The minimum single dose (LOEL) of deoxynivalenol that induced vomiting (emesis) in groups of three to six pigs weighing 9–10 kg was 0.1 mg/kg bw when given by oral gavage (NOEL, 0.075 mg/kg bw) and 0.05 mg/kg bw when given intrapertioneally (NOEL, 0.025 mg/kg bw) (Forsyth et al., 1977).

The minimum single emetic doses of deoxynivalenol and 15-acetyldeoxynivalenol in groups of three Yorkshire pigs weighing 10–15 kg were 0.050 and 0.075 mg/kg bw, respectively, when given either by gavage or intraperitoneally. After gavage, three of 15 pigs given the 15-acetyl metabolite and four of 15 given deoxynivalenol showed emesis at all doses from 20 to 200 µg/kg bw. After intraperitoneal administration, nine of 15 pigs showed emesis at all doses. The NOELs were 0.025 mg/kg bw for deoxynivalenol and 0.050 mg/kg bw for the 15-acetyl metabolite after either oral intubation or intraperitoneal injection (Pestka et al., 1987a).

In a pilot study, the median emetic dose (ED₅₀) of purified deoxynivalenol, administered by gavage to groups of two to six Yorkshire pigs weighing 28–51 kg was 0.088 mg/kg bw (Young et al., 1983). Additional data provided by D. Prelusky, Agriculture and Agrifoods, Canada, indicated that the LOEL in this study was 0.07 mg/kg bw (for the two pigs that responded, and the NOEL was 0.035 mg/kg bw.

The ED₅₀ values after single doses of deoxynivalenol (> 96%pure) administered by cannula into the stomach or intravenously to groups of four to six Yorkshire pigs, weighing 12–25 kg, which had fasted for 4 h, were 75 and 20 mg/kg bw by the two routes, respectively (Prelusky & Trenholm, 1993). Additional data provided by D. Prelusky, Agriculture and Agrifoods, Canada, indicated that after oral administration the LOEL was 0.05 mg/kg bw for the one pig that responded, and the NOEL was 0.025 mg/kg bw. After intravenous administration, the NOEL was 0.015 mg/kg bw.

In studies in which young pigs (7.5 kg) received feed containing heavily conta-minated mouldy corn, a dietary concentration of approximately 20 mg/kg (equal to 0.8 mg/kg bw) caused emesis, whereas no emesis was observed at 12 mg/kg feed (equal to 0.6 mg/kg bw) (Young et al., 1983).

In a study of feed inoculated with corn, no emesis was observed in 34-kg pigs at a dietary concentration of 14 mg/kg (equal to 0.42 mg/kg bw) (Friend et al., 1984).

Similarly, in studies in dogs and cats given a diet containing contaminated wheat, emesis occurred at doses of 0.45 and 0.4 mg/kg bw, respectively. The NOEL was 0.3 mg/kg bw for both species (for further experimental details, see section 2.2.2) (Hughes et al., 1999).

The minimum single emetic doses (LOELs) of deoxynivalenol and 15-acetyl-deoxynivalenol given subcutaneously to 6-month-old dogs weighing 2–3 kg were 0.1 and 0.2 mg/kg bw, respectively (Yoshizawa & Morooka, 1974).

The Committee noted that, in the studies described above, emesis occurred in pigs at much lower doses when deoxynivalenol was given by gavage than when it was given in the feed. This difference was attributed to a bolus effect of gavage. The usual exposure of humans would be comparable to administration in the feed.

At a dose of 46 mg/kg bw given by gavage, deoxynivalenol damaged the cells lining the gastrointestinal tract of 4-week-old ICR mice. Ulcers and cell infiltration were observed in the forestomach, and necrosis of immature crypt cells, cell infiltration in the mucosa, and cystic changes in the crypts were observed in the small intestine (Ito et al., 1993).

2.2.2 Short-term studies of toxicity

The results of these studies are summarized in Table 3.

Mice: In a study lasting about 18 weeks, groups of 80 Swiss-Webster weanling male mice were fed either an ‘uncontaminated’ wheat diet containing deoxynivalenol at 0.05 mg/kg or a contaminated wheat diet containing 6.3 mg/kg, equivalent to approximately 0.9 mg/kg bw per day. The mice were killed serially between 7 and 137 days. A 10% decrease in body weight was seen which was related in part to decreased food intake. There were no significant pathological findings and only slight changes in haematological parameters, probably related to body weight. The NOEL was 0.9 mg/kg bw per day (Arnold et al., 1986a).

Groups of four BALB/c mice were fed diets containing deoxynivalenol at a concentration of 2.5, 5, 10, 20, or 50 mg /kg, equal to 0.35, 0.67, 1.3, 2.7, and 6.5 mg/kg bw per day, for 7 days (repeated in three replicate trials). Food intake was decreased at all doses. At 1.3 mg/kg bw, decreased body-weight gain, decreased thymus weight, and decreased cardiac protein synthesis were seen. When feeding was continued until day 30, cardiac lesions (calcified pericarditis foci) were also observed at this dose (Robbana-Barnat et al., 1987).

Groups of 10–12 young female ICR mice were fed diets containing deoxynivalenol at a concentration of 0, 4, 8, 12, or 16 mg/kg, equivalent to 0.6, 1.2, 1.8, and 2.4 mg/kg bw. Reduced feed intake was observed at the three higher doses and reduced growth at all doses (Rotter et al., 1992).

Groups of eight male and eight female outbred ICR mice aged 3 weeks and weighing 16–18 g, housed singly, were fed diets containing deoxynivalenol at a concentration of 0, 2, 4, or 8 mg/kg for 14 days. The authors calculated that the actual intakes were 0, 0.37, 0.76, and 1.5 mg/kgbw per day for the males and 0, 0.41, 0.81, and 1.6 mg/kg bw per day for the females. Food consumption was measured on days 7 and 14 of the experiment, and individual body weights were recorded on days 0, 7, and 14. Feed efficiency was calculated for both weeks of the experiment. Blood samples were collected at the end of the experiment and analysed for erythrocyte count, haemoglobin concentration, erythrocyte volume fraction, mean corpuscular volume, mean corpuscular haemoglobin, and mean corpuscular haemoglobin concentration. The feed consumption of the deoxynivalenol-fed animals was significantly reduced (p < 0.05), especially for the males, during each of the two 7-day test periods. For animals of each sex and during both weeks, the amounts consumed by animals at 2 and 4 mg/kg were similar. The weight gains of all treated males were significantly reduced after 7 days of exposure (p < 0.001), but only the males fed 8 mg/kg of diet gained less weight than the controls during week 2. Females at this concentration weighed less than the controls after the first week of exposure but gained more weight than the controls during the second week. The erythrocyte counts of treated animals were significantly lower than in the controls (p < 0.05), but the effect was seen primarily at 2 mg/kg of diet. The LOEL was 0.37 mg/kg bw per day and the NOEL was 0.81 mg/kg bw per day (Rotter et al., 1994a).

Groups of 24 Swiss-Webster-derived male weanling mice were given deoxy-nivalenol at a dose of 0, 0.75, 2.5, or 7.5 mg/kg bw per day by gavage for 35 days. Two control groups were given the solvent or were untreated. Most of the animals given the two higher doses died during the study. At 2.5 mg/kg bw per day, lesions were found in the spleen, thymus, lymph nodes, and gastrointestinal tract. Bone-marrow and haematological parameters were also affected at this dose. Decreased food consumption, decreased body weight, decreased relative weights of the thymus and heart, and increased relative weight of the stomach were seen at all doses (Arnold et al., 1986a).

Groups of eight weanling female B6C3F₁ mice were fed diets containing deoxy-nivalenol at a concentration of 0.5, 2, 5, 10, or 25 mg/kg, equivalent to 0.07, 0.28, 0.7, 1.4, and 3.5 mg/kg bw per day for 56 days. Decreased body-weight gain was seen at doses > 0.28 mg/kg bw per day . The liver and kidney were affected at doses > 0.7 mg/kg bw per day (Forsell et al., 1986).

Groups of 10 weanling female B6C3F₁ mice were fed diets containing 15-acetyldeoxynivalenol at a concentration of 0.5, 2, 5, 10, or 25 mg/kg, equivalent to 0.07, 0.28, 0.7, 1.4, and 3.5 mg/kg bw per day for 56 days. Decreased body-weight gain was seen at doses > 0.7 mg/kg bw per day, and the weights of the spleen and kidney were decreased (Pestka et al., 1986).

Rats: Groups of 50 male Sprague-Dawley rats were fed either an ‘uncontami-nated’ wheat diet containing deoxynivalenol at 0.05 mg/kg or a contaminated wheat diet containing 6.23 mg/kg, estimated to be equivalent to 0.5 mg/kg bw. Rats were killed serially up to 91 days. A 10% decrease in body weight was observed which was related in part to decreased food intake. There were no significant pathological findings; the slight changes in haematological parameters were probably related to the changes in body weight (Arnold et al., 1986a).

Groups of 25 male and female weanling Sprague-Dawley rats were fed diets containing purified deoxynivalenol to provide a dose of 0.25, 0.5, or 1 mg/kg bw per day for 60 days. Body weight decreases, attributed in part to reduced food intake, were seen at all doses in females and at the highest dose in males (both 5% less than controls, but statistically significant). Thymidine uptake was decreased in the jejunum and spleen of males at the two highest doses, although the effect was significant only at the highest dose. The average decreases in spleen were 2.6% for male controls, 2.8% at 0.25 mg/kg bw per day, 1.8% at 0.5 mg/kg bw per day, and 0.9% at 1.0 mg/kg bw per day; and 19%, 17%, 12%, and 9.2%, respectively, in the jejunum. No notable differences in thymidine labelling were observed in the oesophagus of males or in the spleen, jejunum, or oesophagus of females. No changes were observed in organ weights, haematological or bone-marrow variables, sequential multichannel autoanalyser variables, or histological appearance (Arnold et al., 1986b).

In a 90 day-study, groups of 10 male Sprague Dawley rats weighing 200 g were fed diets containing deoxynivalenol at 20 mg/kg, equivalent to about 1 mg/kg bw per day. At this dose, there was no feed refusal, but a 10% decrease in body-weight gain was observed. Serum enzyme activity, haematological end-points, histopathological appearance, and liver detoxication systems were unaffected (Morrissey et al., 1985).

Poultry: Chickens tolerate deoxynivalenol at a concentration of at least 5 mg/kg of diet, equivalent to about 0.45 mg/kg bw per day. At concentrations up to 5 mg/kg feed, in fact, some beneficial effects on food consumption and weight gain were observed in Leghorn chickens (up to 28 days) and broilers (up to 45 days). When laying Leghorn hens were fed diets containing deoxynivalenol at 0.7 mg/kg for 70 days, no effect was found on feed intake, body weight, egg production, egg yield, or the number of cracked eggs. With increasing doses of deoxynivalenol up to 5.2 mg/kg of feed for 168 days, egg and shell weight and shell thickness decreased. Turkey poults given feed containing deoxynivalenol at 0.5 mg/kg for 14 days showed slightly reduced feed intake and weight gain, which were not statistically significant (Trenholm et al., 1984). The Committee noted that insufficient experimental detail was provided.

Many studies have shown that the performance of chicken broilers and turkey poults is little affected by concentrations of deoxynivalenol up to about 16 mg/kg; some of the more recent studies are summarized in Table 3.

Mink: Mink given a choice between uncontaminated feed and deoxynivalenol-contaminated feed displayed a preference for the uncontaminated feed at a concentration of deoxynivalenol as low as 0.28 mg/kg. However, when no choice was available, the mink readily consumed feed containing deoxynivalenol at concentrations up to 1.2 mg/kg (equal to 0.20 mg/kg bw per day) with no apparent ill effects over a 28-day period. The study suggests that the sensitivity of mink is close to that of pigs and greater than that of rats and chickens (Gibson et al., 1993).

Cats: Wheat naturally contaminated with deoxynivalenol to a concentration of 37 mg/kg was used to manufacture feed containing the toxin at 0, 1, 2, 4, 6, 8, or 10 mg/kg, equivalent to 0, 0.05, 0.1, 0.2, 0.3, 0.4, and 0.5 mg/kg bw per day and fed to groups of two to seven mature American shorthair cats, 1–9 years old, for 14 days. The wheat also contained 15-acetyldeoxynivalenol at 1 mg/kg, but none of the other Fusarium toxins (detection limits, 0.1–0.5 mg/kg). The concentration of deoxy-nivalenol in the feed was unchanged after manufacture, indicating that the toxin was stable during conventional extrusion processing. The feed intake of the cats was significantly reduced when it contained deoxynivalenol at a concentration > 7.7 ± 1.1 mg/kg (equivalent to 0.38 mg/kg bw per day). Vomiting was common at the highest concentration (Hughes et al., 1999).

Dogs: The same diets were fed to groups of 2–14 mature male and female beagle or Brittany dogs, 1–7 years old for 14 days, equivalent to doses of deoxynivalenol of 0, 0.075, 0.15, 0.3, 0.45, 0.6, and 0.75 mg/kg bw per day. Dogs previously fed deoxynivalenol-contaminated food preferentially selected uncontaminated food, whereas dogs not previously exposed to deoxynivalenol-contaminated food consumed equal quantities of contaminated and uncontaminated food. The digestibility of the food was not affected when it contained 6 mg/kg. Feed intake was significantly reduced at concentrations of deoxynivalenol > 4.5 ± 1.7 mg/kg, equivalent to 0.34 mg/kg bw per day. Vomiting was common at the two higher doses (Hughes et al., 1999).

Maize infected with Gibberella zeae, in which about 25% of the kernels were seen to be damaged and containing deoxynivalenol at a concentration of 12 mg/kg, was added to the feed of pigs weighing 20–45 kg. Feed consumption was reduced, by 20% at a concentration of 3.6 mg/kg to 90% at 40 mg/kg. Weight loss was associated with the feed refusal. Feed was refused more frequently when it contained naturally contaminated maize than when equal concentrations of the pure compound. This indicates that additional factors are involved in the feed refusal response of pigs (Forsyth et al., 1977).

Pigs of two age groups, weighing 8 and 60 kg, were fed rations containing wheat naturally contaminated with deoxynivalenol at a concentration of 1–4.2 mg/kg (equal to 0.036, 0.09, and 0.18 mg/kg bw per day) for 21 days and 42 days, respectively. The younger animals had decreased feed intake at the highest dose, but their weight gain was not affected. In the older animals, feed intake and body-weight gain were affected at 0.09 mg/kg bw. Histological examination revealed no significant lesions or abnormalities in the tissues examined (Pollman et al., 1985).

Groups of six 49-day-old, 14-kg, castrated male pigs were fed a ration to which naturally contaminated wheat had been added, resulting in a deoxynivalenol concentration of 5 (actual, 4.5) mg/kg of feed, equal to 0.2 mg/kg bw per day (using a factor of 0.05 rather than 0.04), for 28 days. Feed intake and growth were not decreased, but mild renal nephrosis was observed in two pigs. An additive or greater interaction was observed when the diets also contained fumonisin B₁ at 100 (actual, 47–56) mg/kg of feed (Harvey et al., 1996).

In young pigs weighing 30–80 kg and pregnant gilts, ingestion of a diet containing deoxynivalenol at concentrations > 2 mg/kg of feed, equivalent to > 0.08 mg/kg bw per day, resulted in decreased feed consumption and reduced weight gain. Pigs could ingest feed containing deoxynivalenol at up to 2 mg/kg of feed without serious adverse effects (Trenholm et al., 1984). The Committee noted that insufficient experimental details were provided for an assessment of the study.

Four trials were conducted in young pigs to evaluate the effect of deoxynivalenol-contaminated maize on performance. Mouldy maize containing deoxynivalenol at 875 mg/kg and zearalenone at 3.9 mg/kg was mixed with clean maize and other ingredients to provide feeds containing deoxynivalenol at concentrations ranging from 0.14 mg/kg (control) to 230 mg/kg. A dietary concentration of approximately 20 mg/kg caused emesis, 12 mg/kg caused almost complete feed refusal, and 1.3 mg/kg of feed (equivalent to 0.06 mg/kg bw per day) caused a significant reduction in feed intake and rate of weight gain. No lesions attributable to deoxynivalenol were observed in pigs fed up to 43 mg/kg feed for 21 days. Alterations in various serum characteristics were observed in pigs fed deoxynivalenol, but the effects could not be separated from those that result from low intake of food (Young et al., 1983).

Groups of six 84-day-old, 38-kg pigs were fed a ration to which naturally contaminated maize had been added, resulting in a deoxynivalenol concentration of 2.5 mg/kg, equivalent to 0.1 mg/kg bw per day, for 35 days. Decreased feed intake and growth were observed (Friend et al., 1992).

Groups of six to eight castrated male Yorkshire pigs, 6–7 weeks old and weighing 13 kg, were fed diets containing deoxynivalenol at a concentration of 0 (control), 0.95, 1.8, or 2.8 mg/kg for 28 days, equal to 0, 0.08, 0.13, and 0.18 mg/kg bw per day. The deoxynivalenol in the diet was from naturally contaminated maize, which also contained 15-acetyldeoxynivalenol (at about 25% the concentration of deoxynivalenol) and zearalenone (at about 4%). Feed consumption and body weight were recorded on days 2, 4, 7, 14, 21, and 28 of the study. Blood samples, collected on days 0, 7, and 28, were analysed for thyroxine, tri-iodothronine uptake, cortisol, and haematological parameters. Serum electrophoresis was conducted on samples collected on day 28. At sacrifice, the weights of the thyroid, thymus, spleen, and kidneys were recorded, and the stomachs were scored for colour, thickness, and inflammation. During the first 2 weeks of the experiment and also overall, food intake was decreased as the dietary deoxynivalenol concentration increased. Intake during the last 7 days was similar in all groups except that receiving 2.8 mg/kg of diet. A dose-related reduction was seen in weight gain in treated pigs during the first 7 days, but the overall daily gain over 28 days was similar in all groups. The absolute and relative weights of the thyroid were significantly lower (p <0.02 and p < 0.05, respectively) in pigs given the diets containing 1.8 and 2.8 mg/kg. The thyroxine concentration increased in response to increasing dietary concentrations of deoxy-nivalenol after 7 and 28 days (p < 0.017), and the albumin concentration increased in pigs fed increasing concentrations of deoxynivalenol at 28 days (p = 0.013). The alpha-globulin concentration showed a dose-related linear decrease (p = 0.016), and the albumin:globulin ratio in treated pigs was higher than in controls (p = 0.009). The NOEL was 0.08 mg/kg bw per day (Rotter et al., 1994b).

The effects of feeding a diet contaminated with deoxynivalenol at 4 mg/kg on performance and blood parameters were studied for 42 days in groups of eight male castrated Yorkshire pigs weighing 18 kg. On the basis of feed intake, the intake of deoxynivalenol was 0.26 and 0.16 mg/kg bw per day at the beginning and end of the experiment, respectively. Blood samples were collected weekly from all animals. Controls fed ad libitum and in pairs with treated animals were used to distinguish between differences in feed intake and effects of the deoxynivalenol-containing diet. Pigs fed the contaminated diet had on average a 20% lower feed intake and 13% lower weight gain than the controls fed ad libitum, but these parameters were similar in the pair-fed groups. At necropsy, no differences were found in absolute and relative organ weights, but the fundic region of the stomach of pigs fed the deoxynivalenol diet was more corrugated than that of either of the controls. When compared with both set of controls, the serum protein concentration and beta-globulin levels were reduced, although these differences had disappeared by the end of the 6-week experiment (Rotter et al., 1995).

A 100-day feeding trial was conducted to evaluate the effect of including deoxynivalenol-contaminated oats in the feed of groups of seven to nine growing pigs with initial weight of 25 kg to provide concentrations of 0.5, 1, 2, and 4 mg/kg in the complete diets, equivalent to 0.02, 0.04, 0.08, and 0.16 mg/kg bw per day. Performance was recorded as weight gain, feed intake, efficiency of feed use, and carcass quality. Restricted feeding was compared to feeding ad libitum. At the highest concentration of deoxynivalenol, feed intake, weight gain, and efficiency of feed use were decreased throughout the experiment. The groups fed diets containing the two highest concentrations of deoxynivalenol showed a dose-related decrease in weight gain during the first 8 weeks on experimental diets. No effects were observed in groups fed diets containing deoxynivalenol at 0.5 or 1 mg/kg. The carcass quality was not affected at any concentration (Bergsjø et al., 1992). Since the control diet contained deoxynivalenol at 0.1–0.4 mg/kg, changes due to low added concentrations of deoxynivalenol could not be detected.

Oats naturally contaminated with deoxynivalenol were included in feed mixtures at graded levels and given to groups of 7–11 female or castrated male growing pigs (59 days old, 21 kg) for 95 days. The concentrations of deoxynivalenol were 0, 0.7, 1.7, and 3.5 mg/kg of complete feed mixture given ad libitum, equal to 0.04, 0.1, and 0.2 mg/kg bw per day. Feed consumption, body-weight gain, weight at slaughter, biochemical and haematological data including serum immunoglobulin (Ig) A, clinical condition, and pathological and histopathological effects post mortem were recorded. The group that received the highest dose had significantly decreased body-weight gain throughout the experiment, decreased weight at slaughter, and reduced feed use efficiency. At the same concentration, the weight of the liver was increased, the concentrations of serum protein and albumin were decreased, and packed blood cell volume and serum calcium and phosphorus concentrations fell transiently. At the two higher doses, a statistically significant, dose-related decrease in daily feed consumption was observed. No other effects on haematological, biochemical, or immunological parameters were observed. Carcass quality was not affected in any group. The authors concluded that significant effects in growing pigs can be observed at a dietary deoxynivalenol concentration of 1.7 mg/kg, originating from naturally contaminated oats included in a diet that was otherwise adequate and contained only minor traces of other mycotoxins (zearalenone, 0.75 mg/kg of feed) (Bergsjø et al., 1993a).

Groups of three castrated male Yorkshire pigs weighing 28 kg were given diets containing pure deoxynivalenol at a concentration of 4.7 mg/kg, maize inoculated with various strains of F. graminearum providing deoxynivalenol at 2.1–5.2 mg/kg of diet, uncontaminated maize, or naturally contaminated wheat ad libitum for 56 days. The feed intake of pigs receiving the diet containing pure deoxynivalenol at 0.19 mg/kg was reduced by 29% after 1 week and 18% after 7 weeks, and their weight gain was reduced by 27% and 20%, respectively, although the differences were not significant. The reductions in feed consumption and weight gain were generally greater in pigs given the diets inoculated with maize, reaching 40% and 37% after 7 weeks on a diet containing deoxynivalenol at 5.2 mg/kg (0.2 mg/kg bw per day). The difference was attributed to factors such as other fungal metabolites and differences in storage of the maize. No emesis occurred (Foster et al., 1986).

The toxic effects of deoxynivalenol were examined in castrated male Yorkshire pigs weighing 10–13 kg given feed into which deoxynivalenol was incorporated at a concentration of 0, 1, or 3 mg/kg, either as the purified toxin or as naturally contaminated maize, for 32 days. The estimated intakes of deoxynivalenol were 0.08 and 0.24 mg/kg bw per day of purified toxin and 0.09 and 0.22 mg/kg bw from the naturally contaminated feed. The diet also contained 7% 15-acetyldeoxynivalenol and 3% nivalenol. Growth performance and blood biochemical and haematological parameters were monitored throughout the study. At the higher concentrations, significantly reduced feed consumption and body-weight gain were evident soon after the start of feeding. While the weight gain of pigs fed the diet containing purified deoxynivalenol recovered after several days, the values for pigs fed the naturally contaminated diet remained depressed throughout the study. These observations might reflect the presence of other, unidentified toxic compounds in the naturally contaminated grain. Generally, the blood chemical parameters of pigs fed the contaminated diets were not different from those of controls, with the exception of reduced serum concentrations of alpha-globulin (significant at the highest concentration of either pure deoxynivalenol or naturally contaminated maize at day 32) and possibly increased cortisol concentrations in animals receiving the highest concentration in either diet. The effect of deoxynivalenol on the alpha-globulin fraction might have been independent of the feed refusal syndrome associated with this toxin. Alterations in several haematological end-points, including a higher erythroyte count, erythrocyte volume fraction, and platelet count, occurred sporadically at 3 mg/kg of either diet; however, these effects could not be separated from the influence of decreased feed intake and were of limited value in diagnosing the effects of dietary deoxynivalenol in pigs (Prelusky et al., 1994).

Groups of three to six pigs weighing 60 kg were used to study the health effects of purified deoxynivalenol and ochratoxin A in their feed, singly or in combination, and the presence of residues 90 days after intake. The pigs received diets containing ochratoxin A at 0.1 mg/kg with deoxynivalenol at 1 mg/kg, equivalent to 0.004 mg of ochratoxin A and 0.04 mg of deoxynivalenol per kg bw, respectively; ochratoxin A alone at 0.1 mg/kg; or deoxynivalenol alone at 1 mg/kg. Two controls received feed containing neither ochratoxin A nor deoxynivalenol. The pigs that received mycotoxins in their feed did not show clinical or haematological changes. The pigs that received both mycotoxins had hyperaemia in the gastric mucosa, and changes in the tubular epithelium were observed in one animal in each treated group. Few pathological lesions were found, but the Committee noted that there were few animals in the study. The observed antibody titres against pseudorabies (Aujeszky disease or ‘mad itch’), as a measure of effects on the immune system, suggest that non-specific defence mechanisms were not affected. The mean concentration of ochratoxin A in the kidneys of animals treated with both toxins was about 50% higher than that in the group given ochratoxin A alone, indicating a possible interaction. The concentration of ochratoxin A also appeared to be slightly increased in muscle of animals receiving both mycotoxins (Lusky et al., 1998).

Semi-synthetic potato-based, grain-free diets containing deoxynivalenol purified from inoculated rice cultures at a concentration of 0, 0.3, 0.6, or 1.2 mg/kg were fed with restriction to groups of nine sows weighing 10 kg over a period of 8 weeks, equivalent to 0, 0.012, 0.024, and 0.048 mg/kg bw per day. Body weight, biochemical and haematological end-points including serum IgA and insulin-like-growth factor-I, crude protein content in faeces, general condition, and pathological and histological findings post mortem were recorded. No significant effect was seen on weight gain, and the maximum body weight in all groups was about 30 kg. These results correspond to the results of clinical chemical and histological investigations. Serum IgA levels were increased by about 30% at the two higher doses, but these changes were not statistically significant (Götz-Schröm et al., 1998). Additional details were provided by M. Lauber (University of Hohenheim), who noted that the water content of the diet was high and may have affected feed consumption and weight gain.

The same group conducted three further unpublished trials. In the first, groups of five castrated male pigs weighing 10 kg were fed, with restriction, semi-synthetic potato-based, grain-free diets containing deoxynivalenol purified from inoculated rice cultures at a concentration of 0 or 4 mg/kg or wheat-based diets consisting of clean wheat or wheat inoculated with Fusarium culmorum, for 4 weeks, equivalent to an intake of deoxynivalenol of 0 or 0.16 mg/kg bw per day. The inoculated wheat also contained 5.8–12% 3-acetyldeoxynivalenol, 15-acetyldeoxynivalenol, and small quantities of nivalenol. The second and third trials were similar to the first, except that the diets were fed ad libitum and the dietary concentrations of deoxynivalenol were 0, 4, or 6 mg/kg diet, equivalent to 0, 0.16, and 0.24 mg/kg bw per day. The results were presented in graphical form. In the first trial, the maximum weight at the end of 4 weeks was about 28 kg in all groups; in the second trial, a dose-related decrease in body-weight gain was seen at each week, with weights of 35 kg for controls, 30 kg at 4 mg/kg, and 26 kg at 6 mg/kg at the end of 4 weeks of dosing. In the third trial, the maximum weight at the end of 4 weeks was about 30 kg in all groups. De-epoxidation of deoxynivalenol to the de-epoxy metabolite in faeces was more extensive with the wheat-based than the potato-based diets (information provided by M. Lauber, University of Hohenheim). The Committee noted that data on food intake were not provided, and only limited statistical analysis was done. Animals gained less weight on the potato-based diet than on the wheat-based diet, even when fed ad libitum.

Ruminants: The feed consumption of 10 non-lactating dairy cows decreased slightly when a wheat–oats diet containing deoxynivalenol at 6 mg/kg was fed at a rate of 1 mg/kg bw per day, with hay offered ad libitum. In surveys of Canadian grains carried out during the early 1980s, the deoxynivalenol content (maximum, 8.5 mg/kg) in eastern Canadian wheat was probably not high enough to account for reports of feed refusal, vomiting, and reproductive problems in livestock operations. This conclusion is based partly on the fact that formulated diets contain a maximum of 70–80% wheat. Consequently, the actual deoxynivalenol content of diets fed to farm animals would be much lower (Trenholm et al., 1984). The Committee noted that insufficient experimental details were provided.

Ruminants tend to be less susceptible to the effects of deoxynivalenol in feed concentrate, which is usually provided in addition to hay. Recent studies are summarized in Table 3.

Equidae: The results of one study are shown in Table 3.

Primates: In a limited study, changes in haemostasis were seen after single oral administration of deoxynivalenol at 1, 5, 10, 25, or 50 mg/kg bw or oral administration of deoxynivalenol at 1 or 5 mg/kg bw per day for 2 weeks to groups of one or two Macaca rhesus monkeys, with normalization of blood coagulation parameters within 45–60 days (Fomenko et al., 1991).

Groups of two male and two female infant cynomolgus monkeys were given pure deoxynivalenol at a dose of 0, 1, 2, or 5 mg/kg bw per day in milk by gavage for 200 days. Two males at the highest dose died during the first week of the study. In these animals, the histological effects, including atrophy of the thymus and spleen, were similar to those seen in rodents given deoxynivalenol, and the relative weight of the thymus was reduced. No significant pathological findings were observed in any other monkeys (F. Iverson, Health Canada, personal communication, 1986).

2.2.3 Long-term studies of toxicity and carcinogenicity

Mice: In a 2-year study, groups of 50 male and 50 female B6C3F₁ mice were given diets containing deoxynivalenol (purity, > 95%; no 3-acetyl- or 15-acetyldeoxy-nivalenol) at a concentration of 0, 1, 5, or 10 mg/kg, equal to 0, 0.1, 0.5, and 1.1 mg/kg bw per day in males and 0, 0.1, 0.7, andor 1.6 mg/kg bw per day in females. Survival was not significantly affected. Average daily food consumption was unchanged in females, but that of males was significantly reduced by about 8% at the two higher doses. The graphical presentation of body-weight changes indicated that the decreases in body weight (and in body-weight gain) at 500 days were 8.7% (13%) at 1 mg/kg, 21% (32%) at 5 mg/kg, and 38% (56%) at 10 mg/kg of diet in females and 1% (1.6%), 6.8% (11%), and 21% (33%) for males, respectively. Females, showed a 56% increase in serum IgA and a < 10 % increase in IgG at 5 and 10 mg/kg of feed, and there were sporadic changes in haematological and clinical chemical end-points; however, these changes were considered not to be biologically relevant. The relative weight of the liver was decreased in males at 5 and 10 mg/kg; at 10 mg/kg, the relative weight of the spleen was decreased and the relative weight of the testis significantly increased. No increase in the incidence of preneoplastic or neoplastic changes was observed. In fact, there was a statistically significant, dose-related decrease in the incidences of preneoplastic and neoplastic lesions in the liver and in that of non-neoplastic lesions affecting large islets of the pancreas. In the liver, this negative trend probably resulted from the known positive correlation between body weight and the appearance of spontaneous hepatic neoplasms in this strain of mouse. The NOEL was 1 mg/kg of diet, equal to 0.1 mg/kg bw per day (Iverson et al., 1995).

Deoxynivalenol was tested for its potential to initiate or promote skin tumours in a two-stage regimen in female Sencar mice. Initiation was tested by applying a single topical dose of 200 µg followed by multiple treatments with the promoter 12-O-tetradecanoylphorbol 13-acetate. The test for promotion involved initiation with the carcinogen 7,12-dimethylbenz[a]anthracene followed by multiple treatments with 50 µg of deoxynivalenol. Appropriate control groups were included. The mice were observed for 26 weeks, and skin tumours were counted. Deoxynivalenol was neither an initiator nor a promoter. When it was tested as an initiator, no statistically significant difference was found in the cumulative number of tumours or the number of tumour-bearing mice. When it was administered as a promoter, no tumours were observed. Histopathological examination of the skin showed that deoxynivalenol induced mild diffuse squamous hyperplasia, but there was no progression of the lesion to neoplasia (Lambert et al., 1995).

2.2.4 Genotoxicity

The results of studies for genotoxicity summarized in Table 4 indicate that deoxynivalenol did not cause gene mutation in vitro but that it caused chromosomal aberration in vitro and in vivo.