Allyl alcohol

Endpoint summary

Key value for chemical safety assessment

Effects on fertility

Description of key information

In a rat dominant lethal test (method equivalent to OECD 478) allyl alcohol did not affect male reproductive parameters,including testis weight and sperm count. In the rat OECD 414 study, two dams in both the 35 and 50 mg/kg/day groups had entirely resorbed litters that is judged to be not secondary to maternal toxicity. The NOAEL from this study would be 10 mg/kg/day In the rabbit OECD 414 study does exposed to 20 mg/kg/day had evidence with interference of the maintenance of pregnancy. The NOAEL from this study would be 10 mg/kg/day.

Effect on fertility: via oral route

Endpoint conclusion:

adverse effect observed

Dose descriptor:

NOAEL

10 mg/kg bw/day

Additional information

In the reproductive and developmental toxicity screening key study conducted in accordance with OECD Guideline 421, male and female Crj: CD(SD) rats were dosed daily with allyl alcohol at 2, 8 and 40 mg/kg (by oral gavage). Males were dosed 14 days pre-mating, then during the mating period until necropsy; females were dosed for 14 days pre-mating, then during mating, gestation period and delivery up to day 3 of lactation. At 40 mg/kg/day, hepatotoxicity was evident in both sexes, together with cases of squamous epithelium hyperplasia in the forestomach of males; thymic atrophy plus hyperplasia of ovarian luteal cells was seen in females, as was mean oestrus cycle prolongation (dioestrus extension: considered secondary to deteriorated general condition). Histopathology found no abnormalities of reproductive organs and no treatment-related effect on reproductive performance was seen. In high-dose males, no effects on testis or epididymis weight, or histopathological changes in testes or epididymides were observed. In females of this high dose group, oestrous cycle length was significantly prolonged, with extended dioestrus. A single case of total litter loss (mainly between birth and Day 4 post parturition) was observed. NOEL values of 40 mg/kg/day (males) and 8 mg/kg/day (females and offspring) were concluded.

In 14-week rodent oral toxicity studies (methods similar to OECD 408: maximum allyl alcohol dosages 25 mg/kg/day in rats, 50 mg/kg/day in mice), male reproductive organ weights showed no effects of allyl alcohol treatment. Spermatid counts and measurements of concentration and motility of epididymal spermatozoa also showed no treatment-related changes, and histopathology examinations of the clitoral, prostate and preputial glands, ovary, testis and uterus from high-dose rats and mice found no treatment-related changes.

 

In an OECD 416-equivalent 2-generation study, rats were dosed orally with acrolein (the major allyl alcohol metabolite) at 1, 3 or 6 mg/kg/day over a prolonged period (P generation 93 – 130 doses, F1 generation 104-125 doses). Forestomach and glandular stomach lesions were seen at 3 and 6 mg/kg/day, but no effects were seen in reproductive organs. Among P and F1 males and females, 1 mg/kg/day represented the NOAEL for systemic toxicity. No effects on reproductive parameters were seen, except for reduced F1 pup bodyweights during the lactation period (a possible effect of maternal toxicity and/or reproductive toxicity). No other indications of reproductive toxicity were seen and histopathological investigations found no changes in reproductive organs of P or F1 animals. Based on evidence of systemic effects (stomach lesions, reduced bodyweight gains), NOAELs for adult toxicity (P and F1, both sexes) were 1 mg/kg/day; NOAELs for reproduction were 6 mg/kg/day in P and F1 males (maximum tested dosage) , 3 mg/kg/day in P and F1 females (assuming F1 bodyweight effects during lactation were a reproductive effect). The NOAEL for offspring was 3 mg/kg/day (based on the F1 pup bodyweight reductions).

In the OECD 414 rat study, two dams in both the 35 and 50 mg/kg/day groups (that survived to necropsy) had entirely reabsorbed litters. In addition, another 35 mg/kg/day dam that died during the treatment period had an entirely reabsorbed litter. It has to be noted that additional signs of significant maternal toxicity was noted at these dose levels, particularly in the one and six dams that died during the treatment period at the 35 and 50 mg/kg/day dose levels, respectively. Since the incidence of postimplantation loss within the remaining dams that survived to necropsy (in these groups) was not increased, this suggests that the toxic effect was most likely on the dam, rather than the conceptus. For this reason, it is suggested that the classification and labeling exercise be considered for reproductive (fertility) toxicity rather than for developmental toxicity. The OECD 414 rabbit study used a high dose level of 20 mg/kg/day and significant mortality occurred at this dose level. Evidence of an interference with the maintenance of pregnancy in this study was limited to dark red intrauterine contents in the animals that died and abortion of one dead fetus in one of the dams that survived to necropsy. It is interesting in this study that the rabbits did not demonstrate any evidence of liver toxicity when examined at necropsy. The endpoint of concern for GHS reproductive and developmental toxicity classification and labeling is the increased incidence of postimplantation loss, either as total litter loss/abortion, increased early or late resorptions or decreased litter size (in the rat 421 study). While it is normal to expect a small number of postimplantation losses within individual litters within a study, the presence of total litter loss is a rare event and certainly the presence of more than one dam/group with total litter loss is very rare. Within the classification guidance document, this endpoint is discussed as “pregnancy outcomes”, “fertility” or “death of the developing organism” and as such can be considered either a reproductive or developmental outcome. Under both scenarios, the loss of implantations during pregnancy is considered a serious outcome and worthy of consideration for classification and labeling.

Short description of key information:
Allyl alcohol has shown no evidence of toxicity to reproductive organs in subchronic rodent toxicity studies (rat and mouse 14-week studies, methods equivalent to OECD 408). In a rat dominant lethal test (method equivalent to OECD 478) allyl alcohol did not affect male reproductive parameters, including testis weight and sperm count. In an OECD 421 rat screening test for reproductive and developmental toxicity, histopathological and functional evidence of the effect of the test subsance on altered ovarian function was observed. The major allyl alcohol metabolite acrolein provides evidence of inactivity in a rat 2 -generation study (method equivalent to OECD 416), showing no adverse effects indicative of selective reprotoxicity. Data from the two OECD 414 studies, although designed to assess developmental effects, provide evidence of total litter loss which can be considered either a reproductive or developmental outcome.

Effects on developmental toxicity

Description of key information

Both an OECD 421 screening study and and OECD 414 study in rats found no evidence of teratogenicity or foetotoxicity (unrelated to maternal toxicity). In rabbits an OECD 414 study and the associated dose range finding study also found no evidence of teratogenicity or foetotoxicity (unrelated to maternal toxicity). These negative findings are supported by a rabbit teratology study performed with the major allyl alcohol metabolite, acrolein: again no evidence of teratogenicity or selective foetotoxicity were seen.

Effect on developmental toxicity: via oral route

Dose descriptor:

NOAEL

10 mg/kg bw/day

Additional information

The OECD 421 reproductive and developmental toxicity screening key study previously described found no induction of foetal malformations and no significant increase in postimplantation losses at dosages up to 40 mg/kg/day.

In an OECD 414 developmental toxicity test, groups of 25 mated female rats were dosed orally with allyl alcohol at 10, 35 or 50 mg/kg/day on gestation days 6 – 19. Severe systemic toxicity at the two higher dosages was characterised by clinical signs, deaths (6 at 50 mg/kg/day, 1 at 35 mg/kg/day), reduced bodyweight gains and macroscopic liver effects, leading to a maternal LOAEL of 10 mg/kg/day. Post-implantation losses were increased at 35 and 50 mg/kg/day, dosages at which maternal toxicity was also observed, but no developmental variations or malformations related to treatment were seen among foetuses from all test groups: 10 mg/kg/day therefore also represented an NOAEL for developmental toxicity.

 

In another OECD 414 developmental toxicity test, a second species, namely New Zealand White rabbits was used. In this study groups of 25 mated female rabbits were gavaged (oral exposure) with allyl alcohol at 5, 10 and 20 mg/kg bw/day. Exposures were once daily on gestation days 7 – 28. Severe toxicity was observed at the highest dose with the death of two dams and a spontaneous abortion. Food consumption was decreased and lower body weight gains were observed to be treatment related. In addition to maternal toxicity, foetal weights from the 20 mg/kg bw/day group were also significantly smaller than controls. However no foetal malformations or developmental variations were associated with the test substance. Based on this data the NOAEL for maternal and developmental toxicity was 10 mg/kg bw/day.

 

A developmental screening dose range finding study was conducted in rabbits prior to the definitive OECD 414 study. 5 female rabbits/group were exposed to 0, 10, 15, 25 or 35 mg/kg bw/day allyl alcohol from gestation days 7 – 28. All animals in the 35 mg/kg bw/day group were found dead or euthanized on gestation day 11. In the 25 mg/kg bw/day group, 2 animals were found dead or euthanized on gestation day 13. Body weight, gravid uterine weight, and food consumption was significantly lower in the 35 mg/kg bw/day group and the 25 mg/kg bw/day animals terminated early, while those animals that survived were statistically the same as controls. No effects were observed at 10 or 15 mg/kg bw/day. No teratogenic effects were observed in this study and the only developmental effect was a decrease in foetal weight at 25 mg/kg bw/day, the same dose that resulted in nearly 50% maternal mortality. The data from this limited study would support a NOAEL for maternal and developmental toxicity of 15 mg/kg bw/day.

 

In a rabbit study of developmental toxicity using methods comparable to OECD 414, does were dosed orally with the allyl alcohol metabolite acrolein at 0.1, 0.75 or 2 mg/kg/day from days 7 to 19 of presumed gestation, after range-findings studies saw major reactions including high mortality (respectively 25% and 100%), abortion, gastric ulcerations and clinical signs in rabbits dosed at 4 and 6 mg/kg/day. Although slight increases in resorptions were seen at 2 mg/kg/day, these were not statistically significant and were associated with maternal toxicity (transient weight loss); all reproductive parameters were unaffected by treatment. No evidence of selective foetotoxicity or teratogenicity was detected: the incidence of foetal malformations and variations in test groups was consistent with that among both concurrent and historical control rabbits, with NOEL values for both foetotoxicity and maternal toxicity being between 0.75 and 2 mg/kg/day. It was concluded that acrolein was not a developmental toxicant or teratogen at dosages not showing maternal toxicity.

Consideration of acute oral toxicity data (rat LD50 99 – 105 mg/kg, rabbit LD50 72 mg/kg [CI 42 -125 mg/kg]) do not suggest any marked inter-species difference in AA toxicity. The demonstrated mechanism of action for the target organ toxicity of allyl alcohol (dependent on alcohol dehydrogenase mediated conversion to acrolein) is one which can be expected to be common to mammalian species, and the observed toxicity profile of allyl alcohol in animal studies is indicative of conversion to acrolein followed by localised tissue damage due to the reactivity (alkylating properties) of this metabolite. The primary site for acrolein release is the liver, supplemented perhaps by limited release in the stomach, and there is no evidence for distal transmission of acrolein and consequent and tissue damage away from the liver or site of primary entry. Tests of effects on reproduction performed with the allyl alcohol metabolite, acrolein, used dosages sufficient to cause overt parental toxicity (2-generation rat study) without finding evidence of selective toxicity to reproduction.

Justification for classification or non-classification

Evidence from the OECD 421 study suggests a direct ovarian effect, with abnormal estrus cycles noted in 4 of 12 female rats during the two-week premating period and luteal cell hypertrophy in 5 of 12 female rats at necropsy, in the 40 mg/kg/day group. In addition, there was one total litter loss in the 40 mg/kg/day group, although it was not possible to determine if this occurred during pregnancy or postnatally. It should be noted that the same hormones that allow for maintenance of pregnancy are required for normal lactation. Although these findings were not severe enough to alter the pregnancy rate or fertility rate in the 421 study, they do provide histopathological and functional evidence of AA altering ovarian function. Furthermore, the specific cells within the ovary (luteal cells) that were affected are cells that perform a critical function in the maintenance of pregnancy. In the OECD 414 rat study, two dams in both the 35 and 50 mg/kg/day groups (that survived to necropsy) had entirely reabsorbed litters. In addition, another 35 mg/kg/day dam that died during the treatment period had an entirely reabsorbed litter. It has to be noted that additional signs of significant maternal toxicity was noted at these dose levels, particularly in the one and six dams that died during the treatment period at the 35 and 50 mg/kg/day dose levels, respectively. Since the incidence of postimplantation loss within the remaining dams that survived to necropsy (in these groups) was not increased, this suggests that the toxic effect was most likely on the dam, rather than the conceptus. For this reason, it is suggested that the classification and labeling exercise be considered for reproductive (fertility) toxicity rather than for developmental toxicity. The OECD 414 rabbit study used a high dose level of 20 mg/kg/day and significant mortality occurred at this dose level. Evidence of an interference with the maintenance of pregnancy in this study was limited to dark red intrauterine contents in the animals that died and abortion of one dead fetus in one of the dams that survived to necropsy. It is interesting in this study that the rabbits did not demonstrate any evidence of liver toxicity when examined at necropsy. The endpoint of concern for GHS reproductive and developmental toxicity classification and labeling is the increased incidence of postimplantation loss, either as total litter loss/abortion, increased early or late resorptions or decreased litter size (in the rat 421 study). While it is normal to expect a small number of postimplantation losses within individual litters within a study, the presence of total litter loss is a rare event and certainly the presence of more than one dam/group with total litter loss is very rare. Within the classification guidance document, this endpoint is discussed as “pregnancy outcomes”, “fertility” or “death of the developing organism” and as such can be considered either a reproductive or developmental outcome. Under both scenarios, the loss of implantations during pregnancy is considered a serious outcome and worthy of consideration for classification and labeling.

 

According to the GHS classification criteria, total litter loss (adverse pregnancy outcome) is considered a major manifestation of reproductive toxicity (UN GHS Section 3.7.1.2, EU CLP Section 3.7.1.3). It is not possible to completely rule out the total litter loss in the rat OECD 414 study as a “non-specific secondary consequence” of other toxic effects. However, the presence of histopathological changes in the OECD 421 study in the specific ovarian cells required to maintain pregnancy and the evidence of functional perturbation as a possible consequence of those changes (i.e. altered estrus cyclicity) suggests a direct effect of AA exposure. The data as reported supports a Classification and Labeling of Category 2 “Suspected human reproductive toxicant” based upon the increased incidence of total pregnancy loss in the rat OECD 414 study and evidence from the OECD 421 study providing supporting evidence of altered ovarian function.

 

A Category 1B designation is not supported given the lack of corroboration in the rabbit studies within the available information. Under the EU CLP regulation (Regulation (EC) No. 1272/2008), the classification of allyl alcohol would be Category 2 (H361f – Suspected of damaging fertility).