Pentachlorophenol (PCP) Risk Assessment and Cancer Risk

Pentachlorophenol (PCP)

RoseyFields conducts pentachlorophenol (PCP) risk assessments using the following methodology, which tracks the critical study selection, dose conversion, uncertainty factor application, and toxicity value derivation underlying the current noncancer and cancer assessments for this chemical, and describes how those values are applied in practice.

1. Chemical Identity, Regulatory Background, and Exposure Sources

Pentachlorophenol (PCP; CASRN 87-86-5) is a chlorinated phenol that was widely used in the United States as a wood preservative, particularly for utility poles, railroad ties, and fence posts, and historically also saw use as a general-use biocide, herbicide, defoliant, and mold inhibitant. Commercial PCP is sold in two principal grades: technical-grade PCP (tPCP, typically ~90% purity) and analytical-grade PCP (aPCP, higher purity), a distinction that matters toxicologically because technical-grade material carries higher levels of contaminants — chiefly chlorinated dibenzo-p-dioxins and dibenzofurans — that can influence the severity of toxic and carcinogenic responses observed in some studies. Because PCP is well absorbed via the oral, inhalation, and dermal routes, occupational exposure historically occurred primarily among sawmill workers, wood treaters, and other tradespeople through dermal contact and inhalation of treated-wood dust or vapor, while the general population has historically been exposed at much lower levels through contact with treated wood, contaminated soil and groundwater near treatment facilities, and, to a lesser extent, food and drinking water.

PCP has been the subject of EPA toxicity assessment since 1987, when it was first entered into the Integrated Risk Information System (IRIS) database with an oral reference dose based on a chronic rat study (Schwetz et al., 1978). The chemical’s regulatory profile was substantially revised in 2010, when EPA updated both the noncancer reference dose and the cancer assessment based on a fuller toxicological database, including a chronic dog study that became the new principal study for the noncancer value (Mecler, 1996) and a re-derivation of the cancer slope factor using benchmark dose modeling rather than the earlier linear extrapolation from tumor incidence alone. RoseyFields’ methodology below reflects this current (2010) EPA toxicity value derivation, which remains the operative basis for PCP risk assessment absent a more recent EPA update.

2. Point-of-Departure (POD) Selection

2.1 Noncancer POD: The Mecler (1996) Chronic Dog Study

The current noncancer point of departure is drawn from a 52-week chronic toxicity study in which technical-grade PCP (90.9% purity) was administered by gelatin capsule to four beagle dogs per sex per dose at 0, 1.5, 3.5, or 6.5 mg/kg-day (Mecler, 1996). This study was selected as the principal study over the earlier rat study (Schwetz et al., 1978) because it identified hepatotoxicity at substantially lower doses and provided a more complete characterization of the dose-response relationship for liver injury, including interim histopathology, clinical chemistry, and organ-weight measurements at 13, 26, 39, and 52 weeks.

The dog study identified clear, dose-related hepatic injury. At the highest dose (6.5 mg/kg-day), one male and one female were euthanized in extremis due to hepatic insufficiency, with histologic evidence of hepatocellular swelling and degeneration, fibrosis, bile duct hyperplasia, and cirrhosis-like changes. At all three doses tested, relative liver weight was significantly increased in both sexes (14-66% in males and 37-94% in females across the 1.5-6.5 mg/kg-day range), and serum alkaline phosphatase (ALP) was elevated 1.9- to 6.8-fold across doses and sexes. Alanine aminotransferase (ALT) was near control levels at 1.5 mg/kg-day but rose to 2.8- to 3.1-fold above control at 3.5 mg/kg-day and 3.9- to 8.8-fold above control at 6.5 mg/kg-day. Microscopically, hepatocellular pigmentation, cytoplasmic vacuolization, and chronic inflammation increased in incidence and severity with dose.

The study authors identified a NOAEL of 3.5 mg/kg-day and a LOAEL of 6.5 mg/kg-day based on overt morphologic liver effects. EPA, however, determined that the dose-related progression of hepatic lesions across all three tested doses, together with the severity of effects and morbidity observed at the top dose, warranted treating the lowest dose tested, 1.5 mg/kg-day, as the LOAEL, based on dose-related increases in hepatocellular pigmentation, cytoplasmic vacuolation, and chronic inflammation, together with statistically significant increases in relative liver weight (and, in females, absolute liver weight) and increased serum enzyme activity at that dose. Because hepatic effects were present at every dose level tested, a NOAEL could not be identified from this study, and the POD used for RfD derivation is therefore a LOAEL of 1.5 mg/kg-day.

This finding is corroborated by a large body of supporting literature identifying the liver as the primary noncancer target of PCP across species. Chronic oral studies in rats show hepatic pigment accumulation, chronic inflammation, and cystic degeneration (Schwetz et al., 1978; NTP, 1999), while chronic studies in mice show necrosis, cytomegaly, chronic active inflammation, and bile duct lesions, with more severe liver toxicity than in rats at comparable doses, potentially reflecting species differences in PCP biotransformation (NTP, 1989). Kidney pigment deposition in the proximal convoluted tubules has also been observed in rats (Schwetz et al., 1978), and decreased spleen weight has been reported in mice (NTP, 1989), rats (Bernard et al., 2002), and calves (Hughes et al., 1985).

2.2 Cancer POD: NTP (1989) Mouse Bioassay

The cancer POD is derived from a 2-year dietary carcinogenicity bioassay in male B6C3F1 mice exposed to technical-grade PCP (tPCP) at average daily doses of 0, 18, or 35 mg/kg-day (NTP, 1989). This study identified statistically significant, dose-related increases in hepatocellular adenomas or carcinomas (7/32 in controls versus 26/47 and 37/48 at the low and high doses, respectively) and adrenal benign or malignant pheochromocytomas (0/31 in controls versus 10/45 and 23/45 at the low and high doses), both with statistically significant trends by Cochran-Armitage test and statistically significant differences from control by Fisher’s exact test at both dose levels. Because PCP produced tumors at multiple sites, EPA used a bootstrap resampling procedure (Efron and Tibshirani, 1993) to combine the liver and adrenal tumor response into a single joint dose-response analysis, generating simulated incidence data at each dose group and refitting the multistage model to each simulated dataset to characterize the distribution of the combined benchmark dose. The POD for the cancer slope factor — the LED10, the 95% lower confidence bound on the dose associated with 10% extra cancer risk — was estimated at 0.25 mg/kg-day (central estimate ED10 of 0.34 mg/kg-day) from this combined analysis.

3. Pharmacokinetic Modeling and Dose Conversion

Unlike some chemical assessments that rely on physiologically based pharmacokinetic (PBPK) models to convert administered animal doses into internal (blood or tissue) dose metrics before extrapolating to humans, the current PCP noncancer assessment does not employ a PBPK model. EPA explicitly noted that a description of PCP toxicokinetics in either dogs or humans is limited, and in the absence of adequate data to quantify species-specific absorption, distribution, metabolism, and excretion differences, the interspecies extrapolation was instead handled entirely through a default 10-fold interspecies uncertainty factor (discussed in Section 4 below) rather than through a mechanistic dose-conversion model. Reported LOAEL and NOAEL values for the dog study are therefore expressed and applied directly as administered (external) doses in mg/kg-day, without a body-weight scaling adjustment for the noncancer POD.

For the cancer assessment, dose conversion between mice and humans was addressed using a default cross-species body-weight scaling factor of BW3/4 (three-quarters power scaling), applied under the assumption that equivalent lifetime average daily doses scaled to body weight raised to the three-quarter power produce equivalent risks across species. This scaling approach, standard EPA practice for cancer dose-response extrapolation, was used to translate the mouse LED10 into a human-equivalent dose for slope factor derivation, rather than a PBPK model calibrated to PCP-specific toxicokinetic parameters. RoseyFields’ practice notes that this represents a source of uncertainty distinct from chemical-specific data limitations: because BW3/4 scaling is a generic default rather than a PCP-calibrated model, it does not capture any PCP-specific differences in absorption or metabolism between mice and humans, and analysts should treat cross-species dose conversion for PCP as a source of residual, non-quantified uncertainty in both the noncancer and cancer values.

4. Uncertainty Factor Derivation

The chronic oral RfD was derived by applying a total uncertainty factor (UF) of 300 to the LOAEL of 1.5 mg/kg-day identified in the Mecler (1996) dog study. Each component factor and its rationale is summarized below.

4.1 Interhuman (Intraspecies) UF — UFH = 10

A full 10-fold factor was applied to account for variability in susceptibility across the human population, including sensitive subgroups, because current data are inadequate to quantitatively characterize human-to-human variability in PCP toxicokinetics or toxicodynamics.

4.2 Interspecies UF — UFA = 10

A full 10-fold factor was applied to account for uncertainty in extrapolating from dogs to humans. As discussed in Section 3, because no PBPK model or other quantitative interspecies dosimetric adjustment was available or applied for the noncancer endpoint, this default factor was used to cover both toxicokinetic and toxicodynamic interspecies differences in full, rather than a reduced factor covering toxicodynamics alone (which would be typical practice when a dosimetric adjustment factor has already addressed the toxicokinetic component).

4.3 LOAEL-to-NOAEL UF — UFL = 3

Because a NOAEL could not be identified in the principal study (hepatic effects were observed at every dose tested, including the lowest), a LOAEL-to-NOAEL extrapolation factor was required. EPA applied a reduced factor of 3 (rather than the full default of 10) because the effects defining the LOAEL at 1.5 mg/kg-day — increased incidence of hepatocellular pigmentation, minimal chronic inflammation, and increased relative liver weight, accompanied by less-than-twofold changes in serum ALT and ALP — were judged to be of minimal toxicological significance, i.e., a comparatively mild LOAEL rather than one defined by frank, severe toxicity.

4.4 Subchronic-to-Chronic UF — UFS = 1 (not applied)

No factor was applied for extrapolation from subchronic to chronic exposure duration because the principal study (Mecler, 1996) was itself a chronic (52-week) exposure protocol, so this extrapolation step was unnecessary.

4.5 Database Uncertainty Factor — UFD = 1 (not applied)

No additional factor was applied for database deficiencies. EPA’s rationale was that the PCP database is unusually complete for a legacy chemical: it includes human epidemiological data, chronic toxicity studies in rats, mice, and dogs, subchronic studies across multiple species, dedicated neurological, endocrine, reproductive, and developmental toxicity studies, and a two-generation reproductive toxicity study (Bernard et al., 2002), such that no additional default penalty for missing study types was warranted.

4.6 Composite UF

The composite uncertainty factor is the product of the four applied components: UF = UFH (10) × UFA (10) × UFL (3) = 300. (UFS and UFD both equal 1 and do not change the product.) This total UF of 300 is applied to the LOAEL of 1.5 mg/kg-day to derive the chronic oral RfD, as described in Section 5.

5. Derivation of the Reference Dose (RfD) and Reference Concentration (RfC)

The chronic oral RfD is calculated as the POD divided by the composite UF:

RfD = 1.5 mg/kg-day ÷ 300 = 0.005 mg/kg-day.

This value supersedes the prior (1987) RfD of 0.03 mg/kg-day, which had been based on a NOAEL of 3 mg/kg-day for liver and kidney pigmentation in a chronic rat study (Schwetz et al., 1978) with a UF of 100. The six-fold reduction in the current RfD relative to the 1987 value reflects both the lower LOAEL identified in the more sensitive dog study and the larger composite UF (300 versus 100), driven primarily by the addition of the LOAEL-to-NOAEL extrapolation factor, which was not needed under the earlier NOAEL-based derivation.

EPA rated overall confidence in this RfD as medium, with medium confidence in the principal study and high confidence in the supporting database. The medium study confidence reflects the small group sizes in the Mecler (1996) study (four dogs/sex/dose) and the limited resolution of the dose-response curve at low doses, since incidence of the key liver lesions rose from 0% in controls to 100% at the lowest dose tested and remained at 100% at the two higher doses — meaning the study could bound a LOAEL but could not itself establish where a true no-effect threshold lies. High database confidence reflects the breadth of supporting acute, subchronic, chronic, developmental, reproductive, and carcinogenicity data across species described in Sections 2 and 7.

No inhalation RfC has been derived for PCP. EPA’s assessment concluded that adequate chronic or subchronic inhalation toxicity data in animals are not available to support quantitative derivation of an RfC, so this value remains a data gap rather than a quantitative output; RoseyFields’ methodology therefore relies on the oral RfD, adjusted through route-to-route extrapolation assumptions where an inhalation exposure pathway must be evaluated for a specific site, rather than on a chemical-specific inhalation toxicity value.

6. Carcinogenicity Classification and Cancer Slope Factor

6.1 Weight-of-Evidence Classification

Under EPA’s 2005 Guidelines for Carcinogen Risk Assessment, PCP is classified as “likely to be carcinogenic to humans” by all routes of exposure. This classification, which updates the chemical’s earlier B2 (“probable human carcinogen”) classification from 1991, rests on four converging lines of evidence: (1) oral carcinogenicity studies in mice showing hepatocellular adenomas and carcinomas and adrenal pheochromocytomas in both sexes, and hemangiomas/hemangiosarcomas in females (NTP, 1989); (2) evidence of carcinogenicity in male rats, showing malignant mesotheliomas and nasal squamous cell carcinomas (Chhabra et al., 1999; NTP, 1999); (3) human epidemiological evidence of increased risk of non-Hodgkin’s lymphoma and multiple myeloma, some evidence of soft tissue sarcoma, and limited evidence of liver cancer (Demers et al., 2006; Hardell et al., 1995, 1994; Kogevinas et al., 1995); and (4) mechanistic evidence of hepatocellular tumor-promoting activity (Umemura et al., 1999, 2003a, 2003b) and lymphoma- and skin-adenoma-promoting activity in mice (Chang et al., 2003). Because tumors were observed at sites remote from the point of oral contact (liver, adrenal gland, circulatory system, nose) and because PCP is well absorbed by all exposure routes, EPA applied the classification to all routes of exposure rather than limiting it to oral exposure alone, notwithstanding the absence of inhalation- or dermal-route carcinogenicity data.

The human epidemiological evidence is anchored by the British Columbia sawmill worker cohort (Demers et al., 2006), which used a validated cumulative dermal chlorophenol exposure score and found a graded exposure-response relationship for non-Hodgkin’s lymphoma and multiple myeloma, together with at least a doubling of liver cancer risk across most exposure categories. Supporting case-control evidence for non-Hodgkin’s lymphoma and soft tissue sarcoma comes from a set of Swedish studies (Hardell et al., 1995, 1994; Kogevinas et al., 1995), although a later case-control study conducted after PCP had been banned in Sweden did not find an association (Hardell and Eriksson, 1999), a discrepancy potentially attributable to a shorter latency period between the more recent exposures and disease onset. EPA’s assessment separately examined whether the observed cancer associations could be attributable solely to dioxin/furan contaminants in technical-grade PCP rather than PCP itself, and concluded this was unlikely: specific furan congeners are not elevated in the blood of PCP workers relative to the general population (Collins et al., 2007), the cancer risk patterns in dioxin-exposed cohorts (Steenland et al., 2004) differ from those seen in PCP-exposed cohorts, and the PCP associations with specific cancers are generally stronger than corresponding associations with dioxin or other chlorophenol exposures in studies measuring both (Demers et al., 2006; Kogevinas et al., 1995).

6.2 Oral Slope Factor Derivation

The current oral slope factor is 4 × 10⁻¹ per mg/kg-day, superseding the 1991 value of 1.2 × 10⁻¹ per mg/kg-day (which had been derived using simple linear extrapolation from pooled hepatocellular and hemangiosarcoma tumor incidence in female mice, NTP, 1989). The current value was derived using a multistage model fit to the combined (bootstrap-pooled) liver and adrenal tumor incidence data from male mice (NTP, 1989), as described in Section 2.2, with linear extrapolation from the LED10 POD of 0.25 mg/kg-day. The slope of linear extrapolation from the central estimate (ED10 = 0.34 mg/kg-day) is 0.1 ÷ 0.34 mg/kg-day = 2.9 × 10⁻¹ per mg/kg-day; slope factors for individual tumor sites considered separately in the male mouse ranged from 1.5 × 10⁻¹ to 2.9 × 10⁻¹ per mg/kg-day, and the combined bootstrap analysis was used specifically because relying on any single tumor site would likely understate PCP’s overall carcinogenic potential given the multiplicity of tumor types observed. EPA cautions that this slope factor should not be applied to exposures exceeding the POD of 0.25 mg/kg-day, since above that dose the fitted multistage model, rather than the linear low-dose extrapolation, better characterizes the known dose-response relationship.

For drinking water applications, EPA derived a unit risk of 1 × 10⁻⁵ per µg/L, and, assuming default water consumption of 2 L/day by a 70-kg adult, calculated corresponding drinking water concentrations of 9 µg/L, 0.9 µg/L, and 0.09 µg/L associated with upper-bound excess lifetime cancer risks of 1 in 10,000, 1 in 100,000, and 1 in 1,000,000, respectively. No inhalation unit risk (IUR) is available for PCP, consistent with the absence of inhalation toxicity data noted in Section 5.

Supporting mechanistic data for the cancer classification include genotoxicity findings showing that, while PCP has not produced mutations in prokaryotic test systems, it can induce oxidative DNA and protein damage in subcellular systems and human cells in vitro, along with gene mutation and recombination in fungi, clastogenic effects in mammalian systems, and weak evidence of transplacental mutation in mice; PCP’s metabolite tetrachlorohydroquinone (TCpHQ) has similarly been shown to cause oxidative DNA damage in vitro and in vivo. Equivocal cytogenetic findings (sister chromatid exchange, chromosomal aberrations) have been reported in PCP-exposed workers (Ziemsen et al., 1987; Bauchinger et al., 1982; Schmid et al., 1982), and immune system effects — including mitogen-response stimulation in low-dose rat studies (Blakley et al., 1998) and impaired lymphocyte mitogen response in exposed humans (Daniel et al., 1995) — together with porphyria observed in both PCP-exposed workers (Cheng et al., 1993) and animals (Kimbrough and Linder, 1978), add biological plausibility to the epidemiological cancer signal.

7. Practical Application by RoseyFields Risk Assessors

RoseyFields applies the RfD and cancer slope factor described above within standard EPA risk assessment methodology to evaluate site-specific or product-specific PCP exposures for both regulatory submissions and litigation support.

7.1 Hazard Quotient and Hazard Index (Noncancer Risk)

For noncancer effects, RoseyFields calculates a chronic daily intake (CDI, in mg/kg-day) for the exposure scenario at issue — reconstructed from site data (soil, groundwater, or air concentrations), exposure frequency and duration, body weight, and relevant intake or contact rates — and divides it by the RfD of 0.005 mg/kg-day to obtain a Hazard Quotient (HQ = CDI ÷ RfD). An HQ at or below 1 indicates the estimated exposure is not expected to result in appreciable risk of the critical effect (hepatotoxicity) or other noncancer effects captured within the RfD’s built-in uncertainty margin; an HQ exceeding 1 signals that the margin of exposure relative to the POD is inadequate and warrants further evaluation. Because the same source (e.g., contaminated groundwater or soil) may present PCP exposure through multiple pathways (ingestion, dermal contact, inhalation of volatilized or particulate-bound PCP), RoseyFields sums pathway-specific HQs into a Hazard Index (HI) for the exposure scenario as a whole, consistent with EPA’s default assumption of dose additivity across exposure routes for a shared critical effect (hepatotoxicity). Because the RfD is built on a LOAEL rather than a NOAEL, and study confidence in the principal study is only medium, RoseyFields’ practice is to treat HQ values modestly above 1 with particular caution given the comparatively mild nature of the LOAEL effects and the limited dose-response resolution at low doses noted in Section 5, rather than treating 1.0 as a bright-line threshold.

7.2 Excess Lifetime Cancer Risk (ELCR)

For carcinogenic risk, RoseyFields multiplies the reconstructed lifetime average daily dose (LADD, in mg/kg-day) by the oral slope factor (4 × 10⁻¹ per mg/kg-day) to estimate excess lifetime cancer risk (ELCR = LADD × SF), consistent with the linear low-dose extrapolation approach underlying the slope factor, and confirms that the reconstructed dose does not exceed the POD of 0.25 mg/kg-day, above which the slope factor is not considered valid for extrapolation. For drinking water exposure pathways specifically, RoseyFields can instead apply the drinking water unit risk (1 × 10⁻⁵ per µg/L) directly to a measured or modeled water concentration, or reference the associated benchmark concentrations (9, 0.9, and 0.09 µg/L for 10⁻⁴, 10⁻⁵, and 10⁻⁶ risk levels, respectively) as a rapid screening comparison before undertaking a full site-specific dose reconstruction.

7.3 Exposure and Dose Reconstruction

Because litigation and regulatory matters frequently involve historical exposures (e.g., decades-old wood-treatment operations or legacy contamination), RoseyFields’ dose reconstruction work typically integrates historical use records, environmental sampling or fate-and-transport modeling, and, where available, biomonitoring data, to estimate exposure concentrations and durations for the specific population or individual at issue. Given the absence of a PCP-specific PBPK model (Section 3), internal dose is not reconstructed from blood or urine biomarkers using a mechanistic toxicokinetic model; instead, RoseyFields’ reconstructions rely on external (administered) dose estimates paired with the administered-dose-based RfD and slope factor, consistent with how both values were themselves derived.

7.4 Litigation and Regulatory Support

In litigation contexts, RoseyFields uses this methodology to support or evaluate causation and damages analyses in toxic tort and environmental contamination matters, including expert reports addressing whether reconstructed exposures plausibly exceeded levels associated with appreciable noncancer or cancer risk, and to respond to opposing experts’ dose-response assumptions. In regulatory contexts, the same RfD and slope factor inform screening levels, remediation goals, and permit conditions at PCP-contaminated sites (for example, former wood-treatment facilities), and support comment submissions or petitions where a party seeks to challenge or defend a proposed cleanup standard. Because the current toxicity values reflect a substantial 2010 revision (a six-fold reduction in the RfD and an increase in the slope factor relative to pre-2010 values), RoseyFields’ practice is to confirm, for any given matter, which vintage of toxicity value was in effect at the time relevant regulatory decisions or exposures occurred, since retrospective application of the current values to historical regulatory decisions can be a significant point of dispute.

8. Key Uncertainties and Sensitive Populations

Several uncertainties bear directly on how RoseyFields’ risk assessors should qualify conclusions drawn from these toxicity values in a specific matter.

Limited low-dose resolution in the principal study. As noted in Section 5, the key liver lesions in the Mecler (1996) study were already at or near 100% incidence at the lowest dose tested, so the true shape of the dose-response curve below 1.5 mg/kg-day is not empirically characterized; the RfD’s reliance on a LOAEL-to-NOAEL factor of 3 (rather than the full default of 10) reflects EPA’s judgment that the LOAEL effects were mild, but this remains a judgment call rather than an empirically resolved threshold.

No chemical-specific interspecies dosimetric adjustment. Because no PBPK model was available for the dog-to-human extrapolation, the full default 10-fold interspecies UF was applied without any offsetting toxicokinetic adjustment, which may be conservative or non-conservative depending on whether dogs and humans in fact handle PCP similarly — a question the assessment could not resolve for lack of data (Section 3).

Reproductive, developmental, endocrine, and immune sensitivity. Although these outcomes were not selected as the critical effect for RfD derivation, several study findings suggest they occur at doses not greatly above the hepatotoxic LOAEL and may be relevant to specific sensitive subpopulations. A two-generation reproductive study found decreased fertility, delayed puberty, testicular effects, decreased litter size and viability, and decreased pup weight at 30 mg/kg-day, coincident with parental systemic toxicity (Bernard et al., 2002), while developmental studies in rats identified fetal sensitivity at doses below those causing maternal toxicity (5-13 mg/kg-day in fetuses versus 30-43 mg/kg-day in dams) (Schwetz et al., 1974; Welsh et al., 1987). Thyroid hormone suppression (decreased T3 and T4) has been observed across several species at doses as low as 1-3 mg/kg-day (Jekat et al., 1994; Hughes et al., 1985; Beard et al., 1999a, 1999b; Beard and Rawlings, 1998; Rawlings et al., 1998), and immune suppression (impaired humoral and cell-mediated immunity) has been reported in mice at doses as low as 2-4 mg/kg-day (Kerkvliet et al., 1985a, 1985b; Exon and Koller, 1983) and in humans with blood PCP levels exceeding 10-20 µg/L (Daniel et al., 1995; McConnachie and Zahalsky, 1991). None of these outcomes independently drove the RfD, but they inform RoseyFields’ judgment about which populations — pregnant individuals, children, and those with pre-existing thyroid or immune conditions — may warrant additional scrutiny in a case-specific weight-of-evidence analysis, particularly where reconstructed exposures approach or exceed these lower-bound effect levels.

Neurotoxicity findings at comparatively low doses. Motor activity and startle-response changes have been observed in mice at doses as low as 36-102 mg/kg-day (NTP, 1989), and morphological nerve changes have been reported in rats following prolonged drinking-water exposure (Villena et al., 1992), with some biochemical brain effects considered transient in another study (Savolainen and Pekari, 1979). Because most neurotoxicity data were generated using technical-grade material or PCP of unspecified purity, it is possible that some reported neurotoxic effects reflect co-occurring dioxin/furan contaminants rather than PCP itself, a distinction RoseyFields flags when neurotoxicity is at issue in a specific matter.

Contaminant confounding in the carcinogenicity database. As discussed in Section 6, EPA concluded that dioxin/furan contamination in technical-grade PCP is unlikely to fully explain the observed human cancer associations, but this remains an area of residual uncertainty, particularly for site-specific matters where the historical grade and contaminant profile of the PCP used may differ from that studied in the key epidemiological cohorts.

Route-to-route extrapolation for inhalation exposure. Given the absence of an inhalation RfC or IUR (Sections 5 and 6.2), any RoseyFields evaluation of inhalation exposure pathways necessarily relies on route-to-route extrapolation from the oral toxicity values, introducing additional uncertainty beyond that already embedded in the oral RfD and slope factor themselves — a limitation that should be disclosed explicitly wherever inhalation exposure is a material pathway in a given matter.

References