Registration Dossier

Diss Factsheets

Administrative data

Key value for chemical safety assessment

Additional information

Summary of in vitro studies

The overall picture presented by thein vitro assay results available is that at least in some test systems (including Ames tests and in vitro chromosome aberration and sister chromatid exchange studies in mammalian cells) (Watabe et al., 1978a; De Meester et al., 1981; Jantunen et al., 1986; Norppa et al., 1983b; Chakrabarti et al., 1993), styrene does posses some genotoxic potential in vitro. Metabolic activation (presumably to styrene oxide) is required for this activity (Watabe et al., 1978a; Brunnemann et al., 1992; Jantunen et al., 1986; Norppa et al., 1983b; Chakrabarti et al., 1993).

 

A literature search from 1998 to January 2010 overlapping the date of the UK RAR (June 2008) was conducted. Overall, this did not lead to a modification of the conclusions reached by the UK RAR and reinforced some of the aspects elaborated above.

Fontaine et al. (2004) reported positive results by the comet assay in freshly isolated murine hepatocytes (and in human lymphocytes) at non-cytotoxic concentrations of styrene. Co-incubation with the CYP inhibitor SKF-525A strongly attenuated the effect in the hepatocyte cultures.

The effects of styrene on cell viability, cell proliferation, DNA damage and its repair and on colony formation in soft-agar (as an indication for carcinogenicity) were investigated in mesothelial and epithelial alveolar cell in vitro (Strafella et al., 2009). Styrene inhibited cell proliferation and induced low cytotoxicity in mesothelial cells, but induced cell proliferation in alveolar cells. DNA damage was not induced in both cell types but DNA repair was inhibited in mesothelial cells. No effect on colony formation was reported.

Several in vitro studies investigated the possible impact of polymorphism in humans on gentoxicity. Bernardini et al. (2002) studied the induction of sister chromatid exchanges (SCE) by styrene in human lymphocyte cultures from donors with polymorphism of glutathione S-transferase (GSTM1 and GSTT1). High concentrations of styrene (1.5 mM) induced SCE in cultures of all donors and SCE induction was significantly higher in subjects lacking both GSTM1 and GSTT1. If only one of these enzymes was missing, the SCE induction was intermediate between those subjects having either both or non of these genes. The authors suggest that the concurrent lack of GSTM1 and GSTT1 may increase the genotoxic effect of styrene on human cells. The potential importance of the detoxification pathways for styrene oxide (epoxide hydrolase and glutathione S-transferase) in respect to the genotoxicity of styrene was investigated by Laffon et al. (2003a). Human lymphocytes from donors with different activities of these detoxifying enzymes were incubated with styrene

oxide and the induction of micronuclei and DNA damage (alkaline comet assay) were measured. The results indicated that polymorphism of epoxide hydrolase and GSTP1 influences the induction of cytogenetic and DNA damage by styrene oxide. High activities of these enzymes led to relatively low genotoxic effects of styrene oxide. Laffon et al. (2003) incubated styrene itself with mononuclear leukocytes of 30 human donors. An influence of polymorphism on genotoxicity as measured by the alkaline comet assay was suggested for the activating enzymes CYP1A1 and CYP2E1 and for the detoxifying GSTP1. In conclusion, these in vitro data suggest that polymorphism of toxifying or detoxifying enzymes may modulate the genotoxicity of styrene.

 

In the literature update until Oct. 01, 2015, a study on DNA methylation was identified. Although this is not related to genotoxicity, it is mentioned here as an indication for a non-genotoxic mode of action. Tabish et al. (2012) measured global DNA methylation in vitro in human lymphoblastoid TK6 cells by liquid chromatography-mass spectrometry of 5-methyl-2-deoxycytidine. In the presence of S9-mix styrene induced global DNA hypomethylation as a potential indication for a non-genotoxic effect. But when examining methylations at low, mid, and high exposure concentrations the effects are less clear.

Summary of in vivo studies in experimental animals

Styrene has been exhaustively studied in clastogenicity studies up to dose levels producing severe toxicity in some cases. There is no convincing evidence of styrene clastogenicity in experimental animals when the quality of the studies and the plausibility of the test results are considered. Equivocal results were obtained after exposure to high doses causing lethality (Simula and Priestly, 1992; Norppa, 1981). However, overall, negative results were obtained fromin vivochromosome aberration and micronucleus studies in the rat (Sinha et al., 1983; Simula and Priestly, 1992; Kligerman et al., 1993; Preston and Abernethy, 1993), hamster (Norppa et al., 1980a) and the mouse (Loprieno et al., 1978; Sbrana et al., 1983; Sharief et al., 1986; Kligerman et al., 1992; 1993; Engelhardt et al., 2003) following single or repeated exposures to styrene up to concentrations and/or doses causing systemic toxicity, via the inhalation (Norppa et al., 1980a; Sinha et al., 1983 Kligerman et al., 1992; 1993; Preston and Abernethy, 1993; Engelhardt et al., 2003), oral (Loprieno et al., 1978; Sbrana et al., 1983) and intraperitoneal route (Sharief et al., 1986; Simula and Priestly, 1992) in the tissues examined (bone marrow, peripheral lymphocytes, splenocytes and whole blood). Furthermore, a recently published micronucleus test in bone marrow cells of mice conforming to the current OECD guideline was clearly negative (Engelhardt et al., 2003).

 

The general pattern of SCE results in wide range of tissues examined (lymphocytes, spleenocytes, bone marrow, alveolar macrophages, regenerating liver cells) from both the rat (Simula and Priestly, 1992; Kligerman et al., 1993; Preston and Abernethy, 1993) and the mouse (Conner et al., 1979; 1980a; Sharief et al., 1986; Simula and Priestly, 1992; Kligerman et al., 1992; 1993) following inhalation (Conner et al., 1979; 1980a; Preston and Abernethy, 1993; Kligerman et al., 1992; 1993)or i.p exposure (Sharief et al., 1986; Simula and Priestly, 1992) to styrene has been positive (Conner et al., 1979; 1980a; Simula and Priestly, 1992; Kligerman et al., 1992; 1993). However, it is important to note that in most cases concomitant chromosome aberration and/or micronucleus assays involving the same animals and in some cases the same tissues were carried out and that negative results were obtained for these indicators of chromosome damage. Therefore, this clearly reduces the significance of the SCE findings in relation to mutagenicity.

The binding of styrene metabolites to DNA was very low and did not indicate any specificity for the target tissue (mouse lung) (Cantoreggi and Lutz, 1993; Boogard et al., 2000; Vodicka et al., 2001; Otteneder et al., 2002). Induction of alkali-labile single-strand breaks has also been producedin vivoin mice exposed to styrene by intraperitoneal application (Solveig-Walles and Orsen, 1983; Vaghef et al., 1998) while after inhalation exposure equivocal results were obtained. When rats were exposed to styrene by inhalation DNA stran breaks were not observed in the comet assay (Kligerman et al., 1993). Again the significance of these findings is unclear, given the repeated failure of styrene to demonstrate mutagenic activity in standard clastogenicity assays.

In contrast to the weakly positive findings in indicator tests detecting SCEs, DNA strand breaks and DNA adducts, an in vivo UDS test performed in accordance with international guidelines did not reveal a genotoxic effect of styrene in mouse liver (Clay, 2004).

Overall, based on standard regulatory tests, there is no convincing evidence that styrene possesses significant mutagenic/clastogenic potential in vivo from the available data in experimental animals.

 

A literature search from 1998 to January 2010 overlapping the date of the UK RAR (June 2008) was conducted. Overall, this did not lead to a modification of the conclusions reached by the UK RAR and reinforced some of the aspects elaborated above.

In a comprehensive review of the genotoxicity tests in animals Speit and Henderson (2005) concluded that styrene is weakly positive in indicator tests (DNA adducts, DNA strand breaks, and SCE), but there is no convincing evidence for clastogenicity. A similar conclusion was reached by Nestmann et al. (2005) when assessing genotoxicity/mutagenicity of styrene from in vitro tests, animal tests, and results obtained in exposed workers. The evaluations by both of these reviews correspond to that of the UK Risk Assessment Report.

The formation of guanine DNA adducts described by Vodicka et al. (2001) in mice was reconfirmed by Vodicka et al. (2006) after inhalation of styrene (750 and 1500 mg/m³; 6 h/d; continuously up to 3 weeks). The guanine DNA adducts increased linearly with the exposure duration in the lung, but no adducts were measured in the liver at the end of the 3 week exposure period. In addition, guanine DNA adducts derived from styrene oxide were found in the urine of mice exposed to styrene. Urinary adenine DNA adducts were identified by Mikes et al. (2009) (in addition to guanine adducts) after exposure of mice to 600 and 1200 mg/m³ over 10 consecutive days (6h/d).

A new literature search up to March 2013 was carried out. The following paper was identified; the new data basically supported non-genotoxicity/mutagenicity of styrene and SO in rats after inhalation.

Gate et al. (2012) exposed rats to styrene (75, 300, 1000 ppm) orstyrene-7,8-oxide(SO) by inhalation (25, 50, 75 ppm) 6 h/d, 5 d/week over 4 weeks.Micronuclei in circulating reticulocytes using the very sensitive flow cytometry procedure and DNA strand breaks in leukocytes by the Comet assay were studied at the end of the 3rdand 20thdays of exposure. Blood was collected from the carotid artery to measure SO directly coming from the lung avoiding first-pass hepatic metabolism. After SO exposure the SO blood concentrations increased nearly linearly and 1000 ppm styrene led to blood SO concentrations between those after 25 and 50 ppm SO. Reticulocytes in blood were significantly decreased after exposure to styrene and SO and this effect was more pronounced after 3 days of exposure that after 20 days. Thus the two compounds or their metabolites had reached the bone marrow. Neither styrene nor SO induced a significant increase of micronuclei frequency in reticulocytes or DNA strand breaks in white blood cells. However, in the presence of formamidopyridine DNA glycosylase (Fpg), an enzyme able to recognize and excise oxidized DNA bases, a significant increase of DNA damage was observed at the end of the 3rdday (but not after the 20thday) of treatment with styrene but not with SO. But no dose response relationship was observed at styrene exposures between 75 to 1000 ppm. According to the authors the results obtained with the Fpg-modified Comet assay may suggest oxidative stress in white blood cells after styrene exposure. In summary, inhalation exposure to styrene and SO over up to 20 days did not lead to an increase in micronuclei or DNA strand breaks in blood reticulocytes of rats. At the dose levels used the bone marrow was reached. Positive results in the Comet assay using Fpg are difficult to interpret: they were obtained only with styrene on day 3 (but not on day 20 or with SO) and did not show a dose response relationship.

 

Summary of human studies

A large number of studies have been published which have aimed to investigate the genotoxic potential of styrene in humans by examination of various endpoints in styrene exposed workers. Very low levels of DNA adducts were found in some styrene exposed workers but it has been stated that such low levels should be viewed with caution. There is also some evidence of DNA damage (SSBs) induced in styrene exposed workers. Both these endpoints are indicative of exposure but are not necessarily associated with heritable effects. The results of several studies on another indicator endpoint of unclear health significance, SCEs, did not provide evidence of a positive response, despite these being induced in animals exposed to styrene. There are also many studies investigating endpoints (gene mutations, chromosome aberrations and micronuclei) known to lead to heritable effects. The number of studies assessing gene mutation is very limited and no conclusions can be drawn from them (Bigbee et al., 1986; Tates et al., 1994; Vodicka et al., 2003). Although 5 studies appear to present evidence that styrene may be weakly clastogenic in humans (Brenneret al., 1991; Tomanin et al., 1992; Anwar and Shamy, 1995; Somorovska et al., 1999; Laffon et al., 2002), there are 12 robust negative studies also (Watanabe et al., 1983; Hansteen et al., 1984; Pohlava and Sram, 1985; Mäki-Paakkanen , 1987; Jablonicka et al., 1988; Hagmar et al., 1989; Mäki-Paakkanen et al., 1991; Norppa et al., 1991and Sorsa et al., 1991a [both referring to the same study population]; Yager et al., 1993; Van Hummelen et al., 1994; Karakaya et al., 1997; Teixeira et al., 2004). Together with a lack of evidence of a dose-response relationship and the negative response for induction of micronuclei when studied concurrently in two of the positive chromosome aberration studies, no clear conclusion on in vivo clastogenicity of styrene in humans can be made.

 

A more detailed evaluation of the results for clastogenicity in humans is given in the UK Risk Assessment Report at the end of the section “chromosome aberration and micronuclei”:

In summary, 30 independent publications reporting analysis of clastogenicity in styrene exposed workers are available. These include 23 analyses of chromosome aberrations and 16 of micronuclei. Many of these studies are considered to be inconclusive due either to insufficient detail, possibility of confounding exposures or technical deficiencies such as inadequate harvest time, insufficient cells or individuals. However, of the more robust studies, 8 chromosome aberrations studies are clearly negative (Watanabe et al., 1983; Hansteen et al., 1984; Pohlava and Sram, 1985; Mäki-Paakkanen , 1987; Jablonicka et al., 1988; Hagmar et al., 1989; Mäki-Paakkanen et al., 1991; Norppa et al., 1991and Sorsa et al., 1991a [both referring to the same study population]) and only 3 (Tomanin et al., 1992; Anwar and Shamy, 1995; Somorovska et al., 1999) demonstrate weak positive responses. In none of the positive investigations was the response dose related (against personal monitoring data) and in two studies the concurrent assessment of micronuclei, using the sensitive cytocholasin B method, was negative (Tomanin et al., 1992; Anwar and Shamy, 1995). It is difficult to rationalise why positive responses were found in 4 studies but not the others. These were apparently well-conducted, and the exposures were moderate (from 6 up to 104 ppm) but not excessively high compared to other negative studies (up to 139 ppm).

Ten micronuclei studies were clearly negative (Mäki-Paakkanen , 1987; Hagmar et al., 1989; Mäki-Paakkanen et al., 1991; Norppa et al., 1991and Sorsa et al., 1991a [both referring to the same study population]; Tomanin et al., 1992; Yager et al., 1993; Van Hummelen et al., 1994; Anwar and Shamy, 1995; Karakaya et al., 1997; Teixeira et al., 2004), 2 gave a positive response (Brenner et al., 1991; Laffon et al., 2002) whilst the remainder were inconclusive. In the study reporting a weak positive response (Brenneret al., 1991) the possibility of other confounding exposures in the workplace was not addressed. Seven studies using the more reliable cytocholasin B method were negative (Mäki-Paakkanen et al., 1991; Norppa et al., 1991and Sorsa et al, 1991a [both referring to the same study population]; Tomanin et al., 1992; Yager et al., 1993; Van Hummelen et al., 1994; Anwar and Shamy, 1995; Teixeira et al., 2004), including two where an increase in chromosome aberrations was reported (Tomanin et al., 1992; Anwar and Shamy, 1995). However, one well-conducted study using cytocholasin B indicated a positive response (Laffon et al., 2002).

Overall, given the extensive number of negative studies reported and the lack of dose-response relationships, the 3 chromosome aberration studies and the 2 micronuclei studies considered as providing evidence of a weak positive response, on balance, do not present conclusive evidence that styrene can cause chromosome breakage in humans.

 

A more detailed evaluation of the results for the induction of SCE in humans is given in the UK Risk Assessment Report at the end of the section “sister chromatid exchanges”:

In summary, SCEs are one of the major reported positive endpoints in animal studies and are also induced in in vitro assays. In humans, there are 11 clearly negative studies (Meretoja et al., 1978b; Hansteen et al., 1984; Watanabe et al., 1981; 1983; Mäki-Paakkanen , 1987; Kelsey et al., 1990; Mäki-Paakkanen et al., 1991; Norppa et al., 1991and Sorsa et al., 1991a [both referring to the same study population]; Brenner et al., 1991; Van Hummelen et al., 1994; Holz et al., 1995), 2 investigations (Laffon et al., 2002 and Teixeira et al., 2004) indicating a marginal effect and 9 which are inconclusive, mainly due to inadequate statistics, technical deficiencies (scoring anomalies), possible exposure to other confounding factors and inadequate cell or individual numbers. Overall, on balance, there is no convincing evidence that styrene increases the frequency of SCEs in humans. Furthermore, the lack of convincing data on the induction of SCEs in humans casts doubt that the positive results claimed for other endpoints are in fact due to styrene since the sensitivity of this endpoint to styrene exposure has been demonstrated in animal and in vitro systems.

 

Overall, given the lack of evidence of consistent relationships between exposure levels and study outcome, the lack of any consistent profile of endpoints and the absence of information on the relevance of the types of adducts seen and their mutagenic potential in vivo, there is no convincing evidence that styrene has shown mutagenic activity in humans.

A more detailed evaluation of the results for DNA single-strand breaks in humans is given in the UK Risk Assessment Report at the end of the section “DNA single-strand breaks”:

In summary, a number of different assay types have been used to detect DNA strand breaks, including the DNA unwinding assay and the comet assay. The former was used by Mäki-Paakkanen et al (1991), Brenner et al (1991) and Shamy et al (2002). A small increase in SSBs was reported in all studies. No statistically significant correlation was seen between the levels of damage and urinary metabolites or blood styrene levels in the first two studies, although a correlation was claimed in the third. Walles et al(1993) used the more sensitive alkaline elution assay to compare increases in SSBs at the end of shifts relative to pre-shift values. Although an increase was seen, this approach without the use of a concurrent control population is seriously flawed as diurnal changes (including smoking as demonstrated in this study) can affect this endpoint. Therefore, the increase in damage cannot be ascribed to styrene without an adequate control group. Other studies have used the comet assay to investigate DNA damage (Vodicka et al., 1995, 1999, 2002a, Somorovska et al., 1999, Laffon et al., 2002, Migliore et al., 2002, Buschini et al., 2003). It is not clear how independent the first 4 publications are since the paper of Vodicka et al (2002a) appears to be an overview of all of the other papers and there is clear overlap between study groups and data. Four of the studies appear to indicate that styrene induces DNA damage (Somorovska et al., 1999, Vodicka et al., 1999, Migliore et al., 2002, Buschini et al., 2003) although there are some deficiencies in two of them (internal data inconsistency in Somorovska et al., 1999 and lack of viability data in Vodicka et al., 1999). The positive response reported by Laffon et al (2002) appears to be due predominantly to smoking. There is no evidence of a relationship to exposure levels in any study. These assays are considered to be indicators of damage and various parameters which are not considered to be genotoxic such as exercise, diurnal sampling period and vitamin C intake have been previously shown to induce responses (e.g. Mölleret al., 2000). Thus, although there is some evidence for DNA damage in styrene exposed workers, the studies have not been consistent, the levels of damage are not related to exposure levels and the significance of any positive response is unclear.

 

A more detailed evaluation of the results for DNA adducts in humans is given in the UK Risk Assessment Report at the end of the section “DNA (and protein) adducts”:

In summary, there are several studies which have shown the presence of styrene-related DNA adducts in humans. The majority of these have been conducted in lymphocytes or mononuclear cells (Vodicka et al., 1994; 1995). Studies using granulocytes (Vodicka et al., 1994; 1995) or monocytes (Holz et al., 1995) have been very weakly positive or negative, presumably a reflection of the quick turnover of these cell types which makes them unable to indicate cumulative exposure. Some of the studies have failed to identify the adducts detected. However, to date, adducts at N2-guanine (Horvath et al., 1994), O6-deoxyguanosine (Vodicka et al., 1993; 1994; 1995; 1999), 1-adenine (Koskinen et al., 2001) have been identified. The major adduct seen in animal studies, N7-styrene oxide-guanine, has not been detected so far in human studies, presumably partly due to its short half-life. The adducts seen, which may be minor but stable adducts, are present at low levels and some comparisons have been made of their significantly lower levels in relation to those found in butadiene exposed workers (Koskinen et al., 2001). Marczynskiet al (1997) studied oxidative adducts in styrene exposed workers and proposed this as the mechanism of styrene genotoxicity. However, the possibility of confounding environmental exposures to oxidative mutagens has not been adequately characterised in their work. Overall, there is good evidence that styrene induces DNA adducts in humans but the significance of these in terms of subsequent mutagenicity is not proven.

 

A literature search from 1998 to January 2010 overlapping the date of the UK RAR (June 2008) was conducted. Overall, this did not lead to a modification of the conclusions reached by the UK RAR and reinforced some of the aspects elaborated above.

In a comprehensive review of the genotoxicity data in humans Henderson and Speit (2005) basically came to the same conclusion as the UK Risk Assessment Report. Using criteria derived from the IPCS guidelines they concluded for the different endpoints:

-        the data are not convincing for the induction of gene mutations

-        as regards clastogenicity, there is a predominant lack of induction of micronuclei but conflicting responses in chromosome aberration assays

-        numerous studies on sister chromatid exchanges do not provide evidence of a clear positive response, despite these being induced in animals

-        there is evidence that both DNA adducts and single strand breaks are induced in styrene workers. But these types of damage do not necessarily result in heritable changes

-        there is evidence that metabolism of styrene in humans is affected by genetic polymorphism and polymorphism affects the outcome of in vitro mutagenicity studies.

In summary, they reach the conclusion that there is no clear evidence that styrene exposure in workers results in detectable levels of mutagenic damage taking into consideration dose-response relationships and the lack of a common profile of positive responses for the various endpoints.

Vodicka et al. (2006a) listed the cytogenetic studies in humans reported since 1994 without giving more detailed interpretations or assessments of their validity. In total they found:

-        chromosomal aberrations, 7 studies: two clearly positive on sufficiently large cohorts, one positive on a population with a limited size; two positive but inconclusive due to the co-exposure to other genotoxicants, the remaining two clearly negative.

-        Micronuclei, 8 studies: one positive, two confounded by co-exposure to other genotoxicants, one inconclusive due to an unsuitable control group, four negative.

-        SCE, 9 studies: two clearly positive, four moderately positive (limited size of the investigated cohorts); one positive but confounded by a very small population and by lack of exposure data, two clearly negative.

 

- As regards mutation assays:

Kuricova et al. (2005) investigated the influence of polymorphisms of various DNA repair genes on styrene induced genotoxicity. After stratification for the different DNA repair genetic polymorphisms it was shown that the polymorphism in exon 23 of the XPD gene modulates the level of the HPRT mutant frequency.

 

- As regards cytogenetic assays (chromosome aberrations and micronuclei):

Teixeira et al. (2004) studied 28 reinforced plastic workers in comparison to 28 control persons. For exposure assessment urinary metabolites of styrene were determined and personal air sampling was carried out at the time of investigation. There was a broad range of exposure: 2-91 ppm. There was no significant difference for the incidence of micronuclei in exposed as compared to control workers.

Vodicka et al. (2004) investigated chromosomal aberrations and micronuclei in peripheral blood lymphocytes of 84 exposed workers (mean exposure 81.3 mg/m³) and 42 control persons. They further studied a possible correlation with different exposure parameters like styrene in air or blood or urinary metabolites (mandelic acid, phenylglyoxylic acid, 4-vinyl phenol conjugates, phenyl hydroxyethyl mercapturic acids). The incidence of chromosomal aberrations did not show a difference between exposed and control workers and for micronuclei a statistically significant increase was only observed in the 35 workers of the one factory with the highest exposure (mean 112.4 mg/m³). There was no correlation between the incidence of chromosomal aberrations and the different exposure metrices and only a moderate relationship for micronuclei.

In the blood of tire plant workers exposed mainly to styrene (up to 13.2 mg/m³) and butadiene (up to 5.8 mg/m³) no influence of polymorphism of CYP1A1, CYP2E1, and glutathione transferase was found on chromosomal aberrations or single strand breaks. On the other hand, individuals with a low epoxide hydrolase activity exhibited higher chromosomal aberration frequencies (Vodicka et al.,2004a).

Kuricova et al. (2005) investigated the influence of polymorphisms of various DNA repair genes on styrene induced genotoxicity. After stratification for the different DNA repair genetic polymorphisms it was shown that the polymorphism in exon 23 of the XPD gene modulates the level of chromosomal aberrations. De Palma et al. (2003) reported that clastogenicity as measured by the incidence of micronuclei in laminators of the glass-fibre reinforced plastics industry seems to be modulated by NQO1 polymorphism and aneuploidogenic effects (centromer positive micronuclei) by GSTM1 polymorphism (publication in Italian language).

A higher incidence of micronucleated lymphocytes and nasal epithelial cells was found by Godderis et al. (2004) in 44 workers exposed to styrene (exposure rang 0-36.6 ppm; duration of exposure 7-455 months) as compared to 44 matched referents. The airborne exposure was calculated from urinary mandelic acid. The incidence of micronucleated lymphocytes was influenced by polymorphism of the XRCC1 and XRCC3 repair genes.

In the new literature search up to March 2013 a study was identified that critically reviews the methodology of the micronucleus test with human nasal cells. This has to be taken into consideration when interpreting the results of Godderis et al. (2004):

Knasmueller et al. (2011) reviewed the micronucleus (MN) assay (and other genotoxicity tests) in human nasal cells. The study of Godderis et al. (2004) is briefly mentioned without any assessment. Regarding the MN assay with human nasal cells the authors conclude thatlymphocytes,buccal and nasalcellsareingeneralequally sensitive, butstudieswithnasalcellsarebyfar less validated andstandardizedthan theones withbloodand buccal cells.Theimpactage,gender,lifestyle,dietandsmokingisstillunclear. Otherimportantissuesarethesampling siteandthetiming ofcellcollectionafterexposure.Broadvariationsin thebaselinelevels,differencesofresultsobtained invarious studies aswellasthelackof informationon potentialconfoundingfactorsindicatetheneed forfurtherstandardization of the experimental protocols. In summary, this review emphasizes the need to further standardize the micronucleus test with nasal cells in humans.

Migliore et al. (2006a) investigated 95 exposed workers (mean exposure 37.1 mg/m³; range 2-535) in comparison to 98 controls. The percentage of smokers and the number of cigarettes smoked per day was higher in the exposure group. Chromosomal aberrations (chromosome type) and micronuclei (both centromeric positive and negative) were significantly increased in the exposed population. The possible influence of polymorphism of epoxide hydrolase, glutathione S-transferase (GSTT1, GSTM1, GSTP1), and N-acetyltransfeerase was also investigated. The incidence of micronuclei was associated with GSTT1 polymorphism in the exposed workers.

Migliore et al (2006) investigated the relationship between genotoxic biomarkers in somatic and germ cells (comet assay). 42 exposed workers were compared to 25 referents, but no information is available on matching. By analysis of urinary metabolites the airborne exposure was estimated to be above 20 ppm (TWA). Micronuclei in blood lymphocytes arising from chromosomal breakage and whole chromosomes (identified by centromeric fluorescence staining) were significantly increased in the exposed workers. The cytogenetic biomarkers in somatic and germ cells (comet assay, see below) were interrelated but only in the control and not in the exposed group when analyzed separately. No relationship was apparent with the exposure parameters.

 

- As regards sister chromatid exchanges:

Teixeira et al. (2004) studied 28 reinforced plastic workers in comparison to 28 control persons. For exposure assessment urinary metabolites of styrene were determined and personal air sampling was carried out at the time of investigation. There was a broad range of exposure: 2-91 ppm. The possible influence of polymorphism in the GSTM1, GSTT1, GSTP1, and CYP2E1 gene was analyzed in addition. There was a significant difference for SCEs in exposed as compared to control workers. While the SCE response in exposed workers was not influenced by genetic polymorphism, the baseline level in the control group was modulated by the GSTP1 and CYP2E1 genes.

 

- As regards DNA single-strand breaks:

Vodicka et al. (2004) investigated single-strand breaks (comet assay) in peripheral blood lymphocytes of 84 exposed workers (mean exposure 81.3 mg/m³) and 42 control persons. They further studied a possible correlation with different exposure parameters like styrene in air or blood or urinary metabolites (mandelic acid, phenylglyoxylic acid, 4-vinyl phenol conjugates, phenyl hydroxyethyl mercapturic acids). The incidence of single-strand breaks did not show a difference between all exposed and control workers and was even significantly decreased in the 35 workers of the one factory with the highest exposure (mean 112.4 mg/m³). There was an inverse correlation between single-strand breaks and the different exposure metrices.

Kuricova et al. (2005) investigated the influence of polymorphisms of various DNA repair genes on styrene induced genotoxicity. After stratification for the different DNA repair genetic polymorphisms it was shown that the polymorphism in exon 23 of the XPD gene modulates the level of DNA single-strand breaks.

In the comet assay no significant difference was found by Godderis et al. (2004) in 44 workers exposed to styrene (exposure rang 0-36.6 ppm; duration of exposure 7-455 months) as compared to 44 matched referents. The airborne exposure was calculated from urinary mandelic acid.

In a follow-up study to Migliore et al. (2002), Migliore et al (2006) investigated the relationship between biomarkers in somatic and germ cells. 42 exposed workers were compared to 25 referents, but no information is available on matching. By analysis of urinary metabolites the airborne exposure was estimated to be above 20 ppm (TWA). As exposure data and results of the comet assay were very similar in both publications it may be assumed that there was a substantial overlap in the study populations. Primary DNA damage in germ cells as investigated by the comet assay was significantly higher in the styrene exposed workers. The frequency of aneuploidy and diploidy in sperm was investigated by FISH analysis using specific DNA probes for the centromeric regions of sex chromosomes and chromosome 2. No difference was found between exposed and control workers. The cytogenetic biomarkers in somatic (micronulei in peripheral lymphocytes, see above) and the DNA damage in germ cells were sowed a significant positive correlation, but only in the control and not in the exposed group when both groups were analyzed separately. But no relationship was apparent for DNA damage with the exposure parameters. As regards the missing effect of styrene exposure on numerical alterations of chromosomes in spermatozoa this study confirms the data of Naccarati et al. (2003) in a smaller population of exposed workers (N= 18) and referents (N=13). For comparison purpose, there is only one study measuring single strand breaks in testes. Solveig-Walles and Orsen (1983) treated mice with a very high intraperitoneal dose of 880 mg/kg bw and observed a small increase of single-strand breaks by the DNA unwinding technique after 4 and 24 h. It s not clear whether this effect has any statistical or biological significance. Speit et al. (2008) critically reviewed the use of the comet assay as an indicator test for germ cell genotoxicity. They concluded that many different modifications of the comet assay are in use but a standard protocol has not yet been established. High and variable background levels of DNA effects were reported and there still is the need for standardization and validation. The authors conclude that it seems to be premature to use biomonitoring data with human sperm for hazard identification or classification. Thus, no clear conclusion can be drawn from the isolated findings of Migliore at al. (2002; 2006).

Fracasso et al. (2009) studied DNA single- and double-strand breaks by the alkaline and neutral version of the comet assay. 34 styrene exposed workers (mean exposure 47 ppm) were compared to 29 matched referents. Single- and double-strand breaks were significantly increased in the exposed workers.

 

- As regards DNA (and protein) adducts:

Kuricova et al. (2005) investigated the influence of polymorphisms of various DNA repair genes on styrene induced genotoxicity. After stratification for the different DNA repair genetic polymorphisms it was shown that the polymorphism in exon 23 of the XPD gene moderately affects the level of DNA adducts.

No significant difference in the level of N-terminal hemoglobin adducts was found by Godderis et al. (2004) in 44 workers exposed to styrene (exposure rang 0-36.6 ppm; duration of exposure 7-455 months) as compared to 44 matched referents. The airborne exposure was calculated from urinary mandelic acid.

Teixeira et al. (2007) studied the possible influence of genetic polymorphism (epoxide hydrolase, CYP2E1, GSTT1, GSTP1, GSTM1) on the formation of N-terminal valine hemoglobin adducts. 75 exposed and 77 control workers were enrolled. The level of exposure to styrene was 30.4 ppm (mean) with a range of 0.5-114 ppm. Hemoglobin adducts were significantly increased in the exposed population and a modulation was by the activity of epoxide hydrolase was observed.

Biomarkers of nucleic acid oxidation and the expression of the human 8-oxoguanine DNA N-glycosylase 1 repair gene were investigated by Manini et al. (2009) in exposed laminators (exposure level 107.4 +/-66.7 mg/m³) in comparison to unexposed control workers. Biomarkers of nucleic acid (guanine) oxidation were determined in urine and DNA of white blood cells. The urinary biomarkers of oxidative stress correlated with measures of exposure (mandelic acid, phenylglyoxylic acid, 4-vinylphenol in urine) but there was a negative correlation between exposure parameters and oxidative DNA products in blood. Styrene exposed workers had a significantly lower oxidative DNA damage in white blood cells compared to controls. This contrasts to findings of Marczynski et al. (1997) who reported increased 8-hydroxy-2-deoxyguanosine in white blood cells in a small study of exposed boat builders. Furthermore, in workers significantly higher levels of the DNA repair enzyme (base excision repair system) were observed. The authors concluded that styrene exposure seems to be associated with oxidative damage to nucleic acids, especially the DNA, and with an induction of the base excision repair system.

A new literature search up to March 2013 was carried out. The following papers were identified; the new data do not change the overall interpretation given in the UK Risk Assessment Report.

Rueff et al. (2009) reviewed genotoxicity data in humans and their application for biomonitoring for effects. The advantage of such biomarkers is that they integrate besides exposure, all known factors mediating individual variability in absorption, distribution, metabolism and excretion. Such studies should deal with the analysis of individual genotypes associated with the metabolic fate and DNA damage of styrene (metabolizing as well as DNA repair enzymes). As regards genotoxicity, the authors conclude that styrene exposure has been reported to cause an increase in DNA and hemoglobin adducts and in the frequency of chromosomal aberrations; there is less evidence for an association between styrene exposure and the frequency of SCE and no compelling evidence for micronuclei formation in human studies. The evidence for carcinogenicity is summarized as follows: “Evidence of the carcinogenicity of styrene has been reported in studies with mice. Epidemiologic evidence does not suggest a causal association styrene and any forms of cancer in humans.However,the possibility of a small elevation of risk for one or more cancers cannot be ruled out.” In summary, by this review it is concluded that styrene leads to DNA adducts and chromosomal aberrations, but there is less evidence for SCE and no compelling evidence for micronuclei formation. There is no causal association with cancer but a small elevation of risk cannot be ruled out.

Jin et al. (2009) studied the frequency of micronuclei in lymphocytes in 54 workers exposed to styrene from a chemicals and plastic plant of a petrochemical company and in 25 controls by the cytokinesis-blocked micronucleus test (CB-MNT). A high exposure group (N=24; 7.6-24 mg/m³) and a low exposure group (N=20; 4.1-7.5 mg/m³) were defined. The micronucleus rate of the high exposure group was statistically significantly higher than that of the control group. In addition the authors report an increasing rate along with an increase in the length of service. But this is certainly an over-interpretation as the mean rate increases from 4.3 %o (1-10 years of exposure) to 4.4 %o (10-20 years) and to 4.5 %o (>20 years). These findings have to be interpreted with caution as the study suffers from several shortcomings, e.g. no information on potential co-exposures to other substances or when and how often the exposure measurements were made, no indication whether the micronuclei determinations were made blinded. In summary, an increased frequency of micronuclei is reported by the cytokinesis-blocked micronucleus test (CB-MNT) in workers with relatively low styrene exposure. Due to insufficiencies in reporting his study must be interpreted with caution.(Publication in Chinese language, only English abstract)

Teixeira et al. (2010) determined micronuclei (by the cytokinesis-block micronucleus assay), SCE rates and DNA damage by the Comet assay in 52 fibreglass reinforced plastics workers in comparison to 54 controls. Only non-smokers (as verified by a questionnaire) were included. The mean air concentration of styrene was 29.9 ppm (measured by personal sampling) and the mean value of mandelic acid + phenylglyoxylic acid in urine was 419 mg/g creatinine. None of the workers used protective devices. For SCEs higher rates (p<0.01) were observed in the exposed workers as compared to controls and there was a significant correlation with the exposure parameters (for both p<0.01). Micronuclei rates and DNA damage in the Comet assay showed slight and non-significant increases related to exposure. In summary, workers exposed to a mean of 30 ppm showed significantly increased, exposure related SCE rates. For micronuclei and DNA damage in the Comet assay slight and non-significant increases were observed.

Hanova et al. (2010) investigated single strand breaks (SSB), micronuclei, and DNA repair capacity on 71 styrene exposed hand lamination workers and 51 control individuals. Styrene concentrations at workplace and in blood characterized occupational exposure. The workers were divided into low (below 50 mg/m3) and high (above 50 mg/m3) styrene exposure groups. Single strand breaks were determined by the comet assay including the endonuclease III modification specific for abasic sites and oxopyrimidines. DNA damage and DNA repair capacity (after γ-irradiation) were analyzed in peripheral blood lymphocytes by Comet assay. A significant negative correlation was observed between SSBs and styrene concentration at workplace, but there was no difference between the exposure groups with regard to endonuclease III sensitive single strand breaks. The capacity to repair irradiation-induced DNA did not show a dose response relationship. Micronuclei were not affected by styrene exposure. No associations were observed between γ-irradiation specific DNA repair rates and mRNA expression levels of three important DNA repair genes (XRCC1, hOGG1 and XPC). The mRNA expression levels of these repair genes were negatively correlated with styrene concentrations in blood and at workplace and positively with SSBs. In summary, as regards genotoxic effects in exposed workers there was no exposure related increase in single strand breaks or the incidence of micronuclei in peripheral blood lymphocytes. No clear dose response associations were noted forγ-irradiation specific DNA repair rates and mRNA expression levels of DNA repair genes.

Mikes et al. (2010) analysed urine samples from humans occupationally exposed to styrene, with mandelic acid levels ranging from 400 to 1145 mg/g creatinine and from 68 to 400 mg/g creatinine for high and low exposure group, respectively, for N3 adenine DNA adducts, namely 3-(2-hydroxy-1-phenylethyl)adenine (N3αA) and 3-(2-hydroxy-2-phenylethyl)adenine (N3ßA). Peaks corresponding to N3αA and/or N3ßA were found in 7/9 end-of-shift samples of the high exposure group and in 6/19 end-of-shift samples of the low exposure group and the concentration of N3αA+N3ßA amounted to 2.8±1.6 pg/mL and 1.8±1.3 pg/mL, respectively. As interfering peaks were detected also in some control urine samples the method was insufficiently specific for biological monitoring. Comparing the excretion of N3αA+N3ßA to that reported previously in mice it can be estimated that at the same absorbed dose, humans excreted not more than 1/30 of the amount of adenine adducts excreted by mice. Therefore, the damage to DNA caused by styrene 7,8-oxide was much lower in humans than in mice. In summary, the DNA adducts quantified in the urine of exposed workers were by a factor of 30 lower than those found in the urine of mice. This further underlines the difference in metabolism between humans and mice.

Wongvijitsuk(2011) studied 50 styrene exposed reinforced-fibreglass plastics workers in comparison to 40 unexposed controls. The exposed workers were subdivided into low (<10 ppm, N=16), medium (10-20 ppm, N=13) and high exposure groups (>20 ppm, N=14). In addition, there was exposure to benzene at mean concentrations of about 5-15 ppb. DNA strand breaks (measured by the alkaline Comet assay), 8-hydroxydeoxyguanosine (8-OHdG) and DNA repair capacity (determined by the amount of dicentric chromosomes and chromosome deletions after irradiation) were determined in peripheral leukocytes. DNA strand breaks and 8-hydroxydeoxyguanosine were significantly increased in all exposure groups in a dose related manner while a significantly lower DNA repair capacity was observed. The expression of CYP2E1 was higher in all exposure groups (significant at medium and high exposure) as compared to the controls as was the expression of the DNA repair genes h0GG1 and XRCC1 (significant for all groups). There was no correlation between benzene exposure and DNA damage or repair capacity. In summary, it is concluded that already at exposure levels below 20 ppm styrene leads to genotoxic effects (DNA strand breaks and 8-hydroxydeoxyguanosine), DNA repair capacity is decreased and the expression of DNA repair genes is increased.

Costa et al. (2012) studied workers in fiberglass-reinforced plants (75 exposed workers compared to 77 controls) for genotoxicity in relation to genetic polymorphism (CYP2E1, microsomal epoxide hydrolase, glutathione S-transferase). The mean styrene concentrations in air were 30.4 ppm and the mean concentrations of urinary metabolites were 443 mg/g creatinine (MA+PGA), both exceeding the threshold limit value (20 ppm) and the biological exposure index (400 mg/g creatinine) of ACGIH (2003). Significantly higher SCE frequency rates were observed in exposed workers, while the frequency of micronuclei by the cytokinesis-block micronucleus assay (CBMN) was not markedly modified by exposure. The tail length (TL) in the Comet assay was higher in exposed than in controls with a p value close to significance (p=.058). Although TL tended to be higher in almost all exposed subgroups (males/females; smokers/non-smokers) the difference did not reach statistical significance. With respect to genetic polymorphism, only elevated microsomal epoxide hydrolase had an influence leading to an increase in SCE levels in exposed worker. But this finding is not plausible, as this enzyme detoxifies styrene oxide, the genotoxic metabolite of styrene. In summary, in this study a significant increase of SCE, a numerical (but not statistically significant) increase in tail length in the Comet assay and no effect on the incidence of micronuclei in the blood of workers exposed to mean styrene air concentrations of 30.4 ppm was observed.

In a review Speit (2013) casts doubt on the usefulness of the cytokinesis-block micronucleus assay (CBMN) as an indicator of genotoxicity in human biomonitoring and questions the relevance of many published data. Based on the principle of the assay, increased micronuclei (MN) frequencies in binucleated cells (BNC) are mainly due to MN produced during the cultivation period. The sensitivity of the assay is limited because cytochalasin B (cyt-B) is added relatively late during the culture period and therefore the BNC that are scored do not always represent cells that have completed one cell cycle only. Due to this delay damaged cells can be eliminated or DNA damage be repaired. A comparison with the in vitro CBMN leads to the conclusion that that it is highly unlikely that DNA damage produced in vivo is the cause for increased MN frequencies after exposure to genotoxic chemicals. He concludes the MN scored in the CBMN assay are mainly formed in vitro and MN already produced in vivo do not substantially contribute. The experimental approach reduces the sensitivity of the assay as shown by comparison with in vitro studies. To his assessment increased MN frequencies can only be expected after in vivo exposure to ionizing radiation and radiomimetic chemicals or after very high exposure to chemical mutagens producing persistent DNA lesions. In summary, based on methodological considerations this publication questions the applicability of the CBMN for the assessment of genotoxic effects in exposed workers.

A new literature search up to March 2013 was carried out and the following paper on formation on protein adducts was identified:

Dai et al. (2010) developed an analytical method to quantify protein adducts to cysteine residues from styrene oxide. They could differentiate between α- and ß-adducts. After ip application of styrene (50 and 400 mg/kg bw/d over 5 days) both adducts were found at 400 but not at 50 mg/kg bw/d in the blood of mice.

In the literature update until Oct. 01, 2015, the following new information was found:

In the study of Sati et al. (2011) mentioned above, parameters for oxidative stress were compared for 34 styrene exposed male workers from a plastic factory (no further details) employed between 1-10 years with those in 30 age and sex matched controls (no further details). Ferric-reducing ability of plasma and GSH in blood were reduced while thiobarbituric acid reactive substances were increased. It is to be noted that the matching criteria for the control subjects are very sparse. In addition, although smoking was mentioned to have been analyzed as potential confounder in the statistical analysis, no indication is given about verification of smoking status. Importantly, no exposure data are available and the small group size is noted.

In the study of Helal and Elshafi (2012) mentioned above, 40 male workers (age 18-33 years) from an Egyptian plastic factory (no further details given) were compared with 40 unexposed controls from the same employer matched for age, sex, socioeconomic status, and smoking habits (no indication how smoking status was verified). All subjects participated voluntarily but no information is given for the selection criteria from the total workforce or for the participation rate. The styrene exposed workers worked for 12 h/d with 1 day off without any protection equipment. Urine was taken during workshift but no further specification is available and time of blood sampling is unknown. Styrene blood levels were 1117±64,52µg/l and urinary mandelic acid 246±21,60µmol/l, but these data are difficult to interpret because of uncertainties about sampling times. The frequency of chromosomal aberrations was significantly increased in the exposed as compared to controls and there was significant positive correlation between some types of chromosomal aberrations and exposure duration. The long working shift for the workers is noted as well as the unclear participation rate.

Hanova et al. (2011) extended the investigation of Hanova et al. (2010) based on the same subjects and results from the comet assay. They measured mRNA expression of the cell cycle genes (TP53, BCL2, BAX and p21CDKN1A) and correlated them with exposure, single strand breaks in the comet assay andγ-irradiation specific DNA repair. There was a statistically significant negative correlation between exposure and mRNA expression of TP53, BCL2, and BAX, and a positive correlation for mRNA expression of p21CDKN1A. Single stand breaks and endonuclease III sensitive sites were significantly increased with mRNA expression of TP53 and BCL2 but decreased with increasing mRNA expression of p21CDKN1A.γ-Irradiation specific DNA repair increased with p21CDKN1AmRNA levels. These correlations are difficult to interpret in terms potential exposure related effects of styrene. According to the authors, these results suggest a possible relationship between styrene exposure, DNA repair and gene expression of cell cycle genes.


Short description of key information:
A large number of studies have been published which have aimed to investigate the genotoxic potential of styrene in humans by examination of various endpoints in styrene exposed workers. Overall, given the lack of evidence of consistent relationships between exposure levels and study outcome, the lack of any consistent profile of endpoints and the absence of information on the relevance of the types of adducts seen and their mutagenic potential in vivo, there is no convincing evidence that styrene has shown mutagenic activity in humans. Hence, information from studies in experimental animals and other systems needs to be considered.

The overall picture presented by the in vitro assay results available is that at least in some test systems, styrene does posses some genotoxic potential in vitro. Metabolic activation (presumably to styrene oxide) is required for this activity. Styrene has been exhaustively studied in clastogenicity studies in animals up to toxic dose levels in some cases. There is no convincing evidence of styrene clastogenicity when the quality of the studies and the plausibility of the test results are considered. Equivocal results were obtained after exposure to high doses causing lethality.
Overall, negative results were obtained from in vivo chromosome aberration and micronucleus studies in the rat, hamster and the mouse following single or repeated exposures to styrene up to concentrations and/or doses causing systemic toxicity, via the inhalation, oral and intraperitoneal route in the tissues examined.

Endpoint Conclusion:

Justification for classification or non-classification

Overall, given the lack of evidence of consistent relationships between exposure levels and study outcome, the lack of any consistent profile of endpoints and the absence of information on the relevance of the types of adducts seen and their mutagenic potential in-vivo, there is no convincing evidence that styrene has shown mutagenic activity in humans.

The overall picture presented by the in-vitro assay results available is that at least in some text systems (including tests in vitro chromosome aberration studies in mammalian cells), styrene does posses some genotoxic potential in-vitro. Metabolic activation (presumably to styrene oxide) is required for this activity.

In summary, the available data in-vivo in experimental animals suggest that styrene is weakly positive in indicator tests detecting SCEs, DNA stand breaks and DNA adducts. In contrast, an in vivo UDS test performed in accordance with international guidelines did not reveal a genotoxic effect of styrene in mouse liver.

Overall, based on standard regulatory tests, there is no convincing evidence that styrene possesses significant mutagenic/clastogenic potential in vivo from the available data in experimental animals.

Therefore, a classification for mutagenicity according to EU-criteria (67/548/EEC) and to GHS-criteria (1272/2008/EC) is not warranted for styrene.