Registration Dossier

Administrative data

Key value for chemical safety assessment

Effects on fertility

Description of key information

For fertility, there are three non-standard studies in animals, one on female rats (Belyaeva, 1967) and one on male rats and mice (Omura et al., 2002); and one human occupational report study (Belyaeva, 1967). At a first glance, the inhalation study in female rats by Belyaeva (1967) appears to suggest that Sb trioxide might have an adverse effect on fertility after repeated exposure to 250 mg/ m³. However, the results must be regarded as inconclusive and cannot be used for risk assessment, since the study report does not provide a valid description of the overall experimental conditions and the purity of the test substance. Furthermore, since it is well-established that rats are particularly sensitive to inert particle overload (in contrast to humans) which occurs in sub-chronic studies at levels approximately 50-fold below the level in Belyaeva (1967), it can reasonably be assumed that the rats in this study suffered from a massive, non-substance-specific impairment of their respiratory system by overload of lung macrophages and breakdown of their lung clearance, so that any adverse effects can easily be explained as of secondary nature. Significant pulmonary damage, impaired pulmonary function (hypoxia) and secondary changes related to hypoxia (e.g. renal lesions, erythroid hyperplasia) have subsequently been noted in recent chronic inhalation studies with rats and mice at exposure levels ranging from 3 to 30 mg/m3(NTP, 2017).


The gavage study on male rats and mice (Omura, 2002) showed no testicular toxicity after 4 weeks repeated exposure up to 1200 mg/ kg bw. The human case report study on women occupationally exposed to Sb trioxide indicates that Sb trioxide might affect the fertility of female workers. However, this study is inconclusive due to the lack of information on the control group, the exposure situation and the overall workplace environment. Based on these fertility studies in animals and humans, no conclusion on female fertility can be derived.


However, a 90-day oral feeding study in male and female rats with Sb trioxide reported no effects on reproductive organs up to a dose of 1686 mg/kg in males and 1879 mg/kg in females. The effects of Sb trioxide, Sb potassium tartrate and sodium hexahydroxoantimonate upon the fertility of rats and/or mice have been evaluated after oral exposure (Hext, et. al., 1999; Omura et al. 2002; Hansen, 2014a), and i.p. injection (Dieter, 1992). No significant adverse functional or structural impacts upon the reproductive systems of male or female animals have been observed. The NOAEL for effects upon fertility via oral exposure (adjusted for the Sb content of compounds evaluated) is in excess of 1000 mg/kg bw/day.


Data evaluating fertility impacts after inhalation exposure are lacking, but the overall profile of Sb compounds indicates low potential for reproductive toxicity. Fertility effects via inhalation exposure would not be expected given the high oral NOAEL for fertility impacts, coupled with the lack of developmental impacts from high (6.3 mg/m3) inhalation exposure to Sb trioxide (Schroeder, 2003).


In summary, focusing on the very few GLP-compliant, guideline (Klimisch score 1) studies thatshould be used as the basis for effects assessment and classification, there are no definitive guideline studies which assessreproductive function, and data are limited to evaluations performed on chronic toxicity studies. Such studies are available on two Sb 3+ compounds: Sb trioxide and Sb potassium tartrate, and one Sb 5+ compound: sodium hexahydroxoantimonate:

·       In 90-day studies with intraperitoneal injection of Sb potassium tartrate up to 24 mg/kg bw/d for 3 days/week in rats and mice, there were no adverse effects on estrous cycles or sperm count, morphology or motility and no adverse effect on histopathology of the reproductive organs.

·       In a 90-day dietary study in the rat with Sb trioxide, no adverse histopathology was reported for reproductive organs at dosages which exceed the limit dose (1686 mg/kg bw/d in males and 1879 mg/kg bw/d in females).

·       A further published 4-week intermittent dosing study with Sb trioxide and Sb potassium tartrate in rats and mice reported no effect on reproductive organ weight, histopathology, sperm count, motility and morphology at target dosages of 12 or 1200 mg/kg bw/d Sb trioxide and 27.4 mg/kg bw/d Sb potassium tartrate.

Effect on fertility: via oral route
Endpoint conclusion:
no adverse effect observed
Dose descriptor:
1 686 mg/kg bw/day
Study duration:
Quality of whole database:
The effects of antimony trioxide, potassium antimony tartrate and sodium hexahydroxoantimonate upon the fertility of rats and/or mice have been evaluated after oral exposure (Hext, et. al., 1999; Omura et al. 2002; Hansen, 2014a), and i.p. injection (Dieter, 1992). No significant adverse functional or structural impacts upon the reproductive systems of male or female animals have been observed. The quality of the database is medium to high; but could be improved further (cf. scientific opinion in Section 13).
Effect on fertility: via inhalation route
Endpoint conclusion:
no study available
Effect on fertility: via dermal route
Endpoint conclusion:
no study available

Effects on developmental toxicity

Description of key information

The developmental toxicity of oral exposure of rabbits to Sb metal powder (Hansen, 2017) and inhalation exposure of rats and mice to Sb trioxide (Schroeder, 2003) has been studied with little evidence of significant developmental deficits. Oral exposure studies of rabbits to Sb metal powder were complicated by maternal toxicity at the higher levels of Sb exposure. These exposures were associated with lethality, gastric irritation and reductions in maternal food intake and maternal body weight gain. Reductions in fetal weights, increased post-implantation losses and delayed vertebral ossification were observed in parallel to this maternal toxicity and have been documented to result from reduced maternal food intake and growth in the rabbit (Cappon et al., 2005). No impacts of Sb, independent of maternal toxicity, were observed.


Hansen (2014) evaluated the effects of oral gavage with 100, 300 and 1000 mg/kg bw/d sodium hexahydroantimonate upon developmental toxicity in Sprague Dawley rats. No evidence of maternal toxicity was evident up to 1000 mg/kg bw/d of sodium hexahydroantimonate (Sb 5+) although slight retardation of fetal skeletal ossification was observed at 300 and 1000 mg/kg bw/d. The significance of this mild effect is difficult to assess, particularly when contrasted with the observations from injection studies with other Sb 5+ compounds projected to yield systemic Sb 5+ concentrations approximately 10 - 100-fold higher than those achieved via oral gavage. This observed delayed ossification can provisionally be dismissed as not relevant as per Carney and Kimmel (2007).


A separate but growing body of literature has focused upon the therapeutic intravenous administration of Sb 5+ compounds such as meglumine antimoniate for the treatment of leishmaniasis - a parasitic disease state endemic in tropical countries that in and of itself poses health risk to pregnant women and their fetuses. After more than 5 decades of therapeutic use of pentavalent antimonials, a small number of anecdotal case studies have suggested embryotoxicity during treatment for leishmaniasis. Rodent studies have thus evaluated the developmental toxicity of meglumine antimoniate administered to the rat by injection (Miranda et al., 2006; Coelho et al, 2014):


Miranda et al. (2006) evaluated the developmental toxicity of daily subcutaneous injections of 75 – 300 mg Sb(V) /kg bw/day throughout gestation. Reductions in fetal weight and some skeletal and soft tissue abnormalities were evident at 300 mg/kg bw/day and attenuated at 150 mg/kg bw/day, yielding a NOAEL of 75 mg/kg bw/day. The authors suggest that maternal toxicity was not evident in the study but reductions in maternal weight gain during gestation were evident at the highest dose tested, and do indicate the presence of maternal toxicity.


Coelho et al. (2014) expanded upon the work of Miranda et al (2006), essentially conducting an extended one-generation reproductive test study (EOGRTS). Subcutaneous injection of 75 – 300 mg Sb(V)/kg bw/day to pregnant rats was conducted from day 0 to 20 of gestation and through parturition and lactation to post-natal day 21. Embryotoxicity, manifesting as decreases in pup weight and reductions in litter size, were noted to have occurred in the absence of maternal toxicity. However, suppression of maternal weight gain (minus the weight of the uterus) was observed and indicates the subtle onset of maternal toxicity at 300 mg/kg bw/day. Impacts upon pregnancy outcome had an apparent NOAEL of 150 mg Sb 5+/kg bw/d. Post-natal developmental impacts were present but minor. Other than modest decrements in female pup weight gain and exploratory behavior in female pups, no detrimental impacts of treatment were observed on neurobehavioural development, sexual maturation, male fertility or female reproductive performance. The authors concluded that meglumine antimoniate was a weak developmental toxicant and further postulated that the minor effects seen in offspring were potentially the result of decrements in maternal weight gain during gestation that had subsequent negative impacts upon lactation and pup nutrition.


The subcutaneous injection studies with meglumine antimoniate are relevant to pharmacological applications in the treatment of leishmaniasis. Coelho et al. (2014) further note that the injection doses used in the rat reproduction studies are much higher than those employed therapeutically in the treatment of leishmaniasis in humans– typically on the order of 20 mg Sb (V)/kg bw/d – and yield rat blood antimony levels approximately 100-fold higher than therapeutic human blood antimony levels. The systemic Sb levels achieved via s.c. injection are also strikingly higher than systemic levels that will result from physiological routes of exposure. For example, blood Sb levels of 160 µg/g result one hour after s.c. injection of 300 mg/kg bw meglumine antimoniate in the rat (Miranda et al., 2006). In contrast, oral administration of rodents at similar dosing rates is indicated to yield no detectable increases in Sb levels in blood (Coelho et al., 2014). Rodent blood Sb levels after injection are also approximately 10,000-fold higher than the blood Sb levels of humans occupationally exposed to Sb compounds (Wu and Chen, 2017). The mild developmental impacts observed after administration of antimony compounds by injection are thus elicited by systemic antimony levels many-fold higher than can conceivably be produced via physiological routes of exposure due to a combination of extremely limited uptake and/or irritating and emetic effects of highly soluble compounds (Dieter et al., 1991).


Sb compounds administered via physiological exposure routes do not produce developmental toxicity independent of maternal toxicity. The NOEC for effects from inhalation exposure is in excess of 6.3 mg/m3for Sb trioxide. 


In summary, as regardsdevelopmental toxicity, no adverse effects on embryofetal development were observed in a rat inhalation study with Sb trioxide, despite the expected adverse lung pathology in dams. Studies in rats and rabbits using oral administration, effects reported includematernal toxicity, especially in groups exposed to high/excessive quantities of Sb, demonstrated by mortalities, abortions, reduced weight gain and periods of negligible food intake in the dams,with reduced food intake in some individual animals at the intermediate dose level,andretardation of fetal ossification. In rats treated with sodium hexahydroxoantimonate, fetal effects were limited to ossification delays (which are generally accepted as reversible).

Effect on developmental toxicity: via oral route
Endpoint conclusion:
no adverse effect observed
Quality of whole database:
The developmental toxicity of oral exposure of rabbits to antimony metal powder (Hansen, 2017) and inhalation exposure of rats and mice to antimony trioxide (Schroeder, 2003) has been studied with little evidence of significant developmental deficits. The quality of the whole database is high, but the maternal toxicity and delayed ossifications observed merit further research (cf. scientific opinion in Section 13).
Effect on developmental toxicity: via inhalation route
Endpoint conclusion:
adverse effect observed
Dose descriptor:
6.3 mg/m³
Study duration:
Quality of whole database:
two key studies are available; a prenatal developmental toxicity study in rats (which is reliable with minor restrictions (RL=2)) and a rabbit PND study (according to OECD 414). The overall quality of the database is therefore high.
Effect on developmental toxicity: via dermal route
Endpoint conclusion:
no study available

Toxicity to reproduction: other studies

Additional information

In order to assess the potential reproductive toxicity of Sb substances, it is important to understand the mechanism(s) by which Sb compounds seem to cause some level of adverse effects in a number of studies, and the chemical species involved in such a response. The following hypothesis have been put forward based on the available reproductive toxicity information:


Maternal toxicityhas been demonstrated by excessive exposure to Sb substances. As indicated previously, extensive rapid dissolution of Sb substances can produce a disruption of gastrointestinal tract function. It is necessary therefore to distinguish between: i) the adverse effects of disruption of normal food intake, including secondary effects on reproductive outcome, and ii) direct reproductive toxicity.


Published studies investigating fetal effects in an array of Sb substances and routes have limitations (not least the excessive exposure profile achieved by parenteral administration of Sb in the investigation of Sb 5+ compounds designed for medicinal use and not relevant for non-therapeutic human exposures). These available data are also not fully GLP or OECD guideline compliant, and even for the most recent studies, do not provide sufficient detail on maternal adverse effects, and there are no individual data records to assess how such effects may influence embryofetal development.


It is not clear how exposure to Sb influencesfetus ossificationnor whether the delay is reversible, in that normal ossification takes place by weaning. Unless studies include a dedicated assessment of the Ca metabolism of animals exposed to Sb substances, the actual role of Sb and the resulting developmental toxicity assessment cannot be confirmed.


The limited Klimisch score 1 relevant, reliable studies together with the less reliable ones, leave a number of questions unaddressed. There is little indication of effects on reproduction or fertility, but this statement has a relatively low confidence level, and is shadowed by at least two uncertainties: maternal toxicity vs direct developmental toxicity, and the influence of Sb exposure on the fetus’ Ca metabolism and its skeletal ossification.


The overall confidence/uncertainty of the weight of evidence assessment would be increased/decreased upon verification of i) the maternal toxicity of Sb substances; ii) the influence of Sb in the Ca metabolism and ossification delays, and the reversibility of this effect; and iii) the developmental effects of Sb observed independent of any sign of maternal toxicity.


To address these knowledge gaps, the following research options appear relevant:

1)    Array of 2-weekdose-range finding/tolerance studiesin rats in a comparative analysis that permits selection of the most representative test item/source substances for further oral testing (based on the general reaction of the test animals to the treatments/substances, their specific gastric tolerance/irritation, and the systemic uptake/exposure levels obtained). In light of Sb’s reported emetic properties, this step is important before any in vivo oral study is considered. Sb have known emetic properties, but rats cannot vomit, so the assessment has to be done on the basis of their general reaction and specific gastric tolerance/irritation observations. This will also involve selection of the most appropriate (non-standard?) vehicles, demonstrated to cause no local irritancy either;

2)    OECD 422 in vivo Combined Repeat Dose Screening Reproductive studieson one or two Sb substances to screen for toxicokinetics and relevant systemic and reproductive toxicity parameters, including maternal effects of Sb exposures, and Ca metabolism to investigate the cause of the ossification delays and their (ir)reversibility.

Maternal toxicity should be determined on the basis of weight gain adjusted for gravid uterus weight and examination of the frequency of periods of zero/low food intake. The study should include weighing and detailed histopathology of the reproductive organs and the thyroid.

The effects on the fetus ossification should be assessed by using double staining (i.e. cartilage and bone tissues). Visceral and skeletal abnormalities should be examined in 50:50 of fetuses (not the standard 33:67). The (ir)reversibility of the eventual delayed ossification should be explored by examining some post-delivery pups.

In any case, the study should be designed to explore possible endocrine disrupting effects too, and enable derivation of both maternal and fetal NOAEL. For this, individual data should be recorded, not only means or medians, and results should be compared with historical control data, not only data from the concurrent control group.

3)    OECD 414 in vivo Prenatal Developmental Toxicity studieson one or more Sb substances (including the most representative test item/source substance) if necessary. Study design will have to maximize the information that can be collected from the observations, along the lines of specifications provided for the OECD 422 studies above;

4)    Additional higher tier oral studyto clarify any pending reproductive toxicity properties, if necessary.


In the meantime, there are three possible values to be considered for the chemical safety assessment:

·       NOAEL oral for Antimony trioxide: 1686 mg/kg bw/d) (subchronic; rat)

·       NOAEL oral for Sodium hexahydroxoantimonate: 1000 mg/kg bw/d) (subchronic; rat)

·       NOEC inhalation for Antimony trioxide: 6.3 mg/m³ (chronic; rat)

Among these values, the NOAEL oral for Sb trioxide will be selected for use in the chemical assessment. The NOAEL for Sodium hexahydroxoantimonate appears to be conservative but this is the result of testing artifacts (lower doses of sodium hewahydroxoantimonate were administered in the study) and is not truly representative of a worst-case situation. Furthermore, starting from an oral NOAEL will enable a more robust calculation and extrapolation of DNELs.


Justification for classification or non-classification

Cf. Scientific opinion on reproductive toxicity in Section 13 for complete weight of evidence and read-across assessments.


Based on the available evidence and the assessment of it made above, the available data (none of which is an OECD 421 study) support the conclusion that Sb substances are not considered to impact fertility; and provisionally supports that Sb substances do not have a direct adverse effect on the development of the conceptus. In light of this, no reproductive toxicity classification is warranted.


The level of confidence / uncertainty of the reproductive toxicity assessment of Sb substances could however be improved / decreased by conducting the following research:

o         Array of 2-week dose-range finding/tolerance studies;

o         OECD 422 in vivo Combined Repeat Dose Screening Reproductive studies;

o         OECD 414 in vivo Prenatal Developmental Toxicity studies, if necessary;

o         Additional higher tier oral study, if necessary.


The research strategy developed by the International Antimony Association, which supports REACH registrants with their Registration and Evaluation obligations, already foresees the above research options.