18
Tetrakis(hydroxymethyl) Phosphonium Salts

THIS chapter reviews the physical and chemical properties, toxicokinetics, toxicological, epidemiological, and exposure data on tetrakis(hydroxymethyl) phosphonium chloride (THPC),1 which is one of the tetrakis(hydroxymethyl) phosphonium salts. Although other tetrakis(hydroxymethyl) phosphonium salts have been used as flame retardants, the subcommittee chose to focus its assessment on THPC because it has a large toxicology database and is the most toxic of the phosphate salts. The subcommittee used the toxicity and exposure information on THPC to characterize the health risk from exposure to THPC. The subcommittee also identified data gaps and recommended research relevant for determining the health risk from exposure to THPC.

1  

It is important to note that THPC is polymerized onto fabric in combination with amine compounds and may undergo chemical changes that alter its chemical properties and toxicity. It is also reasonable to assume that oxidized forms of THPC may be present in or on aged THPC-treated fabric. However, the chemical species of THPC and the average ratio of these compounds present in treated fabric are not known at this time.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 417
Toxicological Risks of Selected Flame-Retardant Chemicals 18 Tetrakis(hydroxymethyl) Phosphonium Salts THIS chapter reviews the physical and chemical properties, toxicokinetics, toxicological, epidemiological, and exposure data on tetrakis(hydroxymethyl) phosphonium chloride (THPC),1 which is one of the tetrakis(hydroxymethyl) phosphonium salts. Although other tetrakis(hydroxymethyl) phosphonium salts have been used as flame retardants, the subcommittee chose to focus its assessment on THPC because it has a large toxicology database and is the most toxic of the phosphate salts. The subcommittee used the toxicity and exposure information on THPC to characterize the health risk from exposure to THPC. The subcommittee also identified data gaps and recommended research relevant for determining the health risk from exposure to THPC. 1   It is important to note that THPC is polymerized onto fabric in combination with amine compounds and may undergo chemical changes that alter its chemical properties and toxicity. It is also reasonable to assume that oxidized forms of THPC may be present in or on aged THPC-treated fabric. However, the chemical species of THPC and the average ratio of these compounds present in treated fabric are not known at this time.

OCR for page 417
Toxicological Risks of Selected Flame-Retardant Chemicals PHYSICAL AND CHEMICAL PROPERTIES The physical and chemical properties of THPC are summarized in Table 18–1. THPC is a water-soluble tetrakis(hydroxymethyl) phosphonium salt (IARC 1990) that is a common ingredient in commercial flame-retardant (FR) formulations. TABLE 18–1 Physical and Chemical Properties for Tetrakis(hydroxymethyl) Phosphonium Chloride Properties Value Reference Chemical formula C4H12O4PCl HSDB 1999 Structure Loewengart and Van Duuren 1977 CAS Registry # 124–64–1 HSDB 1999 Synonyms THPC HSDB 1999 Molecular weight 190.58 HSDB 1999 Physical state Crystalline solid; sold as 80% aqueous solution IARC 1990; HSDB 1999 Color 80% aqueous solution is straw-colored or clear and colorless NTP 1991; Hazleton UK 1992 Solubility Soluble in water, methanol, ethanol; less than 1 mg/mL in DMSO, insoluble in ether, reaction with acetone NTP 1991 Vapor pressure 80% aqueous solution: 1.0 mm Hg at 25 °C NTP 1991 PH 1 for aqueous solution of unspecified concentration; 2 for 80% aqueous solution Hazleton UK 1991; Hazleton UK 1992 Melting point 154°C Grasseli and Ritchey 1975 Boiling point 80% aqueous solution: 118°C NTP 1991 Density (water=1) 80% aqueous solution: 1.322 g/cm3 at 17.8°C; 1.34 g/cm3 at 20°C NTP 1991; Hazleton UK 1992

OCR for page 417
Toxicological Risks of Selected Flame-Retardant Chemicals OCCURRENCE AND USE THPC is produced by the reaction of formaldehyde, phosphine, and hydrochloric acid (IARC 1990). THPC and its sulfur salt THPS are the predominant FR chemicals used for cotton apparel, especially children’s sleepwear (NTP 1987). THPC was the most widely used FR for cotton fabrics in the 1950s. About 1,000–5,000 tons of THPC was used in the United States in 1987 (NTP 1987). TOXICOKINETICS Absorption Dermal No studies were identified that investigated the dermal absorption of THPC or other tetrakis(hydroxymethyl) phosphonium salts by humans. Dermal application of THPC to rats resulted in body weight loss and death (Aoyama 1975), indicating that THPC is absorbed by this route. Ulsamer et al. (1980), citing a 1953 report by the Wisconsin Alumni Research Foundation, stated that THPC can be absorbed through the skin in large amounts (1.5 gm/kg). It is not clear whether this is a derived amount or one based on animal data. The subcommittee could not locate a copy of the 1953 report. Inhalation No studies were identified that investigated the absorption of THPC following inhalation exposures. Oral Acute, subchronic, and chronic toxicity studies in rats and mice provide indirect evidence that THPC is absorbed through the gastrointestinal tract and becomes systemically bioavailable (see Hazard Identification section). Distribution No studies were identified that investigated the distribution of THPC or of other tetrakis(hydroxymethyl) phosphonium salts in humans or laboratory

OCR for page 417
Toxicological Risks of Selected Flame-Retardant Chemicals animals following dermal, inhalation, or oral exposure. Since the liver has been shown to be a target organ for THPC toxicity (see NTP 1987), it can be inferred that THPC is distributed systemically. Metabolism No studies were identified that investigated the metabolism of THPC or of other tetrakis(hydroxymethyl) phosphonium salts. Excretion No studies were identified that investigated the excretion of THPC or of other tetrakis(hydroxymethyl) phosphonium salts. HAZARD IDENTIFICATION2 Dermal Exposure Irritation No skin reactions were observed in 100 volunteers (23 males, 77 females) aged 9–63 yr who were exposed to THPC-treated fabric for 72 hr (Osbourn 1971). Volunteers were treated topically (location not specified) with THPC-treated fabric patches, some of which were moistened with distilled water. THPC was found to be non-irritating to the skin in 38 male volunteers who were dermally exposed to fabric patches containing Proban® 210, a THPC-based FR (Albright and Wilson 1982, as cited in IPCS 2000). The THPC content of the fabric patches was not reported. The fabric patches were applied to the forearms of the volunteers and covered for 48 hr. The test sites were then uncovered and examined for skin reactions 50 hr, 90 hr, 1 wk, and 2 wk after exposure. Moderate to severe skin reactions were observed in male white rats and rabbits treated topically for 8 d with 0.75 mL of 15%, 20%, or 30% aqueous THPC (Aoyama 1975). In rats, skin redness was observed starting on d 4 for 2   In this section, the subcommittee reviewed data on toxicity of tetrakis(hydroxymethyl) phosphonium salts, including the toxicity assessment prepared by the U.S. Consumer Product Safety Commission (Bittner 1999).

OCR for page 417
Toxicological Risks of Selected Flame-Retardant Chemicals rats treated with 15% and 20% THPC, while redness was observed beginning on d 2 in rats treated with 30% THPC. Skin redness in the 30% dose group became very intense on d 6 followed by partial hair loss on d 7 and death on d 9. Rabbits treated topically for 6 d with 1 mL of 15% or 30% aqueous THPC developed skin redness on d 2–3, which became severe by no later than d 6. Skin necrosis developed on d 3–12. Total hair loss occurred in both dose groups by d 11–13, but hair began to regrow in both dose groups by d 18–19. Toxicity was comparatively more severe in rabbits treated with 30% THPC. Ulsamer et al. (1980), citing the 1953 report by the Wisconsin Alumni Research Foundation, state that THPC is a mild skin irritant in the female rat and caused lethality, skin sloughing, and hyperemia after dermal application of >1.5 g/kg (species not given). The 1953 report was not located by the subcommittee. Sensitization There are reports of textile contact dermatitis occurring in children wearing nightdresses treated with THPC-based FRs (Martin-Scott 1966). A 2 1/2-yr-old girl developed pruritis, enlarged regional glands, and blisters after wearing THPC-treated sleepware over a 4-mo period. A 6-yr-old girl developed an urticarial rash reaction after wearing THPC-treated nightwear for a second time. Both girls developed skin reactions after exposure to patches treated with Proban. Symptoms of sensitization subsided in both girls after the THPC-treated dresses were no longer worn. Negative skin reactions were observed in 100 other children given patch tests with no known exposure to THPC and in children with atypical infantile eczema (Martin-Scott 1966). No skin sensitization reactions were observed upon re-challenge of 100 volunteers (23 males, 77 females) aged 9–63 yr exposed to THPC-treated fabric patches. Initially, the volunteers were exposed to the treated patches for 48 hr and the patches were then removed. Two wk after initial exposure, the volunteers were then re-challenged with THPC-treated fabric patches held in place for 72 hr. The patches were then removed and assessed for local skin reactions to the treatment. THPC was not a sensitizer in albino guinea pigs tested using the Buehler method (Industrial Bio-test Labs 1975). Ten animals were treated topically on their shaved backs with nine consecutive applications of patches containing 0.5 mL of a 1% v/v dilution of commercial THPC (THPC concentration not reported). Each patch was applied for 5 hr. The animals were then rechallenged with THPC-treated patches 2 wk after the induction period and the application sites were examined for skin reactions at 24 and 48 hr after re-challenge.

OCR for page 417
Toxicological Risks of Selected Flame-Retardant Chemicals Systemic Effects Moderate to severe skin effects characterized by tissue changes were observed in male white rats and rabbits treated topically for 8 d with 0.75 mL of 15%, 20%, or 30% aqueous THPC (Aoyama 1975). Treatment was intended to be for 20 d in both species but was discontinued after 8 d in rats because of severe weight loss and after 6 d in rabbits due to severe skin reactions. Rabbits continued to be observed until d 20. In rats, histological examination of the skin showed atrophy, enhanced keratinization of the epidermis, and degeneration of the hair roots in all treated animals. In rabbits, histological examination of the skin showed severe subepidermal fibrosis without regeneration of epidermis papillae. A high rate of deaths occurred in white mice treated repeatedly on their tails with aqueous extracts from fabric treated with a THPC-based FR (Afanas’eva and Evseenko 1971). Mice were treated with the extracts daily for 21 d. The extracts reportedly contained formaldehyde, hydrogen chloride, and organophosphorus compounds (not identified). The authors did not describe how the extracts were prepared, the sex of the animals used, or the number of animals tested. The authors reported that 50–70% of the treated animals died over the course of the experiment and exhibited weight loss and changes in the appearance of their fur. Tail-skin irritation was evident after 10–12 d of treatment and many of the tails fell off. Immunological Effects Afanas’eva and Evseenko (1971) reported that many (number not reported) of the mice treated with aqueous extracts from fabrics treated with a THPC-based FR developed leukopenia. The authors note that in addition to exposure to THPC, the extracts also contained formaldehyde, hydrogen chloride, and organophosphorus compounds and the exact composition of the extracts was not reported. Therefore, it is not possible to determine whether dermal exposure to THPC itself was the cause for the increased incidence of leukopenia in this study. Neurological Effects Mice treated with aqueous extracts from fabrics treated with a THPC-based FR became sluggish, had “reduced working capacity” for static work, and 20–40% lower cholinesterase activity levels (Afanas’eva and Evseenko 1971).

OCR for page 417
Toxicological Risks of Selected Flame-Retardant Chemicals Developmental Effects No studies were identified that investigated the toxic effects of THPC on reproduction or development following dermal exposure. Cancer Loewengart and Van Duuren (1977) found that THPC had tumor-promoting activity in a mouse skin carcinoma assay. Female ICR/Ha Swiss mice (20/group) were first topically treated once in a shaved area of the interscapular region with the tumor initiator 7, 12-dimethylbenz[a]anthracene (DMBA) and were then topically treated at the same site with THPC (2 mg in 0.1 mL dimethyl sulfoxide [DMSO]), three times a wk for 400 d. Three of 20 animals developed skin papillomas. All three tumors progressed to squamous cell carcinomas over the course of the assay. In the same experiment, one skin papilloma developed among 20 mice that received topical applications of THPC (2 mg in 0.1 mL of DMSO) 3 times/wk but without pretreatment with a tumor initiator. The authors did not comment on the significance of this finding, but the report notes that none of the 20 animals receiving no treatment with any compounds developed skin tumors over the 400 d experimental period. No tumors were observed in animals treated initially with extracts from THPC-treated cloth and then topically treated with the tumor promoter phorbol myristate acetate (PMA) (2.5 µg in 0.1 mL acetone), 3 times/wk for 400 d. Similarly, no tumors were observed in animals treated initially with THPC extracts followed by dermal applications of acetone (0.1 mL), 3 times/wk for 400 d. In comparison, 19 of 20 animals treated initially with a single application of DMBA followed by treatment with PMA 3 times/wk for 400 d produced papillomas in 19 of 20 animals and squamous cell tumors developed in 9 of 20 animals. No tumors were reported in animals treated initially with DMBA followed by DMSO treatment 3 times/wk for 400 d. The initiating effect of DMSO was investigated for comparison. The authors reported that the use of DMSO as a solvent for THPC produced results that were unusually low in all respects as compared with acetone. Van Duuren et al. (1978) found no evidence of carcinogenicity in female ICR/Ha Swiss mice treated topically with 2 mg THPC dissolved in 0.2 mL acetone/water, 3 times/wk for 496 d. All animals were killed at study termination. Skin, liver, and kidneys were examined microscopically for all animals that died prematurely and for 20% of the animals surviving to study completion. The incidence of tumors in THPC-treated mice was not significantly different from negative controls and animals treated with acetone.

OCR for page 417
Toxicological Risks of Selected Flame-Retardant Chemicals Inhalation Exposure No human or animal studies were identified that investigated the toxic effects of THPC following inhalation exposure. Oral Exposure Systemic Effects Acute oral toxicity data for THPC are summarized in Table 18–2. The single-dose LD50 for THPC ranges between 161–282 mg/kg in rats and 280–600 mg/kg in mice. Administration of a single, oral dose of THPC to rats, once a day, for 14 consecutive days resulted in decreased body weight gain at dose levels ≥18.8 mg/kg and changes in physical appearance and death at dose levels of ≥75 mg/kg. Administration of a single dose of THPC to mice, once a day, for 14 consecutive days resulted in decreased body weight gain at dose levels ≥18.8 mg/kg and all animals died when treated with 300 mg/kg. Ishizu (1975) reported dose-related increases in serum GOT and GPT activities and histological changes in the liver in both males and females of mice and rats given THPC in their drinking water for 1 mo. The author did not report how often THPC solutions were replaced. Therefore, there is some uncertainty regarding the actual concentrations of THPC that animals were exposed to in this study. Apparently, THPC undergoes oxidation in water, particularly at low concentrations (Hazleton UK 1992). Thake et al. (1982) found that female F-344 rats gavaged with 80 mg THPC/kg-d for 45 or 90 d had decreased weight gain, lack of response to external stimuli, stiff gait, and paresis. Many of the rats also had abnormal neurobehavior and degeneration of the peripheral nerves (see Neurological Effects section). No treatment-related effects were reported among female rats similarly treated with 40 mg THPC/kg-d. These data were presented in an abstract and a follow-up study was not located in the searched literature. Table 18–3 summarizes the results of subchronic (13-wk) and chronic (103-wk) toxicity studies for THPC in rats and mice conducted by NTP (1987). In the 13-wk study, all high-dose (60 mg/kg-d) male rats and five of 10 female rats died over the course of the study due to THPC treatment-related effects. Two of 10 males and one of 10 female rats died in the 15 mg/kg-d group, apparently due to gavage errors. Liver vacuolization was apparent in males at dose levels≥7.5 mg/kg-d and in females≥15 mg/kg-d. Paresis and axonal degeneration was apparent at 60 mg/kg-d. In the 13-wk study in mice, deaths, reduced body weights, paresis, and axonal degeneration occurred in both sexes

OCR for page 417
Toxicological Risks of Selected Flame-Retardant Chemicals TABLE 18–2 Short-Term Oral Toxicity Studies on Tetrakis(hydroxymethyl) Phosphonium Chloride Species, Sex, Number Duration, Route Dose (mg/kg-d) Effect Reference Rats, NS, F, NS 1 dose, gavage NS “Oral toxicity” of 250 mg/kg Anonymous 1953a, as cited in Ulsamer et al. 1980 Rat, F344, F, NS 40 (N) 80 (L) 45 or 90 d 45 or 90 d No effect Decreased weight gain, neurotoxicity (behavioral, histological) Thake et al. 1982b Rat, F344, M/F, 5/sex/dose 1 dose, gavage (2-wk observation) 75–1,200 LD50 of 185 mg/kg for M LD50 of 161 mg/kg for F NTP 1987 Mouse, B6C3F1, M/F, 5/sex/dose 1 dose, gavage (2-wk observation) 75–1,200 LD50 between 300–600 mg/kg for M LD50 of 280 mg/kg for F NTP 1987 Rat, F344 rat, M/F, 5/sex/dose 14 d, gavage (1-wk observation) 9.4 18.8 37.5 75 150 No effect Decreased body weight gain: 6% for M Decreased body weight gain: 11% for M Rough coats, arched backs, 2 of 5 M died Yellow/tan or mottled red livers NTP 1987 Mouse, B6C3F1, M/F, 5/sex/dose 14 d, gavage (1-wk observation) 18.8 37.5 75 150 300 Decreased body weight gain: 3% for M, 7% for F Decreased body weight gain: 6% for M, 7% for F Decreased body weight gain: 6% for M, 9% for F 3–6% body weight loss (18–20% decreased body weight) All animals died NTP 1987 Rat and Mouse 1 mo, drinking water 20–200 ppmc Liver histological and enzyme changes; decreased body weight gain at≥20 ppm, dose-related in both species Ishizu 1975b F, female; M, male; N, no-observed-adverse-effect level identified by reviewer; L, lowest-observed-adverse-effect level identified by reviewer. aUnpublished, non-Good Laboratory Practice study. bPublished, non-Good Laboratory Practice study. cEstimated dose for rats is 2.8–28 mg/kg-d (assume water intake of 0.049 L/d and body weight of 0.35 kg), and estimated dose for mice is 3.8–19 mg/kg-d (assume water intake of 0.0057 L/d and body weight of 0.03 kg).

OCR for page 417
Toxicological Risks of Selected Flame-Retardant Chemicals TABLE 18–3 Subchronic and Chronic Gavage Studies on Tetrakis(hydroxymethyl) Phosphonium Chloride Conducted by NTP (1987) Species, Strain, Sex, Number Duration Dose (mg/kg-d) Effect Rat, F 344, M/F, 10/sex/dose 5 d/wk, 13 wk 3.75 (N) 7.5 (L) 15 30 60 No effects at 3.75 mg/kg-d; Liver vacuolization1 in 8 of 10 M at 7.5 mg/kg-d; Liver vacuolization in 9 of 10 M, 8 of 10 F; liver necrosis2 of minimal severity in 9 of 10 M, 7 of 10 F; Liver vacuolization, necrosis in 10 of 10 M, F; decreased body weight gain in M; Most died; paresis, incoordination, axonal degeneration; liver vacuolization 10/10 M, F, liver necrosis in 7 of 10 M, 8 of 10 F Mouse, B6C3F1, M/F, 10/sex/dose 5 d/wk, 13 wk 1.5 4.5 (N) 15 (L) 45 135 No effects No effects Liver vacuolization in 10/10 M, F Liver vacuolization in 10/10 M, F; necrosis in 10/10 M Most died, decreased weight gain; paresis, incoordination, axonal degeneration; liver vacuolization in 10/10 M, 9/10 F; liver necrosis in 8/10 M, 7/10 F Rat, F344, M/F, 50/sex/dose 5 d/wk, 103 wk 3.75 (L) 7.5 Rough hair coats, diarrhea; liver vacuolization in 9/50 M (c=0/50), 11/50 F (c=3/50); liver cystic degeneration in 23/50 M (c=12/50); spleen hematopoiesis in 9/50 F (c=3/50) Rough hair coats, diarrhea; F had increased mortality after wk 70; liver vacuolization in 23/49 M (c= 0/50), 25/50 F (c=3/50); liver cystic degeneration in 26/50 M (c=12/50); spleen hematopoiesis in 15/50 F (c=3/50) Mouse, B6C3F1, M, 49 or 50/dose 5 d/wk, 103 wk 7.5 (L) 15 Rough hair coats; diarrhea; liver vacuolization in 39/49 M (c=0/49) Rough hair coats; diarrhea; liver vacuolization in 44/50 M Mouse, B6C3F1, F, 50/dose   15 30 Rough hair coats; diarrhea; liver vacuolization in 42/50 F (c=0/49) Rough hair coats; diarrhea; liver vacuolization in 48/50 F; thyroid hyperplasia 1Liver vacuolization: hepatocellular periportal cytoplasmic vacuolization. 2Liver necrosis: periportal hepatocellular necrosis. c, control group; F, female; M, male; N, no-observed-adverse-effect level identified by reviewer; L, lowest-observed-adverse-effect level identified by reviewer. at 135 mg/kg-d. Hepatocyte periportal cytoplasmic vacuolization occurred in males and females at≥15 mg/kg-d.

OCR for page 417
Toxicological Risks of Selected Flame-Retardant Chemicals In the 2-yr study, high-dose female rats had lower survival than controls starting at wk 70 (21/50 vs. 37/50 for controls). No effects on survival were apparent in male rats or male or female mice. Body weights in both species were not affected by treatment with THPC. Clinical toxicity was apparent at all dose levels in both species characterized by rough hair coats and diarrhea. Dose-related increases in hepatocyte cytoplasmic vacuolization occurred in the treatment groups of both species. In rats, cystic degeneration of the liver and spleen hematopoiesis was observed in male and females, respectively. Increased incidence of thyroid follicular cell hyperplasia occurred in female mice at dose level of 30 mg THPC/kg-d, but were not considered treatment-related because it is a common degenerative lesion in aging rodents. No neuropathological findings were apparent in either species. LOAELs for liver lesions in this study were 3.75 mg/kg-d for rats and 7.5 mg/kg-d for mice. However, hepatocellular vacuolization can be a reversible lesion (Harada et al. 1999). Progression of the liver lesions in the NTP (1987) study to fibrosis or hyperplasia, which occurs with severe and persistent vacuolization, was not reported. Immunological Effects No studies were identified that investigated the immunological effects of THPC following oral exposure. Neurological Effects It was reported that female rats given 80 mg THPC/kg-d by gavage developed peripheral nerve (sciatic, tibial, muscle branches, and plantar) degeneration that was characterized as mild after 45 d of exposure and mild to severe after 90 d of exposure (Thake et al. 1982). These rats also had reduced spontaneous motor activity and reduced forelimb and hindlimb grip strength. No neurological effects were observed in females given 40 mg THPC/kg-d for 45 or 90 d. The THPC solvent and the number of animals tested were not specified. Neurotoxic effects were seen in both rats and mice in the 13-wk studies conducted by NTP (1987). High-dose animals (60 mg/kg-d for rats and 135 mg/kg-d for mice) had paresis and incoordination of the rear limbs as well as rough hair coats, hunched backs, diarrhea, lethargy, and axonal degeneration. Axonal degeneration was seen in 2 of 10 high-dose female rats and in at least 19 of 20 high-dose mice and was characterized by swollen axon sheaths, missing or fragmented axons, and proliferation of neurolemma cells in the sciatic nerve, dorsal roots of the caudal spinal nerves, and spinal cord tracts.

OCR for page 417
Toxicological Risks of Selected Flame-Retardant Chemicals In the full-scale study, maternal toxicity was apparent in dams treated with ≥6 mg/kg-d (Hazleton UK 1992). No maternal or fetal toxicity occurred at 2 mg/kg-d. There were no differences in uterine implantation rates, fetal weights, or sex ratio as compared with controls in any dose groups with the exception of one total resorption occurring in the 36 mg/kg-d dose group. The incidences of fetal malformations and external and visceral variations were not different in treated versus control offspring. Fewer of the offspring from treated animals had incomplete ossification of the 5th or 6th sternebra (p < 0.01) while a greater number of high-dose fetuses had supernumerary thoraco-lumbar ribs (83.1% vs. 62.6% for controls). The incidence of lowering of the pelvic girdle attachment position was significantly elevated in the offspring from high-dose dams (46.1% vs. 13.0% for controls; p<0.01). The incidence of this malformation was also elevated in offspring from dams treated with 6 or 18 mg/kg-d (37.0% and 37.1%, respectively), but this was statistically significant only in the latter group (p<0.05). Cancer No evidence of carcinogenicity was reported for rats and mice given THPC by gavage 5 d/wk for 103 wk (NTP 1987) (see Table 18–3). The incidence of mononuclear cell leukemia was increased in low-dose male rats (19/50, 25/50, and 16/50 for control, low, and high doses). However, this increase was not considered treatment-related because of a lack of dose-response and was statistically significant only in the life-table analysis. Animals were tested at or near the maximum tolerated dose (MTD) as evidenced by increased deaths in high-dose females and liver toxicity at all dose levels in both species. Genotoxicity Negative results were obtained when testing aqueous THPC in the Salmonella mutagenicity test using strains TA98 and TA100 with metabolic activation (Kawachi et al. 1980), or strains TA98, TA100, TA1535, and TA1537, with or without exogenous activation (MacGregor et al. 1980; Zeiger et al. 1987). The eluent (0.1 mL) of THPC-treated fabric samples incubated in physiological saline at 37°C for 18–24 hr (5 g fabric in 30 mL saline) also yielded negative results in Salmonella strains TA98, TA100, TA1537, and 1538, with or without exogenous metabolic activation (Huntingdon Research 1976). Mutations were not induced in silkworms treated with THPC, but no experimental details were provided (Kawachi et al. 1980). THPC-induced mutations in mouse lymphoma cells without the use of exogenous metabolic activation

OCR for page 417
Toxicological Risks of Selected Flame-Retardant Chemicals (Myhr et al. 1990). DMSO extracts of THPC-treated cotton fabrics were mutagenic to V79 hamster lung cells and also transformed BHK cells and BALB/3T3 cells, with and without metabolic activation (Ehrlich et al. 1980). Pure DMSO was not mutagenic when similarly tested. No difference in results were observed between dyed and undyed THPS-treated fabrics. Matthews et al. (1993) obtained an equivocal response in three transformation assay with A-31-1–13 BALB/C-3T3 cells (a positive response was obtained in one and a negative response in two). THPC was positive for genotoxicity in the Bacillus subtilis rec assay, with or without metabolic activation (Kawachi et al. 1980). THPC induced chromosome aberrations in CHO cells without metabolic activation using either a conventional or delayed harvest protocol (18 hr instead of 10 hr), and with metabolic activation using the conventional harvest protocol (Loveday et al. 1989). Equivocal results were obtained for chromosome aberrations in hamster lung fibroblast cells in vitro without metabolic activation, and for rat bone marrow chromosome aberrations in vivo (Kawachi et al. 1980). THPC induced SCEs in cultured CHO cells with or without metabolic activation (Loveday et al. 1989). In vitro, THPC has been shown to form a stable adduct with the 2-amino group of guanosine (Van Duuren et al. 1978). QUANTITATIVE TOXICITY ASSESSMENT It is important to note that toxicity assessment values developed for THPC (such as RfDs) may not be a direct measure of the toxicity of THPC-treated fabric in a “real-world” exposure situation. This is because it is unknown whether any free THPC is present after polymerization onto the fabric. Information on the types, levels, and toxicities of THPC derivatives formed during the polymerization process is not available. An evaluation of THPC toxicity, however, is necessary as it provides a knowledge base for assessing the toxicological risks associated with this class of flame-retardant chemicals. Noncancer Dermal Assessment There are no adequate dermal toxicity studies for deriving a dermal RfD for THPC. Long-term dermal studies conducted by Loewengart and Van Duuren (1977) and Van Duuren et al. (1978) only reported on the tumor findings, and therefore, could not be used for RfD development. Studies by Aoyama (1975) are of inadequate duration (6 d in rabbits and 8 d in rats).

OCR for page 417
Toxicological Risks of Selected Flame-Retardant Chemicals Inhalation RfC No studies were identified that investigated the noncancer effects of THPC inhalation, therefore an inhalation reference concentration (RfC) could not be determined. Since THPC reacts (i.e., polymerized) directly with the fabric, it is expected that human inhalation exposure would be primarily to polymer derived from THPC attached to microscopic fibers of fabric. However, the subcommittee could not locate data that verify this hypothesis or toxicity data for THPC-treated fibers. In the absence of relevant inhalation exposure data, the subcommittee chose to extrapolate inhalation RfCs from oral RfDs The subcommittee, however, recognizes that it is not an ideal approach and also recognizes that the estimated RfC levels might be considerably different than actual levels (if inhalation data were available). Extrapolating from one route of exposure (oral) to another (inhalation) requires specific knowledge about the uptake kinetics into the body by each exposure route, including potential binding to cellular sites. The subcommittee believes that its extrapolation of oral RfDs to inhalation RfCs is highly conservative; it assumes that all of the inhaled compound is deposited in the respiratory tract and completely absorbed into the blood. The NRC Committee on Toxicology (NRC 1985) has used this approach when inhalation exposure data were insufficient to derive inhalation exposure levels. The subcommittee believes that such an approach is justified for conservatively estimating the toxicological risk from exposure to THPC, and the RfC should be used as an interim or provisional level until relevant data become available for the derivation of an inhalation RfC. For the purpose of estimating a hazard index for the inhalation of THPC, a provisional inhalation RfC of 0.0105 mg/m3 was derived from the oral RfD for THPC using Equation 7 in Chapter 3. Oral RfD Of the available THPC studies, the most appropriate one to use for RfD derivation is the NTP (1987) 2-yr bioassay for THPC (see Table 18–3). Rats were dosed with 0, 3.75, or 7.5 mg/kg-d; male mice with 0, 7.5, or 15 mg/kg-d; and female mice with 0, 15, or 30 mg/kg-d (103 wk, 5 d/wk). In the NTP (1987) study, the liver appears to be the critical target organ in both species. A dose-related increase in periportal hepatocyte cytoplasmic vacuolization occurred in both sexes of rats and mice, and was statistically significant at all dose levels tested. A clear dose-response was seen for the

OCR for page 417
Toxicological Risks of Selected Flame-Retardant Chemicals incidence of periportal hepatocyte cytoplasmic vacuolization in male and female rats. Similar results were observed in mice, but the increase in incidence was not linear. Liver toxicity occurred at the lowest dose tested in rats at 3.75 mg/kg-d which equates to 2.68 mg/kg-d when adjusted for discontinuous exposure (i.e., multiplied by 5/7). The LOAEL of 2.68 mg/kg-d for liver toxicity in the rat was divided by the composite uncertainty factor (UF) of 1000 to yield an oral RfD of 0.003 mg/kg-d (see Table 18–5). The composite uncertainty factor was comprised of the following uncertainty factors: a factor of 3 (UFL) was applied for LOAEL to NOAEL extrapolation. This uncertainty factor was reduced from the default of 10 to 3 because a minimal LOAEL was established due to the low frequency (18% in males, 22% in females) and low severity of the critical response (not life-threatening, possibly reversible). A factor of 3 (UFD) was used for less than complete data (i.e., the toxicity database includes a chronic study in two species and a developmental toxicity study). A factor of 10 (UFH) was used to account for intraspecies differences, and a factor of 10 (UFA) was used for interspecies differences. Cancer IARC (1990) has determined that the tetrakis(hydroxymethyl) phosphonium salts, including THPC, are not classifiable as to their carcinogenicity to humans (Group 3), based on inadequate (i.e., lack of) evidence of carcinogenicity in animals and no data from human studies. TABLE 18–5 Oral Reference Dose for Tetrakis(hydroxymethyl) Phosphonium Chloride Critical Effect Species Effect Level (mg/kg-d) Uncertainty Factors RfD (mg/kg-d) Reference Liver toxicity Rat LOAEL: 2.68 UFA: 10 UFH: 10 UFL: 3 UFD: 3 Total: 900 0.003 NTP (1987) LOAEL, lowest-observed-adverse-effect level; RfD, reference dose; UFA, interspecies variability; UFH, intraspecies variability; UFL, NOAEL (no-observed-adverse-effect level) for critical effect not determined; UFD, inadequate or deficient toxicity database.

OCR for page 417
Toxicological Risks of Selected Flame-Retardant Chemicals Dermal Treatment-induced neoplasia was not found in the 496-d dermal carcinogenicity study conducted in female mice by Van Duuren et al. (1978). The results of an earlier dermal carcinogenicity study by the same group, in which a low incidence of skin papilloma was found, were compromised by solvent effects (Loewengart and Van Duuren 1977). The subcommittee concluded that data are inadequate to determine the carcinogenicity of THPC. Inhalation No studies were identified that investigated the carcinogenic effects of THPC following inhalation exposure. The subcommittee concluded that data are inadequate to assess the carcinogenicity of THPC via the inhalation route. Oral No evidence of carcinogenicity of THPC was found in rats or mice given THPC by gavage for 103 wk (NTP 1987). The subcommittee concluded that THPC is not likely to be carcinogenic via the oral route of exposure. EXPOSURE ASSESSMENT AND RISK CHARACTERIZATION This assessment assumes that exposure is to unreacted THPC and does not assess exposure to chemically altered forms of THPC that may form during the application of THPC to upholstery fabrics. Noncancer Dermal Exposure Dermal exposure to THPC was estimated using the dermal exposure scenario described in Chapter 3. This exposure scenario assumes that an adult spends 1/4th of his or her time sitting on furniture upholstery treated with commercial THPC. It also assumes that 1/4th of the upper torso is in contact with the upholstery and that clothing presents no barrier.

OCR for page 417
Toxicological Risks of Selected Flame-Retardant Chemicals The subcommittee concluded that THPC is an ionic substance and essentially not absorbed through the skin. In addition, THPC is likely to polymerize after application to the upholstery fabric; therefore, exposure to THPC is not likely to occur and it should not pose a toxic hazard by the dermal route of exposure when used as an FR in furniture upholstery. Inhalation Exposure Particles Inhalation exposure estimates for THPC were calculated using the exposure scenario described in Chapter 3. This scenario assumes that a person spends 1/4th of his or her lifetime in a 30 m3 room containing 30 m2 of THPC-treated fabric and the room is assumed to have an air-change rate of 0.25/hr. It is also assumed that 50% of the THPC present in 25% of the surface area of the treated fabric is released over 15 yr and 1% of released particles are small enough to be inhaled. Particle exposure was estimated using Equations 4 and 5 in Chapter 3. The subcommittee estimated an upholstery application rate (Sa) for THPC of 4.5 mg/cm2. The release rate (µr) for THPC from upholstery fabric was estimated to be 2.3×10−7/d (see Chapter 3, Equation 5) yielding a room airborne particle concentration (Cp) of 1.7 µg/m3 and a short time-average exposure concentration of 0.43 µg/m3. The time-averaged exposure concentration for particles was calculated using Equation 6 in Chapter 3. Division of the time-average exposure concentration of 0.43 µg/m3 by the provisional RfC for THPC of 0.0105 mg/m3 gives a hazard index of 4.1×10−2. These findings suggest that under this worst-case exposure scenario, inhalation of THPC particles from furniture upholstery is not likely to pose a noncancer toxicological risk to humans. Vapors In addition to the possibility of release of THPC in particles from worn upholstery fabric, the subcommittee considered the possibility of the release of THPC by evaporation. This approach is described in Chapter 3, and uses an exposure scenario similar to that described above for exposure to THPC particles. The rate of flow of THPC vapor from the room is calculated using Equations 8–11 in Chapter 3. A saturated vapor concentration (Cν) of 10,300 mg/m3 was estimated for THPC. The application density (Sa) for THPC in the treated upholstery was estimated as 4.5 mg/cm2.

OCR for page 417
Toxicological Risks of Selected Flame-Retardant Chemicals Using the parameters described, the equilibrium room-air concentration of THPC was estimated to be 8,700 mg/m3. The short-term time-average exposure concentration for THPC was estimated as 2,175 mg/m3 using Equation 12 in Chapter 3 and the equilibrium room-air concentration for THPC. It was also estimated that this air concentration could be maintained for approximately 16 hr. Division of the exposure estimate of 2,175 mg/m3 by the provisional inhalation RfC of 0.0105 mg/m3 yields a hazard index of 2.2×105. These results indicate that if all of the THPC is released from the fabric into the air, THPC could be a toxic risk to persons entering the room. In reality, any FR that evaporated so rapidly would be useless in preventing upholstery flammability. Either THPC is much more strongly bound to the fabric than assumed in this scenario (so that the parameter γ in the analysis above is substantially less than unity), or the chemical is transformed during the application process. In either case, the emission rate would likely be controlled by some process other than diffusion through a boundary layer of air, as assumed here. The subcommittee believes that this exposure scenario provides no useful information about the potential toxicity of THPC vapors to humans associated with the emission of THPC vapors from treated furniture upholstery. Therefore, further investigation should be carried out to determine if exposure to THPC by this route poses a toxic risk to humans. These results suggest that the vapor inhalation scenario is unrealistic for THPC-treated furniture in a residential setting because evaporative loss of all THPC over 16 hr could not occur under normal conditions. Since THPC is chemically cross-linked within the treated upholstery, the vapor pressure of THPC is assumed to be the vapor pressure for the polymerized form. In the absence of any published data, it is assumed that the polymerized form of THPC will have a vapor pressure approaching zero. Thus, the vapor inhalation noncancer risk from THPC-treated fabric may be assumed to be minimal if not zero. Oral Exposure The assessment of noncancer toxicological risk for oral exposure to THPC is based on the oral exposure scenario described in Chapter 3. This scenario assumes a child is exposed to THPC by sucking on 50 cm2 of fabric treated with THPC, 1 hr/d for two yr. The subcommittee estimated an upholstery application rate (Sa) for THPC of 4.5 mg/cm2. Oral exposure was calculated using Equation 15 in Chapter 3. The extraction rate (µw) for THPC was estimated to be 0.001 based on laundry data (Horrocks et al. 1992). The worst-case, average oral daily dose for THPC was estimated as 0.00094 mg/kg-d. Division of the oral daily dose estimate of 0.00094 mg/kg by the oral

OCR for page 417
Toxicological Risks of Selected Flame-Retardant Chemicals RfD for THPC of 0.003 mg/kg-d results in a hazard index of 0.313 and indicates that oral exposure to THPC is not likely to be a toxic risk under the given exposure scenario and conditions. Cancer Dermal The evidence for the dermal carcinogenicity of THPC is equivocal. Van Duuren et al. (1978) found no evidence for the dermal carcinogenicity of THPC in mice. However, Loewengart and Van Duuren (1977) found equivocal evidence for the carcinogenicity of THPC when administered in combination with a known tumor promoter or initiator. The subcommittee concluded that data are inadequate to determine human carcinogenic potential by the dermal route. Inhalation No adequate data are available in humans or laboratory animals to assess the carcinogenicity of THPC vapors or particles containing THPC. Oral No evidence of carcinogenicity was found in rats or mice following chronic oral administration of THPC (see Systemic Effects section). The subcommittee concluded that THPC, used as an FR in upholstery fabric, is not likely to pose a cancer risk by the oral exposure route. RECOMMENDATIONS FROM OTHER ORGANIZATIONS Regulatory standards or guidelines have not been established for THPC or any of the other tetrakis(hydroxymethyl) phosphonium salts. DATA GAPS AND RESEARCH NEEDS Key information is needed on the types, amounts (including ratios), and toxicity of THPC derivatives present in THPC-treated cloth. It is important to

OCR for page 417
Toxicological Risks of Selected Flame-Retardant Chemicals note that THPC polymerizes within the fiber and fabric structure and may also react with other FR-formulation components present, so it might undergo other chemical changes that would alter its chemical properties and toxicity. It is also highly likely that oxidized forms of THPC will be present in or on the aged THPC FR-treated fabric. The subcommittee recommends that research be conducted to determine whether new chemical species are formed, and if so, to identify those chemical species. Hazard indices for the inhalation of particles and oral exposure to THPC are less than one and therefore, these route of exposure are not anticipated to be a concern. Dermal exposure to THPC through contact with treated material is not expected to occur since THPC is chemically bound to the fabric. REFERENCES Afanas’eva, L.V., and N.S.Evseenko. 1971. Hygienic assessment of fireproof fabrics treated with an organophosphorus impregnating compound based on tetrahydroxymethylphosphonium chloride. Gig. Sanit. 36(3): 102–103. Albright and Wilson (Albright and Wilson, Ltd.). 1982. Study to Determine the Primary Irritancy of Untreated Fabric, Proban 210 Treated Fabric, Proban NX Treated Fabric, Laundered Proban NX Treated Fabric in Human Volunteers. J.R.Jackson, study director for Albright and Wilson Ltd. Study dated November 30, 1982. In: Product Summary: Proban® Tetrakis (hydroxyethyl)phosphonium compound with urea. Albright and Wilson Americas, Inc. Glen Allen, VA. Albright and Wilson (Albright and Wilson, Ltd.). 1989. Composition and Chemistry of Tetrakishydroxymethylphosphonium Salts. TSCATS Fiche# OTS0519137, Doc# 89–890000200. Albright & Wilson, Ltd., West Midlands, UK. Anonymous. 1953. Wisconsin Alumni Research Foundation Report to Oldbury Electrical Co. Aoyama, M. 1975. Effect of anti-flame treating agents on the skin. Nagoya Med. J. 20(1):11–19. Bittner, P. 1999. Toxicity Review for Tetrakis(hydroxymethyl) Phosphonium Salts Precondensate with Urea. Memorandum, dated March 6, 1999, from Patricia Bittner, Toxicologist, Division of Health Sciences, to Ronald Medford, Assistant Executive Director for Hazard Identification and Reduction, U.S. Consumer Product Safety Commission, Washington, DC. Ehrlich, K., A.Hulett, and T.Turnham. 1980. Mammalian cell culture mutagenicity and carcinogenicity testing of dimethyl sulfoxide extracts of flame retardant-treated cotton fabrics. J. Toxicol. Environ. Health 6(2):259–271. EPA (U.S. Environmental Protection Agency). 1986. Guidelines for Carcinogen Risk Assessment. Fed. Regist. 51(185):33992–34003. EPA (U.S. Environmental Protection Agency). 1996. Proposed guidelines for carcinogen risk assessment. Fed. Regist. 61(Apr. 23):17960–18011.

OCR for page 417
Toxicological Risks of Selected Flame-Retardant Chemicals EPA (U.S. Environmental Protection Agency). 1999. Carcinogen Risk Assessment Guidelines. Draft for SAB discussion. [Online]. Available: http://www.epa.gov/nceawww1/raf/SABamtg.pd Grasseli, J.G., and W.M.Ritchey, eds. 1975. Pp. 120 in CRC Atlas of Spectral Data and Physical Constants for Organic Compounds, 2nd Ed., Vol. IV. Cleveland, OH: CRC Press. Harada, T., A.Enomoto, G.A.Boorman, and R.R.Maronpot. 1999. Liver and gallbladder. Pp. 131 in Pathology of the mouse. R.R.Maronpot, G.A.Boorman, and B.W. Gaul, eds. Vienna, IL: Cache River Press. Hazleton UK. 1991. THPC: Oral (Gavage) Range-finding Study in the Pregnant Rabbit. Report No. 6470–254/28. Hazelton UK, North Yorkshire, UK. Hazleton UK. 1992. THPC: Oral (Gavage) Teratology Study in the Rabbit. Report No. 6700–254/29. Hazelton UK, North Yorkshire, UK. Horrocks, A.R., J.Allen, S.Ojinnaka, and D.Price. 1992. Influence of laundering on durable flame retarded cotton fabrics. 1. Effect of oxidant concentration and detergent type. J. Fire Sci. 10(4):335–351. HSDB (Hazardous Substances Data Bank). 1999. Hazardous Substances Data Bank. MEDLARS Online Information Retrieval System, National Library of Medicine. [Online]. Available: http://sis.nlm.nih.gov/cgi-bin/sis/htmlgen?HSDB Retrieved August 16, 1999. Huntingdon Research (Huntingdon Research Center). 1976. Evaluation of Material Samples for Their Mutagenic Potential Utilizing the Ames Test Methodology. HRC Study #608675, dated August 27, 1976. In: Product Summary: Proban® Tetrakis (Hydroxyethyl)phosphonium Compound with Urea. Albright and Wilson Americas, Inc. Glen Allen, VA. IARC (International Agency for Research on Cancer). 1990. Tetrakis(hydroxymethyl) phosphonium salts. Pp. 95–107 in IARC Monographs on the Evaluation of Carcinogenic Risk of Chemicals to Humans, Vol. 48, Some Flame Retartdants and Textile Chemicals, and Exposures in the Textile Manufacturing Industry. Lyon, France: IARC Press. Industrial BIO-TEST Laboratories, Inc. 1975. Skin Sensitization Tests with Five Samples in Albino Guinea Pigs. IBT No. 601–06278. Doc# 878212099; Fiche# OTS0205956. IPCS (International Programme on Chemical Safety). 2000. Environmental Health Criteria 218, Flame Retardants: Tris(2-Butoxyethyl) Phosphate, Tris(2-Ethylhexyl) Phosphate and Tetrakis (Hydroxymethyl) Phosphonium Salts. Geneva: World Health Organization. Ishizu, S. 1975. Toxicity of organophosphorus fire retardants. Kobunshi 24:788–792. Kawachi, T., T.Komatsu, T.Kada, M.Ishidate, M.Sasaki, T.Sugiyama, and Y. Tazima. 1980. Results of recent studies on the relevance of various short-term screening tests in Japan, In: The Predictive Value of Short-term Screening Tests in Carcinogenicity Evaluation. Appl. Methods Oncol. 3:253–267. Loewengart, G., and B.L.Van Duuren. 1977. Evaluation of chemical flame retardants for carcinogenic potential. J. Toxicol. Environ. Health 2(3):539–546. Loveday, K.S., M.H.Lugo, M.A.Resnick, B.E.Anderson, and E.Zeiger. 1989.

OCR for page 417
Toxicological Risks of Selected Flame-Retardant Chemicals Chromosome aberration and sister chromatid exchange tests in Chinese hamster ovary cells in vitro: II. Results with 20 chemicals. Environ. Mol. Mutagen. 13(1):60–94. MacGregor, J.T., M.J.Diamond, L.W.Mazzeno, Jr., and M.Friedman. 1980. Mutagenicity tests of fabric-finishing agents in Salmonella typhimurium: Fiber-reactive wool dyes and cotton flame retardants. Environ. Mutagen. 2(3):405–418. Martin-Scott, I. 1966. Contact textile dermatitis (with special reference to fire proof fabrics). Br. J. Dermatol. 78(12):632–635. Matthews, E.J., J.W.Spalding, and R.W.Tennant. 1993. Transformation of BABL/c-3T3 cells: V. Transformation responses of 168 chemicals compared with mutagenicity in Salmonella and carcinogenicity in rodent bioassays. Environ. Health. Perspect. Suppl. 101(Suppl. 2):347–482. Myhr, B., D.McGregor, L.Bowers, C.Riach, A.G.Brown, I.Edwards, D.McBride, R.Martin, and W.J.Caspary. 1990. L5178Y mouse lymphoma cell mutation assay results with 41 compounds. Environ. Mol. Mutagen. 16(Suppl. 18):138–167. NTP (National Toxicology Program). 1987. Toxicology and Carcinogenesis of Tetrakis(hydroxymethyl) Phosphonium Sulfate (THPS) (CAS No. 5566–30–8) and Tetrakis(hydroxymethyl) Phosphonium Chloride (TMPC) (CAS No. 124–64–1) in F344/N Rats and B6C3F1 Mice (Gavage Studies). NTP Technical Report No. 296. NIH Publication No. 87–2552. NTP (National Toxicology Program). 1991. Tetrakis(hydroxymethyl) Phosphonium Chloride. NTP Chemical Health and Safety Data. [Online]. Available: http://ntp-server.niehs.nih.gov/cgi/iH_Indexes/Chem_H&S/iH_Chem_H&S_Frames.html retrieved 8/99. Osbourn, R.A. 1971. Report of Human Patch Tests on Substance: THPC® Safety Finishes, Cotton Knit Finish No. THPOH-NH3. Hazelton Laboratories, Inc. Project No. 157–188. TSCATS Doc# 88–920001111; Fiche# OTS0537009. Thake, D., P.Kurtz, and S.Carter. 1982. Neurotoxic evaluation of tetrakis hydroxymethyl phosphonium chloride in Fischer 344 rats. [Abstract No. 407]. Toxicologist 2(1):115. Ulsamer, A.G., R.E.Osterberg, and J.McLaughlin, Jr. 1980. Flame-retardant chemicals in textiles. Clin. Toxicol. 17(1):101–131. Van Duuren, B.L., G.Loewengart, I.Seidman, A.C.Smith, and S.Melchionne. 1978. Mouse skin carcinogenicity tests of the flame retardants tris(2, 3-dibromopropyl)-phosphate, tetrakis(hydroxymethyl)phosphonium chloride, and polyvinyl bromide. Cancer Res. 38(10):3236–3240. Zeiger, E., B.Anderson, S.Haworth, T.Lawlor, K.Mortelmans, and W.Speck. 1987. Salmonella mutagenicity tests: III. Results from the testing of 255 chemicals. Environ. Mutagen. 9 (Suppl. 9): 1–109.