2018 年 6 巻 3 号 p. 109-117
Recently, a long-term (1-year) dog toxicity study has not been a mandatory toxicity study for application of agricultural chemical in the United States (US) and the European Union (EU). This study was conducted to propose a guide for making science-based judgement on the necessity of long-term dog toxicity study, which is one of required toxicity studies at toxicological evaluation in Japanese pesticide regulation system. In order to carry out the proposal we analyzed the results of toxicity studies including subacute (3-month) toxicity study in dogs or toxicity studies in other species in the pesticide evaluation reports published by the Food Safety Commission of Japan (FSCJ), the responsible regulatory body for toxicological evaluation of pesticides in food. In the analysis of evaluation reports of 286 pesticides ADI (acceptable daily intake) of 93 pesticides (32.5%) were established based on dog studies. The ADIs of 74 pesticides among them, however were not considered to have a big influence if the long-term dog toxicity study was omitted. With regard to the other four agents the possibility that the long-term dog study becomes unnecessary was considered by adding detailed examination. With respect to the remaining 15 agents, we could not judge that long-term dog study were unnecessary. The analysis indicated that the dog long term test could be omitted in most cases. On the other hand, it should be considered carefully necessity of the long-term dog study when the toxicological profiles observed in dogs and rats were different, when the toxicity susceptibility in dogs was considered high, when no no-observed-adverse-effect level (NOAEL) is specified in subacute toxicity study in dogs or when bioaccumulation in dogs is concerned. We also noted that the studies already conducted for pesticide registered previously should be used for their hazard evaluation.
Assessors of pesticides have often encountered in large species differences of toxicities between rodents and non-rodents at their toxicological evaluations. It is not hard to imagine for them that the wide species difference in pharmacokinetics of agricultural chemicals is one of the major factors. However, there is little information on pharmacokinetics in dogs because the major species to clarify metabolic pathways of chemicals submitted by applicants is rat, but not dog. As for a different viewpoint, contribution to animal welfare, 3Rs (reduction, refinement and replacement) in using experimental animals in toxicity study is also an important subject in recently.
Various types of toxicity studies using rodents or non-rodents have been required for safety assessment of pesticides. Dogs are thought to be susceptible to toxicity of certain number of pesticides and have been regarded as the second animal species (non-rodent) to be used for animal experiments. Recently, a long-term (1-year) dog toxicity study has not been a mandatory toxicity study for application of agricultural chemical in the United States (US) of America and the European Union (EU)1,2. The US Environmental Protection Agency (EPA) requires it when the elimination of the pesticide from body is slow and its accumulation is high.
In Japan, for about one third of pesticide evaluated in FSCJ, their ADIs, human health based guidance values for long-term exposure, have been established based on the results in chronic toxicity studies in dogs (Fig. 1). Conversely, this meant that dog toxicity studies were not critical to establish the ADI for remaining pesticides. However, there has not been no criterion which indicates appropriate tier approaching methods at toxicological evaluation of method appropriately, considering animal species difference and administration period in Japan. Such situation in toxicological evaluation of pesticides is considered not to be adapted to toxicological progress on the species difference or international trend of toxicological evaluation of pesticides.
Species or type of toxicity study used for ADI setting in Japan. The label in chart indicate species, the number of pesticides, and their percentage in the left figure. In the right figure, the label indicates chronic study, the number of pesticides, and its percentage.
Chronic, chronic study; Repro/Dev, reproductive or developmental toxicity study: Subacute, subacute toxicity study; Two compounds that were not established their ADIs were excluded.
Given the current circumstance, we should appreciate that it is not scientific to simply eliminate the dog’s chronic toxicity study only by the international cooperation aspect alone, and that the assessors scientifically verified the past evaluation in Japan. Therefore to establish new approach considering species difference is an urgent task in toxicological evaluation of pesticides in Japan. Ultimately, revision of dog toxicity studies according to Japanese situation should be done based on the verification. In the present study, we analyzed the outcome of toxicological studies described in the monographs of pesticide evaluated in Japan comprehensively to propose tiered approach to judge necessity and/or unnecessity of long-term study in dogs based on the species difference of toxicological profiles and dose-responses more scientifically. The proposal approach will be reduced the number of animals to be tested necessary for evaluating the toxicity of agricultural chemicals, and as a result contribute to 3Rs.
In this study, we analyzed the results of all of the toxicity studies conducted and their ADIs as well as the types of toxicity study used ADI setting, using the evaluation reports of 286 pesticides already performed their toxicological evaluations by FSCJ and published its contents by them3).
1. Analysis on the Experimental Designs and Points of Departure of Subacute and Long-term Toxicity Studies in Dogs and RatsThe first step of analysis was to check experimental designs including the common ratio of dosing or administration route of test substance, and NOAEL/low-observed-adverse-effect level (LOAEL) of subacute (3-month) and long-term (1-year) toxicity studies of dogs and rats of each pesticide using evaluation reports published by FSCJ.
2. Analysis on Toxicological Necessity of Dog Long-term Toxicity StudyThe second step was to review the necessity of a long-term study in dogs, we comprehensively analyzed the major toxicities, target organs, NOAEL/LOAEL values of dog and rat studies, and endpoints at LOAELs, and compared the results in dog studies with those in rats.
Regarding analysis on the NOAEL and LOAEL values obtained in each toxicity study, we referred to the criteria on the presence or absence of difference of the values between long-term and subacute studies in dogs, or between rats and dogs in subacute studies by Dellarco et al.4). Briefly, Dellarco et al.4) reported that the toxicological profiles of subacute and long-term toxicity studies in dogs were almost consistent among 70 of 110, and that in comparison of these NOAEL and LOAEL between subacute and long-term studies, the differences of values were within 1.5 times. They reported that 2% of the total pesticides examined were judged to require long-term toxicity study in USEPA4,5,6). Kobel. et al.6) reported that a ratio of ADI or NOAELs within 2 times among animal species could be judged no difference due to general variation in toxicity test such as biological reactivity and dose setting4,5). Both reports described above concluded that approximately 3 times difference of NOAEL or LOAELs did not affect ADI when integrated the relationship between the toxicity findings of each pesticide, dose relationship of the findings and their common ratios4,5,6). With reference to these reports we decided to judge that there is no substantial difference when these values are less than 2 times in comparison of NOAEL and LOAEL with between subacute and long-term or between animal species.
The common ratios, which mean fold intervals between dosage, used in dose setting for toxicity studies are shown in Table 1. The common ratios which were 3 to 10 were major types (approximately 50%) for both of dogs and rat studies. The wide intervals between dosages which were set 10 times or more were observed in more than 30% of subacute or long-term toxicity studies in both species. Especially, seventeen studies showed the wide intervals exceeding 20 in rats. In dogs, such wide range intervals were found in 20 studies, the results showing that ratio in rats might be slightly larger than dogs.
| Common ratios used in dog or rat studies | No. of pesticides (%) | |||
| Dogs | Rats | |||
| Short-term studies | ||||
| 3 or more and less than 5 | 71 | (24.8) | 72 | (25.2) |
| 5 or more and less than 10 | 69 | (24.1) | 88 | (30.8) |
| More than 10 | 71 | (24.9) | 93 | (32.5) |
| Sub-classification of the common ratios more than 10 | ||||
| [10] | 55 | 67 | ||
| [11a-20b] | 8 | 9 | ||
| [21-30] | 2 | 8 | ||
| [31-40] | 4 | 1 | ||
| [41-50] | 0 | 3 | ||
| [51-] | 2 | 5 | ||
| Short-term studies in rats or dogs
not conducted |
51 | (17.8) | 12 | (4.2) |
| Total | 286 | (100) | 286 | (100) |
| Long-term studies | ||||
| Less than 3 | 23 | (8.0) | 13 | (4.5) |
| 3 or more and less than 5 | 67 | (23.4) | 61 | (21.3) |
| 5 or more and less than 10 | 90 | (31.5) | 92 | (32.2) |
| More than 10 | 94 | (32.9) | 118 | (41.3) |
| Sub-classification of the common ratios more than 10 | ||||
| [10] | 72 | 80 | ||
| [11-20] | 12 | 21 | ||
| [21-30] | 3 | 8 | ||
| [31-40] | 3 | 2 | ||
| [41-50] | 1 | 1 | ||
| [51-] | 3 | 6 | ||
| Long-term studies in rats or dogs
not conducted |
12 | (4.2) | 2 | (0.7) |
| Total | 286 | (100) | 286 | (100) |
# : fold intervals between dosages.
a,b : a-b indicates the range of common ratios.
Each NOAELs of subacute study and long-term studies for 286 pesticides was compared (Fig. 2) in rats and dogs, respectively. A tendency that NOAELs in long-term toxicity study were lower than those in subacute toxicity study were somewhat increased in rats compared to dogs.
Ratios of NOAELs for subacute toxicity to those for chronic toxicity in studies of dogs or rats. The X label indicates the ratios.
<0.1, the ratio was less than 0.1; 0.1<0.2, the ratio was 0.1 or more and less than 0.2; 0.5<1.67, the ratio was 0.1 or more and less than 1.67; 1.67<5, the ratio was 1.67 or more and less than 5; 5<10, the ratio was 5 or more and less than 10; 10<30, the ratio was 10 or more and less than 30; 30<, the ratio was 30 or more.
Each NOAEL of the 286 pesticides undergoing the toxicity studies in dogs and rats was compared (Table 2). The lower value of the subacute and the long-term studies was used as NOAEL for each animal in this analysis. Pesticides whose NOAEL of dogs was 5 times or lower than rats, and NOAELs with 5 times or more than rats were 11.3% and 14.4%, respectively, and their difference between animals in the overall distribution was not clear.
| The ratio of NOAEL in dogs to that in rats | The number of pesticides | (%) |
| More than 10 | 28 | 10.1 |
| More than 5, and less than 10 | 12 | 4.3 |
| More than 1.5, and less than 5 | 68 | 24.6 |
| More than 0.67, and less than 1.5 | 73 | 26.4 |
| More than 0.2, and less than 0.67 | 64 | 23.2 |
| More than 0.1, and less than 0.2 | 17 | 6.2 |
| Less than 0.1 | 14 | 5.1 |
| Total | 276 |
Among the 286 pesticides analyzed in the present study, toxicological findings observed dog long-term toxicity studies were critical endpoints for setting ADI in 93 pesticides (32.5%) (Table 3). Of these 93 agents, the same dose levels as the NOAEL of the dog studies were found in the 6 pesticides obtained from the toxicity studies using other species than dogs. In comparison with the results of subacute and long-term toxicity studies in dogs, 28 pesticides showed less than 2 times of ratios of NOAEL for short-term toxicity to long-term ones. Similarly, there were 10 pesticides showed less than 2 times of ratios of LOAEL for long-term toxicity to that for subacute toxicity in dogs. NOAELs or LOAELs for long-term toxicity in dogs were compared for corresponding values in long-term toxicity studies in rats. There were 16 pesticides which ratios of NOAELs for dogs to those for rats were less than 2 times, and there were 5 pesticides which ratios of LOAELs in dogs to those in dogs were less than 2 times. For the total of 65 pesticides described above, it is assumed that the ADI values derived from other toxicity studies do not differ greatly even if data of long-term dog studies have not been obtained. Therefore, it could be judged that the long-term dog study is not necessarily required for them.
| Sub-classification | The number of pesticides |
| Same levels of NOAEL obtained from other study(ies) as long-term study in dogs | 6 |
| Similar levels of NOAEL obtained from sub-acute toxicity study in dogs to long-term toxicity study in dogs* | 28 |
| Similar levels of LOAEL for sub-acute toxicity study in dogs to long-term toxicity study in dogs* | 10 |
| A difference in NOAEL between long-term toxicity study in rats and dogs was less than 2 times | 16 |
| A difference in LOAEL between long-term toxicity study in rats and dogs was less than 2 times | 5 |
| More than 2 times of difference in NOAEL (if specified NOAEL, LOAEL was used for comparison) between long-term toxicity study in dogs and toxicity study compared | 28 |
*, The difference of NOAEL among them was less than 2.
The pesticide evaluation reports were reviewed in detail for 28 pesticides whose ratios of NOAEL or LOAEL (if not specified NOAEL) values in the dog long-term toxicity study were twice or higher than the corresponding values of the dog subacute or rat long-term toxicity studies (Table 4). As for the 6 pesticides, the ratios of NOAELs of dog long-term toxicity study to those in dog subacute toxicity studies, or to rat long-term toxicity studies were nearly doubled in 5 compounds (Ametrin, Carpromid, Cloransulam-methyl, Thidiazuron, Pyridaben), LOAEL being in 1 compound (TCMTB). The ratios were, Carpromid (2.2 times to dog subacute toxicity study), Ametrin (2.9 times to rats, Thidiazuron (2.0 times to rats), Cloransulam-methyl (2.0 times to rats), Pyridaben (2.0 times to dog subacute or 2.1 times to rats). In TCMTB, the ratio of LOAEL for dog long-term study to NOAEL for rat long-term study was 2.1 times. Considering their experimental designs including the common ratios of these studies, the gaps of dose showing toxicity between the long-term study in dogs and short-term study in dogs or long-term toxicity studies in rats were considered small. With regard to Pyriproxyfen and Dinotefuran, the NOAELs of rat two-generation reproductive toxicity study and/or rabbit development toxicity study are considered to be applicable for the second option for setting ADI, because their NOAELs were less than twice than long-term toxicity stud in dogs (Table 4). Comparison of NOAEL or LOAEL values of Ametoctradin was considered to possess little toxicological significance due to very weak toxicity of this compound (LOAEL: 800 mg/kg body weight or higher). Therefore, even if none of the chronic toxicity data in dogs were obtained from these 9 pesticides described above, the influence on the ADI setting was considered to be limited.
In the cases of Imicyafos, Aldoxycarb and Ethoprofos, their toxicity profile was inhibition of ChE activity, and the endpoints at their LOAELs were also the inhibitory effects of ChE activity (Table 4). This type of toxicological profile is a clear critical endpoint. Therefore, we could minimize differences of NOAEL or LOAEL in dog long-term toxicity study from those in dog subacute toxicity study, or rat long-term toxicity study if continuous and comprehensive analysis on ChE activity or related items are conducted under considering species differences.
Fentrazamide showed common hepatotoxicity between dogs and rats, and more than 3-fold difference was found in the ratios of NOAELs or LOAELs in dogs to those in rats (LOAEL data are not shown) (Table 4). The critical endpoint for ADI establishing was ALP increase found in female dogs only. ALP isozymes are derived from not only liver but also bone or other tissues. In addition, a dog-specific isozyme, corticosteroid ALP, has been reported to be increased by transportation stress7,8,9). Therefore, the increased ALP level in dogs is considered to result in large physiological fluctuation of individual differences of body and bone growth until approximately 12 months of age. Some stress due to treatment with pesticides might be related. Therefore, if statistically significant increase in ALP without other changes which indicated hepatotoxicity was observed in a dog study, it is important to check comprehensively the ALP change whether the increase was caused by hepatotoxicity or not using clinical pathology including ALP isozymes measurements or histopathological examination. Such an additional study approach would reduce differences among animal species in the case of Fentrazamide.
3. Pesticides That Could Not Be Judged Unnecessary for Long-term Dog Study from the Analysis of the Pesticide Evaluation ReportThe common toxicological findings of Cyantraniliprole in dogs and rats were found in the liver, but the dog’s long-term NOAEL and LOAEL differed more than 3-fold from those in the subacute toxicity study in dogs or chronic toxicity study in rats (Table 4). This result indicated that it was not thought that the long-term toxicity study in dogs should be unnecessary.
In Alachlor, diarrhea, salivation and hemosiderin deposition observed in the long-term toxicity study in dogs were the critical endpoints for setting ADI. These findings however were not seen in the dog at 6 months treatment. The findings at LOAEL of Flumeturam were hepatitis, renal tubular atrophy and other findings. In the case of Flumeturam, their subacute toxicity study in dogs was not conducted (Table 4). Therefore, it was unknown whether these findings found in long-term toxicity studies are also observed in short-term studies or not. In addition, the ratios of LOAEL in dog study to NOAEL in the rat chronic toxicity study in Alachlor and Flumeturam were 5 and 2 times, respectively. The ratios of NOAEL for dogs to that in rats were 2.5 and 5-times, respectively. The critical endpoint for establishing ADI of Pyrifluquinazon was mononuclear cell infiltration in olfactory epithelium of nasal cavity in long-term toxicity study in dogs, but the same finding was not detected in dog subacute and rat long-term toxicity studies. The critical findings for establishing ADI of Tiadinil were soft stools and depressed body weight gain observed in dog subacute toxicity study as well, but NOAEL and LOAEL both had a dose difference of 5 times. Additional study should be added to confirm whether these changes of these two compounds have a characteristic for chronic toxicity in dogs.
In comparing rats and dogs long-term toxicity study, NOAEL ratios of Lactofen, MCPA and Heptachlor, and LOAEL ratio of Ethoxyquin were more than 2 times, i.e. 2.5 (=2/0.79), 3.4 (=0.7/0.19), 6 (=0.15/0.025) and 4.8 (=12/2.5) times respectively. Moreover, for MCPA and Heptachlor, NOAELs of long-term toxicity study in dogs were 2.4 (=0.06/0.025) and 5.2 (=1/0.19) times lower than NOAELs of subacute toxicity studies in dogs. For other two pesticides (Lactofen and Ethoxyquin), we could not compare the NOAELs of long-term and subacute toxicity studies in dogs due to lack of data. Data of all these 4 pesticides evaluated in FSCJ were obtained from the evaluation reports (thus, not original) of overseas agency and FSCJ could not access original data of toxicity studies reports. Also, different values of ADIs were sometimes established at national or international toxicological evaluations. Regarding the ADI of Ethoxyquin and Heptachlor, more than 5 times different values between the evaluation agencies were establish, and ADI of MCPA is 2 times different between agencies. With these four agents, we could not judge the necessity of long-term toxicity study in dogs only with the pesticide evaluation report investigated in this analysis.
Imazapic, Propyrisulfuron and Quinclorac showed more than 8-fold difference in the ratios of NOAELs (Propyrisulfuron and Quinclorac) or LOAEL (Imazapic) in dog’s long term toxicity to those in rat long-term toxicity studies. No NOAEL was obtained from the dog long-term study of Imazapic. There were no toxicity found all 3 pesticides in rats, and the NOAELs of rat long-term toxicity study were the dose highest tested. Toxicities at their LOAELs in dog studies were muscular degeneration in great muscle, diaphragm and abdominal wall, necrosis and inflammation found in Imazapic; anemia in Propyriisulfuron, increased kidney weight gain in Quinclorac, respectively. A subacute study in dogs has not been conducted in Imazapic. 4 weeks-toxicity study was conducted in Quinclorac as subacute study but FSCJ could not determine the NOAEL due to that number of dogs per group was two and not enough.
LOAEL and/or NOAEL ratios of Clofenset in dog long-term study to corresponding values in rats were more than 8 times each. The major toxicity findings in dog were inflammation of epididymis, degeneration of seminiferous tubules, and aspermia in dog while the target organs in rats were lung, liver and kidney in Clofenset. Six month but not 1-year toxicity study was conducted in dogs in Norflurazone. In comparison with the 6-month toxicity study with the long-term study in rats, the NOAEL and/or LOAEL ratios in dog study to corresponding values in rats were more than 10 times.
Data of Cinidon-ethyl was very limited for toxicological evaluation. The evaluation was conducted on the basis of reports for toxicological evaluation reports in overseas agency where the NOAEL with extremely few toxicity information, and the NOAEL ratio between long-term toxicity in dogs and subacute or long-term was 5-fold difference.
Consequently, these six pesticides (Imazapic, Propyrisulfuron, Quinclorac, Norflurazone, Clofenset and Cinidon-ethyl) could not be determined the necessity or unnecessity of long-term toxicity study in dogs in this analysis.
4. Summary of This AnalysisIn summary, the necessity of 286 pesticides on the long-term (1-year) toxicity study in dog study was analyzed. Long-term toxicity studies in dogs of 93 pesticides were used as critical points for establishment of ADI. Data of these 93 pesticides were reviewed in detail focusing on necessity of long-term dog study. Seventy-four pesticides were considered that their ADI could be established even if no long-term toxicity study in dogs were conducted. In addition, due to their clear toxicity a possibility that dog chronic study could be omitted was suggested for 4 compounds if detailed analysis could be done. On the other hand, the remaining 15 pesticides (5.2%) were judged impossible to delete long-term toxicity study in dogs based on the data of toxicity described in their evaluation reports. We could not found any similarity in the biological effects or chemical structure of these pesticides.
The results of this analysis suggest that dogs should be recognized a more sensitive species for toxicity detection if clear toxic dose in the subacute toxicity study in dogs are lower than that in rats. The toxic profile of the agent is different and that even if the toxic target organs are the same, it is considered that a long-term test would be necessary. It is also considered that a long-term toxicity study but not further subacute toxicity study in dogs would be a priority if no NOAELs are set in a subacute (3-month) dog studies. In addition, if a toxicological concern for bioaccumulation or different kinetics is raised by toxicological profiles in dogs, careful judgement based on relevance to humans is necessary to consider whether long-term toxicological study in dogs.
This research was commissioned under the “2015 Cabinet Office Research for Assessment of the Effect of Food on Human Health: Development of the new tiered approach for toxicity studies of pesticide considering species difference in “toxicity profile” and “toxicity dose-response” of necessity of 1-year study in dogs- and carcinogenicity in mice” (Topic No. 1501).
The authors have no conflict of interest.
