Download PDF
December 6, 2022

European General Court Annuls Harmonized Classification and Labeling of Titanium Dioxide

Bergeson & Campbell, P.C.

On November 23, 2022, the Court of Justice of the European Union (EU) announced a decision of the General Court that annuls the 2019 harmonized classification and labeling of titanium dioxide as a carcinogenic substance by inhalation in certain powder forms. According to the press release, the European Commission (EC) “made a manifest error in its assessment of the reliability and acceptability of the study on which the classification was based and, second, it infringed the criterion according to which that classification can relate only to a substance that has the intrinsic property to cause cancer.”


In 2016, the competent French authority submitted a proposal to the European Chemicals Agency (ECHA) to classify titanium dioxide as a category 1B carcinogenic substance (carcinogenic to humans). In 2017, ECHA’s Committee for Risk Assessment (RAC) adopted an opinion classifying titanium dioxide as a category 2 carcinogen (suspected human carcinogen), including the hazard statement “H 351 (inhalation).” On the basis of RAC’s Opinion, the EC adopted Regulation 2020/217, implementing the harmonized classification and labeling of titanium dioxide, recognizing that that substance was suspected of being carcinogenic to humans, by inhalation, in powder form containing one percent or more of particles of a diameter equal to or less than ten micrometers (µm). The transition period for adoption of these changes ended October 1, 2021. The applicants, in their capacity as manufacturers, importers, downstream users, or suppliers of titanium dioxide, brought actions before the General Court for the partial annulment of Regulation 2020/217.

Findings of the Court

The General Court held that the requirement to base the classification of a carcinogenic substance on reliable and acceptable studies was not satisfied. According to the press release, in recognizing that the results of a scientific study on which it based its opinion on the classification and labeling of titanium dioxide were sufficiently reliable, relevant, and adequate for assessing the carcinogenic potential of that substance, RAC committed “a manifest error of assessment.” Specifically, to verify the degree of lung overload of titanium dioxide particles in that scientific study to assess carcinogenicity, RAC used a density value corresponding to the density of unagglomerated primary particles of titanium dioxide, “which is always higher than the density of the agglomerates of nano-sized particles of that substance.” The press release states that in so doing, RAC did not take into account all the relevant factors to calculate the lung overload during the scientific study at issue, namely the characteristics of the particles tested in that scientific study, the fact that those particles tend to agglomerate, and the fact that the density of the agglomerates of particles was lower than the particle density and that, for that reason, those agglomerates occupied more volume in the lungs. Thus, according to the General Court, RAC’s findings that the lung overload in the scientific study at issue was acceptable were implausible. Consequently, when the EC based the contested regulation and the harmonized classification and labeling of titanium dioxide on the RAC Opinion, thus following RAC’s conclusion as to the reliability and acceptability of the results of the scientific study at issue, “it made the same manifest error of assessment as the RAC.”

The General Court found that the contested classification and labeling “infringed the criterion according to which the classification of a substance as carcinogenic can apply only to a substance that has the intrinsic property to cause cancer.” Under Regulation (EC) No 1272/2008, harmonized classification and labeling of a substance as carcinogenic may be based only on the intrinsic properties of the substance that determine its intrinsic capacity to cause cancer. The General Court interpreted the concept of “intrinsic properties,” stating that, although that concept does not appear in Regulation (EC) No 1272/2008, “it must be interpreted in its literal sense as referring to the ‘properties which a substance has in and of itself’, which is consistent, inter alia, with the objectives and purpose of harmonised classification and labelling under that regulation.” The General Court notes that the contested classification and labeling are intended to identify and notify a carcinogenic hazard of titanium dioxide, which, in the RAC Opinion, was classified as “non-intrinsic in a classical sense.” According to the press release, the General Court states that that “non-intrinsic in the classical sense” nature stems from several factors, referred to both in that opinion and in the contested regulation. The carcinogenicity hazard is linked solely to certain respirable titanium dioxide particles when they are present in a certain form, physical state, size, and quantity; it occurs only in lung overload conditions; and corresponds to particle toxicity.

The General Court concluded that, by upholding the conclusion contained in the RAC Opinion that the mode of action of carcinogenicity on which RAC relied could not be regarded as intrinsic toxicity in the classical sense, but which had to be taken into consideration in the context of harmonized classification and labeling under Regulation (EC) No 1272/2008, the EC committed a manifest error of assessment. According to the press release, the General Court states that the examples of classification and labeling of other substances, relied on to compare them with the classification and labeling of titanium dioxide, “illustrate only cases in which, even though the form and size of the particles were taken into account, certain properties specific to the substances were nevertheless decisive for their classification, which does not correspond to the case here.”


It is unclear if the EC and/or member states will appeal. If not, the ruling will take effect at the expiration of a two month and ten-day period from the notification of the judgment (i.e., approximately February 10, 2023). It is also unclear what impact the decision will have in the United Kingdom (UK) unless the Health and Safety Executive (HSE) chooses to revise the current classification that was adopted into its legislation at the time of Brexit, or if a separate court case will be brought in the UK. HSE may opt to propose revisions to align with the General Court decision.

As for the decision’s impact more broadly, a few points merit note. First, the decision marks a victory for science and the titanium dioxide community’s resolve to challenge flawed science. That the European Court of Justice overturned the classification and explained its reasoning as clearly and emphatically as it did is encouraging. The decision was based on scientific questions presented under the regulatory language of Regulation (EC) No. 1272/2008 for classification, labeling, and packaging (CLP) of substances. This is the first known case where a harmonized classification and labeling mandate has been overturned, no small feat given the precautionary framework established within the EU under the CLP regulation. Stakeholders that amended labeling and safety data sheet content to align with the original mandatory deadline of October 1, 2021, will need to consider, if the decision is not appealed, what additional revisions are needed once these changes are enacted. It is unclear how this decision may impact decision making on comparable substances under other regulatory frameworks, such as the U.S. Toxic Substances Control Act (TSCA) and the U.S. Occupational Safety and Health Administration’s (OSHA) Hazard Communication (Haz Com).

Second, it is unclear what effect the General Court’s ruling will have on the broader question of lung overload in the regulatory risk assessment context. The term “lung overload” or “lung particle overload” has been described as the “impairment of particle clearance from the deep lung after inhalation of high concentrations of poorly solubility, low toxicity [PSLT] materials exemplified by titanium dioxide and carbon black.” The rat is particularly sensitive to effects, including cancer, resulting from lung overload. There is not, however, consensus in the scientific community about the human relevance of effects observed in rats due to lung overload.

Two relevant expert workshops have, however, weighed in on this point. For example, in 1998, the International Life Sciences Institute (ILSI) convened an expert workshop titled “The Relevance of the Rat Lung Response to Particle Overload for Human Risk Assessment: A Workshop Consensus Report.” One of the questions posed to the experts was “What would be an appropriate uncertainty factor (UF) to account for interspecies extrapolation?” In response, the experts stated at pages 12-13: “For both neoplastic and fibrogenic endpoints in the rat, associated with PSP [poorly soluble particles] exposures, the work group proposed that the TD [toxicodynamic] component of the interspecies UF be reduced from a factor of 3 to 1, given that chronic active inflammation in the rat appears to be a more sensitive response than in other species, including humans.” More recently, an expert workshop on the hazards and risks of PSLT particles noted on page 59 that “There was consensus that the rat is more sensitive than other species and humans in the lung response to PSLT for inflammation, epithelial hyperplasia, and fibrosis [emphasis in original].” Despite these conclusions, regulatory agencies have not adopted a reduced TD component of the interspecies UF from 3 to 1.

Third, beyond the question of human relevance is the question about the appropriate dose metric to use when assessing the potential risks to PSLTs from effects observed in rat inhalation studies where lung overload was achieved. The General Court’s decision does not provide clarity on this point, although the U.S. Environmental Protection Agency (EPA) has weighed in on this. For example, EPA’s 1994 document titled “Methods for Derivation of Inhalation Reference Concentrations and Application of Inhalation Dosimetry” states “a more appropriate dose metric for particle exposures may be to take into account clearance of the deposited dose and thereby calculate the retained dose and the dose rate to extrarespiratory tissues.” Despite this, EPA has not consistently applied this guidance. For example, EPA used deposited dose as the dose metric in its recent evaluation of C.I. Pigment Violet 29, a PSLT particle, yet it used retained dose as the dose metric when developing a chemical category for PSLT polymers.

Fourth, an important item to note is that the General Court concluded that “the Commission made a manifest error in its assessment of the reliability and acceptability of the study [Heinrich et al., 1995] on which classification was based…” This may have relevance under TSCA because EPA included the Heinrich et al. (1995) study as part of its TSCA chemical category on “Respirable, Poorly Soluble Particulates.” Under amended TSCA Section 26(h)-(i), EPA is required to use the best available science and weight of scientific evidence when making decisions under TSCA Sections 4, 5, and 6. Therefore, the General Court’s decision may have an impact on EPA’s use of this study as a point of departure for use in its risk assessments under TSCA Sections 5 and 6.

In conclusion, the General Court’s decision is a win for science under Regulation (EC) No. 1272/2008. We note that compelling scientific arguments alone will not ensure success under regulatory challenges, unless the scientific arguments are prepared within the context of regulations that allow for such support. The General Court’s decision does, however, provide compelling evaluations under Regulation (EC) No. 1272/2008 that would aid with informing potential challenges under other regulatory frameworks such as TSCA. As noted above, the requirement for “reliability and acceptability” of data under Regulation (EC) No. 1272/2008 is comparable to the requirements of best available science under TSCA Section 26(h). Therefore, we encourage interested persons to read the General Court’s decision and to consider more broadly how such a decision could aid with framing potential scientific challenges under other regulatory frameworks that include scientific standards.

It also may suggest that International Agency for Research on Cancer (IARC) Monograph 93 that concludes “sufficient evidence in experimental animals for the carcinogenicity of titanium dioxide,” based on similar scientific concepts, could be revisited. This would have a significant impact on the classification of substances and mixtures in other jurisdictions that rely on IARC as a source for these types of determinations (e.g., OSHA and Health Canada).