JUDGMENT OF THE GENERAL COURT (Sixth Chamber, Extended Composition)

19 November 2025 ( * )

( REACH – Substance di-tert-butyl 1,1,4,4-tetramethyltetramethylene diperoxide – Compliance check of registrations – Request for additional toxicity studies – Article 41 of Regulation (EC) No 1907/2006 – Manifest error of assessment – Error of fact – Proportionality – Decision of the Board of Appeal of ECHA – Admissibility of the arguments directed against the decision of ECHA )

In Case T‑1122/23,

Nouryon Functional Chemicals BV, established in Amsterdam (Netherlands),

Arkema GmbH, established in Düsseldorf (Germany),

Pergan Hilfsstoffe für industrielle Prozesse GmbH, established in Bocholt (Germany),

United Initiators GmbH, established in Pullach im Isartal (Germany),

represented by R. Cana and Z. Romata, lawyers,

applicants,

v

European Chemicals Agency (ECHA), represented by T. Basmatzi, F. Becker and L. Bolzonello, acting as Agents,

defendant,

THE GENERAL COURT (Sixth Chamber, Extended Composition),

composed, at the time of the deliberations, of M.J. Costeira, President, M. Kancheva, U. Öberg, P. Zilgalvis (Rapporteur) and E. Tichy-Fisslberger, Judges,

Registrar: M. Zwozdziak-Carbonne, Administrator,

having regard to the written part of the procedure,

further to the hearing on 13 March 2025,

gives the following

Judgment

1 By their action under Article 263 TFEU, the applicants, Nouryon Functional Chemicals BV, Arkema GmbH, Pergan Hilfsstoffe für industrielle Prozesse GmbH and United Initiators GmbH, seek the annulment of Decision A-009-2022 of the Board of Appeal of the European Chemicals Agency (ECHA) of 19 September 2023 (‘the contested decision’), by which the Board of Appeal dismissed in part their appeal brought against the decision of ECHA of 8 June 2022 on the compliance check of the registration dossier for the substance di-tert-butyl 1,1,4,4-tetramethyltetramethylene diperoxide (‘the substance concerned’) adopted under Article 41 of Regulation (EC) No 1907/2006 of the European Parliament and of the Council of 18 December 2006 concerning the Registration, Evaluation, Authorisation and Restriction of Chemicals (REACH), establishing a European Chemicals Agency, amending Directive 1999/45/EC and repealing Council Regulation (EEC) No 793/93 and Commission Regulation (EC) No 1488/94 as well as Council Directive 76/769/EEC and Commission Directives 91/155/EEC, 93/67/EEC, 93/105/EC and 2000/21/EC (OJ 2006 L 396, p. 1, and corrigendum OJ 2007 L 136, p. 3; ‘the REACH Regulation’) (‘the initial decision’).

Background to the dispute

2 The applicants are registrants of the substance concerned under the REACH Regulation.

3 They registered the substance concerned between 2011 and 2013 in the tonnage band of 100 to 1 000 tonnes per annum, which corresponds to the volume of manufacture or import referred to in Annex IX to the REACH Regulation.

4 Following a compliance check which ECHA carried out in accordance with Article 41 of the REACH Regulation, that agency adopted the initial decision, which required the applicants to submit, by 15 September 2025, information on an extended one-generation reproductive toxicity study (‘EOGRTS’) under Section 8.7.3. of Annex IX to that regulation, including, inter alia, cohort 1A (reproductive toxicity), cohort 1B (reproductive toxicity) without extension to mate the cohort 1B animals to produce the F2 generation, cohorts 2A and 2B (developmental neurotoxicity), along with studies into learning and memory functions as described in paragraph 37 of test guideline (TG) 426 of the Organisation for Economic Co-operation and Development (OECD), corresponding to European Union test method B.53.

5 On 8 September 2022, the applicants lodged an appeal against the initial decision, pursuant to Article 91(1) of the REACH Regulation.

6 By the contested decision, the Board of Appeal dismissed the appeal in part as regards the requirement for the applicants to submit information on an EOGRTS by oral route in rats, including cohorts 1A and 1B (without extension to include the F2 generation) under column 1 of Section 8.7.3. of Annex IX to the REACH Regulation (‘EOGRTS with a basic design’), as well as additional cohorts 2A and 2B under column 2 of Section 8.7.3. of Annex IX to the REACH Regulation. The Board of Appeal set 26 December 2026 as the deadline for the applicants to submit that information.

Forms of order sought

7 The applicants claim that the Court should:

– annul the contested decision in so far as it requires them to submit information on an EOGRTS with a basic design including additional cohorts 2A and 2B under column 2 of Section 8.7.3. of Annex IX to the REACH Regulation;

– order ECHA to pay the costs.

8 ECHA contends that the Court should:

– dismiss the action;

– order the applicants to pay the costs.

Law

9 The applicants raise five pleas in law.

10 The first plea alleges infringement of column 1 of Section 8.7.3. of Annex IX to the REACH Regulation, in that the contested decision required the applicants to submit information on an EOGRTS with a basic design.

11 The second plea alleges infringement of the principle of proportionality and of Article 25 of the REACH Regulation in that the contested decision required the applicants to submit information on an EOGRTS with a basic design.

12 The third plea alleges that ECHA made manifest errors of assessment and infringed the principles of legal certainty and of the protection of legitimate expectations by requiring the applicants to submit information on an EOGRTS with a basic design.

13 The fourth plea alleges that ECHA infringed the second paragraph of column 2 of Section 8.7.3. of Annex IX to the REACH Regulation and the principle of proportionality by failing to examine whether the obligation that the EOGRTS include additional cohorts 2A and 2B was proportionate.

14 The fifth plea alleges that ECHA made manifest errors of assessment and infringed the principles of legal certainty and of the protection of legitimate expectations by requiring that the EOGRTS include additional cohorts 2A and 2B.

Preliminary observations on the intensity of the Court’s review

15 As a preliminary point, it must be noted that ECHA has a broad discretion in the application of Article 41 of the REACH Regulation in so far as it is called upon to carry out complex scientific and technical assessments (see, to that effect and by analogy, judgment of 22 November 2017, Commission v Bilbaína de Alquitranes and Others , C‑691/15 P, EU:C:2017:882, paragraph 34 and the case-law cited).

16 However, the exercise of that discretion is not excluded from review by the Court. The Court has consistently held that, in the context of such a review, the EU judicature must verify whether the relevant procedural rules have been complied with, whether the facts admitted by the administrative authority have been accurately stated and whether there has been a manifest error of appraisal or a misuse of powers (see, to that effect and by analogy, judgment of 18 July 2007, Industrias Químicas del Vallés v Commission , C‑326/05 P, EU:C:2007:443, paragraph 76 and the case-law cited).

17 In that regard, in order to establish that the administrative authority made a manifest error in assessing the facts such as to justify the annulment of the contested measure, the evidence adduced by the applicant must be sufficient to make the factual assessments used in that measure implausible. Without prejudice to that examination of plausibility, it is not for the Court to substitute its assessment of highly complex facts for that of the author of the contested measure (see, to that effect, judgments of 12 December 1996, AIUFFASS and AKT v Commission , T‑380/94, EU:T:1996:195, paragraph 59, and of 19 September 2019, Arysta LifeScience Netherlands v Commission , T‑476/17, EU:T:2019:618, paragraph 87 and the case-law cited).

Details of the applicants’ arguments directed against the initial decision

18 In the application, some of the applicants’ arguments are directed against the initial decision.

19 ECHA contends that those arguments must be regarded as inadmissible or ineffective on the ground that an action for annulment of a decision of the Board of Appeal relates solely to the legality of that decision.

20 In that regard, it should be recalled that, under Article 94(1) of the REACH Regulation, an action may be brought before the General Court or the Court of Justice, in accordance with Article 263 TFEU, contesting a decision taken by the Board of Appeal or, in cases where no right of appeal lies before the Board of Appeal, by ECHA.

21 It follows that, in the case of decisions against which an appeal may be brought before the Board of Appeal, the EU judicature intervenes, where appropriate, only to verify the final outcome of an internal appeal, that is, to examine the decision taken at the end of the appeals procedure before the Board of Appeal, and therefore the decision of the Board of Appeal (see, to that effect and by analogy, judgments of 28 January 2016, Heli-Flight v EASA , C‑61/15 P, not published, EU:C:2016:59, paragraph 81, and of 7 September 2022, BNetzA v ACER , T‑631/19, EU:T:2022:509, paragraph 26).

22 Furthermore, in the context of such an appeal, the Board of Appeal is to confine itself to examining whether the arguments put forward by the applicant are such as to demonstrate the existence of an error vitiating the contested decision, and it is not for it to examine whether or not, at the time when it adjudicates on the appeal, in the light of all the relevant elements of law and facts, in particular scientific issues, a new decision with the same operative part as the decision contested before it may be lawfully adopted (see, to that effect, judgment of 20 September 2019, BASF Grenzbach v ECHA , T‑125/17, EU:T:2019:638, paragraph 59 and 60).

23 Where the examination of the pleas put forward by the applicant in the course of the proceedings shows that an ECHA decision is vitiated by error, under Article 93(3) of the REACH Regulation, it is for the Board of Appeal to decide whether to refer the case to the competent body of that agency or whether it exercises any power falling within the competence of that agency (judgment of 20 September 2019, BASF Grenzach v ECHA , T‑125/17, EU:T:2019:638, paragraph 117).

24 It follows that the relevant provisions of the REACH Regulation must be interpreted as meaning that the decisions taken by the Boards of Appeal replace the decisions initially taken by ECHA and that, consequently, the subject matter of the action for annulment must be regarded as being the decision of the Board of Appeal which dismissed the internal appeal brought against the initial decision (see, to that effect and by analogy, judgments of 28 January 2016, Heli-Flight v EASA , C‑61/15 P, not published, EU:C:2016:59, paragraph 84; of 20 September 2019, Germany v ECHA , T‑755/17, EU:T:2019:647, paragraph 59; and of 7 September 2022, BNetzA v ACER , T‑631/19, EU:T:2022:509, paragraph 27).

25 Consequently, the pleas in law and arguments in the present action which are based on irregularities relating to the initial decision and which cannot be interpreted as being directed also against the contested decision must be rejected as ineffective, since the Court can rule only on the legality of the latter decision (see, to that effect and by analogy, judgments of 11 December 2014, Heli-Flight v EASA , T‑102/13, EU:T:2014:1064, paragraph 32, and of 7 September 2022, BNetzA v ACER , T‑631/19, EU:T:2022:509, paragraph 27).

26 However, since that decision is based on the grounds stated in the initial decision, and indeed confirms those grounds, whether implicitly or explicitly, all the pleas in law and arguments in the present action that are directed against those grounds must be found to be fully effective for the purpose of reviewing the legality of the contested decision (see, to that effect and by analogy, judgment of 7 September 2022, BNetzA v ACER (T‑631/19, EU:T:2022:509, paragraph 28).

27 It is in the light of those considerations that the applicants’ arguments directed against the initial decision must be examined.

The first plea in law, alleging infringement of column 1 of Section 8.7.3 . of Annex IX to the REACH Regulation, in that the contested decision required the applicants to submit information on an EOGRTS with a basic design

28 Under Article 12(1)(d) of the REACH Regulation, registrants who manufacture or import a substance in quantities of 100 tonnes or more per year are required to provide the information specified in Annexes VII and VIII to that regulation and testing proposals for the provision of the information specified in Annex IX to that regulation.

29 The standard information requirements for an EOGRTS were laid down in the REACH Regulation by Commission Regulation (EU) 2015/282 of 20 February 2015 amending Annexes VIII, IX and X to the REACH Regulation as regards the Extended One-Generation Reproductive Toxicity Study (OJ 2015 L 50, p. 1). The EOGRTS is a ‘modular’ study, that is, it consists of a basic design to which additional elements, such as cohorts, can be added.

30 In that regard, it is apparent from column 1 of Section 8.7.3. of Annex IX to the REACH Regulation that an EOGRTS with a basic design must be carried out if the available repeated dose toxicity studies (for example, 28-day or 90-day studies, OECD 421 or 422 screening studies) indicate adverse effects on reproductive organs or tissues or reveal other concerns in relation with reproductive toxicity.

31 In the initial decision, ECHA considered, in essence, that the significantly reduced cauda epididymal spermatid count in animals in the group receiving the highest dose (150 milligrams per kilogram of body weight per day (mg/kg bw/day)) in the OECD TG 408 study was a concern in relation to reproductive toxicity within the meaning of column 1 of Section 8.7.3. of Annex IX to the REACH Regulation.

32 Before the Board of Appeal, the applicants argued, in essence, that ECHA had made a number of errors of assessment of the effects observed in the OECD TG 408 study, which should have been regarded as incidental and of no toxicological relevance. They also stated that ECHA had not complied with its own guidance by assessing only the existence of effects and not their potentially adverse nature.

33 The Board of Appeal found, adopting the same ground as that of the initial decision, that the effects observed in the OECD TG 408 study satisfied the conditions set out in column 1 of Section 8.7.3. of Annex IX to the REACH Regulation. In that regard, the Board of Appeal found that a change in the cauda epididymal spermatid count did not constitute an adverse effect on reproductive organs or tissues because sperm is not a reproductive organ or tissue as such. The Board of Appeal nevertheless found that the results of the OECD TG 408 study were sufficient to raise other concerns in relation to reproductive toxicity.

34 The applicants submit, in the first place, that the conclusion that the reduction in the cauda epididymal spermatid count does not constitute an adverse effect on reproductive organs or tissues, since sperm is neither a reproductive organ nor a reproductive tissue, amounts to a manifest error of assessment.

35 According to the applicants, sperm is present and measured in the testes and epididymis, which are, by definition and according to even the most basic scientific principles, reproductive organs (the epididymis being the tube that carries sperm from the testes), and so any effect on sperm is an effect on the reproductive organs or tissues in which the sperm is present (testes or epididymis).

36 ECHA disputes that line of argument.

37 In that regard, it should be noted that, as follows from paragraph 33 above, the obligation to carry out an EOGRTS with a basic design was motivated, inter alia, by the fact that the results of the OECD TG 408 study were sufficient to raise other concerns regarding reproductive toxicity, within the meaning of the second condition in column 1 of Section 8.7.3. of Annex IX to the REACH Regulation.

38 The Board of Appeal thus did not base the contested decision, in so far as it required the applicants to submit information on an EOGRTS with a basic design, solely on the fact that the available studies indicated adverse effects on reproductive organs or tissues, within the meaning of the first condition in column 1 of Section 8.7.3. of Annex IX to the REACH Regulation.

39 It must therefore be held that the applicants’ argument concerning the existence of a possible adverse effect on reproductive organs or tissues is not directed against a ground of the contested decision that in itself supports the operative part of that decision.

40 In those circumstances, that argument must be rejected as ineffective in so far as it is not capable of calling into question the ground of the contested decision that is based on the OECD TG 408 study and that in itself supports the operative part of that decision.

41 In the second place, the applicants submit that ECHA’s conclusion that the considerable reduction in the epididymal spermatid count raises another concern in relation to reproductive toxicity is vitiated by an error of law and by a manifest error of assessment.

42 ECHA disputes that line of argument.

43 First, the applicants assert, in essence, that the Board of Appeal erred in law in taking the view that ECHA was not under an obligation to demonstrate the existence of adverse effects when requiring information on an EOGRTS on the basis of other concerns regarding reproductive toxicity.

44 In that respect, it must be noted that, as follows from paragraph 30 above, an EOGRTS with a basic design must be carried out if the available repeated dose toxicity studies indicate adverse effects on reproductive organs or tissues or reveal other concerns regarding reproductive toxicity.

45 At the outset, it should be noted that it is apparent from the use of the coordinating conjunction ‘or’ that the condition relating to the indication of adverse effects on reproductive organs or tissues and the condition relating to the emergence of other concerns regarding reproductive toxicity are alternatives.

46 It follows that the wording of column 1 of Section 8.7.3. of Annex IX to the REACH Regulation precludes the interpretation that the condition relating to the emergence of other concerns regarding reproductive toxicity requires the indication of adverse effects to be demonstrated, since that demonstration is required only for the first condition laid down by that provision.

47 According to settled case-law, where the wording of a provision is unambiguous, an interpretation in the light of its legal context cannot have the result of depriving the clear and precise wording of that provision of all effectiveness (see, to that effect, judgments of 8 December 2005, ECB v Germany , C‑220/03, EU:C:2005:748, paragraph 31, and of 28 February 2008, Carboni e derivati , C‑263/06, EU:C:2008:128, paragraph 48).

48 In the present case, since it is clear from the wording of column 1 of Section 8.7.3. of Annex IX to the REACH Regulation that the condition relating to the emergence of other concerns regarding reproductive toxicity does not depend on the demonstration of the indication of adverse effects, the applicants’ arguments based on the interpretation of that provision in the light of its origin or context must be rejected, since such an interpretation would result in the clear and precise wording of that provision being deprived of all effectiveness.

49 The same applies to the argument that a subjective concern cannot trigger an EOGRTS in the absence of decisive objective mechanisms, since that argument is not capable of calling into question the clear wording of column 1 of Section 8.7.3. of Annex IX to the REACH Regulation.

50 It is also necessary to reject the applicants’ argument based on the interpretation of the second paragraph of column 2 of Section 8.7.3. of Annex X to the REACH Regulation that was carried out in the judgment of 29 March 2023, Nouryon Industrial Chemicals and Others v Commission (T‑868/19, EU:T:2023:168, paragraphs 103 to 105), from which, they state, it is apparent that the emergence of a concern presupposes the demonstration of sufficiently serious or severe adverse effects.

51 First, the drafting of the provision interpreted in the judgment of 29 March 2023, Nouryon Industrial Chemicals and Others v Commission (T‑868/19, EU:T:2023:168), differs from column 1 of Section 8.7.3. of Annex IX to the REACH Regulation and, second, the passages cited by the applicants, relating to the concept of ‘particular concern’, have no bearing on whether the concept of ‘concern in relation with reproductive toxicity’ requires the demonstration of adverse effects. Moreover, paragraphs 103 and 104 of that judgment do not support the applicants’ line of argument. In particular, paragraph 103 of that judgment states that it follows from the very wording of the second paragraph of column 2 of Section 8.7.3. of Annex X to the REACH Regulation, and in particular from the word ‘concern’, which, in the context in question, means ‘worry’, that, for such a concern to exist, information of a certain nature held by the registrants or by the competent authority must establish that the substance in question has developmental neurotoxicity effects, irrespective of effects that result from a more general toxicity, or even merely gives reasonable grounds for fearing that that substance might have those effects.

52 Second, the applicants argue that ECHA has not proved the existence of serious and severe effects in the present case.

53 However, since it follows from a literal interpretation of column 1 of Section 8.7.3. of Annex IX to the REACH Regulation that ECHA is not required to prove the existence of adverse effects, that argument must be rejected as ineffective.

54 Lastly, in paragraphs 59 to 61 of the application, the applicants set out the content of the initial decision, argue that ‘none of the findings upon which [ECHA] relies as a trigger for the EOGRTS meet the requirements in … Annex [IX to the REACH Regulation], because, as will be explained below …’, and assert that they ‘have demonstrated’ that the findings of the studies analysed in the initial decision, and then in the contested decision, were of no toxicological relevance.

55 ECHA considers that those arguments, directed against the initial decision, should be rejected as inadmissible or, in any event, ineffective.

56 However, it must be noted that those arguments consist of a restating of the arguments put forward during the administrative procedure and that they overlap with the arguments put forward against the contested decision, with the result that they will be addressed in the context of the examination of the third plea.

57 In the light of the foregoing and subject to the arguments referred to in paragraph 54 above, which will be taken into account below, the first plea must be rejected.

The second plea in law, alleging infringement of the principle of proportionality and of Article 25 of the REACH Regulation in that the contested decision required the applicants to submit information on an EOGRTS with a basic design

58 The applicants submit, in essence, that the contested decision was adopted in breach of the principle of proportionality and of Article 25 of the REACH Regulation in so far as an EOGRTS with a basic design was not necessary and was not the most appropriate study to address the concerns alleged under Section 8.7.3. of Annex IX to the REACH Regulation.

59 ECHA disputes that line of argument.

60 The applicants submit, in the first place, that ECHA infringed the principle of proportionality by taking the view that that principle did not apply and by failing to exercise its discretion, in accordance with that principle, when it considered that there were other concerns regarding reproductive toxicity requiring the applicants to submit information on an EOGRTS under column 1 of Section 8.7.3. of Annex IX to the REACH Regulation.

61 In that regard, it must be noted that the principle of proportionality, which is one of the general principles of EU law, requires that measures adopted by EU institutions do not exceed the limits of what is appropriate and necessary in order to attain the objectives legitimately pursued by the legislation in question; when there is a choice between several appropriate measures, recourse must be had to the least onerous, and the disadvantages caused must not be disproportionate to the aims pursued (see judgment of 7 March 2013, Rütgers Germany and Others v ECHA , T‑96/10, EU:T:2013:109, paragraph 133 and the case-law cited).

62 In the present case, under Article 10(a)(vi) of the REACH Regulation, read in conjunction with column 1 of Section 8.7.3. of Annex IX to that regulation, a registrant’s registration dossier, for the manufacture or import of substances referred to in that annex, must, in principle, contain information in the form of summaries of studies on an EOGRTS with a basic design under ‘standard information requirements’ if the available repeated dose toxicity studies (for example, 28-day or 90-day studies, or OECD TG 421 or 422 screening studies) indicate adverse effects on reproductive organs or tissues or reveal other concerns regarding reproductive toxicity.

63 Furthermore, it should be noted that column 1 of Annexes VII to X to the REACH Regulation sets out the studies in respect of which a registrant is required to provide summaries, depending on the quantities manufactured or imported for which the registrant intends to register the substance in question. Column 2 of Annexes VII to X to the REACH Regulation lists the specific rules on adapting or replacing the required standard information by other information, provided at a different stage or adapted in another way (specific adaptations). Where the conditions set out in column 2 of Annexes VII to X to the REACH Regulation allow specific adaptations to be proposed, a registrant must clearly indicate that fact and state the reasons for each adaptation under the appropriate heading in the registration dossier. In addition to those specific adaptations, a registrant may refrain from conducting a study where they can demonstrate that the conditions set out in Annex XI to the REACH Regulation are met (general adaptations).

64 Furthermore, the second paragraph of Annex IX to the REACH Regulation states that column 1 of that annex establishes the standard information required for all substances manufactured or imported in quantities of 100 tonnes or more in accordance with Article 12(1)(d) of that regulation.

65 It follows from those provisions that, where available repeated dose toxicity studies indicate adverse effects on reproductive organs or tissues or reveal other concerns regarding reproductive toxicity, ECHA is under an obligation to require registrants, for substances manufactured or imported in quantities of 100 tonnes or more, to provide information on an EOGRTS with a basic design, unless an adaptation (specific or general) can be applied.

66 Since it is common ground between the parties that no general or specific adaptation was applicable in the present case, ECHA was therefore under an obligation, after finding that the available repeated dose toxicity studies revealed other concerns regarding reproductive toxicity, to require the applicants to provide information on an EOGRTS with a basic design.

67 Contrary to what the applicants claim, ECHA was therefore not under an obligation to examine whether the EOGRTS was necessary and whether it was the most appropriate study to address the concerns alleged under Section 8.7.3. of Annex IX to the REACH Regulation (see, to that effect and by analogy, judgments of 29 March 2023, Nouryon Industrial Chemicals and Others v Commission , T‑868/19, EU:T:2023:168, paragraphs 69 and 70, and of 28 June 2023, Polynt v ECHA , T‑207/21, not published, EU:T:2023:361, paragraph 110).

68 Since the EU legislature decided that the measure requiring the provision of information on an EOGRTS with a basic design was proportionate, it cannot be alleged that ECHA infringed the principle of proportionality by adopting that measure.

69 Moreover, the applicants have not alleged that the REACH Regulation is unlawful in the light of the principle of proportionality.

70 The fact that ECHA had a margin of discretion in assessing the complex question of whether the available repeated dose toxicity studies revealed indications of adverse effects on reproductive organs or tissues or other concerns regarding reproductive toxicity does not call into question that conclusion.

71 The existence of discretion concerns the issue of whether the conditions laid down by the EU legislature in Section 8.7.3. of Annex IX to the REACH Regulation are satisfied and must be distinguished from the issue of the consequences attached by the EU legislature to the fulfilment of those conditions, in respect of which ECHA has no discretion.

72 The applicants’ arguments in that regard must therefore be rejected.

73 As regards the applicant’s argument alleging infringement of Article 25 of the REACH Regulation, it should be noted, as a preliminary point, that the EU legislature established, as the main purpose of the obligation to register laid down in Article 6 of the REACH Regulation, the objective of ensuring a high level of protection of human health and the environment. The means by which to achieve that objective is, as recital 19 of the REACH Regulation states, the registration obligation imposed on manufacturers and importers, which includes the obligation to generate data on the substances which they manufacture or import, to use those data to assess the risks related to those substances and to develop and recommend appropriate risk management measures (judgment of 7 July 2009, S.P.C.M. and Others , C‑558/07, EU:C:2009:430, paragraphs 45 and 46).

74 It is true that the objective of ensuring animal protection is also pursued by the REACH Regulation, in particular by Article 13(1) and Article 25(1) thereof. According to that latter provision, testing on vertebrate animals for the purposes of that regulation is to be undertaken only as a last resort (judgment of 28 June 2023, Polynt v ECHA , T‑207/21, not published, EU:T:2023:361, paragraph 107).

75 However, the information requirements set out in Annexes VII to X to the REACH Regulation confirm that animal testing cannot be avoided in every case. In some cases, only testing on vertebrate animals will provide sufficient scientific information to enable measures to be taken to protect human health and the environment (judgment of 28 June 2023, Polynt v ECHA , T‑207/21, not published, EU:T:2023:361, paragraph 108).

76 Thus, since, in the present case, ECHA could not exercise any discretion as to the consequences attached to the fulfilment of the conditions laid down by the legislature in Section 8.7.3. of Annex IX to the REACH Regulation, it cannot be criticised for not having accepted the step-by-step approach proposed by the applicants consisting of first carrying out an OECD TG 421 study and, therefore, for having infringed Article 25 of the REACH Regulation by imposing an approach requiring the sacrifice of a higher number of vertebrate animals (see, to that effect and by analogy, judgment of 28 June 2023, Polynt v ECHA , T‑207/21, not published, EU:T:2023:361, paragraph 111).

77 The applicants’ arguments in that regard must therefore be rejected.

78 In the light of the foregoing, the second plea must be rejected.

The third plea in law, alleging that ECHA made manifest errors of assessment and infringed the principles of legal certainty and of the protection of legitimate expectations by requiring the applicants to submit information on an EOGRTS with a basic design

79 The applicants argue, in essence, that ECHA made a manifest error of assessment in taking the view that the information in the registration dossier for the substance concerned indicated ‘other concerns’ or ‘adverse effects’ such as to allow ECHA to require an EOGRTS with a basic design.

80 ECHA disputes that line of argument.

81 As a preliminary point, it must be noted that the applicants’ argument directed against the initial decision, alleging that they demonstrated on the basis of the raw data in the study reports that the findings relating to the cauda epididymal spermatid count in the context of the OECD TG 408 study were incidental and were of no toxicological or other relevance, in view of all other unchanged male reproductive parameters, is based on the applicants’ reference to their comments on the draft initial decision set out in the annex.

82 However, it is not for the Court to seek and identify in the annexes the pleas and arguments on which it may consider the action to be based, since the annexes have a purely evidential and instrumental function (see judgment of 11 July 2007, Asklepios Kliniken v Commission , T‑167/04, EU:T:2007:215, paragraph 40 and the case-law cited).

83 That argument must therefore be rejected, without it being necessary to determine whether it refers to grounds of the initial decision which are expressly or implicitly set out in the contested decision.

84 In the contested decision, the part of the operative part which imposes on the applicants the obligation to submit, by 26 December 2026, information taken from an EOGRTS with a basic design is based on the finding that the OECD TG 408 study reveals another reproductive toxicity concern, but also on the finding that the OECD TG 407 study indicates the existence of adverse effects.

85 As regards the results of the OECD TG 407 study, it should be noted that, in the contested decision, the Board of Appeal found that that study showed testicular tubular degeneration or atrophy at all dose levels (20, 60 and 200 mg/kg bw/day) compared to the control group and that those effects constituted adverse effects on reproductive tissues.

86 After acknowledging that the OECD TG 407 study did not make it possible to ascertain whether the effects observed were statistically significant or whether there was a dose-response relationship, the Board of Appeal found, first, that the relatively low absolute number of animals in each group which showed those effects resulted from the low statistical power of the study, that is, from the fact that the size of the group included only five male animals per treatment group and, second, that the fact that certain effects had also occurred in one animal in the control group did not mean that the effects observed in all the affected animals in all the other groups had to be disregarded.

87 In the first place, the applicants claim to have demonstrated during the administrative procedure that the results of the OECD TG 407 study were of no toxicological or other relevance because they were observed at only a limited severity in a small number of animals, including one animal in the male control group, and with no dose-response relationship. The applicants add that they have demonstrated that those results could not be attributed to the treatment of the animals with the substance concerned. Furthermore, such results were not, according to the applicants, observed at comparable levels in the OECD TG 408 study with a three times longer treatment duration.

88 The applicants also state that the Board of Appeal acknowledged that the OECD TG 407 study did not make it clear whether the effects observed were statistically significant and whether there was a dose-response relationship. They also state that ECHA relies on arbitrary and random considerations which lead it to the conclusion that the effects observed in all the affected animals in all the groups other than the control group cannot be disregarded.

89 In particular, the applicants argue that ECHA made a manifest error of assessment by failing to assess whether the results of the OECD TG 407 study could be attributed to the substance concerned and by simply and summarily accepting that the ‘adverse effects’ condition was met because the ‘effects observed … cannot be disregarded’.

90 According to the applicants, such a simplistic and superficial conclusion makes the trigger in Section 8.7.3. of Annex IX to the REACH regulation devoid of any meaning and is in blatant disregard of the obligations imposed on ECHA. Furthermore, in their view, that conclusion fails to take account of the arguments put forward by them, to the effect that the OECD TG 407 and TG 408 conclusions, which do not even align, should be rejected from both a scientific and a legal perspective.

91 In that regard, it must be noted that the applicants’ line of argument consists, in essence, in criticising the low statistical importance of the OECD TG 407 study and in disputing the relevance of the results of that study on the ground that certain effects also occurred in one animal in the control group.

92 However, the applicants merely make general and unsubstantiated assertions and refer, in general terms, to their observations made during the administrative procedure and to the fact that they have, in their view, demonstrated that the conclusions of that study could not be attributed to the treatment of the animals with the substance concerned.

93 In the absence of more specific evidence, it must be held that those arguments are not such as to render implausible the Board of Appeal’s assessments and must, therefore, be rejected.

94 In the second place, the applicants claim, with regard to the OECD TG 407 and TG 408 studies, that, in its initial decision, ECHA took into account only the statistical significance of those studies, ignoring the other relevant parameters referred to in the ECHA Guidance on Information Requirements and Chemical Safety Assessment, dated June 2017, thus infringing the principle of legal certainty and the principle of legitimate expectations.

95 However, as the Board of Appeal notes in paragraph 65 of the contested decision, the extracts from the Guidance on Information Requirements and Chemical Safety Assessment produced by the applicants provide a purely indicative, not mandatory, list of factors that may be taken into consideration in the assessment of the conditions under Section 8.7.3. of Annex IX to the REACH Regulation.

96 Consequently, the fact that ECHA gave more weight to one of the criteria listed in that guidance, namely that of the statistical significance of the studies, is not such as to demonstrate an infringement of that guidance and, therefore, an infringement of the principle of the protection of legitimate expectations or an infringement of the principle of legal certainty.

97 The arguments directed against the grounds relating to the OECD TG 407 study must therefore be rejected.

98 In those circumstances, it must be noted that, according to case-law, where the operative part of a decision is based on several pillars of reasoning, each of which would in itself be sufficient to justify that operative part, that decision should, in principle, be annulled only if each of those pillars is vitiated by an illegality. In such a case, an error or other illegality which affects only one of the pillars of reasoning cannot be sufficient to justify annulment of the decision at issue because that error could not have had a decisive effect on the operative part adopted by the institution (judgment of 15 September 2021, France v ECHA , T‑127/20, not published, EU:T:2021:572, paragraph 33).

99 In the present case, as follows from paragraph 84 above, the operative part of the contested decision, in so far as it requires the applicants to submit information on an EOGRTS with a basic design, is based on the finding that the OECD TG 408 study reveals another reproductive toxicity concern, but also on the finding that the OECD TG 407 study indicates the existence of adverse effects on reproductive tissues.

100 Since, for the reasons set out in paragraphs 91 to 97 above, the applicants have not succeeded in calling into question the lawfulness of the statement of reasons based on the OECD TG 407 study, and since that statement of reasons is in itself sufficient to justify the part of the operative part of the contested decision requiring the applicants to submit information on an EOGRTS with a basic design, the applicants’ arguments relating to the ground based on the OECD TG 408 study are not such as to lead to the annulment of the contested decision.

101 Those arguments must therefore be rejected as ineffective.

102 In the light of the foregoing, the third plea must be rejected.

103 The Court considers it appropriate to deal with the fifth plea before the fourth plea.

The fifth plea in law, alleging that ECHA made manifest errors of assessment and infringed the principles of legal certainty and of the protection of legitimate expectations by requiring that the EOGRTS include additional cohorts 2A and 2B

104 The applicants submit, in essence, that the conditions for triggering a request for additional cohorts 2A and 2B were not satisfied.

105 It should be recalled that, under the second paragraph of column 2 of Section 8.7.3. of Annex IX to the REACH Regulation, an extended one-generation reproductive toxicity study including cohorts 2A and 2B (developmental neurotoxicity) and/or cohort 3 (developmental immunotoxicity) is to be proposed by the registrant or may be required by ECHA in case of particular concerns on (developmental) neurotoxicity or (developmental) immunotoxicity justified, inter alia, by specific mechanisms or modes of action of the substance with an association to (developmental) neurotoxicity and/or (developmental) immunotoxicity (for example, cholinesterase inhibition or relevant changes in thyroidal hormone levels associated to adverse effects).

106 In the contested decision, the Board of Appeal noted that, in the initial decision, that condition had been found to have been met as regards cohorts 2A and 2B (developmental neurotoxicity) on account of the results of two studies.

107 The OECD TG 408 study showed a greater incidence of minimal or diffuse hypertrophy of the follicular epithelium at all dose levels compared to the control group.

108 The study from the database on chemical products in Japan (‘the JECDB study’) showed a significant increase, in relative and absolute terms, in thyroid weights and follicular thyroid hyperplasia in males and females at the highest dose (1 000 mg/kg bw/day).

109 Furthermore, the Board of Appeal stated that ECHA had inferred from this that those effects showed that the substance concerned could affect the functioning of the thyroid and, consequently, thyroid hormone levels and the neurological development of the foetus.

110 In the first place, the applicants submit that the findings regarding thyroid weight, observed in the JECDB study and the OECD TG 408 study, do not constitute a ‘particular concern’ that could be relied upon as a trigger for the study of additional cohorts.

111 The applicants complain, in particular, that ECHA presumed the existence of a particular concern for a thyroid-related mechanism or mode of action of the substance concerned, even though that presumption has, in their view, no basis.

112 In that regard, the applicants state that they argued in their comments on the draft initial decision, as regards the conclusions drawn on the basis of the JECDB study, that the translation of the report of that study was not always clear, that the difference as regards the highest doses used in that study and in the other repeated dose studies (namely, respectively, 150 and 200 mg/kg bw/day in the OECD TG 407 and TG 408 studies compared to 1 000 mg/kg bw/day in the JECDB study) was remarkable, that the study report did not include a statement on good laboratory practices and that the year in which the JECDB study was carried out was not known. The applicants also claimed that that study was included in the dossier only as a study in support of the OECD TG 408 study and that, in the light of their comments, it should be regarded as an unknown reliability study according to the ‘Klimisch’ classification.

113 According to the applicants, ECHA cannot rely, as a factor triggering a request for additional cohorts, on studies which have limitations as regards, for example, their reliability or relevance.

114 In addition, for the applicants, ECHA did not address their comments or explain how it could rely on conclusions based on a low quality study.

115 In that regard, the Board of Appeal found that the conclusions of the OECD TG 408 study were sufficient to show that the substance concerned could cause (developmental) neurotoxicity effects, with the result that the applicants’ argument that the JECDB study should be disregarded was, in the Board of Appeal’s view, ineffective.

116 The Board of Appeal added, for the sake of completeness, that the results of a study of unknown or limited reliability could still be informative, provided that its results were carefully assessed, and then it explained why the results of the JECDB study could be taken into account to support the results of the OECD TG 408 study.

117 In the circumstances in which the Board of Appeal took into account the limitations of the JECDB study and drew conclusions from those limitations, taking the view that that study could be taken into account only in support of the OECD TG 408 study, the applicants’ arguments do not demonstrate that the Board of Appeal’s assessment of the JECDB study is implausible or that it did not take the applicants’ comments into account.

118 Those arguments must therefore be rejected.

119 In the second place, as regards the OECD TG 408 study, the applicants submit that ECHA made manifest errors of assessment and breached its duty to take all relevant information into account by not considering the information submitted by the applicants.

120 First, the applicants state that, in the OECD TG 408 (90-day) study, minimal to mild diffuse follicular thyroid hypertrophy was noted both in the animals treated, at all dose levels, and in the control group. The applicants add that, in the JECDB study, slight to moderate diffuse follicular thyroid hyperplasia was observed in the high-dose group (tested at 1 000 mg/kg bw/day, that is, a dose almost seven times higher than that used in the OECD TG 408 90-day study) at the end of the treatment period, and that minimal diffuse follicular thyroid hyperplasia was observed at that dose in only one out of five males and in one out of five females at the end of the 14-day recovery period, which, according to the applicants, indicates the reversibility of the finding.

121 The applicants rely, in that regard, on a report by the European Society of Toxicologic Pathology (Huisinga, M., Bertrand, L., Chamanza, R., and Others, ‘Adversity Considerations for Thyroid Follicular Cell Hypertrophy and Hyperplasia in Nonclinical Toxicity Studies: Results From the 6th ESTP International Expert Workshop’, Toxicologic Pathology , 2020, pp. 920 to 938) (‘the ESTP report’), which highlighted the physiological capacity of the thyroid gland to compensate for hormonal and metabolic changes. The applicants submit that, according to the ESTP report, diffuse follicular thyroid cell hypertrophy and hyperplasia, in the absence of other morphological changes such as focal hyperplasia or neoplasia, should not be regarded as intrinsically adverse effects.

122 Consequently, according to the applicants, even if the histopathological findings are limited in the current studies, ECHA could not conclude, in the initial decision, that ‘[the substance concerned] perturbs thyroid hormone signalling’. Moreover, since no hormone measurements were performed with the substance concerned in the relevant studies, ECHA could not, in the applicants’ view, validly conclude on the disturbance of thyroid hormone signalling. The Board of Appeal disregarded the evidence which, from that perspective, was in the form of the ESTP report.

123 In that regard, the Board of Appeal found that the question that arose in the present case was not whether the effects observed in the OECD TG 408 study had to be regarded as inherently adverse in themselves, but whether the results of that study gave reasonable grounds for considering that the substance concerned could cause developmental neurotoxicity effects.

124 It should be noted that, as the Board of Appeal found, under the second paragraph of column 2 of Section 8.7.3. of Annex IX to the REACH Regulation, an EOGRTS including cohorts 2A and 2B (developmental neurotoxicity) may be required by ECHA in case of particular concerns related to (developmental) neurotoxicity justified, inter alia, by specific mechanisms or modes of action of the substance with an association to (developmental) neurotoxicity.

125 It follows that, as ECHA submits, the conclusion of the ESTP report that follicular thyroid cell hypertrophy and hyperplasia, in the absence of other morphological changes such as focal hyperplasia or neoplasia, should not be considered to be intrinsically adverse effects, could be regarded as irrelevant.

126 An EOGRTS including cohorts 2A and 2B is justified where there are specific mechanisms or modes of action of the substance with an association to (developmental) neurotoxicity, with the result that ECHA was not required to demonstrate the existence of adverse effects.

127 Furthermore, as regards the argument that there were no direct thyroid hormone level measurements, it must be noted that the applicants have not specifically disputed the finding that ECHA could reasonably infer from histopathological changes in the thyroid that the functioning of the thyroid could be affected, that an impaired functioning of the thyroid could lead to changes in thyroid hormone levels in the mother and that those changes in hormone levels could cause (developmental) neurotoxicity effects in the foetus.

128 The applicants’ arguments therefore do not demonstrate that the contested decision is vitiated by a manifest error of assessment in that regard and must, as a result, be rejected.

129 Second, for the applicants, the results of the OECD TG 408 study did not give reasonable grounds for considering that the substance concerned could cause developmental neurotoxicity effects.

130 The applicants submit that the contested decision reveals a lack of precision in ECHA’s decision-making process. According to them, the OECD TG 408 (90-day) study indicates the presence of minimal diffuse hypertrophy of the follicular epithelium in the thyroid glands in five out of ten males and in three out of ten females in the control group. In the applicants’ submission, the Board of Appeal, by stating that the effects measured by that study were ‘seen in one animal in the control group’, therefore made an error of fact, which cannot be regarded as a mere clerical error. The applicants explain that, because of that error, ECHA failed to take into account the clear evidence that the minimal diffuse hypertrophy of the follicular epithelium in the thyroid glands was not treatment-related because it was also observed in numerous animals in the control group.

131 According to the applicants, by stating that ‘even in the presence of effects in a single animal in the control group, the OECD TG 408 study shows a greater incidence of minimal or diffuse hypertrophy of the follicular epithelium at all dose levels in the groups of treated animals compared to the control group’, the Board of Appeal repeated the error regarding the control group and made another error by referring to ‘minimal or diffuse hypertrophy’ when the recorded finding in that study was ‘diffuse follicular epithelial hypertrophy’ described as ‘minimal’.

132 In that regard, it should be noted that the Board of Appeal stated that the OECD TG 408 study showed histopathological changes in the thyroid consisting of diffuse follicular thyroid hypertrophy at all dose levels (15, 50 and 150 mg/kg bw/day) in both male and female rats.

133 Furthermore, according to the Board of Appeal, the fact that the effects observed in the OECD TG 408 study were not clearly linked to the dose administered and were also observed in a single animal in the control group did not invalidate that assessment. Even if it could not be demonstrated that the effects observed were linked to the dose administered, in the context of that study, they could nevertheless constitute reasonable grounds for considering that the substance concerned could cause (developmental) neurotoxicity effects.

134 The Board of Appeal added that the fact that certain effects were observed in a single animal in the control group did not mean that similar effects observed in all other groups could be disregarded since the OECD TG 408 study showed a greater incidence of minimal or diffuse hypertrophy of the follicular epithelium at all dose levels in the groups of treated animals compared to the control group.

135 In that regard, it must be noted that, as the applicants submit without being contradicted by ECHA, the contested decision is vitiated by an error relating to the number of control group animals in which effects were observed.

136 By referring to effects in a single animal in the control group, whereas a diffuse hypertrophy of the follicular epithelium had been observed in five out of ten males and in three out of ten females in that group, the Board of Appeal vitiated the contested decision by an error as to the accuracy of the facts within the meaning of the case-law referred to in paragraph 16 above.

137 Contrary to what ECHA maintains, that error, which led the Board of Appeal to find that effects had been observed in 5% of animals in the control group (1 out of 20) instead of 40% of the animals in that group (8 out of 20), cannot, on account of its seriousness, be regarded as a mere clerical error which could have been remedied by replacing the reference to ‘one animal’ with the reference to ‘some animals’.

138 In accordance with the case-law referred to in paragraph 16 above, that error as to the accuracy of the facts vitiates, in itself, the Board of Appeal’s finding that the OECD TG 408 study showed a greater incidence of minimal or diffuse hypertrophy of the follicular epithelium at all dose levels in the groups of treated animals compared to the control group.

139 As a result of that error as to the accuracy of the facts, which the EU judicature must specifically review when ECHA exercises its broad discretion (see, to that effect and by analogy, judgment of 4 May 2023, ECB v Crédit lyonnais , C‑389/21 P, EU:C:2023:368, paragraph 55 and the case-law cited), the conclusion that the OECD TG 408 study showed histopathological changes in the thyroid, consisting of diffuse follicular thyroid hypertrophy at all dose levels (15, 50 and 150 mg/kg bw/day) in both male and female rats, is based on materially incorrect facts.

140 Contrary to ECHA’s submissions, it cannot be ruled out that, if that error had not been made, the Board of Appeal would not have concluded that the substance concerned had developmental neurotoxicity effects.

141 In addition, as follows from the case-law referred to in paragraph 16 above, the existence of an error as to the material accuracy of the facts that were relied on by ECHA means that there is no need for the Court to verify whether there has been any manifest error in the assessment of those facts. The EU judicature cannot review the legality of the assessment of incorrect facts.

142 Moreover, such a review would amount, for the Court, to substituting its assessment of highly complex facts for that of ECHA in a manner contrary to the case-law referred to in paragraph 17 above and would be contrary to the applicants’ right to effective judicial protection; they were not, in any event, in a position to understand the assessment of incorrect facts that was made in the contested decision.

143 Furthermore, ECHA’s reference to the grounds of the initial decision also does not make up for the error vitiating the grounds of the contested decision.

144 In accordance with the case-law cited in paragraph 24 above, the Board of Appeal’s decision replaced the initial decision, with the result that the grounds of the initial decision which were not reproduced as grounds in the contested decision cannot be regarded as included in the grounds of the latter decision.

145 That is all the more so since the grounds of the initial decision which were not reproduced in the contested decision cannot be challenged before the Court, with the result that to regard those grounds as forming part of the statement of reasons for the contested decision would prejudice the applicants’ right to effective judicial protection.

146 The contested decision is therefore vitiated by a material error as to the facts assessed by the Board of Appeal.

147 Accordingly, the fifth plea must be upheld and the action against the contested decision must be upheld in part in so far as that decision requires that the EOGRTS on which the applicants must submit information is to include additional cohorts 2A and 2B (developmental neurotoxicity).

148 Since the fifth plea has been upheld, there is no longer any need to rule on the fourth plea.

Costs

149 Under Article 134(3) of the Rules of Procedure of the General Court, where each party succeeds on some and fails on other heads, the parties are to bear their own costs. However, if it appears justified in the circumstances of the case, the General Court may order that one party, in addition to bearing his own costs, pay a proportion of the costs of the other party.

150 In the present case, since the applicants and ECHA have been partially unsuccessful, each party must be ordered to bear its own costs.

On those grounds,

THE GENERAL COURT (Sixth Chamber, Extended Composition)

hereby:

1. Annuls Decision A-009-2022 of the Board of Appeal of the European Chemicals Agency (ECHA) of 19 September 2023 in so far as that decision requires that the extended one-generation reproductive toxicity study on which Nouryon Functional Chemicals BV, Arkema GmbH, Pergan Hilfsstoffe für industrielle Prozesse GmbH and United Initiators GmbH must submit information is to include additional cohorts 2A and 2B (developmental neurotoxicity);

2. Dismisses the action as to the remainder;

3. Orders ECHA, Nouryon Functional Chemicals, Arkema, Pergan Hilfsstoffe für industrielle Prozesse and United Initiators each to bear their own costs.

Costeira

Kancheva

Öberg

Zilgalvis

Tichy-Fisslberger

Delivered in open court in Luxembourg on 19 November 2025.

V. Di Bucci

M. van der Woude

Registrar

President

* Language of the case: English.