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

15 May 2024 (*)

(Medicinal products for human use – Suspension of national marketing authorisations for medicinal products for human use containing the active substance hydroxyethyl starch (HES), solutions for infusion – Action for annulment – Direct concern – Partial inadmissibility – Obligation to state reasons – Error of law – Manifest error of assessment – Precautionary principle – Proportionality – Article 116 of Directive 2001/83/EC)

In Case T‑416/22,

Fresenius Kabi Austria GmbH, established in Graz (Austria), and the other applicants whose names are listed in the annex, (1) represented by W. Rehmann and A. Knierim, lawyers,

applicants,

v

European Commission, represented by M. Escobar Gómez and A. Sipos, acting as Agents,

defendant,

supported by

Ireland, represented by A. Joyce, M. Tierney, M. Browne and D. O’Reilly, acting as Agents, and by P. McCann, Senior Counsel, and E. O’Callaghan, Barrister-at-Law,

and by

European Medicines Agency (EMA), represented by S. Marino, S. Drosos and M. van Egmond, acting as Agents,

interveners,

THE GENERAL COURT (Fifth Chamber, Extended Composition),

composed of J. Svenningsen, President, C. Mac Eochaidh, J. Laitenberger (Rapporteur), J. Martín y Pérez de Nanclares and M. Stancu, Judges,

Registrar: A. Marghelis, Administrator,

having regard to the order of 18 October 2022, Fresenius Kabi Austria and Others v Commission (T‑416/22 R, not published, EU:T:2022:636),

having regard to the written part of the procedure,

further to the hearing on 14 November 2023,

gives the following

Judgment

1        By their action based on Article 263 TFEU, the applicants, Fresenius Kabi Austria GmbH and the other legal persons whose names are listed in the annex, essentially seek annulment of Commission Implementing Decision C(2022) 3591 final of 24 May 2022 concerning, in the framework of Article 107p of Directive 2001/83/EC of the European Parliament and of the Council, the marketing authorisations of medicinal products for human use which contain the active substance ‘hydroxyethyl starch (HES), solutions for infusion’ following an assessment of a post authorisation safety study (‘the contested decision’), in that it requires the Member States concerned to suspend the national marketing authorisations (‘MAs’) for the medicinal products referred to in Annex I thereto (‘the medicinal products concerned’).

I.      Background to the dispute

2        The applicants are part of the worldwide Fresenius Kabi group, which itself belongs to the Fresenius group, which specialises in healthcare and manufactures and distributes, inter alia, medicinal products containing hydroxyethyl starch (HES) as an active substance.

3        The applicants hold MAs for some of the medicinal products concerned.

4        The medicinal products concerned, which have been authorised at national level, are mainly indicated for the treatment of hypovolaemia (low blood volume) caused by acute (sudden) blood loss, when treatment with alternative infusion solutions known as ‘crystalloids’ alone is not considered sufficient.

5        Since 2013, the medicinal products concerned have been the subject of several assessments of their risk-benefit balance, particularly in view of an increased risk of renal impairment and mortality, when using those medicinal products, for patients with sepsis or burns or patients who are critically ill.

6        On 19 December 2013, the European Commission adopted Implementing Decision C(2013) 9793 (final), concerning, in the framework of Articles 31 and 107i of Directive 2001/83/EC of the European Parliament and of the Council, the [MAs] of HES containing medicinal products, solutions for infusion. By that decision, addressed to the Member States, the Commission decided that the Member States concerned should amend those MAs on the basis of the scientific conclusions of the Coordination Group for Mutual Recognition and Decentralised Procedures – Medicinal Products for Human Use of the European Medicines Agency (EMA) (‘the CMDh’). According to that decision, risk minimisation measures (‘the RMMs’) were to be established, notably in the form of new contraindications and new warnings, as well as a reduction in the dosage of those medicinal products.

7        On 17 July 2018, the Commission adopted Implementing Decision C(2018) 4832 final, concerning, in the framework of Article 107i of Directive 2001/83/EC of the European Parliament and of the Council, the [MAs] of medicinal products for human use which contain the active substance [HES], solutions for infusion. By that decision, addressed to the Member States, the Commission decided that the Member States concerned should amend those MAs on the basis of the scientific conclusions of the CMDh by establishing, in particular, additional RMMs on the grounds that the initial RMMs were not sufficiently respected in clinical practice and that the medicinal products concerned continued to be administered to patients with contraindications. Those additional RMMs included limiting the supply of the medicinal products concerned to healthcare professionals who had undergone specific mandatory training, as well as more visible warnings on the packaging.

8        On 10 February 2022, the Pharmacovigilance Risk Assessment Committee (‘the PRAC’), which is the EMA committee responsible for assessing safety issues related to medicinal products for human use, adopted an assessment report in which it concluded by a majority that non-adherence to the product information continued to exist, despite the significant additional RMMs implemented in 2018. The PRAC also concluded that the medicinal products concerned continued to be used in populations with contraindications and therefore at greater risk of serious harm, including mortality. The PRAC indicated that no additional RMMs or combination of RMMs could be identified to sufficiently ensure safe use of the medicinal products concerned. The cases of non-adherence to the product information were, in its view, not only due to a lack of awareness, but appeared to be the result of an intentional choice on the part of prescribers. As a result, the PRAC concluded that the risk-benefit balance of the medicinal products concerned was unfavourable and recommended that the MAs for those products be suspended.

9        On 23 February 2022, the CMDh, acting in accordance with the first subparagraph of Article 107q(2) of Directive 2001/83/EC of the European Parliament and of the Council of 6 November 2001 on the Community code relating to medicinal products for human use (OJ 2001 L 311, p. 67), as amended by Directive 2010/84/EU of the European Parliament and of the Council of 15 December 2010, as regards pharmacovigilance (OJ 2010 L 348, p. 74), considered, by majority decision, that the MAs for the medicinal products concerned should be suspended.

10      On 24 May 2022, the Commission, seised in accordance with the fifth subparagraph of Article 107q(2) of Directive 2001/83, as amended, adopted the contested decision, addressed to the Member States, by which it decided that the Member States concerned should suspend the MAs for the medicinal products concerned on the basis of the scientific conclusions set out in Annex II to that decision.

11      Article 3 of the contested decision provides that Member States may exceptionally, under certain conditions and for a maximum period of 18 months from the date of adoption of that decision, postpone implementation of the suspension.

II.    Form of orders sought

12      The applicants claim that the Court should:

–        annul the contested decision in so far as it orders the Member States to suspend the MAs for the medicinal products referred to in Annex I to that decision;

–        in the alternative, annul the contested decision in so far as it orders the Member States to suspend the MAs for the medicinal products which they market and which are referred to in Annex I to that decision;

–        order the Commission to pay the costs.

13      The Commission contends that the Court should:

–        dismiss the action as inadmissible in so far as it seeks the suspension of MAs for HES-containing products for which the applicants are not the MA holders or, in the alternative, dismiss the action as inadmissible in its entirety;

–        in any event, dismiss the action for annulment as unfounded in its entirety;

–        order the applicants to pay the costs.

14      Like the Commission, Ireland contends that the Court should:

–        dismiss the action as inadmissible in so far as it seeks the suspension of MAs for HES-containing products for which the applicants are not the MA holders or, in the alternative, dismiss the action as inadmissible in its entirety;

–        in any event, dismiss the action for annulment as unfounded in its entirety.

15      The EMA contends that the Court should:

–        dismiss the action as unfounded;

–        order the applicants to pay the costs.

III. Law

A.      The admissibility of the action

16      The Commission, supported by Ireland, contends, without formally raising a plea of inadmissibility, that the action is inadmissible.

17      In the first place, the Commission submits that the action should be declared inadmissible in so far as the applicants seek annulment of the contested decision in respect of products other than those for which they hold MAs without providing any evidence that they are acting on behalf of the MA holders concerned.

18      In the second place, the Commission contends, more specifically, that the second head of claim is inadmissible, since the applicants are not directly concerned by the contested decision, which is addressed to the Member States. It follows from Article 3 of the contested decision that the Member States have a margin of discretion in implementing that decision in that they may postpone the suspension based on their own substantive discretionary assessment. In addition, the contested decision includes implementing measures, namely national decisions suspending MAs, thereby implementing the contested decision.

19      The applicants submit that their action is admissible.

20      In accordance with the fourth paragraph of Article 263 TFEU, any natural or legal person may, under the conditions laid down in the first and second paragraphs of that article, institute proceedings against an act addressed to that person or which is of direct and individual concern to him or her, and against a regulatory act which is of direct concern to him or her and does not entail implementing measures.

21      In this case, the present action can be declared admissible only in so far as the applicants are directly and individually concerned by the contested decision, by virtue of the second hypothesis referred to in the fourth paragraph of Article 263 TFEU.

22      In the first place, as regards the condition that a natural or legal person must be individually concerned by the act of the European Union which is the subject matter of the action, it should be recalled that natural or legal persons can claim to be individually concerned only if the provision at issue affects them by reason of certain attributes peculiar to them or by reason of a factual situation which differentiates them from all other persons and thereby distinguishes them individually in the same way as the addressee (judgments of 15 July 1963, Plaumann v Commission, 25/62, EU:C:1963:17, p. 107, and of 3 October 2013, Inuit Tapiriit Kanatami and Others v Parliament and Council, C‑583/11 P, EU:C:2013:625, paragraph 72).

23      The contested decision singles out the applicants in a manner analogous to that of an addressee, since they are holders of some of the MAs for the medicinal products concerned and are designated as such in Annex I to the contested decision. It must therefore be held that the applicants are individually concerned by the contested decision.

24      In the second place, the condition that a natural or legal person must be directly concerned by the act of the European Union which is the subject matter of the action, as laid down in the fourth paragraph of Article 263 TFEU, requires that two criteria be cumulatively met. First, the act must have a direct effect on the applicant’s legal position. Secondly, it must not leave any discretion to those to whom it is addressed, who are responsible for its implementation, the latter being purely automatic and resulting from EU rules alone without the application of other intermediate rules (see, to that effect, order of 19 June 2008, US Steel Košice v Commission, C‑6/08 P, not published, EU:C:2008:356, paragraph 59, and judgment of 6 November 2018, Scuola Elementare Maria Montessori v Commission, Commission v Scuola Elementare Maria Montessori and Commission v Ferracci, C‑622/16 P to C‑624/16 P, EU:C:2018:873, paragraph 42).

25      The same is true where the opportunity for the addressees not to give effect to the EU act is purely theoretical and their intention to act in conformity with it is not in doubt (see judgment of 5 May 1998, Dreyfus v Commission, C‑386/96 P, EU:C:1998:193, paragraph 44 and the case-law cited).

26      Any act, whether of a regulatory or other nature, may, in principle, directly concern an individual and thus directly affect its legal situation, irrespective of whether it entails implementing measures (judgment of 12 July 2022, Nord Stream 2 v Parliament and Council, C‑348/20 P, EU:C:2022:548, paragraph 74).

27      In the present case, the contested decision alters the legal position of the applicants in that it obliges the Member States to suspend the MAs of which they are holders. It is apparent from paragraph 22 of the rejoinder that the Commission does not dispute that the applicants meet the first criterion mentioned in paragraph 24 above. It challenges only the second criterion mentioned in that paragraph, asserting that the Member States have a margin of discretion in implementing the contested decision.

28      As to the question whether a contested measure leaves no discretion to those to whom it is addressed and who are responsible for its implementation, the mere fact that the contested measure is to be the subject of implementing measures for the purposes of its implementation does not necessarily imply the existence of a margin of discretion in relation to the addressees of such a measure (see, to that effect, judgment of 12 July 2022, Nord Stream 2 v Parliament and Council, C‑348/20 P, EU:C:2022:548, paragraph 96).

29      The existence of a margin of discretion must necessarily be assessed in the light of the specific legal effects which are the subject of the action and which may actually have an effect on the legal position of the person concerned. Thus, it is necessary to examine the legal effects produced by that act’s provisions, as referred to in the action, on the situation of the person pleading the right to bring proceedings, pursuant to the second limb of the fourth paragraph of Article 263 TFEU (see, to that effect, judgment of 12 July 2022, Nord Stream 2 v Parliament and Council, C‑348/20 P, EU:C:2022:548, paragraphs 97 and 98 and the case-law cited).

30      In that regard, the Commission relies on Article 3 of the contested decision, which provides that, under certain conditions, a Member State may exceptionally postpone the suspension of the MAs concerned for a period not exceeding 18 months.

31      It is, admittedly, true that, by virtue of that provision, the Member States enjoy a certain margin of discretion as regards a possible postponement of the implementation of the contested decision and, thus, as regards the time at which they must suspend the MAs for the medicinal products concerned. By contrast, they have no discretion as regards the obligation to suspend those MAs, and must do so at the latest on expiry of a period of 18 months from the date of adoption of the contested decision, with the result that the applicants cannot avoid that obligation (see, to that effect and by analogy, judgment of 12 July 2022, Nord Stream 2 v Parliament and Council, C‑348/20 P, EU:C:2022:548, paragraphs 105 and 110).

32      It follows from the foregoing that the contested decision leaves the Member States no discretion as to the suspension of the MAs at issue, which is purely automatic and derives solely from EU rules. Consequently, it directly concerns the applicants in so far as they are the holders of those MAs.

33      The case-law relied on by the Commission (judgments of 13 March 2018, European Union Copper Task Force v Commission, C‑384/16 P, EU:C:2018:176, paragraphs 47 to 59; of 13 March 2018, Industrias Químicas del Vallés v Commission, C‑244/16 P, EU:C:2018:177, paragraphs 54 to 66; and order of 14 February 2019, Associazione GranoSalus v Commission, T‑125/18, EU:T:2019:92, paragraphs 74 to 85) is not capable of calling that conclusion into question. That case-law is not relevant in the present case in so far as it concerns the third hypothesis of the fourth paragraph of Article 263 TFEU (see judgments of 13 March 2018, European Union Copper Task Force v Commission, C‑384/16 P, EU:C:2018:176, paragraph 45; of 13 March 2018, Industrias Químicas del Vallés v Commission, C‑244/16 P, EU:C:2018:177, paragraph 52; and order of 14 February 2019, Associazione GranoSalus v Commission, T‑125/18, EU:T:2019:92, paragraph 65).

34      In the third place, although the applicants seek annulment of the contested decision not only in so far as it concerns them, but also in so far as it concerns the other MA holders for medicinal products listed in Annex I thereto, they neither establish nor allege that they have standing to act on behalf of those other holders, so that, even assuming that the contested decision were annulled, it could be annulled only in part in so far as that decision concerns the applicants. Thus, in any event, the annulment sought could produce its effects only in respect of the medicinal products for which the applicants hold MAs (see, to that effect, judgment of 23 September 2020, BASF v Commission, T‑472/19, not published, EU:T:2020:432, paragraph 29 and the case-law cited).

35      Furthermore, the applicants’ argument that the contested decision can be annulled only in its entirety and that any declaration of invalidity would have ‘erga omnes’ effect is incorrect and must therefore be rejected.

36      The action must therefore be declared admissible only in respect of those medicinal products for which the applicants hold MAs and the claims for annulment must be rejected as inadmissible in respect of the remainder on the ground that the applicants lack standing to seek annulment of the contested decision in so far as it concerns MAs for which they are not the holders.

B.      Submission of the defence

37      The applicants claim that the Commission’s defence is out of time. The Commission was served with the application on 27 July 2022. The defence was not lodged until 13 October 2022, whereas the time limit expired on 10 October 2022, in accordance with Article 81(1), Article 60 and Article 58(2) of the Rules of Procedure of the General Court.

38      According to the first paragraph of Article 6 of the Decision of the General Court of 11 July 2018 on the lodging and service of procedural documents by means of e-Curia (OJ 2018 L 240, p. 72), procedural documents are served via e-Curia on holders of an access account in cases concerning them. In accordance with the third paragraph of Article 6 of that decision, the procedural document is served at the time when the addressee (representative or assistant) requests access to that document.

39      In the present case, since the application was sent to the Commission via the e-Curia application on 27 July 2022 and that institution requested access to that procedural document on 3 August 2022, the time limit for submitting the defence expired on 13 October 2022.

40      Consequently, by filing the defence on 13 October 2022, the Commission complied with that time limit.

C.      The substance

41      In support of their action, the applicants rely on four pleas in law. The first plea alleges infringement of Article 116 of Directive 2001/83, as amended by Directive 2010/84. The second plea alleges infringement of the precautionary principle. The third plea alleges infringement of the principle of proportionality. The fourth plea alleges a failure to state reasons.

1.      The fourth plea in law, alleging failure to state reasons

42      The applicants claim that the contested decision is vitiated by a failure to state reasons in that it is contradictory.

43      In the first place, the applicants claim that Annex III to the contested decision lays down a condition for lifting the suspension of MAs which, if the reasoning underlying that decision is followed, would be impossible to fulfil.

44      In that regard, the applicants point out that, according to that annex, in order to obtain the lifting of the suspension, MA holders must, first, produce convincing scientific evidence attesting to a favourable risk-benefit balance for a well-defined population and, secondly, propose a series of RMMs sufficiently protecting patients who run an increased risk of serious damage to their health.

45      First, the applicants claim that the use of the medicinal products concerned in accordance with the terms of the MA already demonstrates the existence of a favourable risk-benefit balance on the basis of the scientific data currently available. According to the scientific conclusions set out in Annex II to the contested decision, the restrictions adopted in 2013 are not sufficient, since healthcare professionals are not complying with them. The applicants claim that to accept such an argument, relied on in order to reject the additional RMMs that they have proposed, would mean that it would generally be impossible to obtain the lifting of a suspension, since any alleged off-label use, even if deliberate, would systematically preclude the adoption of additional RMMs.

46      Secondly, the applicants claim that the scientific conclusions set out in Annex II to the contested decision explicitly rejected the implementation of other RMMs. Thus, the contested decision de facto orders a revocation of the MAs. The conditions for lifting the suspension could never be satisfied.

47      Thirdly, the applicants claim that the contested decision does not indicate the reasons why the decision taken in 2013 could not be maintained pending the results of two phase IV clinical trials in surgery and traumatology (‘the PHOENICS and TETHYS studies’) ordered in 2013 so as to satisfy the condition relating to the production of further evidence.

48      In the second place, the applicants claim that the contested decision is contradictory because of the exception provided for in Article 3 thereof.

49      First, it is contradictory for that provision to require a Member State which makes use of that exception and postpones the suspension of an MA to continue to comply with the RMMs introduced in 2018 despite the fact that those measures are not effective, as set out in the scientific conclusions in Annex II to the contested decision.

50      Secondly, the fact that that provision requires such a Member State to suspend supplies to an accredited establishment if it appears that that establishment is not complying with the RMMs demonstrates that the Commission considers that the additional restrictions on supplies, as proposed by the applicants, constitute effective additional RMMs, contrary to what is stated in the scientific conclusions set out in Annex II to the contested decision.

51      Thirdly, the Commission based the decision to suspend the MAs concerned on public health considerations in the Member States. At the same time, a Member State could, on the basis of Article 3 of the contested decision, exceptionally postpone the suspension for a period of only 18 months, if that was deemed necessary in the light of public health considerations in that Member State.

52      The Commission, supported by Ireland and the EMA, contends that the fourth plea in law should be rejected.

53      The obligation to state reasons constitutes an essential procedural requirement which must be distinguished from the question whether the statement of reasons is well founded, which is a question of the substantive legality of the contested measure (judgments of 22 March 2001, France v Commission, C‑17/99, EU:C:2001:178, paragraph 35, and of 20 September 2019, ICL-IP Terneuzen and ICL Europe Coöperatief v Commission, T‑610/17, EU:T:2019:637, paragraph 47).

54      According to well-established case-law, the statement of reasons required by the second paragraph of Article 296 TFEU must be appropriate to the nature of the measure at issue and must show clearly and unequivocally the reasoning of the institution which adopted the measure, so as to enable the persons concerned to ascertain the reasons for the measure and the competent court to exercise its review function. The requirement to state reasons must be assessed in the light of the circumstances of the case, in particular the content of the act, the nature of the reasons given and the interest which the addressees or other persons affected by the act within the meaning of the fourth paragraph of Article 263 TFEU may have in receiving explanations. It is not required that the statement of reasons specify all the relevant matters of fact and law, since the question whether the statement of reasons meets the requirements of Article 296 TFEU must be assessed with regard not only to its wording but also to its context and to all the legal rules governing the matter in question (see judgments of 2 April 1998, Commission v Sytraval and Brink’s France, C‑367/95 P, EU:C:1998:154, paragraph 63 and the case-law cited, and of 28 March 2017, Rosneft, C‑72/15, EU:C:2017:236, paragraph 122 and the case-law cited).

55      Moreover, the participation of interested parties in the procedure for drawing up an act may reduce the requirement to state reasons, since it contributes to informing them (judgments of 21 July 2011, Etimine, C‑15/10, EU:C:2011:504, paragraph 116, and of 1 February 2013, Polyelectrolyte Producers Group and Others v Commission, T‑368/11, not published, EU:T:2013:53, paragraph 101).

56      In the present case, it is apparent from recital 2 of the contested decision that the PRAC concluded that non-adherence to the product information continued to exist, despite the significant RMMs implemented following the referral procedure completed in 2018. According to recital 3 of the contested decision, the PRAC took into account the seriousness of the safety concerns and the fact that the proportion of patients who are exposed to those risks in the absence of effective RMMs could have significant public health consequences, including potentially increased mortality. According to recital 4 of the contested decision, the PRAC concluded that the risk-benefit balance of the medicinal products concerned was no longer favourable and therefore recommended that the MAs for those medicinal products be suspended.

57      Moreover, it follows from recital 5 and Article 1 of the contested decision that that decision was adopted on the basis of the position of the majority of the Member States represented within the CMDh, set out in a document entitled ‘Scientific conclusions’, which is reproduced in Annex II to the contested decision and thus forms an integral part thereof. According to those conclusions, it was appropriate to adopt, in the interests of the European Union, a decision suspending the MAs for the medicinal products concerned.

58      The scientific conclusions set out on 10 pages the reasons why it was considered that the risk-benefit balance of the medicinal products concerned was no longer favourable. In particular, they show that the implementation of additional RMMs was envisaged, but that it was considered that no other RMM could be identified to ensure the safe use of the medicinal products concerned.

59      The applicants’ criticism of the statement of reasons for the contested decision is directed in particular at the alleged contradictions and inconsistencies in that decision.

60      As regards the argument that the conditions for lifting the suspension laid down in Annex III to the contested decision are contradictory, since they could never be fulfilled if it is considered that the scientific conclusions explicitly reject the efficacy of RMMs, it should be noted that Annex III to the contested decision states as follows:

‘For the suspension to be lifted, the [MA] [h]older(s) shall provide robust scientific evidence showing a positive benefit-risk balance in a clinically relevant patient population(s), together with a set of [RMMs] that can sufficiently protect patients at an increased risk of serious harm from being exposed to HES solutions for infusions.’

61      Thus, according to that annex, applicants may provide all the evidence likely to demonstrate a favourable risk-benefit balance for each medicinal product concerned in a clinically relevant patient population. In particular, that evidence may cover all the risks and benefits of each medicinal product concerned and new RMM proposals. Furthermore, it is in no way apparent from the contested decision or from the scientific conclusions set out in Annex II thereto that it is logically and necessarily excluded that, in the future, RMMs may be identified which can sufficiently protect patients presenting a high risk of serious harm resulting from exposure to the medicinal products concerned. Recital 3 of the contested decision refers only to the absence of effective RMMs. The scientific conclusions state that the PRAC concluded that no RMM or combination of RMMs that could sufficiently ensure the safe use of the medicinal products concerned ‘could be identified’, the use of the past tense indicating that that finding was valid in the past up to the time when the PRAC adopted its conclusion.

62      As regards the exception provided for in Article 3 of the contested decision, recital 6 of that decision states that, exceptionally, Member States may temporarily postpone the suspension of the MAs concerned, provided that certain conditions are met to protect patients and that the previously agreed RMMs are maintained and monitored. Article 3 of the contested decision mentions, in particular, as conditions that may justify a postponement of the suspension that such a postponement is deemed necessary in the light of public health considerations in the Member State concerned (Article 3(a) of the contested decision) and several other conditions designed to ensure adherence to the RMMs.

63      Contrary to what the applicants submit, those conditions do not conflict with the suspension provided for in the contested decision, which is based on public health considerations and the fact that the RMMs were not considered sufficient to render the risk-benefit balance of the medicinal products concerned favourable.

64      The public health considerations justifying a postponement of the suspension may be different from those justifying the suspension and may relate, for example, to the need for the healthcare sector to prepare for the suspension of MAs during the postponement period, without however justifying a postponement for an unlimited period.

65      With regard to the RMMs, it should be noted that they were deemed insufficient, in particular because the initial RMMs had not been sufficiently complied with. Furthermore, a drug utilisation study evaluated in 2022 showed that the rate of non-adherence to the RMMs varied considerably between the different sites in the different Member States that took part in the study. Thus, the Commission’s approach of making the possibility for a Member State to postpone the suspension of the MAs concerned for a limited period subject to conditions designed to ensure adherence to the RMMs and of establishing an obligation to suspend the supply of an accredited establishment where it appears that the latter is not complying with the RMMs is not inconsistent with the underlying consideration of the contested decision as noted in paragraph 63 above.

66      Finally, the applicants’ argument that the contested decision should have stated the reasons why the decision taken in 2013 could not be maintained, in particular pending the results of the PHOENICS and TETHYS studies, does not demonstrate a failure to state reasons. First, as is apparent from recital 2 of the contested decision and from the scientific conclusions set out in Annex II thereto, the contested decision was taken following the assessment of the last drug utilisation study, which evaluated the effectiveness of the RMMs adopted in 2018. Secondly, if the applicants consider that the data from the PHOENICS and TETHYS studies are likely to favour the benefit-risk balance of the medicinal products concerned, they have the opportunity to submit them to the competent authority for it to assess whether the conditions for lifting the suspension are met, as set out in Annex III to the contested decision. Thirdly, the question whether the Commission should have awaited the results of those two studies before adopting the contested decision is a question of the merits of the contested decision and will be examined in the context of the third plea in law, alleging breach of the principle of proportionality.

67      Similarly, the applicants’ argument that the use of the medicinal products concerned in accordance with the terms of the MA demonstrates a favourable risk-benefit balance on the basis of the scientific data currently available is also relevant to the merits of the contested decision and will be examined in the context of the first and second pleas in law.

68      In conclusion, the contested decision, including the scientific conclusions set out in Annex II thereto, sets out in detail the reasons on which it is based. The statement of reasons for the contested decision discloses, in particular, the reasoning of the institution which adopted the measure, in such a way as to enable the applicants, who moreover participated in the procedure leading to the adoption of the contested decision, to ascertain the justification for the measure taken and to enable the Court to exercise its review.

69      Accordingly, the fourth plea in law must be rejected as unfounded.

2.      The first plea in law, alleging infringement of Article 116 of Directive 2001/83, as amended, and the second plea in law, alleging infringement of the precautionary principle

70      The applicants claim that the contested decision is unfounded, since the conditions laid down in Article 116 of Directive 2001/83, as amended, empowering the competent authorities to suspend MAs for medicinal products are not satisfied. The scientific conclusions on which the contested decision is based do not make it possible to establish an unfavourable risk-benefit balance in the context of the use of the medicinal products concerned within the approved indications.

71      The Commission, supported by Ireland and the EMA, contends that there was neither an error with regard to compliance with the legal conditions laid down in Article 116 of Directive 2001/83, as amended, nor a manifest error of assessment in the evaluation of the available data on the safety of the medicinal products concerned, since the scientific data corroborated the conclusion that the risk-benefit balance of those products could not be regarded as favourable.

(a)    The first part of the first plea in law, alleging an error of law resulting from an erroneous interpretation of the concept of ‘risk-benefit balance’ in Article 116 of Directive 2001/83, as amended

72      The applicants claim that Article 116 of Directive 2001/83, as amended, allows an MA to be withdrawn or suspended only if the product concerned proves harmful in the context of its intended use, in other words if it does not present a favourable risk-benefit balance in relation to its on-label use.

73      In that regard, the applicants submit that the risk-benefit balance, as defined in Article 1(28a) of Directive 2001/83, as amended by Directive 2004/27/EC of the European Parliament and of the Council of 31 March 2004 (OJ 2004 L 136, p. 34), is assessed primarily in the light of the intended use of the medicinal product, which is defined by the MA indications and the related restrictions on use also mentioned in the MA. According to Article 1(28) of Directive 2001/83, as amended by Directive 2004/27, the term ‘risk’ encompasses any risk relating to the quality, safety or efficacy of the medicinal product as regards patients’ health or public health. In accordance with recital 7 of Directive 2001/83, the concepts of harmfulness and therapeutic efficacy can only be examined in relation to each other and only have a relative significance assessed in the light of the state of progress of science and taking into account the intended purpose of the medicinal product. The quality, efficacy and safety of the medicinal product is always linked to its intended use, which is examined during the authorisation procedure. No medicinal product can be safe and effective if it is not used as indicated.

74      The Commission, supported by Ireland and the EMA, contends that the first part of the first plea in law should be rejected.

75      Under the first paragraph of Article 116 of Directive 2001/83, as amended, the competent authorities are to suspend, revoke or vary an MA if the view is taken that the medicinal product is harmful, that it lacks therapeutic efficacy, that the risk-benefit balance is not favourable or that the medicinal product’s qualitative and quantitative composition is not as declared.

76      Those conditions for modification, suspension or revocation of an MA are alternative and not cumulative (judgments of 11 December 2014, PP Nature-Balance Lizenz v Commission, T‑189/13, not published, EU:T:2014:1056, paragraph 41, and of 19 September 2019, GE Healthcare v Commission, T‑783/17, EU:T:2019:624, paragraph 44). They must, moreover, be interpreted in accordance with the general principle established by the case-law, according to which the protection of public health must unquestionably take precedence over economic considerations (judgments of 19 April 2012, Artegodan v Commission, C‑221/10 P, EU:C:2012:216, paragraph 99, and of 19 September 2019, GE Healthcare v Commission, T‑783/17, EU:T:2019:624, paragraph 44).

77      In the present case, the question whether the first part of the first plea is well founded depends on the interpretation of the concept of ‘risk-benefit balance’.

78      As a general rule and as follows from settled case-law, in interpreting a provision of EU law, 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 (see judgment of 7 May 2019, Germany v Commission, T‑239/17, EU:T:2019:289, paragraph 40 and the case-law cited), bearing in mind, however, that no interpretation may exceed the limits of clear and precise wording (judgment of 16 December 2020, American Airlines v Commission, T‑430/18, EU:T:2020:603, paragraph 109; see, to that effect, judgment of 15 July 2010, Commission v United Kingdom, C‑582/08, EU:C:2010:429, paragraph 51 and the case-law cited). The origins of a provision of EU law may also provide information relevant to its interpretation (see judgment of 16 March 2023, Towercast, C‑449/21, EU:C:2023:207, paragraph 31 and the case-law cited).

(1)    The literal interpretation of the concept of ‘risk-benefit balance’ in Article 116 of Directive 2001/83, as amended

79      The concept of ‘risk-benefit balance’ is defined in Article 1(28a) of Directive 2001/83, as amended, as ‘an evaluation of the positive therapeutic effects of the medicinal product in relation to the risks as defined in point 28, first indent’.

80      The first indent of Article 1(28) of that directive defines the concept of ‘risks related to use of the medicinal product’ as ‘any risk relating to the quality, safety or efficacy of the medicinal product as regards patients’ health or public health’.

81      None of the provisions cited in the preceding paragraphs explicitly includes or excludes consideration of the risks resulting from the off-label use of a medicinal product in the assessment of its risk-benefit balance under the first paragraph of Article 116 of Directive 2001/83, as amended.

82      By contrast, the wording of the first indent of Article 1(28) of Directive 2001/83, as amended, is general in that it refers to ‘any risk’ to the health of the patient or to public health relating to the quality, safety or efficacy of the medicinal product. Similarly, that provision does not restrict the concept of ‘safety of the medicinal product’ to specific uses. Thus, that provision does not exclude risks resulting from the off-label use of a medicinal product from the concept of risks relating to the safety of that medicinal product.

83      It follows from a literal interpretation of the concept of ‘risk-benefit balance’ in Article 116 of Directive 2001/83, as amended, that it is sufficiently broad to allow account to be taken of the risks associated with the off-label use of a medicinal product.

(2)    The contextual interpretation of the concept of ‘risk-benefit balance’ in Article 116 of Directive 2001/83, as amended

84      First, Article 23(2) of Directive 2001/83, as amended, lays down certain information obligations incumbent on each holder of an MA for a medicinal product.

85      While it is true that that provision does not as such provide for the suspension of an MA on the grounds of an unfavourable risk-benefit balance resulting from consideration of the risks associated with the off-label use of a medicinal product, the fact remains that it explicitly lays down the obligation for the holder of the MA to communicate to the competent national authority ‘data on the use of the medicinal product where such use is outside the terms of the [MA]’. That obligation to provide information would, to a large extent, be rendered meaningless if the competent authority were unable to take such data into account and draw regulatory consequences from them.

86      Moreover, it follows from that provision, and more particularly from the link between, first, the end of the first sentence of the second paragraph and, secondly, the end of the second sentence of the second paragraph, that ‘data on the use of the medicinal product where such use is outside the terms of the [MA]’ may be ‘new information which might influence the evaluation of the benefits and risks of the medicinal product concerned’.

87      Secondly, Article 101(1) of Directive 2001/83, as amended, describes the objective of the pharmacovigilance system and the scope of the information which it serves to collect. It follows from that provision that the concept of ‘risks … as regards patients’ or public health’ also covers risks resulting from ‘use outside the terms of the [MA]’. Neither Directive 2001/83 nor any other provision of EU law contains any indication that that concept should have, in the context of Article 101 of Directive 2001/83, a different scope from that in the context of Article 116 of that directive, in that, in the context of the former provision, it would cover the risks associated with the off-label use of the medicinal product concerned and, in the context of the latter provision, it would not cover such risks.

88      Moreover, it follows from Article 101(2) of Directive 2001/83, as amended, that the Member States may take into account all the information collected within the framework of the pharmacovigilance system, including information relating to the risks associated with the off-label use of a medicinal product, in order to examine options for preventing or reducing the risks and, if necessary, to take regulatory measures concerning the MA concerned. That provision contains no indication that the suspension or revocation of an MA would be excluded as a matter of principle from the measures that Member States could take in order to address the risks associated with the off-label use of a medicinal product.

89      Thirdly, in the second paragraph of Article 22 of Directive 2001/83, as amended by Directive 2010/84, which admittedly concerns a situation different from that in the present case, namely the decision to grant an MA subject to certain conditions, the legislature expressly laid down the conditions under which information on the efficacy and safety of a medicinal product relating to its use ‘under normal conditions’ may be limited.

90      Thus, the fact that Article 116 of Directive 2001/83, as amended, contains no reference to ‘normal conditions’ of use corroborates the Commission’s interpretation that the concept of ‘risks related to use of the medicinal product’ also covers the risks associated with its off-label use.

91      It follows from a contextual interpretation of the concept of ‘risk-benefit balance’ in Article 116 of Directive 2001/83, as amended, that that concept also covers the risks associated with the off-label use of a medicinal product.

92      The applicants’ argument that Article 23(2) and Article 101 of Directive 2001/83, as amended, make provision for only pharmacovigilance obligations imposed on MAs holders and not for the suspension of an MA, is not capable of calling that conclusion into question. Those provisions are not examined as the legal basis for the contested decision, but as part of a contextual interpretation of the concept of ‘risk-benefit balance’ in Article 116 of Directive 2001/83, as amended.

(3)    The interpretation of the concept of ‘risk-benefit balance’ taking into account the objective of Article 116 of Directive 2001/83, as amended

93      It follows from recital 2 of Directive 2001/83 that safeguarding public health is its essential objective (see judgment of 5 May 2011, Novo Nordisk, C‑249/09, EU:C:2011:272, paragraph 37 and the case-law cited).

94      Article 116 of Directive 2001/83, as amended, pursues that objective, in so far as it imposes an obligation on the competent authorities to suspend, revoke or vary MAs where the risk-benefit balance of a medicinal product is considered to be unfavourable.

95      As noted in paragraph 76 above, Article 116 of Directive 2001/83, as amended, must be interpreted in accordance with the general principle established by the case-law, according to which the protection of public health must unquestionably take precedence over economic considerations.

96      In order to ensure the effective pursuit of that objective, the competent authorities must be able to take into account information relating to all the risks that a medicinal product poses to public health, including those associated with off-label use. The off-label use of a medicinal product may pose risks to public health similar to those associated with its on-label use. Off-label use of a medicinal product is not uncommon. It is a professional decision taken by a medical practitioner who assesses the benefits and risks involved. That practitioner must therefore be informed as fully as possible.

97      Consequently, the applicants’ argument that suspending the MA for a safe and effective medicinal product when used in accordance with the indications does not, by definition, meet the objective of safeguarding public health must be rejected.

98      It follows from an interpretation of the concept of the risk-benefit balance, taking account of the objective of Article 116 of Directive 2001/83, as amended, that that concept also covers the risks associated with the off-label use of a medicinal product.

99      In conclusion, it follows from a literal, contextual and purposive interpretation of the first paragraph of Article 116 of Directive 2001/83, as amended, that the Commission did not infringe that provision by taking account of the risks posed by the off-label use of the medicinal products concerned when assessing their risk-benefit balance.

100    That conclusion is corroborated by the preparatory work for Directive 2010/84, which amended, in particular, Article 116 of Directive 2001/83. In particular, it appears from Annex I to the Commission’s working document of 10 December 2008 accompanying the proposal for a directive amending Directive 2001/83 that the concept of ‘normal conditions of use’ has been deleted ‘as this is not defined and could be interpreted as restricting regulatory action in the case of a major public health issue related to off-label use ([for example,] in children)’.

(4)    The applicants’ other arguments

101    None of the other arguments raised by the applicants is capable of calling into question the conclusion that the concept of the risk-benefit balance set out in Article 116 of Directive 2001/83, as amended, also covers the risks associated with the off-label use of a medicinal product.

102    First, as regards the applicants’ argument that the MA holders for a medicinal product are not responsible for its off-label use, which, in their view, follows from the case-law, it is sufficient to note that that circumstance does not preclude the taking into account of the risks associated with the off-label use of a medicinal product in order to assess the risk-benefit balance of that medicinal product in the framework of the exercise of the power to adopt regulatory measures. The contested decision neither presupposes nor suggests that the applicants and the other MA holders of the medicinal products concerned are responsible for the off-label use of those medicinal products.

103    Secondly, the argument based on recital 7 of Directive 2001/83 is irrelevant, since that recital refers to the original text of Directive 2001/83, including Article 116 thereof, which did not provide for the alternative on which the contested decision is based, namely that an MA is suspended, withdrawn or amended where it is considered that the risk-benefit balance is not favourable.

104    Thirdly, the argument that EU law does not govern the off-label use of medicinal products is not relevant either. The fact that EU law does not define the conditions for the off-label use of medicinal products does not mean that the risks resulting from such use cannot or must not be taken into account in the framework of the exercise of the European Union’s power to adopt regulatory measures.

105    Fourthly, the applicants’ argument that the misuse of a medicinal product could be countered only at national level cannot succeed either. The fact that national law provides for consequences relating to the misuse of a medicinal product does not prevent the authorities of the European Union from being able, or even obliged, to exercise the power entrusted to them, with diligence, to adopt regulatory measures, including in order to deal with the risks resulting from off-label use.

106    Consequently, the first part of the first plea in law must be rejected.

(b)    The second part of the first plea in law, alleging a manifest error of assessment, and the second plea in law, alleging breach of the precautionary principle

107    The applicants refer to the case-law relating to the precautionary principle, according to which that principle authorises the competent authorities, where there is uncertainty, to take appropriate measures to prevent certain potential risks to human health, safety and the environment, without having to wait until the reality and seriousness of those risks have been fully demonstrated. According to that case-law, it would be for the competent authority to prove that the conditions for suspending an MA had been met, which the Commission and the EMA failed to do in the present case. To that end, the Commission would be required to produce serious and conclusive evidence giving rise to reasonable doubt as to the safety of the medicinal product concerned. Such evidence is totally lacking in the present case.

108    In that regard, the applicants raise, in particular, arguments aimed at calling into question the establishment of the risks, the alleged failure to take into account the benefits of the medicinal products concerned resulting from a medical need and the conclusions relating to the absence of effective RMMs.

109    The Commission, supported by Ireland and the EMA, contends that the second part of the first plea in law and the second plea in law should be rejected.

110    The precautionary principle, which constitutes a general principle of EU law, authorises the competent authorities, where there is uncertainty, to take appropriate measures in order to prevent certain potential risks for public health, safety and the environment without having to wait until the reality and seriousness of those risks become fully apparent (see judgment of 19 September 2019, GE Healthcare v Commission, T‑783/17, EU:T:2019:624, paragraph 45 and the case-law cited; see also, to that effect, judgment of 3 December 2015, PP Nature-Balance Lizenz v Commission, C‑82/15 P, not published, EU:C:2015:796, paragraph 21).

111    Consequently, in accordance with the precautionary principle, the risks to health which the grounds mentioned in the first paragraph of Article 116 of Directive 2001/83 aim to prevent need not be specific, but only potential (judgment of 19 September 2019, GE Healthcare v Commission, T‑783/17, EU:T:2019:624, paragraph 46; see also, to that effect, judgments of 10 April 2014, Acino v Commission, C‑269/13 P, EU:C:2014:255, paragraph 59, and of 3 December 2015, PP Nature-Balance Lizenz v Commission, C‑82/15 P, not published, EU:C:2015:796, paragraph 23).

112    In that system, the first paragraph of Article 116 of Directive 2001/83 confers rights on undertakings holding MAs, since it guarantees that they will retain their MAs until the existence of one of the conditions for amending, suspending or withdrawing them is established (see, to that effect, judgment of 19 April 2012, Artegodan v Commission, C‑221/10 P, EU:C:2012:216, paragraph 96). It follows that, as regards the burden of proof, it is for the competent authority, in this case the Commission, to establish that the conditions relating to the revocation, suspension or modification of an MA, set out in Article 116 of Directive 2001/83, are met (see, to that effect, judgments of 7 March 2013, Acino v Commission, T‑539/10, not published, EU:T:2013:110, paragraph 79, and of 19 September 2019, GE Healthcare v Commission, T‑783/17, EU:T:2019:624, paragraph 47).

113    Having regard to the precautionary principle, the Commission may nevertheless confine itself to providing serious and conclusive evidence which, without ruling out scientific uncertainty, gives rise to reasonable doubt as to the safety of the medicinal product concerned, its therapeutic effect, the existence of a favourable risk-benefit balance or the qualitative and quantitative composition declared (judgments of 3 December 2015, PP Nature-Balance Lizenz v Commission, C‑82/15 P, not published, EU:C:2015:796, paragraph 23, and of 19 September 2019, GE Healthcare v Commission, T‑783/17, EU:T:2019:624, paragraph 48).

114    However, the adoption of a decision to vary, suspend or revoke an MA for a medicinal product is justified only where that decision is substantiated by new, objective, scientific or medical data (see judgment of 19 September 2019, GE Healthcare v Commission, T‑783/17, EU:T:2019:624, paragraph 49 and the case-law cited).

115    In that regard, the competent authority is required to indicate the main scientific reports and expert opinions on which it relies and to specify, in the event of significant divergence, the reasons why it departs from the conclusions of the reports or expert opinions produced by the companies concerned. That obligation is particularly important in cases of scientific uncertainty. It is a question of proceeding in an adversarial and transparent manner, in order to guarantee that the substance considered has undergone an in-depth and objective scientific assessment based on a comparison of the most representative scientific opinions with the scientific arguments put forward by the pharmaceutical laboratories concerned (judgment of 19 September 2019, GE Healthcare v Commission, T‑783/17, EU:T:2019:624, paragraph 50).

116    As regards the judicial review exercised by the Court, the present action is in the medico-pharmacological field, which is of a complex technical and scientific nature.

117    According to the case-law, when the Commission is called upon to carry out complex technical or scientific assessments, it has a broad discretion. In the context of its judicial review, the EU judicature must determine whether the relevant procedural rules have been complied with, whether the facts established by the Commission are correct and whether there has been a manifest error of appraisal of those facts or a misuse of powers (see judgment of 9 September 2010, Now Pharm v Commission, T‑74/08, EU:T:2010:376, paragraph 111 and the case-law cited).

118    As regards the scope of the review of the scientific assessments, it should be borne in mind that the Court cannot substitute its own assessment for that of the CMDh, which followed the PRAC’s recommendation. Its judicial review is exercised only over the lawfulness of its operation and over the internal consistency and reasoning of the CMDh’s opinion. As regards the latter aspect, the Court is only entitled to verify whether the opinion contains a statement of reasons from which it is possible to ascertain the considerations on which the opinion is based, and whether it establishes a comprehensible link between the medical and/or scientific findings and its conclusions (see, to that effect, judgments of 11 December 2014, PP Nature-Balance Lizenz v Commission, T‑189/13, not published, EU:T:2014:1056, paragraph 52, and of 19 September 2019, GE Healthcare v Commission, T‑783/17, EU:T:2019:624, paragraph 51).

119    However, in the present case, as is clear from the first paragraph of Article 1 of the contested decision, the Commission, after consulting the Standing Committee on Medicinal Products for Human Use (recital 7 of the contested decision), based that decision on the scientific conclusions of the CMDh, that committee for its part agreeing with the overall conclusions of the PRAC and its reasons justifying the recommendation. The CMDh’s scientific conclusions are set out in Annex II to that decision and thus form an integral part of the contested decision. Accordingly, it must be held that the judicial review incumbent on the Court, in particular the examination as to whether there has been a manifest error of assessment, must be exercised over those scientific conclusions as a whole (see, to that effect and by analogy, judgments of 5 December 2018, Bristol-Myers Squibb Pharma v Commission and EMA, T‑329/16, not published, EU:T:2018:878, paragraph 98, and of 19 December 2019, Vanda Pharmaceuticals v Commission, T‑211/18, EU:T:2019:892, paragraph 54).

120    In the first place, the procedure which led to the adoption of the contested decision follows several assessments which took place in 2013, in the context of a referral procedure in 2012 under Article 31 of Directive 2001/83, in the context of a referral procedure in 2013 under Article 107i of that directive, as amended, and in 2018, in the context of a referral procedure under that article. As can be seen from the scientific conclusions set out in Annex II to the contested decision, in the context of the 2012 and 2013 referrals, it was established that the medicinal products concerned were associated with an increased risk of mortality for patients suffering from sepsis, with renal failure or who are critically ill. In that regard, the PRAC referred to three randomised controlled trials (‘RCTs’), namely the 6S (Perner et al., 2012), VISEP (Brunkhorst et al., 2008) and CHEST (Myburgh et al., 2012) studies and two meta-analyses, namely the Zarychanski et al., 2013, and Perel, Roberts, and Ker, 2013, studies. According to the 6S and VISEP studies, treatment with HES-containing products was associated with an increased risk of mortality at 90 days. Those results were also confirmed by the two meta-analyses. The RCTs uniformly demonstrated the existence of an increased risk of renal replacement therapy or renal failure after treatment with HES-containing products.

121    It is clear from the file, and more particularly from the PRAC’s report in the context of the 2013 referral, that that committee considered that, for patients suffering from sepsis or septic shock, the risks of increased mortality outweighed the limited benefits of the medicinal products concerned. In addition, the PRAC found that ‘evidence is not available … to indicate that [the benefits of the] use [of the medicinal products] in other indications [outweighed] the risks’.

122    As a result, RMMs have been established, notably in the form of new contraindications, including for patients suffering from sepsis, renal failure or who are critically ill, and of new warnings.

123    In the second place, the two drug utilisation studies imposed in the context of the 2013 referral revealed in 2017 and 2018 a significant failure to comply with those initial RMMs. It appears from the file, and more particularly from the PRAC’s assessment report in the context of the 2018 referral, that that committee considered that the benefits of the medicinal products concerned in the context of on-label use were modest and that the claimed clinical usefulness of the medicinal products concerned did not outweigh the risks of increased mortality and renal failure for patients who are critically ill or suffering from sepsis who continued to be exposed to those medicinal products. As a result, the PRAC recommended the suspension of the MAs for the medicinal products concerned already in 2018.

124    Nevertheless, as stated in the scientific conclusions on which the contested decision is based, in 2018 the Commission decided that the medicinal products concerned could remain on the market provided that a set of additional RMMs was implemented in order to ensure that the medicinal products concerned were not used for patients at risk of serious harm. The performance of a new drug utilisation study to assess the efficacy of those new RMMs was imposed as a condition of the MA, since compliance with the new procedure and those RMMs was considered essential in order to ensure a favourable risk-benefit balance of the medicinal products concerned.

125    In the third place, the PRAC concluded in 2013 and confirmed in 2018 that additional safety studies were needed for elective surgery and trauma patients as the risks may be lower in those patients. In that regard, Annex IV to Commission Implementing Decision C(2013) 9793 (final) states the following:

‘[The MA holder(s)] should perform two phase IV [RCTs] with an appropriate control and clinically meaningful endpoints to demonstrate the efficacy and safety in the perioperative setting and trauma.’

126    Although, according to that annex, the final study reports were expected to be available at the end of 2016, the results of the studies were not available at the time of the procedure leading to the adoption of the contested decision.

127    In the fourth place, it is apparent from the scientific conclusions set out in Annex II to the contested decision that the main purpose of the 2022 PRAC assessment report was the latest drug utilisation study.

128    However, on that occasion, the PRAC also considered additional evidence provided by the holders of the MAs concerned, including new studies, namely the studies by Gupta, 2021, Suzuki, 2020, Kwak, 2018, Nizar, 2020, Mahrouse, 2021, and Lee, 2021. In that regard, the PRAC concluded that those data did not call into question the benefits and risks of the use of the medicinal products concerned, as established, and did not contain any meaningful information relating to a potential change in their safety profile.

129    With regard to the drug utilisation study evaluated in 2022, the PRAC assessed its results as well as the responses provided by the MA holders. On that basis, the committee concluded that product information was still not being adhered to despite the wide range of additional RMMs implemented in 2018. In particular, the PRAC expressed serious concerns about what it considered to be a persistently high rate of non-adherence to contraindications, constituting 6.6% of non-adherent prescriptions, namely 3.5% of prescriptions for critically ill patients, 2.2% of prescriptions for patients with renal failure and around 1% of prescriptions for patients with sepsis. In addition, the results of that study showed that 7 of the 18 patients suffering from sepsis and treated with an HES solution had, in addition to sepsis, an additional contraindication. Furthermore, the PRAC was concerned about the high overall rate of non-adherence to the product information in two Member States.

130    The PRAC also indicated that the restriction on the dose administered and the duration of treatment was generally well respected. However, it considered that the scientific data also included some evidence demonstrating harm for patients treated with doses in line with existing recommendations. Consequently, it concluded that it was not possible to conclude that HES-containing products were safe for patients with contraindications when the dose administered and the duration of treatment adhered to the recommendations.

131    In addition, the PRAC considered whether the introduction of additional RMMs could reduce non-adherence to the product information for the medicinal products concerned. As noted in paragraph 8 above, the PRAC concluded that no additional RMMs or combination of RMMs could be identified that would sufficiently ensure the safe use of the medicinal products concerned.

132    In the fifth place, the PRAC and, subsequently, the CMDh concluded that the risk-benefit balance of the medicinal products concerned was no longer favourable since the risks associated with the use of those medicinal products outweighed the benefits. Consequently, the PRAC and, subsequently, the CMDh recommended suspending the MAs for the medicinal products concerned. In the contested decision, the Commission, after consulting the Standing Committee on Medicinal Products for Human Use, followed that recommendation.

133    In the light of the considerations set out in paragraphs 110 to 119 above and in the light of the arguments put forward by the applicants, it is therefore necessary to examine whether the applicants’ arguments demonstrate that the scientific conclusions set out in Annex II to the contested decision are vitiated by a manifest error of assessment concerning the evaluation of the risk-benefit balance and, consequently, that the contested decision based on those conclusions is unlawful.

(1)    Establishing the risks associated with the off-label use of medicinal products

134    In the first place, the applicants claim that the risk-benefit balance of the medicinal products concerned is favourable. There are numerous clinical data confirming the efficacy and safety of those medicinal products. Periodic safety update reports regularly establish a robust safety profile. Moreover, the PRAC itself confirmed in 2021 that the risk-benefit balance remained the same. The use of the medicinal products concerned only gave rise to safety concerns for patients suffering from sepsis, renal failure or who are critically ill. No scientific evidence of an unfavourable risk-benefit balance with regard to other patient populations (in surgery and traumatology, for example) existed in 2013 and was therefore not examined. The applicants claim that there are no safety signals with regard to on-label use. In support of that claim, the applicants rely on numerous data, including recent scientific publications, guidelines from the Deutsche Gesellschaft für Anästhesiologie und Intensivmedizin (German Society for Anaesthesiology and Intensive Care Medicine) and other medical guidelines. The scientific conclusions do not take that evidence into account. Nor does the drug utilisation study evaluated in 2022 suggest that the risk-benefit balance of the medicinal products concerned is unfavourable in the context of on-label use.

135    In the second place, the Commission does not sufficiently distinguish between off-label use in critically ill patients and off-label use in the context of other indications not associated with safety risks.

136    In the third place, the applicants claim that the risk-benefit balance considered favourable in 2013 for the revised indication should have been examined for elective surgery and trauma patients. The PHOENICS and TETHYS studies were carried out respectively in order to provide further evidence in those patient populations. Pending the clinical trial reports for those two studies, there was no clinical or scientific evidence to suggest that the risk-benefit balance of the medicinal products concerned was unfavourable for on-label use.

137    In their observations on the statement in intervention submitted by the EMA, the applicants state that the PHOENICS and TETHYS studies are now available. Those studies prove the safety and efficacy of HES-containing products and that the contested decision was unjustified. Moreover, the conditions for lifting the suspension provided for in Annex III to the contested decision are met, as the Commission itself appears to acknowledge in points 88 and 89 of its rejoinder, thanks to the data resulting from those studies.

138    The PHOENICS study, relating to elective surgery patients, demonstrated that a modern balanced HES solution is ‘not inferior’ to an appropriate electrolyte solution in terms of safety. It also demonstrated that the first product was more effective than the second in terms of haemodynamic stabilisation and fluid balance.

139    The TETHYS study, relating to trauma patients, showed that modern balanced HES solution was not inferior to the electrolyte solution with regard to the primary composite endpoint of 90-day mortality and 90-day renal failure.

140    In the fourth place, the applicants submit that the contested decision is based on the results of the CHEST study carried out in 2012, the reliability of which was called into question, inter alia, by an ad hoc committee of experts of the EMA and which gave rise to criticism within the scientific community. The results of that study are no longer relevant.

141    First, the volumes of HES used in that study and in the VISEP and 6S studies are much greater than those used in current clinical practice.

142    Secondly, those studies no longer reflect current clinical practice in that the current median duration of treatment, as a result of the measures implemented in 2013 and 2019, are very short and limited to a single prescription for almost all patients.

143    Thirdly, the re-analysis of the CHEST study is flawed, partly because it was carried out by virtually the same authors who conducted the original study without the original data ever being communicated, as the associate editor of the British Medical Journal, Mr Peter Doshi, highlighted in the article provided in Annex A.28 to the reply. The study by Kajdi et al, 2014, cited by the PRAC in its 2018 assessment report, does not concern either the CHEST study or HES.

144    The Commission, supported by the EMA, disputes those arguments.

145    In the first place, the Court finds that the contested decision is based on the consideration that the risk-benefit balance of the medicinal products concerned is not favourable because of the risks resulting from the off-label use of those medicinal products. In that regard, it is apparent from the scientific conclusions on which the contested decision is based, and more particularly from the overall conclusion, that the PRAC considered that that ratio was no longer favourable in view of the seriousness of the safety problems and the proportion of patients with contraindications, including patients who are critically ill, with renal failure or suffering from sepsis, who nevertheless continued to be exposed to risks resulting from the use of medicinal products containing HES, including an increased risk of mortality. It also follows that the PRAC considered that adherence to the RMMs established in 2013 and 2018 was essential in order to ensure a favourable risk-benefit balance of the medicinal products concerned.

146    The fact that the contested decision is based on the risks associated with off-label use, in particular for patients with contraindications, complies with the first paragraph of Article 116 of Directive 2001/83, as amended, as follows from the considerations set out in paragraphs 77 to 106 above. The applicants’ arguments to the effect that the risk-benefit balance of the medicinal products concerned would be favourable if account had been taken only of on-label use of the medicinal products concerned are therefore ineffective (see, to that effect, judgment of 21 September 2000, EFMA v Council, C‑46/98 P, EU:C:2000:474, paragraph 38).

147    In the second place, the argument that the Commission did not sufficiently distinguish between off-label use in critically ill patients and off-label use in other indications not associated with safety risks must be rejected. As noted in paragraph 129 above, the PRAC expressed serious concerns about a high rate of non-adherence to contraindications. In addition, the PRAC noted that, given that serious harm had been demonstrated for patients with sepsis, renal failure or who are critically ill, the significant proportion of use of HES solutions for infusions in those populations raised significant public health concerns.

148    In the third place, as regards the risks for patients suffering from sepsis, with renal failure or who are critically ill, it is clear from the considerations set out in paragraph 120 above that the latter were established in the context of the 2012 and 2013 referrals on the basis of three studies and two meta-analyses. Moreover, as noted in paragraph 128 above, additional evidence provided by the MA holders in question, including new studies, was also examined. According to the scientific conclusions on which the contested decision is based, the latter also did not justify a different conclusion as to the benefits and risks of the medicinal products concerned.

149    None of the applicants’ arguments is capable of calling those conclusions into question.

(i)    The absence of adverse safety signals

150    It appears from the file that the PRAC, in the context of the 2018 referral, examined and rejected the argument of the MA holders concerned based on the alleged absence of adverse safety signals for the safety of the medicinal products concerned. In that regard, the PRAC noted that those data were particularly difficult to interpret. Given the nature of the medicinal products concerned, the fact that they had been on the market for decades, the situations in which they were used, namely emergency situations in which patients were treated with multiple therapies, and the possibility of a delay between acute exposure to HES and the onset of renal failure or death, the PRAC expected a considerable rate of ‘under-reporting’. The committee therefore concluded that, given the limitations of spontaneous reporting, the risks involved had been established and confirmed on the basis of data from RCTs and not spontaneous reporting.

151    The applicants put forward only one argument in that regard. In their view, ‘over-reporting’ should be expected in view of the fact that the medicinal products concerned are used in situations in which patients are treated with multiple therapies and because of the dissemination of several direct communications to healthcare professionals.

152    That argument does not demonstrate that the data resulting from spontaneous reporting are more reliable than the RCTs and meta-analyses on which the scientific conclusions are based and is not likely to demonstrate that the decision to establish the risks on the basis of the RCTs and meta-analyses rather than the data resulting from spontaneous reporting was manifestly erroneous.

(ii) Reliability of the CHEST study

153    First of all, it appears from the file that the PRAC, in its 2018 assessment report, considered concerns about the conduct of the CHEST study expressed by some stakeholders. The PRAC also indicated that, in its previous analysis, it had highlighted the potential limitations of that study. Nevertheless, in the same assessment report, the PRAC concluded that, in the independent analysis of that study conducted by the Duke Clinical Research Institute, minor differences were observed in some secondary and tertiary outcomes but that the latter did not affect the conclusions. The PRAC considered that that new analysis was sufficiently reassuring and that it confirmed the main results of the CHEST study.

154    In that regard, it is admittedly true that, on page 26 of the 2018 assessment report, it is incorrectly stated that the analysis conducted by the Duke Clinical Research Institute was carried out by ‘Kajdi, M.E., et al.’ in ‘2014’. However, that is a clerical error, as can be seen in particular from the reference to the authors and year of publication ‘Patel, A., et al., 2017’, which then appears on the same page and in the list of references in that report.

155    Next, the mere fact that the CHEST study has been criticised within the scientific community is not capable of demonstrating that the PRAC’s assessment based on the abovementioned analysis is manifestly erroneous, in particular because it takes into account certain limitations of that study.

156    Finally, the applicants do not raise any argument calling into question the content of that new analysis or its assessment by the PRAC, but confine themselves to challenging the independence of that analysis. However, it is clear from the article by Mr Doshi provided in Annex A.28 to the reply, on which the applicants rely, that only three of the eight authors of the independent analysis had also participated in the CHEST study and that, according to the New England Journal of Medicine which published the analysis, those authors of the CHEST study were involved in the new independent analysis of that study solely in order to confirm that the data had been correctly received and that the data elements had been correctly identified. In those circumstances, the applicants have not demonstrated that the PRAC could not rely on the analysis in question.

157    The applicants’ arguments concerning the reliability of the CHEST study must therefore be rejected.

(iii) The lack of data based on the revised dosage

158    The applicants’ argument that the results of the CHEST study and those of the VISEP and 6S studies are no longer relevant in that they no longer reflect current clinical practice with regard to dosage must be rejected. It was decided in 2013, on the basis of those studies in particular, that certain patients should not receive any treatment with the medicinal products concerned. Moreover, the applicants did not challenge that decision at the time. Accordingly, they cannot challenge that decision in the context of the new procedure which led to the adoption of the contested decision.

159    It is true that it is for the Commission to establish that the conditions relating to the suspension of the MAs concerned, set out in Article 116 of Directive 2001/83, are fulfilled (see, to that effect, judgment of 7 March 2013, Acino v Commission, T‑539/10, not published, EU:T:2013:110, paragraph 79) and that the adoption of a decision to suspend MAs is justified only if that decision is supported by objective and new scientific or medical data (see judgment of 19 September 2019, GE Healthcare v Commission, T‑783/17, EU:T:2019:624, paragraph 49 and the case-law cited). However, that circumstance does not mean that the Commission is not entitled to use data which have been used in previous proceedings in order to support the same conclusions. In the present case, the Commission did not depart from the previous conclusions relating to the risks established adopted in those proceedings. On the contrary, the contested decision is based on the new conclusion that the RMMs are not sufficiently effective to render the risk-benefit balance favourable. In that regard, the Commission relied on new data, namely those obtained from the latest drug utilisation study relating to the efficacy of RMMs.

160    Consequently, it must be held that the Commission was right to rely in the contested decision on certain findings and conclusions of the previous proceedings.

(iv) PHOENICS and TETHYS studies

161    The PHOENICS and TETHYS studies, the reports of which were provided by the applicants in Annexes A.29 and A.30 to their observations on the pleadings in intervention submitted by Ireland and the EMA, filed at the Registry of the General Court on 15 March 2023, were completed respectively on 6 July and 25 June 2022. The reports submitted are dated 16 and 17 February 2023 respectively.

162    According to the case-law, in the context of an action for annulment based on Article 263 TFEU, the lawfulness of European Union measures must be assessed on the basis of the facts and the law as they stood at the time when those measures were adopted (judgment of 20 September 2019, PlasticsEurope v ECHA, T‑636/17, EU:T:2019:639, paragraph 217).

163    The reports of the PHOENICS and TETHYS studies were not available when the contested decision was adopted, namely on 24 May 2022, and could not therefore be taken into consideration by the Commission.

164    Consequently, the applicants’ arguments based on those two studies are ineffective in the context of the present dispute.

(2)    The medical need and alternative treatments

165    The applicants maintain that the alternative treatments to the medicinal products concerned are less evaluated, potentially less safe and do not satisfy the existing medical need. The scientific conclusions do not take that into account. HES is safe and better than crystalloids, as demonstrated in the Chappell et al. 2021 study, extracts of which are provided in Appendix A.25 to the application. Excessive use of crystalloids is associated with increased health risks for patients. In addition, alternative synthetic colloids have been the subject of fewer studies than HES and are probably not better alternatives.

166    The Commission disputes those arguments.

167    In that regard, it appears from the file that the PRAC, in the context of the 2018 referral under Article 107i of Directive 2001/83, as amended, considered and discussed whether the suspension of the MAs for the medicinal products concerned would give rise to an unmet medical need. In particular, the PRAC took into account the information submitted by the Member States and the Kingdom of Norway, as well as the opinions of the holders of MAs for the medicinal products concerned.

168    More specifically, the PRAC examined the arguments raised by eight Member States to the effect that the suspension of the MAs for the medicinal products concerned could have an impact on national clinical practices, since the medicinal products concerned met a medical need on their territory at the time. In that regard, the PRAC examined and rejected, in particular, the argument that the use of the medicinal products concerned could be relevant for patients with an increased risk of fluid overload and tissue oedema, given the claimed volume-saving effect.

169    The PRAC also examined the arguments raised by six Member States to the effect that the risk-benefit balance of alternatives to the medicinal products concerned was no better. In that regard, it found that this claim was not supported by relevant data. In particular, the PRAC examined two studies, namely the 2016 Ripollés et al. study and the 2018 Ertmer et al. study submitted by a Member State. It concluded that the possibility of a medical need not being met in the event of suspension of the MAs for the medicinal products concerned had not been established.

170    Furthermore, as noted in paragraph 128 above, it is apparent from the scientific conclusions on which the contested decision is based that the PRAC considered additional evidence provided by the holders of the MAs at issue, in particular the studies by Gupta, 2021, Suzuki, 2020, Kwak, 2018, Nizar, 2020, Mahrouse, 2021 and Lee, 2021, in which HES-containing solutions were compared with either albumin or crystalloids. More specifically, with regard to the 2021 Chappell meta-analysis on which the applicants rely in support of their argument, the PRAC noted that it had already assessed most of the clinical trials comparing HES-containing solutions with crystalloids in that meta-analysis that had been published prior to the 2017 referral. In its opinion, the four articles published after 2018 did not identify an increased risk of renal toxicity for patients treated with products containing 6% HES. As already noted in paragraph 128 above, the PRAC concluded that those data did not call into question the benefits and risks of the medicinal products concerned as they had been established and that they did not contain any meaningful information relating to a potential change in their safety profile.

171    Finally, it is also apparent from the scientific conclusions on which the contested decision is based that the CMDh examined and rejected certain contributions from third parties, received after the adoption of the PRAC recommendation, relating in particular to the safety profile and the relevance of the medicinal products concerned in the therapeutic options for the treatment of hypovolaemia. As the Commission points out, without being contradicted on that point by the applicants, in those contributions the third parties asserted that HES-containing products presented a better risk-benefit balance and better availability than alternative treatments.

172    As stated in the scientific conclusions on which the contested decision is based, the CMDh concluded that those observations did not affect the overall conclusions and recommendation of the PRAC.

173    The applicants merely put forward an alternative view of medical need and alternative treatments, referring in paragraphs 59 and 60 of the application to a number of publications in support of their argument. However, their argument in the application remains general and does not refer to the detailed considerations of the PRAC and the CMDh, and the applicants do not explain to what extent those considerations are manifestly erroneous.

174    Consequently, their argument that the alternatives to the medicinal products concerned do not offer a better risk-benefit balance and that the scientific conclusions on which the contested decision is based do not take that into account cannot succeed.

(3)    The effectiveness of the RMMs

175    In the first place, the applicants claim that the drug utilisation study evaluated in 2022 shows that the revised dosage is almost 100% complied with. Moreover, that study tends to indicate that the contraindications are complied with by the vast majority of healthcare professionals who took part in the study, with the exception of two establishments in Italy which have since lost their accreditation to use the medicinal products concerned. It appears that, in most countries, prescriptions made for the medicinal products concerned despite a contraindication are the exception rather than the rule. According to the applicants, such off-label use cannot be extrapolated to other countries that did not participate in the drug utilisation study evaluated in 2022. In addition, the scientific conclusions showed that the highest rate of non-adherence, namely that observed in Belgium, was mainly linked to the use of low doses during caesarean deliveries, which was not associated with any particular risk for patients, but which had benefits, as shown by several studies and publications and the S 3 guideline ‘intravascular volume therapy in adults’ adopted by 14 German companies.

176    In the second place, the applicants submit that the effectiveness of the RMMs is confirmed by the fact that the total number of units of medicinal products sold has significantly decreased.

177    In the third place, the applicants claim that non-adherence to the RMMs should not be taken into account, since they are not responsible for it and EU law does not apply.

178    The Commission, supported by the EMA, disputes those arguments.

(i)    The representativeness of the sample taken into account for the drug utilisation study evaluated in 2022

179    Studies such as the drug utilisation study at issue are carried out by default on a sample representative of the target population, as the EMA rightly points out.

180    Thirty-two hospitals in nine Member States – Belgium, the Czech Republic, France, Germany, Hungary, Italy, Poland, Spain and the Netherlands – took part in that drug utilisation study.

181    That sample, composed of those nine Member States, was proposed by the applicants themselves, as is apparent from the study protocol. The scientific conclusions on which the contested decision is based highlight that the PRAC examined the representativeness of that sample, that it considered that it represented a widespread distribution in the European Union and that the representativeness of the sample was confirmed by the fact that the exclusion of the site with the largest number of patients, in a post hoc sensitivity analysis, had no relevant effect on the overall results. The PRAC concluded that the drug utilisation study evaluated in 2022 was representative of its main clinical use in the European Union.

182    The mere fact that the rates of non-adherence to the product information, including contraindications, varied significantly between different sites in different Member States is not capable of demonstrating that the sample was not representative or that a larger sample should have been chosen.

183    Since the applicants did not raise any other arguments in that regard, there is no reason to call into question the conclusion, set out in the scientific conclusions on which the contested decision is based, that the sample taken into account in the context of the drug utilisation study evaluated in 2022 was representative.

(ii) The absence of effective RMMs

184    As regards adherence to the RMMs put in place and the absence of effective RMMs to address the risks identified, it is apparent from recital 2 of the contested decision that the PRAC concluded that non-adherence to the product information continued to exist, despite the significant additional RMMs implemented following the referral procedure completed in 2018. Furthermore, it appears from recital 3 of the contested decision that it also concluded that, at the time of adoption of the contested decision, there were no RMMs capable of effectively improving that situation. That finding concerns both the RMMs in place and the additional RMMs proposed.

185    In particular, the PRAC considered several additional proposed RMMs, namely a modification of the product information and controlled access programme to supply HES-containing products only to accredited hospitals and healthcare professionals trained in the safe use of the products, a letter of commitment, revised training material, annual recertification and mandatory post-training tests for healthcare professionals, annual re-certification of hospitals, more communication from the holders of MAs to healthcare professionals, and delivery of the medicinal products concerned in four Member States restricted to hospitals with a non-adherence rate of less than 20% on the basis of information on the indication and contraindications to be entered in a database for each patient treated.

186    In that regard, the PRAC noted that information obtained during the present referral procedure indicated that non-adherence to the product information was not only due to a lack of awareness of RMMs, which in its view probably rendered ineffective the implementation of some proposed RMMs, such as communication and training. The PRAC also considered that increased complexity could even have the opposite result, namely less adherence to product information. In the PRAC’s view, due to the expected further reduction in the number of accredited sites and the limited interest of sites in participating in drug utilisation studies, conducting another study to measure adherence to the proposed additional revised RMMs might not provide meaningful results, making it impossible to measure whether future patients would be treated in accordance with what is stated in the product information.

187    None of the applicants’ arguments is capable of demonstrating a manifest error of assessment by the Commission.

188    In the first place, it is necessary to reject, for the reasons set out in paragraphs 79 to 105 above, the applicants’ argument that the Commission should not have taken into consideration the non-adherence to the RMMs in place at the time the contested decision was adopted, since they were not responsible for it and EU law was not applicable.

189    In the second place, as regards the applicants’ argument that the effectiveness of the RMMs in place at the time of adoption of the contested decision is confirmed by the fact that the total number of units of medicinal products sold had decreased significantly, it must be noted that it is apparent from the scientific conclusions on which the contested decision is based that the PRAC took account of the decrease in sales of the medicinal products concerned in the European Economic Area. In that regard, the PRAC considered that it was apparent from the sales figures that, despite that decrease, a considerable number of patients were still exposed to the medicinal products concerned, also taking into account the fact that the reduced dosage was generally adhered to and, consequently, that the quantities used per patient were reduced. According to the PRAC, that fact gave rise to concern in view of the high rates of non-adherence to key restrictions, such as the indication and contraindications. The applicants do not put forward any arguments that call those considerations into question.

190    In the third place, it is true that the case file shows that the last drug utilisation study showed a reduction, described as ‘considerable’ in that study, in the rates of non-adherence to the product information and, in particular, the rate of non-adherence to contraindications. Nevertheless, it is clear from the scientific conclusions on which the contested decision is based that, in particular, the rate of non-adherence to contraindications was still considered to be too high.

191    In that regard, the scientific conclusions emphasise that the level of non-adherence should be proportional to the established risks. Given the seriousness of the harm demonstrated for patients with sepsis, renal failure or who are critically ill, the mere fact that the rate of non-adherence to contraindications has evolved positively is not sufficient to render implausible the assessment that the significant proportion of use of HES solutions for infusions in those populations still raised significant public health concerns.

192    Having regard to the considerations set out in paragraph 116 et seq. above, and in particular the importance of the objective pursued, namely the safeguarding of public health, and the seriousness of the risks established, which include a significant risk of death, that assessment is not manifestly erroneous. Moreover, the EU Courts are not entitled to call into question such an assessment by themselves establishing a rate of non-adherence to the contraindications which, in their view, would still be acceptable in the light of the risks to public health.

193    In the fourth place, with regard to the proposed additional RMMs, according to the first subparagraph of Article 104(2) of Directive 2001/83, as amended by Directive 2010/84, MA holders are to use the pharmacovigilance system to evaluate all information scientifically, consider options for risk minimisation and prevention and take appropriate measures, as necessary.

194    As a result, it was incumbent on the applicants to propose additional effective RMMs in order to respond to the risks identified and, in particular, to the risks to public health resulting from the use of the medicinal products concerned for patients with contraindications.

195    The applicants do not raise any argument capable of calling into question the consideration that non-adherence to the product information was not due to a lack of knowledge on the part of healthcare professionals and, consequently, the conclusion that additional RMMs consisting of more training would probably not be effective (see paragraph 186 above).

196    Moreover, as regards more specifically the RMM proposed by the applicants in an oral explanation of 7 February 2022, namely that of supplying the medicinal products concerned in four Member States only to hospitals with a non-adherence rate of less than 20% on the basis of information relating to the indication and contraindications to be entered in a database concerning each patient treated, it follows from the scientific conclusions on which the contested decision is based that the MA holders had not submitted an assessment of its technical feasibility. According to those conclusions, the PRAC retained its doubts as to the potential impact of that RMM on risk reduction, its feasibility and the fact that that RMM would result in an additional administrative burden in the circumstances of clinical practice in which decisions must be taken quickly.

197    In view of the fact that it was incumbent on the applicants to propose additional RMMs likely to be effective in order to address the risks identified, it was without committing a manifest error of assessment that the Commission adopted the contested decision without awaiting, once again, the results of a further attempt to implement additional RMMs based on an approach similar to that of the RMMs already in place.

198    Consequently, the second part of the first plea in law and, therefore, the first and second pleas in law are rejected in their entirety.

3.      The third plea in law, alleging breach of the principle of proportionality

199    The applicants claim that the contested decision infringes the principle of proportionality.

200    In the first place, they argue that suspending the MAs for the medicinal products concerned does not contribute to patient safety, but has the opposite effect. Alternative medicinal products do not offer a better risk-benefit balance in the light of the scientific data available. Guidelines for the use of HES-containing medicinal products, such as that provided in Appendix A.19 of the application, provide guidance on the safe use of those medicinal products and confirm their therapeutic value. HES is safe and has advantages over alternative crystalloids and synthetic colloids, which have been the subject of fewer studies than HES and are unlikely to be better alternatives.

201    In the second place, the applicants claim that the contested decision is disproportionate in so far as the failure to comply with the terms of the MAs could be sufficiently and effectively remedied by adopting additional RMMs, such as those they proposed in the course of the proceedings. Those measures were designed to dissuade healthcare professionals from the off-label use of the medicinal products concerned and to ensure that any healthcare establishment that did not comply with the RMMs would no longer be supplied with them. The argument that non-adherence to the restrictions was the result of a deliberate choice on the part of healthcare professionals could not be used to reject the adoption of such measures, since any off-label use was the responsibility of those professionals. The applicants point out that, in the present case, off-label use in contraindications has been significantly reduced as a result of direct communication to healthcare professionals, training and monitoring.

202    In the third place, the applicants submit that the suspension is disproportionate and inappropriate in that it rendered obsolete the PHOENICS and TETHYS studies which were in progress. In their view, it was appropriate to wait, in the absence of reports of adverse safety signals, for the results of those studies, which were intended to bring to light new elements of assessment concerning the safety and effective use of the product.

203    In the fourth place, the applicants claim that the suspension ordered at EU level is disproportionate in so far as, according to the PRAC’s own conclusions, there is no off-label use at EU level for patients suffering from sepsis or who are critically ill. Non-compliance observed in one Member State could not be extrapolated to other Member States, as the findings in the nine selected Member States differ significantly.

204    The Commission, supported by Ireland and the EMA, contends that the third plea in law should be rejected.

205    The principle of proportionality is recognised by settled case-law as one of the general principles of EU law. By virtue of that principle, acts of the EU institutions must not exceed the limits of what is appropriate and necessary in order to achieve the objectives legitimately pursued by the measure in question, it being understood that, 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 (judgments of 13 November 1990, Fedesa and Others, C‑331/88, EU:C:1990:391, paragraph 13; of 5 May 1998, United Kingdom v Commission, C‑180/96, EU:C:1998:192, paragraph 96; and of 11 December 2014, PP Nature-Balance Lizenz v Commission, T‑189/13, not published, EU:T:2014:1056, paragraph 111).

206    As regards judicial review of those conditions in the present case, it should be borne in mind that, according to well-established case-law, decisions concerning the application of criteria relating to the efficacy, safety and quality of a medicinal product are the result of complex medical and pharmacological assessments which are subject to limited judicial review. Where an EU institution is called upon to carry out complex assessments, it has a wide discretion, the exercise of which is subject to judicial review limited to verifying whether the measure in question is vitiated by a manifest error or misuse of powers or whether the competent authority has manifestly exceeded the limits of its discretion (see, to that effect, judgment of 23 September 2020, BASF v Commission, T‑472/19, not published, EU:T:2020:432, paragraph 109 and the case-law cited).

207    In order to assess whether the principle of proportionality has been observed in the field of public health, account must be taken of the fact that human health and human life rank foremost among the assets and interests protected by the FEU Treaty (see judgment of 19 December 2019, Vanda Pharmaceuticals v Commission, T‑211/18, EU:T:2019:892, paragraph 154 and the case-law cited).

208    The Court has also held that, given the exclusive nature of the criteria of safety, efficacy and quality enshrined in the EU system for harmonising the grant and management of MAs for medicinal products, it is solely in the light of those criteria that the proportionality of a measure suspending or withdrawing an MA must be assessed. It follows that the relevant interests in the context of the review of proportionality are the same as those related to the protection of public health, taken into consideration in the application of the relevant legislation (see judgment of 19 December 2019, Vanda Pharmaceuticals v Commission, T‑211/18, EU:T:2019:892, paragraph 155 and the case-law cited).

(a)    The objective pursued by the contested decision

209    As regards, first of all, the objective of the contested decision, it is apparent from recital 3 thereof that it was adopted by the Commission in order to address significant risks to public health. That objective is consistent with the essential objective of Directive 2001/83, as set out in recital 2 thereof.

210    The applicants’ argument that the contested decision does not contribute to patient safety, given the therapeutic value of the medicinal products concerned and the fact that the risk-benefit balance of alternative medicinal products is no better, does not call that conclusion into question. The suspension of the MAs concerned does contribute to ensuring the protection of public health, in particular in that it prevents patients with certain contraindications from being treated with those medicinal products and incurring serious risks to their health, including a risk of death.

(b)    The existence of less restrictive measures

211    Next, in the light of the arguments put forward by the applicants, it is necessary to examine whether, in the present case, there existed another appropriate but less restrictive measure.

(1)    The adoption of additional RMMs

212    The applicants consider that the risks in question and, more particularly, the risk of non-adherence to the contraindications, could be sufficiently and effectively remedied by the adoption of additional RMMs.

213    First, the first paragraph of Article 116 of Directive 2001/83, as amended, provides for three possibilities for a competent authority when it finds that the risk-benefit balance of a medicinal product is not favourable. In such a situation, the competent authority may suspend, revoke or vary the MA in question. By providing for those three options of differing intensity, the EU legislature has taken account of the principle of proportionality.

214    Thus, the competent authority must have recourse, as a first step, to modification of the MA, which is the least restrictive of the measures provided for in the first paragraph of Article 116 of Directive 2001/83, as amended, provided that such modification constitutes an appropriate measure.

215    Secondly, any measures which are less restrictive than the measure adopted may be taken into consideration in the context of the proportionality test only if they are equally suitable for achieving the objective pursued by the contested measure adopted by the EU institution (see, to that effect, judgments of 14 December 2004, Arnold André, C‑434/02, EU:C:2004:800, paragraph 55; of 14 December 2004, Swedish Match, C‑210/03, EU:C:2004:802, paragraph 56; and of 4 May 2016, Philip Morris Brands and Others, C‑547/14, EU:C:2016:325, paragraph 180).

216    It has already been found, in paragraph 184 et seq. above, that the conclusions relating to the absence of effective additional RMMs are not vitiated by any manifest error of assessment. Consequently, the applicants’ argument that the non-compliance found could be sufficiently and effectively remedied by the adoption of additional RMMs can also not succeed in the context of the plea alleging breach of the principle of proportionality.

(2)    The PHOENICS and TETHYS studies

217    In the present case, Annex III to the contested decision explicitly provides for the possibility for the MA holders of the medicinal products concerned to provide robust evidence demonstrating a favourable risk-benefit balance in a clinically relevant patient population, in order to lift the suspension. As a result, and as noted in paragraph 66 above, if the data from the PHOENICS and TETHYS studies prove likely to favour the risk-benefit balance of the medicinal products concerned, the applicants will be able to submit them with a view to having the suspension lifted. Moreover, it is clear from paragraphs 125 and 126 above that the performance of those studies was a condition of maintaining the MA and that their final results were supposed to be available at the end of 2016. In those circumstances, the Commission cannot be criticised for having failed to observe the principle of proportionality by adopting the contested decision without waiting for the results of those studies to become available. If such a complaint were accepted, it would in fact be sufficient for any MA holder to defer the performance of a safety study that is a condition for keeping a medicinal product on the market in order to have the lack of availability of the results of such a study noted and, where appropriate, accepted.

218    Consequently, the applicants’ arguments that the contested decision is disproportionate and inadequate because it was adopted before the results of the PHOENICS and TETHYS studies were available are unfounded.

(c)    The existence of disadvantages disproportionate to the aims pursued

219    Finally, it is necessary to consider, in the light of the arguments raised by the applicants, whether the contested decision causes disproportionate disadvantages in relation to the aims pursued.

(1)    The suspension of MAs for the medicinal products concerned at EU level

220    With regard to the applicants’ argument that the suspension of the MAs for the medicinal products concerned at EU level is disproportionate inasmuch as there is no off-label use at EU level for patients suffering from sepsis or who are critically ill, the Court finds as follows.

221    As is apparent from the considerations set out in paragraph 179 et seq. above, it was concluded, in the scientific conclusions on which the contested decision is based, without committing a manifest error of assessment, that the sample taken into account in the drug utilisation study evaluated in 2022 was representative of the European Union. That conclusion implies that it was representative of all the Member States, including those in which no site had participated in that study and those in which the data from that study showed considerably lower rates of non-adherence to the product information, in particular contraindications, than in others.

222    The mere fact that, according to that study, the rates of non-adherence to the product information, in particular contraindications, varied according to the different sites in the different Member States that took part in the drug utilisation study evaluated in 2022, did not therefore prevent the Commission from adopting an EU-wide measure.

223    Consequently, the applicants’ argument that the contested decision is disproportionate in so far as the medicinal products concerned are not used throughout the European Union for patients suffering from sepsis or who are critically ill must be rejected.

(2)    The benefits of the medicinal products concerned

224    Nor does the contested decision cause disproportionate disadvantages in relation to the aims pursued as a result of the fact that the suspension of the MAs for the medicinal products concerned also prevents their use for patients who do not have contraindications. It follows from the considerations set out in the context of the examination of the second part of the first plea in law that the contested decision is not vitiated by a manifest error of assessment either as regards the conclusion that the risk-benefit balance of the medicinal products concerned is no longer favourable or as regards an alleged unmet medical need, in the event of suspension of the MAs.

225    Consequently, the applicants’ argument based on the therapeutic value of the medicinal products concerned cannot succeed either in the context of the plea alleging breach of the principle of proportionality.

(3)    The possibility of lifting the suspension

226    The applicants’ argument, raised in the context of the fourth plea in law, that the contested decision de facto orders a revocation of the MAs cannot be accepted in the context of the third plea in law either.

227    As noted in paragraphs 60 and 61 above, that argument is based on the erroneous assumption that the conditions for lifting the suspension laid down in Annex III to the contested decision could never be met. In reality, those conditions are formulated in a broad and open manner in order to allow the applicants to provide all the evidence likely to demonstrate a favourable risk-benefit balance of the medicinal products concerned.

(4)    The possibility of postponing the implementation of the suspension of the MAs

228    Nor does the possibility, provided for in Article 3 of the contested decision, of temporarily postponing the suspension of the MAs concerned demonstrate the existence of an imbalance between the disadvantages associated with the suspension of the MAs and its aims.

229    First, that possibility remains subject to certain conditions aimed in particular at protecting patients and ensuring that RMMs are respected (see paragraph 62 above). Thus, the risks to public health resulting from the off-label use of medicinal products, particularly for patients with contraindications, are minimised as far as possible.

230    Secondly, that possibility is an expression of the principle of proportionality, in so far as it allows a postponement in those Member States where a transitional period is still deemed necessary, having regard to public health considerations specific to the Member State concerned.

231    It follows from all the foregoing that all the arguments raised by the applicants do not show that the contested decision exceeds the limits of what is appropriate and necessary to achieve the objective pursued.

232    Consequently, the third plea in law must be rejected and, therefore, the action as a whole must be dismissed.

IV.    Costs

233    Under Article 134(1) of the Rules of Procedure, the unsuccessful party is to be ordered to pay the costs if they have been applied for in the successful party’s pleadings. Since the applicants have been unsuccessful, they must be ordered to bear their own costs and to pay those incurred by the Commission, including the costs relating to the interlocutory proceedings, in accordance with the form of order sought by the latter.

234    Under 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. Therefore, Ireland and the EMA are ordered to bear their own costs.

On those grounds,

THE GENERAL COURT (Fifth Chamber, Extended Composition),

hereby:

1.      Dismisses the action;

2.      Orders Fresenius Kabi Austria GmbH and the other applicants listed in the annex to bear their own costs and pay those incurred by the European Commission, including those relating to the interlocutory proceedings;

3.      Orders Ireland and the European Medicines Agency (EMA) to bear their own costs.

Svenningsen Mac Eochaidh Laitenberger

Martín y Pérez de Nanclares Stancu

Delivered in open court in Luxembourg on 15 May 2024.

V. Di Bucci M. van der Woude

Registrar President

Table of contents

I. Background to the dispute

II. Form of orders sought

III. Law

A. The admissibility of the action

B. Submission of the defence

C. The substance

1. The fourth plea in law, alleging failure to state reasons

2. The first plea in law, alleging infringement of Article 116 of Directive 2001/83, as amended, and the second plea in law, alleging infringement of the precautionary principle

(a) The first part of the first plea in law, alleging an error of law resulting from an erroneous interpretation of the concept of ‘risk-benefit balance’ in Article 116 of Directive 2001/83, as amended

(1) The literal interpretation of the concept of ‘risk-benefit balance’ in Article 116 of Directive 2001/83, as amended

(2) The contextual interpretation of the concept of ‘risk-benefit balance’ in Article 116 of Directive 2001/83, as amended

(3) The interpretation of the concept of ‘risk-benefit balance’ taking into account the objective of Article 116 of Directive 2001/83, as amended

(4) The applicants’ other arguments

(b) The second part of the first plea in law, alleging a manifest error of assessment, and the second plea in law, alleging breach of the precautionary principle

(1) Establishing the risks associated with the off-label use of medicinal products

(i) The absence of adverse safety signals

(ii) Reliability of the CHEST study

(iii) The lack of data based on the revised dosage

(iv) PHOENICS and TETHYS studies

(2) The medical need and alternative treatments

(3) The effectiveness of the RMMs

(i) The representativeness of the sample taken into account for the drug utilisation study evaluated in 2022

(ii) The absence of effective RMMs

3. The third plea in law, alleging breach of the principle of proportionality

(a) The objective pursued by the contested decision

(b) The existence of less restrictive measures

(1) The adoption of additional RMMs

(2) The PHOENICS and TETHYS studies

(c) The existence of disadvantages disproportionate to the aims pursued

(1) The suspension of MAs for the medicinal products concerned at EU level

(2) The benefits of the medicinal products concerned

(3) The possibility of lifting the suspension

(4) The possibility of postponing the implementation of the suspension of the MAs

IV. Costs

*      Language of the case: English.

1 The list of other applicants is only attached to the version notified to the parties.