Chlorinated paraffins: chapter 5

5. Human health risk assessment

5.1 Population exposure

The following presentation is limited to identified recent data considered critical to quantitative estimation of exposure of the general population in Canada to chlorinated paraffins and, hence, to assessment of "toxic" under Paragraph 64(c) of the Canadian Environmental Protection Act, 1999 (CEPA 1999). Other sources of data that were also identified but were not directly relevant to estimation of exposure in Canada include Peters et al. (2000), Borgen et al. (2000, 2002) and Lahaniatis et al. (2000).

The degree of confidence in data on the concentrations of chlorinated paraffins in various media varies considerably, depending upon the nature of the analysis. To the extent possible, estimates of intake have been based on higher-confidence analyses by high-resolution gas chromatography (HRGC)/electron capture negative ion high-resolution mass spectrometry (ECNI-HRMS), due to its higher mass resolving power and selectivity. However, such information is limited solely to determination of short-chain chlorinated paraffins (SCCPs) in human breast milk (Tomy, 1997), fish (Muir et al., 1999) and media that contribute less to human exposure, including ambient air (Tomy, 1997), surface water (Tomy, 1997) and sediment (Muir et al., 2001). For all chlorinated paraffins, either concentrations in surface water and sediment, or the limits of detection for these media, were used as surrogates for concentrations in drinking water and soil, respectively, in estimating intake.

Indeed, data on concentrations of chlorinated paraffins in foodstuffs are extremely limited. While additional data on the concentrations of SCCPs, medium-chain chlorinated paraffins (MCCPs) and long-chain chlorinated paraffins (LCCPs) in foods in the United Kingdom (Campbell and McConnell, 1980b) reported in an early investigation reviewed in the PSL1 assessment (Campbell and McConnell, 1980a) were acquired and are presented in Table 8, they are considered, at best, to be semi-quantitative, owing to limitations of the methodology available at that time. Analysis was based on liquid-solid adsorption chromatography, which has now largely been replaced by micro-analytical techniques and quantification by visual reference to spots appearing on thin-layer chromatographic plates.

Table 8: Concentrations of short-chain, medium-chain and long-chain chlorinated paraffins in foodstuffs
 Food group Short- and medium-chain chlorinated paraffins Long-chain chlorinated paraffins
Dairy

0.3 µg/g

mean of 13 samples of dairy products in U.K.
C10-20 (SCCPs and MCCPs)
(Campbell and McConnell, 1980a)

0.19 µg/g

1 sample of cheese in U.K.
C20-30
(Campbell and McConnell, 1980a)

Fats

0.15 µg/g

mean of 6 samples of vegetable oils and derivatives
C10-20 (SCCPs and MCCPs)
(Campbell and McConnell, 1980a)

0.05 µg/g

detection limit in analysis of 1 sample of lard in U.K.
C20-30
(Campbell and McConnell, 1980b)

Fruits

0.025 µg/g

mean of 16 samples of fruits and vegetables in U.K.
C10-20 (SCCPs and MCCPs)
(Campbell and McConnell, 1980a)

0.025 µg/g

1 sample of peach fruit in U.K.
C20-30
(Campbell and McConnell, 1980a)

Vegetables

0.025 µg/g

mean of 16 samples of fruits and vegetables in U.K.
C10-20 (SCCPs and MCCPs)
(Campbell and McConnell, 1980a)

0.025 µg/g

1 sample of potato crisps in U.K.
C20-30

(Campbell and McConnell, 1980a)

Cereal products

SCCPs

0.13 µg/g

one reported concentration for "Chlorowax 500C" in enriched white bread in market basket survey carried out by U.S. Food and Drug Administration (KAN-DO Office and Pesticides Team, 1995); average molecular formula is C12H19Cl7, with 60-65% chlorine content (w/w) (IPCS, 1996)

0.05 µg/g

detection limit in analyses of corn flakes in U.K.
C20-30
(Campbell and McConnell, 1980b)

Cereal products

SCCPs/MCCPs

0.05 µg/g

detection limit in analysis of 1 sample of corn flakes in U.K.
C10-20 (SCCPs and MCCPs)
(Campbell and McConnell, 1980b)

0.05 µg/g

detection limit in analyses of corn flakes in U.K.
C20-30
(Campbell and McConnell, 1980b)

Meat and poultry

0.099 µg/g

1 sample of bacon in U.K.
C10-20 (SCCPs and MCCPs)
(Campbell and McConnell, 1980b)

0.05 µg/g

detection limit in analysis of 1 sample each of ox liver and beef in U.K.
C20-30
(Campbell and McConnell, 1980b)

Fish Note: Campbell and McConnell (1980b) presented data for combined SCCPs and MCCPs. Data for fish identified in Bennie et al. (2000), Muir et al. (1999) and Tomy and Stern (1999) were presented as separate analyses. no data identified
Fish

SCCPs

2.630 µg/g (wet weight); analysis of whole samples of carp from Hamilton Harbour; C10-C13 (Muir et al., 1999)

0.0588 µg/g; lake trout, Niagara-on-the-Lake (Muir et al., 1999)

0.0726 µg/g; lake trout, Port Credit (Muir et al., 1999)
 
0.502 µg/g; carp (n = 3) (Bennie et al., 2000)

1.47 µg/g; trout (n = 10) (Bennie et al., 2000)

1.8 µg/g (estimated); perch, Detroit River (Tomy and Stern, 2000)

no data identified
Fish

MCCPs

1.23 µg/g; mean of 10 samples of whole trout from western Lake Ontario (Bennie et al., 2000)

0.393 µg/g; carp (n = 3) (Bennie et al., 2000)

82 ng/g in perch; 904 ng/g in catfish (Tomy and Stern, 1999)

0.008 µg/g (estimated); perch, Detroit River (Tomy and Stern, 2000)

no data identified
Eggs no data identified no data identified
Foods primarily sugar

0.025 µg/g

1 sample of strawberry jam in U.K.
C10-20 (SCCPs and MCCPs)
(Campbell and McConnell, 1980b)

0.05 µg/g

detection limit in 1 sample of strawberry jam in U.K.
C20-30
(Campbell and McConnell, 1980b)

Mixed dishes no data identified no data identified
Nuts and seeds no data identified no data identified
Soft drinks, alcohol, coffee, tea

0.05 µg/g

detection limit in analyses of beverages in U.K.
(Campbell and McConnell, 1980a)

0.05 µg/g

detection limit in analysis of 1 sample each of beer and tea in U.K.
C20-30
(Campbell and McConnell, 1980b)

5.1.1 Short-chain chlorinated paraffins

Tomy (1997) determined SCCPs (C10-13, 60-70% chlorine) in 24-hour air samples collected daily during a 4-month period in the summer of 1990 in Egbert, Ontario, a "rural site northwest of Toronto," by high-resolution gas chromatography/electron capture negative ion Hhigh-resolution mass spectrometry (HRGC/ECNI-HRMS) (Muir et al., 1999). Concentrations ranged from 65 to 924 pg/m³. Although a summary statistic of 543 pg/m³ was reported, it was not specified whether this was a mean or median value. Egbert has also been reported to be near an "industrialized area" (Muir et al., 2000). Lower concentrations of SCCPs have been identified at other sites in Canada (Halsall et al., 1998; Stern et al., 1998; Bidleman et al., 1999, 2000, 2001; Muir et al., 2001).

Concentrations of SCCPs (C10-13, 52% chlorine) ranged from 11 to 17 µg/kg in human breast milk in Canada (Tomy, 1997). Analyses were carried out by HRGC/ECNI-HRMS. No additional details were reported.Footnote 2

Muir et al. (1999) analysed whole fish samples for SCCPs (C10-13) and detected 2630 ng/g (wet weight) in carp from Hamilton Harbour, 58.8 ng/g (wet weight) in lake trout from Niagara-on-the-Lake and 72.6 ng/g (wet weight) in lake trout from Port Credit. The quantification was by gas chromatography (GC)/ECNI-HRMS. Lower concentrations were reported in an earlier study (Muir et al., 1996).

In a market basket survey (KAN-DO Office and Pesticides Team, 1995)Footnote 3 of 234 ready-to-eat foods, which represented approximately 5000 food types in American diets, "Chlorowax 500C"Footnote 4 was detected once, in enriched white bread, at a concentration of 0.13 µg/g. Food items were screened by gas or liquid chromatography using ion-selective detectors. Findings were confirmed by unspecified analysis.

Concentrations of SCCPs have been identified in blubber of aquatic mammals such as ringed seal, beluga and walrus (Tomy et al., 2000Footnote 5; Bennie et al., 2000Footnote 6). The samples were from animals in Greenland, the Canadian Arctic and the St. Lawrence River. A mean concentration of 46 100 ng/g (n = 15) was reported for beluga from the St. Lawrence River/Gulf of St. Lawrence. Concentrations in ringed seals from Ellesmere Island ranged from 370 to 770 ng/g. Jansson et al. (1993) detected SCCPs in biota in Sweden, including fish and both terrestrial and marine mammals. Analysis was by GC/mass spectrometry (MS).

Data on concentrations of SCCPs in drinking water in Canada or elsewhere were not identified. The maximum concentration of SCCPs (C10-13, 50-70% chlorine) in the Red River, at a site remote from industrialized areas, was 0.05 µg/L (Tomy, 1997).Footnote 7 Analyses were by HRGC/ECNI-HRMS. A lower concentration was reported in surface water from Lake Ontario (Muir et al., 2001).

Concentrations of SCCPs in soil in Canada or elsewhere were not identified. The concentrations in surface sediment in harbours in Lake Ontario ranged from 5.9 to 290 ng/g dry weight (Muir et al., 2001). Analyses were by HRGC/ECNI-HRMS.

Upper-bound estimates of intake of SCCPs for the general Canadian population and the assumptions upon which they are based are presented in Table 9. For each age group in the Canadian population, virtually all of the estimated intake is from food. The upper-bound estimated intake of breast-fed infants was 1.7 µg/kg-bw per day, and that of formula-fed infants was 0.01 µg/kg-bw per day. For the remaining age groups, intakes ranged from 5.1 µg/kg bw per day for adults over 60 years of age to 26.0 µg/kg-bw per day for infants who were not formula fed (i.e., those being introduced to solid foodsFootnote 8).

Table 9: Upper-bounding estimated average daily intake of short-chain chlorinated paraffins by the population of Canada
(estimated intake (µg/kg-bw per day)
Route of exposure 0-6 monthsFootnote a
breast fedFootnote b
0-6 months
formula fedFootnote c
0-6 months
not formula fedFootnote d
0.5-4 yearsFootnote e 5-11 yearsFootnote f 12-19 yearsFootnote g 20-59 yearsFootnote h 60+ yearsFootnote i
Ambient airFootnote j <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001
Indoor airFootnote k <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001
Drinking waterFootnote l 1.7 0.005 0.001 0.001 0.001 <0.001 <0.001 <0.001
FoodFootnote m 1.7 0.005 25.96 24.26 16.44 9.02 7.18 5.14
SoilFootnote n 0.001 0.001 0.001 0.002 0.001 <0.001 <0.001 <0.001
Total intake 1.7 0.01 25.97 24.26 16.44 9.02 7.18 5.14

Canadian data incorporated within this estimate include high-confidence values in fish (whole carp determined by GC/ECNI-HRMS) and data on breast milk, for which details of sampling and analysis were not reported. Estimated intake of SCCPs in fish represents up to 58% of the total daily intake. The intake from dairy products, which accounts for 89.9% of the intake of infants not formula fed, is based upon limited sampling and analysis -- considered semi-quantitative only -- of dairy products in the United Kingdom, reported in 1980. Probably the most representative estimates of intake are those from cereals, which are based upon data reported in an American market basket survey, carried out from 1982 to 1991; however, intake from this foodstuff constitutes <0.1% of total estimated intake, and analytical methods were not specified.

Intake of SCCPs by a potentially higher-exposure subgroup of Inuit for whom the primary source of food is subsistence hunting and fishing (Kuhnlein, 1989; Kinloch et al., 1992) was also estimated, based on data on concentrations of SCCPs in blubber from marine mammals in Canada (Tomy et al. 2000) and less specific data (including both SCCPs and MCCPs) for terrestrial and marine mammals from Sweden (Jansson et al., 1993). On the basis of these data, the estimated intake of an Inuit adult, namely 1.47 µg/kg-bw per day, is well within the range of values estimated above for the general population.

5.1.2 Medium-chain chlorinated paraffins

MCCPs were detected by HRGC/low-resolution mass spectrometry (LRMS) in effluent (13 µg/L) from a chlorinated paraffin manufacturing plant in Canada in 1993, but not in surface water or sediment (Metcalfe-Smith et al., 1995). MCCPs were detected in three samples of carp from Hamilton Harbour in 1996 by low-resolution GC/MS (mean 0.393 µg/g; range 0.276-0.563 µg/g) (Bennie et al., 2000). Similarly, MCCPs were detected in the homogenized (whole) samples of 10 trout collected from western Lake Ontario in 1996 (mean 1.23 µg/g; range 0.257-4.39 µg/g) (Bennie et al., 2000).

Upper-bounding estimates of intake for MCCPs and the assumptions on which they are based are presented in Table 10. For each age group, virtually all of the estimated intake is from food, which, in turn, is based almost entirely upon the limited data reported by Campbell and McConnell (1980a,b). The highest intake estimated (25.5 µg/kg-bw per day) was for infants not formula fed.

Table 10: Upper-bounding estimated average daily intake of medium-chain chlorinated paraffins by the population of Canada
Route of exposure 0-6 monthsFootnote a.1
formula fedFootnote b.1
0-6 monthsFootnote a.1
not formula fedFootnote c.1
6 months-
4 yearsFootnote d.1
5-11 yearsFootnote e.1 12-19 yearsFootnote f.1 20-59 yearsFootnote g.1 60+ yearsFootnote h.1
Ambient airFootnote i.1 - - - - - - -
Indoor airFootnote j.1 - - - - - - -
Drinking waterFootnote k.1 0.05 0.01 0.01 0.01 <0.01 <0.01 <0.01
FoodFootnote l.1 0.05 25.48 18.48 11.64 6.3 4.69 3.47
SoilFootnote m.1 0.01 0.01 0.02 0.01 <0.01 <0.01 <0.01
Total intake 0.07 25.51 18.51 11.65 6.3 4.69 3.47

5.1.3 Long-chain chlorinated paraffins

Upper-bounding estimates of total intake of LCCPs and associated assumptions are presented in Table 11. As for SCCPs and MCCPs, for each age group, virtually all of the estimated intake is from food. The highest intake estimated (16.8 µg/kg-bw per day) was for infants not formula fed. In addition to the limitations of the analytical methodology noted previously, these estimates are further limited in that estimates for five of the eight food groups are based upon the limit of detection in that survey (Campbell and McConnell, 1980a,b).

Table 11: Upper-bounding estimated average daily intake of long-chain chlorinated paraffins by the population of Canada
(estimated intake µg/kg-bw per day)
Route of exposure 0-6 monthsFootnote a.2
breast fedFootnote b.2
0-6 monthsFootnote a.2
not formula fedFootnote c.2
6 months - 4 yearsFootnote d.2 5-11 yearsFootnote e.2 12-19 yearsFootnote f.2 20-59 yearsFootnote g.2 60+ yearsFootnote h.2
Ambient airFootnote i.2 - - - - - - -
Indoor airFootnote j.2 - - - - - - -
Drinking waterFootnote k.2 0.05 0.01 0.01 0.01 <0.01 <0.01 <0.01
FoodFootnote l.2 16.81 9.66 5.61 3.04 2.12 1.73
SoilFootnote m.2 0.01 0.01 0.02 0.01 <0.01 <0.01 <0.01
Total intake 0.07 16.83 9.69 5.63 3.04 2.12 1.73

5.2 Hazard characterization and dose-response analyses

A limited number of studies on the toxicity of SCCPs have been reported in the period following release of the PSL1 assessment. Most of these studies were conducted to investigate the mode of action of carcinogenicity for the tumours observed in the National Toxicology Program (U.S.A.) (NTP) (1986a) bioassay, which were liver tumours in both sexes of rats and mice, kidney tumours in male, but not female, rats and thyroid tumours in rats and mice (females only). For several of these more recent studies, results have been reported in abstracts or summaries only: Elcombe et al. (1994) (abstract), Elcombe et al. (2000) (summary) and Warnasuriya et al. (2000) (abstract). For only one of the relevant investigations has a full published account been identified (Wyatt et al., 1993). While secondary accounts of (possibly) other studies investigating mode of action of tumour induction in assessments have been reported by the European Commission (2000), the U.S. National Research Council (U.S. NRC, 2000) and the Australian National Industrial Chemicals Notification and Assessment Scheme (NICNAS, 2001), they are not further considered here, owing to lack of availability or confirmation of subsequent publication (Jackson, 2001).

Few data relevant to the assessment of the toxicity of either MCCPs or LCCPs were identified for the period to the release of the PSL1 assessment report. The following presentation is limited to those considered critical to hazard characterization or dose-response analyses for effects in the general population and, hence, to assessment of "toxic" under Paragraph 64(c) of CEPA 1999. Other sources of non-critical data identified but not included were DuPont (1995), Kato and Kenne (1996) and Warngard (1996).

In view of the absence of recent toxicological data that impact on critical aspects, the dose-response analyses for MCCPs and LCCPs presented here reflect primarily those developed in the PSL1 Assessment Report released under CEPA 1988.

5.2.1 Short-chain chlorinated paraffins

A- Liver

Increased liver weight, hepatocellular hypertrophy, peroxisomal proliferation and increased S-phase activity in hepatocytes were reported in Fischer 344 rats administered SCCPs for up to 90 days (presumably by gavage) at dose levels up to 1000 mg/kg-bw per day (Elcombe et al., 1994; abstract). Lower doses administered were not specified, and quantitative dose- or sex-specific data and analyses were not presented.

Elcombe et al. (2000) administered Chlorowax 500C (C10-13; 58% chlorine) to male and female Fischer 344 rats by gavage in corn oil for up to 90 days, at dose levels of 0, 312 or 625 mg/kg-bw per day. In both sexes, liver weight was increased, accompanied by peroxisomal proliferation (as indicated by an increase in cyanide-insensitive palmitoyl coenzyme A [CoA] oxidation) and increased thyroxine (T4)-uridine diphosphoglucose glucuronosyl transferase (UDPGGT). (The effects were, presumably, observed at both dose levels.) These effects were not observed in male Dunkin Hartley guinea pigs similarly administered 0, 500 or 1000 mg/kg-bw per day for 14 consecutive days. The numbers of animals exposed were not specified, and quantitative dose- or sex-specific data and analyses were not presented in this summary account.

Wyatt et al. (1993) exposed groups of five male rats (Alpk:APfSD strain) each by gavage for 14 days to 0, 10, 50, 100, 250, 500 or 1000 mg/kg-bw per day to two SCCPs (Chlorowax 500C: C10-13, 58% chlorine; or Cereclor 56L, C10-13: 56% chlorine). For the 58% chlorine SCCPs, both absolute and relative liver weights were significantly increased in a dose-related manner, at doses of 100 mg/kg-bw per day or greater. Peroxisomal fatty acid β-oxidation activity (indicated by palmitoyl CoA oxidation) was significantly increased at 250 mg/kg-bw per day and greater (irregular dose-response). For the 56% chlorine SCCPs, the pattern of response for absolute liver weight was irregular; however, relative liver weight was increased in a dose-related manner, significantly at 50 mg/kg-bw per day and greater. Palmitoyl CoA oxidation was significantly increased only at the highest dose.

In similarly exposed male mice (Alpk:APfCD-1 strain), for the 58% chlorine SCCPs, there was a dose-related increase in relative liver weight and palmitoyl CoA oxidation, both significant at 250 mg/kg-bw per day and greater (Wyatt et al., 1993). For the 56% chlorine SCCPs, both absolute and relative liver weights were significantly increased in a dose-related manner at doses of 100 mg/kg-bw per day or greater. Palmitoyl CoA oxidation was significantly increased in a dose-related manner at 250 mg/kg-bw per day and greater.

The only other relevant investigation identified was an in vitro study in which SCCPs inhibited gap junction intercellular communication in rat liver cells (Kato and Kenne, 1996; Warngard et al., 1996).

B- Kidney

Increased proximal tubular cell eosinophilia (suggestive of a protein overload, but not necessarily α2u globulin) and regenerative focal basophilic tubules, as well as increased S-phase activity in the proximal tubular cells, were reported in male, but not female, rats administered up to 1000 mg SCCPs/kg-bw per day for up to 90 days (other dose levels were not specified) (Elcombe et al., 1994). These observations were reported in an abstract; neither quantitative data nor statistical analyses were presented.

Elcombe et al. (2000) also investigated renal effects in F344 rats and guinea pigs administered 0, 312 or 625 mg SCCPs/kg-bw per day for up to 90 days. In the male rats only, there was chronic protein nephropathy, associated with regenerative hyperplasia and increased DNA synthesis (S-phase activity), presumably at both dose levels. There was "some limited evidence" for an involvement of α2u globulin. These changes were not observed in the guinea pigs. Again, neither quantitative data nor statistical analyses were presented in this summary account.

Warnasuriya et al. (2000) exposed male and female rats by gavage for 28 days to 625 mg SCCPs (C12; 60% chlorine)/kg-bw per day. There was an increase in α2u globulin and cell proliferation in the kidney of males only. Data from individual rats indicated that increased cell proliferation was directly correlated with the increase in α2u globulin. Five different isoelectric isoforms of α2u globulin were identified by Western blotting in the control male kidney, and all five were increased in the treated males. These observations were reported in an abstract; neither quantitative data nor statistical analyses were presented.

C- Thyroid

Elcombe et al. (1994) reported that exposure of rats to SCCPs for up to 90 days resulted in induction of T4-glucuronosyl transferase activity, accompanied by a decrease in plasma T4 and an increase in thyroid stimulating hormone (TSH). Thyroid follicular cell hypertrophy and hyperplasia were also observed. Increased S-phase activity in the thyroid follicular cells was also reported. The maximum dose was 1000 mg/kg-bw per day; other dose levels were not specified. This study was reported as an abstract; neither quantitative data nor statistical analyses were presented.

In male and female Fischer 344 rats exposed by gavage in corn oil to 0, 312 or 625 mg/kg-bw per day for up to 90 days, there were decreases in plasma T4, increases in plasma TSH and thyroid follicular cell hypertrophy and hyperplasia in both sexes, changes that were not observed in male guinea pigs (Elcombe et al., 2000). Quantitative data and statistical analyses were not presented in this summary account.

Gavage administration of 6.8 mg/kg-bw per day commercial C10-13 (71% chlorine) to female Sprague-Dawley rats for 14 days had no effect upon thyroid hormonal T4 levels or microsomal enzyme activity (Hallgren and Darnerud, 1998).

In male rats (Alpk:APfSD strain) exposed by gavage for 14 days to two SCCPs (Chlorowax 500C: C10-13, 58% chlorine; or Cereclor 56L, C10-13: 56% chlorine), for which examination of thyroid function was restricted to the control and high-dose groups (1000 mg/kg-bw per day), both free and total T4 were significantly reduced, TSH was significantly increased and the capability of liver microsomes to glucuronidate T4 was significantly increased in exposed animals (Wyatt et al., 1993). No differences in levels of free or total triiodothyronine (T3) were observed for either SCCPs. A significant increase in glucuronosyl transferase activity with p-nitrophenol was observed only from microsomes from rats exposed to the C10-13 (58% chlorine) compound.

5.2.2 Medium-chain chlorinated paraffins

A subchronic dietary study with MCCPs in rats (Poon et al., 1995) was initiated by Health Canada in response to the research needs identified in the PSL1 assessment of chlorinated paraffins (Government of Canada, 1993a). Sprague-Dawley rats (10 per sex per group) were fed diets containing 0, 5, 50, 500 or 5000 ppm for 13 weeks. The dose levels calculated by the authors on the basis of weekly food consumption were 0, 0.4, 3.6, 36 and 363 mg/kg-bw per day for males and 0, 0.4, 4.2, 42 and 419 mg/kg-bw per day for females. The protocol included serum biochemistry, hematology, hepatic enzyme activities, urinary enzyme activity, organ weights and histopathology. Mild, adaptive histological changes were detected in the liver of rats of both sexes at the two highest doses (lowest observed effect level (LOEL) = 36 mg/kg-bw per day) and in the thyroid of males at 36 mg/kg-bw per day and greater and of females at 4.2 mg/kg-bw per day and greater (no-observed-adverse-effect level (NOAEL) = 0.4 mg/kg-bw per day). Minimal changes were observed in the renal proximal tubules of males at the highest dose and in the inner medulla of females at the two highest doses.

5.2.3 Long-chain chlorinated paraffins

No critical data relevant to the assessment of the toxicity of LCCPs were identified for the period since the PSL1 assessment was released.

5.3 Human health risk characterization

5.3.1 Short-chain chlorinated paraffins

A- Hazard characterization
Genotoxicity

Requisite criteria for assessing the weight of evidence for hypothesized modes of induction of tumours addressed below include the criterion that SCCPs are not DNA-reactive. Recent data on genotoxicity reported since the PSL1 assessment was released have not been identified. Limited available data reviewed within the PSL1 assessment indicated that SCCPs were clastogenic in in vitro assays, although they had not been clastogenic or mutagenic in a limited number of in vivo assays.

Based on review of the available data, including two additional unpublished studies in which no increases in revertant colonies in five strains of Salmonella and no increases in mutant colonies in Chinese hamster V79 cells were reported in the secondary account, it was concluded that "as a group, SCCPs are not mutagenic" (European Commission, 2000).

Liver

It has been hypothesized that SCCPs cause liver tumours in rodents secondary to peroxisome proliferation. Peroxisome proliferation involves activation of a nuclear receptor in rodent liver, the peroxisome proliferator activated receptor, α isoform (PPARα). The activated PPARα interacts with regulatory elements of the DNA to initiate transcription of genes for increased peroxisomal enzyme activity and cell proliferation characterized by morphological and biochemical changes in the liver. These changes include increased liver weight through both hepatocyte hypertrophy and hyperplasia, increased number and size of peroxisomes, increased activity (up to 40-fold) of peroxisomal enzymes (especially those involved in peroxisomal fatty acid oxidation) and induction of microsomal fatty acid oxidation through the CYP4A subfamily of cytochrome P-450 isozymes. Minimum criteria for characterizing peroxisome proliferation are considered to include hepatomegaly, enhanced cell proliferation and an increase in hepatic acyl-CoA oxidase and/or palmitoyl-CoA oxidation levels.

In the NTP bioassay (NTP, 1986a; Bucher et al., 1987) reported in the PSL1 assessment, increases in benign liver tumours were observed in both SCCPs-exposed rats (312 and 625 mg/kg-bw per day) and mice (125 and 250 mg/kg-bw per day), with males of both species being considerably more sensitive. This pattern of induction of liver tumours by SCCPs is consistent with that for other peroxisome proliferating hepatocarcinogens, such as di(2-ethylhexyl)phthalate.

Available data on the role of peroxisome proliferation in the etiology of hepatic effects and liver tumours induced by SCCPs are restricted to one study for which there is a published manuscript (Wyatt et al., 1993) and two investigations reported only in summary (Elcombe et al., 2000) or abstract form (Elcombe et al., 1994). Significant, dose-related increases in both absolute and relative liver weights accompanied at higher doses by increases in palmitoyl CoA oxidation in male Alpk:APfSD rats and Alpk:APfCD-1 mice exposed to two SCCPs, reported by Wyatt et al. (1993), are consistent with the observations in rats of Elcombe et al. (1994, 2000). Also, to the extent to which the more recent and better-documented study of Wyatt et al. (1993), with more extensive characterization of dose-response, can be compared with the earlier investigations of Elcombe et al. (1994, 2000), for which only summary reports are available, observations on dose-response for increases in liver weight and palmitoyl CoA oxidation in rats in these investigations are also consistent (increases in relative liver weight in rats were significant at ≥50 mg/kg-bw per day and palmitoyl CoA oxidation at ≥250 mg/kg-bw per day; comparable values for mice were 100 mg/kg-bw per day and 250 mg/kg-bw per day).

Therefore, although characterization of exposure-response was limited in the NTP bioassay to only two dose levels, evidence to date indicates that tumours in both rats and mice occur only at doses at which peroxisome proliferation and associated morphological and biochemical effects have been observed in shorter-term studies (Wyatt et al., 1993; Elcombe et al., 1994, 2000).

Additional weight of evidence for concordance might have been afforded through consideration of sex-related differences in peroxisome proliferation in shorter-term mechanistic studies. Unfortunately, this aspect was not investigated in the well-reported study by Wyatt et al. (1993) in which only male rats and mice were exposed; moreover, the limited extent of reporting in Elcombe et al. (1994, 2000) precludes consideration of relevant data in this context, if such data were, indeed, collected. Recovery studies would also have been informative, since peroxisome proliferation is initiated rapidly after treatment with a proliferator begins, attains a maximal response in a few weeks and is maintained only in the continued presence of the proliferator. Consistent with a receptor-mediated response, the process is reversible.

While there have been no carcinogenesis bioassays for SCCPs in species other than rats and mice, the variation in species sensitivity to peroxisome proliferation reported by Elcombe et al. (2000) is consistent with that observed for other peroxisome proliferators. Rats and mice are uniquely responsive to the morphological and biochemical effects of peroxisome proliferators, while Syrian hamsters exhibit intermediate responsiveness. This is consistent with marked interspecies variations in the expression of PPARα.

Additional published documentation of existing relevant studies is desirable. Also, investigation of additional aspects of concordance would strengthen the weight of evidence for causality for the purported association between peroxisome proliferation and liver tumours induced by SCCPs. However, although there are limitations of the identified information, data are strongly suggestive that peroxisome proliferation plays a role in the etiology of liver damage and hepatic tumours associated with exposure to SCCPs. Although additional evidence for the weight of causality for liver tumours is desirable, a TDI based on hepatic effects in experimental animals is considered to be protective for carcinogenicity.

Kidney

It has been hypothesized that the kidney tumours observed following exposure of male rats to SCCPs are a species- and sex-specific response attributable to α2u globulin nephropathy and hence not relevant to humans. This mode of induction of renal tumours, which is relatively well characterized, involves binding to α2u globulin, a protein specific to male rats. This binding renders the protein more resistant to proteolytic degradation, which causes its accumulation in renal proximal tubule cells (manifested as hyaline droplets on histopathological examination), resulting in cell death and regenerative proliferation. Sustained cell proliferation leads to a low but significant incidence of renal tubular tumours.

Minimum criteria for establishment of α2u globulin nephropathy as a basis for tumour development include lack of genotoxicity and observation of requisite precursor lesions and tumours in male rats only. Confirmation of requisite precursor lesions is based not only on histopathological observations such as excessive accumulation of hyaline droplets in renal proximal tubule cells, subsequent cytotoxicity and single-cell necrosis of the tubular epithelium and sustained regenerative tubular cell proliferation in the presence of continued exposure, but also on explicit identification of the protein accumulating in tubule cells as α2u globulin, along with demonstrated reversible binding of the relevant chemical or metabolite to α2u globulin (U.S. EPA, 1991; IARC, 1999).

In the NTP bioassay (NTP, 1986a; Bucher et al., 1987) reported in the PSL1 assessment, renal tubular cell adenomas were observed in male rats at both doses (312 and 625 mg/kg-bw per day), although the increase was significant (p < 0.05) only at the lower dose. Characterization of exposure-response was limited, therefore, in the NTP bioassay to only two dose levels.

Available data on the mode of induction of kidney tumours in male rats by SCCPs are restricted to three investigations reported only in summary or abstract format (Elcombe et al., 1994, 2000; Warnasuriya et al., 2000). In Elcombe et al. (1994, 2000), regenerative focal basophilic tubules and increased S-phase activity in the proximal tubular renal cells were observed in male, but not female rats and considered by the authors to constitute "limited evidence" of the role of α2u globulin. More recently, the presence of α2u globulin was confirmed using immunohistochemical techniques, although no details of methodology were provided (Warnasuriya et al., 2000).

Owing to the inadequate characterization in abstracts of even administered doses, in some cases with quantitative data on effects and analyses not being reported, there is very limited documentation to serve as a basis for conclusion that renal tumours occur only at doses at which either chronic protein nephropathy associated with regenerative hyperplasia and increased DNA synthesis (Elcombe et al., 2000) or α2u globulin is observed (Warnasuriya et al., 2000).

While information is strongly suggestive that the kidney tumours observed in male rats are attributable to hyaline droplet formation, a male rat-specific phenomenon not relevant to humans, additional published documentation of available studies is clearly desirable as a basis for consideration of the weight of evidence of mode of induction of kidney tumours. Although additional confirmation is desirable, atolerable daily intake (TDI) based on renal effects in experimental animals is considered to be protective for carcinogenicity.

Thyroid

There are a variety of non-DNA-reactive compounds that cause thyroid tumours in rats associated with decreased circulating thyroid hormone levels due to increased hepatic metabolism (particularly Phase II conjugating enzymes such as uridine diphosphate (UDP) glucuronosyl transferases [UDPGTs] and glutathione S-transferases) and clearance. These compounds induce hepatic glucuronidation of thyroid hormones and increase biliary excretion of the conjugated hormones, resulting in decreased circulating T3 and T4 levels. As a result of the hypothyroid state, TSH levels increase and cause sustained thyroid follicular cell hyperplasia, leading to tumour formation.

While the basic physiology and feedback mechanisms of the hypothalamic-pituitary-thyroid axis are qualitatively similar across species, quantitative differences make rodents more sensitive than humans to development of thyroid cancer for which the sole mode of action is thyroid-pituitary disruption (U.S. EPA, 1998). These include the lack of a high-affinity thyroid binding globulin in rats relative to humans (Dohler et al., 1979), which likely affects the turnover of the hormone. With a more rapid turnover of T4, there is a generalized increased activity of the pituitary-thyroid axis in rats compared with humans, which correlates with increased susceptibility to thyroid gland neoplasia.

Minimum criteria for establishment of this mode of action as a basis for tumour development include evidence of increases in thyroid growth and hormonal changes (the latter including reduction in circulating serum T4 and T3 and an increase in thyroid stimulating hormone (TSH) levels within days or a few weeks of exposure). Evidence of increases in thyroid growth is provided by measured increases in absolute or relative thyroid weight, histological indication of cellular hypertrophy and hyperplasia, morphometric determination of alteration in thyroid cellular components and changes in proliferation of follicular cells detected by DNA labelling or mitotic indices (U.S. EPA, 1998).

In the NTP bioassay (NTP, 1986a; Bucher et al., 1987) reported in the PSL1 assessment, increases in follicular cell adenomas and carcinomas (combined) were observed in female rats only, at 312 and 625 mg/kg-bw per day, and in female mice only, at 250 mg/kg-bw per day.

Available data relevant to assessment of the weight of evidence of induction of thyroid tumours in rats by SCCPs are limited to one study for which there is a published manuscript (Wyatt et al., 1993) and two investigations for which only a published summary report (Elcombe et al., 2000) or abstract (Elcombe et al., 1994) is available. In the study for which a complete account was published, effects on the thyroid were considered only in the control and highest dose groups; the administered dose for the latter was considerably greater than those in the NTP bioassay associated with thyroid tumours (that is, 1000 mg/kg-bw per day versus 312 and 625 mg/kg-bw per day). In addition, in the abstract and summary accounts, quantitative data on effects or analyses were not presented. For example, Elcombe et al. (2000) reported only that male and female Fischer 344 rats were exposed by gavage in corn oil for up to 90 days at dose levels of 0, 312 or 625 mg/kg-bw per day and that "there were decreases in plasma thyroxine, increases in plasma TSH concentration and thyroid follicular cell hypertrophy and hyperplasia in both sexes." There are extremely limited data, therefore, to serve as a basis for consideration of concordance of dose-response between thyroid tumour induction and precursor effects in shorter-term studies, such as thyroid growth and hormonal changes. In a single additional study for which a full account is available (Hallgren and Darnerud, 1998), the dose level  at which effects on thyroid hormonal T4 levels or microsomal enzyme activity were not observed were much less than those administered in the NTP bioassay; as a result, these are not additionally meaningful in this context.

As a result, although data from the studies reported by Elcombe et al. (1994, 2000) and Wyatt et al. (1993) fulfil the criteria for tumour induction by thyroid disruption in part, it should be noted that these data are insufficient as a basis for analysis of dose-response for concordance with that for thyroid tumours. Also, recovery in the absence of continued exposure has not been investigated. In view of the limitations of both reporting and dose-response analyses, therefore, there is considerable uncertainty in attributing observed thyroid tumours to thyroid-pituitary disruption, to which rodents are more sensitive than humans.

B- Risk characterization

Available data relevant to consideration of the weight of evidence for proposed modes of induction of liver, kidney and thyroid tumours associated with exposure to SCCPs, although limited, are suggestive that tolerable intakes that protect for non-neoplastic precursor effects will likely also be protective for cancer. However, owing principally to limited investigation of aspects such as recovery and inadequate documentation of relevant studies, there is considerable uncertainty in drawing this conclusion, particularly for the thyroid tumours. In recognition of this uncertainty, both neoplastic and non-neoplastic effects are considered here.

IPCS (1996) derived a TDI of 100 µg/kg-bw per day for non-neoplastic effects of SCCPs on the basis of the lowest reported No-Observed-Effect Level (NOEL) of 10 mg/kg-bw per day in a 13-week study in rats (IRDC, 1984a). At the next higher dose in the critical study (100 mg/kg-bw per day), there were increases in liver and kidney weight and hypertrophy of the liver and thyroid. In IPCS (1996), an uncertainty factor of 100 was applied in the development of the TDI to account for interspecies variation (×10) and intraspecies variation (×10). The potential for progression of lesions following longer-term exposure was not explicitly addressed in the development of the TDI. This is balanced to some degree by the relatively large margin between the NOEL and the LOEL (10-fold) in the critical study and the minimal severity of the effects at the next higher concentration; however, there is some justification for considering a somewhat lower value for the TDI.

On the basis of multistage modelling of the tumours with highest incidence (hepatocellular adenomas or carcinomas [combined] in male mice) in the carcinogenesis bioassay with SCCPs, IPCS (1996) also estimated the dose associated with a 5% increase in tumour incidence (Tumorigenic Dose05 [TD05]) to be 11 mg/kg-bw per day (amortized for period of administration).

The upper-bound estimate of exposure for the age group with greatest exposure to SCCPs (i.e., 26 µg/kg-bw per day) is within the range of the IPCS (1996) TDI, for which there is some justification for considering a somewhat lower value, to take into account potential progression of the lesions in longer-term studies.

The margin between the upper-bound estimate of exposure for the age group with greatest exposure to SCCPs and the Tumorigenic Dose (TD05) (i.e., 440) is also considered inadequate in view of the uncertainty concerning mode of induction of tumours.

5.3.2 Medium-chain chlorinated paraffins

A TDI developed on the basis of the NOAEL (0.4 mg/kg-bw per day) in the more recent subchronic study conducted by Health Canada (Poon et al., 1995) would be similar to that derived for the PSL1 assessment (that is, 6 µg/kg-bw per day).

Several of the highly uncertain bounding estimates of total daily intake of MCCPs from drinking water, food and soil for the general population of Canada exceed the TDI (6 µg/kg-bw per day) for non-neoplastic effects. Indeed, for infants not formula fed, the total daily intake of MCCPs (that is, 25.5 µg/kg-bw per day) exceeds the TDI by up to 4-fold.

5.3.3 Long-chain chlorinated paraffins

None of the highly uncertain bounding estimates of total daily intake of LCCPs from drinking water, food and soil for the general population of Canada exceeds the TDI (71 µg/kg-bw per day) for non-neoplastic effects. However, for infants not formula fed, the total daily intake of LCCPs (16.8 µg/kg-bw per day) is within the same order of magnitude as the TDI.

5.4 Uncertainties and degree of confidence in human health risk characterization

There is low confidence in the upper-bounding estimates of exposure to all chlorinated paraffins. The estimates of intake for most age groups in the general Canadian population are based almost entirely upon limited sampling of foodstuffs in the United Kingdom, which were published in 1980. Methodology for analysis in this study is considered inadequate by present-day standards, and, as such, the data can be regarded at best as semi-quantitative. Reported concentrations represented both SCCPs and MCCPs, and, as a result, intake of the individual groups of chlorinated paraffins (SCCPs, MCCPs and LCCPs) from these sources has been overestimated.

The estimates of intake for SCCPs are based in part upon the results of more recent surveys, for which methods of analysis were more reliable (i.e., quantification by GC/ECNI-HRMS). Concentrations of SCCPs determined by HRMS were available for ambient air, water and samples of carp from Hamilton Harbour (intake from fish represented 38-58% of estimated total intake of SCCPs, although fish accounts for, at most, 4% of the total daily intake of food across the six age groups).

However, it is not possible to quantify the extent of overestimation of exposure based on the earlier, likely less selective analytical methodology, owing to lack of comparable data. Moreover, results based on analysis of the same samples by LRMS versus HRMS have been inconsistent, with levels of SCCPs being 1-2 orders of magnitude less for the latter in samples of whale blubber (Bennie et al., 2000; Tomy et al., 2000) and trout (Muir et al., 1999; Bennie et al., 2000) but slightly greater for the high-resolution analysis in carp (Muir et al., 1999; Bennie et al., 2000).

There is minimal confidence in the upper-bounding estimates of exposure to MCCPs. These estimates are based in large part upon concentrations reported in a limited number of foodstuffs in the United Kingdom, which were published in 1980. More recent, although limited, data on concentrations in trout analysed by LRMS were included in the calculation of upper-bounding estimates.

There is minimal confidence in the upper-bounding estimates of exposure to LCCPs. These estimates are based entirely upon concentrations reported in a limited number of foodstuffs in the United Kingdom, which were published in 1980. Furthermore, concentrations in foods were represented by the limits of detection for five of eight food groups in the calculations of daily intake.

There is a low degree of confidence in the database of toxicological studies that serves as the basis for the assessment of the weight of evidence for mode of induction of tumours by SCCPs, for which only one published complete report (Wyatt et al., 1993) is available and for which it has not been possible to identify published accounts for reported pre-publication manuscripts reviewed in previous assessments. Results in the only fully documented study provide most meaningful support for the purported role of peroxisome proliferation in induction of liver tumours in rats and mice.

There is a moderate degree of confidence in the database of toxicological studies upon which the TDI for MCCPs is based, for which studies on chronic toxicity or carcinogenicity are lacking. The database for LCCPs is more complete, including a well-documented carcinogenicity bioassay in rats and mice.

Page details

Date modified: