LL2-04 hindsight check: Too much to swallow: PCE contamination of mains water (Ozonoff; panels by Rudén and Onasch), Late lessons from early warnings: science, precaution, innovation (EEA Report No 1/2013), Ch. 4, pp. 76–91#

Checked 25 September 2026. The check covers what happened between publication (January 2013) and September 2026 that bears on the chapter’s claims, evidence, predictions and recommendations. Page numbers are report pages (PDF page minus 2). All web sources were accessed on 25 September 2026.

Author context. The chapter’s author, David Ozonoff (Boston University), is a co-investigator on the Boston University epidemiology of the Cape Cod pipe-lining episode, before and after 2013. Two of those papers disclose that “At the request of the Commonwealth of Massachusetts, Dr. Ozonoff was a witness at the Johns-Manville Corporation bankruptcy hearing”. They also disclose that he has “on occasion, testified in personal injury cases involving exposure to tetrachloroethylene and trichloroethylene” (Spence et al., Environ Health, 2 June 2008). The chapter mentions neither. This does not make the account wrong: litigation is a plausible source of the documentary detail in the uncited Providence narrative. It does bear on how the uncited claims (Claims 2 and 4) should be weighed.

Annex 3 note. Not applicable. The chapter is new in the 2013 report and is not one of the 2001 cases updated in Annex 3.

Access note. The web-search quota for this session had run out before this check began. Everything below was verified by fetching primary sources directly: - the US Federal Register (API and full texts), eCFR, EPA IRIS, the NTP Report on Carcinogens, EUR-Lex and CARB; - PubMed/PMC abstracts and full texts; - Crossref, PubChem (used for EU and Japanese hazard classifications) and CourtListener.

Several sources could not be reached: - the IARC Volume 106 PDF (access denied); - the NCBI Bookshelf copy of Volume 106, which returned a CAPTCHA page (not bypassed); - mass.gov (403); - ATSDR/CDC PDFs (403); - the ECHA website (firewall).

Where a point depends on a secondary source, the text says so.

Overview#

The chapter’s central factual story has held up: an untested product, a missed anomaly, and long-running contamination of drinking water. The harm it could only gesture at is now much better documented. Its forward-looking claims have fared unevenly.

What has been vindicated or strengthened - The episode is better quantified. About 660 miles of vinyl-lined asbestos-cement (VL/AC) pipe went into 91 Massachusetts towns, and about 750 miles into roughly 100 communities across Massachusetts and Rhode Island. PCE levels in 1980 ranged from 1.5 to 7,750 µg/L, depending on flow. Remediation by flushing and bleeding aimed only at 40 µg/L, eight times today’s US limit, and in some towns contamination persisted “through the 1990s” (Aschengrau et al. 2016; Aschengrau et al. 2018). - The harm is now documented. Studies published since 2013 on people exposed through these pipes report associations between early-life exposure and: - drug-use disorder; - bipolar disorder and post-traumatic stress disorder (PTSD); - placental abruption and placenta-related stillbirth; - cleft lip and delayed time to pregnancy.

Many other outcomes showed no association. EPA’s 2024 rule cites this neurodevelopmental evidence. - Neurotoxicity is now the regulatory driver. EPA’s IRIS (2012), its TSCA risk evaluation (2020) and its 2024 rule all make chronic neurotoxicity the most sensitive endpoint. This matches the pre-1970 warnings the chapter emphasises: neurological effects at the lowest doses in Dow’s own work, and slow excretion and accumulation (p. 82). - Occupational limits still lag. The US OSHA limit is still 100 ppm, taken from a 1967 consensus standard. EPA’s 2024 workplace limit is 0.14 ppm. EPA’s 2022 risk determination quotes OSHA’s own description of its 1970s limits as “outdated and inadequate”. This strongly confirms the chapter’s point that occupational standards lag drinking-water standards (p. 88). - Assessors still disagree. Assessors reading the same PCE data continue to reach different verdicts, as Rudén’s panel argued (pp. 84–85): - the EU classifies PCE as a suspected carcinogen (Carc. 2); - Japan’s GHS classification puts it in Category 1B (presumed); - IARC rates it 2A (probable); - EPA calls it likely; - the US National Toxicology Program (NTP) calls it reasonably anticipated.

A 2022 systematic review by a litigation-linked consultancy concluded there was no PCE–lymphoma association, the reverse of the EPA, IARC and ATSDR readings. - The California phase-out happened. California’s air regulator (CARB) states that perc “will no longer be used in dry cleaning operations by January 1, 2023”, and its grant programme closed that day (Onasch panel, p. 87).

What was wrong, overstated or overtaken - The “on the cusp” prediction was wrong (p. 88). No major body has since declared PCE a confirmed human carcinogen. IARC’s October 2012 re-evaluation kept PCE at 2A and moved only its relative, TCE, to Group 1. EPA’s February 2012 IRIS assessment settled on “likely”. Both were public before the chapter appeared, and the chapter mentions neither. - “No sound precautionary strategy” (p. 88) was fair in 2013 but has been partly overtaken. The 2016 reform of US chemicals law (TSCA) produced, by December 2024, a federal rule that: - bans consumer uses; - bars new PCE dry-cleaning machines from June 2025 and ends PCE dry cleaning by December 2034; - sets a workplace limit roughly 700 times below OSHA’s.

The rule has not settled the matter. Within five months EPA told a court it would reconsider it. In July 2025 it asked whether a higher limit should apply, and in July 2026 it pushed key workplace deadlines to 2027. - “History will continue as in the past” (p. 88) is only partly borne out. The interpretive dispute continues. But action came anyway, through a statute with explicit decision rules and a non-cancer endpoint. Whether that action lasts now turns on regulatory politics more than on new science. - The narrative contains small slips. - The chapter’s dates (installation 1969–1979; “accidentally discovered” in 1976 after “over 700 miles” had been laid) differ from later accounts. Boston University papers give installation as May 1968 to March 1980. - They date regulators’ awareness to early 1980. - An EPA–state follow-up to the 1976–77 National Organics Monitoring Survey found the residual PCE (Larsen et al. 1983). - About 750 miles was the final total for two states, not the length laid by 1976. - The worst-case exposure is far above anything measured. The body-burden illustration (p. 83) assumes 125 ppm (125,000 µg/L). That is about 16 times the highest concentration later measured in the pipes, and several orders of magnitude above typical values.

How to read the lessons now. The mechanism lessons are the most durable: - monitoring blind to novel contaminants (p. 77); - anomalies explained away (p. 77); - durable infrastructure prolonging harm (p. 76); - occupational protection lagging environmental protection (p. 88); - assessment divergence on shared data (pp. 84–85).

Each has independent later support. The diagnostic claim is that science alone cannot close the dispute, and that explicit decision triggers matter (pp. 76, 88). Post-2013 events support it: action came through a legal trigger and a sidestep to a different endpoint, not through scientific consensus on cancer. The two uncited claims should be presented as the author’s informed account, not as established fact. These are the Johns-Manville conduct narrative (p. 77) and “artificially and purposefully creating doubt” (p. 88).

Claim-by-claim assessment#

Claim 1: PCE leached from PCE-applied linings in a “new and relatively untested” pipe installed 1969–1979; “accidentally discovered” in 1976 after more than 700 miles had been laid in New England; some supplies “still today require continuous remediation” (pp. 76–77)#

Original claim. The pipe was installed in 1969–1979 and “is now known to have caused widespread PCE contamination of water supplies in the US state of Massachusetts”. The chapter cites Demond 1982, MDEE 1982 and Larsen et al. 1983 (p. 77). “Not until 1976, when over 700 miles of this pipe had been installed in New England water distribution systems, was it accidentally discovered that PCE had been leaching” (p. 77). “Some public water supplies are still today contaminated with PCE and require continuous remediation” (p. 77). The chapter gives no exposure, health-outcome or cost figures (digest caveat).

Subsequent developments

Scale and dates. The Boston University group, with Ozonoff as co-author, has published consistent accounts since 2008: - Installation. VL/AC pipe was installed “From May 1968 through March 1980”. The liner was “cured” by drying for two days before delivery (Spence et al. 2008). - Length. “Approximately 660 miles of vinyl-lined asbestos-cement pipes (VL/AC) had been installed in 91 Massachusetts cities and towns” (Aschengrau et al., Ann Glob Health, 2016). Across two states, “approximately 750 miles of VLAC pipes had been installed in about 100 cities and towns” in Massachusetts and Rhode Island (Aschengrau et al., Environ Health, 6 Nov 2018). - The liner. It was a vinyl-toluene resin (Piccotex, Johns-Manville) sprayed as a slurry in PCE, on the assumption that “because PCE is volatile it was assumed it would evaporate completely during the drying process” (2016 review).

Discovery. The chapter’s 1976 date and the later papers’ 1980 date refer to different events: - The Boston University papers date the state’s awareness to early 1980. “When the Massachusetts Department of Environmental Protection (DEP) became aware of the problem in early 1980, regulators collected drinking water samples” (Spence 2008). “More than a decade lapsed before it was discovered” (Aschengrau 2018). - The 1983 journal account the chapter cites says that “in a follow-up to the National Organics Monitoring Survey, the USEPA and several New England states found residual PCE” (Larsen, Love and Reynolds, J AWWA, April 1983; abstract via Crossref). That survey ran in 1976–77, so a mid-1970s first detection through a general survey fits “accidentally discovered”.

The “over 700 miles” figure is close to the eventual two-state total, not the length in place in 1976.

Exposure levels. - Massachusetts samples from 1980 ranged “from 1.5 to 80 μg/l in medium and high flow locations and 1600 to 7750 μg/l in low flow locations”. In Rhode Island, eleven systems exceeded the 1980 action level of 40 µg/L (Aschengrau 2018). - A set of 88 historical pre-remediation samples had a median of 0.5 µg/L, a mean of 66 µg/L and a maximum of 2,432 µg/L. Almost half were non-detects (Spence 2008). - The current US limit (maximum contaminant level) is 5 µg/L.

Remediation. - Method. “Digging up and replacing the VL/AC pipes was prohibitively expensive, so a program of flushing and bleeding the water distribution system was instituted in the most problematic areas”. Some dead-end layouts were changed and some pipes replaced (2016 review; Spence 2008). - Target. The target was 40 µg/L, derived from an EPA advisory level. According to Spence, the state guidance that set it assumed “the problem was not a long-term one”. - Duration. “PCE contamination persisted in public water supplies of selected towns through the 1990s because the target level for remediation was 40 μg/L”. By 2018, “reported levels in MA and RI are now below” 5 µg/L (Aschengrau 2018). - Cost. No aggregate remediation cost has been located. - Current status. Whether any supply still needs continuous bleeding in 2026 could not be verified, because mass.gov was inaccessible.

Health outcomes. Before 2013 the Boston University case-control work had already reported an elevated leukaemia risk above the 90th exposure percentile (Aschengrau et al. 1993, adjusted OR 5.84, 95% CI 1.37–24.91). Since 2013 the Cape Cod Health Study and a Massachusetts–Rhode Island case-control study have reported the following: - Neurological and psychiatric outcomes. “The strongest associations were seen with illicit drug use, bipolar disorder, and post-traumatic stress disorder” (2016 review). In a 2020 analysis, drug-use-disorder criteria had an adjusted RR of 1.4 (95% CI 1.0–1.8) with no dose-response (Aschengrau et al., Environ Health, Sept 2020). - Reproductive outcomes. “delayed time-to-pregnancy, and increased risks of placental abruption, stillbirths stemming from placental dysfunction, and certain birth defects”. There were no associations with pregnancy loss, birth weight or gestational duration (Aschengrau et al., Environ Sci Process Impacts, 2020). - Birth defects. First-trimester exposure above 40 µg/L was associated with cleft lip (OR 3.8, 95% CI 1.2–12.3), with imprecise elevations for spina bifida and hypospadias (Aschengrau 2018). - Null findings. There was no association with obesity, diabetes, cardiovascular disease, hypertension or several vision outcomes. Imprecise elevations were reported for cancer and epilepsy (Aschengrau et al. 2015). - Outside the pipe episode. A 2026 meta-analysis of 21 studies of prenatal PCE exposure found a borderline association with spontaneous abortion (RR 1.28, 95% CI 1.00–1.63) and “very low to low” certainty across outcomes (Menon et al., Environ Res, Aug 2026).

Regulatory uptake. EPA’s December 2024 rule lists “neurodevelopmental outcomes from prenatal and early childhood exposure to PCE such as increased affinity of engaging in drug, alcohol, and tobacco use as a teen or adult” among PCE’s neurotoxic effects (89 FR 103560, 18 Dec 2024). That is the Cape Cod finding. The episode’s own epidemiology has thus fed into national regulation.

Verdict: partly held up. - The core facts hold, and the scale is somewhat larger than stated (two states, about 750 miles). - The health dimension, absent from the chapter, has been substantially strengthened by later work. - The dates need correcting: installation ran 1968–1980, and the regulatory discovery was in 1980. - No remediation cost is available. The “still today require continuous remediation” claim was not verifiable for 2026.

Implication for weight. The case can carry the lesson that durable infrastructure prolongs and hides harm (p. 76). One refinement comes from the record: an interim clean-up threshold set on the assumption that the problem was temporary extended exposure for another decade or more. The Cape Cod findings rest on modelled historical exposures and self-reported outcomes, and several associations are imprecise. They show plausible harm at real exposure levels, not a settled burden.

Claim 2: Johns-Manville saw odour and taste problems in 1966 and 1968 trials, blamed incomplete curing, and there is “no indication” it ever tested for residual PCE (p. 77)#

Original claim. The 1966 trial gave water “a slight chemical taste and odour”. In 1968 the air in the pipe smelled of chloroform or dry-cleaning fluid. Company representatives attributed this to a pipe “kept covered in brown paper and therefore not ‘cured’ completely”. “There is no indication that the taste and odour incident prompted Johns-Manville to investigate … whether PCE remained in the liner” (p. 77). The account is uncited.

Subsequent developments - No public record. No published judgment, public litigation record or regulatory history describing the Providence trials was found. CourtListener opinion searches for “vinyl-lined” with tetrachloroethylene, for “Piccotex”, and for “Johns-Manville” with tetrachloroethylene and pipe returned nothing relevant (CourtListener search). MDEE (1982) and Demond (1982) are not online. - Consistent later accounts. Later accounts by the Boston University group, with Ozonoff as co-author, agree with the chapter’s picture of an untested assumption that PCE would flash off during drying (2016 review; 2020 review). But they are not independent of the chapter’s author. - Larsen et al. (1983). Also cited by the chapter, this account describes “the reasons why its production was discontinued” and the remedies (flushing, continuous bleeders). Only the abstract was accessible, and it does not mention the Providence trials. - The exposure model. Spence et al. (2008) built their exposure model on “Johns Manville specifications for the perchloroethylene suspension (30% Piccotex® and 70% PCE)” and on an estimate that “6% of PCE remained in the liner at installation”. The paper does not say whether Johns-Manville knew of that residual at the time. - Author’s role. As noted above, the author testified for Massachusetts at the Johns-Manville bankruptcy hearing (Spence 2008 disclosure). That testimony is the likely route to the documentary detail, and it was not disclosed in the chapter.

Verdict: unclear. Nothing found contradicts the account, but nothing independent confirms it either.

Implication for weight. The mechanism lesson survives as a well-formed hypothesis: routine tests that cannot see a new kind of contaminant, and an anomaly explained away rather than investigated (p. 77). The Johns-Manville specifics should be cited as “according to Ozonoff (2013), drawing on records from the Massachusetts proceedings”, not as established history.

Claim 3: Enough pre-1970 evidence existed to warrant caution; harm was foreseeable, “not merely a statement based on hindsight” (pp. 79–83)#

Original claim. Several strands of pre-1970 evidence argued for caution: - hookworm-therapy side effects and individual susceptibility (pp. 79–80); - occupational poisonings within “accepted” limits (pp. 80–81); - Dow’s own findings of slow excretion, accumulation and neurological effects at the lowest exposures (p. 82); - Smyth et al. (1969) on more-than-additive toxicity (p. 82); - recognition since the 1940s–50s that solvents threaten groundwater (p. 83).

A worst case of 125 ppm in water and 2 L a day implies a body burden of about 1 g, “close to the dose used to treat hookworm” (p. 83). Harm was therefore foreseeable (p. 83).

Subsequent developments - No new historiography. No post-2013 historical study re-examining the pre-1970 PCE literature was located, and the chapter’s historical sources were not re-checked. - Neurotoxicity confirmed as the lead hazard. The strand the chapter most emphasises, chronic low-level neurological effects, is now the hazard regulators treat as most important: - EPA’s IRIS assessment (last updated 10 Feb 2012) bases its reference dose and concentration on “Nervous, Ocular” critical effects from occupational studies of reaction time, cognition and colour vision (EPA IRIS, tetrachloroethylene). - EPA’s IRIS authors: “Neurotoxicity was identified as a sensitive noncancer health effect, occurring at low exposures: a conclusion supported by multiple studies” (Guyton et al., EHP, Feb 2014). - The 2024 TSCA rule: “The most sensitive health effect driving the unreasonable risk of PCE and selected as the basis for this rule is neurotoxicity from chronic exposure” (89 FR 103560). - Harm at real exposures. The Cape Cod findings (Claim 1) suggest harm at actual pipe exposures, well below the chapter’s worst case. - The worst case overshoots. The arithmetic is internally correct: 250 mg a day divided by an elimination constant of 0.25 per day gives about 1,000 mg. But the concentration it assumes is far above anything measured. At the highest measured 1980 value (7,750 µg/L), the same model gives a steady-state burden of about 60 mg. At the 40 µg/L remediation threshold it gives about 0.3 mg. The “close to the hookworm dose” comparison therefore does not describe any documented exposure, although the chapter labels it a worst case.

Verdict: partly held up. The qualitative foreseeability argument stands. Later science has strengthened it by confirming the very endpoint (chronic neurotoxicity) that the pre-1970 literature flagged, and by documenting effects at real exposures. The quantitative illustration overstates plausible exposure by more than an order of magnitude. “Foreseeable” remains an interpretive judgement that post-2013 sources neither test nor dispute.

Implication for weight. Several lessons are well supported: - acceptable risk belongs to a use and its beneficiaries, not to the substance (pp. 80, 83); - familiarity and beneficial use breed a false presumption of safety (pp. 78, 81–82); - susceptibility, accumulation and synergy defeat simple threshold thinking (pp. 79–82).

Use the body-burden arithmetic only with the measured concentrations beside it.

Claim 4: No evidence Johns-Manville concealed PCE information (unlike asbestos); indifference is the plausible explanation; after detection it denied any health hazard to avoid damages (pp. 83–86)#

Original claim. “there is no such evidence in relation to PCE … the proposition that Johns-Manville was merely indifferent to these dangers is a plausible explanation” (p. 84). “once the problem was detected, it denied that there were any health hazards — a necessary position if it were to avoid paying damages” (pp. 83–84). “nobody made them care” (p. 86).

Subsequent developments - No new evidence either way. No post-2013 documentary evidence on Johns-Manville’s knowledge of residual PCE was found, whether of concealment or of indifference. - Consistent with a mistaken assumption. The Boston University accounts describe a production assumption (PCE would evaporate during two-day curing), which fits the indifference or negligence reading better than concealment. - Liability. Johns-Manville’s 1982 bankruptcy, triggered by asbestos, framed the liability question. The author’s testimony at that hearing (Spence 2008) confirms that the Commonwealth pursued the matter there. No public record of the outcome for PCE claims was located. - Liability as a lever. The chapter’s recommendation was that liability is the “acid test” of what makes a company care (p. 86). The later record adds a parallel case of after-the-fact liability for solvent-contaminated drinking water, the Camp Lejeune Justice Act of 2022 (Pub. L. 117-168, s. 804). That case involved TCE and also PCE: reconstructed PCE at the Tarawa Terrace plant peaked at a monthly average of 183 µg/L (Maslia et al., Water, Oct 2016). It supports the claim that liability tends to arrive decades after exposure; it says nothing about Johns-Manville.

Verdict: unclear. The claim is not contradicted, and it is carefully hedged (“plausible”). No independent evidence has emerged.

Implication for weight. The general lesson carries moderate weight. Indifference, without concealment, is sufficient for harm when nobody with standing has reason to look, and only levers that change a firm’s incentives work (pp. 84, 86). The Johns-Manville specifics rest on one informed participant’s reading.

Claim 5: PCE is “now on the cusp of being declared a confirmed human carcinogen in some major national markets” (p. 88)#

Original claim. Quoted in full above (p. 88). The chapter also quotes Karstadt (1998) on IARC’s raising of the PCE evaluations in Volume 63 (1995) to sufficient evidence in animals and limited evidence in humans (p. 88).

Subsequent developments. Status as of September 2026:

The two decisions most relevant to “on the cusp”, EPA’s IRIS (February 2012) and IARC’s re-evaluation (October 2012, Lancet Oncology summary December 2012), were both public before the report appeared in January 2013. Both stopped short of “confirmed”, and the chapter cites neither.

Verdict: overturned. Thirteen years on, no major regulator or classifier has moved PCE to “known” or “confirmed”. The upgrade the chapter may have had in view happened to TCE, not PCE.

Implication for weight. Do not cite this passage as foresight. It is better read as evidence for the chapter’s other claim: PCE’s carcinogenicity verdict is stuck in the “probable/likely/suspected” band while evidence accumulates slowly (Claims 8 and 10). Regulation came through the non-cancer endpoint instead (Claim 6).

Claim 6: Forty years after the Massachusetts pipes, “a sound precautionary strategy for continued exposure to PCE has still not been initiated”; occupational standards lag drinking-water standards (p. 88)#

Original claim. “Thus 40 years after the hard lesson of the water mains in Massachusetts, a sound precautionary strategy for continued exposure to PCE has still not been initiated” (p. 88). “fairly strict community drinking water standards have been established, although occupational standards have lagged behind”, partly because of “a weak labour movement” (p. 88).

Subsequent developments

United States: new law, a rule, then reconsideration. - Designation (2016). The June 2016 Lautenberg amendments to TSCA required risk evaluations. PCE was among the first ten chemicals designated (81 FR 91927, 19 Dec 2016). - Risk evaluation (2020). It found unreasonable risk for specific conditions of use (85 FR 82474, 18 Dec 2020). - Revised determination (2022). It found unreasonable risk for PCE “as a whole chemical substance” and dropped the assumption that workers always wear protective equipment. In part this was because “many of OSHA’s chemical-specific permissible exposure limits largely adopted in the 1970’s are described by OSHA as being ‘outdated and inadequate for ensuring protection of worker health’” (87 FR 76481, 14 Dec 2022). - Final rule, 18 December 2024 (89 FR 103560; effective 17 Jan 2025): - prohibits consumer uses; - phases out dry cleaning over ten years. Use in machines acquired after 16 June 2025 is prohibited, third-generation machines are barred after 20 December 2027, and all dry-cleaning use ends after 19 December 2034; - applies a Workplace Chemical Protection Program to most remaining uses, which together cover more than 80% of production volume. Its core is an existing-chemical exposure limit (ECEL) of 0.14 ppm (8-hour time-weighted average), based on chronic neurotoxicity.

EPA estimated annualised costs of USD 43.43 million, and monetised benefits of USD 32.6–84.6 million from cancer reductions alone (2% discount rate). - Air rule (January 2025). EPA closed its Clean Air Act technology review of dry-cleaning emissions without amendment, “given the recently finalized action under [TSCA] which has instituted a 10-year phaseout” (90 FR 1041, 7 Jan 2025). - Court challenge. Petitions for review were consolidated in the Fifth Circuit as FabriClean Supply v. EPA, No. 25-60006, filed 6 Jan 2025 and still open. Petitioners include the American Chemistry Council, Olin Corporation and state chemistry councils, and also the NGO Center for Environmental Health (CourtListener docket). - Reconsideration. On 12 May 2025 EPA told the court it would reconsider the rule. Its July 2025 request for comment asks specifically whether the 0.14 ppm limit should give way to “the acute non-cancer exposure limit of 0.50 ppm … and the lifetime cancer exposure limit of 0.47 ppm”. It also asks which uses might be moved from prohibition to workplace controls, and about industrial dry cleaning (90 FR 35858, 30 July 2025). - Delayed deadlines. A final rule of 28 July 2026 extended the non-federal deadlines (91 FR 47145):

EPA “intends to publish a separate Notice of Proposed Rulemaking to potentially amend aspects of the PCE Final Rule in the future”. No such proposal appears in the Federal Register as of 25 September 2026.

United States: occupational and drinking-water limits. - Occupational. OSHA’s limit remains 100 ppm (8-hour time-weighted average), with a 200 ppm ceiling and a 300 ppm peak. Its source is ANSI Z37.22-1967 (29 CFR 1910.1000, Table Z-2, eCFR as of Aug 2026). That is about 700 times EPA’s ECEL. - Drinking water. The limit remains 5 µg/L. In 2024 EPA dropped PCE and TCE as candidates for revision, abandoning the carcinogenic volatile organic compounds (cVOC) group-rule plan that dated from 2010. It gave as reasons “resource limitations, competing workload priorities … as well as limited potential health benefits” (89 FR 59623, 23 July 2024).

California. California’s perc phase-out in dry cleaning was completed on 1 January 2023 (Claim 9).

European Union. - Occupational limit. An indicative occupational exposure limit of 20 ppm (138 mg/m³, 8 hours) and 40 ppm (15 minutes), with a skin notation, was set by Commission Directive (EU) 2017/164 of 31 January 2017 (EUR-Lex). - Drinking water. The limit remains 10 µg/L for the sum of PCE and TCE (Directive (EU) 2020/2184, Annex I Part B; EUR-Lex). - Dry cleaning. No EU-wide ban on PCE in dry cleaning was identified. A 2021 review states that “None of the EU countries have banned the use of PERC in dry cleaning”. It reports that France, where over 90% of shops used PERC, was phasing out PERC machines in residential buildings by 2022 (Ceballos et al., Front Public Health, Mar 2021; secondary source). EU restriction activity under the REACH regulation could not be checked because the ECHA site was blocked.

Verdict: partly held up. - The occupational-lag claim has been strengthened: OSHA’s 1967-derived limit remains, and a sister agency has formally called it inadequate. - The “no sound strategy” claim was accurate in 2013. It was then partly overtaken: the United States adopted a comprehensive phase-out and exposure-limit rule in 2024. - That rule is under reconsideration and its deadlines are slipping, so whether a durable strategy exists remains unclear. - The drinking-water standards the chapter called “fairly strict” have not been tightened. EPA declined to revise them in 2024.

Implication for weight. The chapter’s explanation of the lag, political salience with a weak labour movement set against high-profile childhood-cancer cases (p. 88), fits the later record but was not tested by any source found. The post-2013 US sequence supports the chapter’s institutional diagnosis: protection followed a new statutory trigger, not new science. It also adds a lesson the chapter did not draw: protective rules can be reopened within months when administrations change.

Claim 7: The chemical industry has been “artificially and purposefully creating doubt” and is “buying extra time” (p. 88)#

Original claim. “The chemical industry has been active and aggressive in countering new information through the strategy of artificially and purposefully creating doubt and uncertainty in the minds of decision-makers … the industry is essentially buying extra time (and creating continuing exposure and disease) by this strategy” (p. 88). The chapter gives no citation.

Subsequent developments. The documentary evidence found since 2013 shows sustained, industry-resourced counter-science and litigation that follow the lines the chapter describes. None of it documents intent for PCE specifically. - Rodent tumours dismissed as irrelevant to humans. A 2024 mode-of-action review co-authored by the Science Director of the Halogenated Solvents Industry Alliance (HSIA), “a trade association that represents producers and users of PCE”, concludes that mouse liver tumours arise through PPARα activation and that “human liver cancer risk to PCE is not likely” (Klaunig, Bevan and Gollapudi, Toxicol Ind Health, Mar 2024). This re-runs an argument already considered and rejected. In the 2010 National Research Council (NRC) peer review of EPA’s draft, a dissent took that position, “However, the NRC peer-review committee as a whole did not support these conclusions”. EPA judged “Evidence is insufficient for the hypothesized MOAs evaluated: PPARα activation” (Guyton et al. 2014). This is the pattern the chapter describes: “sophisticated toxicological arguments as to why PCE could be a carcinogen in rodents but not a carcinogen in humans” (p. 88). - Lymphoma. A 2022 systematic review by the consultancy Gradient concluded that “the evidence does not support an association between PCE exposure and NHL”, contrary to EPA, IARC and ATSDR. Its disclosure states: “Gradient has conducted work on PCE in general, and this topic specifically, in the context of litigation and TSCA assessment reviews … This review has not been registered and no protocol was prepared” (Goodman et al., Glob Epidemiol, June 2022). Independent commentators did not dismiss it: - three academic epidemiologists (Boston University, Stanford, NYU) called it “a strong improvement over the current reviews on the same topic” while urging quantitative bias analysis (Fox et al., Oct 2022); - Savitz used it to illustrate legitimate “key choices in evidence synthesis” (Savitz, Oct 2022). - TCE heart defects. Researchers from HSIA, Dow, Olin and Exponent published a TCE drinking-water study reporting no heart defects. Its purpose was to rebut a study that EPA had relied on (DeSesso et al., Birth Defects Res, June 2019; comment by Runyan et al. and reply in the same issue). - Litigation and reconsideration. Industry petitioners challenged the 2024 rule (Claim 6), and the 2025 reconsideration invites higher exposure limits. Both are lawful forms of participation. The reconsideration has so far pushed non-federal workplace compliance dates back by six to about 18 months.

Verdict: partly held up. - The behavioural claim holds and is better documented now. Industry-funded or litigation-linked science has continued to press the arguments the chapter predicted, and compliance dates have slipped. - The intent claim (“artificially and purposefully”) remains asserted, not shown. No internal documents were located. - Some industry-linked work has been judged methodologically serious by independent epidemiologists. So the line between manufactured doubt and genuine disagreement is not as clean as the chapter implies.

Implication for weight. Use the technology-neutral mechanism, which is well supported: uncertainty favours the better-resourced party, because a party with high stakes can keep funding alternative readings (p. 86). Present intent as the author’s allegation. The fairest way to state what these episodes show is that sponsor interests shape which questions get studied and how evidence is synthesised, and that this is visible in disclosures. That is a documented pattern, not a proven strategy of deception.

Claim 8: Rudén panel on divergent risk assessments of TCE and PCE; the remedy is transparent evidence-weighting and consistent terminology (pp. 84–85)#

Original claim. Of 29 TCE carcinogenicity assessments (1973–97), 8 of the 10 “animal risk only” conclusions came from international organisations or industry, and 8 of the 9 “human risk from animal data” conclusions came from government or academia. Divergence persisted even in 1995–96, when all assessors had the same data. IARC (1995) read the PCE non-Hodgkin lymphoma (NHL) studies as positive; ECETOC (1999), the industry-funded European Centre for Ecotoxicology and Toxicology of Chemicals, read them as negative. “Evaluators must communicate better about the approach they use to evaluate the strength of evidence … [and] employ clear and consistent terminology” (p. 85).

Subsequent developments

TCE converged. - IARC Group 1 (2012/2014). - EPA IRIS “carcinogenic to humans” (2011). - NTP “Known to be a human carcinogen” (15th Report on Carcinogens, Dec 2021). - EU Carc. 1B (H350).

For TCE, the divergence Rudén documented narrowed decisively after 2011.

PCE did not converge. - Classifications still span four categories: EU “suspected” (Carc. 2), Japan “presumed” (1B), IARC “probable” (2A), EPA “likely”, NTP “reasonably anticipated” (see Claim 5). - The NHL question the panel singled out is still unresolved. - EPA’s 2012 assessment judged bladder cancer and NHL to have “the strongest databases” (Guyton 2014). - IARC’s 2012 “limited evidence” rested mainly on bladder cancer (Vlaanderen 2014). - A Nordic cohort found “some evidence indicative of an excess risk of cancer of the liver and NHL” (Vlaanderen et al., OEM, 2013). - A US National Cancer Institute case-control study found PCE “not associated with NHL” (Callahan et al., OEM, 2018). - Gradient’s 2022 review found no association. - Weed’s 2022 commentary framed the whole episode as a question of “the reliability of causal claims” when “different investigators apply these methods to the same evidence and come up with different causal assessments” (Weed, Glob Epidemiol, Oct 2022). - Endpoint choice is now itself the dispute. EPA’s 2025 reconsideration asks whether to regulate on chronic neurotoxicity (0.14 ppm), acute effects (0.50 ppm) or cancer (0.47 ppm) (Claim 6). - The sector pattern persists. Industry-linked and trade-association authors continue to reach the “animal risk not relevant to humans” and “no human association” readings, while government and academic assessors mostly do not. This fits Rudén’s sector pattern. No formal post-2013 re-tabulation for PCE was found.

The transparency remedy has been widely adopted in method, not in outcome. Since 2013, evidence-weighting has been formalised in: - the National Academies review of IRIS (NASEM 2014); - EFSA’s weight-of-evidence guidance (EFSA Journal 2017;15(8):4971); - IARC’s revised Preamble (Samet et al., JNCI, 2020); - Rudén’s own group’s SciRAP tool for study evaluation (Beronius et al., J Appl Toxicol, 2018); - EPA’s TSCA systematic-review protocol, itself reviewed critically by the National Academies (NASEM 2021).

The Gradient review and its critics show that systematic methods relocate the disagreement to quality-scoring and causal-criteria choices rather than eliminate it. Savitz: algorithmic assessments’ “drawbacks include a failure to incorporate the nuances of study methods”.

Verdict: held up. The core observation, that the same data yield different verdicts depending on assessment policy and institution, is strongly confirmed for PCE through 2026. It has been overcome for TCE, where human evidence strengthened. The recommended remedy has been adopted widely, but it has not produced convergence on PCE.

Implication for weight. This is one of the chapter’s strongest transferable lessons (digest insight 8: strong for divergence, moderate for cause). The PCE–TCE contrast adds a refinement: divergence closes when a decisive new body of human evidence arrives (TCE and kidney cancer), not through better procedures alone. Where human evidence stays weak, better procedures move the dispute rather than end it.

Claim 9: Onasch panel: wet cleaning can handle 100% of garments; USD 2,749 annual savings and a payback of under 4.5 years in one Massachusetts shop; about 85% of US cleaners use PCE; California phasing out PCE machines by 1 January 2023; slow adoption without regulation (p. 87)#

Original claim. As stated above (p. 87), including “over 150 dedicated wet cleaners operate in California”. The panel also claims that “the shift to wet cleaning from solvent-based cleaning has been slow, especially where regulations phasing out solvent use do not exist”.

Subsequent developments - California. CARB states that under the 2007 amendments “Perc will no longer be used in dry cleaning operations by January 1, 2023” (CARB programme page; About). The staged deadlines were: no new perc machines from 1 January 2008; removal from co-residential sites and of machines 15 years or older by 1 July 2010; and “Perc machines may no longer be used” from 1 January 2023 (CARB Notice 2015-2). The USD 10,000-per-shop AB 998 grants for water-based and CO₂ systems ran from 2004 and “ended on January 1, 2023” (CARB AB 998 page). Compliance was not independently audited in the sources found. - Nationally, PCE share and substitution. - The ~85% figure was probably dated by 2013. A 2021 review gives ~20,600 US dry-cleaning shops (2017), with “60–65% of dry cleaners use PERC as their primary solvent … and most of the remainder use a high-flashpoint hydrocarbon” (Ceballos et al. 2021). - EPA’s 2024 rule estimated that “6,000 dry cleaners still use PCE, a majority of which are small businesses”. It expected only about 60 PCE machines to remain in use by the end of the phase-out, because “the purchase of new PCE dry cleaning machines has been in decline” (89 FR 103560). - The main substitute has therefore been hydrocarbon solvent, not wet cleaning. CARB itself notes limited health data for hydrocarbons, uterine tumours in rats from the siloxane D5, and toxicity concerns for a glycol-ether solvent. - One substitute, n-propyl bromide, became a textbook “regrettable substitution”, with neurological illness in New Jersey dry cleaners (Ceballos 2021). - Wet cleaning where it was promoted. - In King County, Washington, a 2010 survey found that 69% of shops still used PERC and that cost was the main barrier. With reimbursement, 27 of the approximately 65 remaining PERC shops had switched to professional wet cleaning by October 2020. The programme stopped funding solvent alternatives because of hazardous-waste and flammability concerns (Ceballos 2021). - Minneapolis became “the first PERC-free city” in January 2018 (Ceballos 2021, secondary). - No independent replication of the single-shop Bellingham cost figures was found. Commenters on EPA’s rule “noted the viability of professional wet cleaning” (89 FR 103560). - Europe. “60–90% of dry cleaning shops use PERC, depending on the country”, and no EU country had banned it by 2021 (Ceballos 2021, secondary).

Verdict: held up. The California phase-out took effect on schedule. Adoption outside regulated jurisdictions was slow and heavily dependent on grants. Regulation (California, Minneapolis, EPA’s 2024 ban on new machines) drove change. Two qualifications apply: the market mainly shifted to other organic solvents rather than to wet cleaning, and the 85% share was already high for 2013.

Implication for weight. The lesson that viable substitutes spread slowly without regulation or subsidy (digest insight 12) is now moderately well supported by several jurisdictions, not one shop. The record adds a caution the panel did not state: a phase-out that does not specify the replacement tends to produce substitution to the cheapest drop-in alternative, whose hazards may be poorly characterised.

Claim 10: “the history of PCE will continue in the future as it has in the past”, because interpretation is “irresolvable within science itself” (pp. 76, 88)#

Original claim. “continued argument over how to interpret the scientific evidence is irresolvable within science itself … there are no overarching criteria from the philosophy of science that can force a solution. Whether the problem is a failed duty of care or a lack of clarity about what evidence will trigger action, the history of PCE will continue in the future as it has in the past. The science has not been hidden. It has been ineffective in guiding and catalysing action” (p. 88). The remedy: “Mechanisms are needed to force the production, sharing and publication of information about exposure and effects” plus duty-of-care requirements (p. 88).

Subsequent developments - The pattern was broken in 2024, but not by resolving the cancer dispute. EPA acted on a non-cancer endpoint, chronic neurotoxicity (colour vision, cognition), for which it judged the human evidence adequate. It acted under a statute (TSCA 2016) that: - forces risk evaluation on a schedule; - forbids weighing cost when determining risk; - requires attention to “potentially exposed or susceptible subpopulations”.

That is close to what the chapter asked for: “clarity about what evidence will trigger action” and forced production of exposure information (p. 88). - Part of the evidential basis came from the pipe episode itself. The Cape Cod neurodevelopmental findings are among the effects EPA listed. - California acted earlier, on a statutory trigger. CARB’s listing of perc as a toxic air contaminant in 1991 required it to consider controls. That trigger led to the 2007 phase-out decision. - The argument continues, on different ground. Within months EPA reopened the choice of endpoint and limit (0.14 vs 0.47 vs 0.50 ppm). Industry petitions and the reconsideration cite feasibility and monitoring limits, not only science. The reconsideration followed a change of administration. The 2025 notice grounds it partly in a deregulatory executive order (EO 14219), not in new toxicological evidence. - The cancer question itself remains in the band the chapter described. Mixed NHL studies, contested mode-of-action arguments and slowly accumulating bladder-cancer evidence all persist (Claims 5, 7 and 8).

Verdict: partly held up. The diagnosis that the interpretive dispute is irresolvable by science alone held: cancer verdicts still diverge, and endpoint choice is now openly contested. The prediction that nothing would change was overturned in the United States, and in California before it. In both places action came through explicit legal triggers, and in the federal case through a less contested endpoint. How long that action lasts is uncertain.

Implication for weight. The strongest reading of the chapter is prescriptive rather than predictive. It argued that decision triggers must be made explicit because evidence will not settle the matter (digest insight 9), and post-2013 events are consistent with this. Two refinements follow from the record: - Choice of endpoint. Action can proceed on whichever adverse effect has the clearest human evidence, without waiting for the most feared effect to be confirmed. - Durability of triggers. A trigger is only as durable as the institutional settlement behind it.

Cross-cutting observations for using this section as a lens#

1. The event produced its own evidence. The pipes created a “natural experiment” (Aschengrau 2016). Epidemiology built on it became part of the national regulatory case 40 years later (89 FR 103560). The mechanism in technology-neutral form: when a deployed product exposes a population in an irregular, recorded pattern, the records of the deployment become the evidence base. This happened here only because regulators kept detailed water-system records (2016 review), which supports the chapter’s call to force the production of exposure information (p. 88).
2. Interim thresholds lengthen exposure. The 1980 remediation target of 40 µg/L was set on the assumption that “the problem was not a long-term one” (Spence 2008). It let contamination persist into the 1990s. This extends the chapter’s durable-infrastructure point (p. 76).
3. Action came through a different endpoint, not a settled dispute. Decades of argument over whether PCE causes cancer did not produce action; a statutory trigger applied to neurotoxicity did. This supports the chapter’s claim that science alone cannot close the argument (pp. 76, 88). It also shows that when one question is stuck, rules can be justified on a less contested adverse effect.
4. Occupational protection lags, and the lag is officially admitted. OSHA’s 1967-derived 100 ppm limit stands beside EPA’s 0.14 ppm. EPA’s 2022 determination quotes OSHA calling its own limits “outdated and inadequate” (p. 88 vindicated).
5. Who bears the transition cost. Both EPA and the 2021 review stress that US dry cleaners are mostly small, immigrant-owned businesses; about 80% of the workforce identifies as a racial or ethnic minority (Ceballos 2021). The costs of both the hazard and the phase-out fall on them. The chapter’s alternatives panel (p. 87) does not address this distributional dimension.
6. Author positionality. The chapter’s author also contributed to the downstream epidemiology and gave expert evidence in the episode’s legal proceedings (Spence 2008 disclosure). His insider knowledge probably underlies the most vivid, and uncited, parts of the narrative (pp. 77, 83–86). Those parts should be flagged as a participant’s account.
7. Slips to correct when citing: - installation “1969–1979” should be May 1968 to March 1980; - “over 700 miles” by 1976 should be about 750 miles in total by 1980, across Massachusetts and Rhode Island; - discovery was “1976” (first detections following the EPA organics survey) and 1980 (the state’s systematic sampling and remediation); - the “on the cusp” prediction (p. 88) ignored the 2012 IARC and EPA decisions; - the “85%” PCE share (p. 87) was probably out of date by 2013.

Sources#

US federal regulatory documents (Federal Register, eCFR, IRIS) - EPA. Designation of Ten Chemical Substances for Initial Risk Evaluations Under TSCA. 81 FR 91927, 19 Dec 2016. https://www.federalregister.gov/documents/2016/12/19/2016-30468/designation-of-ten-chemical-substances-for-initial-risk-evaluations-under-the-toxic-substances - EPA. Perchloroethylene (PCE); Final TSCA Risk Evaluation; Notice of Availability. 85 FR 82474, 18 Dec 2020. https://www.federalregister.gov/documents/2020/12/18/2020-27880/perchloroethylene-pce-final-toxic-substances-control-act-tsca-risk-evaluation-notice-of-availability - EPA. Perchloroethylene (PCE); Revision to TSCA Risk Determination. 87 FR 76481, 14 Dec 2022. https://www.federalregister.gov/documents/2022/12/14/2022-27129/perchloroethylene-pce-revision-to-toxic-substances-control-act-tsca-risk-determination-notice-of - EPA. National Primary Drinking Water Regulations; Results of EPA’s Fourth Review of Existing Drinking Water Standards. 89 FR 59623, 23 July 2024. https://www.federalregister.gov/documents/2024/07/23/2024-15807/national-primary-drinking-water-regulations-announcement-of-the-results-of-epas-fourth-review-of-existing-drinking-water-standards - EPA. Perchloroethylene (PCE); Regulation Under TSCA (final rule). 89 FR 103560, 18 Dec 2024 (effective 17 Jan 2025). https://www.federalregister.gov/documents/2024/12/18/2024-30117/perchloroethylene-pce-regulation-under-the-toxic-substances-control-act-tsca - EPA. NESHAP: National Perchloroethylene Air Emission Standards for Dry Cleaning Facilities Technology Review (final). 90 FR 1041, 7 Jan 2025. https://www.federalregister.gov/documents/2025/01/07/2024-31223/national-emission-standards-for-hazardous-air-pollutants-national-perchloroethylene-air-emission - EPA. Perchloroethylene (PCE); Regulation Under TSCA; Request for Comment. 90 FR 35858, 30 July 2025. https://www.federalregister.gov/documents/2025/07/30/2025-14429/perchloroethylene-pce-regulation-under-the-toxic-substances-control-act-tsca-request-for-comment - EPA. Procedures for Chemical Risk Evaluation Under TSCA (proposed rule). 90 FR 45690, 23 Sept 2025 (noted for context; not relied on). https://www.federalregister.gov/documents/2025/09/23/2025-18431/procedures-for-chemical-risk-evaluation-under-the-toxic-substances-control-act-tsca - EPA. PCE and CTC; Regulation Under TSCA; Compliance Date Extensions (proposed). 91 FR 14790, 27 Mar 2026. https://www.federalregister.gov/documents/2026/03/27/2026-05977/perchloroethylene-pce-and-carbon-tetrachloride-ctc-regulation-under-the-toxic-substances-control-act - EPA. PCE and CTC; Regulation Under TSCA; Compliance Date Extensions (final). 91 FR 47145, 28 July 2026. https://www.federalregister.gov/documents/2026/07/28/2026-15192/perchloroethylene-pce-and-carbon-tetrachloride-ctc-regulation-under-the-toxic-substances-control-act - OSHA. 29 CFR 1910.1000, Table Z-2 (Tetrachloroethylene, Z37.22-1967). eCFR, version of 1 Aug 2026. https://www.ecfr.gov/current/title-29/subtitle-B/chapter-XVII/part-1910/subpart-Z/section-1910.1000 - EPA IRIS. Tetrachloroethylene (Perchloroethylene), CASRN 127-18-4, last updated 10 Feb 2012. https://iris.epa.gov/ChemicalLanding/&substance_nmbr=106

Court record - FabriClean Supply v. EPA, No. 25-60006 (5th Cir.), filed 6 Jan 2025 (docket via CourtListener). https://www.courtlistener.com/docket/69566264/fabriclean-supply-v-epa/

Hazard classifications - IARC. Trichloroethylene, Tetrachloroethylene, and Some Other Chlorinated Agents. IARC Monographs Vol. 106, 2014 (landing page; PDF not accessible). https://publications.iarc.who.int/Book-And-Report-Series/Iarc-Monographs-On-The-Identification-Of-Carcinogenic-Hazards-To-Humans/Trichloroethylene-Tetrachloroethylene-And-Some-Other-Chlorinated-Agents-2014 - Guha N et al. (2012). Carcinogenicity of trichloroethylene, tetrachloroethylene, some other chlorinated solvents, and their metabolites. Lancet Oncol 13:1192–1193. https://doi.org/10.1016/s1470-2045(12)70485-0 - NTP. Report on Carcinogens, 15th ed. (Dec 2021): Tetrachloroethylene profile. https://ntp.niehs.nih.gov/sites/default/files/ntp/roc/content/profiles/tetrachloroethylene.pdf ; Trichloroethylene profile. https://ntp.niehs.nih.gov/sites/default/files/ntp/roc/content/profiles/trichloroethylene.pdf - PubChem, Tetrachloroethylene (CID 31373), GHS Classification section (EU CLP Annex VI harmonised H351; ECHA C&L; NITE-CMC). https://pubchem.ncbi.nlm.nih.gov/compound/31373#section=GHS-Classification - NITE-CMC (Japan). Tetrachloroethylene GHS classification, FY2009 revision. https://www.chem-info.nite.go.jp/chem/english/ghs/09-mhlw-2094e.html

European Union - Commission Directive (EU) 2017/164 of 31 January 2017 (fourth list of indicative occupational exposure limit values). https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:32017L0164 - Directive (EU) 2020/2184 of 16 December 2020 on the quality of water intended for human consumption (recast), Annex I Part B. https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:32020L2184

California - CARB. Phase Out of Perchloroethylene from the Dry Cleaning Process (programme page and “About”). https://ww2.arb.ca.gov/our-work/programs/phase-out-perchloroethylene-dry-cleaning-process ; https://ww2.arb.ca.gov/our-work/programs/phase-out-perchloroethylene-dry-cleaning-process/about - CARB. Non-Toxic Dry Cleaning Grant Program (AB 998). https://ww2.arb.ca.gov/resources/documents/non-toxic-dry-cleaning-incentive-program-ab998 - CARB. Dry Cleaning Notice 2015-2: Alternative Solvents: Health and Environmental Impacts. https://ww3.arb.ca.gov/toxics/dryclean/notice2015_alt_solvents.pdf

The Massachusetts/Rhode Island pipe episode and its epidemiology - Larsen CD, Love OT, Reynolds G (1983). Tetrachloroethylene leached from lined asbestos–cement pipe into drinking water. J AWWA 75(4). https://doi.org/10.1002/j.1551-8833.1983.tb05104.x (abstract via Crossref) - Aschengrau A et al. (1993). Cancer risk and tetrachloroethylene-contaminated drinking water in Massachusetts. Arch Environ Health 48:284–292. https://doi.org/10.1080/00039896.1993.9936715 - Webler T, Brown HS (1993). Exposure to tetrachloroethylene via contaminated drinking water pipes in Massachusetts: a predictive model. Arch Environ Health 48:293–297. https://doi.org/10.1080/00039896.1993.9936716 - Spence LA et al. (2008). Evaluation of the Webler-Brown model for estimating tetrachloroethylene exposure from vinyl-lined asbestos-cement pipes. Environ Health 7:24. https://doi.org/10.1186/1476-069X-7-24 (full text: https://pmc.ncbi.nlm.nih.gov/articles/PMC2432054/) - Aschengrau A et al. (2015). Long-term health effects of early life exposure to PCE-contaminated drinking water. Environ Health 14:36. https://doi.org/10.1186/s12940-015-0021-z - Aschengrau A et al. (2016). Long-term neurotoxic effects of early-life exposure to tetrachloroethylene-contaminated drinking water. Ann Glob Health 82:169–179. https://doi.org/10.1016/j.aogh.2016.01.013 (full text: https://pmc.ncbi.nlm.nih.gov/articles/PMC4916338/) - Aschengrau A et al. (2018). Modeled exposure to tetrachloroethylene-contaminated drinking water and the occurrence of birth defects. Environ Health 17:75. https://doi.org/10.1186/s12940-018-0419-5 - Aschengrau A et al. (2018). Modeled exposure to tetrachloroethylene-contaminated drinking water and the risk of placenta-related stillbirths. Environ Health 17:58. https://doi.org/10.1186/s12940-018-0402-1 - Aschengrau A et al. (2020). Reproductive and developmental health effects of prenatal exposure to tetrachloroethylene-contaminated drinking water. Environ Sci Process Impacts 22:555–566. https://doi.org/10.1039/c9em00590k (full text: https://pmc.ncbi.nlm.nih.gov/articles/PMC7937243/) - Aschengrau A et al. (2020). Drug use disorder following early life exposure to PCE-contaminated drinking water. Environ Health 19:99. https://doi.org/10.1186/s12940-020-00638-2 - Menon PV et al. (2026). Association between prenatal exposure to tetrachloroethylene and adverse birth outcomes: systematic review and meta-analysis. Environ Res 307:125496 (epub 17 Aug 2026). https://doi.org/10.1016/j.envres.2026.125496

Cancer evidence and assessment disputes - Vlaanderen J et al. (2013). Occupational exposure to TCE and PCE and the risk of lymphoma, liver, and kidney cancer in four Nordic countries. Occup Environ Med 70:393–401. https://doi.org/10.1136/oemed-2012-101188 - Guyton KZ et al. (2014). Human health effects of tetrachloroethylene: key findings and scientific issues. Environ Health Perspect 122:325–334. https://doi.org/10.1289/ehp.1307359 (full text: https://pmc.ncbi.nlm.nih.gov/articles/PMC3984230/) - Vlaanderen J et al. (2014). Tetrachloroethylene exposure and bladder cancer risk: a meta-analysis of dry-cleaning-worker studies. Environ Health Perspect 122:661–666. https://doi.org/10.1289/ehp.1307055 - Callahan CL et al. (2018). Case-control investigation of occupational exposure to chlorinated solvents and non-Hodgkin’s lymphoma. Occup Environ Med 75:415–420. https://doi.org/10.1136/oemed-2017-104890 - DeSesso JM et al. (2019). Trichloroethylene in drinking water throughout gestation did not produce congenital heart defects in Sprague Dawley rats. Birth Defects Res 111:1217–1233. https://doi.org/10.1002/bdr2.1531 - Goodman JE, Ticknor RC, Zhou J (2022). Systematic review of perchloroethylene and non-Hodgkin’s lymphoma. Glob Epidemiol 4:100077. https://doi.org/10.1016/j.gloepi.2022.100077 ; Response. Glob Epidemiol 4:100091. https://doi.org/10.1016/j.gloepi.2022.100091 - Savitz DA (2022). Commentary on methodologic choices in synthesizing epidemiologic evidence to assess perchloroethylene and NHL. Glob Epidemiol 4:100089. https://doi.org/10.1016/j.gloepi.2022.100089 - Fox MP, Mathur MB, Matthay EC (2022). Commentary: Quantifying the impact of bias … the case of perchloroethylene and NHL. Glob Epidemiol 4:100090. https://doi.org/10.1016/j.gloepi.2022.100090 - Weed DL (2022). Commentary: On the reliability of causal claims. Glob Epidemiol 4:100087. https://doi.org/10.1016/j.gloepi.2022.100087 - Klaunig JE, Bevan C, Gollapudi B (2024). Assessment of the mode of action of perchloroethylene-induced mouse liver tumors. Toxicol Ind Health 40:272–291. https://doi.org/10.1177/07482337241240188

Evidence-synthesis frameworks (Claim 8) - National Academies (2014). Review of EPA’s Integrated Risk Information System (IRIS) Process. https://doi.org/10.17226/18764 - EFSA Scientific Committee (2017). Guidance on the use of the weight of evidence approach in scientific assessments. EFSA Journal 15(8):4971. https://doi.org/10.2903/j.efsa.2017.4971 - Beronius A et al. (2018). Testing and refining the Science in Risk Assessment and Policy (SciRAP) web-based platform. J Appl Toxicol 38:1460–1470. https://doi.org/10.1002/jat.3648 - Samet JM et al. (2020). The IARC Monographs: updated procedures for modern and transparent evidence synthesis in cancer hazard identification. J Natl Cancer Inst 112:30–37. https://doi.org/10.1093/jnci/djz169 - National Academies (2021). The Use of Systematic Review in EPA’s Toxic Substances Control Act Risk Evaluations. https://doi.org/10.17226/25952 (landing page only consulted)

Dry cleaning and alternatives - Ceballos DM et al. (2021). Perchloroethylene and dry cleaning: it’s time to move the industry to safer alternatives. Front Public Health 9:638082. https://doi.org/10.3389/fpubh.2021.638082 (full text: https://pmc.ncbi.nlm.nih.gov/articles/PMC7973082/). Used as a secondary source for the EU, France, Minneapolis and national solvent shares.

Camp Lejeune (context for Claim 4) - Maslia ML et al. (2016). Reconstructing historical VOC concentrations in drinking water … at a U.S. military base. Water 8:449. https://doi.org/10.3390/w8100449 - Bove FJ et al. (2024). Cancer incidence among Marines and Navy personnel and civilian workers exposed to industrial solvents in drinking water at Camp Lejeune. Environ Health Perspect 132:107008. https://doi.org/10.1289/EHP14966 (consulted; mixed TCE/PCE exposure, not used for PCE-specific inference) - Honoring our PACT Act of 2022, Pub. L. 117-168, s. 804 (Camp Lejeune Justice Act of 2022). Cited from knowledge; the text was not fetched in this check.

Access notes. The following were not accessible, and claims that depend on them are flagged in the text: - the IARC Vol. 106 full text (PDF denied; NCBI Bookshelf CAPTCHA, not bypassed); - mass.gov pages on current PCE monitoring in affected Massachusetts towns (403); - ATSDR’s 2017 Camp Lejeune causality assessment (403); - ECHA substance pages (web-application firewall); - MDEE (1982) and Demond (1982), the chapter’s primary sources for the pipe history (not online).