interpret and extrapolate to humans and are not supported by previous safety conclusions of international food safety authorities. The current understanding of the gut microbiota is that the structure is highly dynamic and is heavily influenced by the diet. Thus, the results of these studies may not necessarily suggest a safety concern, but rather reflect an adaptive response of the gut microbiota to an external stressor. Future research will need to further elucidate the mechanisms of metabolic disease in rodents and humans and establish clinically relevant and reliable endpoints to assess changes in gut microflora.

Risk Assessment Paradigm for Glutamate.

BACKGROUND: Re-evaluation of the use of glutamic acid and glutamate salts (referred to as glutamate hereafter) by the European Food Safety Authority (EFSA) proposed a group acceptable daily intake (ADI) of 30 mg/kg body weight (bw)/day. SUMMARY: This ADI is below the normal dietary intake, while even intake of free glutamate by breast-fed babies can be above this ADI. In addition, the pre-natal developmental toxicity study selected by EFSA, has never been used by regulatory authorities worldwide for the safety assessment of glutamate despite it being available for nearly 40 years. Also, the EFSA ignored that toxicokinetic data provide support for eliminating the use of an uncertainty factor for interspecies differences in kinetics. Key Messages: A 3-generation reproductive toxicity study in mice that includes extensive brain histopathology, provides a better point of departure showing no effects up to the highest dose tested of 6,000 mg/kg bw/day. Furthermore, kinetic data support use of a compound-specific uncertainty factor of 25 instead of 100. Thus, an ADI of at least 240 mg/kg bw/day would be indicated. In fact, there is no compelling evidence to indicate that the previous ADI of "not specified" warrants any change.

Preclinical safety evaluation of the human-identical milk oligosaccharide lacto-N-tetraose.

Lacto-N-tetraose (LNT) is one of the most abundant oligosaccharides that are endogenously present in human breast milk. To simulate the composition of human breast milk more closely, commercial infant formula can be supplemented with human-identical milk oligosaccharides, which are manufactured structurally identical versions of their naturally occurring counterparts. As part of the safety evaluation of LNT, in vitro genotoxicity tests and a subchronic oral gavage toxicity study (in neonatal Sprague-Dawley rats) were conducted. In the subchronic study, LNT was administered at dose levels of 0, 1,000, 2500 or 4000 mg/kg body weight (bw)/day, once daily for at least 90 days, followed by a 4-week treatment-free period. An identically comprised reference control group received fructooligosaccharides powder (a non-digestible oligosaccharide used in infant formula) at 4000 mg/kg bw/day, to allow for direct comparison against the high-dose LNT group. LNT was non-genotoxic in the in vitro tests. There were no compound-related adverse effects in the 90-day study; therefore, 4000 mg/kg bw/day (the highest feasible dose) was established as the no-observed-adverse-effect-level. These results support the safe use of LNT in infant formula and as a food ingredient, at levels not exceeding those found naturally in human breast milk.

Safety evaluation of water-soluble palm fruit bioactives.

Water-soluble palm fruit bioactives, derived from the aqueous stream of palm oil processing, have shown anti-diabetogenic effects in rodent models. To assess the safety of potential incorporation of this polyphenol-containing material in food, in vitro bacterial reverse mutation and in vitro chromosome aberration assays were conducted along with a 90-day subchronic toxicity study in Sprague-Dawley rats. Water-soluble palm fruit bioactives were inactive in the Ames and in vitro chromosome aberration assays up to the limit doses of 5000 µg/plate and 5000 µg/mL, respectively. In the 90-day feeding study, water-soluble palm fruit bioactives were administered via gavage at doses 0, 500, 1000 or 2000 mg/kg body weight/day. No significant effects were noted on body weight, food consumption, hematology, clinical chemistry, organ weights, and histopathological examination. The No Observable Adverse Effect Level was considered to be 2000 mg/kg body weight/day, the highest dose tested. These data provide evidence to support the safe use of water-soluble palm fruit bioactives in food or food ingredients.

Lipid-soluble green tea extract: Genotoxicity and subchronic toxicity studies.

To assess the potential safety of lipid soluble green tea extract, also referred to as lipid soluble tea polyphenols (LSTP), a series of genotoxicity tests were conducted, including an Ames, in vivo mouse micronucleus, and in vivo mouse sperm abnormality test. The toxicity of LSTP was evaluated in 90- and 30-day feeding studies. LSTP did not show mutagenic activity in the Ames test and no genotoxic potential in the in vivo assays at doses up to 10 g/kg body weight (bw). In the 90-day feeding study, LSTP was given in the diet at levels providing 0, 0.125, 0.25, or 0.50 g/kg bw/day. No significant effects were noted on body weight, food consumption, hematology, clinical chemistry, organ weights, and histopathological examination. The no-observed-adverse-effect level (NOAEL) was therefore considered to be 0.50 g/kg bw/day, the highest dose tested. Likewise, dosing of SD rats by gavage for 30 days also showed no adverse effects of growth, hematology, clinical chemistry, organ weights, or histopathology at doses of 0.58, 1.17, and 2.33 g/kg bw/day. The NOAEL in the 30-day study was considered to be the highest dose tested. These data provide evidence to support the safe use of LSTP in food.

Clinical safety evaluation of marine oil derived from Calanus finmarchicus.

Marine oils are rich in polyunsaturated fatty acids (PUFAs), including docosahexaenoic and eicosapentaenoic acid. These PUFAs are associated with health benefits and additional sustainable sources of marine oils are desirable. One of the source organisms is Calanus finmarchicus, a copepod endemic to the North Atlantic. PUFAs in the lipid fraction of this organism are largely in the form of wax esters. To assess the safety of these wax esters as a source of PUFAs, a randomized, double-blinded, placebo-controlled clinical trial was conducted whereby 64 subjects consumed 2 g Calanus oil in capsule form daily for a period of one year. A group of 53 subjects consumed placebo capsules. At baseline, 6-, and 12-months, series of evaluations were conducted, including: vital signs, clinical chemistry and hematological evaluations, and adverse event reporting. Food intake and physical exercise were controlled by means of a questionnaire. There were no effects on Calanus oil treatment on any of the safety parameters measured. A slight increase in the incidence of eczema was reported in the Calanus oil group, but the response was minor in nature, not statistically significant after controlling for multiple comparisons, and could not be attributed to treatment.

Steviol glycosides in purified stevia leaf extract sharing the same metabolic fate.

The safety of steviol glycosides is based on data available on several individual steviol glycosides and on the terminal absorbed metabolite, steviol. Many more steviol glycosides have been identified, but are not yet included in regulatory assessments. Demonstration that these glycosides share the same metabolic fate would indicate applicability of the same regulatory paradigm. In vitro incubation assays with pooled human fecal homogenates, using rebaudiosides A, B, C, D, E, F and M, as well as steviolbioside and dulcoside A, at two concentrations over 24-48 h, were conducted to assess the metabolic fate of various steviol glycoside classes and to demonstrate that likely all steviol glycosides are metabolized to steviol. The data show that glycosidic side chains containing glucose, rhamnose, xylose, fructose and deoxy-glucose, including combinations of α(1-2), ß-1, ß(1-2), ß(1-3), and ß(1-6) linkages, were degraded to steviol mostly within 24 h. Given a common metabolite structure and a shared metabolic fate, safety data available for individual steviol glycosides can be used to support safety of purified steviol glycosides in general. Therefore, steviol glycosides specifications adopted by the regulatory authorities should include all steviol glycosides belonging to the five groups of steviol glycosides and a group acceptable daily intake established.

Chemical-specific adjustment factors (inter-species toxicokinetics) to establish the ADI for steviol glycosides.

The acceptable daily intake (ADI) of commercially available steviol glycosides is currently 0-4 mg/kg body weight (bw)/day, based on application of a 100-fold uncertainty factor to a no-observed-adverse-effect-level value from a chronic rat study. Within the 100-fold uncertainty factor is a 10-fold uncertainty factor to account for inter-species differences in toxicokinetics (4-fold) and toxicodynamics (2.5-fold). Single dose pharmacokinetics of stevioside were studied in rats (40 and 1000 mg/kg bw) and in male human subjects (40 mg/kg bw) to generate a chemical-specific, inter-species toxicokinetic adjustment factor. Tmax values for steviol were at ∼8 and ∼20 h after administration in rats and humans, respectively. Peak concentrations of steviol were similar in rats and humans, while steviol glucuronide concentrations were significantly higher in humans. Glucuronidation in rats was not saturated over the dose range 40-1000 mg/kg bw. The AUC0-last for steviol was approximately 2.8-fold greater in humans compared to rats. Chemical-specific adjustment factors for extrapolating toxicokinetics from rat to human of 1 and 2.8 were established based on Cmax and AUC0-last data respectively. Because these factors are lower than the default value of 4.0, a higher ADI for steviol glycosides of between 6 and 16 mg/kg bw/d is justified.

Safety evaluation of the human-identical milk monosaccharide, l-fucose.

l-Fucose is a natural monosaccharide present in mammals where it is found predominantly as an O-glycosidically linked component of glycoproteins, glycolipids, and oligosaccharides. It is also present in its free form in human breast milk (human milk monosaccharide). l-Fucose plays important roles in the development of the immune and nervous systems and is involved in cognitive function and memory formation. The human-identical milk monosaccharide l-fucose is therefore proposed for use in infant formulas to better simulate the free saccharides present in human breast milk. As part of the safety evaluation of l-fucose, a subchronic dietary toxicity study preceded by an in utero phase was conducted in Sprague-Dawley rats. l-Fucose was without maternal toxicity or compound-related adverse effects on female reproduction and general growth and development of offspring at a maternal dietary level up to 1%, equivalent to a dose of 1655 mg/kg body weight (bw)/day. During the subchronic phase, no compound-related adverse effects were observed in first generation rats at dietary levels of up to 1% (highest level tested), corresponding to doses of 516 and 665 mg/kg bw/day in males and females, respectively. l-Fucose was non-genotoxic in a series of in vitro genotoxicity/mutagenicity tests. These results support the safe use of l-fucose in infant formula and as a food ingredient at levels equivalent to those present in human breast milk.

In vitro metabolism of rebaudioside B, D, and M under anaerobic conditions: comparison with rebaudioside A.

The hydrolysis of the steviol glycosides rebaudioside A, B, D, and M, as well as of steviolbioside (a metabolic intermediate) to steviol was evaluated in vitro using human fecal homogenates from healthy donors under anaerobic conditions. Incubation of each of the rebaudiosides resulted in rapid hydrolysis to steviol. Metabolism was complete within 24h, with the majority occurring within the first 8h. There were no clear differences in the rate or extent of metabolism of rebaudioside B, D, or M, relative to the comparative control rebaudioside A. The hydrolysis of samples containing 2.0mg/mL of each rebaudioside tended to take slightly longer than solutions containing 0.2mg/mL. There was no apparent gender differences in the amount of metabolism of any of the rebaudiosides, regardless of the concentrations tested. An intermediate in the hydrolysis of rebaudioside M to steviol, steviolbioside, was also found to be rapidly degraded to steviol. The results demonstrate that rebaudiosides B, D, and M are metabolized to steviol in the same manner as rebaudioside A. These data support the use of toxicology data available on steviol, and on steviol glycosides metabolized to steviol (i.e., rebaudioside A) to substantiate the safety of rebaudiosides B, D, and M.

Genotoxicity of pyrroloquinoline quinone (PQQ) disodium salt (BioPQQ™).

The genotoxic potential of pyrroloquinoline quinone (PQQ) disodium salt (BioPQQ™) was evaluated in a battery of genotoxicity tests. The results of the bacterial mutation assay (Ames test) were negative. Weak positive results were obtained in 2 separate in vitro chromosomal aberration test in Chinese hamster lung (CHL) fibroblasts. Upon testing in an in vitro chromosomal aberration test in human peripheral blood lymphocytes, no genotoxic activity of PQQ was noted. In the in vivo micronucleus assay in mice, PQQ at doses up to 2,000 mg/kg body weight demonstrated that no genotoxic effects are expressed in vivo in bone marrow erythrocytes. The weak responses in the in vitro test CHL cells were considered of little relevance under conditions of likely human exposure. PQQ disodium was concluded to have no genotoxic activity in vivo.

Alternatives to animal testing in toxicity testing: Current status and future perspectives in food safety assessments.

The development of alternative methods to animal testing has gained great momentum since Russel and Burch introduced the "3Rs" concept of Reduction, Refinement, and Replacement of animals in safety testing in 1959. Several alternatives to animal testing have since been introduced, including but not limited to in vitro and in chemico test systems, in silico models, and computational models (e.g., [quantitative] structural activity relationship models, high-throughput screens, organ-on-chip models, and genomics or bioinformatics) to predict chemical toxicity. Furthermore, several agencies have developed robust integrated testing strategies to determine chemical toxicity. The cosmetics sector is pioneering the adoption of alternative methodologies for safety evaluations, and other sectors are aiming to completely abandon animal testing by 2035. However, beyond the use of in vitro genetic testing, agencies regulating the food industry have been slow to implement alternative methodologies into safety evaluations compared with other sectors; setting health-based guidance values for food ingredients requires data from systemic toxicity, and to date, no standalone validated alternative models to assess systemic toxicity exist. The abovementioned models show promise for assessing systemic toxicity with further research. In this paper, we review the current alternatives and their applicability and limitations in food safety evaluations.