Review of the Styrene Assessment in the National Toxicology Program 12th Report on Carcinogens (2014)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

2 Review of the Styrene Profile in the National Toxicology Program 12th Report on Carcinogens To address part of its statement of task, the committee reviewed the sub- stance profile for styrene in the National Toxicology Program (NTP) 12th Re- port on Carcinogens (RoC) (NTP 2011a). The committee’s review was informed by many documents, including those in Table 1-1, and by comments submitted by stakeholder organizations and members of the general public. The committee also looked in detail at the primary literature cited in the background document for styrene and other literature identified by June 10, 2011 (the date on which the 12th RoC was released). To guide its review, the committee considered whether NTP described and conducted the literature search appropriately, whether all the relevant literature identified during the literature search was cited and sufficiently described in the background document, whether NTP had se- lected the most informative studies to support the listing determination, and whether NTP’s arguments supported its conclusions. This chapter is organized to follow the headings and organizational struc- ture of the substance profile as presented in the 12th RoC (NTP 2011a). Those headings are “Cancer Studies in Humans”, “Cancer Studies in Experimental Animals”, “Metabolism of Styrene”, and “Studies on Mechanisms of Carcino- genesis”. The committee also reviewed the section in the substance profile that presents properties, use, and production of and exposure to styrene. On the basis of its review and analysis of the 12th RoC substance profile for styrene, the committee ends the chapter with a discussion about whether the evidence pre- sented by NTP in the background document and the substance profile support the listing of styrene as “reasonably anticipated to be a human carcinogen”. CARCINOGENICITY NTP began the substance profile for styrene with a clear statement of its conclusions—that styrene is reasonably anticipated to be a human carcinogen. That conclusion was based on limited evidence of carcinogenicity from studies 32 

Review of the Styrene Profile in the NTP 12th Report on Carcinogens 33 in humans, sufficient evidence from studies in experimental animals, and sup- porting mechanistic data. The committee finds this paragraph to be necessary, informative, and succinct. Cancer Studies in Humans The “Cancer Studies in Humans” section of the substance profile for sty- rene considers whether the epidemiologic literature published by June 10, 2011, provides limited evidence of human carcinogenicity or whether that evidence reaches the level of being sufficient for such a listing. Overall, the background document and the substance profile include appropriate literature reviews and identify the most informative studies (NTP 2008, 2011a). The text and tables in the background document clearly describe and critique the major strengths and limitations of the key epidemiologic studies, and the background document itself presents accurate data summaries with a few minor exceptions, which are men- tioned below. The committee concludes that the description and analysis of literature presented in the background document and the substance profile support NTP’s classification of styrene in the 12th RoC as “reasonably anticipated to be a hu- man carcinogen”, as will be discussed in this section. The committee’s assess- ment is based on the following RoC listing criterion: “there is limited evidence of carcinogenicity from studies in humans which indicates that causal interpreta- tion is credible but that alternative explanations, such as chance, bias, or con- founding factors, could not adequately be excluded” (NTP 2011b). Neither the background document nor the substance profile was explicit about how NTP defined the terms limited and sufficient in the context of the epidemiology evi- dence. So, the committee used its professional judgment to develop and apply a set of factors that it used to evaluate the credibility of evidence on the human carcinogenicity of styrene. As described in Chapter 3, those factors were high estimates of relative risks or its surrogates; exposure–response relationships for any reliably established exposure metric; consistency of observations among independent cohort studies of the reinforced-plastics industry or between cohort and case–control studies; and at least two informative studies in independent populations or with varied study designs. The committee judged the evidence to be limited if the epidemiology evidence was credible but chance, bias, and con- founding could not be excluded. The evidence was judged to be sufficient if the epidemiology evidence was credible and chance, bias, and confounding could be excluded as an alternative explanation for the observed association. Lymphohematopoietic Cancers, Including Leukemia, Non-Hodgkin Lymphoma, Hodgkin Lymphoma, and Multiple Myeloma The classification of lymphohematopoietic cancers has evolved in recent decades, so there are some inconsistencies over time in the same cohort analyses 

34 Review of the Styrene Assessment in the NTP 12th Report on Carcinogens and inconsistencies among studies. For example, there have been changes in the diagnosis and classification of non-Hodgkin lymphoma—some cases of non- Hodgkin lymphoma would have been considered Hodgkin lymphoma before the 1980s (Banks 1992). According to a 1989 revision of the French–American– British classification criteria for leukemia, the M6 subtype of acute myeloid leukemia could be diagnosed if less than 30% of all nucleated cells in the bone marrow are blast cells (Bennett et al. 1985). Some of those cases would previ- ously have been classified as refractory anemia with excess blast cells in trans- formation with a major erythroid component (Bennett et al. 1982), two condi- tions that were not considered malignancies. The grouping of “all lymphohematopoietic cancers” includes many bio- logically distinct diagnoses in humans (NRC 2011), so it is not ideal to consider lymphohematopoietic cancers as constituting a homogeneous entity for etiologic assessment. However, the committee is limited by the methods used in the exist- ing studies and is unable to separate lymphohematopoietic cancers into finer categories if the investigators of the studies did not separate them. That has been a challenge for many types of exposures in the epidemiologic literature and is not specific to studies of styrene exposure. In addition, lymphohematopoietic cancers are infrequent, so individual studies may have low statistical power, as reflected by low numbers of observed and expected cancer cases or deaths. Giv- en those considerations, the committee thinks that to move forward with the review it is acceptable to consider lymphohematopoietic cancers as a group. The committee briefly summarizes below the evidence of carcinogenicity from tradi- tional epidemiologic studies. On the basis of the studies available to NTP by June 10, 2011, the commit- tee judges there to be limited but credible evidence that exposure to styrene in occupational setting is associated with an increased frequency of lymphohema- topoietic cancers. The evidence comes primarily from two occupational-cohort studies of workers in the reinforced-plastics industry in Europe (Kogevinas et al. 1994; Kolstad et al. 1994). Kogevinas et al. (1994) compared workers in the reinforced-plastics indus- try with different levels of styrene exposure and showed that longer time since first exposure (at least 10 years vs less than 10 years) was associated with a sig- nificantly higher mortality due to lymphohematopoietic cancers combined. Compared with workers who had an average exposure of less than 60 ppm (sev- en deaths), the mortality rate ratios (MRRs) in those who had an average expo- sure of 110–119 ppm, 120–199 ppm, and at least 200 ppm were 3.11, 3.08, and 3.59, respectively, with a p value of 0.019 in a test of linear trend. Kolstad et al. (1994) found that the standardized incidence ratio (SIR) for lymphohematopoietic cancers combined in workers at companies producing reinforced plastics was 1.20 (95% confidence interval [CI] 0.98–1.44). When the analysis was stratified by year of first employment, those who were first em- ployed during 1964–1970 had a significantly higher incidence of lymphohema- topoietic cancers (SIR = 1.32, 95% CI 1.02–1.67), whereas the SIRs for those first employed during 1971–1975 and during 1976–1988 were lower. That is 

Review of the Styrene Profile in the NTP 12th Report on Carcinogens 35 consistent with a possible exposure–response relationship, inasmuch as histori- cal personal air samples from this cohort showed that average styrene concentra- tions decreased from 180 ppm during 1964–1970 to 43 ppm during 1976–1988 (Jensen et al. 1990). It should be noted that the Kogevinas et al. (1994) study included approximately one-third of the subjects in the Kolstad et al. (1994) study who worked at plants where the main product during the study period was reinforced plastics, although the analyses of the former were of mortality and included both male and female workers while the latter analyzed cancer inci- dence and excluded female workers. Delzell et al. (2006) reported that exposure to styrene was associated with leukemia in a cohort of North American synthetic-rubber industry workers. Alt- hough the finding lends some credence to a possible leukemogenic role of occu- pational exposure to styrene, the committee does not give the study as much weight as NTP did, because of the workers’ concomitant exposure to 1,3- butadiene, a known carcinogen, and the difficulty of teasing out the influence of 1,3-butadiene in assessing the independent effect of styrene (NTP 2011a). Solid Tumors On the basis of the studies published by June 10, 2011, the committee judged that there was limited but credible evidence that exposure to styrene in the occupational setting is associated with an increase in the frequency of some cancers in addition to those of the lymphohematopoietic system. The substance profile for styrene reported credible evidence of a cause–effect relationship be- tween styrene exposure and cancers of the esophagus and pancreas (NTP 2011a). The committee thinks that the data regarding an association of styrene with kidney cancer should have been discussed in the substance profile. Of the studies reviewed by NTP, four cohort studies of the reinforced- plastics industry that covered subjects and controls in Washington state (Ruder et al. 2004), the United States (Wong et al. 1994), Denmark (Kolstad et al. 1994), and combined European nations (Kogevinas et al. 1994) were the most informative. The strengths of those studies and the associations observed are credible because the studies were of high quality, of varied design (mortality and incidence), and consistent in their findings of associations of styrene with these cancers, especially when internal comparisons—many with an apparent expo- sure–response relationship—were presented. Esophagus The primary evidence of esophageal tumors is from cohort mortality stud- ies in the reinforced-plastics industry (Kogevinas et al. 1994; Wong et al. 1994; Ruder et al. 2004), in which high exposures to styrene (especially from the 1940s through the 1970s) have been documented. The full study cohort of Ruder et al. (2004) in Washington state had an elevated rate of esophageal cancers 

36 Review of the Styrene Assessment in the NTP 12th Report on Carcinogens (standardized mortality ratio [SMR] = 2.30, 95% CI 1.19–4.02) on the basis of 12 deaths. The US cohort of Wong et al. (1994) had a similarly elevated SMR of 1.92 (95% CI 1.05–3.22) on the basis of 14 deaths. Those working in the high- exposure activity of open-mold processing had an SMR of 3.57 (two cases with exposure of at least 100 ppm-years,1 95% CI not reported). In an internal analy- sis in the European study by Kogevinas et al. (1994), exposure–response associ- ations were observed in the most highly exposed subcohort (workers engaged in lamination; SMR = 1.81, 95% CI 0.87–3.34, 10 deaths). In the group with high cumulative exposure, the SMR rose to 5.82 (95% CI 1.0–33.9) after at least 20 years since first exposure. Pancreas The substance profile for styrene appropriately noted observations of high SMRs for pancreatic tumors, especially in high-exposure subcohorts of three of the reinforced-plastics cohorts, although the 95% CIs around some of the SMRs included 1.0. In Ruder et al. (2004), the SMR was 1.43 (95% CI 0.78–2.41) for the Washington state comparison group that was based on 14 deaths, and the SMR in high-exposure workers was 1.88 (95% CI 0.51–4.81). Kolstad et al. (1995) reported the incidence rate ratio for high-exposure workers to be 2.20 (95% CI 1.1–4.5) on the basis of 17 incident cases. Kogevinas et al. (1995) re- ported the SMR in laminators to be 1.48 (95% CI 0.76–2.58) on the basis of 12 deaths. In the latter study, after at least 20 years since first exposure, the SMR was 2.05 (95% CI 0.58–7.29), and the cumulative positive exposure trend had a p value of 0.068. Kidney There were multiple reports of associations of kidney cancers with styrene exposure in the reinforced-plastics industry, and some exhibit exposure– response relationships. Specifically, the relatively small study of Ruder et al. (2004) found elevated SMRs for kidney cancer (SMR = 1.43, 95% CI 0.57– 2.95) on the basis of seven deaths, but in the high-exposure group the SMR was 3.60 (95% CI 0.98–9.20) on the basis of four deaths. The US study by Wong et al. (1994) found an SMR of 1.75 (95% CI 0.98–2.89) on the basis of 15 deaths; the association was higher in workers exposed for more than 2 years to open- mold processing: an SMR of 4.57 (95% CI not given). In the European cohort of Kogevinas et al. (1994), an exposure–response relationship of MRRs with in- creasing cumulative exposure to styrene was observed (p for trend = 0.12). Those studies did not attain the traditional level of statistical significance for 1 “ppm–years” is the cumulative exposure calculated by multiplying the number of years by the (average) concentration in parts per million. 

Review of the Styrene Profile in the NTP 12th Report on Carcinogens 37 kidney cancer. However, kidney cancer should also have been mentioned in the styrene substance profile as having some evidence of styrene carcinogenicity. A case–control study of renal-cell cancer and occupational exposures, including exposure to styrene, was published in February 2011 (Karami et al. 2011), after the background document was issued but before the publication of the substance profile. That study should have been included in the NTP evaluation for styrene. It is discussed in more detail in Chapter 3. Genetic Damage The background document and the substance profile for styrene (NTP 2008, 2011a) cite informative studies that assessed the genetic damage caused by styrene. The brief "Genetic Damage" section under "Cancer Studies in Hu- mans" summarized pertinent findings regarding adducts, single-strand DNA breaks, and an elevated frequency of chromosomal aberrations in workers ex- posed to styrene. Genetic damage may not result in clinical disease, but such information may inform underlying mechanisms of carcinogenicity. For this reason, NTP could consider describing the evidence of genetic damage in the section "Studies of Mechanisms of Carcinogenesis" rather than the section “Cancer Studies in Humans” if the background document and the substance pro- file for styrene are updated for a future edition of the RoC. Cancer Studies in Experimental Animals In the background document for styrene, NTP summarized findings from several studies in which a carcinogenic response was evaluated in mice or rats after administration of styrene by various routes (inhalation, ingestion via ga- vage or drinking water, and injection) (NTP 2008). The committee is not aware of any important studies of styrene carcinogenicity in animals that were availa- ble before June 10, 2011, that were not included in the background document. The characterization of styrene carcinogenicity findings in mice and rats as pre- sented in the substance profile reflects the state of knowledge as of June 10, 2011. In the substance profile, NTP correctly focused on the key animal studies that provide evidence for and against styrene carcinogenicity (NTP 2011a). The substance profile states that lung tumors were not observed in styrene-treated rats and briefly summarizes equivocal findings regarding mammary gland tu- mors. Findings of lung tumors in CD-1 mice after inhalation exposure (Cruzan et al. 2001) and supporting data from a study of perinatal styrene exposure in mice (Ponomarkov and Tomatis 1978) are appropriately described, as are the negative findings. In the National Cancer Institute oral (gavage) study (NCI 1979), alveolar and bronchiolar adenomas and carcinomas combined were increased significant- ly in male B6C3F1 mice compared with concurrent study controls: concurrent 

38 Review of the Styrene Assessment in the NTP 12th Report on Carcinogens controls, 0/20 (0%); low dose, 6/44 (14%); high dose, 9/43 (21%). The authors noted the absence of lung tumors in the controls. Historical vehicle controls from the same laboratory (control male mice from similar studies in the same laboratory that also received only the vehicle, corn oil) were available for com- parison and also showed no lung tumors (0/40). However, the authors consid- ered the historical vehicle control population too small to provide useful per- spective. A much larger number of historical untreated controls (controls from similar studies in the same laboratory that received no treatment) had an average lung tumor incidence of 12% (32/271) with incidences as high as 20% in two studies. If the concurrent control male mice in the NCI styrene bioassay had a lung tumor incidence similar to the average in the historical untreated controls, the incidences in the styrene-exposed animals would probably not have been significantly increased. Because the lung tumor incidence was increased only when compared with concurrent (and not historical untreated) controls, the NCI study authors concluded that the male mouse lung tumor findings were sugges- tive but that no convincing evidence of carcinogenicity was found. In an effort to provide further insight into whether the incidence of lung tumors in the concurrent controls in the NCI oral bioassay was unusually low, NTP obtained historical vehicle control mouse lung tumor incidences from other laboratories used for NCI bioassays. The tumor incidence data came from stud- ies that were carried out at about the same time as NCI (1979), under the same protocol, and with animals from the same source. NTP reported in the back- ground document that the incidence of combined lung tumors in those historical vehicle controls was 4% (11/273) and concluded that the incidence in the con- current controls in NCI’s styrene gavage study was not unusually low. Because the increase in lung tumors in male mice was significant when compared with concurrent controls, NTP interpreted the study as positive, and it is so described in the substance profile. The issue of controls is not discussed in the substance profile and appears only in the background document. The appropriate controls for the lung tumor incidences observed in male mice in the NCI oral bioassay constitute a pivotal issue in interpretation of the study as positive on one hand or negative, equivocal, or inconclusive on the oth- er hand. It is a basic facet of the design of any scientific experiment that control animals need to be selected in such a way as to minimize, to the extent possible, variables that might influence the results other than the variables being specifi- cally studied. With regard to cancer bioassays, it is well established that many characteristics of the conduct of a study unrelated to the treatment being evalu- ated can affect tumor rates (Haseman et al. 1984; Festing and Altman 2002; Keenan et al. 2009). Some characteristics related to the genetic makeup of the experimental animals used in the bioassay are the strain and substrain of the experimental animals, the supplier of the experimental animals, the breeding colony and the subpopulation of the colony from which the experimental ani- mals were derived, procedures used for monitoring and controlling genetic drift in breeding populations, and the genetic homogeneity of animals used in the experiment. Husbandry practices are also important, both at the suppliers’ sites 

Review of the Styrene Profile in the NTP 12th Report on Carcinogens 39 and at the experimental sites. Characteristics that can influence bioassay results related to husbandry practices include ventilation air (changes, filtration, moni- toring, or exhaust), caging (structure or cleaning), bedding (frequency of re- placement, source, chemical composition, or outgas monitoring), diet (availabil- ity to animals, vendor, nutrient composition, chemical composition, or quality- monitoring procedures), drinking water (availability to animals, sources, purity, or quality-monitoring procedures), and treatment vehicle (volume or dose, ven- dor, chemical composition, or quality-monitoring procedures). Controls and treated animals must be matched with respect to each of those characteristic to the greatest extent possible for the study to yield valid comparisons. That is best accomplished by using concurrent controls. If historical vehicle controls or his- torical untreated controls are used for comparison, the extent to which they might differ from treated animals with respect to the characteristics listed above must be considered. NTP has been criticized for using historical vehicle control data from oth- er laboratories to determine that the lung tumor incidences in the NCI styrene bioassay were not unusually low (Rhomberg et al. 2013). It is easy for studies conducted in different laboratories, even under the same experimental protocol, to vary in subtle but important respects (as outlined above) and consequently to yield different tumor incidences. Therefore, drawing historical controls from other laboratories is seldom justified (Haseman et al. 1984). The committee con- siders the comparison of concurrent controls in the NCI styrene oral bioassay (NCI 1979) with historical vehicle control data from other laboratories to be of little value. The same concern applies to comparison with historical untreated controls in the NCI bioassay (NCI 1979). The vehicle alone can influence tumor rates, and historical vehicle controls and historical untreated controls cannot be considered equivalent (Haseman et al. 1984). Therefore, in the case of the NCI styrene bioassay, the interpretive value of comparison with historical untreated controls is also of limited value. Although limited in number, the historical ve- hicle controls from the same laboratory at about the same time are most relevant and are consistent with the concurrent controls. The committee finds that the use of concurrent controls reported by NCI (1979) is appropriate. Metabolism of Styrene The section “Metabolism of Styrene” in the substance profile provides in- formation on metabolites of styrene and on the specific CYP450 enzymes that are probably involved (NTP 2011a). Styrene, through formation of active me- tabolites, is thought to be capable of inducing genotoxic and cytotoxic effects. The metabolism of styrene is complex, and multiple metabolites are formed. One, styrene-7,8-oxide, is known to be genotoxic (see below), but the genotoxi- city of others has not been thoroughly investigated. The contribution of specific metabolites to the genotoxic and cytotoxic effects of styrene may be organ- specific and the metabolites primarily responsible for cytotoxicity may not be 

40 Review of the Styrene Assessment in the NTP 12th Report on Carcinogens the same as those responsible for genotoxicity. An explicit statement of that cir- cumstance would have enhanced the clarity of the metabolism section of the substance profile. As indicated in the substance profile, the strongest evidence of carcinogenicity in laboratory animals comes from studies in mice. In the mouse, lung tumors have been observed after either inhalation or oral exposure to sty- rene; this target-organ specificity is thought to be due to pulmonary activation of styrene in the mouse (see the discussion below and also the sections on Cancer Studies in Experimental Animals and Lung Cytotoxicity in Mice). Although the relevance of the mouse lung tumor response to humans has been questioned, the committee concludes that there was insufficient information when the substance profile was completed, so the mouse lung tumor response should not be exclud- ed as irrelevant. Thus, information on styrene metabolism in both the human and mouse is important for establishing a robust understanding of the carcinogenicity of this compound. The metabolism section of the substance profile provides a succinct overview of some aspects of styrene metabolism and, in general, is written clear- ly. The references provided are appropriately interpreted and described in the background document for styrene. However, the substance profile does not pro- vide complete citations for the information that is presented; for example, the substance profile states without citation that over 90% of styrene is metabolized to styrene-7,8-oxide. Information is also provided on the state of knowledge regarding the specific CYP450s associated with styrene metabolism to styrene- 7,8-oxide and their distribution in mice and humans, and some information is provided on metabolites other than styrene-7,8-oxide or downstream metabolites of styrene that have been detected in mice and humans. A clear focus of the metabolism section of the substance profile is on the formation of styrene-7,8-oxide. Although not explicitly stated in the substance profile, that focus is presumably due to the facts that styrene-7,8-oxide is geno- toxic and that styrene-7,8-oxide–protein and styrene-7,8-oxide–DNA adducts have been detected after styrene exposure. However, it is not known whether styrene-7,8-oxide is key to the cytotoxic response or whether it is the sole geno- toxic metabolite of styrene. Minimal information is provided on the formation of 4-vinylphenol (presumably through the intermediate styrene-3,4-oxide), which is identified merely as a minor pathway. Evidence available at the time of the release of the substance profile indi- cates that the role of metabolism may be more complex than portrayed in the substance profile for styrene. Specifically, although 4-vinylphenol is a minor metabolite of styrene, it is considerably more potent than styrene or styrene-7,8- oxide in inducing lung injury (Carlson et al. 2002; Carlson 2004; Cruzan et al. 2005), and it has been suggested that the aromatic ring–derived metabolites are important in the pulmonary toxicity of styrene in mice (Cruzan et al. 2005; Cru- zan et al. 2009). To the extent that cytotoxicity and reparative cell proliferation play a critical role in pulmonary carcinogenesis in the mouse, a comprehensive description of styrene metabolism should include information on all metabolites that are thought to be important for cytotoxic responses. 

Review of the Styrene Profile in the NTP 12th Report on Carcinogens 41 The cellular balance of activation and detoxification pathways is an im- portant consideration relative to target-organ sensitivity to any metabolically activated toxicant, including styrene. The substance profile includes some in- formation on glutathione S-transferase in this regard. Information is also availa- ble from mouse studies on the potentially important role of epoxide hydrolase (Carlson 2010b), but that information is not included in the substance profile. The cellular response to a metabolically activated compound depends critically on both the activation rate and the detoxification rate. It is possible that the bal- ance between activation and detoxification may differ among species and among target organs within a species. Thus, in accordance with fundamental toxicology principles, a comprehensive toxicologic evaluation requires detailed understand- ing of the full metabolic spectrum. The incompleteness of the information on metabolic detoxification pathways in the substance profile detracts from the quality of this section. In summary, the substance profile for styrene provides much information on its metabolism. The information that is provided is correct, but the profile is not complete in its presentation of published information on styrene metabolism. There is a lack of clarity in the potential complexity of metabolism relative to styrene’s cytotoxic and genotoxic effects. The information on styrene-metabolite phase II detoxification pathways is incomplete, and this weakens the scientific balance of this section of the substance profile. That information was included in the background document for styrene but was not provided clearly in the sub- stance profile itself. Studies on Mechanisms of Carcinogenesis The section “Studies on Mechanisms of Carcinogenesis” in the substance profile for styrene (NTP 2011a) and supporting information in the background document (NTP 2008) summarize the mechanistic events that might link styrene exposure to cancer in experimental animals and humans. The mechanistic evi- dence on styrene and its major metabolites that was available to NTP is exten- sive and comes from a variety of studies in diverse model systems and from exposed humans. Although neither the substance profile nor the background document provides the exact search strategy that was used in collecting the evi- dence on the mechanisms of carcinogenesis, these documents present a bal- anced, comprehensive, and thorough review of the literature on the subject. Evi- dence tables and narrative descriptions of each study were used in the background document to present mechanistic evidence from primary studies and meta-analyses, and the committee finds the presentation of information to be inclusive and balanced. The background document and the substance profile are in agreement with Part B of the styrene expert review panel (Phillips et al. 2008) that “the mecha- nisms of styrene carcinogenicity are not fully understood” (NTP 2011a, p. 385). Several potential mechanisms have been studied and are identified as separate 

42 Review of the Styrene Assessment in the NTP 12th Report on Carcinogens subsections below. Overall, it is clearly stated that the carcinogenicity of styrene depends on its metabolism to styrene-7,8-oxide and other reactive intermediates and that such metabolism occurs in both rodents and humans. Styrene-7,8-oxide has been listed as reasonably anticipated to be a human carcinogen since the 10th RoC (NTP 2002). Even though species, tissue, and individual differences in metabolic capacity or in enzymes involved in styrene metabolism have been reported, strong evidence presented in the substance profile and the background document for styrene suggests that mechanistic events that may lead to carcino- genesis (such as genotoxicity) occur in both exposed rodents and humans. Fur- thermore, the listing correctly states that multiple mechanistic events may occur and that they are “not necessarily mutually exclusive” (NTP 2011a, p. 385). The substance profile for styrene identifies three modes of action: geno- toxicity, cytotoxic effects, and immunosuppression. Genotoxicity is identified as relevant to all cancer sites, and cytotoxicity and immunosuppression are identi- fied as site-specific. Those are reasonable categories, but they are somewhat inconsistent with the presentation and categorization of the mechanistic evi- dence in the background document and in Part B of the styrene expert panel report (Phillips et al. 2008). The executive summary of the background docu- ment, the body of the background document, and the substance profile present three styles of categorization of the same evidence; this may create some confu- sion with regard to the relative weight assigned by NTP to the mechanistic evi- dence in the overall cancer hazard classification of styrene. First, the executive summary of the background document includes separate sections on “genetic damage” and “mechanistic data”, even though the latter section mentions “geno- toxic pathway” as an important mechanistic event. Second, the body of the background document creates a poorly rationalized separation between sections “5.4 Genetic and related effects” and “5.5 Mechanistic studies and considera- tions”, and the latter also contains subsection “5.5.1 Genotoxicity”. Third, the substance profile for styrene, while maintaining the separate subsection “Geno- toxicity” in accord with the executive summary and the main background docu- ment, emphasizes lung cytotoxicity in mice (information consistent with the background document subsection “5.5.4 Cytotoxic effects of styrene on mouse lung”) and immunosuppression. Those mechanisms have not been clearly sepa- rated from among the likely mechanisms of styrene carcinogenesis in the back- ground document. In contrast, “5.5.2 Gene expression and apoptosis” and “5.5.3 Oxidative stress” mechanisms are not mentioned in the substance profile. Alt- hough such inconsistencies are immaterial for the purpose of the cancer hazard classification of styrene in the 12th RoC, the committee suggests that NTP pro- vide greater concordance among various documents to avoid confusion. Genotoxicity Both the background document and the substance profile identify genotox- icity of styrene as an important carcinogenesis mechanistic event that applies to 

Review of the Styrene Profile in the NTP 12th Report on Carcinogens 43 multiple potential target tissues and that operates in both experimental animals and humans exposed to styrene. The evidence that metabolism of styrene to sty- rene-7,8-oxide and other electrophiles that are capable of covalently binding to DNA, of forming adducts, and of causing other types of DNA damage that result in mutations and higher order cytogenetic damage is extensive and has been thoroughly reviewed and clearly presented. In the background document, over 20% of the entire document is dedicated to the presentation and critical evalua- tion of the data pertinent to this mechanism. The substance profile concludes correctly that styrene-associated “genotoxicity [is] relevant to all types of can- cer” and that “a causal relationship between styrene exposure and cancer in hu- mans is . . . supported by the finding of DNA adducts and chromosomal aberra- tions in lymphocytes from exposed workers” (NTP 2011a, p. 383). Similar conclusions regarding the role of metabolism and evidence of genotoxicity were drawn in Part B of the styrene expert panel report (Phillips et al. 2008). Hence, the overall role of genotoxicity of styrene as a key mechanistic event is well documented, supported by experimental evidence, and consistent with the rec- ommendation of the expert panel. Although the genotoxicity information in the substance profile for styrene is correct and comprehensive, the appropriateness of specific references selected to substantiate most of the arguments is somewhat difficult to ascertain because the literature base that is used to support these mechanisms and each specific event is large. The committee recognizes that no single study or even a collec- tion of publications can represent the extent and diversity of mechanistic evi- dence; it therefore suggests that references included in this section be labeled as representative of similar studies. Perhaps the best way that the evidence can be summarized and presented is that depicted in “Table 5-18. Genetic and related effects of styrene” in the background document (NTP 2008). Such a systematic presentation of evidence may be amenable to being used as a guide for the brief summary included in the substance profile. The document could be further im- proved by pointing out, for each effect or model system, which evidence comes from studies of styrene and which from studies of styrene-7,8-oxide. As noted above, the role of styrene metabolism is necessary for its genotoxicity; however, because human evidence of genotoxic effects comes from subjects exposed to styrene, not to its metabolites, it is clear that styrene is metabolized in humans to genotoxic intermediates. Questions still remain as to which enzymatic systems may be responsible and as to the quantitative differences in styrene metabolism among species or individuals in the same species. Thus, the detailed attention devoted to human evidence in this and other sections concerned with genotoxici- ty end points is warranted and appropriate. Lung Cytotoxicity in Mice The objective of the section “Lung Cytotoxicity in Mice” of the substance profile for styrene is to address the mechanistic events by which exposure to 

44 Review of the Styrene Assessment in the NTP 12th Report on Carcinogens styrene leads to the formation of tumors, and this objective was clearly ad- dressed by NTP. Establishing the mechanistic events for compounds—such as styrene, whose toxic and carcinogenic potential appears to require metabolic transformation by biologic tissues—depends on several issues. Some of those issues are identification of target tissues and cell populations; definition of the cellular response to exposure; establishment of the pattern of toxicity on the basis of exposure characteristics, including route, duration, and concentration; definition of variations in cytotoxic response between species and between strains of species; comparison of changes in cytotoxic response produced by altering the function of relevant enzyme systems through either chemical inhibi- tors or molecular alteration; and comparison of changes in cytotoxic response produced by the parent compound and its potentially toxic metabolites. NTP identified 26 references that were published before the release date of the substance profile for styrene, June 10, 2011, and were relevant to lung cyto- toxicity in mice and the role of specific activating enzyme systems and styrene metabolites. All except one (Cruzan et al. 2009) were cited and discussed in the background document for styrene. The substance profile for styrene addresses the most relevant of the studies, including that by Cruzan et al. (2009). For is- sues on which studies disagreed, such as the cytotoxic response in Clara and alveolar type II cells in the lungs of rats, all the relevant studies identified in the background document are cited and discussed in the substance profile for sty- rene. With respect to the cytotoxicity of different styrene metabolites, the most detailed, rigorous, and informative studies cited in the background document are discussed in the substance profile. However, in an independent literature search the committee identified six additional studies that could have added strength to the discussion of the mechanisms by which styrene produces cytotoxicity in the lungs (Harvilchuck and Carlson 2009; Harvilchuck et al. 2008, 2009; Carlson 2010a,b; Meszka-Jordan et al. 2009). Thus, the background document could be strengthened by providing a full evaluation of the strengths and limitations of the literature relevant to lung cytotoxicity and the potential roles of specific me- tabolites and activating enzymes. The studies are discussed in more detail in Chapter 3. Of the material that was included in the background document, NTP se- lected the most appropriate for inclusion in the substance profile for styrene. As emphasized in the metabolism section, cytotoxicity produced by a bioactivated compound is the result of the balance between enzymatic capability for activat- ing the compound and cellular capacity for detoxifying the reactive metabolites. When the metabolites are transported by the circulation, organs that have little activation capability may also become targets if their detoxification systems are overwhelmed. The substance profile for styrene could have been strengthened by broadening the discussion to include studies that assess oxidative stress pro- duced by styrene and its metabolites, modulation of styrene toxicity by detoxifi- cation pathways, and toxicity produced in other organ systems, such as the gas- trointestinal, urinary, and lymphohematopoietic systems. Broadening that discussion may provide insights into species differences (for example mouse vs 

Review of the Styrene Profile in the NTP 12th Report on Carcinogens 45 rat vs human) related to styrene sensitivity and mechanisms of target organ sen- sitivity to styrene or styrene-7,8-oxide. Immunosuppression An intact immune system is critical if an organism is to recognize or coun- teract the damaging effects of chemical substances, infectious agents, or neo- plastic cells. Epidemiologic studies have shown that several chemical substanc- es, particularly vapor chemicals, may alter immune function in humans and animals (Weill et al. 1975; Aranyi et al. 1986; Boverhof et al. 2013). Immunode- ficiency or immunosuppression is theorized to be a possible mechanism by which exposure to a chemical substance might lead to carcinogenesis. In the substance profile, NTP included a brief section on immunosuppres- sion. The substance profile indicates that CYP2E1, the enzyme involved in con- verting styrene to styrene-7,8-oxide, is expressed in lymphocytes and hemato- poietic stem cells (Kousalova et al. 2004; Siest et al. 2008); this suggests that cytotoxicity might occur in these cells and might damage the immune system after exposure to styrene. Genotoxicity has been detected in peripheral lympho- cytes of styrene-exposed workers (Biro et al. 2002), and NTP discusses this point in several places: the section “Genotoxicity” in the substance profile and the corresponding sections “Genetic and Related Effects” and “Mechanistic Studies and Considerations” in the background document. The committee found that NTP provided inadequate evidence and few citations to support the argu- ment that exposure to styrene may cause immunosuppression. Basic but critical data pertaining to immunity—such as the number of lymphocytes, the weight of lymphoid organs, the function of systemic and localized lymphoid organs, and effects on innate vs specific immunity after styrene exposure in experimental animals or humans—were not fully reviewed in the background document or the substance profile for styrene. In addition, the committee identified some typo- graphic errors in the substance profile for styrene that should be corrected. In the sentence in the substance profile that states, “Veraldi et al. (2006) concluded that there was immediate evidence for the immunotoxicity of styrene oxide”, “im- mediate” should be changed to “intermediate”, and the sentence should refer to styrene, not styrene-7,8-oxide. PROPERTIES The section “Properties” of the substance profile for styrene details major physicochemical characteristics of the compound, including chemical stability, reactivity, and flammability (NTP 2011a). Overall, this brief section serves its purpose well and provides necessary information on the chemical itself. The substance profile also includes information on styrene-7,8-oxide, so a reference should be made to the substance profile for styrene-7,8-oxide (NTP 2011c). 

46 Review of the Styrene Assessment in the NTP 12th Report on Carcinogens USE The substance profile and background document provide a comprehensive review of industrial uses of styrene (NTP 2008, 2011a). This section makes it clear that styrene is used primarily in the production of polymer products and resins and that humans may therefore come into contact with styrene through a variety of consumer products and diverse manufacturing processes. Overall, the section demonstrates that “a significant number of persons residing in the United States are exposed,” which is one of the requirements for a substance to be listed in the RoC. PRODUCTION The section “Production” in the substance profile covers the chemical pro- cesses that are used to manufacture styrene and provides estimates of domestic production and import and export volumes (NTP 2011a). Styrene is a high- volume production chemical, and its manufacture is increasing. This section further supports the notion of potentially wide exposure to styrene in the United States inasmuch as nearly 40 lb of styrene were produced per person in the Unit- ed States in 2006. EXPOSURE The section “Exposure” of the background document is critical for the in- terpretation of much of the epidemiologic data (NTP 2008). The substance pro- file makes it clear that exposure to styrene has been documented in both occupa- tional settings and the general population (NTP 2011a). In addition, smoking is identified as an important source of exposure to styrene, and it is correctly noted that workers who are exposed to styrene may also suffer effects of nonoccupa- tional sources of styrene, such as cigarette smoke, outdoor and indoor air, food, and water. Exposure sources, duration, frequency, and concentrations; measures of exposure (for example, average, duration, or cumulative exposure); and bi- omarkers of exposure assist in determining the linkages between exposure and adverse health effects. Exposures may occur in the workplace and in community settings, although sources are not the same and occupational exposures may differ from environmental exposures in intensity, duration, frequency, and route of exposure. Because of these differences, occupational epidemiology studies sometimes inform environmental epidemiology studies. Such information pro- vides critical context for the hazard classification of any chemical, including styrene. The background document for styrene contains relevant information and is comprehensive and well organized. The committee concludes that ample evi- dence of widespread exposure to styrene justifies its consideration for listing in the RoC. 

Review of the Styrene Profile in the NTP 12th Report on Carcinogens 47 The committee identified several revisions that would improve the clarity and conciseness of the presentation of the information. First, different types of industries may be compared in a quantitative manner, and the most logical way of presenting the information is perhaps by magnitude of exposure. Where in- formation on smoking status is available, special attention should be paid to dif- ferences between smokers and nonsmokers in the risks posed by workplace ex- posure. Time trends, if any, need to be identified. A clear description of methods of exposure assessment is especially important for the interpretation and clarifi- cation of the exposure component in the epidemiologic studies. Second, the sub- section “General population” should be divided into exposures from smoking, including second-hand exposure (that is, environmental tobacco smoke), and exposures from other sources. NTP could also consider including information on whether there is a relationship between pack-years of smoking and styrene ex- posure and on the magnitudes of styrene exposure from tobacco-smoking vs exposure in various occupations. Third, NTP could consider including a discus- sion of the effects of various other sources of potential exposure in the non- smoking general population. This section devotes considerable space to the dis- cussion of food as a source of styrene, but it is listed last among all possible sources. All possible units of exposure (μg/kg of body weight, μg/m3, μg/L, and ppm) are mentioned. To improve clarity, the units of exposure could be stand- ardized. Fourth, the subsection “General population” could begin by pointing out that nearly all members of the general population, not only those exposed in a workplace or through smoking, have detectable styrene in their biologic fluids (for example, blood and breast milk). That information makes it clear not only that there is a high potential that “a significant number of persons residing in the United State are exposed” to styrene but that there are detectable concentrations of styrene in most people tested. REGULATIONS AND GUIDELINES This section in the substance profile provides a comprehensive list of rules, regulations, and advisory notices that pertain to styrene (NTP 2011a). Many government agencies in the United States have set quantitative limits of styrene exposure in various scenarios (such as the Food and Drug Administra- tion’s maximum permissible level of styrene in bottled water or the Occupation- al Safety and Health Administration’s ceiling concentration and permissible exposure limit for styrene) and many agencies regulate the production, use, dis- tribution, and disposal of styrene. The level of detail provided on these varied greatly, and it is not clear whether the appropriate source of the information can be easily identified. For example, the Department of Homeland Security regula- tion is identified by a clear reference to the Code of Federal Regulations, where- as other regulations are noted without proper links to the appropriate documents. If this section is to provide not only information but proper references to other sources, this part of the discussion could be improved in future editions of the RoC. 

48 Review of the Styrene Assessment in the NTP 12th Report on Carcinogens SUGGESTED REVISIONS FOR THE LISTING OF STYRENE IN THE REPORT ON CARCINOGENS Through its review of the background document and substance profile for styrene, the committee identified several revisions that could be made to im- prove future iterations of the listing of styrene in the RoC (see Table 2-1). The committee recognizes that the Office of the Report on Carcinogens is managing an open and public process and that it must produce RoC editions that are scien- tifically sound, consistent, and timely. Its comments and suggestions for future revisions are not intended to lead to additional layers of complexity or to delay future editions of the RoC. Addressing the suggestions in Table 2-1 would add clarity and improve the presentation of information in NTP’s assessment of sty- rene, but making the revisions would not likely change the overall conclusion of carcinogenicity presented in the substance profile. The committee identified two overarching improvements that could be made to the presentation of information in the background document and sub- stance profile. The first observation pertains to NTP’s identification of relevant literature in support of the assessment of styrene. Although the committee did not identify many relevant studies that NTP missed in its assessment of styrene, there was no discussion of the literature-search strategy in the background doc- ument. The committee was able to obtain information about the literature searches directly from NTP (Bucher 2013; see Appendix C). There are several organizations who have developed or have commented on best practices for evi- dence identification (Higgins and Green 2008; CRD 2009; AHRQ 2011; IOM 2011; NRC 2014). Describing NTP’s literature identification process in greater detail in the background document for styrene, including inclusion and exclu- sion criteria, the date of the search, the publication dates searched, and the roles of experts involved in reviewing the literature, would have provided additional transparency to the development of the background document. In spite of that deficiency, the essential literature appears to be cited and discussed appropriate- ly in the substance profile and the background document. The second area of improvement identified by the committee pertains to the way in which NTP used the listing criteria to reach the determination that styrene is “reasonable anticipated to be a human carcinogen”. The substance profile states that styrene is “reasonably anticipated to be a human carcinogen based on limited evidence of carcinogenicity from studies in humans, sufficient evidence of carcinogenicity from studies in experimental animals, and support- ing data on mechanisms of carcinogenesis” (NTP 2011b). The committee is clear about NTP’s determination of “sufficient evidence of carcinogenicity from studies in experimental animals” on the basis of the listing criteria instruction that experimental animal evidence is considered sufficient if there is an increase in “(1) multiple species or to multiple tissue sites, or (2) by multiple routes of exposure, or (3) to an unusual degree with regard to incidence, site, or type of 

Review of the Styrene Profile in the NTP 12th Report on Carcinogens 49 tumor, or age at onset” (NTP 2008, p. v). However, the background document and substance profile are less clear how NTP came to its conclusion that the listing criteria for “limited evidence of carcinogenicity from studies in humans” (NTP 2008, p. v) had been fulfilled. The introductory section of the 12th RoC describes the multiple layers of expert judgment involved in the RoC process, but neither the background document nor the substance profile for styrene trans- parently describe the way in which the different studies and the attributes of those studies were integrated to support a conclusion of limited evidence. The background document would be more transparent if it stated the considerations that were used to evaluate evidence from studies in humans (such as study popu- lation characteristics, approach used for exposure assessment, potential for bias and confounding, and precision of estimate of effect); if it described the struc- tured approach that experts used in their application of the listing criteria to reach a conclusion; and if it explicitly defined limited evidence and sufficient evidence in terms of the number of studies and attributes of those studies that would lead to such conclusions. NRC (2014) discusses organizing principles and qualitative and quantitative approaches for integrating evidence that NTP might find helpful in its application of the RoC listing criteria to the scientific literature for styrene. CONCLUSIONS The committee concludes that NTP correctly determined that styrene should be considered for listing in the RoC. There is sufficient evidence of ex- posure to a significant number of persons residing in the United States to war- rant such consideration. NTP adequately documented that exposure to styrene occurs in occupational settings and in the general public regardless of smoking status. After conducting a scientific review of the styrene assessment presented in the NTP 12th RoC, the committee finds that the overall conclusion reached by NTP in 2011, that styrene is “reasonably anticipated to be a human carcinogen”, was appropriate. The following points of the listing criteria support NTP’s con- clusion:  “There is limited evidence of carcinogenicity from studies in humans” (NTP 2008). Publications available to NTP as of June 10, 2011, provided limited but credible evidence that exposure to styrene is associated with lymphohematopoietic, pancreatic, and esophageal cancers. The most in- formative human epidemiologic studies that support that conclusion are those by Ruder et al. (2004), Wong et al. (1994), Kolstad et al. (1994), and Kogevinas et al. (1994). The evidence is limited in that “chance, bias, or confounding factors could not be adequately excluded” (NTP 2008). 

50 TABLE 2-1 Suggested Clarifications and Updates for the Styrene Substance Profile and Background Document in Future Editions of the Report on Carcinogens Sections in the Substance Profile for Styrene Suggested Revisions Carcinogenicity  Add a discussion of the literature-search strategy in the background document for styrene. Cancer Studies in Humans  Include a review of Karami et al. (2011) and add a discussion of the evidence of an association of styrene with kidney cancer on the basis of that study and findings in the reinforced-plastics industry cohort studies.  Include and discuss a case–control study of non-Hodgkin lymphoma by Cocco et al. (2010).  Describe the evidence of genetic damage in the section “Studies of Mechanisms of Carcinogenesis” rather than the section “Cancer Studies in Humans”. Metabolism  Include a more complete presentation of information on styrene metabolism in the substance profile, particularly with respect to detoxification pathways. Studies on Mechanisms of Carcinogenesis  Provide greater concordance among the headings of the executive summary of the background document, the body of the background document, and the substance profile.  Include information in the substance profile on the potential role of oxidative damage to DNA and the important role that DNA repair, or individual variability in DNA-repair response, may play in the genotoxicity of styrene to augment and strengthen the overall genotoxicity evidence.  Consider including studies that assess o toxicity produced in organ systems other than the respiratory system, such as the gastrointestinal, urinary, and lymphohematopoietic systems. o toxic responses in animals whose bioactivation potential has been modified by gene manipulation or administration of specific inhibitors. o cellular oxidative stress responses to styrene or its metabolites. 

o the role of cellular antioxidant pools in modulating the toxic response. o the role of enzymatic detoxification pathways in modulating the toxic response. o the effect of modulation of bioactivation and detoxification pathways and antioxidant pools on the cytotoxic response in whole animals or specific organs. Provide a more complete review of data pertaining to immunity, such as the number of lymphocytes, the weight of lymphoid organs, the function of systemic and localized lymphoid organs, and effects on innate vs specific immunity after styrene exposure in experimental animals or humans. Properties  Add a reference in the substance profile for styrene for the substance profile for styrene- 7,8-oxide. Exposure  Consider comparing different types of industries in a quantitative manner, perhaps by magnitude of exposure.  Where information on smoking status is available, special attention should be paid to differences between smokers and nonsmokers in the risks posed by workplace exposure.  Identify time trends, if any.  Add a clear description of methods of exposure assessment to support the interpretation and clarification of the exposure component in the epidemiologic studies.  Divide the subsection “General population” into exposures from smoking, including second- hand exposure (that is, environmental tobacco smoke), and exposures from other sources.  Include information on whether there is a relationship between pack-years of smoking and styrene exposure and on the magnitudes of styrene exposure from tobacco-smoking vs exposure in various occupations.  Include a discussion of the effects of various other sources of potential exposure in the nonsmoking general population.  All possible units of exposure (μg/kg of body weight, μg/m3, μg/L, and ppm) are mentioned and clarity could be improved by standardizing the units of exposure. (Continued) 51 

52 TABLE 2-1 Continued Sections in the Substance Profile for Styrene Suggested Revisions  Make a point at the beginning of the subsection “General population” that nearly all members of the general population, not only those exposed in a workplace or through smoking, have detectable styrene in their biologic fluids (for example, blood and breast milk). Regulations and Guidelines  Provide more information on regulations and include references to the sources. 

Review of the Styrene Profile in the NTP 12th Report on Carcinogens 53  “There is sufficient evidence of carcinogenicity from studies in exper- imental animals” (NTP 2008). Literature published by June 10, 2011, provided sufficient evidence that “there is an increased incidence of . . . a combination of malignant and benign tumors” (NTP 2008) in experi- mental animals induced by styrene administered by multiple routes of exposure (inhalation and oral gavage). The most informative experi- mental animal studies that support that conclusion are studies in mice (NCI 1979; Cruzan et al. 2001).  “There is convincing relevant information that the agent acts through mechanisms indicating it would likely cause cancer in humans” (NTP 2008). Literature published by June 10, 2011, provided convincing evi- dence that genotoxicity is observed in cells from humans who were ex- posed to styrene. That evidence is derived from a large body of publica- tions. In addition, styrene-7,8-oxide “was listed in a previous Report on Carcinogens as . . . reasonably anticipated as a human carcinogen” (NTP 2008). Styrene-7,8-oxide, a compound that is structurally related to styrene, is a major metabolite of styrene in both experimental ani- mals and humans; it was first listed in the 10th RoC (NTP 2002) as rea- sonably anticipated to be a human carcinogen. REFERENCES AHRQ (Agency for Healthcare Research and Quality). 2011. The Effective Health Care Program Stakeholder Guide. Publication No. 11-EHC069-EF. Rockville, MD: Agency for Healthcare Research and Quality. Aranyi, C., W.J. O’Shea, J.A. Graham, and F.J. Miller. 1986. The effects of inhalation of organic chemical air contaminats on murine lung host defenses. Fundam. Appl. Toxicol. 6(4):713-720. Banks, P.M. 1992. Changes in diagnosis of non-Hodgkin's lymphomas over time. Cancer Res. 52(19 suppl.):5453s-5455s. Bennett, J.M., D. Catovsky, M.T. Daniel, G. Flandrin, D.A. Galton, H.R. Gralnick, and C. Sultan. 1982. Proposals for the classification of the myelodysplastic syndromes. Br. J. Haematol. 51(2):189-199. Bennett, J.M., D. Catovsky, M.T. Daniel, G. Flandrin, D.A. Galton, H.R. Gralnick, and C. Sultan. 1985. Proposed revised criteria for the classification of acute myeloid leukemia. A report of the French-American-British Cooperative Group. Ann. In- tern. Med. 103(4):620-625. Berthiaume, S., and K.L. Ring. 2006. Styrene. P. 694.3000A in Chemical Economics Handbook Menlo Park, CA: SRI Consulting. Biró, A., E. Pállinger, J. Major, M.G. Jakab, T. Klupp, A. Falus, and A. Tompa. 2002. Lymphocyte phenotype analysis and chromosome aberration frequency of workers occupationally exposed to styrene, benzene, polycyclic aromatic hydrocarbons or mixed solvents. Immunol. Lett. 81(2):133-140. Boverhof, D.R., S.M. Krieger, J.A. Hotchkiss, K.E. Stebbins, J. Thomas, and M.R. Woolhiser. 2013. Assessment of the immunotoxic potential of trichloroethylene and percholoroethylene in rats following inhalation exposure. J. Immunotoxicol. 10(3):311-320. 

54 Review of the Styrene Assessment in the NTP 12th Report on Carcinogens Bucher, J.R. 2013. Follow-up Questions Submitted by the NAS Committee on Review of the Formaldehyde Assessment in the NTP 12th RoC and the NAS Committee on Review of the Styrene Assessment in the NTP 12th RoC, April 2, 2013. Carlson, G. 2004. Influence of selected inhibitors on the metabolism of the styrene me- tabolite 4-vinylphenol in wild-type and CYP2E1 knockout mice. J. Toxicol. Envi- ron. Health A 67(12):905-909. Carlson, G.P. 2010a. Depletion by styrene of glutathione in plasma and bronchioalveolar lavage fluid of non-Swiss albino (NSA) mice. J. Toxicol. Environ. Health A 73(11):766-772. Carlson, G.P. 2010b. Metabolism and toxicity of styrene in microsomal epoxide hydro- lase-deficient mice. J. Toxicol. Environ. Health A 73(24):1689-1699. Carlson, G.P., M. Ullman, N.A. Mantick, and P.W. Snyder. 2002. 4-Vinylphenol-induced pneumotoxicity and hepatotoxicity in mice. Toxicol. Pathol. 30(5):565-569. Cocco, P., A. t'Mannetje, D. Fadda, M. Melis, N. Becker, N., S. de Sanjose, L. Foretova, J. Mareckova, A. Staines, S. Kleefeld, M. Maynadie, A. Nieters, P. Brennan, and P. Boffetta. 2010. Occupational exposure to solvents and risk of lymphoma sub- types: Results from the Epilymph case-control study. Occup. Environ. Med. 67(5):341-347. CRD (Centre for Reviews and Dissemination). 2009. Systematic Reviews: CRD’s Guid- ance for Undertaking Reviews in Health Care. York, UK: York Publishing Ser- vices, Ltd [online]. Available: http://www.york.ac.uk/inst/crd/pdf/Systematic_Rev iews.pdf [accessed Feb. 24, 2013] Cruzan, G., J.R. Cushman, L.S. Andrews, G.C. Granville, K.A. Johnson, C. Bevan, C.J. Hardy, D.W. Coombs, P.A. Mullins, and W.R. Brown. 2001. Chronic toxici- ty/oncogenicity study of styrene in CD-1 mice by inhalation exposure for 104 weeks. J. Appl. Toxicol. 21(3):185-198. Cruzan, G., G.P. Carlson, M. Turner, and W. Mellert. 2005. Ring-oxidized metabolites of styrene contribute to styrene-induced Clara-cell toxicity in mice. J. Toxicol. Envi- ron. Health A 68(3):229-237. Cruzan, G., J. Bus, M. Banton, R. Gingell, and G. Carlson. 2009. Mouse specific lung tumors from CYP2F2-mediated cytotoxic metabolism: An endpoint/toxic response where data from multiple chemicals converge to support a mode of action. Regul. Toxicol. Pharmacol. 55(2):205-218. Delzell, E., N. Sathiakumar, J. Graff, M. Macaluso, G. Maldonado, and R. Matthew. 2006. An Updated Study of Mortality Among North American Synthetic Rubber Industry Workers. HEI Research Report No. 132. Boston, MA: Health Effects Institute [online]. Available: http://pubs.healtheffects.org/getfile.php?u=366 [accessed Feb. 21, 2014] Festing, M.F., and D.G. Altman. 2002. Guidelines for design and statistical analysis of experiments using laboratory animals. ILAR J. 43(4):244-258. Harvilchuck, J.A., and G.P. Carlson. 2009. Effect of multiple doses of styrene and R- styrene oxide on CC10, bax, and bcl-2 expression in isolated Clara cells of CD-1 mice. Toxicology 259(3):149-152. Harvilchuck, J.A., R.J. Zurbrugg, and G.P. Carlson. 2008. CC10 mRNA and protein ex- pression in Clara cells of CD-1 mice following exposure to styrene or its metabo- lites styrene oxide or 4-vinylphenol. Toxicol. Lett. 183(1-3):28-35. Harvilchuck, J.A., X. Pu, J.E. Klaunig, and G.P. Carlson. 2009. Indicators of oxidative stress and apoptosis in mouse whole lung and Clara cells following exposure to styrene and its metabolites. Toxicology 264(3):171-178. 

Review of the Styrene Profile in the NTP 12th Report on Carcinogens 55 Haseman, J.K., J. Huff, and G.A. Boorman. 1984. Use of historical control data in carcinogenicity students in rodents. Toxicol. Pathol. 12(2):126-135. Higgins, J.P.T., and S. Green, eds. 2008. Chochrane Handbook for Systematic Reviews of Interventions. Chichester, UK: John Wiley & Sons. IOM (Institute of Medicine). 2011. Finding What Works in Health Care: Standards for Systematic Reviews. Washington, DC: The National Academies Press. Jensen, A.A., N.O. Breum, J. Bacher, and E. Lynge. 1990. Occupational exposures to styrene in Denmark 1955-88. Am. J. Ind. Med. 17(5):593-606. Karami, S., P. Boffetta, P. Brennan, P.A. Stewart, D. Zaridze, V. Matveev, V. Janout, H. Kollarova, V. Bencko, M. Navratilova, N. Szeszenia-Dabrowska, D. Mates, J.P. Gromiec, R. Sobotka, W.H. Chow, N. Rothman, and L.E. Moore. 2011. Renal cancer risk and occupational exposure to polycyclic aromatic hydrocarbons and plastics. J. Occup. Environ. Med. 53(2):218-223. Keenan, C., S. Elmore, S. Francke-Carroll, R. Kemp, R. Kerlin, S. Peddada, J. Pletcher, M. Rinke, S.P. Schmidt, I. Taylor, and D.C. Wolf. 2009. Best practices for use of historical control data of proliferative rodent lesions. Toxicol. Pathol. 37(5):679- 693. Kogevinas, M., G. Ferro, A. Andersen, T. Bellander, M. Biocca, O. Coggon, V. Gennaro, S. Hutchings, H. Kolstad, I. Lundberg, E. Lynge, T. Partanen, and R. Saracci. 1994. Cancer mortality in a historical cohort study of workers exposed to styrene. Scand. J. Work Environ. Health 20(4):251-261. Kolstad, H.A., E. Lynge, J. Olsen, and N. Breum. 1994. Incidence of lymphohematopoietic malignancies among styrene-exposed workers of the reinforced plastics industry. Scand. J. Work Environ. Health 20(4):272-278. Kolstad, H.A., K. Juel, J. Olsen, and E. Lynge. 1995. Exposure to styrene and chronic health effects: Mortality and incidence of solid cancers in the Danish reinforced plastics industry. Occup. Environ. Med. 52(5):320-327. Kousalova, L., E. Anzenbacherova, J. Baranova, P. Anzenbacher, I. Skoumalova, and J. Vondrakova. 2004. Presence of cytochrome P450 enzymes in human CD34+ haematopoietic progenitor cells. Gen. Physiol. Biophys. 23(2):251-257. Meszka-Jordan, A., R. Mahlapuu, U. Soomets, and G.P. Carlson. 2009. Oxidative stress due to (R)-styrene oxide exposure and the role of antioxidants in non-Swiss albino (NSA) mice. J. Toxicol. Environ. Health A 72(10):642-650. NCI (National Cancer Institute). 1979. Bioassay of Styrene for Possible Carcinogenicity. Technical Report No. 185. NIH 79-1741. National Cancer Institute, National Insti- tute of Health, Bethesda, MD [online]. Available: http://ntp.niehs.nih.gov/ntp/ht docs/LT_rpts/tr185.pdf [accessed Feb. 21, 2014]. NRC (National Research Council). 2011. Review of the Environmental Protection Agen- cy’s Draft IRIS Assessment of Formaldehyde. Washington, DC: The National Academies Press. NRC (National Research Council). 2014. Review of EPA’s Integrated Risk Information System (IRIS) Process. Washington, DC: The National Academies Press. NTP (National Toxicology Program). 2002. Report on Carcinogens, 10th Ed. U.S. De- partment of Health and Human Services, Public Health Service, National Toxicol- ogy Program, Research Triangle Park, NC. NTP (National Toxicology Program). 2008. Report on Carcinogens Background Docu- ment for Styrene, September 29, 2008. U.S. Department of Health and Human Services, Public Health Service, National Toxicology Program, Research Triangle Park, NC [online]. Available: http://ntp.niehs.nih.gov/NTP/roc/twelfth/2010/ FinalBDs/Styrene_Final_508.pdf [accessed Aug. 14, 2013]. 

56 Review of the Styrene Assessment in the NTP 12th Report on Carcinogens NTP (National Toxicology Program). 2011a. Styrene. Pp. 383-392 in Report on Carcino- gens, 12th Ed. U.S. Department of Health and Human Services, National Toxicol- ogy Program, Research Triangle Park, NC [online]. Available: http://ntp.niehs.nih. gov/ntp/roc/twelfth/profiles/Styrene.pdf [accessed Aug. 14, 2013]. NTP (National Toxicology Program). 2011b. Listing Criteria [online]. Available: http://n tp.niehs.nih.gov/?objectid=47B37760-F1F6-975E-7C15022B9C93B5A6 [accessed Feb. 24, 2014]. NTP (National Toxicology Program). 2011c. Styrene-7, 8-oxide. Pp. 383-392 in Report on Carcinogens, 12th Ed. U.S. Department of Health and Human Services, Nation- al Toxicology Program, Research Triangle Park, NC [online]. Available: http://ntp.niehs.nih.gov/ntp/roc/twelfth/profiles/StyreneOxide.pdf [accessed Aug. 14, 2013]. Phillips, D., S. Snedeker, S. Eustis, M. Stone, P. Infante, E.M. Ward, G. Matanoski, G.S. Yost, S.S. Que Hee, L. Zeise, and T.J. Smith. 2008. Part B – Recommendation for Listing Status for Styrene and Scientific Justification for the Recommendation. Styrene Expert Panel Report [online]. Available: http://ntp.niehs.nih.gov/NTP/ roc/twelfth/2008/ExpertPanelMtgs/Styrene_PanelReportPartB_508.pdf [accessed July 17, 2013]. Ponomarkov, V., and L. Tomatis. 1978. Effects of long-term oral administration of sty- rene to mice and rats. Scand. J. Work Environ. Health 4(suppl. 2):127-135. Rhomberg, L., J.E. Goodman, and R.L. Prueitt. 2013. The weight of evidence does not support the listing of styrene as “reasonably anticipated to be a human carcinogen” in NTP’s Twelfth Report on Carcinogens. Hum. Ecol. Risk Assess. 19(1):4-27. Ruder, A.M., E.M. Ward, M. Dong, A.H. Okun, K. Davis-King. 2004. Mortality patterns among workers exposed to styrene in the reinforced plastic boatbuilding industry: An update. Am. J. Ind. Med. 45(2):165-176. Siest, G., E. Jeannesson, J.B. Marteau, A. Samara, B. Marie, M. Pfister, and S. Visvikis- Siest. 2008. Transcription factor and drug-metabolizing enzyme gene expression in lymphocytes from healthy human subjects. Drug Metab. Dispos. 36(1):182-189. Veraldi, A. A.S. Costantini, V. Bolejack, K. Miligi, P. Vinesis, and H. van Loveren. 2006. Immunotoxic effects of chemicals: A matrix for occupational and environmental epidemiological studies. Am. J. Ind. Med. 49(12):1046-1055. Wong, O., L.S. Trent, and M.D. Whorton. 1994. An updated cohort mortality study of workers exposed to styrene in the reinforced plastics and composites industry. Occup. Environ. Med. 51(6):386-396. Weill, H., J. Salvaggio, A. Neilson, B. Butcher, and M. Ziskind. 1975. Respiratory effects in toluene diisocyanate manufactore: a multidisciplinary approach. Environ. Health Perspect. 11:101-108 .