State of the Science Report - Part 4

Phthalate Substance Grouping
Medium-Chain Phthalate Esters

Chemical Abstracts Service Registry Numbers
84-61-7; 84-64-0; 84-69-5; 523-31-9; 5334-09-8; 16883-83-3; 27215-22-1; 27987-25-3; 68515-40-2; 71888-89-6

Environment Canada
Health Canada
August 2015

Table of Contents

• Appendix A: Structural identity and physical chemical properties of analogue substances
• Appendix B: Physical and chemical properties for the substances in the medium-chain phthalate subgroup
• Appendix C: Results of Level III fugacity modelling (EQC 2011) for the medium-chain phthalate esters in the phthalate substance grouping
• Appendix D: Bioaccumulation
• Appendix E: Toxicity values
• Appendix F-1: Estimates of daily intake
• Appendix F-2: Methodology - Dietary intakes
• Appendix G: Methodology - Biomonitoring intakes
• Appendix H: Summary of toxicokinetics of medium-chain phthalates (MCPs)
• Appendix I: Supporting information of the chronic toxicity and carcinogenicity of BBP
• Appendix J: Description and Application of the Downs and Black Scoring System and Guidance for Level of Evidence of An Association
• Back to the State of the Science Report - Part 1

Appendix A: Structural identity and physical chemical properties of analogue substances

Top of Page

Appendix B: Physical and chemical properties for the substances in the medium-chain phthalate subgroup

Top of Page

Appendix C: Results of Level III fugacity modelling (EQC 2011) for the medium-chain phthalate esters in the phthalate substance grouping

Top of Page

Appendix D: Bioaccumulation

Top of Page

Appendix E: Toxicity values

Top of Page

Appendix F-1: Estimates of daily intake

Top of Page

Appendix F-2: Derivation of dietary intakes

Occurrence data - DIBP and DCHP

Occurrence data for DIBP and DCHP were obtained from an American total diet study (Schecter et al. 2013), and any data gaps were filled using data from a British total diet study (Bradley et al. 2013b). Occurrence data for these two phthalates, in food, that was reported as less than the analytical LOD were assigned values of ½ LOD. However, a value of 0 (zero) was assigned to all samples within a broad food category when no phthalates were detected above the LOD in any sample in that category.

Food consumption data and matching to occurrence data

The phthalate concentrations in individual foods were matched to consumption figures for these foods from the Canadian Community Health Survey (CCHS) B Cycle 2.2 on Nutrition (Statistics Canada 2004) to generate distributions of phthalate exposure for various age-sex groups. The CCHS included 24-hour dietary recall information for over 35 000 respondents of all ages across Canada.

If a food line item belonged to a recipe that was matched to a set of the assayed foods, then the associated phthalate levels matched to the recipe were assigned to the ingredient. Otherwise, if the food line item itself matched to a set of the assayed foods then the phthalate levels matched to the food line item were assigned. For DIBP and DCHP, 989 foods and 23 recipes were matched with assayed foods.

Body weight information

For the purpose of determining per kilogram body weight exposure estimates, infant body weights were set to the mean body weights, as derived from the body weight data from the United States Department of Agriculture Continuing Survey of Food Intakes by Individuals (CSFII; 1994-96, 1998). For all age groups, body weights reported in the CCHS, whether measured or self-reported, were used and, where missing, were imputed using the median for the corresponding age-sex group and quintile of energy intake.

Probabilistic exposure assessment

For each food consumed by a respondent in the CCHS survey, phthalate concentrations were randomly selected from the matching list of assayed values. For each individual respondent, exposure estimates from each food were summed, generating a distribution of exposure for all respondents. This was repeated 500 times (500 iterations) to model the variability of the distribution of exposures due to the variability of the phthalates levels. For each age-sex group, the median and 90^th percentile exposures were derived from the empirical distribution generated by the 500 iterations.

Top of Page

Appendix G: Derivation of daily intakes for DIBP based on biomonitoring

P4 pregnant women and MIREC CD+ infants:

DIBP_{daily intake} (µg/kg bw • day) = [C_Sum (moles/g Cr) × CER × MW of DIBP] / [FUE_Sum × BW ]

Where,

C _Sum (moles/g Cr):

Sum of molar concentrations of metabolites,

CER:

24 hour creatinine excretion rate (estimated using the Mage Equation),

[FUE _Sum:

Sum of FUE of the metabolites = 0.91,

MW of DIBP:

278

Step 1: Conversion of concentrations

C_metabolite (moles/g Cr) = C_metabolite(µg/g Cr) / MW_metabolite

C_MIBP (moles/g Cr) = C_MIBP (µg/g Cr) / 222 g/mol

C_2OH-MIBP (moles/g Cr) = C_2OH-MIBP (µg/g Cr) / 239 g/mol

Step 2: Sum the concentrations from Step 1

C_Sum (moles/g Cr) = Σ C_MIBP + C_2OH-MIBP

Step 3: Sum FUEs

FUEs for MIBP and 2OH - MIBP are 0.71 and 0.195, respectively. Therefore, the sum would be 0.91.

Step 4: Compute DI for DIBP using Equation 1.

CHMS

Statistical analysis: The data were analyzed with SAS 9.2 (SAS Institute Inc., USA) and SUDAAN 10.0.1 software (RTI International, USA). Variance estimates were produced using bootstrap weights, taking into account the 11 degrees of freedom for cycle 1 and 13 degrees of freedom for cycle 2, as suggested in the CHMS data user guide. All analyses were weighted using the CHMS cycle 1 survey weights (phthalate subsample) and CHMS cycle 2 survey weights (environmental urine subsample) in order to be representative of the Canadian population. Phthalate concentrations that were below LOD were assigned a value of LOD/2.

Estimation of creatinine excretion rate (CER): For each study, the participant creatinine excretion rate was calculated using the Mage equations (Huber et al. 2010). The adiposity adjustment (discussed in the supplemental information [Huber et al. 2010]) was applied for all participants, and the body surface area adjustment was applied for children under the age of 18. Median BMIs by age for the adiposity adjustment were computed using the entire CHMS sample. The CHMS phthalate subsample dataset had 174 children who exceeded height limits in the Mage equations (186 cm for males and 172 cm for females). The Mage equations were applied directly to the observed heights in order to extrapolate creatinine excretion rates for these participants. The predicted excretion rates for these individuals appeared to be reasonable despite the extrapolation.

Daily intake estimation: The daily intake of DIBP, based on urinary concentrations of the monoester MIBP, was estimated for each participant using the following equation (David et al. 2000; Koch et al. 2007):

Equation 1:

Daily intake (µg/kg bw/day) = [UC_Cr (µg/g Cr) × CER (g/day) / BW (kg) × FUE] × [MW_D×MW_M]

The fractional urinary excretion (FUE) is defined as the fraction of the diester exposure dose excreted as metabolites in urine, calculated on a mole basis. For the calculation, an FUE of 0.71 for MIBP was used (Koch et al. 2012). MW_D and MW_M are the molecular weights of the diester (DIBP: 278 g/mol) and the monoester (MIBP: 222 g/mol), respectively.

Arithmetic and geometric means, and selected percentiles along with their 95% confidence intervals of daily intake, were produced for the Canadian population by age group, sex and fasting status. Descriptive statistics were computed using SUDAAN proc DESCRIPT and SAS proc SURVEYREG.

Top of Page

Appendix H: Summary of toxicokinetics of medium-chain phthalates (MCPs)

A review of the available literature indicates that almost all in vivo studies on the toxicokinetics of medium-chain phthalates (MCPs) have been conducted via oral and dermal routes (only one inhalation study was found). Several in vitrostudies were found, mainly investigating dermal absorption (cell diffusion), intestinal absorption (everted gut) and metabolism (microsomal preparations, tissue homogenates from liver, kidney, intestines, testes, plasma and purified enzymes). Most studies have been conducted with Di-(2-ethylhexyl) phthalate (DEHP) in rats, but some studies have examined the toxicokinetics of medium-chain phthalates in other rodents and non-rodent species.

Oral route

There is evidence that phthalates, regardless of chain length, are absorbed from the gastrointestinal (GI) tract after oral exposure. However, several studies have shown that the extent of absorption of phthalates in the GI tract of rats was not linear with increasing dose, likely due to saturation of the mechanism of uptake or of the diester hydrolysis, particularly for the phthalates with long carbon chains on the ester linkage. Similarities and differences are seen across the phthalates with respect to metabolism and chain length. The smaller phthalates undergo hydrolysis to their respective monoester in the GI tract and are excreted without further metabolism. Larger phthalates, such as medium-chain phthalates, undergo hydrolysis in the GI tract to their respective monoester but can also undergo further oxidative metabolism to other metabolites and be excreted as such or as conjugates.

Absorption

Three human studies regarding oral absorption of DEHP were found. There is variability in the rates of absorption reported in these low-dose studies (greater than or equal to 70% over 44 hours at 0.005-0.65 mg/kg [Koch et al. 2005], 70-89% over 36 hours at 3 mg/person [Kurata et al. 2012a] and 11-25% over 24-58 hours at less than 0.5 mg/kg [Schmid and Schlatter 1985]). All of the studies in animals were conducted at higher doses (2.90-2800 mg/kg) and, due to possible saturation at high doses, may not be directly comparable. However, at low doses (2.9 mg/kg in rat; Daniel and Bratt 1974) and at relatively low doses in other animals (50-100 mg/kg in monkeys, marmosets, rats, dogs and pigs), the absorption rates (56-66% and 30-50%, respectively, in urine) were in the same range as those reported in humans at very low doses (Short et al. 1987; Ikeda et al. 1980; Rhodes et al. 1986; Lhuguenot et al. 1985). Consequently, the data available do not provide evidence of strong differences between the absorption of DEHP in the GI tract of humans and of other mammals. See Table 1 for a summary of the absorption rates of DEHP in animals and humans.

Similar to DEHP, human studies for other phthalates (DBP, BBP and DIBP) were conducted at very low doses (less than 1.3 mg/kg), while animal studies were usually conducted at doses higher than 50 mg/kg, with the exception of one study with BBP (Eigenberg et al. 1986a). BBP is the only phthalate with data at low doses in both humans and animals, and a comparison of their absorption rates indicates that absorption is similar in both species (67-84% over 24 hours in humans vs. 70-80% over 96 hours in rats) (Anderson et al. 2001; Eigenberg et al. 1986a). For DBP, the values obtained in humans (69-92% over 8-48 hours at 0.255-5 mg/kg) were also comparable to those obtained in rats at the lowest tested doses (77-96% over 24-48 hours at 100 mg/kg) (Williams and Blanchfield 1975; Fennellet al. 2004; Seckin et al. 2009; Anderson et al. 2001; Koch et al. 2012). A more recent study by Koch et al. (2012) using one male volunteer determined that approximately 92% of the orally administered dose of DBP was excreted within the first 24 hours, while only less than 1% of the dose was excreted in urine after 48 hours. See Table E-1 for a summary of the absorption rates of other phthalates studied in animals and humans.

Several studies have shown that the absorption of phthalates in the GI tract of rats and marmosets was not linear with increasing dose. This might be due to saturation of the mechanism of uptake or of the diester hydrolysis. At environmental levels, DBP is most likely absorbed as MBP in the GI tract of rats due to the high lipase enzyme activity in situ. At relatively high doses (100-250 mg/kg), however, direct absorption of the unhydrolyzed phthalate most likely occurs due to enzymatic saturation (Silva et al. 2007a). Saillenfait et al. (1998) also showed that at a dose of 500 mg/kg of DBP, 60% of DBP was absorbed, whereas only 48% was absorbed at a dose of 1500 mg/kg (based on urinary excretion over 48 hours). An even greater discrepancy was reported by Eigenberg et al. (1986a), with 70-80% absorption at 2-200 mg/kg and only 22% at 2000 mg/kg (based on urinary excretion over 96 hours) for BBP.

Similar observations were done with DEHP administered to marmosets, with a 2-fold increase in absorption for a 20-fold increase in dose (Rhodes et al. 1986). More recently, Kurata et al. (2012a) observed a ~3.5-fold drop in absorption of DEHP with a 25-fold increase in dose in marmosets based on plasma concentrations (AUC_all). With BBP administered to rats, saturation seemed to occur between 475 and 780 mg/kg, since absorption rates were 58, 54, 43 and 30% (based on urinary excretion over 24 hours) at 150, 475, 780 and 1500 mg/kg of BBP, respectively (Nativelle et al. 1999).

Absorption also seems to differ with respect to the age of the animal. Plasmatic MEHP levels were measured after repeated exposure to DEHP (1000 mg/kg/day) in rats of different ages (25, 40 and 60 days old) for 14 days. The mean plasmatic AUC of MEHP in the youngest age group was reported to be twice as high as in the two older groups (Sjoberg et al.1986). It was suggested by Sjoberg et al. (1985a) that absorption is greater in young rats when DEHP is orally administered. The authors proposed that this may be related to the higher relative proportion of intestinal tissue to body weight and to the higher blood flow through the intestinal tissue in young rats compared to older rats. This may occur in humans since blood flow decreases with age, but experimental evidence on age differences in absorption is non-existent. However, recent work by Kurata et al. (2012a) examining the toxicokinetics of DEHP in 3-month-old and 18-month-old marmosets did not detect any age-related differences in absorption as measured by plasma concentrations.

Distribution

Distribution of medium-chain phthalate compounds after oral absorption was studied in vivo in several rodent (rat and mice) and non-rodent (dog, pig, marmoset, cynomolgus monkey and human) species mainly for two phthalates: DBP and DEHP. Overall, it appears as though adipose tissues, absorptive organs and excretory organs are the major initial repositories for the dialkyl esters, with distribution through the body at varying levels according to the phthalate administered, the dose and the species used. Several studies have also examined the distribution of phthalates in pregnant animals and fetuses. Most human studies refer to biomonitoring of phthalates in serum, amniotic fluid or breast milk within the general population (environmental exposure).

A dietary study conducted in rats (1000 ppm labelled DBP in diet) showed particularly high radioactivity in the liver but also in kidney and adipose tissue. The radioactivity persisted after 96 hours in the adipose tissue, while it disappeared rapidly from the other tissues after termination of exposure. Data also suggested an accumulation, after four weeks of exposure (compared to one day) in testes (1.6 vs. 0.3 μg/g) and in adipose tissue (11.2 vs. 8.35 μg/g) in rats (Williams and Blanchfield 1975). More recent work using rats has shown that DBP is rapidly distributed (distribution half-life of 5.77 min) after administration (30 mg/kg, i.v.) and undetectable in the plasma with low cumulative fecal excretion after oral administration (100 mg/kg, oral gavage) after 48 hours (Chang et al. 2013).

In studies conducted in animals that were exposed repeatedly, there was no evidence of accumulation of the monoester of DEHP in plasma. Phokha et al. (2002) reported no cumulative effect on the area under the curve (AUC) of the monoester MEHP in rats after repeated oral administrations of DEHP (500 mg/kg/day, in aqueous emulsion). In a study by Clewell et al. (2009), peak MBP concentrations in maternal and fetal plasma from rats exposed to DBP (0, 50, 100 and 500 mg/kg, in corn oil) on GD12-19 were 67 and 55% lower than those after a single dose (0 or 500 mg/kg, in corn oil).

Tissue distribution of DEHP may be governed by its lipophilicity since higher concentrations occurred in adipose tissue compared to liver in rats administered DEHP (0 or 5000 ppm in diet) for 13 weeks (Poon et al. 1997). At the end of the study, the levels in adipose tissue were 23 ppm (males) and 31 ppm (females), compared to a barely detectable level (3 ppm) in the liver for both sexes.

Although most studies have shown that the liver and kidneys are initially the most common retention repositories in rats (Williams and Blanchfield 1974, 1975), radioactivity accumulation in muscles was also documented. Tanaka et al. (1975) showed that after a single oral administration of 500 mg/kg of radiolabelled DEHP, the distribution of radioactivity was in the following order after 3 hours: intestine (C_max = 51%) greater than muscle (4.86%) greater than liver (2.75%) greater than other organs. Excretion of DEHP appeared to be delayed in adipose tissue.

In a species comparison study, a single oral dose of labelled DEHP (50 mg/kg) administered to rats, dogs and miniature pigs showed that distribution of radioactivity in pigs and rats was similar (high radioactivity in the liver and the adipose tissues at 4 hours, with a steady decline thereafter). Dogs showed a different pattern, with radioactivity initially high in the liver and muscles, whereas radioactivity in the adipose tissue was very low (Ikeda et al. 1980). In another study, oral exposure of rats to ¹⁴C-DEHP labelled in the phenyl ring (2000 mg/kg/day) for 14 days demonstrated the magnitude of tissue distribution in the following order: liver (205 μg/g of DEHP equivalent) greater than kidney (105 μg/g) greater than blood (60 μg/g) greater than testes (40 μg/g) (Rhodes et al. 1986).

Distribution appeared different in monkeys exposed under the same conditions as described above; their tissue concentrations were lower (10-15% of the amounts in rats) and the levels of DEHP equivalent were as follows: testes (3.75 μg/g) greater than kidney (3 μg/g) greater than liver (2.5 μg/g) greater than blood (1 μg/g) (Short et al. 1987). In juvenile and adult marmosets, the highest radioactivity of orally administered ¹⁴C-DEHP was found in the kidney 2 hours after dosing and was attributed to be the result of urine excretion (Kurata et al. 2012a). There was no abnormal distribution of radioactivity in the testis or other male reproductive organs.

Potential distribution to the fetus and infant

Daily oral administration of DEHP (750 mg/kg bw, in corn oil) in rats showed that the parent compound and its metabolites cross the placental barrier and reach the fetal gonads (Stroheker et al. 2006). However, fetal livers contained a major part of radioactivity (20-31%) and gonad levels were low (2-5%). More recently, Hayashi et al. (2012) measured hepatic MEHP levels in pregnant mice and their offspring and found that concentrations of this metabolite were 1.5 times higher in the liver of pregnant dams than those of postpartum mice. Further, MEHP concentrations in foetuses were 1.7 times higher than in pups at the same dose levels of DEHP (0.05%).

Fennell et al. (2004) showed the presence, in female rats, of MBP and its glucuronide levels in maternal plasma, foetal plasma and amniotic fluid. MBP concentrations were two- to four-fold higher in maternal plasma than in foetal plasma. In amniotic fluid, MBP is initially the major metabolite, but 24 hours after oral administration, MBP glucuronide became the major metabolite. The half-life reported was very different between free MBP (6-11 hours) and MBP glucuronide (up to 64 hours) (Fennell et al. 2004). A non-linear increase in MBP was observed in both maternal plasma (by ten-fold) and fetal plasma (by eight-fold), while the dose increased by only five-fold in rats administered DBP (100 or 250 mg/kg/day) on GD12-18 (Fennell et al. 2004). That MEHP and MBP are primarily unconjugated the day after the last dosing (while in maternal urine, both free and conjugated monoesters are important) was confirmed by Calafat et al. (2006a).

Since many studies indicate that DBP and its metabolites are rapidly cleared from the body, it was previously suggested that it was unlikely that DBP would be stored in maternal tissues and released during pregnancy and lactation (Foster et al. 1982; Tanaka et al. 1978; Williams and Blanchfield 1975). Indeed, Saillenfait et al. (1998) showed that the amount of radioactivity in the embryo peaked at 0.12% of the total administered dose at 6 hours post-dosing and rapidly declined to undetectable levels thereafter, following a single oral dose of 1500 mg/kg [¹⁴C]-DBP to pregnant rats on GD14.

The effect of repeated doses on the distribution of DBP metabolites in maternal tissues and amniotic fluid was studied by Clewell et al. (2009). Pregnant rats were exposed to DBP (0, 50, 100 and 500 mg/kg, in corn oil) on GD12-19. MBP concentrations in the amniotic fluid were reduced with repeated doses of DBP (at 500 mg/kg). MBP-glucuronide, however, was not decreased. In fact, the MBP-glucuronide concentrations in amniotic fluid were consistently higher in the repeated-dose study than in the single-dose study. Maternal liver MBP levels were also reduced after multiple exposures (the C_max for MBP after multiple doses was 72% of the value at single dose). Maternal liver MBP-glucuronide concentrations were not significantly different in the single- and repeated-dose groups.

The effect of dose on distribution of radioactivity from ¹⁴C-DBP was also studied by Saillenfaitet al.(1998). Rats were administrated a single oral dose of ¹⁴C-DBP (500 or 1500 mg/kg, in mineral oil). In all tissues studied (maternal kidneys, liver, ovaries, stomach, intestine and uterus), C_max and AUC for MBP were higher at 1500 mg/kg than at 500 mg/kg. However, an increase of AUC was disproportionate in embryo and amniotic fluid, with an eight-fold increase in AUC_0-∞ (interestingly, the high dose was embryotoxic), suggesting a more pronounced embryonic exposure to MBP at high doses. This was confirmed in a study by Calafat et al. (2006a). In this study, oral administration of DBP (0, 11, 33, 100 and 300 mg/kg/day) to pregnant rats showed that increasing doses resulted in increased concentration of metabolites in the amniotic fluid. There was also an exponential relationship between MEHP levels in amniotic fluid and maternal urine. A strong correlation between MEHP levels in amniotic fluid and maternal DEHP dose was reported when dams were administered DEHP (0, 11, 33, 100 or 300 mg/kg, in corn oil); pups were likely to receive some intact DEHP (Calafat et al. 2006a).

Silva et al. (2007b) showed a linear dose-related increase of serum concentrations of MBP and its oxidative metabolites (mono-n-hydroxybutyl phthalate [MHBP] and mono-(3-carboxypropyl) phthalate [MCPP]). There was a non-linear dose-related increase of MBP concentration in fetal amniotic fluid, the concentration in amniotic fluid being increased by approximately 10-fold (mean 1.4 μg/mL and 13.4 μg/mL, respectively), while the dose administered (100 and 250 mg/kg/day) differed by only 2.5-fold. However, the detection of MBP in amniotic fluid does not provide definitive evidence of MBP crossing the placenta (or of DBP metabolism in the fetus). According to Fennell et al. (2004), the rapid appearance of MBP and the delay in appearance of its glucuronide could indicate fetal metabolism of MBP at a much slower rate than by the mother, if in fact MBP glucuronide does not cross the placenta. Alternatively, it could be an indication that the MBP glucuronide does cross the placenta, but at a much slower rate than MBP.

When considering potential distribution of medium-chain phthalates during lactation, the increasing fat solubility of longer-chain phthalates may facilitate their higher segregation into maternal milk, since lipophilic chemicals readily partition into high fat materials (Main et al. 2006; Kluwe 1982). In a study where female rats received three daily administrations of DEHP (2000 mg/kg) on days 15-17 of lactation, DEHP was not detected in dam's plasma, whereas milk concentrations of DEHP were very high (216 μg/mL). This may be explained by the association of DEHP with lipoproteins in the plasma and with lipids in rat milk, or by uptake of lipoproteins by the mammary gland for milk synthesis (Dostal et al. 1987).

Species differences in distribution of phthalates during pregnancy have been examined in a study conducted in rats and marmosets. This work suggested that MEHP tissue burden may be smaller in marmoset fetuses than in those of rats, since at a similar daily dose (500 mg/kg), C_max and normalized AUC of MEHP in marmoset blood were up to 7.5- and 16-fold lower, respectively, than in rats (Kessler, et al. 2004).

Humans

In the general population, monoesters of phthalates were found in serum with a glucuronide distribution pattern similar to that in urine (Silva et al. 2003). This is in line with the results obtained in an experimental study conducted in one man administered a single oral dose of DEHP (Koch et al. 2005). However, a recent study by Kessler et al. (2012) found surprisingly high concentrations of the parent DEHP in the blood of male volunteers compared with animal data, and MEHP was detected almost immediately (15 min) after ingestion.

Several monoesters (MEP, MBP and MEHP) were found in human amniotic fluid at 2- to 3-fold lower levels than in serum (Silva et al. 2004a). These results were in accordance with those reported in experimental studies (Fennell et al. 2004).

Calafat et al. (2004) suggested that phthalate metabolites may be present in breast milk and, therefore, can be transferred to the nursing child. Indeed, several phthalate diesters have been found to be present in samples of human milk. The most commonly detected compounds were DEHP, DBP, DIBP, BBP and DEP (as well as their monoesters) (Latini et al. 2009; review by Fromme et al. 2011; Guerranti et al. 2012). In the study conducted by Fromme et al. (2011), DCHP was also detected in 17% of milk samples. Oxidative metabolites of DEHP and DINP were not detected (Latini et al. 2009; Fromme et al. 2011; Guerranti et al. 2012).

Metabolism

The metabolism of DEHP has been extensively studied. There appears to be a consensus that the metabolism of other phthalate compounds is qualitatively similar. Briefly, the diester is first hydrolyzed into a monoester (before, during and/or after absorption). The monoester can then either be i) hydrolyzed into phthalic acid, ii) conjugated to be further excreted, or iii) metabolized into primary and secondary hydroxyl products that can further be oxidized to yield diacids. Thus, the metabolites of phthalate diesters are monoesters, phthalic acid and products of the oxidative metabolism. These metabolites may be conjugated before excretion or be excreted under their free form (see Table H-2 for summary). Some studies have shown that metabolic pathways may saturate at high doses.

Metabolic pathways

Although metabolism is not exactly the same for all phthalate diesters, a metabolic pathway for the most common plasticizers, those having saturated alkyl groups, has been postulated based on the identification of metabolites produced in vivo and excreted in urine (Albro 1986). The obligatory first step is the hydrolysis of the parent to its monoesters by a non-specific lipase (esterase) found in several organs and tissues, particularly in the pancreas, intestinal mucosa, liver, skin and lung (Albro and Lavenhar 1989). Enzymes capable of hydrolyzing phthalate diesters have been found in human saliva (Silva et al. 2005b), breast milk (Calafat et al. 2004) and serum (Kato et al. 2003).

Hydrolysis of the phthalate diester to the monoester may occur in many tissues (e.g., small intestine, skin, pulmonary tract, liver and kidney), but more extensive metabolism may be limited to some tissues, especially the liver (Kluwe 1982). In most cases, especially with large molecules, a rapid hydrolysis of phthalate diesters to their respective monoesters takes place before absorption, and the phthalic moiety is further distributed into the body (Ono et al. 2004). Studies conducted in rats with DEHP indicate that a large proportion of diester administered orally was hydrolyzed to monoester in the intestine (Phokha et al. 2002). However, a study with DEHP, DnOP and DCHP reported a faster rate of hydrolysis in the presence of small intestine contents than with caecal contents, suggesting that such diesters are hydrolyzed by esterases of both bacterial and mammalian origin (Rowland et al. 1977).

During the second phase of metabolism, the monoester can either be i) hydrolyzed into phthalic acid by microsomal esterase, ii) conjugated by UDP-glucuronyltransferase, or iii) metabolized into primary and secondary hydroxyl products by microsomal monooxygenase (analogous to the cytochrome P450 associated fatty acid ω- and ω-1 hydroxylase). The primary and secondary hydroxyl products resulting from the latter pathway are subject to oxidation by the alcohol dehydrogenase (ADH) and aldehyde dehydrogenase (ALDH), respectively, leading to diacids or ketoacids. Finally, the diacids are subject to α- and β-oxidation to yield shorter diacids (Albro et al. 1973a,b; Albro 1986; Albro and Lavenhar 1989). See Figure H-1 below for a postulated metabolic pathway for DEHP as an example.

Figure H-1. Figure showing postulated pathways for metabolism of DEHP in mammals (adapted from Albro 1986)

DEHP is by far the most studied phthalate diester (Koch et al. 2005; Kessler et al. 2012; Anderson et al. 2012; Kurata et al. 2012a). Other diesters have similar metabolic steps, but the involvement of each pathway may differ from one substance to another. It is considered that, after oral ingestion, DEHP is hydrolyzed by acid lipases in the stomach, followed by immediate resorption of the monoester (Kessler et al. 2012). The peak plasma level of the parent DEHP lagging one hour after ingestion is attributed to its structural similarity with lipids. Lipid resorption does not start before gastric emptying and the formation of an emulsion with bile. MEHP is assumed to be resorbed into the portal blood as it binds preferentially to serum albumin. Hayashi et al. (2012) measured hepatic MEHP levels in pregnant mice and their offspring along with enzyme activities of lipase and uridine 5'-diphosphate-glucuronosyltransferase (UGT). UGT activity appeared to be 1.5-fold higher in the liver of pregnant dams than postpartum ones. This was potentially reflective of the higher MEHP levels measured in pregnant dams compared to postpartum mice, based on the hypothesis that some MEHP is conjugated with uridine 5'diphosphate (UDP)-glucuronide by the catalytic action of UGT and is excreted in the urine (Albro and Lavenhar 1989). The remaining MEHP is also excreted directly in the urine or is oxidized by cytochrome P450A (CYP4A) (Hayashi et al. 2012).

Kessler et al. (2012) also found that there was significant variability in time between the four volunteers tested and that individual blood burdens of DEHP and MEHP can be estimated from the DEHP dose. Further, the mean AUC/D of DEHP was 50 and 100 times higher than in rats and marmosets, respectively (Kessler et al. 2004). This can be explained by the species differences in intestinal resorption and hydrolysis [see Figure H-2, and DEHP and other medium-chain phthalates in Table H-2].

Figure H-2. The metabolism of DEHP is illustrated in this figure (adapted from Koch et al. 2005). DEHP is rapidly metabolized to its monoester, MEHP, which is further extensively modified by various side-chain hydroxylation and oxidation reactions. The major metabolites of DEHP are bolded

Recent studies by Kurata et al. (2012a,b) determined that there are significant species differences between humans, marmosets and rats in the ratio of excreted conjugated and non-conjugated forms (G/F ratio) of the secondary metabolites of DEHP in urine. The G/F ratios for humans and marmosets were similar (77.6-84.2% and 87.7%, respectively) for the glucuronidated form of metabolites in urine after 24 hours, compared to only 11.2% G/F ratio in rats (Kurata et al. 2012b). For rats, the majority of the secondary metabolites were in their free form (87.4% G/F ratio).

Information on the metabolism of other medium-chain phthalates is not as detailed, but the pathways are qualitatively similar to the one described for DEHP. The toxicokinetics of DIBP was recently evaluated by Koch et al. (2012) using one human male volunteer. It appears as though mono-iso-butylphthalate (MIBP) is the major urinary metabolite of DIBP (about 70% of the administered dose), followed by 2OH-mono-iso-butylphthalate (2OH-MIBP) (19%) and 3OH-mono-iso-butylphthalate (3OH-MIBP) (0.69%) after 24 hours. Therefore, the oxidized metabolites account for about 20% of the overall dose.

BBP is an asymmetric diester that can potentially form equal amounts of monobutylphthalate (MBP) and monobenzylphthalate (MBeP). However, a higher proportion of MBP (29-34%) compared to MBzP (7-12%) was reported in rats after repeated oral administrations of BBP; in this study, the major metabolite was hippuric acid (51-56%) and there was no glucuronide (Nativelleet al. 1999). The metabolic pathways of BBP in rats proposed by these authors are illustrated in Figure 5. In contrast, in humans administered BBP orally (253 and 506 µg, single dose), there was a preferential cleavage of the butyl ester link (leading to more MBzP) and there was little further metabolism of MBzP; MBzP was thus the major urinary metabolite in humans (Andersonet al. 2001). In a more recent study, the species differences in BBP hydrolysis by liver microsomes were investigated among humans, monkeys, dogs, rats and mice. It was found that the hydrolysis activities of BBP to MBP in monkey, rat and mouse liver microsomes were 28-, 22- and 44-fold higher than that in human liver microsomes, although the activity of dog liver microsomes was comparable to that in human liver microsomes. In contrast, the hydrolysis activities of BBP to MBzP in monkey, rat and mouse liver microsomes were 34, 9.3 and 12% of that in human liver microsomes, respectively, whereas the activity in dog liver microsomes was 1.6-fold higher than that in human liver microsomes. The authors proposed that the hydrolysis of BBP to monoester phthalates in mammalian liver microsomes could be classified into two types: MBzP greater than MBP type for humans and dogs, and MBP greater than MBzP type for monkeys, rats and mice. Since the formation profile of MBP and MBzP in liver microsomes of dogs highly paralleled that of human liver microsomes, it was also suggested that the properties of carboxylesterase isoform(s) of dogs involved in BBP hydrolysis could be much more similar to those of humans than other animal species (Takahara et al. 2014).

Figure H-3. Proposed routes of BBP metabolism in female Wistar rats (adapted from Nativelle et al. 1999). The six metabolites of BBP (n°1 to 6) recovered in urines are reported in this figure

Induction and saturation of metabolic pathways

Several studies documented an induction of metabolic pathway after repeated exposure to phthalate diesters. For example, DnOP and DCHP were shown to induce the hepatic activity of monoxygenases involved in their own metabolism after oral absorption (Lake et al. 1982; Poon et al. 1997). In rats, MEHP blood levels were lower and their half-life was shorter after repeated oral administrations of DEHP (1055 mg/kg bw/day) than after a single administration (Sjoberg et al. 1986), suggesting an induction of MEHP hydrolysis. Daniel and Bratt (1974) suggested that the DEHP/MEHP ratio (reflecting hydrolysis of DEHP) in the GI tract may be altered by induction or inhibition of pancreatic lipase. Induction of metabolic pathway appears to be consistent during pregnancy as well. After repeated oral exposure of pregnant rats to DBP, maternal and fetal plasma MBP levels were consistently lower compared to single doses, suggesting induced MBP metabolism (Clewell et al. 2009).

This effect is not limited to rodents. In Cynomolgus monkeys, hydrolysis and ω-oxidation were induced by repeated oral doses of DEHP, as reflected by increased urinary levels of mono(2-ethyl-5-carboxypentyl) phthalate (5cx-MCPP) and decreased levels of mono(2-ethyl-3-carboxypropyl) phthalate (MECPrP) (Short et al. 1987). β-oxidation was increased after repeated dietary administrations of DEHP in the same study using rats, as reflected by a decreased urinary level of 5cx-MCPP. However, there was an increased level of MECPrP (Shor et al. 1987).

Conversely, the metabolism of phthalate diesters also appears to be saturable at several steps. Metabolism can be saturated at hydrolysis of the diester as seen in a study showing that after oral administration of a high dose of DEHP (2800 mg/kg), unchanged DEHP was recovered in blood. Saturation was suggested to occur before (in the gut content) or after absorption through the GI tract (Teirlynck and Belpaire 1985).

Metabolism of the monoester was also suggested to become saturated after oral administration of a single dose of DEHP. Kinetics of MEHP metabolism was slower at 1000 mg/kg compared to 30 mg/kg (AUC in blood was only twice as high at 1000 mg/kg; delayed time point for maximal blood levels) in female Sprague-Dawley rats (Kessler et al. 2004). In pregnant rats, the metabolism of MEHP also appeared to be lower at 500 mg/kg compared to 30 mg/kg (the blood AUCs for DEHP were comparable, while the normalized blood AUC for MEHP was higher at the highest dose) (Kessler et al. 2004).Further along metabolism, during the oxidation of downstream products, it was shown that β-oxidation (5cx-MEPP to MECPrP) appeared to be saturated in rats administered DEHP through their diet (greater than or equal to 6000 ppm in diet) (Short et al. 1987).

Saturation was also seen in other phthalates. Metabolism appeared to be saturated at 780 and 1500 mg/kg/day in female rats administered BBP orally for three days. At these doses, urinary elimination of metabolites (hippuric acid, MBP, MBeP, phthalic acid) as the percentage of the administered dose (43 and 30% at 780 and 1500 mg/kg, respectively) was lower than at the low dose (54-58% at 150 and 475 mg/kg/day) (Nativelle et al. 1999).

In pregnant rats administered DBP (single dose: 50, 100 or 250 mg/kg), glucuronidation of MBP appeared saturated at 250 mg/kg since the time to reach maximal plasma concentration of MBP and MBP-glucuronide was longer than at 50 or 100 mg/kg (MBP: 2 hours vs. 0.5 hour, MBP-glucuronide: 2 hours vs. 1 hour). There was a non-linear increase of AUC for MBP (disproportionate increase: by 10-fold vs. at 50 mg/kg) and the maternal and fetal plasmatic concentrations exhibited two peaks, one at 0.5 hour (followed by a decrease at 1 hour) and an absolute peak at 2 hours (maternal plasma) or 4 hours (fetal plasma) (Fennell et al. 2004).

Metabolic differences related to species, age and inter-individual variation

Some studies have shown that there are some species differences in the metabolism of phthalates. For example, lipase, which transforms DEHP into MEHP, may play a predominant role in interspecies variability of DEHP metabolism. Enzymatic activities involved in the metabolism of DEHP differ between primates (marmosets) and rodents (rats and mice) (Ito et al. 2005). It was shown that lipase activity in various tissues (liver, small intestine, kidney and lung) was lower in marmosets than in rats or mice by at least one order of magnitude. Lipase activity was found to be higher in the small intestine than in the liver of both rats (by 1.7-fold) and mice (by 4.3-fold). In contrast, lipase activity was 1.6-fold higher in the liver of marmosets compared to the small intestine. Similarly, the ratio V_max/K_m for lipase activity in the liver of marmosets (1.38) was dramatically lower than in rats (227) or mice (333). Hepatic UGT activity was also lower (2- to 3-fold) in marmosets compared to rodents. However, ADH and ALDH activities were generally similar or higher in marmosets, suggesting that ω- or ω-1 oxidized metabolites of MEHP (by CYP4A) are more difficult to further metabolize in rats and mice compared to marmosets (Ito et al. 2005). Overall, the activity of marmoset lipases appears to be much less than that of the rat and this may explain the different metabolite patterns between these two species during urinary excretion (Rhodes et al. 1986; Kurata et al. 2012a). Kurata et al. (2012a) postulated that the secondary metabolites of DEHP appeared to be promptly conjugated and excreted in marmosets (as observed in humans) and, therefore, this species would be a good analogue to measure toxicity of phthalates in humans because conjugation may potentially reduce the bioactivity of the metabolites by reducing their bioavailability.

In Ito et al. (2014), the activities of the same four DEHP-metabolizing enzymes were measured in the livers of 38 human subjects of various ages and in eight 129/Sv male mice. Microsomal lipase activity was significantly lower in humans than in mice regardless of sex, age or race differences. The V_max/K_m value in humans was one-seventh of that in mice. Microsomal UGT activity in humans was a sixth of that in mice, and cytosolic ALDH activity for 2-ethylhexanal in humans was one-half of that in mice. In contrast, ADH activity for 2-ethylhexanol was two-fold higher in humans than in mice. The total amount of DEHP urinary metabolites and the concentration of MEHP were much higher in mice than in the U.S. general population based on data reported in the 2003-2004 U.S. National Health and Nutrition Examination Survey, regardless of the similar estimated DEHP intake between these mice and the human reference population. However, mono(2-ethyl-5-oxo-hexyl)phthalate (5oxo-MEHP) and mono(2-ethyl-5 carboxypentyl)phthalate (5cx- MEPP) levels in urine were higher in humans than in mice (Ito et al. 2014).

In vitro hepatic studies (for DMP, DEP, DBP, DnOP, DEHP and DCHP) have also revealed quantitative species differences in the phthalate diester hydrolase activity, with higher alkaline esterase activities in non-human primates (baboons) than in rats, and higher activity in rats than in ferrets. Similar studies conducted with intestinal mucosal cell preparations also indicated a higher activity in baboons than in rats, and higher activity in rats than in ferrets. However, these values were not strictly comparable on an interspecies basis because the intestinal sections used (30-40 cm) may refer to different intestinal regions in rats, baboons and ferrets (Lake et al. 1977a).

The enzymatic activities of esterase and β-glucuronidase in the liver, intestinal mucosa and testes were also examined and compared between rats and hamsters for DBP metabolism (Foster et al. 1983). It was shown that esterase activity (which transforms MBP into phthalic acid) in the liver and intestinal mucosa was 2- and 1.3-fold higher, respectively, in hamsters than in rats. In contrast, the β-glucuronidase activity in the testes of rats was higher (by 2.2- to 6.5-fold) than in hamsters.

To facilitate excretion, metabolites may be conjugated. The rates of conjugation were found to vary between species (Lake et al. 1976; Albro et al. 1982; Egestad et al. 1996). It is noticeable that rats present the particularity of not conjugating the metabolites of DEHP. To compensate, three to six oxidative steps occur to produce metabolites with carboxyl groups on the side chain (Albro et al. 1982).

Glucuronidation of phthalate metabolites may also be affected by life stage. Rat fetuses do not have a functional glucuronidation pathway at GD17 (Calafat et al. 2006a). A slower fetal metabolism of MBP compared to maternal glucuronidation was suggested by the results obtained by Fennell et al. (2004). After oral administration of DBP (50 or 100 mg/kg) to pregnant rats on GD20, MBP appeared rapidly in both maternal and fetal plasma (maximal concentration reached 0.5 and 1 hour after administration, respectively), but there was a delay in the appearance of MBP-glucuronide in fetal plasma (time to reach maximal concentration: 4 hours vs. 1 hour in maternal plasma). These results could indicate either a slower fetal metabolism of MBP (vs. maternal glucuronidation) if MBP-glucuronide does not cross the placenta, or that MBP-glucuronide crosses placenta at a much slower rate than MBP (Fennell et al. 2004). This might have significant impact on the level of toxicity caused by phthalates during fetal development.

In relation to gender differences in metabolism in humans, a recent study by Anderson et al. (2012) measuring low- and high-dose administrations of DEHP to ten male and ten female volunteers showed that there was no statistically significant difference in the excretion kinetics or the metabolite composition between males and females, but there was a considerable amount of inter-individual variability. Ito et al. 2014 has proposed that the inter-individual variation in the metabolism of DEHP in humans may be greater that the inter-species differences between mice and humans based on the variability in the measurement of four enzymes involved in DEHP metabolism in the livers of human subjects and male mice (10- to 26-fold for inter-individual variation vs. 2- to 7-fold for inter-species variation).

Chemical-specific factors affecting metabolism

Significant work has been done investigating whether there is a relationship between the molecular weight, chain length, chemical structure and/or lipophilic characteristics of phthalates and their metabolism in rodents (in vivo and in vitro), primates (in vitro) and humans (in vitro).

In vitro studies conducted with rat liver and kidney homogenates have shown that there is a direct relationship between the molecular weight of phthalate diesters (DMP, DBP, DnOP and DEHP) and their metabolic rates in these organs. In both the liver and kidney, hydrolysis to the monoester was faster for the diesters for a lower-molecular-weight phthalate (metabolic rate ranking: DMP greater than DBP greater than greater than DnOP greater than DEHP) than the larger phthalate (Kaneshima et al. 1978a). Similarly, the metabolism of phthalate diesters by intestinal mucosal preparations was shown to be inversely related to the alkyl side chain length of phthalates (DMP greater than DEP greater than DBP greater than DnOP). This relationship was observed with rat and baboon intestinal mucosa cell preparations, and with human duodenum and jejunum preparations (Lake et al. (1977a).

In vivo studies examining the second phase of metabolism were conducted on rats administered phthalate diesters orally. Results have shown that hydrolytic monoesters are more likely to be the ultimate metabolites of the small phthalate diesters (e.g., DBP) than of the comparatively larger C8+ phthalates (Albro and Lavenhar 1989; Albro and Moore 1974; Albroet al.1973; Calafat et al. 2006b; McKee et al. 2002). In vitro metabolism studies conducted with rat liver and kidney homogenates also suggested a possible relationship between the molecular structure of phthalate diesters and their metabolic rates in these organs. It was shown that hydrolysis of DNOP, an n-alkyl C8 phthalate diester, was faster than hydrolysis of DEHP, a branched C8 diester (Kaneshima et al. 1978a).

The lipophilicity of a phthalate appears to also play a role in its metabolism. The affinity of phthalate diesters for purified rat liver carboxylesterases (pI 5.6 and pI 6.2/6.4) has been shown to increase (i.e., decreasing K_m values) with increasing lipophilicity (K_ow) diester compounds (K_mvalues ranking: DMP greater than DEP greater than DBP greater than DIBP). For the reaction rates (V_max), a similar relationship was observed for esterase pI 5.6; however, for esterase pI 6.2/6.4, there was no evident link between Vmax and log Kow (Mentlein and Butte 1989).

Excretion

Urine is the major route of elimination for medium-chain phthalate diesters and their metabolites. In all species and for all phthalate compounds for which data were available, the metabolites present in urine are both free and glucuronidated, except for DEHP in rats (metabolites only present under the free form). The pattern of urinary excretion of DEHP in humans can be illustrated by the results of Dirven et al. (1993b), who reported that 26% of the metabolites quantified were MEHP, 52% were products of a (ω-1)-hydroxylation of MEHP and 22% were the product of a ω-hydroxylation.

As urine is the most important route of excretion for most phthalates, it is largely used to conduct human biomonitoring in order to estimate exposure to phthalates. Generally, the metabolites allow for the identification of the parent compound, but some metabolites are common to several compounds and are thus poor biomarkers of exposure to their precursor phthalate diester. For instance, mono(3-carboxypropyl) phthalate (MCPP), a major metabolite of DnOP, is also recovered at varying levels in the urine of rats administered DIOP, DINP, DIDP, DEHP and DBP (Calafat et al. 2006b). These authors also suggested that the hydrolytic monoesters of the larger (C8+) phthalates are poor biomarkers of exposure in rats and, although there may be differences in metabolism among species, the lower molar ratios of the hydrolytic monoesters of these phthalates compared to those of the oxidative metabolites may explain the relatively low frequency of detection of the hydrolytic metabolites in humans.

Fecal excretion represents both the part of the compound not absorbed by the GI tract and the part of the compound excreted in the bile and not further reabsorbed. Fecal excretion may be an important route of excretion, depending on the parent compounds, the dose (higher fecal excretion when metabolism is saturated) and the route of administration. For instance, for DEHP, the feces contained relatively high concentrations of unoxidized MEHP, while the more polar metabolites (e.g., diacids and hydroxyl acids) were in much higher relative abundance in the urine in rodents and monkeys (Albro and Lavenhar 1989). At doses not associated with metabolic saturation, fecal excretion is generally less important than urinary excretion for most phthalates.

Biliary excretion was shown to occur for a limited number of phthalate compounds, i.e. DMHP, DEHP, DIDP and DBP (Sato et al. 1984; Ikeda et al. 1980; Daniel and Bratt 1974; General Motors Research Laboratories 1983; Tanaka et al. 1978). Generally, bile contains the monoester (free or glucuronidated), which can be reabsorbed in the intestine. Biliary elimination of phthalates was demonstrated using bile-cannulated animals. In rats, Daniel and Bratt (1974) reported that after an oral dose of 2.6 mg/kg of labelled DEHP, 14% of the administered dose was recovered in bile after four days. In dogs, biliary excretion was detected one day post-dosing at 10% of administered dose after repeated oral administration (50 mg/kg/day) of DEHP (Ikeda et al. 1980).

Kluwe (1982) suggested that hepato-biliary excretion may saturate at high doses or that it may happen only at a given period of time after absorption. These hypotheses were based on results by Tanaka et al. (1978) and Daniel and Bratt (1974) for DBP and DEHP, respectively. The suggestion of delayed biliary excretion is based on the finding that only 10% of an intravenous dose of 50 mg/kg DBP was recovered in bile in 5 hours, in comparison to 44% of an oral dose of 60 mg/kg in 24 hours. Biliary excretion may be followed by further reabsorption in the intestine (and finally, urinary excretion of the reabsorbed part). Enterohepatic recirculation of DEHP is suggested by the observation that only 8% of an oral dose (gavage) of 1.0 g/kg DEHP was isolated from feces as DEHP metabolites (another explanation, less likely, would be that biliary excretion is not a major route of elimination in this dose range) (Kluwe 1982).

Differences in excretion related to species and age

Most toxicokinetic studies are conducted in male rats, but there is some information acquired on other rodents or primate species. There appears to be similarities among species for the metabolic pathways, resulting in the excretion of similar metabolites. However, there may be species-related differences in the importance of each metabolic route.

The studies allowing for interspecies comparisons were essentially conducted with DEHP. The metabolic pathways extensively described in the rat (hydrolysis to MEHP and further oxidative metabolism (ω-, (ω-1)- and β-oxidations) were also found to occur in other species (Albro et al. 1981; Albro et al. 1982; Lake et al. 1976; Rhodes et al. 1986; Short et al. 1987), including humans (Silva et al. 2006). A study conducted in marmoset indicated a urinary metabolite pattern qualitatively similar to that in the rat, but quantitatively different (marmoset excreted principally conjugated metabolites derived from ω-1 oxidation) (Rhodes et al. 1986).

The major difference in urinary excretion of phthalate metabolites seems to be the level of conjugation. Although conjugation facilitates excretion by increasing its hydrophilicity, the conjugation of DEHP metabolites is negligible in rats, while it is important in other species (Frederiksen et al. 2007). Among the six species (rats, mice, guinea pigs, monkeys, humans and hamsters) studied by Albro et al. (1982), all except rats excreted metabolites under conjugated forms. Monkeys appeared to be the best model for elimination of phthalates from humans since both have a similar pattern of urinary excretion (high excretion of MEHP and mostly conjugated metabolites) (Albroet al.1982). Egestadet al.(1996) imparts an additional precision on the form excreted in mice. The form is not only combined with glucuronide, but also with β-glucose, a phenomenon that is not observable in guinea pigs and (humans) infants.

The effect of age on urinary excretion of DEHP and metabolites was studied in rats aged 60 and 25 days administered labelled DEHP by gavage (Sjoberg et al. 1985b). The authors observed a decreased rate of excretion in the 60-day-old rats compared to younger rats (26 and 44% of radioactivity in urine within 72 hours, respectively). No unchanged DEHP or MEHP was found in urine.

Inhalation route

There is limited information on medium-chain phthalate absorption via inhalation. In humans, an occupational study demonstrated that DEHP can be absorbed through the lungs (Dirven et al. 1993a). These authors measured DEHP concentrations in the air by personal air sampling of nine workers in a PVC boot factory and found these individuals were exposed to a maximum of 1.2 mg/m³ DEHP. They were able to demonstrate an increase in the urinary concentrations of all four metabolites of DEHP measured in the workers.

Dermal route

Absorption

A summary of in vitro and in vivo dermal absorption fluxes, Kp's, and % absorbed for medium-chain phthalates is presented in tables 3 and 4, respectively.

Data obtained from in vivo and in vitrostudies have shown that short-chain phthalates have higher absorption through rat and human skin than longer-chain phthalates (Scott et al. 1987; Elsisi et al. 1989; Mint and Hotchkiss 1993; Mint et al. 1994). Data obtained in vitro show a decrease in steady state absorption rates and extent of absorption, as the molecular weight and the lipophilicity of phthalates increase (Mint and Hotchkiss 1993; Mint et al. 1994; Payan et al. 2001). In vivo studies, conducted in rats, also observed that the extent of absorption (based on urinary excretion and retention in tissues) increases with increasing molecular weight and lipophilicity, and reaches a maximum with DBP. It then decreases as the molecular weight and lipophilicities increase (Elsisi et al. 1989).

In vivo data obtained in humans by Janjua et al. (2007) and Janjua et al. (2008) have also shown that DBP has a slower rate of absorption than DEP (based on urinary excretion and serum samples), suggesting a possible relationship with molecular weight or side chain length in humans. In this two-week study conducted in humans (26 healthy Caucasian males), subjects received whole-body topical applications of control basic cream formulation (dermal load: 2 mg/cm²), once per day for five consecutive days, followed by five daily topical applications of the same cream containing 2% (V/V) DEP and 2% (V/V) DBP (as well as 2% butyl paraben). Blood and urine were collected during the study and analyzed for levels of MEP and MBP. Two hours after the first application of the cream containing DEP, serum concentration of MEP peaked at 1001 µg/L (corresponding to 6.9 mg) and decreased to 23 µg/L after 24 hours just before the following application. The total percent of DEP absorbed from blood MEP concentrations is approximately 10%. Maximum dermal absorption for DBP from blood concentration could not be evaluated since concentration of MBP peaked over a longer period of time, and the authors started collecting blood at less frequent times (every 1 hour for 4 hours vs. every 24 hours subsequently). However, over the whole period of data collection (120 hours), serum concentrations of MEP were consistently higher than serum concentrations of MBP, indicating that DBP is probably absorbed though skin at less than 10%. In urine, the average dermal absorption for DEP and DBP, estimated from daily recovery of MEP and MBP, was 5.8 and 1.82%, respectively. However, significant interindividual and daily variations were observed, with a maximum dermal absorption in volunteers corresponding to approximately 13 and 6% of the applied DEP and DBP dose, respectively (Janjua et al. 2007, 2008; NICNAS 2011).

In vitro experiments conducted with rat and human epidermises have also shown that human skin is less permeable than rat skin to phthalate diesters (Table 3). Therefore, the use of rats as a model for dermal phthalate absorption in humans may overestimate dermal bioavailability.

Distribution

Distribution after dermal exposure to medium-chain phthalates was studied in vivo in rats and guinea pigs; retention in the skin was also documented in in vitro studies (diffusion cells). These studies show that skin may constitute a reservoir and that, similar to oral administration, phthalates are distributed throughout the body at varying levels according to the compound, the applied dose and the species.

Phthalate diesters (5-8 mg/cm²; skin not washed after exposure) applied topically to the dorsal side of rats revealed that part of the dose remained at the site of application (i.e., retained in the skin) (Elsisi et al. 1989). For all diesters, distribution in tissues after seven days was generally low (less than 1% in each tissue), except for BBP (4.6% in muscle) and DEHP (1.1% in skin other than at the site of application and 1.1% in muscle). The ranking of phthalate diester distribution in tissues was muscle greater than skin greater than fat for DIBP, DEHP and BBP, and skin greater than muscle greater than fat for DBP.

Dermal retention of DBP was also studied in vitro with rat and human skin. The results confirmed that skin may play the role of a reservoir for these diesters, and showed that retention in skin was 3- to 6-fold higher in rats compared to humans. With DBP, half of the applied dose (54%) remained on the surface of human skin, compared to 42% of rat skin. The fraction present in the skin was 4% in human skin and 21% in rat skin, respectively (Mint and Hotchkiss 1993; Mint et al. 1994).

Distribution of DBP, after dermal administration, was well documented by Payan et al. (2001). The authors found that after application of ¹⁴C-DBP (10 μL/cm²) on rat skin, DBP penetrated rapidly and diffused into the stratum corneum and/or epidermis, which constituted a reservoir. From this reservoir, DBP was slowly hydrolyzed by skin esterases before reaching systemic circulation. Less than 2% of unchanged DBP was present in the plasma of male haired rats, while MBP and MBP-glucuronide accounted for 61-88% of plasma radioactivity. Apparent plasma elimination of ¹⁴C was slightly lower in male than in female haired rats, and radioactivity in plasma decreased 3-fold faster in hairless male rats than in haired male rats. It is also noteworthy that in hairless male rats, the fraction of the applied dose remaining in the carcass and skin (less than 5%) was lower than in haired male rats (14-18%).

An in vivo study conducted in female hairless guinea pigs administered DEHP (34 nmol/cm²; skin washed 24 hours after exposure) indicated that seven days after administration, 5% of the applied dose was present in the dosed area of the skin and 4% was present in other body tissues (Ng et al. 1992). The authors also conducted an in vitro study with higher doses (35, 153 and 313 nmol/cm²) applied on the skin of guinea pigs in a diffusion cell (receptor fluid: HHBSS). They reported higher skin retention (about 41%, 38% and 36%, respectively) 24 hours after application and following skin washing. Use of non-viable skin resulted in a lack of metabolism.

Dermal absorption of medium-chain phthalates for risk assessment

As presented above, medium-chain phthalates are absorbed through the skin of rodents, rabbits and humans, but shorter-chain phthalates have higher rates of absorption through rat and human skin than longer-chain phthalates. Recent in vivo and in vitro studies have also shown that absorption of medium-chain phthalates, such as DBP and DEHP, through human skin is lower than through animal skin. This difference could be explained by species differences, such as difference in skin permeability as demonstrated in in vitro studies, and/or other factors related to the different methodology used in the various studies. Considering the maximum percentage of the applied dose recovered in serum for DEP-a short-chain phthalate that is expected to be more dermally available than other medium-chain phthalates-in a study conducted in humans (Janjua et al. 2008), it is expected that the dermal bioavailability for medium-chain phthalates in humans is not likely to be greater than 10%. Additionally, dermal absorption of many medium-chain phthalates (DBP, DIBP, BBP and DEHP) has also been evaluated by various other agencies (Danish EPA, ECHA, NICNAS). A summary of dermal absorption values assigned and main rationales is presented in Table 5. Overall, dermal absorption for these medium-chain phthalates has been proposed to be 10% or less by these agencies.

Given the lack of available data on dermal absorption for some medium-chain phthalates (B84P, B79P), it is therefore proposed that dermal absorption for these diesters is assumed to be 10%. The assignment of 10%, as a default for B84P and B79P, is based on Janjua et al. (2008), which showed a maximal dermal absorption of approximately 10% for DEP and less than 10% for DBP (a medium-chain phthalate with lower molecular weight and log Kow than B84P and B79P) in humans. This default of 10% is also reinforced by the assignment of a dermal absorption of 10% and less, by other agencies, for other medium-chain length phthalates, such as DIBP, DBP, DEHP and BBP (see above and Table E-5).

For DIBP, rat in vivo data shows that this substance may be absorbed at approximately 50%. However, DIBP is an isomer of DBP and, as mentioned above, Janjua et al. (2008) found that dermal absorption is less than 10% for DBP. Additionally, DIBP has been assigned a dermal absorption of 10% by other agencies (see Table E-5). Therefore, given the two considerations above, it is proposed that DIBP is dermally absorbed at a maximum of 10%.

Top of Page

Appendix I: Supporting information of the chronic toxicity and carcinogenicity of BBP

Chronic toxicity and carcinogenicity data available for BBP has been summarized previously in a Priority Substances List (PSL) Assessment Report published by Environment Canada and Health Canada (2000). Detailed information is available below.

A carcinogenicity bioassay was conducted by the NTP (1982) in F344 rats. Fifty rats per sex per group were administered BBP via diet, at levels of 0, 6000 or 12 000 ppm (0, 300 and 600 mg/kg bw/day, respectively, using a dose conversion by Health Canada [1994]). Females were exposed for 103 weeks. Because of poor survival, all males were sacrificed at weeks 29–30; this part of the study was later repeated (NTP 1997a).

Only females were examined histopathologically. The incidence of mononuclear cell leukemias was increased in the high-dose group (p = 0.011); the trend was significant (p = 0.006) (the incidences for the control, low- and high-dose groups were 7/49, 7/49 and 18/50, respectively). The incidence in the high-dose group and the overall trend remained significant (p = 0.008 and p = 0.019, respectively) when compared with historical control data. The NTP concluded that BBP was "probably carcinogenic for female F344/N rats, causing an increased incidence of mononuclear cell leukemias" (NTP 1982).

However, these results were not repeated in the two-year dietary study in F344/N rats recently completed by NTP (1997a). The average daily doses (reported by the authors) were 0, 120, 240 or 500 mg/kg bw/day for males and 0, 300, 600 or 1200 mg/kg bw/day for females. The protocol included periodic hematological evaluation and hormonal assays and a 15-month interim sacrifice.

There were no differences in survival between exposed groups and their controls (NTP 1997a). A mild decrease in triiodothyronine concentration in the high-dose females at 6 and 15 months and at termination was considered to be related to a non-thyroidal disorder. Changes in hematological parameters were sporadic and minor. In this bioassay, there was no increase in the incidence of mononuclear cell leukemias in female rats, as was reported in the earlier bioassay (NTP 1982), although the level of exposure (600 mg/kg bw/day) at which the incidence was observed in the early bioassay was common to both studies.

At the 15-month interim sacrifice, the absolute weight of the right kidney in the females at 600 mg/kg bw/day and the relative kidney weight in all exposed males were significantly greater than in controls. The severity of renal tubular pigmentation in high-dose males and females was greater than in controls, at both 15 months and 2 years. The incidence of mineralization in kidney in low- and high-dose females at 2 years was significantly less than in controls; severity decreased in all groups of exposed females. The incidence of nephropathy was significantly increased in all groups of exposed females (34/50, 47/50, 43/50 and 45/50 in the control, 300, 600 and 1200 mg/kg bw/day groups, respectively) (see Table 2). The incidence of transitional cell hyperplasia (0/50, 3/50, 7/50 and 4/50 in the control, 300, 600 and 1200 mg/kg bw/day groups, respectively) was significantly increased at 600 mg/kg bw/day (NTP 1997a).

At final necropsy, the incidences of pancreatic acinar cell adenoma (3/50, 2/49, 3/50 and 10/50 in the control, 120, 240 and 500 mg/kg bw/day groups, respectively) and pancreatic acinar cell adenoma or carcinoma (combined) (3/50, 2/49, 3/50 and 11/50 in the control, 120, 240 and 500 mg/kg bw/day groups, respectively) in the high-dose males were significantly greater than in the controls and exceeded those in the ranges of historical controls from NTP two-year feeding studies. One carcinoma was observed in a high-dose male; this neoplasm had never been observed in the historical controls. The incidence of focal hyperplasia of the pancreatic acinar cell in the high-dose males was also significantly greater than in the controls (4/50, 0/49, 9/50 and 12/50 in the control, 120, 240 and 500 mg/kg bw/day groups, respectively). Two pancreatic acinar cell adenomas were observed in the high-dose females (NTP 1997a).

The incidences of transitional epithelial papilloma of the urinary bladder in female rats at two years were 1/50, 0/50, 0/50 and 2/50 in the control, 300, 600 and 1200 mg/kg bw/day groups, respectively (NTP 1997a).

The authors concluded that there was "some evidence of carcinogenic activity" in male rats based on the increased incidences of pancreatic acinar cell adenoma and of acinar cell adenoma or carcinoma (combined). There was "equivocal evidence of carcinogenic activity" in female rats based on the marginally increased incidences of pancreatic acinar cell adenoma and of transitional cell papilloma of the urinary bladder (NTP 1997a).

The NTP (1997b) has released a technical report of a study that compared outcomes when chemicals were evaluated under typical NTP bioassay conditions as well as under protocols employing dietary restriction. The experiments were designed to evaluate the effect of dietary restriction on the sensitivity of bioassays towards chemical-induced chronic toxicity and carcinogenicity, and to evaluate the effect of weight-matched control groups on the sensitivity of the bioassays. BBP was included in the protocol; the results were summarized as follows:

Butyl benzyl phthalate caused an increased incidence of pancreatic acinar cell neoplasms in ad libitum-fed male rats relative to ad libitum-fed and weight-matched controls. This change did not occur in rats in the restricted feed protocol after two years. Butyl benzyl phthalate also caused an increased incidence of urinary bladder neoplasms in female rats in the 32-month restricted feed protocol. The incidences of urinary bladder neoplasms were not significantly increased in female rats in any of the two-year protocols, suggesting that the length of the study, and not body weight, was the primary factor in the detection of this carcinogenic response.

Fifty B6C3F₁ mice per sex per group were exposed to 0, 6000 or 12 000 ppm BBP (0, 780 and 1560 mg/kg bw/day, respectively, using a dose conversion by Health Canada, 1994) via diet for 103 weeks (NTP 1982). Approximately 35 tissues were examined histopathologically. The only compound-related sign of exposure was a dose-related decrease (statistical significance not specified) in body weight in both sexes. Survival was not affected, and there was no increased incidence of any neoplasm that was compound-related. As well, non-neoplastic changes were all within the normal limits of incidence for B6C3F₁ mice. The NTP concluded that, under the conditions of the bioassay, BBP "was not carcinogenic for B6C3F₁ mice of either sex."

Top of Page

Appendix J: Description and Application of the Downs and Black Scoring System and Guidance for Level of Evidence of An Association

Evaluation of study quality

A number of systematic approaches for assessing the quality of epidemiologic studies were identified and evaluated. The Downs and Black method was selected based on (1) its applicability to the phthalate database, (2) applicability to multiple study designs, (3) established evidence of its validity and reliability, (4) simplicity, (5) small number of components, and (6) epidemiologic focus. Downs and Black consists of a checklist of 27 questions broken down into the following five dimensions: 1) reporting; 2) external validity; 3) internal validity study bias; 4) internal validity confounding and selection bias; and 5) study power. Overall study quality is based on a numeric scale summed over the five categories. The range of the scale allows for more variability in rating study quality. The 27 questions are applicable to observational study designs including case-control, cohort, cross-sectional, and randomized controlled trials.

Studies retained for assessment were scored for quality using the Downs and Black tool. As previously mentioned, the Downs and Black allows for a range of scores from 27 questions, and each epidemiological study design has a maximum score (the maximum score for cohort studies is 21, case-control studies 18, and cross-sectional studies 17). Studies were divided into quartiles based on the scoring distribution for each study design; the distribution of scores for cohort, case-control and cross-sectional studies appears in Figure J-1. The average scores for cross-sectional and case-control studies were 13.1, whereas cohort studies had higher scores than both other study designs with an average score of 14.4.

Figure J-1 - Distribution of Downs and Black scores by study design

Guidance for levels of evidence of an association

The potential for an association between phthalate exposure and each health outcome was assessed based on strength and consistency as well as the quality of the epidemiology studies, as determined by the Downs and Black scores. Descriptions of the levels of evidence of association are as follows:

1. Sufficient evidence of an association: Evidence is sufficient to conclude that there is an association. That is, an association between exposure to a phthalate or its metabolite and a health outcome has been observed in which chance, bias and known confounders could be ruled out with reasonable confidence. Determination of a causal association requires a full consideration of the underlying biology/toxicology and is beyond the scope of this document.
2. Limited evidence of an association: Evidence is suggestive of an association between exposure to a phthalate or its metabolite and a health outcome; however, chance, bias or confounding could not be ruled out with reasonable confidence.
3. Inadequate evidence of an association: The available studies are of insufficient quality, consistency or statistical power to permit a conclusion regarding the presence or absence of an association.
4. Evidence suggesting no association: The available studies are mutually consistent in not showing an association between the phthalate of interest and the health outcome measured.

Top of Page