| Front page | | Contents | | Previous | | Next |

Siloxanes - Consumption, Toxicity and Alternatives

3 Health evaluation of siloxanes

• 3.1 Data on toxicity of siloxanes
• 3.2 Toxicity of siloxanes
  • 3.2.1 Toxicokinetics
  • 3.2.2 Acute toxicity
  • 3.2.3 Irritation and sensitization
  • 3.2.4 Subacute / subchronic / chronic toxicity
  • 3.2.5 Genetic toxicity
  • 3.2.6 Carcinogenicity
  • 3.2.7 Reproductive toxicity
  • 3.2.8 Endocrine disruption
• 3.3 Conclusion

3.1 Data on toxicity of siloxanes

Information about the toxicity of the siloxanes has been searched in open databases on the Internet and also as a general search based on CAS number, chemical name or just the term “siloxanes”, e.g. in combination with individual terms relevant to toxicity testing and results, using different search engines and meta search engines. Contacts to a few siloxane research university environments have pointed to the same literature as identified from searching the Internet.

As a first step in the data search, a preliminary database screening was carried out for decamethyl cyclopentasiloxane (Annex 6).

The data search has included the following general databases with information on chemical substances and their toxicological effects:

• RTECS;
• TOXNET: TOXLINE, CCRIS, HSDB, IRIS, GENE-TOX, DART / ETIC;
• MEDLINE;
• ScienceDirect (Journals);
• NTP;
• ASTDR.

The screening did not reveal any data on human toxicity, and it was decided not to make similar screenings for other siloxanes, but instead make a short review based on the available original literature.

The main source of information has been the Siloxane Research Program. The program is run by The Silicones Environmental, Health and Safety Council of North America (SEHSC) which is a non-profit trade association comprised of North American silicone chemical producers and importers. The programme was started in 1993 and includes a series of studies examining acute and long-term safety of exposure to the fundamental building blocks of many silicone materials (Meeks 1999). Testing under this programme includes the following type of tests:

Fundamental research:

• Pharmacokinetics;
• Biochemical toxicology.

Descriptive toxicological studies:

• Subacute studies of up to one month of duration;
• Subchronic studies of up to three months of duration;
• Two-year chronic studies to assess carcinogenicity and chronic effects;
• Developmental studies to assess effects on foetal development;
• Two-generation reproductive and fertility studies;
• Immunotoxicity studies.

Human clinical studies:

• Determination of human response and assessment of relevance of animal studies.

Exposure assessment studies:

• Workforce;
• Consumers;
• General public.

Information was specifically searched for the CAS numbers shown in the table below.

Table 3.1 Siloxanes searched by CAS number

Information has primarily been identified for D4, D5 and HMDS which besides polydimethylsiloxane (PDMS) is part of the Siloxane Research Program. For the other three siloxanes which are not part of the program, little or no information has been found. This is also the situation as regards more general information on toxicity related to small linear siloxanes and cyclic siloxanes.

A few studies are focussing on estrogenic and anti-estrogenic properties of D4 and HMDS, but in general data on endocrine disruption end points are scarce.

3.2 Toxicity of siloxanes

Although siloxanes are used in many products including consumer products and have been so for many years, there is relatively little information available about their toxicity apart from the information provided by the Siloxane Research Program. However, siloxanes have generally been regarded as safe in consumer products, but new uses, e.g. in breast implants and focus on reproductive toxicity and possible endocrine disrupting effects have focussed attention on this group of substances.

Of the six siloxanes mentioned in Table 3.1 only D4 is on Annex I to the Substance Directive (67/548/EEC) with a health classification as toxic to reproduction in category 3. The German justification for classification of D4 with regard to carcinogenic, mutagenic and reprotoxic effects is included in the reference list.

D4 is on the list of potential PBT and vPvB (very persistent and very bioaccumulative) substances selected on the basis of screening criteria in the EU (DEPA 2003).

In the following a short review of the findings in literature about the substances are presented. An overview of the studies and their results are presented in Annex 7.

3.2.1 Toxicokinetics

A number of studies in rats using unlabelled or 14C labelled D4 show that the level of absorption following inhalation of this substance is low and independent of gender and dose. The substance is distributed to most tissues and the highest concentrations were found in fat and the lowest in the reproductive tissues. Parent D4 is eliminated via the lungs and metabolised D4 via urine and faeces. The elimination profile from tissues except from fat and lung resembles that from blood and follows a two-compartment model (EPA DCN 86970000024 1996).

The elimination half-life for D4 has been shown to vary from 68 hours in plasma to approximately 150 hours in skin. Higher values are seen in testes. Blood clearance in human volunteers was non-linear and more rapid than by rats, whereas the elimination from the lungs resembled that from rats (BAuA 2001).

D4 has unusual distribution properties that have become apparent after examination of the time course data for blood and tissues using a quantitative physiological model (PBPK). Despite the very high lipophilicity, D4 does not show prolonged retention because of high pulmonary and hepatic clearance coupled with induction of metabolising enzymes at high exposure concentrations (Andersen et al. 2001). This avoids accumulation of free D4.

Pharmacokinetics of D4 administered to rats by inhalation and dermal route are similar and differs from the intravenous and oral route (Sarangapani et al. 2003).

Percutaneous absorption of neat D4 in humans following topical application between 1 and 24 hours has been shown at levels of 0.57 – 1.09% (EPA DCN 86980000153 1998 and EPA DCN 8601000003 2000).

In in vitro studies with percutaneous absorption following 24 hours exposure to 14C-D5 the absorption was found to be 0.8 – 1.08% (EPA DCN 86960000593 1996 and EPA DCN 86970000009 1996).

In rats administered HMDS orally and intravenously no parent HMDS was found in the urine. Metabolites from this linear siloxane appear to be structurally different from those obtained for cyclic siloxane except for the commonly present Me(2)Si(OH)(2).

3.2.2 Acute toxicity

In general the acute toxicity of siloxanes is considered low. LD₅₀ following oral administration of D4 in rats is reported to be more than 4,500 mg/kg and more that 5,000 mg/kg for HMDS. LC₅₀ in rats exposed to D4 was >12.17 mg/l and >48 mg/l when exposed to HMDS (European Commission 2000). LC₅₀ in rats exposed to HMDS for four hours was 15,956 ppm (EPA DCN 86970000724 1997).

LD₅₀ following dermal application of D4 was >2400 mg/kg bw. Several studies with dermal application of HMDS have shown greater LD₅₀ values, but mortality was observed at 10000 mg/kg bw. Toxic effects at 10000 mg/kg included gross pathological findings (lung, kidney, bladder, heart), while clinical findings (altered activity, ataxia, gasping and eschar formation) occurred in small numbers of rabbits. In contrast to rabbits, rats did not produce mortality or signs of toxicity at the dose tested (European Commission 2000).

3.2.3 Irritation and sensitization

D4 and HMDS have been tested on rabbit eyes without signs of irritation. The substances are also found non-irritant on rabbit skin. For both substances one study exists that describes the substances as slightly irritating. There are no further details from these studies (European Commission 2000).

D4 was not sensitizing in guinea pig maximisation test and also not sensitizing in 50 human subjects exposed to repeated insult patch test (European Commission 2000). HMDS has also been tested in guinea pig maximisation test without positive result (European Commission 2000).

3.2.4 Subacute / subchronic / chronic toxicity

D4 administered by oral gavage to rats over 28 days did not cause any immune suppression at doses between 10 and 300 mg/kg (EPA DCN 86980000072 1997). Human volunteers did not show any immunotoxic or proinflammatory/adjuvant effects following ingestion of 12 mg D4 in corn oil for 7 or 14 days (EPA DCN 86990000015 1998). The same result was obtained after inhalation of 10 ppm for one hour and re-exposure after three months (Loony et al. 1998).

Investigation of subacute oral toxicity in rats administered between 25 and 1600 mg/kg per gavage over two week with five applications per week caused increased relative liver weights in female animals at 100 mg/kg and male animals at 400 mg/kg. Absolute liver weight was also increased in female rats at 400 mg/kg. Decreased body weight was seen at the highest concentration in both male and female animals (BAuA 2001).

Rats exposed to D4 at 70 and 700 ppm by inhalation for 28 days, 5 days per week and 6 hours per day in different studies show rapid but reversible increase in liver size, induction of several metabolising enzymes, primarily CYP2B1 and induction of hepatic cytochrome P450 enzymes. D4 appears to be a phenobarbital-like inducer of hepatic microsomal enzymes in Fisher-344 rats (EPA DCN 86970000723 1996; EPA DCN 86970000725 1997; McKim et al. 1998).

Other studies with D4 in Fisher-344 rats exposed by inhalation over 3 months, 5 days per week and 6 hours per day, showed slight reduction in body weight and food intake at the 10.87 mg/kg dose group, slight dose-related increase in absolute and relative lever weight in female rats at 10.87 mg/kg and slight reduction in thymus and ovarian weight in female rats in the two highest dose groups, 5.91 and 10.87 mg/kg. Ovarian atrophy and vaginal mucification was also seen at the highest dose group (EPA DCN 8690000155 1995; EPA DCN 8690000153 1995).

Rats exposed to D5 by inhalation for 28 days, 7 days per week and 6 hours per day at concentrations between 10 and 160 ppm showed no adverse effects on body weight, food consumption or urinalysis. Minor transient changes in haematological serum chemistry and organ weight and a transient increase in liver to body weight and thymus to body weight at 160 ppm. NOEL (histopathological changes) was determined at 10 ppm, NOEL (systemic toxicity) was determined at 75 ppm and NOEL (immunosuppression) was determined at 160 ppm (EPA DCN 86970000385 1996).

Subchronic toxicity studies in rats exposed over three months at doses up to 224 ppm show that the lung is the primary target organ following D5 inhalation (Burns et al. 1998).

Rats exposed to inhalation of HMDS for one month in concentrations between 0.9 and 59.2 mg/l showed moderate increase in focal inflammatory lesions in the lungs in the highest dose group, increase in incidence and severity of renal tubule regeneration in male rats exposed to 12.7 and 59.2 mg/kg, hyaline droplet accumulation, protein casts and granular casts were present in kidneys in several males in the highest 59.2 mg/kg dose group. Other signs of toxicity included minimal hepatocellular hypertrophy in males of the two highest dose groups and a slight increase in pigment accumulation in bile ducts in the high dose group males (EPA DCN 869000048 1997).

Exposure of rats for three months to HMDS in concentrations between 0.33 and 33.3 mg/kg showed also similar histological lesions in the kidneys of males in the three highest dose groups; 4.0, 10.0 and 33.3 mg/kg. NOEL was determined at 1.3 mg/kg for male rats and 33.0 mg/kg for female rats (EPA DCN 86980000182 1998; Cassidy 2001).

Multifocal, subpleural, subacute to subchronic interstitial inflammation were seen in lungs of all groups of rats exposed for three months to inhalation of concentrations between 0.9 and 13.64 mg/kg HMDS. After the recovery period an increase of these finding were still seen in the high dose group (EPA DCN 86980000048 1997).

3.2.5 Genetic toxicity

Both D4 and HMDS have been tested in a number of in vitro studies including Ames test in different strains (with and without metabolic activation), DNA damage and repair test in E. Coli, cytogentic assay in Chinese Hamster Ovary cells, chromosome aberration assay and sister chromatide exchange assay in mouse lymphoma cells, all with and without activation with negative result. In vivo tests have included cytogenetic assay in rat (HMDS) and dominant lethal assay in rat (D4). Both tests were negative (European Commission, 2000). An in vivo chromosome aberration test in rats exposed to 700 ppm D4 was also negative (Vergnes et al. 2000).

The results gave no indication of a genotoxic potential - neither for D4 nor HMDS.

3.2.6 Carcinogenicity

Very little information is available on carcinogenicity of siloxanes. The only information identified is a report from Dow Corning received by EPA with preliminary results from a two-year chronic toxicity and carcinogenicity study in rats exposed to vapour concentrations of 0, 10, 40 or 160 ppm of D5 for 6 hours per day, 5 days per week, for 24 months. The preliminary results show that female rats in the highest dose group had a statistically significant increase of uterine tumours. These findings may indicate that there is a potential carcinogenic hazard associated with D5 (EPA. 2003). Final results are expected in the Spring of 2004.

Other relevant information is related to silicone in breast implants, where IARCH has evaluated that there is evidence suggesting lack of carcinogenicity in humans of breast implants, made of silicone, for female breast carcinoma (IARC 1999).

3.2.7 Reproductive toxicity

Tests to examine reproductive toxicity for siloxanes are primarily available to D4 which is also classified in the EU as Toxic for Reproduction. Cat. 3; R62 (Possible risk of impaired fertility). D4 has been evaluated based on information on toxic effects on the parent animals, toxicity to fertility and developmental toxicity/teratogenicity.

Various effects have been observed in both male and female parent animals in both studies to examine reproductive toxicity and other relevant studies. These effects have included decreased body weight, haematological and clinical-chemical effects, increased liver weight, cellular and subcellular changes, induction of liver enzymes, reduced organ weights (adrenals and thymus), ovarian atrophy and mucification of the vagina. The last two effects are indications of changes in the oestrus cycle. However, the results from the different tests have not shown a consistent picture with regard to the listed effects (BAuA, 2001).

Data on impairment of fertility are related to a reduction of the number of mean corpora lutea and implantation sites and increased post-implantation losses. The effects are phase specific indicating that D4 influences the hormonal cycle. It is suggested that this effect on the ovary is indirect and that the mode of action is related to neuroendocrine mechanisms in rodents which are quite different than those in humans. The rodent estrous cycle is generally four days long but in humans it is much longer (i.e. 28 days). It is this brevity of the estrous cycle that requires the precise control of the neuroendocrine events, the sensory inputs, and the sexual behaviours (Centre Europeen des Silicones 1999).

Studies to investigate developmental toxicity of D4 have not shown significant teratogenic effects (BAuA 2001).

A single generation study in rats exposed to inhalation is available for D5 showing no significant toxicological findings and no effects on reproductive parameters (EPA DCN 86970000006 1996).

3.2.8 Endocrine disruption

Very little information is available on endpoints which are considered relevant to endocrine disruption.

In a uterotrophic assay in immature rats receiving oral doses of D4 and HMDS for 4 days, D4 exhibited weakly estrogenic effects (dose-related increase in uterine weight and epithelial cell height) in both SP and F-344 rats. The substance also showed weak antiestrogenic properties by partially blocking EE (ethinylestradiol) induced uterine weight increases (competitive inhibition of estrogen receptor binding or D4 acting as a partial estrogen agonist). Estrogenic and antiestrogenic effects of D4 were several orders of magnitude less potent than EE, and many times less potent than the weak phytoestrogen CE (EPA DCN 86990000059 1999).

In the same assay HMDS showed no measurable effect on uterine weight when tested as an agonist. When co-administered together with EE, HMDS produced a slight, but statistically significant reduction in absolute uterine weight. The biological relevance of this could not be assessed in the present study (EPA DCN 86990000059 1999).

3.3 Conclusion

Only few siloxanes are described in the literature with regard to health effects, and it is therefore not possible to make broad conclusions and comparisons of the toxicity related to short chained linear and cyclic siloxanes based on this evaluation. Data is primarily found on the cyclic siloxanes D4 and D5 and the small linear HMDS.

The three siloxanes have a relatively low order of acute toxicity by oral, dermal and inhalatory routes and do not require classification for this effect.

They are not shown to be irritating to skin or eyes and are also not found sensitizing by skin contact. Data on respiratory sensitization have not been identified.

Subacute and subchronic toxicity studies show that the liver is the main target organ for D4 which also induces hepatocellular enzymes. This enzyme induction contributes to the elimination of the substance from the tissues. Primary target organ for D5 exposure by inhalation is the lung. D5 has a similar enzyme induction profile as D4. Subacute and subchronic inhalation of HMDS affects in particular the lungs and kidneys in rats.

None of the investigated siloxanes show any signs of genotoxic effects in vitro or in vivo. Preliminary results indicate that D5 has a potential carcinogenic effect.

D4 is considered to impair fertility in rats by inhalation and is classified as a substance toxic to reproduction in category 3 with the risk phrase R62 ('Possible risk of impaired fertility').

The results of a study to screen for estrogen activity indicate that D4 has very weak estrogenic and antiestrogenic activity and is a partial agonist (enhances the effect of the estrogen). It is not uncommon for compounds that are weakly estrogenic to also have antiestrogenic properties. Comparison of the estrogenic potency of D4 relative to ethinylestradiol (steroid hormone) indicates that D4 is 585,000 times less potent than ethinylestradiol in the rat stain Sprague-Dawley and 3.7 million times less potent than ethinylestradiol in the Fisher-344 rat strain. Because of lack of effects on other endpoints designated to assess estrogenicity, the estrogenicity as mode of action for the D4 reproductive effects has been questioned. An indirect mode of action causing a delay of the LH (luteinising hormone) surge necessary for optimal timing of ovulation has been suggested as the mechanism.

Based on the reviewed information, the critical effects of the siloxanes are impaired fertility (D4) and potential carcinogenic effects (uterine tumours in females) (D5). Furthermore there seem to be some effects on various organs following repeated exposures, the liver (D4), kidney (HMDS) and lung (D5 and HMDS) being the target organs.

A possible estrogenic effect contributing to the reproductive toxicity of D4 is discussed. There seems however to be some indication that this toxicity may be caused by another mechanism than estrogen activity.

Effects which based on the reviewed literature do not seem to be problematic are acute toxicity, irritant effects, sensitization and genotoxicity.

| Front page | | Contents | | Previous | | Next | | Top |

Version 1.0 September 2005, © Danish Environmental Protection Agency