JUDGMENT OF THE GENERAL COURT (Fourth Chamber)

29 March 2023 (*) (1)

(REACH – Evaluation of registration dossiers and compliance check of information provided by registrants – Request for further studies for the purposes of the registration dossier for dimethyl ether – Pre-natal developmental toxicity study – Extended one-generation reproductive toxicity study – Dose-range finding study – Article 51(7) of Regulation (EC) No 1907/2006 – Animal testing – Article 25 of Regulation No 1907/2006 – Manifest error of assessment – Proportionality)

In Case T‑868/19,

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

Knoell NL BV, established in Maarssen (Netherlands),

Grillo-Werke AG, established in Duisburg (Germany),

PCC Trade & Services GmbH, established in Duisburg,

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

applicants,

v

European Commission, represented by R. Lindenthal and K. Mifsud-Bonnici, acting as Agents,

defendant,

supported by

Kingdom of Denmark, represented by M. Søndahl Wolff, acting as Agent,

by

Kingdom of the Netherlands, represented by M. Bulterman, A. Hanje and J. Langer, acting as Agents,

by

Kingdom of Sweden, represented by A. Runeskjöld, C. Meyer-Seitz, M. Salborn Hodgson, H. Shev, H. Eklinder, R. Shahsavan Eriksson and O. Simonsson, acting as Agents,

and by

European Chemicals Agency (ECHA), represented by M. Heikkilä, W. Broere, S. Mahoney and N. Herbatschek, acting as Agents,

interveners,

THE GENERAL COURT (Fourth Chamber),

composed, at the time of the deliberations, of S. Gervasoni, President, L. Madise (Rapporteur) and P. Nihoul, Judges,

Registrar: M. Zwozdziak-Carbonne, Administrator,

having regard to the order of 30 April 2020, Nouryon Industrial Chemicals and Others v Commission (T‑868/19 R, not published, EU:T:2020:171), by which the applicants’ application for interim measures was dismissed,

having regard to the written part of the procedure,

further to the hearing on 15 September 2022,

gives the following

Judgment

1        By their action based on Article 263 TFEU, the applicants, Nouryon Industrial Chemicals BV, Knoell NL BV, Grillo-Werke AG and PCC Trade & Services GmbH, seek annulment of Commission Implementing Decision C(2019) 7336 final of 16 October 2019 on the compliance check of a registration of dimethyl ether, adopted on referral by the European Chemicals Agency, on the basis of Article 51(7) of Regulation (EC) No 1907/2006 of the European Parliament and of the Council concerning the Registration, Evaluation, Authorisation and Restriction of Chemicals (REACH) (‘the contested decision’).

 Background to the dispute

2        Dimethyl ether is a chemical substance which serves as an aerosol propellant and as a solvent, used mainly in industrial and consumer aerosol products, including in pharmaceutical applications. It is also used as a chemical building block for another substance (dimethyl sulfate) and in oil extraction and engine fuel.

3        As regards the physical hazards which it presents, dimethyl ether is currently classified as an ‘extremely flammable gas (H220)’ and the label on its container states ‘contains gas under pressure; may explode if heated (H280)’ pursuant to Regulation (EC) No 1272/2008 of the European Parliament and of the Council of 16 December 2008 on classification, labelling and packaging of substances and mixtures, amending and repealing Directives 67/548/EEC and 1999/45/EC, and amending Regulation (EC) No 1907/2006 (OJ 2008 L 353, p. 1).

4        The applicants are manufacturers or importers of dimethyl ether established in the European Union or exclusive representatives acting on behalf of manufacturers of that chemical substance established outside the European Union. In accordance with the principle of ‘no data, no market’ laid down in Article 5 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) (‘the REACH Regulation’), the applicants, together with other registrants, on 30 November 2010, filed with the European Chemicals Agency (ECHA) an application for registration of dimethyl ether for manufactured or imported quantities of 1 000 tonnes or more per year per manufacturer or importer. Akzo Nobel Industrial Chemicals BV, subsequently renamed Nouryon Industrial Chemicals (‘the first applicant’), acted as the lead registrant for the joint registration dossier, in accordance with Article 11 of the REACH Regulation.

5        On 29 March 2016, ECHA initiated a procedure for a compliance check of the registration on the basis of Article 41 of the REACH Regulation.

6        On 26 April 2016, pursuant to paragraph 3 of that article, ECHA sent to the registrants a draft decision requesting them to supplement the registration dossier for dimethyl ether with information relating to the results of two studies, namely a pre-natal developmental toxicity study carried out on rabbits, representing a second animal study species, and an extended one-generation reproductive toxicity study carried out on rats.

7        In accordance with Article 51(1) of the REACH Regulation, ECHA’s draft decision was also communicated to the competent authorities of the Member States and the position of the first applicant, as the lead registrant representing all the registrants, and the positions of those authorities on that draft decision were subsequently brought to their mutual attention.

8        On 19 January 2017, pursuant to Article 51(4) of the REACH Regulation, ECHA sent to the registrants and to the ECHA Member State Committee provided for in Article 76(1)(e) of the REACH Regulation (‘the Member State Committee’) a revised draft decision to which the amendments proposed by certain competent authorities of the Member States were attached. The revised draft decision did not contain any amendment of ECHA’s requests to the registrants, but the reasons for the decision took a position on the comments submitted by the first applicant.

9        On 24 March 2017, the first applicant, as permitted by Article 51(5) of the REACH Regulation, submitted comments on the revised draft decision and on the amendments proposed by certain authorities of the Member States.

10      At its session on 25 and 26 April 2017, the Member State Committee failed to reach a unanimous agreement on the revised draft decision.

11      Following that disagreement, on 13 July 2017, ECHA sent its revised draft decision, together with the dossier, to the European Commission for the latter to adopt a final decision in the case, in accordance with Article 51(7) of the REACH Regulation.

12      On 18 February 2019, the Commission sent a draft implementing decision to all the co-registrants manufacturing or importing the substance in quantities equal to or greater than 1 000 tonnes per year.

13      On 15 March 2019, the first applicant submitted comments on the Commission’s draft decision.

14      On 16 October 2019, the Commission adopted the contested decision.

15      In the operative part of the contested decision, the Commission concluded that the registration of dimethyl ether did not comply with the information requirements as regards two different effects relating to reproductive toxicity, namely effects on pre-natal development and effects on one-generation reproduction (Article 1 of the contested decision). Consequently, in the contested decision, the Commission requires the registrants to provide information on the effects of dimethyl ether, based, in the first place, on a pre-natal developmental toxicity study as referred to in Section 8.7.2. of Annex X to the REACH Regulation (hereinafter, the annexes to the REACH Regulation are referred to solely by their Roman numeral), to be carried out, via inhalation, on a second animal species, namely rabbits (Article 2 of the contested decision) and, in the second place, on an extended one-generation reproductive toxicity study as referred to in Section 8.7.3. of Annex X, to be carried out on rats via inhalation. As regards that second study, the Commission states that a preliminary ‘dose-range finding’ study, conducted for example in accordance with the guideline of the Organisation for Economic Co-operation and Development (OECD) for the testing of chemicals No 421, intended in particular to detect possible narcotic-type effects (namely, sleepiness), must be carried out, in particular in order to determine whether it is necessary, depending on whether or not such effects are observed at any of the concentrations chosen for that extended study, to include, in that study, cohorts 2A and 2B which are intended specifically for the developmental toxicity study (Article 3 of the contested decision). The contested decision requires the applicants to submit, within 36 months of the date of notification of that decision, an updated version of the registration of dimethyl ether to ECHA, together with the results of the requested studies and, where relevant, to submit an update of the chemical safety report (Article 4 of the contested decision).

 Forms of order sought

16      In the application, the applicants claim that the Court should annul the contested decision and order the Commission to pay the costs.

17      The Commission contends that the Court should dismiss the action as unfounded and order the applicants to pay the costs.

18      The Kingdom of Denmark, the Kingdom of the Netherlands, the Kingdom of Sweden and ECHA, intervening in support of the Commission, contend that the Court should dismiss the action as unfounded. The Kingdom of the Netherlands and ECHA also contend that the applicants should be ordered to pay the costs of the proceedings.

 Law

 The first plea in law, alleging that the Commission infringed Article 51(7) of the REACH Regulation by adopting the contested decision which covers aspects on which the Member State Committee reached a unanimous agreement

19      As a preliminary point, it is necessary to note the circumstances in which the Commission was called upon to adopt the contested decision.

20      The contested decision was adopted under the procedural mechanism laid down in Article 51 of the REACH Regulation, which provides:

‘1.      [ECHA] shall notify its draft decision in accordance with Articles 40 or 41, together with the comments of the registrant, to the competent authorities of the Member States.

2.      Within 30 days of circulation, the Member States may propose amendments to the draft decision to [ECHA].

3.      If [ECHA] does not receive any proposals, it shall take the decision in the version notified under paragraph 1.

4.      If [ECHA] receives a proposal for amendment, it may modify the draft decision. [ECHA] shall refer a draft decision, together with any amendments proposed, to the Member State Committee within 15 days of the end of the 30-day period referred to in paragraph 2.

5.      [ECHA] shall forthwith communicate any proposal for amendment to any registrants or downstream users concerned and allow them to comment within 30 days. The Member State Committee shall take any comments received into account.

6.      If, within 60 days of the referral, the Member State Committee reaches a unanimous agreement on the draft decision, [ECHA] shall take the decision accordingly.

7.      If the Member State Committee fails to reach unanimous agreement, the Commission shall prepare a draft decision to be taken in accordance with the procedure referred to in Article 133(3).

8.      An appeal may be brought, [before the Board of Appeal of ECHA], against [ECHA’s] decisions under paragraphs 3 and 6 of this Article.’

21      After ECHA sent its revised draft decision to the Member State Committee (see paragraph 8 above), the latter agreed that it was necessary to require information on studies carried out under Sections 8.7.2. and 8.7.3. of Annex X, as requested by ECHA, namely a pre-natal developmental toxicity study carried out on rabbits, representing a second animal study species, and an extended one-generation reproductive toxicity study carried out on rats. However, it did not reach a unanimous agreement as regards the content of the second of those studies.

22      Unlike the other members of the Member State Committee, eight members of that committee argued that the extended one-generation reproductive toxicity study should include a specific additional study on the developmental neurotoxicity of dimethyl ether, in order to address relevant concerns in relation to humans.

23      In that situation, referring to Article 51(7) of the REACH Regulation, ECHA sent its revised draft decision to the Commission so that the latter could adopt a final decision in the case. If, conversely, the Member State Committee had reached a unanimous agreement on the revised draft decision of ECHA, the latter would have ‘[taken] its decision accordingly’, in accordance with Article 51(6) of the REACH Regulation, in other words ECHA would itself have adopted the final decision.

24      In their first plea in law, the applicants complain that the Commission, in so far as the Member State Committee failed to reach a unanimous agreement on only one specific point of discussion, namely whether or not to add a development neurotoxicity study in the one-generation reproductive toxicity study, infringed Article 51(7) of the REACH Regulation by adopting a decision which also covers aspects on which the Member State Committee reached a unanimous agreement.

25      The applicants submit that that ‘procedural error’ had an effect on their rights, since, if ECHA had adopted a decision on the points on which the Member State Committee had reached a unanimous agreement, the applicants could have brought an appeal before the Board of Appeal of ECHA, before which they have rights other than those which they have before the General Court, including the automatic suspension of the contested decision and the application of a different standard of review.

26      The Commission, supported by the Kingdom of Denmark, the Kingdom of the Netherlands, the Kingdom of Sweden and ECHA, disputes the applicants’ plea.

27      Contrary to what is claimed by the applicants, it is not apparent from Article 51(7) of the REACH Regulation that if a disagreement within the Member State Committee concerns only part of ECHA’s draft decision, ECHA must divide the final decision into one part that would be adopted by it on the basis of paragraph 6 of that article and into another part, the subject matter of the disagreement, which would be adopted by the Commission in accordance with paragraph 7 of that article.

28      Article 51 of the REACH Regulation, which is a procedural article and is entitled ‘Adoption of decisions under [registration] dossier evaluation’, lays down in its various provisions the conditions for the examination of draft decisions prepared by ECHA for that purpose, first by the competent authorities of the Member States and then, where appropriate, by the Member State Committee, and it also determines the conditions for the adoption of the final decisions that are the subject of its heading in different situations. Paragraph 7 of that article refers to the specific situation where there is a lack of unanimous agreement within that committee on the ‘draft decision’ of ECHA, by providing that, in that case, the Commission is to prepare ‘a draft decision’.

29      In interpreting that provision, it is necessary to consider not only its wording but also the context in which it occurs and the objectives pursued by the rules of which it is part (judgments of 17 November 1983, Merck, 292/82, EU:C:1983:335, paragraph 12, and of 19 July 2012, ebookers.com Deutschland, C‑112/11, EU:C:2012:487, paragraph 12). Reference is made in that regard to the literal, contextual (or systematic) and teleological interpretations respectively.

30      In the first place, in the context of a literal approach, it must be observed that that procedural provision does not state that the Commission should prepare a draft decision ‘on the aspects on which the Member State Committee has failed to reach a unanimous agreement’. In addition, Article 41(3) of the REACH Regulation, on the ‘compliance check of registrations’, which, for its part, determines the subject matter of a draft decision and then of a decision, which may be drawn up when such a compliance check is carried out, states in the last sentence that ‘such a decision’, that is to say, the decision following a draft decision, ‘shall be taken in accordance with the procedure laid down in [Article 51]’. None of the wording of the paragraphs of Article 51, where a ‘draft decision’ is mentioned, gives the impression that the purpose of those drafts differs from the purpose referred to in Article 41(3), namely an invitation to provide any information necessary to bring the registration into compliance. That wording therefore supports the interpretation that Article 41(3) of the REACH Regulation refers to the adoption of a single decision at the end of the procedure laid down in Article 51 of that regulation.

31      In the second place, in the context of a contextual approach, it must be observed that Article 51(6) of the REACH Regulation gives ECHA the power to adopt a decision the draft of which has been communicated to the Member State Committee only if, within 60 days of that communication, that committee reaches a unanimous agreement on that draft (see paragraph 20 above). It follows that, in the absence of such unanimous agreement within that period, as in the present case, ECHA loses the power to adopt a decision under Article 51 of the REACH Regulation following a compliance check in respect of a registration and that, consequently, the Commission’s power provided for in Article 51(7) of that regulation covers all of the aspects which have been examined by the Member State Committee, whether or not they gave rise to unanimous agreement within that committee.

32      In the third place, from a teleological perspective, in the light of the principle of good administration laid down in the Charter of Fundamental Rights of the European Union and of the principle of legal certainty, which is a general principle of EU law that requires, inter alia, that those concerned know precisely the extent of the obligations which are imposed on them (see, to that effect, judgment of 29 March 2011, ArcelorMittal Luxembourg v Commission and Commission v ArcelorMittal Luxembourg and Others, C‑201/09 P and C‑216/09 P, EU:C:2011:190, paragraph 68 and the case-law cited), it is more rational that, in the event of a disagreement within the Member State Committee, that is to say, under Article 76 of the REACH Regulation, within one of the components of ECHA, the Commission is to exercise its power over the entire compliance check in respect of the registration under consideration, in order to prevent the preparation, and then second-level review, of assessments regarding the evaluation of the effects and hazards of a chemical substance from being shared amongst several bodies (ECHA and the Commission, the Board of Appeal of ECHA and the General Court, respectively), which might lead to inconsistencies, when those assessments concern the same dossier for the registration of a substance and must retain their overall consistency.

33      Therefore, Article 51(7) of the REACH Regulation can be understood only as meaning that any disagreement within the Member State Committee on an aspect of a draft ECHA decision examined in the context of a compliance check of registrations constitutes a disagreement on that draft considered as a whole, which confers on the Commission the power to prepare a new draft decision evaluating a registration dossier and then to adopt a final decision in that regard under a ‘comitology’ procedure. Consequently, the Commission is right to submit that that provision does not limit its power solely to the specific parts of the draft ECHA decision which is the subject of a disagreement within the Member State Committee, but grants the Commission the power to take a decision on all the aspects addressed in that draft decision.

34      That analysis is not called into question by the applicants’ other arguments.

35      First of all, the applicants state that a disagreement within the Member State Committee has already led ECHA to divide the final decision in cases concerning a compliance check of registrations. However, even though ECHA did so in the past in certain cases, that does not in any way mean that, in the present case, it was required to act in the same way, in particular since it is not bound by its precedents (see, to that effect and by analogy, judgments of 6 May 2008, Parliament v Council, C‑133/06, EU:C:2008:257, paragraph 60, and of 16 January 2020, Iberpotash v Commission, T‑257/18, EU:T:2020:1, paragraph 78) and since ECHA’s decision-making practice cannot determine the review of legality carried out by the Court.

36      The applicants also state that they would have benefited from more guarantees if the final decision, as regards the aspects on which the Member State Committee had reached a unanimous agreement, had been adopted by ECHA. The review carried out by the Board of Appeal of ECHA differs from the review carried out by the General Court and is not limited, as in the case of the latter, to verifying whether there are manifest errors.

37      As the Commission and the Kingdom of the Netherlands submit in essence, the distinction drawn between situations in which the Board of Appeal of ECHA may intervene in the process for the review of an administrative decision requesting registrants to supplement the dossier for the registration of a chemical substance and situations in which there is no provision for that board to intervene, and the consequences which may ensue as regards the scope of that review, stem from the FEU Treaty and from the legislative framework of the REACH Regulation, more specifically from Article 51 thereof, which lays down, in one case, which is the subject of paragraph 6 of that article, an ECHA decision, and in the other case, which is the subject of paragraph 7 of that article, a Commission decision where there is a disagreement within the Member State Committee, that is to say, a disagreement within ECHA. First, the legislature provided for the possibility, for the Board of Appeal, as an administrative body of ECHA empowered to review a first decision adopted by ECHA, to exercise any power which lies within the competence of ECHA or to remit the case to the competent body of ECHA for further action, in accordance with Article 93(3) of the REACH Regulation. Secondly, Article 263 TFEU provides that Commission decisions are to be subject to judicial review by the EU judicature. The different nature of those reviews justifies the procedural differences between them and the different powers of the bodies which exercise them.

38      In that regard, the EU law applicable in the present case has the effect of differentiating between, on the one hand, an administrative review carried out, with regard to a first ECHA decision, by a higher authority, namely the Board of Appeal of ECHA, and, on the other hand, a judicial review in which, in the context of an action for annulment based on Article 263 TFEU, the review is carried out by the EU Courts in respect of a Commission decision. It has already been held that, in an action for annulment brought under Article 263 TFEU, the review carried out by the EU Courts consists, where it involves the assessment of highly complex scientific and technical facts, as may be so in the present case, in reviewing whether they are vitiated by a manifest error, a misuse of powers or whether the author of the decision clearly exceeded the limits of its discretion (see judgment of 20 September 2019, BASF Grenzach v ECHA, T‑125/17, EU:T:2019:638, paragraph 87 and the case-law cited; see also, to that effect, judgment of 15 October 2009, Enviro Tech (Europe), C‑425/08, EU:C:2009:635, paragraph 47). That limitation does not apply to the intervention by the Board of Appeal of ECHA, which is also a component of ECHA, as has already been pointed out. In that intervention, the Board of Appeal does not confine itself to reviewing the legality of the decision taken by ECHA, in view of, in particular, the latter’s discretion, but examines, in the context of the criteria set out in the legislation, whether it is appropriate to review the assessments which ECHA made. That is why the EU legislature made sure to include, in the composition of that Board of Appeal, persons with the necessary technical and scientific expertise to carry out that new assessment and why the nature of that board’s review of the scientific and technical assessments previously carried out within ECHA differs from the nature of a review carried out by the EU judicature (see, to that effect, judgments of 20 September 2019, BASF Grenzach v ECHA, T‑125/17, EU:T:2019:638, paragraphs 88 and 89, and of 20 September 2019, Germany v ECHA, T‑755/17, EU:T:2019:647, paragraph 55). It is not for the Court to call into question that difference and to confer on itself the powers of a body such as the Board of Appeal of ECHA (see, to that effect and by analogy, judgment of 25 July 2002, Unión de Pequeños Agricultores v Council, C‑50/00 P, EU:C:2002:462, paragraphs 44 and 45).

39      Furthermore, although the applicants have not raised a plea of illegality against Article 51(7) of the REACH Regulation, that difference in the review of the assessment of highly complex scientific and technical facts cannot make it possible, contrary to that provision as interpreted in the present judgment (see paragraph 33 above), to restrict the Commission’s power to take a decision, pursuant to that provision, on all the aspects of a draft ECHA decision that has been submitted to the Member State Committee where a disagreement arises within that committee on one or more aspects of that draft decision.

40      Consequently, the first plea must be rejected.

 The second plea in law, alleging that the Commission infringed Article 13(3) of the REACH Regulation and made a manifest error of assessment by requesting tests which run counter to the applicable legal requirements and which are not technically feasible

41      The applicants submit that the Commission infringed Article 13(3) of the REACH Regulation and made a manifest error of assessment by requesting ‘test[s] at concentrations that could produce effects while being safe’ and, more specifically, as regards the extended one-generation reproductive toxicity study,  by requesting ‘a dose level set to induce some toxicity at the highest dose level’.

42      As regards the alleged infringement of Article 13(3) of the REACH Regulation, the applicants submit that, involving tests on animals via inhalation of the substance, the pre-natal developmental toxicity study and the extended one-generation reproductive toxicity study should be carried out in accordance with the recommendations set out in OECD Guidance Document No 39 on acute inhalation toxicity tests (‘OECD GD 39’). It is an international method recognised by the Commission and by ECHA as being appropriate under Article 13(3) of the REACH Regulation, and is, moreover, referred to by various toxicity testing methods set out in the regulation on test methods adopted by the Commission pursuant to that provision (in the present case, Council Regulation (EC) No 440/2008 of 30 May 2008 laying down test methods pursuant to Regulation (EC) No 1907/2006 of the European Parliament and of the Council on the Registration, Evaluation, Authorisation and Restriction of Chemicals (REACH) (OJ 2008 L 142, p. 1, as amended; ‘the regulation on test methods’)).

43      It is apparent from OECD GD 39 that, in the inhalation studies in question, it is recommended that an air concentration of the substance tested not exceed half of its lower explosive limit (‘the LEL’) Since, for dimethyl ether, the LEL is 3.3% concentration, it is recommended that a concentration should not exceed 1.65%. At the same time, it is apparent from a chronic inhalation study in rats carried out by the Haskell laboratory in 1986 (‘the 1986 Haskell study’) that the concentration with no observed adverse effects (‘NOAEC’) for dimethyl ether is 2.5%. In view of those two values, the applicants submit that it is not possible to carry out the requested studies at a concentration producing a certain toxicity, that is to say, at least 2.5%, without disregarding the requirements of the applicable test methods, in this instance the OECD GD 39 requirements, leading to the maximum test concentration for dimethyl ether being set at 1.65%.

44      By failing to take into consideration the technical restrictions applicable to the studies imposed in the contested decision, the Commission also made a manifest error of assessment. Moreover, in view of the risks for animals and technical personnel, the laboratories which the applicants contacted refused to conduct tests at concentrations exceeding the indications in OECD GD 39.

45      The Commission, supported by the Kingdom of Denmark, the Kingdom of the Netherlands, the Kingdom of Sweden and ECHA, disputes the applicants’ plea.

46      As a preliminary point, it must be noted that the plea which those arguments support is only indirectly linked to the challenge to the obligation to carry out the studies requested in the contested decision, which is expressed in other pleas in the action, namely the third to eighth pleas. The purpose of the second plea, now examined, is solely, as is apparent from its wording, to challenge the substance’s concentration levels that are requested in the contested decision in order to conduct those studies.

47      It may be stated, at the outset, that nowhere in the contested decision does the Commission require that concentrations of dimethyl ether for the requested tests be exceeded in such a way as to render those tests dangerous, in breach of the applicable safety rules. It is true that the Commission requests, in the contested decision, in Article 3 of the operative part, that the extended one-generation reproductive toxicity study be conducted at a ‘dose level set to induce some toxicity at the highest dose level’, moreover by reproducing an instruction which appears, in essence, both in Section 21 of test method B 56 of the regulation on test methods as regards that study and in the second paragraph of Section 1.6.3 of test method B 31 of that regulation as regards the pre-natal developmental toxicity study. However, the Commission expresses that request in the general framework applicable to acute inhalation toxicity tests, by which it is bound, as are the registrants. That framework does recommend that a certain concentration of the tested substance not be exceeded based on that substance’s characteristics. The abovementioned two methods, which are explicitly referred to in the operative part of the contested decision, themselves state in their abovementioned provisions: ‘the dose levels should be based on toxic effects, unless limited by the physical/chemical nature of the test chemical’ (test method B 56) and ‘unless limited by the physical/chemical nature or biological properties of the test substance, the highest dose should be chosen with the aim to induce some … toxicity’ (test method B 31).

48      Moreover, in recital 11 of the contested decision, the Commission states that ‘[dimethyl ether] is a gas with explosive properties[;] it must be considered whether it is technically feasible to conduct the required studies without risk of explosion’. In that recital, the Commission states that it is taking into account the fact that, on the basis of OECD GD 39, the registrants stated that tests could be carried out up to a concentration of 1.65% and nevertheless points out that tests at higher concentrations (up to 5%) have already been carried out. It concludes that ‘the available data therefore indicate a possibility to test at concentrations that could produce effects while being safe in relation to the explosive properties’.

49      In that regard, it must be noted that OECD GD 39, the application of which is not disputed by the Commission in its pleadings, states, in paragraph 67, under Section 5.1.4, that ‘in the case of potentially explosive test chemicals, care should be taken to avoid conditions favourable for an explosion’ and that ‘for safety reasons[,] it is generally advisable to not exceed 50% of the [LEL]’. It is apparent from that wording that that limit, in this case 1.65%, is not a universal limit which must not be exceeded under any circumstances. Moreover, in paragraph 61 of the application and paragraph 15 of the reply, the applicants accept that tests on dimethyl ether may be possible up to a maximum of 2%.

50      It follows from the foregoing that the Commission left it to the registrants, of course in conjunction with the laboratories which they might use, to determine the maximum concentration to be used to produce a certain toxicity, but within the limits of the concentrations that might prove dangerous in view of the physicochemical properties of dimethyl ether.

51      It has therefore not been demonstrated in any way that, in the contested decision, the Commission required, contrary to the legally applicable provisions, that dangerous concentrations be reached for acute inhalation toxicity tests.

52      Furthermore, it is apparent from the documents produced by the applicants (Annexes A12 and A13) that there are at least two laboratories which consider that they are capable of carrying out the tests in question at a concentration of 1.65%, or even one of the tests at a concentration of up to 2%. The applicants’ argument calling into question the technical feasibility of the studies requested in the contested decision must therefore be rejected.

53      Consequently, the applicants’ arguments, of a scientific nature, seeking, in essence, to demonstrate that there are no signs of toxicity of dimethyl ether in the test animals up to a concentration of 1.65%, which would mean that the applicants would have to exceed that concentration in order to obtain ‘some toxicity at the highest dose level’, are ultimately ineffective under the present plea, which is based on criticism of requests for tests which, in the applicants’ view, run counter to the legal requirements applicable to the conduct of tests and which are not technically feasible. That will nevertheless be taken into account in the examination of the third plea, since, as the Commission states in the defence, the third plea largely repeats those arguments.

54      It follows from the foregoing that the second plea must be rejected as having no factual basis and that, consequently, contrary to what the applicants argue under that plea, the Commission neither infringed Article 13(3) of the REACH Regulation nor requested technically unfeasible tests by making a manifest error of assessment in that regard.

 The third plea in law, alleging that the Commission made a manifest error of assessment in requiring tests that do not provide any relevant information on dimethyl ether

55      The applicants state that all the available studies concerning dimethyl ether support the conclusion that no toxicity has been observed at concentrations of 1.65% or lower, or even at 2.5%, the latter concentration corresponding to the NOAEC which they put forward. In the reply, they dispute a NOAEC level of 0.5%, referred to in the Commission’s defence in respect of pregnant females and mentioned in the chemical safety report which the applicants submitted in the context of the registration of the substance. The observations leading to that value are not conclusive.

56      Furthermore, in the context of the assessment of the hazards for human health and the management of the relevant risks, which are among the elements covered by Annex I, relating to the general provisions concerning the evaluation of substances and the preparation of chemical safety reports, it appears that the actual uses of the substance by workers and consumers are not likely to cause narcotic effects in humans. Thus, a peak exposure when using hairspray would probably result in an exposure level of 0.07% for a few seconds, whereas, for a NOAEC of 2.5%, the maximum level of exposure which a human may face (DNEL) would be 0.1% for a long-term exposure over a full eight-hour working day, whereas with a peak exposure of a few minutes, there would have to be an exposure level of 8% to 12% to result in acute narcotic effects in humans. In the reply, the applicants add that the risk management measures have demonstrated a considerable degree of safety at the highest possible concentrations of that substance, in particular with a very conservative risk characterisation ratio.

57      Therefore, if the pre-natal developmental toxicity study and the extended one-generation reproductive toxicity study were carried out in accordance with OECD GD 39, they would not produce new or relevant information concerning dimethyl ether. The carrying out of those studies would thus run counter to the requirements under Article 1(1), Article 10(a), Article 12(1), Article 13(4) and Article 41 of the REACH Regulation, and under Note 1 of Annex VI to and recitals 17 and 26 of that regulation, from which it is apparent that, in carrying out a compliance check in respect of registrations, ECHA or the Commission may request only relevant and necessary information and must promote alternative methods to animal testing for the assessment of the hazards associated with substances. The requirements on information to be provided, set out in the annexes to the REACH Regulation, should not be read too strictly in order to meet those objectives and to give them practical effect. In particular, standard information requirements should not be interpreted in such a way as to require registrants to provide information that is manifestly irrelevant.

58      In the reply, the applicants add in particular that, even though the Commission relies, in the defence, on a study based on ‘whole body’ inhalation tests on rats, referred to in recital 15 of the contested decision, during which sluggishness was observed at a concentration of 1% dimethyl ether, the Commission makes a manifest error of assessment in considering that such narcosis may constitute an adverse effect within the meaning of Article 1(3) of the REACH Regulation. According to the applicants, narcosis is adverse when animals are non-responsive, and not when, as occurred during those tests, only some sluggishness is observed for a short period of time after exposure. A simple effect was therefore observed with a concentration of 1%, but not an adverse effect.

59      As regards the Commission’s assertion, expressed in the defence, that toxicological effects could even be detected at a concentration level below 1% for animals exposed during the stages of pregnancy and lactation, the applicants’ counterargument is that, for that to occur, there would have to be a ‘molecular initiating event involving a biological interaction’, which is not possible for a noble gas, to which dimethyl ether is very similar.

60      Again in response to the defence, in which the Commission also states that it requested that a dose-range finding study be carried out prior to the extended one-generation reproductive toxicity study in order to ensure the usefulness of the test relating to developmental neurotoxicity on cohorts 2A and 2B, the applicants submit that that preliminary study itself is pointless, since the observations made on cohorts 1A and 1B at the start or in the context of that extended study as regards narcotic-like effects could provide the same results. In addition, in order to challenge the appropriateness of that study, the applicants state, inter alia, that no instructions were given by the Commission in order to interpret the observations made during that preliminary study and to draw conclusions from it.

61      It must first be observed that, of the studies requested in the contested decision, some seek to obtain standard information which, in all cases, it is necessary to provide in the registration dossier for dimethyl ether under Annex X, which is applicable in view of the declared level of manufacture or importation per year per manufacturer or per importer, in quantities of 1 000 tonnes or more. The applicants do not dispute that that is the case for the extended one-generation reproductive toxicity study in its basic configuration including only cohorts 1A and 1B but they argue, for questions on the interpretation of Annexes IX and X, that that is not the case for the developmental toxicity study carried out on a second animal species; the latter argument is rejected in the context of the examination of the eighth plea (see paragraph 168 below). In particular with regard to the studies which in all cases have to be carried out under Annex X, in so far as they seek to obtain standard information, the applicants consider, in essence, as indicated in paragraph 57 above, that the requirements under that annex should not be applied too strictly in order to prevent registrants from being required to provide information that is manifestly irrelevant by carrying out unnecessary tests on animals.

62      In order to examine that argument, it is necessary, first of all, to note the layout and role of the annexes to the REACH Regulation.

63      Under Article 10 of the REACH Regulation, which refers to the annexes to that regulation in detail, the application for registration must be accompanied by a technical dossier and a chemical safety report.

64      In that regard, that article provides, inter alia, that the technical dossier must contain study summaries or, where applicable, robust (that is to say, detailed) study summaries of the information derived from the application of Annexes VII to XI, and proposals for testing where listed in Annexes IX and X. Article 12 of the REACH Regulation states that, with regard to the studies in question, the technical dossier on the substance is to contain all the relevant physicochemical information available to the registrant (that is to say, regarding the physical and chemical characteristics, for example data on explosivity, inflammability and oxidising potential), toxicological (that is to say, regarding adverse effects for human beings) and ecotoxicological information (that is to say, concerning adverse effects on the environment) and at least certain information. For substances manufactured or imported in quantities of 1 000 tonnes or more per year per manufacturer or per importer, paragraph 1(e) of that article states that such minimum information is that specified in Annexes VII and VIII and testing proposals for the provision of the information are specified in Annexes IX and X.

65      Annex VI, entitled ‘Information requirements referred to in Article 10’, states, in the ‘Guidance note on fulfilling the requirements of Annexes VI to XI’, which is the introductory part of Annex VI:

‘Annexes VI to XI specify the information that shall be submitted for registration and evaluation purposes … For the lowest tonnage level, the standard requirements are in Annex VII, and every time a new tonnage level is reached, the requirements of the corresponding Annex have to be added. For each registration the precise information requirements will differ, according to tonnage, use and exposure. The Annexes shall thus be considered as a whole, and in conjunction with the overall requirements of registration, evaluation and the duty of care.’

66      In Annex X, entitled ‘Standard information requirements for substances manufactured or imported in quantities of 1 000 tonnes or more’, the introductory part states, inter alia:

‘At the level of this Annex, the registrant must submit a proposal and a time schedule for fulfilling the information requirements of this Annex in accordance with Article 12(1)(e).

Column 1 of this Annex establishes the standard information required for all substances manufactured or imported in quantities of 1 000 tonnes or more … Accordingly, the information required in column 1 of this Annex is additional to that required in column 1 of Annexes VII, VIII and IX. Any other relevant physicochemical, toxicological and ecotoxicological information that is available shall be provided. Column 2 of this Annex lists specific rules according to which the registrant may propose to omit the required standard information, replace it by other information, provide it at a later stage or adapt it in another way. …’

67      In Annex IX, relating to ‘Standard information requirements for substances manufactured or imported in quantities of 100 tonnes or more’, the introductory part is worded in a similar way, but in Annex IX there is of course no reference to Annex IX itself as an annex concerning lower quantity levels.

68      As regards, in particular, toxicological information, Annex X includes, in Section 8.7. relating to reproductive toxicity, a section numbered 8.7.2. ‘Developmental toxicity study’ and Section 8.7.3. ‘Extended One-Generation Reproductive Toxicity Study’. Those two sections are already present and listed in Annex IX, the first of which, in that annex, is entitled ‘Pre-natal developmental toxicity study’. However, their provisions are not identical in Annexes IX and X. Those two sections are not covered by Annexes VII and VIII which apply to substances manufactured or imported in quantities of, respectively, 1 tonne or more, and 10 tonnes or more, those quantities being per manufacturer or per importer.

69      It is apparent from that presentation that the information requested from the registrants under Annexes VII to X, in particular the standard information in column 1 of those annexes, which must in all cases be provided except where such information may be adapted under a provision in column 2, is requested incrementally based on the manufactured or imported quantities of substance. It must also be pointed out that those information obligations concern chemical substances and are intended, as set out in Article 1 of the REACH Regulation, to ensure that the hazards associated with those substances, when manufactured, placed on the market and used, are known and that those substances, when used, do not adversely affect human health or the environment. Consequently, in the light of the potential hazards of chemical substances and applying the precautionary principle, but also taking into account the objective of preventing unnecessary testing on vertebrate animals, that principle and that objective both also being referred to in that article, the legislature has already made choices in order to request registrants to carry out studies on vertebrate animals only if those studies appear relevant in view of the relevant quantities of substance. Moreover, Annex XI also provides for possibilities of adaptation which are additional to those provided for in column 2 of Annexes VII to X, which the registrants may put forward if they consider that a study provided for in those annexes serves no purpose.

70      Since the applicants do not challenge the validity of those choices, in other words the legality of the provisions under which, in the contested decision, they were requested to carry out studies, more specifically the provisions in Annex X, and since, under the plea currently being examined, they do not rely on the possibility of adaptation provided for in Annex XI, they cannot validly claim that they are not required to carry out studies which must in all cases be carried out under Annex X, which concerns the obtaining of standard information, on the ground that those studies are irrelevant.

71      The examples, cited in the application, of non-strict interpretations, according to the applicants, of EU acts, those interpretations having been proposed by the Advocates General of the Court (Opinion of Advocate General Trstenjak in Caffaro, C‑265/07, EU:C:2008:250, point 34, and of Advocate General Bobek in European Federation for Cosmetic Ingredients, C‑592/14, EU:C:2016:179, points 77 to 82) do not call into question that conclusion. In those cases, the Advocates General proposed interpretations of provisions the scope of which was discussed by focusing, beyond their wording, on their context and the general objectives of the acts in which they were set out, in order to clarify their scope. In the present case, the applicants’ wish is that certain provisions of the REACH Regulation not be applied even though the conditions for the application of those provisions are met, which is different.

72      Since the applicants’ argument of principle referred to in the last sentence of paragraph 57 above, which advocated a purely legal analysis, has been rejected, it must be observed that, for the remainder, the arguments put forward in support of the third plea seek to call into question the Commission’s assessment relating to the usefulness of the various studies that it requested, in so far as those studies are not in any event mandatory under Annex X, that is to say, the Commission’s assessment relating to the usefulness of the part of the extended reproductive toxicity study including cohorts 2A and 2B in order to assess developmental neurotoxicity and relating to the usefulness of the preliminary dose-range finding study.

73      Such an assessment falls within the category of assessments of highly complex scientific and technical facts by an administrative authority. As has already been pointed out in paragraph 38 above, if the EU judicature is called upon to examine such assessments, it must confine itself to ascertaining whether they are vitiated by a manifest error or a misuse of powers, or whether that authority has manifestly exceeded the limits of its discretion. In that regard, it has been consistently held that, in order to establish that the administrative authority made a manifest error in assessing those elements 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 institution which adopted the 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). In the light of the scientific and technical arguments put forward by the applicants, it is therefore necessary to ascertain whether those arguments render implausible the Commission’s assessment that, by carrying out the studies in conditions intended to ensure that the tests are not dangerous, that is to say by not exceeding a concentration of 1.65%, or even 2%, it is possible that some toxicity may manifest itself at the highest dose (for the sake of simplicity, reference will be made hereinafter only to the value of 1.65%).

74      Before beginning that verification, it is appropriate, however, to state the position taken by the Court with regard to the applicants’ request, expressed in the application, that the Court appoint an independent expert responsible for examining and clarifying certain complex scientific issues, which, moreover, relate to most of the applicants’ pleas, as they stated in the reply. The General Court would have availed itself of that possibility, provided for in Article 25 of the Statute of the Court of Justice of the European Union, only if that had proved necessary in order to decide whether or not certain pleas were well founded, given the nature of the review which it carries out in respect of the assessment of highly complex scientific and technical facts by an administrative authority, referred to in paragraph 73 above. However, as shown in the assessment of the present and the following pleas, that has not proved necessary in the present case.

75      With regard to the applicants’ arguments summarised in paragraphs 55 to 59 above, the Commission, supported by the Kingdom of Denmark, the Kingdom of the Netherlands, the Kingdom of Sweden and ECHA, submits, first of all, that the 2.5% NOAEC put forward by the applicants is not representative. It derives from the 1986 Haskell study (see paragraph 43 above), which is not a reproductive toxicity study such as those requested in the contested decision, but a toxicity study of a different type in which only non-pregnant adult animals were tested. That is not disputed by the applicants, who state, however, that a non-pregnant animal is just as sensitive to narcotic effects as a pregnant animal. The Commission and the Kingdom of Sweden consider, for their part, that the 1986 Haskell study does not provide information on the concentrations capable of giving rise to harmful effects in animals in sensitive stages of life, such as gestation and embryofoetal development.

76      Moreover, as referred to in paragraph 55 above, the Commission states that a NOAEC level of 0.5%, which is therefore below the level of 2.5% put forward by the applicants, was determined for pregnant females and was mentioned in the chemical safety report submitted by the applicants in connection with the registration of the substance. The applicants’ counterargument is that that value is unreliable and that, as a result, they did not include it in the technical registration dossier, in particular because the observations of the study following which that value had been determined are not clear, as is established by a report by a toxicologist from the Nederlandse Organisatie voor toegepast-natuurwetenschappelijk onderzoek (Netherlands Organisation for Applied Scientific Research, TNO), who himself consulted a specialist from the Haskell laboratory.

77      As regards the applicants’ arguments relating to specific uses by humans and relating to the assessment and management of risks in that regard, referred to in paragraph 56 above, the Commission, supported by the Kingdom of Denmark and ECHA, submits that compliance with the standard information requirements laid down by the REACH Regulation is not necessary solely for considerations relating to risk management measures in terms of use, but also for the identification of all the hazards associated with the substance on account of its intrinsic characteristics. The Commission also states that the risk characterisation ratio put forward by the applicants is irrelevant at the hazard assessment stage, which is linked to the intrinsic properties of the substance and is of use solely for management of the risk associated with the substance, which is different.

78      As stated in paragraph 58 above, the Commission relies on a study in which sluggishness was observed in rats at a concentration of 1% dimethyl ether, whereas the applicants submit that it is not possible to infer from that study that there may be harmful effects at a concentration of between 1% and 1.65%, which is the upper limit recommended in OECD GD 39 for tests regarding that substance.

79      Lastly, as regards the applicants’ challenge, summarised in paragraph 59 above, to the fact that dimethyl ether, which is similar to a noble gas, may have toxicological effects at a concentration below 1%, the Commission states that that substance is not a noble gas excluded from registration requirements under the REACH Regulation, that pups are more sensitive than adults at the same level of exposure, that it is therefore likely that the narcotic effects are more harmful to their nervous system than to that of adults and that a ‘biological interaction’ can hardly be excluded since the entire mechanism of toxicity of that substance is not known.

80      It must be stated that that debate does not render implausible the manifestation of effects of ‘some toxicity at the highest dose level’ while respecting the upper concentration limit of 1.65% during tests in studies regarding the reproductive toxicity of dimethyl ether, requested in the contested decision. First, the Court notes that there is precisely a disagreement between the applicants’ experts and those of the Commission or of the Kingdom of Denmark, the Kingdom of the Netherlands, the Kingdom of Sweden or of ECHA and that there is therefore a dispute of a scientific nature. Secondly, the applicants’ arguments do not in themselves clearly exclude the possibility of such effects arising under the conditions referred to above.

81      In that regard, in the first place, in the light of those arguments, it does not appear to be established, contrary to what the applicants argue, that it can be inferred automatically from an inhalation toxicity study not focused on a reproductive toxicity study and carried out on non-pregnant animals that the first concentration at which toxicity is detected in that study, and which makes it possible to infer a NOAEC for that type of studies, will be the same for a reproductive toxicity study carried out on, inter alia, pregnant animals, from which a same NOAEC can be inferred. The applicants refer only to an equivalence of narcotic reactions between pregnant animals and non-pregnant animals, which does not mean that other effects will be revealed at the same concentrations for those two types of animal. A 2.5% NOAEC for reproductive toxicity studies, and therefore the manifest lack of usefulness of tests carried out up to a concentration of 1.65%, cannot therefore necessarily be inferred in the present case from the 1986 Haskell study relied on by the applicants.

82      In the second place, as regards the criticism of the NOAEC level of 0.5% for pregnant females referred to in the chemical safety report submitted in connection with the registration of dimethyl ether, on the ground that the study which led to that value is not clear, it is however apparent from the extracts from the report on that study by a toxicologist of the TNO, reproduced by the applicants in paragraph 17(b) of the reply, that ‘it [was] clear that there was [a slightly decreased response to sound] on the female rat at the 2.0% level and higher’. Therefore, even if the study in question was, as the applicants also state, only a preliminary dose-range finding study, it enabled an effect to be observed at 2% concentration, which makes it possible that that effect already manifests itself between the NOAEC of 0.5% and that concentration of 2%. The applicants also refer, in paragraph 17(d) of the reply, to ‘internal letters from Haskell’ which, ‘as regards the results in the main study’, indicate that the group of rats which received a dose at 2% concentration (rats from Group IV) ‘exhibited decreased response to sound’. A definite lack of usefulness of the tests carried out at a concentration of up to 1.65% cannot therefore be inferred from the comments on those studies.

83      In the third place, as regards the arguments relating to specific uses in humans and relating to the assessment and management of risks in that regard, which seek to demonstrate that, where the substance is used in its industrial, professional or domestic applications, it cannot lead to narcotic effects in humans, the Commission, the Kingdom of Denmark, the Kingdom of the Netherlands, the Kingdom of Sweden and ECHA rightly state that registration of a substance is not intended solely to ensure non-hazardous use in its normal applications, but also to know about the substance and its effects on living organisms and on the environment as such, in other words, to know its intrinsic characteristics, which may require tests recreating conditions that differ from those in its normal applications. In that regard, Annexes VII to X specifically define the information to be provided in order for the intrinsic properties of a substance to be known. Therefore, the fact that a substance is not hazardous for humans in its normal applications, in particular the fact that there are no narcotic effects in humans during such uses, even if established, cannot remove the need to carry out studies which are required under Annexes VII to X, unless an adaptation is possible under Annex XI. ECHA also rightly states that applications of a substance may evolve over time, whereas its intrinsic properties remain the same. In addition, in the present case, the applicants’ demonstration is based in particular on the premise of a NOAEC of 2.5%, which is uncertain for reproductive toxicity studies, as noted in paragraph 81 above.

84      In the fourth place, as regards the criticism of the Commission’s justification that sluggishness was observed in a test on rats at a concentration of 1%, the applicants claim that mere sluggishness does not constitute a harmful effect. However, it must be pointed out that the Commission does not claim that mere sluggishness constitutes a harmful effect, but simply that, if a sluggish effect is observed at a concentration of 1%, it is possible that harmful effects, narcotic or otherwise, may be observed beyond, between that concentration and the concentration of 1.65%. That assessment does not appear to be vitiated by a manifest error and, consequently, the applicants’ arguments concerning the stage at which narcosis is to be regarded as harmful, which is not the case with sluggishness, also do not demonstrate the definite lack of usefulness of the tests requested in the contested decision if they are carried out at a concentration of up to 1.65%.

85      In the fifth place, as regards the applicants’ scientific challenge to the possibility that dimethyl ether might have toxicological effects at a concentration below 1%, to which the Commission replied in the rejoinder, it must be stated that, even if that assertion by the applicants were correct, that would not render implausible the emergence of harmful effects before the maximum concentration of 1.65%. Furthermore, dimethyl ether is not a noble gas. Even if it were similar to a noble gas, as the applicants maintain, that fact alone would not exempt the registrants from having to provide the information required for the registration of dimethyl ether, since the latter does not benefit from the exemption provided for in respect of noble gases in Article 2(7) of the REACH Regulation.

86      As regards the applicants’ arguments put forward in the reply specifically aimed at demonstrating the futility of the preliminary dose-range finding study (see paragraph 58 above), they do not support the third plea currently examined, which alleges that the requested tests do not produce any relevant information on dimethyl ether. That preliminary study seeks only to determine the modalities and scope of the extended one-generation reproductive toxicity study, which has already been established, in the light of the arguments examined in paragraphs 81 to 85 above, as not being manifestly incapable of providing relevant information on dimethyl ether with tests carried out at a concentration of up to 1.65%.

87      In any event, even if those arguments had to be examined in the context of this plea, it must be pointed out that the applicants cast doubt on the need for that preliminary study only in so far as that study seeks to detect narcotic effects which, in the applicants’ view, are equally observable on cohorts 1A and 1B in the context of the extended study itself. According to Article 3 of the operative part of the contested decision, read in the light of recital 16 of that decision, the sole aim of the particular observation of narcotic effects is to assess whether the inclusion of cohorts 2A and 2B in the extended study, in order to assess specifically the developmental neurotoxicity of dimethyl ether, is necessary, whereas the dose-range finding study also, and above all, seeks to ascertain which concentrations to use in the extended study, whether the latter is confined to tests on cohorts 1A and 1B or also includes tests on cohorts 2A and 2B. In view of the uncertainties as to the concentration at which harmful effects of dimethyl ether may be observed in an extended reproductive toxicity study, the preliminary dose-range finding study requested in the contested decision cannot therefore be regarded as manifestly of no use on the basis of the arguments referred to above. The absence of specific instructions in the contested decision on how to conduct and interpret that preliminary study, in particular as regards the observation of narcotic effects, cannot call into question that conclusion, since specialist testing laboratories normally have the scientific competence to organise such studies and interpret their results, as the Commission states, in essence, in the rejoinder.

88      In the light of the foregoing, even though there may be uncertainties as to the concentration at which harmful effects of dimethyl ether could be observed in reproductive toxicity studies and even though there are scientific debates in that regard, it does not appear to be a manifest error of assessment to have requested the tests listed in the contested decision, particularly as the Commission requested that the extended one-generation reproductive toxicity study be preceded by a preliminary dose-range finding study. In that context, it cannot be ruled out that the tests requested in the contested decision demonstrate toxicity below the concentration level of 1.65% which, in OECD GD 39, it is recommended not to exceed. Even if there were no toxicity below that level, those tests would not be futile and would enable the debates referred to above to be settled in part.

89      It follows from the foregoing that the applicants have not demonstrated, contrary to what they claim under their third plea, that the Commission required tests that would not provide any relevant information on dimethyl ether and that the Commission therefore made a manifest error of assessment. The concomitant infringement of Article 1(1), Article 10(a), Article 12(1), Article 13(4) and Article 41 of the REACH Regulation and Note 1 of Annex VI, read in the light of recitals 17 and 26 of that regulation, from which, according to the applicants, it follows that, in carrying out a compliance check of registrations, ECHA or the Commission may request only relevant and necessary information and must promote alternative methods to animal testing for the assessment of hazards associated with substances, is therefore not established either. The third plea must therefore be rejected.

 The fourth plea in law, alleging that the Commission made a manifest error of assessment and infringed column 2 of Section 8.7.3. of Annex X by requiring the addition of cohorts 2A and 2B to the extended one-generation reproductive toxicity study

90      In the fourth plea, the applicants put forward arguments which are divided, in essence, into two parts, one alleging that the Commission erred in law by distorting the scope of the words ‘particular concerns’ in the second paragraph of column 2 of Section 8.7.3. of Annex X, and the other alleging that it made a manifest error of assessment in finding that dimethyl ether presents ‘particular concerns’ associated with neurotoxicity on the basis of the first and third indents of that provision.

91      As a preliminary point, it should be noted that column 1 of Section 8.7.3. of Annex X requires, by way of standard information, an extended one-generation reproductive toxicity study covering cohorts 1A and 1B, carried out on a single species. According to the second paragraph of column 2 of that section, the inclusion of cohorts 2A and 2B in an extended one-generation reproductive toxicity study in order to assess developmental neurotoxicity may be required by ECHA, or where appropriate by the Commission, in the event of particular concerns on (developmental) neurotoxicity justified by any of the following:

–        existing information on the substance itself derived from relevant available in vivo or non-animal approaches (e.g. abnormalities of the [central nervous system], evidence of adverse effects on the nervous or immune system in studies on adult animals or animals exposed prenatally), or

–        specific mechanisms/modes of action of the substance with an association to (developmental) neurotoxicity and/or (developmental) immunotoxicity (e.g. cholinesterase inhibition or relevant changes in thyroidal hormone levels associated to adverse effects), or

–        existing information on effects caused by substances structurally analogous to the substance being studied, suggesting such effects or mechanisms/modes of action.

92      It is apparent from recital 14 and the first paragraph of recital 15 of the contested decision that the Commission re-evaluated the existing in-vivo data as regards the narcotic effects caused by dimethyl ether. On the basis of those data, the Commission stated that the transient narcotic effects of dimethyl ether, namely ataxia, anaesthesia, short jerky respirations, head bobbing, paw waiving, and roving eyeballs had been observed at a dose of 8.4%. Next, the Commission stated that those effects, produced after a single exposure, corresponded to some of the criteria listed for the category ‘STOT SE-cat.3’ in Section 3.8.2.2.2 of Annex I to Regulation No 1272/2008. Finally, the Commission stated that, in the two-week ‘whole-body’ inhalation study, already referred to in paragraphs 58, 78 and 84 above, narcotic effects such as sluggishness had been detected at a concentration of approximately 1% and coordination problems and a lack of response to noise at a concentration of approximately 5%.

93      It is also apparent from the second paragraph of recital 15 of the contested decision that the Commission took into consideration the effects of a substance whose structure is similar to that of dimethyl ether, namely diethyl ether. The Commission noted, in that regard, that slight anaesthetic effects of diethyl ether had been observed at two doses in a 90-day oral study and that diethyl ether had narcotic effects in humans which also corresponded to some of the criteria listed for the category ‘STOT SE-cat.3’ in Section 3.8.2.2.2 of Annex I to Regulation No 1272/2008.

94      On the basis of those observations, the Commission concluded that, although no clear signs of neurotoxicity had already been observed, either for dimethyl ether or for diethyl ether, the narcotic effects observed through exposure to dimethyl ether or diethyl ether were sufficient evidence of harmful effects of dimethyl ether on the nervous system and, therefore, that the conditions laid down in the first and third indents of the second paragraph of column 2 of Section 8.7.3. of Annex X were met.

 The first part, alleging that the Commission erred in law by distorting the scope of the words ‘particular concerns’ in the second paragraph of column 2 of Section 8.7.3. of Annex X

95      The applicants submit that, in so far as the second paragraph of column 2 of Section 8.7.3 of Annex X does not define the concept of ‘particular concerns’ associated with developmental neurotoxicity, that concept should be defined in the light of the other provisions of the REACH Regulation and point R.7.6.2 of Chapter R.7a of the ECHA guidance entitled ‘Guidance on Information Requirements and Chemical Safety Assessment’ (‘the ECHA Guidance’). On those bases, they are ‘strong’ concerns showing a certain level of severity revealed by serious and severe neurotoxicity effects.

96      The Commission, supported by the Kingdom of Denmark, the Kingdom of the Netherlands, the Kingdom of Sweden and ECHA, disputes the applicants’ arguments.

97      First of all, it must be stated that, contrary to what the Commission submits, in essence, in the defence, the three indents in the second paragraph of column 2 of Section 8.7.3. of Annex X do not themselves define in general terms what is to be understood by ‘particular concerns’, but indicate the nature of the information that may raise such concerns, by providing examples of cases that should raise such concerns, as follows from the citation in paragraph 91 above. Nor does the REACH Regulation contain, in any other of its provisions, any definition of what such concerns are.

98      In that regard, it should be noted, as the applicants state, that Article 77(2)(g) and (k) of the REACH Regulation confers on the Secretariat of ECHA the task, inter alia, of ‘providing technical and scientific guidance and tools where appropriate for the operation of [that regulation] in particular … technical and scientific guidance for the application of Article 7 by producers and importers of articles’ and of ‘preparing explanatory information on [that regulation] for other stakeholders’. Given the legislature’s intention, even if it does not constitute a ‘legal opinion’ for economic operators and is not binding on either ECHA or the Commission, a document such as the ECHA Guidance, relied on by the applicants, may be one of the factors to be taken into consideration in interpreting that regulation (see, to that effect, judgment of 10 September 2015, FCD and FMB, C‑106/14, EU:C:2015:576, paragraph 28 and the case-law cited).

99      Next, it may be noted that, according to the ECHA Guidance, in order to constitute a particular concern within the meaning of the second paragraph of column 2 of Section 8.7.3. of Annex X, the concern must be specific to developmental neurotoxicity and reach a certain level of severity. As regards that second aspect, reference is made to ‘serious or severe effects’ and it is concluded that ‘particular concerns’ arise where there is sufficient evidence from which it can reasonably be inferred that the substance in question could be a developmental neurotoxicant.

100    Footnote 153 of the ECHA Guidance, linked to those considerations and relied on by the Kingdom of Sweden, states the following:

‘A serious or severe effect is an effect which has regulatory consequences, i.e. leads to [no observed adverse effect level] values and/or contributes to hazard classification. Thus, a particular concern is an expectation that the substance has (developmental) neurotoxic properties contributing to the regulatory decision making. This also means that [the effects produced by the substance] are not secondary to other systemic toxicity.’

101    It should also be noted that the ECHA Guidance sets out the distinction between the nature of ‘particular concerns’, the existence of which justifies the inclusion of cohorts 2A and 2B in the extended one-generation reproductive toxicity study, and examples of effects caused by the substance that are capable of justifying the existence of those particular concerns. That distinction is already made in the second paragraph of column 2 of Section 8.7.3. of Annex X, as follows from paragraph 91 above. In other words, the ECHA Guidance interprets the REACH Regulation as leaving it to the regulator to determine whether, in a specific case and on the basis of the available information, particular concerns exist.

102    That is also the interpretation set out in paragraph 2 of Section B.56 of the regulation on test methods, according to which the decision to widen the extended one-generation reproductive toxicity study to include cohorts 2A and 2B ‘should reflect existing knowledge for the chemical being evaluated, as well as the needs of various regulatory authorities’.

103    Having noted those interpretations, it appears that, despite the lack of a precise definition as to what is a particular concern on developmental neurotoxicity, within the meaning of the second paragraph of column 2 of Section 8.7.3. of Annex X, it follows from the very wording used in that provision (see paragraph 91 above), 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 neurotoxic 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. Where there is such information, the purpose of the extended one-generation reproductive toxicity study including cohorts 2A and 2B is then to clarify, confirm or disprove the developmental neurotoxic effects of the substance.

104    Therefore, as the Kingdom of the Netherlands submits, in a specific case, it is for the competent authority, in the absence of a spontaneous initiative to that effect by the registrants, to consider, in the light of the existing data and on the basis of the principles referred to in paragraph 103 above, whether concerns on developmental neurotoxicity exist.

105    Thus, in order to reach the conclusion that dimethyl ether gives rise to ‘particular concerns’, the Commission was not required, contrary to what the applicants claim (see paragraph 95 above), to put forward evidence, at that stage, that dimethyl ether has serious and severe neurotoxicity effects. It is sufficient that one of the elements mentioned in the first to third indents of the second paragraph of column 2 of Section 8.7.3. of Annex X exists and gives reasonable grounds for fearing that there are sufficiently serious or severe harmful effects that justify the possibility of developmental neurotoxicity.

106    In view of the foregoing, the first part of the fourth plea must be rejected as unfounded and it must next be examined whether the Commission made a manifest error of assessment in considering that the existing information relating to dimethyl ether or to a substance with a similar structure to dimethyl ether, namely diethyl ether, justified ‘particular concerns’ on developmental neurotoxicity.

 The second part, alleging that the Commission made a manifest error of assessment in considering that dimethyl ether justifies ‘particular concerns’ on developmental neurotoxicity

107    The applicants submit that the Commission made a manifest error of assessment in considering that the existing information on dimethyl ether and its structural analogue, namely diethyl ether, was such as to justify ‘particular concerns’ on the basis of the first and third indents of the second paragraph of column 2 of Section 8.7.3. of Annex X (provisions reproduced in paragraph 91 above).

108    First, according to the applicants, dimethyl ether does not have serious and severe neurotoxicity effects, which the Commission acknowledged by accepting, in recital 15 of the contested decision, that, for dimethyl ether, ‘no clear signs of neurotoxicity have been observed’.

109    Secondly, the applicants state that, in so far as there is no evidence that the narcotic effects of dimethyl ether are more than transient, a number of members of the Member State Committee considered that that substance could not have ‘neurotoxic effects’ in developing organisms, since those effects were necessarily long-term, and, consequently, that the existing information was not such as to raise particular concerns leading to the need to add cohorts 2A and 2B to the extended one-generation reproductive toxicity study.

110    Thirdly, at the stage of the reply, the applicants submit that it is apparent from the report prepared for ECHA in 2017 by a toxicologist, entitled ‘Scientific review on the link between the narcotic effects of solvents and (developmental) neurotoxicity’, that, as an ‘organic solvent’, by its very nature, dimethyl ether cannot be regarded as a substance that produces, even at high concentrations or repeated doses, narcotic effects warranting particular concerns on developmental neurotoxicity.

111    Fourthly, with regard specifically to the application of the third indent of the second paragraph of column 2 of Section 8.7.3. of Annex X, the applicants submit that the Commission has also failed to demonstrate that such concerns are justified by the information on the effects caused by the structural analogue diethyl ether. After assessing the information concerning those effects, the Commission concludes that that substance has only slight anaesthetic effects.

112    The Commission, supported by the Kingdom of Denmark, the Kingdom of the Netherlands, the Kingdom of Sweden and ECHA, disputes the applicants’ arguments.

113    As a preliminary point, it must be noted that the competent authority enjoys, in the circumstances set out in paragraph 103 above, a broad discretion in taking the decision to widen an extended one-generation reproductive toxicity study so as to include cohorts 2A and 2B. The exercise of that discretion is not excluded from judicial review. However, as already stated in paragraph 73 above, in the context of that review, the Court must confine itself to verifying whether there has been a manifest error in the Commission’s assessment of the facts. In order to establish that the Commission made such an error in the assessment of highly complex scientific and technical facts, the evidence adduced by the applicant must be sufficient to render the factual assessment made in the contested decision implausible. Subject to that review of plausibility, it is not the Court’s role to substitute its assessment of the facts for that made by the author of the decision.

114    As regards the applicants’ first argument, according to which the Commission contradicts itself by requesting the inclusion of cohorts 2A and 2B in the extended one-generation reproductive toxicity study when it itself acknowledges that dimethyl ether has not revealed any clear signs of neurotoxicity, it is apparent from the Court’s findings on the first part of the present plea (see paragraph 105 above) that it is sufficient that the data available to the Commission give reasonable grounds for fearing the existence of harmful effects with respect to developmental neurotoxicity. Thus, even though no ‘clear’ signs of neurotoxicity have yet been found, the Commission has not been able to rule out the existence of harmful effects of dimethyl ether associated with developmental neurotoxicity, given the existing data indicating the narcotic effects of dimethyl ether and its structural analogue diethyl ether on the behaviour of the animals tested, such as the data set out in paragraphs 92 and 93 above.

115    As regards the applicants’ second argument, according to which the absence of ‘particular concerns’ is demonstrated by the different approaches adopted at the meeting of the Member State Committee concerning the nature of the narcotic effects of dimethyl ether which justified whether or not to widen the extended one-generation reproductive toxicity study so as to include cohorts 2A and 2B, in particular whether or not those effects were transient in nature, the Court considers that the Commission was entitled to consider, without making a manifest error of assessment, that those differences did not rule out the existence of ‘particular concerns’ which specifically could be confirmed or invalidated by additional tests. It may be observed, as the Commission submits, that the contested decision ultimately received a favourable opinion from the committee established by Article 133 of the REACH Regulation, the 27 members of which expressed their views on the inclusion of cohorts 2A and 2B in the extended one-generation reproductive toxicity study.

116    The applicants’ third argument that, as an organic solvent, dimethyl ether cannot warrant ‘particular concerns’ on the basis of the report referred to in paragraph 110 above, must also be rejected. That report does not rule out a link between the narcotic effects of organic solvents, such as dimethyl ether, and neurotoxicity or developmental toxicity, contrary to what the applicants submit. That report states that ‘investigations for developmental neurotoxicity are often lacking’ (page ix, summary). The report states that, for a long time, ‘little attention was given to effects of organic solvents [on] the developing nervous system’ and that it is now clear ‘that the mature and developing nervous system differ to a large exten[t] thereby making the developing nervous system more prone to the hazardous effects of chemicals including organic solvents’ (page 7 of the report), such as dimethyl ether. Lastly, the report states that ‘it seems very likely that the developing nervous system is more prone to the adverse effects of environmental chemicals than the adult [nervous system]’ (page 47 of the report).

117    As regards the applicants’ fourth argument, concerning the effects produced by the structural analogue diethyl ether, the Commission states in the second paragraph of recital 15 of the contested decision, without being contradicted by the applicants, that that substance causes ‘slight anaesthetic effects’ during animal testing, as well as narcotic effects in humans, classified under Regulation No 1272/2008, referred to in paragraph 3 above, in the category ‘STOT SE-cat.3’.

118    Regulation No 1272/2008 includes, in its classification of the health hazards of the substances to which it applies, a class corresponding to ‘specific target organ toxicity – single exposure’ (‘STOT-SE’). Within that class, three hazard categories, designated 1 to 3, are defined, ranging from the most significant to the least significant. In accordance with Table 3.8.1 in Annex I to Regulation No 1272/2008, substances in category 3 have transient effects on certain organs. It is stated that ‘this category only includes narcotic effects and respiratory tract irritation’, that ‘these are target organ effects for which a substance does not meet the criteria to be classified in Categories 1 or 2 indicated above’, that ‘these are effects which adversely alter human function for a short duration after exposure and from which humans may recover in a reasonable period without leaving significant alteration of structure or function’ and that ‘the substances are classified specifically for these effects as laid down in 3.8.2.2.’. In that regard, the narcotic effects in human beings listed in point 3.8.2.2.2 of Annex I to Regulation No 1272/2008 include drowsiness, narcosis, reduced alertness, loss of reflexes, lack of coordination or vertigo. As regards the effects in animals, lethargy, lack of coordination, loss of righting reflex, and ataxia are mentioned.

119    According to the Court, since such effects, even if they are transient and fall within the category of the least hazardous symptoms, in accordance with Articles 3 and 4 of Regulation No 1272/2008, involve regarding the substance in question as hazardous and classifying and labelling it as having ‘specific target organ toxicity’, the applicants are not entitled to claim that the narcotic effects of diethyl ether identified by the Commission are negligible and not plausible enough to justify particular concerns on the developmental neurotoxicity of dimethyl ether under the third indent of the second paragraph of column 2 of Section 8.7.3. of Annex X. It may, in particular, be noted in that regard that Article 3 of Regulation No 1272/2008 provides that ‘a substance … fulfilling the criteria relating to … hazards …, laid down in … Annex I [to that regulation] is hazardous and shall be classified in relation to the respective hazard classes provided for in that Annex’.

120    In the light of the foregoing assessments, it has not been established that the Commission made a manifest error of assessment in considering that the observed effects were sufficient to enable it to conclude that there were particular concerns on developmental neurotoxicity, justifying the addition of cohorts 2A and 2B to the extended one-generation reproductive toxicity study which, moreover, is dependent on the results of a preliminary dose-range finding study.

121    Consequently, the second part of the fourth plea must be rejected and, therefore, the fourth plea must be rejected in its entirety.

 The fifth plea in law, alleging that the Commission infringed column 1 of Section 8.7.3. of Annex X to the REACH Regulation, and Article 25 of that regulation, by requiring that the extended one-generation reproductive toxicity study be preceded by a preliminary dose-range finding study

122    In their fifth plea, the applicants submit, in essence, that the Commission made several errors of law and a manifest error of assessment in requiring that the extended one-generation reproductive toxicity study be preceded by a preliminary dose-range finding study.

123    In the first place, the applicants submit that, since neither Section 8.7.3. of Annex X, nor Annex X in general, requires an extended one-generation reproductive toxicity study to be preceded by a preliminary dose-range finding study, the request to carry out such a study has no legal basis.

124    In the second place, the applicants submit that, since a pre-natal developmental toxicity study is available, the Commission was not entitled to require a preliminary dose-range finding study. The fourth indent of column 2 of Section 8.7.1. of Annex VIII states that a preliminary dose-range finding study is not required where a pre-natal developmental toxicity study is available.

125    In the third place, the applicants claim that the obligation to carry out a preliminary dose-range finding study runs counter to the principle of limiting and replacing animal testing, laid down in Article 25 of the REACH Regulation.

126    In the fourth place, the applicants submit, in essence, that the Commission made a manifest error of assessment in taking the view that, on the basis of the available data, it was not possible to determine from the outset a concentration to be tested in the extended one-generation reproductive toxicity study. In particular, they submit that the reaction of neuronal cells in young animals to anaesthetic treatment does not, in principle, differ from the reaction of neuronal cells in adult animals to that treatment and that, consequently, the already determined NOAEC should be sufficient to determine which concentrations to use.

127    The Commission, supported by the Kingdom of Denmark, the Kingdom of the Netherlands, the Kingdom of Sweden and ECHA, disputes the applicants’ arguments.

128    It is apparent from recital 16 and the second paragraph of Article 3 of the contested decision that, given the limited available evidence on toxic dose levels for reproduction and the particular physicochemical properties of dimethyl ether, before the extended one-generation reproductive toxicity study is started, a preliminary dose-range finding study must be carried out, in which animals will be carefully monitored in order to detect any narcotic-type effects. It is only if such effects are observed at any of the concentrations chosen for the extended one-generation reproductive toxicity study that cohorts 2A and 2B (developmental neurotoxicity) would have to be included in that study.

129    As regards, in the first place, the alleged lack of a legal basis for requesting that a preliminary dose-range finding study be carried out, it is true that Annex X to the REACH Regulation does not expressly provide that a request to carry out an extended one-generation reproductive toxicity study may include a request to conduct a preliminary dose-range finding study.

130    However, as the Commission, the Kingdom of Sweden and ECHA submit, it must be observed that the preliminary dose-range finding study is not an independent study, but a preliminary study, the purpose of which is to determine the appropriate dose for the main study and, in the present case, to determine whether part of that main study may be omitted. Therefore, the possibility of requesting that such a study take place prior to an extended one-generation reproductive toxicity study stems from Section 8.7.3. of Annex X itself.

131    It may also be noted that, in the contested decision, the carrying out of that preliminary study is requested by reference to OECD Guideline 421 or any ‘similar [study]’. Paragraph 5 of that guideline states that ‘[it] can be used to provide initial information on possible effects on reproduction and/or development, either at an early stage of assessing the toxicological properties of chemicals, or on chemicals of concern’ and that ‘it can also be used as part of a set of initial screening tests for existing chemicals for which little or no toxicological information is available, [or] as a dose range finding study for more extensive reproduction/developmental studies …’. In paragraph 4 of that guideline, it is also stated that it does not replace Guidelines 414 and 443, which concern, respectively, the pre-natal developmental toxicity study and the extended one-generation reproductive toxicity study. OECD Guideline 421 therefore confirms that it may be appropriate to request that a dose-range finding study be carried out prior to an extended one-generation reproductive toxicity study.

132    It may also be noted that the possibility of requesting such a study to be carried out before the extended one-generation reproductive toxicity study is mentioned in the ECHA Guidance, as the Commission states in the rejoinder. It is apparent from point R.7.6.2.3.2., entitled ‘Procedure for adaptations and testing approaches’, on page 481 of Chapter R.7a of that guide, that, where the extended one-generation reproductive toxicity study is envisaged, it is necessary to determine, inter alia, the concentration level of the substance to be tested in the context of that study. The ECHA Guidance provides, in that regard, on page 483, that ‘the dose level selection is assisted by the information from existing studies as well as from specific dose range-finding studies that may need to be conducted’.

133    Section 8.7.3. of Annex X must therefore be interpreted as authorising the Commission to request a preliminary dose-range finding study to be carried out prior to an extended one-generation reproductive toxicity study and it therefore appears that the Commission did not err in law in that regard.

134    In the second place, as regards the argument that, in Section 8.7.1. of Annex VIII, it is stated that a dose-range finding study is not required if a pre-natal developmental toxicity study is already available, it is appropriate to note the following, already mentioned in paragraph 65 above. In accordance with the ‘Guidance note on fulfilling the requirements of Annexes VI to XI’, which is the introductory part of Annex VI, ‘for the lowest tonnage level, the standard requirements are in Annex VII, and every time a new tonnage level is reached, the requirements of the corresponding Annex have to be added’ and ‘for each registration the precise information requirements will differ, according to tonnage, use and exposure’. In addition, the introductory parts of Annexes VIII, IX and X each indicate that ‘the information required in column 1 of this Annex is additional to that required in column 1 of [the preceding annexes]’. It may be inferred from this that Annexes VII to X are not superfluous as regards column 1, in the sense that column 1 of the annex with the highest figure does not repeat all the elements in column 1 of the preceding annexes. However, if the pieces of information to be provided corresponding to the standard requirements in column 1 of the relevant annexes are added to each other when the level of quantity manufactured or imported per year per manufacturer or per importer reaches the level referred to in a given annex, the possible adaptations mentioned in column 2 of those annexes are not retained from one annex to the next, unless they are repeated (the opposite principle). An adaptation may be envisaged for a certain level of manufacture or importation and no longer be envisaged at a higher level.

135    In other words, since, in view of the quantities declared in the present case, the level referred to in Annex X, namely that of substances manufactured or imported per year per manufacturer or per importer in quantities of 1 000 tonnes or more has been reached, the applicants cannot rely on the adaptation possibilities set out in column 2 of Section 8.7. of Annex VIII, which refers to the level of substances manufactured or imported in quantities of 10 tonnes or more, in order to reject a request made to them under Annex X. The applicants’ argument that the Commission infringed that provision is, therefore, unfounded and the error of law alleged in that regard has not been established.

136    In the third place, as regards the argument that the carrying out of a preliminary dose-range finding study disregards the objective, set out in Article 25(1) of the REACH Regulation, of carrying out tests on vertebrate animals only if there is no other solution, the following factors must be taken into consideration.

137    First, the objective of avoiding animal testing must be applied in the light of the other principles underlying the REACH Regulation, in particular in the light of the precautionary principle. Article 1(3) of the REACH Regulation states that its provisions ‘are underpinned by the precautionary principle’. It has been held that that principle entails that, where there is uncertainty as to the existence or extent of risks to human health, protective measures may be taken without having to wait until the reality and seriousness of those risks become fully apparent (see, to that effect, judgments of 5 May 1998, National Farmers’ Union and Others, C‑157/96, EU:C:1998:191, paragraphs 63 and 64, and of 1 October 2019, Blaise and Others, C‑616/17, EU:C:2019:800, paragraph 43 and the case-law cited). It has also been held that a correct application of the precautionary principle in respect of a substance whose effects are not fully determined presupposes (i) identification of the potentially negative consequences for health of the proposed use of the substance at issue, and (ii) a comprehensive assessment of the risk to health based on the most reliable scientific data available and the most recent results of international research (see by analogy, as regards substances used in plant protection products, judgment of 22 December 2010, Gowan Comércio Internacional e Serviços, C‑77/09, EU:C:2010:803, paragraph 75 and the case-law cited). In the present case, the request for a preliminary dose-range finding study in the context of the carrying out of an extended one-generation reproductive toxicity study has made it possible to reconcile the precautionary principle with the requirement to reduce animal testing. As the Commission explains, since no narcotic effect would be found at concentration levels that are compatible with carrying out hazard-free tests, cohorts 2A and 2B would not be included in the extended one-generation reproductive toxicity study.

138    Secondly, the applicants submit that the carrying out of a dose-range finding study would unnecessarily lead to the death of a large number of animals. However, the preliminary dose-range finding study seeks, inter alia, to determine whether the addition of cohorts 2A and 2B is necessary in the extended one-generation reproductive toxicity study. According to what the applicants themselves state, the first of those studies may result in approximately 550 animals being killed, while the second, carried out in full, leads to the death of approximately 1 500 animals. Consequently, if the preliminary study leads to the conclusion that the extended study is not necessary on cohorts 2A and 2B, the latter study will use only approximately 750 animals, the same number being spared, which will have led to at least 200 (750 – 550) animals being preserved. If the preliminary study leads to the opposite conclusion, it is true that the addition of the two studies will lead to the death of a total of up to approximately 2 050 (550 + 1 500) animals, but that will have been justified in the light of the preliminary study, which will not only have demonstrated the need to undertake an in-depth study in order properly to assess the toxic effects of dimethyl ether for reproduction, in particular its neurotoxic effects, but also will have determined the appropriate concentrations for carrying out that in-depth study in a thorough and meaningful manner, probably by reducing the suffering caused to the animals used in that study.

139    Moreover, the applicants have not demonstrated how it would have been possible to sacrifice fewer animals in order to remove the Commission’s doubts concerning the developmental toxicity of dimethyl ether. The applicants have the option and even the obligation to propose adaptations, if they are possible, which would better meet the objective of limiting animal testing than the carrying out of the preliminary dose-range finding study, as will be explained in greater detail in paragraphs 144 and 145 below, which, up to now, they have not done with any success, as follows from paragraphs 149 to 153 below. It follows from the foregoing that the Commission did not infringe Article 25 of the REACH Regulation either.

140    In the fourth place, as regards the manifest error of assessment alleged on the basis of the line of argument summarised in paragraph 126 above, the Court has already pointed out, in the context of the examination of the third plea, that, although the existing studies have not ruled out the presence of harmful effects at a concentration below 1.65%, namely the concentration corresponding to the maximum concentration level of dimethyl ether recommended by OECD GD 39 for the conduct of tests, those studies do not provide up-to-date and relevant information enabling the Commission to determine the levels of toxic doses of dimethyl ether for animals in the pre-natal and post-natal period. Moreover, as the Commission and the Kingdom of Sweden submit, the applicants have not adduced any evidence to show that dimethyl ether has no impact on the development of neuronal cells in young animals. Accordingly, the Commission also did not make a manifest error of assessment in requesting, in the context of the extended one-generation reproductive toxicity study, that a preliminary dose-range finding study be carried out.

141    For all of the foregoing reasons, the fifth plea must be rejected in its entirety.

 The sixth plea in law, alleging that the Commission infringed Article 41 of the REACH Regulation and Annex XI to that regulation, on the ground that the contested decision does not allow the applicants to remedy the non-compliance of the registration of dimethyl ether by submitting adaptations of the studies requested in that decision

142    The applicants complain, in essence, that the contested decision obliges them, as well as the other registrants, to have the studies mentioned in that decision carried out by providing the results of those studies (see paragraph 15 above), without allowing them to provide appropriate information from other sources instead. According to the applicants, following a decision such as the contested decision, adopted pursuant to Article 41 of the REACH Regulation, ECHA must examine any information submitted by the addressees of that decision, as indicated in Article 42 of that regulation. Article 13(1) of the REACH Regulation itself states that ‘information on intrinsic properties of substances may be generated by means other than tests, provided that the conditions set out in Annex XI are met’. Column 2 of Section 8.7. in Annexes IX and X also provides for the possibility of adaptation of what is requested exclusively in the contested decision.

143    In the defence, the Commission submits that both ECHA and the Commission itself have already examined the registrants’ adaptation proposals and that those proposals were not considered appropriate. It is therefore normal that, in the contested decision, there is a request for tests provided for in Annex X. That in no way prevents the registrants, in view of the provisions of the REACH Regulation, from responding to the contested decision by way of information resulting from adaptations of the tests requested in that decision, provided that they take due account of the objections expressed by ECHA and the Commission regarding their earlier adaptation proposals. In the rejoinder, the Commission adds that it was not required to await the results of studies on the structural analogue diethyl ether, which might provide appropriate information on the basis of sources other than the studies requested in the contested decision, in order to adopt that decision.

144    As has already been held, the relevant general provisions of the REACH Regulation and the objective of limiting animal testing set out in those general provisions mean that a registrant whom ECHA has requested to supplement its registration dossier on the basis of a study involving animal testing has, in so far as possible from a scientific and technical perspective, the option and even the obligation to respond to that request by providing appropriate information in the light of the grounds on which that request was based, but from sources other than that study. It has also been held that, in such a situation, ECHA is under a corresponding obligation to check the compliance of such alternative information with the applicable requirements and, more specifically, to determine whether it is to be classified as adaptations in accordance with the rules laid down in the relevant annexes to the REACH Regulation (see, to that effect, judgment of 21 January 2021, Germany v Esso Raffinage, C‑471/18 P, EU:C:2021:48, paragraphs 132 to 136).

145    There is no reason why a different solution should be used where, as in the present case, the decision requesting the registrant to supplement its registration dossier on the basis of a study involving animal testing is adopted, in the context of the procedure laid down in Article 51 of the REACH Regulation on the adoption of decisions under dossier evaluation, not by ECHA, but by the Commission because of the lack of unanimity in the Member State Committee on the draft ECHA decision.

146    In spite of the wording used in the imperative in the contested decision for the purposes of the carrying out of the studies referred to in the operative part of that decision, the latter cannot therefore be interpreted, in its regulatory context well known to the applicants, as prohibiting them and the other registrants from responding to that decision by proposing in the technical dossier, in accordance with the relevant general provisions of the REACH Regulation and its objective of limiting animal testing, appropriate information in the light of the grounds which justified the requests for studies on animals made in that decision, but from sources other than those studies. It must, however, be stated that those adaptations of the tests requested in the contested decision must not be manifestly unreasonable in the light of the possibilities for adaptation provided for in the REACH Regulation, in particular in Annex XI, and in view of the exchanges which have already taken place between the registrants, ECHA and the Commission. Otherwise, ECHA could simply, in order to prevent the procedure from becoming unjustifiably prolonged, once more find that the registration is not compliant, but without having to use the procedure laid down in Article 42(1) of the REACH Regulation, which itself refers in that regard to Article 41 of that regulation (see, to that effect, judgment of 8 May 2018, Esso Raffinage v ECHA, T‑283/15, EU:T:2018:263, paragraphs 62 and 112).

147    It should also be noted that the applicants themselves interpreted the contested decision as permitting them to respond to it by providing information from sources other than the studies requested in that decision, since they stated in the reply that they were awaiting the results of the pre-natal developmental toxicity study carried out on rabbits in respect of the structural analogue diethyl ether, in order to assess whether those results could allow an adaptation of the request for the same study that was made in the contested decision in respect of dimethyl ether.

148    It follows from the foregoing that, contrary to what the applicants submit, the contested decision does not prohibit them from proposing adaptations of the studies requested in that decision. The sixth plea must therefore be rejected.

 The seventh plea in law, alleging that the Commission infringed Article 41 of the REACH Regulation and Annex XI to that regulation on the ground that, in the contested decision, the Commission prematurely rejected any adaptation of the studies requested in that decision

149    According to the applicants, although Article 41(1) of the REACH Regulation invites ECHA and, where applicable, as in the present case, the Commission, to verify that the adaptations submitted in the registrants’ technical dossiers comply with the rules laid down, the Commission took a position too early in the contested decision on any adaptation of the pre-natal developmental toxicity study requested under Section 8.7.2. of Annex X on a second species, namely rabbits. In that decision, the Commission rejected the possibility of adaptation on the basis of the expected results of the pre-natal developmental toxicity study carried out on rabbits in respect of the structural analogue diethyl ether, mainly because of differences in metabolism between that substance and dimethyl ether. The applicants and the other registrants did not include such an adaptation proposal in the technical dossier before the contested decision was adopted. Therefore, ECHA’s draft decisions drawn up in the present case did not contain any assessment in that regard on which the applicants could have commented before the contested decision was adopted. Their procedural rights were thus infringed.

150    The Commission replies that, neither when the application for registration was filed, nor during the years that followed, did the registrants provide the study summaries requested in the contested decision or any adaptation proposals in order to replace those studies. That is why the draft ECHA decision contained a request for additional studies under Annex X. It was in response to that draft, and then later in the procedure, that the applicants claimed that adaptations of those studies were possible. In particular, they argued before the Commission that an adaptation was possible on the basis of the ongoing pre-natal developmental toxicity study carried out on rabbits in respect of the structural analogue diethyl ether. It is therefore normal that the Commission, in the contested decision, adopted a position in that regard, which did not prevent the applicants and the other registrants, when implementing that decision, from proposing such an adaptation and from arguing in favour of it on the basis of renewed arguments.

151    In the light of the documents in the file, the Commission’s arguments are based on accurate factual evidence. Thus, the draft ECHA decision (referred to in paragraph 6 above) indicates that the technical dossier submitted by the applicants does not contain an adaptation as provided for in column 2 of Section 8.7. of Annex X or in the general rules of adaptation set out in Annex XI, which the applicants themselves explain. It was only in response to that draft, and subsequently in the procedure, that the applicants argued that adaptations of the pre-natal developmental toxicity study carried out on a second species were possible, as is demonstrated by the revised draft ECHA decision (referred to in paragraph 8 above), in which the applicants’ arguments for the application of the adaptations provided for in column 2 of Section 8.7. of Annex X are rejected on the ground that the conditions for those adaptations were not met, and, as is also apparent from the contested decision in which those arguments are again rejected on a similar ground (recitals 6 to 10 of the contested decision), as are arguments relating to an adaptation possibility provided for in point 1.5 of Annex XI, based on the extrapolation of the results of the ongoing pre-natal developmental toxicity study carried out on rabbits for the structural analogue diethyl ether, mainly on the ground of the differences in metabolism between that substance and dimethyl ether (recital 13 of the contested decision).

152    It is apparent, first, that the position taken in the contested decision concerning a possible adaptation of the pre-natal developmental toxicity study carried out on rabbits, based on the similar ongoing study concerning the structural analogue diethyl ether, responded to a need to state reasons in the light of the arguments put forward by the applicants and, secondly, in view of what is set out in paragraphs 144 to 146 above, that such a position does not lead to the advance rejection of any adaptation proposal made in the technical dossier following the contested decision in relation to the studies requested in that decision, in particular any proposal that would use the results of the pre-natal developmental toxicity study carried out on rabbits in respect of dimethyl ether which became available in the meantime, where serious arguments are put forward in support of that proposal in addition to those already put forward prior to the adoption of the contested decision.

153    The seventh plea, based on the argument that the Commission prematurely rejected any adaptation of the studies requested in the contested decision, must therefore be rejected.

 The eighth plea in law, alleging that, by requesting that a pre-natal developmental toxicity study be carried out on rabbits, the Commission made a manifest error of assessment, failed to take into account all the relevant information and infringed column 2 of Section 8.7.2. of Annex IX

154    The applicants state that column 2 of Section 8.7.2. of Annex IX contains an adaptation rule in relation to the standard requirement to have a pre-natal developmental toxicity study carried out on one species. That adaptation rule states that ‘the study shall be initially performed on one species’ and that, ‘a decision on the need to perform a study at this tonnage level or the next on a second species should be based on the outcome of the first test and all other relevant available data’. The applicants infer from that provision that that rule also applies ‘to the next level’ to which Annex X relates, namely the level applicable to substances manufactured or imported in quantities of 1 000 tonnes or more per year per manufacturer or per importer. Therefore, a pre-natal developmental toxicity study on a second species should be conducted, including for the level referred to in Annex X, only if the results of the first test and all other relevant available data show that that second study is necessary.

155    According to the applicants, recital 5 of the contested decision should be interpreted as recognising that the conditions set out in column 2 of Section 8.7.2. of Annex IX are not met, in other words the outcome of the first test and all the other relevant available data do not show that that second study is necessary. Thus, by nevertheless requiring that a pre-natal developmental toxicity study be carried out on a second species, in the present case rabbits, the Commission disregarded those conditions and erred in law. In addition, it made a manifest error of assessment by requiring that study after admitting that the available data did not justify it.

156    The Commission’s states, in response, that the pre-natal developmental toxicity study on a second species was requested in the contested decision under Annex X, not Annex IX. The Commission submits that such a study is mandatory in all cases at the level of manufacture or importation to which Annex X relates, on the basis of column 1 of Section 8.7.2. of that annex.

157    The applicants’ eighth plea in fact consists of two parts, one alleging an error of law on account of the alleged infringement of Annex IX, the other alleging a manifest error of assessment.

 The first part, alleging that the Commission erred in law by infringing Annex IX

158    It must be acknowledged that the relationship between the standard requirements and the possible adaptations in respect of Section 8.7.2., relating to developmental toxicity, set out in Annexes IX and X, does not appear to be clear from the outset. As indicated by the applicants, in Annex IX, column 2 of Section 8.7.2. states ‘or [at] the next [level]’, which, taken in isolation, might suggest that the adaptation provided for in that text also applies to Annex X, particularly as the corresponding column 2 of that annex is blank. Furthermore, the wording of column 1 of Section 8.7.2., which concerns the standard requirement, is essentially identical in Annexes IX and X by stating ‘toxicity study, … one species’ which, also taken in isolation, might suggest that, for the two levels to which those annexes relate, it is necessary to draw information from a study carried out on a single species.

159    However, such an interpretation is inconsistent with the general principle of construction and application of Annexes VII to X. Article 12(1)(e) of the REACH Regulation provides that the technical dossier for the registration of a substance manufactured or imported in quantities of 1 000 tonnes or more per year per manufacturer or per importer must contain the information referred to in Annexes VII and VIII and the testing proposals for the provision of the information referred to in Annexes IX and X. As already stated in paragraph 134 above, it may be inferred from that that those annexes are not superfluous as regards column 1, in that column 1 of the annex with the highest figure does not repeat all the elements contained in column 1 of the preceding annexes which nevertheless remain applicable. Otherwise, Article 12(1)(e) of the REACH Regulation would mention only one annex, namely Annex X, for the content of the technical dossier for a substance manufactured or imported in quantities of 1 000 tonnes or more per year per manufacturer or per importer. However, as also stated in paragraph 134 above, the possible adaptations mentioned in column 2 of those annexes are not retained from one annex to the other, unless they are repeated (opposite principle). An adaptation may be possible for a certain level of manufacture or importation and no longer be so at a higher level. That is confirmed in the ‘Guidance note on fulfilling the requirements of Annexes VI to XI’, which is the introductory part of Annex VI, and in the introductory parts of Annexes VIII, IX and X, each of which states that ‘the information required in column 1 of this Annex is additional to that required in column 1 of [of the preceding annexes]’. Consequently, the applicants’ interpretation that the regime for standard requirements and adaptations regarding Section 8.7.2, relating to pre-natal developmental toxicity studies, is the same at the level covered by Annex IX and at the level covered by Annex X would be more convincing in the light of the general principle referred to above if that section did not contain any wording in Annex X, since everything would be set out in Annex IX. Even acknowledging that an annex sometimes ‘repeats’ what is already stated in the preceding annex (which is the case, for example, for column 2 of Section 8.7. which repeats, in Annex X, what is already stated in Annex IX), the applicants’ interpretation is caught by the fact that, for Section 8.7.2., although the wording of column 1 of Annex IX is repeated in column 1 of Annex X, by contrast column 2 does not in any way repeat, in whole or in part, the wording of column 2 of Annex IX, but is blank.

160    It must be concluded from that, in the light of the general principle of construction and application of Annexes VII to X, that the standard requirements and adaptations under Annex X are independent of those in Annex IX. It may already be inferred that the rules in Annex IX for Section 8.7.2. do not make it possible to determine which are the standard requirements and possible adaptations defined in Annex X for that section, which are applicable for a substance manufactured or imported in quantities of 1 000 tonnes or more per year per manufacturer or per importer. In that regard, the provision according to which ‘the study shall be initially performed on one species’ and ‘a decision on the need to perform a study at this tonnage level or the next on a second species should be based on the outcome of the first test and all other relevant available data, highlighted by the applicants, which appears in column 2 of Annex IX, means only that the requirement for a study on a second species for a substance manufactured or imported in quantities between 100 tonnes and 999 tonnes per year per manufacturer or per importer may, where the conditions for carrying out such a study are met, possibly be postponed until such time as the substance comes under the ‘next level’, namely when the substance is manufactured or imported in quantities of 1 000 tonnes or more per year per manufacturer or per importer.

161    The applicant’s allegation that the Commission erred in law for having infringed the provisions of Annex IX has therefore not been established.

 The second part, alleging that the Commission made a manifest error of assessment by requesting a pre-natal developmental toxicity study on a second species when the conditions set out in column 2 of Section 8.7.2. of Annex IX were not met

162    In order to assess this second part of the eighth plea, it must be pointed out at this stage of the analysis that no adaptation equivalent to that provided for in column 2 of Section 8.7.2. of Annex IX applies to Annex X in respect of that section, for the reasons set out in paragraphs 159 and 160 above, in particular because column 2 of Annex X is blank for that section. In order to determine the scope of the obligations based on Annex X for Section 8.7.2. and to determine, at the same time, the discretion which the Commission had in that regard, it is next necessary to ascertain which standard requirement is requested in column 1 of Annex X.

163    As stated in paragraph 158 above, the wording of column 1 for Section 8.7.2. is essentially identical in Annexes IX and X in that it refers to ‘toxicity study … one species’. As already stated in the same paragraph, those texts, read in isolation, might suggest a mere repetition of the same requirement, that is to say, be interpreted as requiring only that a pre-natal developmental toxicity study be conducted on one species, whether the substance concerned is produced or imported at the levels covered by Annex IX or at the levels covered by Annex X.

164    However, in the light of the general principle of construction and application of Annexes VII to X set out in paragraphs 159 and 160 above, which means that column 1 in one of those annexes is not superfluous in relation to column 1 in the other, there would be no sense in repeating the same standard requirement in column 1. As much as the option of adaptation set out in column 2 may be repeated from one annex to another if that option is valid with regard to different standard requirements set out in different annexes, such a repetition is not conceivable for the same standard requirement already set out in column 1 of a preceding annex for a lower level of production or importation. The requirement in column 1 of Section 8.7.2. of Annex X to have a ‘toxicity study … single species’ carried out must therefore be interpreted as differing from the requirement set out in similar terms in column 1 of Annex IX for the same section, which can mean only one thing: both studies in question must each concern a different species. In other words, the requirement in column 1 of Section 8.7.2 of Annex X to have a ‘toxicity study … one species’ carried out must be understood as referring to a study on a species other than that used for the similar study carried out under Annex IX. As no adaptation is provided for in that regard in Section 8.7.2. of Annex X, as stated in paragraph 162 above, it follows that the pre-natal developmental toxicity study carried out on a second species is mandatory where the substance is produced or imported at the levels referred to in Annex X, unless adaptations are possible under the provisions set out elsewhere.

165    That interpretation has, moreover, already been adopted, as the Commission indicates, in the ECHA Guidance document, which indicates in that regard, in paragraph R.7.6.2.3.2., entitled ‘Procedure for adaptations and testing approaches’, on page 490 of Chapter R.7a, that, ‘at … Annex X level, a prenatal developmental toxicity study … conducted on a second species is a standard information requirement in addition to a prenatal developmental toxicity study in a first species that is required at … Annex IX level’, that ‘availability of information on two species allows a more comprehensive evaluation of prenatal developmental toxicity’ and that ‘the prenatal developmental toxicity study in a second species can be omitted, if, taking into account the outcome of the first test and all other relevant available data, an adaptation pursuant to … Annex X, Section 8.7, Column 2 or pursuant to … Annex XI can be justified’. A table in the same guidance, on page 473, summarising the standard information on the subject, for its part states, as regards Annex X, that the pre-natal developmental toxicity study must be carried out on two species.

166    As the Commission also states, the Board of Appeal of ECHA also adopted that interpretation in the decision of 10 October 2013, Lanxess Deutschland GmbH v ECHA (Case A-004-2012, paragraph 73), stating that ‘as a result of the cumulative nature of the requirements contained in Column 1 to the testing Annexes, [it] considers that, pursuant to Section 8.7.2 of Annex X, registrants are required to perform a developmental toxicity study on a species other than the species used in the performance of the pre-natal developmental toxicity study under Column 1 of Section 8.7.2 of Annex IX, unless one or more of the adaptations in Section 8.7 of Annex X or Annex XI apply’.

167    In the light of the applicants’ arguments, which complain that the Commission made a manifest error of assessment in the contested decision by requesting, under Section 8.7.2., a pre-natal developmental toxicity study on a second species under Annex X, when the conditions set out in column 2 of Annex IX for the same section to justify such a study were not met, it must be observed that it follows from paragraphs 164 to 166 above that the Commission had no discretion not to request such a study on the basis of the rules set out for Section 8.7.2., since that study was automatically required under those rules where the substance concerned was registered for produced or imported quantities of 1 000 tonnes or more per year per manufacturer or per importer and since Annex X therefore applied. The Commission was therefore not able to make a manifest error of assessment in that regard. It is true that, in the contested decision, the Commission examined whether adaptations were possible under other rules, in the present case under column 2 of Section 8.7. of Annex X, and under Annex XI (recitals 6 to 10 and recital 13 of the contested decision) and was able to exercise discretion in that regard, but the applicants’ eighth plea does not relate to that aspect.

168    The eighth plea, alleging that the Commission, by requesting that a pre-natal developmental toxicity study be carried out on rabbits, made a manifest error of assessment, failed to take into account all the relevant information and infringed column 2 of Section 8.7.2. of Annex IX, must therefore be rejected.

 The ninth plea in law, alleging infringement of the principle of proportionality and Article 25 of the REACH Regulation

169    The applicants complain that the Commission infringed the principle of proportionality and Article 25 of the REACH Regulation, first, by requesting studies which, in view of the technical safety constraints, are impossible to carry out and which, even if carried out in accordance with the safety rules, would not produce any relevant information and, secondly, by requesting the preliminary dose-range finding study which, in the light of existing data, would prove to be pointless, lead to the death of a large number of animals and cause the applicants to incur significant expenditure. Furthermore, the applicants submit that it is disproportionate to order them to conduct the abovementioned studies where there is a less onerous measure, namely a potential adaptation by read-across to diethyl ether.

170    The Commission disputes the applicants’ arguments.

171    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; see also, to that effect, judgment of 13 November 1990, Fedesa and Others, C‑331/88, EU:C:1990:391, paragraph 13).

172    However, in the context of the ninth plea, the applicants put forward essentially the same criticisms as those already raised in support of the second, third, fifth and seventh pleas. Those criticisms, in particular in so far as they complained of the lack of usefulness of the studies requested in the contested decision, have already been held to be unjustified, either as having no factual basis or on various legal grounds in the light of the provisions of the REACH Regulation.

173    Thus, in essence, it has been found that the Commission had not requested dangerous studies which were impossible to carry out and that the applicants had found laboratories which had agreed to carry out the tests up to the safety limits (see paragraphs 51, 52 and 54 above). The criticism that the Commission made a manifest error of assessment by requesting studies that will not produce any relevant information has been rejected because the studies requested in the contested decision are either in any event mandatory under Annex X, or likely to reveal reproductive toxicity (see paragraphs 70 and 88 above). The criticism that the preliminary dose-range finding study requested in the contested decision will lead to the unnecessary death of a significant number of laboratory animals has been rejected because the principle of proportionality (reflected in the objective of limiting animal testing) must be applied having regard to the precautionary principle, as laid down in the REACH Regulation, and because, whatever the result of that preliminary study, the sacrifice of the animals used for that study would not have been in vain (see paragraphs 137 and 138 above). The criticism that the contested decision prevents the applicants from submitting an adaptation on the basis of the results of a study on diethyl ether has been rejected on the ground that the contested decision does not prevent them from submitting such an adaptation based on that study by way of renewed arguments according to the results of that study (see paragraph 152 above).

174    No error of law or manifest error of assessment has therefore been identified on the basis of those criticisms in the light of the provisions of the REACH Regulation. The applicants do not challenge the legality of the provisions of the REACH Regulation in relation to the principle of proportionality. It appears, moreover, that the legislature applied that principle, as is apparent, for example, from the concomitant taking into account of the objective of limiting animal testing and the precautionary principle, and from the balancing of the various interests at stake, illustrated in particular in recitals 49 to 55, 61 or 69 of that regulation. In those circumstances, those criticisms cannot in the present case serve as a basis for a finding of an infringement of the principle of proportionality.

175    The same is true of the criticism of having to bear significant expenditure, which, moreover, has not been specified. Although expenditure incurred by the applicants is due to the carrying out, requested in the contested decision, of studies which are justified in the light of the provisions of the REACH Regulation, it cannot be held that the Commission imposed disproportionate burdens in that regard.

176    Nor does the applicants’ line of argument support a finding of an infringement of Article 25 of the REACH Regulation, which establishes the objective of limiting testing on vertebrate animals. That infringement has already been alleged by the applicants and has been rejected by the Court in the context of the examination of the fifth plea concerning the dose-range finding study prior to the extended one-generation reproductive toxicity study. In particular, it is noted in paragraph 137 above that the objective of avoiding tests on vertebrate animals, set out in Article 25 of the REACH Regulation, must be applied in the light of the other principles underlying the REACH Regulation and in particular the precautionary principle. It follows from the examination of the fifth plea that, in the present case, that objective and that principle were reconciled and respected.

177    In the light of the foregoing, the ninth plea must be rejected as unfounded.

178    Consequently, the action must be dismissed in its entirety.

 Costs

179    Under Article 134(1) of the Rules of Procedure of the General Court, the unsuccessful party is to be ordered to pay the costs if they have been applied for in the successful party’s pleadings. As the applicants have been unsuccessful, they must be ordered to pay the costs, including those relating to the proceedings for interim relief, in accordance with the form of order sought by the Commission.

180    According to Article 138(1) of the Rules of Procedure, the Member States and institutions which have intervened in the proceedings are to bear their own costs. Under Article 1(2)(f) of the Rules of Procedure, ‘institutions’ means the institutions of the European Union referred to in Article 13(1) TEU and the bodies, offices or agencies established by the Treaties, or by an act adopted in implementation thereof, which may be parties before the General Court. According to Article 100 of the REACH Regulation, ECHA is an EU body. It follows that the Kingdom of Denmark, the Kingdom of the Netherlands, the Kingdom of Sweden and ECHA must bear their own costs.

On those grounds,

THE GENERAL COURT (Fourth Chamber)

hereby:

1.      Dismisses the action;

2.      Orders the applicants to bear their own costs and to pay those incurred by the European Commission, including the costs relating to the proceedings for interim relief;

3.      Orders the Kingdom of Denmark, the Kingdom of the Netherlands, the Kingdom of Sweden and the European Chemicals Agency (ECHA) to bear their own costs.

Gervasoni Madise Nihoul

Delivered in open court in Luxembourg on 29 March 2023.

E. Coulon S. Papasavvas

Registrar President

Table of contents

Background to the dispute

Forms of order sought

Law

The first plea in law, alleging that the Commission infringed Article 51(7) of the REACH Regulation by adopting the contested decision which covers aspects on which the Member State Committee reached a unanimous agreement

The second plea in law, alleging that the Commission infringed Article 13(3) of the REACH Regulation and made a manifest error of assessment by requesting tests which run counter to the applicable legal requirements and which are not technically feasible

The third plea in law, alleging that the Commission made a manifest error of assessment in requiring tests that do not provide any relevant information on dimethyl ether

The fourth plea in law, alleging that the Commission made a manifest error of assessment and infringed column 2 of Section 8.7.3. of Annex X by requiring the addition of cohorts 2A and 2B to the extended one-generation reproductive toxicity study

The first part, alleging that the Commission erred in law by distorting the scope of the words ‘particular concerns’ in the second paragraph of column 2 of Section 8.7.3. of Annex X

The second part, alleging that the Commission made a manifest error of assessment in considering that dimethyl ether justifies ‘particular concerns’ on developmental neurotoxicity

The fifth plea in law, alleging that the Commission infringed column 1 of Section 8.7.3. of Annex X to the REACH Regulation, and Article 25 of that regulation, by requiring that the extended one-generation reproductive toxicity study be preceded by a preliminary dose-range finding study

The sixth plea in law, alleging that the Commission infringed Article 41 of the REACH Regulation and Annex XI to that regulation, on the ground that the contested decision does not allow the applicants to remedy the non-compliance of the registration of dimethyl ether by submitting adaptations of the studies requested in that decision

The seventh plea in law, alleging that the Commission infringed Article 41 of the REACH Regulation and Annex XI to that regulation on the ground that, in the contested decision, the Commission prematurely rejected any adaptation of the studies requested in that decision

The eighth plea in law, alleging that, by requesting that a pre-natal developmental toxicity study be carried out on rabbits, the Commission made a manifest error of assessment, failed to take into account all the relevant information and infringed column 2 of Section 8.7.2. of Annex IX

The first part, alleging that the Commission erred in law by infringing Annex IX

The second part, alleging that the Commission made a manifest error of assessment by requesting a pre-natal developmental toxicity study on a second species when the conditions set out in column 2 of Section 8.7.2. of Annex IX were not met

The ninth plea in law, alleging infringement of the principle of proportionality and Article 25 of the REACH Regulation

Costs

*      Language of the case: English.

1      The present judgment is the subject of publication in extract form.