AUTH/3582/11/21 - AbbVie v Galapagos

Jyseleca (filgotinib) promotional website

  • Received
    15 November 2021
  • Case number
  • Applicable Code year
  • Completed
    09 January 2023
  • No breach Clause(s)
  • Breach Clause(s)
  • Sanctions applied
    Undertaking received
  • Additional sanctions
  • Appeal
    No appeal

Case Summary

AbbVie Ltd complained about claims on the website produced by Galapagos Biotech Ltd relating to Jyseleca (filgotinib) in rheumatoid arthritis (RA).

AbbVie referred to particular claims presented on the Jyseleca website (‘Safety’ section UK-INF-2020-11-0063 and ‘At a glance’ section UK-INF-2020-11-0066). The website stated that it was intended for health professionals in the UK and Ireland and, in AbbVie’s view, the claims at issue were not consistent with the requirements of the Code. AbbVie’s concerns related to the robustness of the safety and sustained efficacy data supporting promotional claims.

AbbVie stated that it engaged in inter-company dialogue with Galapagos to resolve the issues, and had made progress, however, resolution had ultimately been unsuccessful. AbbVie stated that Galapagos had not adhered to the proposed timelines to update or withdraw materials as agreed with AbbVie.

1 ‘Safety’ messaging

Strength of balance website ( – Safety page (UK-INF-2020-11-0063, March 2021, last accessed 15 November 2021)

1.1 General concerns regarding the portrayal of safety data for Jyseleca on the safety page

AbbVie stated that the only adverse data that were shown on the ‘Safety’ tab of the website was what was deemed to be stated as Janus Kinase (JAK) inhibitor associated adverse events, with all other adverse event data, including the most frequently reported adverse events as listed in the Jyseleca SPC not shown at all. Despite the fact that there were RMM [Risk Minimisation Materials] associated with Jyseleca, there was no clear reference to the materials or the safety concerns and information highlighted in those materials which included serious infections, potential effects on male fertility, potential risks of major adverse cardiovascular events, and prescribing in the very elderly (75 years and above), on the safety page.

Other data such as common or serious adverse events were presented on the ‘MOA [Mechanism of Action] & Dosing’ page. However, as this information was in different places on the website, it was important to provide clear direction to health professionals where additional information could be found. AbbVie alleged that the way it was portrayed on the website did not satisfy the requirement that the prescriber should be able to form a reasonable impression of the medicine by reading the information.

The detailed response from Galapagos is given below.

The Panel noted that in its letter of 20 August 2021, Galapagos stated that the website would be undergoing an update following the transition of commercialisation responsibilities from Gilead to Galapagos and whilst there had been no intention to downplay any safety concerns, the safety data detailed on the site would be amalgamated in a single location accessed via the ‘Safety’ section. The Panel further noted that whilst Galapagos did not accept that the absence of a link from the ‘Safety’ page to the ‘Resources’ page constituted a breach of the Code, in the same letter of 20 August 2021 Galapagos agreed that providing a link from the ‘Safety’ section to the risk minimisation materials would increase their visibility and stated that as part of the website update, additional links to RMP materials would be added to the ‘Safety’ section. The Panel noted that Galapagos had committed to complete the changes in November 2021, as outlined in its letter to AbbVie dated 20 October 2021 and acknowledged by AbbVie in its email of 27 October 2021. The Panel therefore considered that at the time the complaint was submitted by AbbVie (15 November 2021), this matter had been settled during inter-company dialogue, and therefore it did not consider the allegations in relation to general concerns regarding the portrayal of safety data for Jyseleca on the safety page as set out in Point 1.1.

1.2 Claim ‘Treatment with JYSELECA was associated with consistently low rates* of JAK inhibitor-associated adverse events up to 52 weeks2’3’.

AbbVie alleged that the claim was not supported by the references to the data in the table below the claim. ‘Low’ was related to adverse events with a frequency <1% and >0.1%. Not all the rates listed in the table fell within this categorization, which was inaccurate. At the foot of the table the explanation for the asterisk with further references was provided which made the interpretation of the table and its data misleading and ambiguous.

The detailed response from Galapagos is given below.

The Panel noted that the asterisk to the claim ‘Treatment with JYSELECA was associated with consistently low rates* of JAK inhibitor-associated adverse events up to 52 weeks2,3’ took the reader to a very small footnote at the bottom of the webpage, beneath the highlighted box which housed the table and a separate footnote, which read ‘Based on AE rates observed as “Uncommon” (<1% and >0.1%) or of lower frequency in the Jyseleca clinical trials’. The Panel noted that the claim at issue was referenced to the Genovese et al poster presented at the EULAR e-congress 2020 and Data on File, whilst the footnote was referenced to the Jyseleca SPC and Genovese et al poster presented at the EULAR e-congress 2020.

The Panel noted Galapagos’ submission that even though the table included some information for adverse events with a higher rate, it did not mean that the information was misleading; the table contained data from individual studies that were included in the integrated safety analysis as stated in the text underneath the table. The Panel noted that beneath the table within the same highlighted box it stated ‘52-week exposure dataset: DARWIN 1 and 2 and FINCH 1-3 studies were integrated to represent the safety of JYSELECA 200mg in controlled clinical studies up to Week 52’. The Panel further noted Galapagos’ submission that the inclusion of these data ensured that the information presented was complete and would enable the recipients to develop their own informed opinion.

The Panel noted that the table would not be viewed in isolation but within the context of the page overall including, particularly, its subheading, ‘Treatment with JYSELECA was associated with consistently low rates* of JAK inhibitor-associated adverse events up to 52 weeks2,3’ which introduced the table and was the claim at issue. The matter was further complicated by the footnote in small print at the bottom of the webpage. It was an established principle of the Code that claims could not be qualified by footnotes if such footnotes were necessary to ensure that a claim complied with the Code. In such circumstances, the footnote should be part of the claim at issue or within its immediate visual field. Whilst noting Galapagos’ submission about the completeness of the data in the table, the Panel considered that some readers, on the balance of probabilities, might nonetheless assume, based on the subheading at issue including the unqualified phrase ‘low rates’, that the adverse events in the table immediately beneath were all of low incidence and that was not necessarily so. In the Panel’s view, this was so, irrespective of the location of the footnote. The Panel considered that the claim in question within the context of the webpage, including the table, was misleading and ambiguous as alleged. A breach of the Code was ruled.

The Panel considered that the implication of the unqualified claim within the context of the webpage was incapable of substantiation and a breach of the Code was ruled.

1.3 Claim ‘Similar observed rates of serious infections, herpes zoster and VTE [Venous thromboembolism] compared to adalimumab2,3’.

AbbVie alleged that with regard to herpes zoster rates compared to adalimumab, the references did not provide sufficient substantiation and might downplay the important safety considerations of herpes zoster rates for rheumatoid arthritis (RA) patients.

Based on the information made available to AbbVie, the company alleged that the claim could not be substantiated by the references provided.

AbbVie noted that the claim was also present on the ‘At a Glance’ page of the website and alleged that it was similarly in breach of the Code.

The detailed response from Galapagos is given below.

The Panel noted Galapagos’ submission that the references accurately portrayed the results of clinical trials and that the publications had undergone peer review and had been subjected to the usual scientific scrutiny. The Panel noted Galapagos’ submission that AbbVie accepted in its complaint letter that the confidence intervals in the stated studies overlapped and that the studies were not powered to compare adverse event differences.

The Panel further noted Galapagos’ submission that the studies showed comparable absolute numbers. Galapagos agreed with AbbVie’s conclusions in this respect and, therefore, in its view, the data from these studies substantiated its use of the term ‘similar’.

The Panel noted that the claim at issue was referenced to Genovese MC, et al. Poster presented virtually at the European League Against Rheumatism (EULAR) 2020 E-Congress, June 3-6, 2020 and data on file. Figure 2 included herpes zoster: 1.4 for filgotinib 200mg and 0.9 for filgotinib 100mg vs adalimumab (0.7) at 52 weeks. The Panel noted that the discussion section stated ‘With filgotinib, EAIR of serious infections and herpes zoster were generally similar to adalimumab and MTX’. No limitations of the study appeared to be included in the poster. The Panel noted that the Data on File included, inter alia, the herpes zoster data for filgotinib 100mg and 200mg QD with no reference to herpes zoster in relation to adalimumab.

The Panel noted Galapagos’ submission that the statement could also be substantiated by more recent publications such as Winthrop and Alves. The Panel noted that it was possible to substantiate a claim by studies supplemental to those cited.

The Panel noted that in Winthrop which, according to the paper, appeared to be received for publication in May 2021 and accepted in September 2021, Exposure-adjusted incidence/event rates for herpes zoster were determined in patients receiving UPA [upadacitinib] (monotherapy or combination therapy) in six randomised phase III trials (data cut- off on 30 June 2020). Herpes Zoster incidence and event rates were also determined in patients receiving MTX monotherapy or adalimumab (ADA) + MTX. Multivariable Cox regression analysis was used to identify herpes zoster risk factors in UPA- treated patients. The Panel noted that filgotinib was not included within the study, upadacitinib was included and appeared to be an AbbVie medicine. The study authors stated that given the lack of direct comparison of head-to-head studies between JAK inhibitors, they were limited in drawing conclusions regarding the relative risk of herpes zoster with UPA as compared with other JAK inhibitors, including filgotinib.

The Panel noted that Alves, a systematic review and network meta-analysis, stated that, compared with filgotinib, adalimumab had an increased risk of herpes zoster infection (4.81; 95% CI, 1.39-16.66). It further stated, however, that although the initial results suggested that filgotinib could have a reduced risk of herpes zoster, the sensitivity analyses did not support those findings. Risk differences between the drugs became statistically non-significant when the sensitivity analysis was conducted.

Whilst the Panel had concerns about the page as a whole, it noted that the allegations appeared to be limited by AbbVie to substantiation of the claim in question by specific studies rather than the broader issues that potentially came within the clauses cited by AbbVie and the matter was considered on that basis.

The Panel considered that whilst neither the data on file the claim was referenced to, or Winthrop, substantiated the claim in question, Genovese et al 2020, the abstract the claim was referenced to and Alves could, on the balance of probabilities, substantiate the claim with regard to similar observed rates of herpes zoster compared to adalimumab. The Panel therefore, on this very narrow ground, ruled no breach of the Code in relation to the claim at issue and substantiation on the safety page and the ‘at a glance’ page. Noting it’s no breach ruling in relation to substantiation, the Panel did not consider that the claim was misleading based on the very narrow allegation that it could not be substantiated, no breach of the Code was ruled.

1.4 Safety and high standards

AbbVie alleged that the issues in Points 1.1, 1.2 and 1.3 raised serious concerns about the balance of safety and efficacy information portrayed regarding Jyseleca in Galapagos’ promotional materials.

AbbVie took patient safety extremely seriously and alleged that the portrayal of the safety profile of Jyseleca was insufficient to enable health professionals to appropriately evaluate the risk-benefit profile of prescribing Jyseleca. The information provided by Galapagos was misleading, incomplete and downplayed the safety concerns associated with the product. In addition to the above mentioned aspects, AbbVie alleged that high standards in preparing promotional materials had not been maintained.

The detailed response from Galapagos is given below.

The Panel noted that it had made no ruling about the matter raised at Point 1.1 above and which was therefore not relevant to the consideration of this point.

The Panel noted its comments and rulings above at Point 1.2, and its concerns regarding the safety page. It considered that, overall, high standards had not been maintained in this regard and a breach of the Code was ruled.

2 Sustained efficacy messaging

Strength of balance website ( – At a Glance page (UK-INF-2020-11-0066, March 2021, last accessed 15 November 2021)

The ‘At a glance’ page was headed ‘SUSTAINED EFFICACY’ followed by ‘In Phase 3 trials, Jyseleca demonstrated’:

ACR20 response as early as Week 2 in 37% of patients (n=475) vs. 15% of patients in the MTX+ placebo group (n=475; p<0.001)2.

ACR70 response in 44% of patients by Week 52 (n=475)2.

DAS28-CRP <2.6 remission sustained up to Week 52 in 54% of patients (n=475)3.

Zero radiographic progression* at Week 52 in 88% of patients (n=475) vs. 82.4% for adalimumab patients (n=325)4.

AbbVie alleged that the information provided fell short of the relevant Code requirements for a number of reasons:

- Firstly, there was a lack of explanation regarding the study description, primary endpoint, and it was not clearly stated that these data were for the filgotinib 200mg dose only.
- Secondly, it had not consistently been made clear what the comparator arm was (which was missing for the 2nd and 3rd bullet points).
- Thirdly, there were no statistical methods and p-values provided for the 2nd, 3rd or 4th bullet points.

In this way, it was not possible ‘At-a-glance’ to determine if the percentages quoted were simply a statement of numerical value, or whether in fact, they had statistical significance and potential clinical benefit to the patient. Because there was a woeful lack of context to these claims, even allowing for the fact that they appeared on a ‘summary’ page, AbbVie believed this was a clear breach of the Code.

The detailed response from Galapagos is given below.

The Panel noted that the four bullet points on the ‘At a glance’ webpage were referenced to three separate data on file. The Panel noted that it did not have these data on file before it.

The Panel noted Galapagos’ submission that the details of the study were provided in detail on the ‘Efficacy’ pages of the website. According to Galapagos, the bullet points beneath the headline claim were accompanied by a statement that clearly referred to the efficacy measures used, ie ACR20 and ACR70 and the ‘Efficacy’ page provided additional information, with tabs with clear graphs showing the evolution of response at various timepoints in FINCH 1, FINCH 2 and FINCH 3. Additional information about the studies was available via a box beneath the graphs entitled ‘Study Details’. These ‘Efficacy’ pages also showed the timepoints at which the response was observed, ie Week 2, and over what period it was observed, ie 52 weeks. Radiographic progression was measured after a specified period, in the case of FINCH 3: 52 weeks, against a baseline examination. The Panel noted that it did not have access to the linked information via the ‘Study Details’ link.

Galapagos also stated that the claim was substantiated by the EPAR which stated that the absolute number of responders in the filgotinib 200mg-group increased from week 12 to week 24 and did not decrease from week 24 to week 52. Galapagos submitted that since the claim was not comparative, it was not necessary to state the comparator. The claims related to Jyseleca, not its relative performance. The Panel further noted Galapagos’ submission that with respect to the lack of statistical method and p values, these were not a requirement of the Code. The statements were factually correct and were substantiated by the additional information on the linked pages. Therefore, the website did not mislead the reader.

The Panel noted Galapagos’ reference to qualifying information elsewhere on the website and considered that each webpage should not be misleading if read in isolation. It was, of course, acceptable to have a page summarising data presented elsewhere but companies had to be particularly careful that such summary pages, nonetheless, complied with the Code and did not rely on data elsewhere to ensure Code compliance.

The Panel noted that whilst the first and last bullet points were comparative, the middle two were not and noted Galapagos’ submission in that regard. Nonetheless, the Panel noted that given the first and fourth bullet points were comparative, some readers might consider the second and third bullet points in that light and assume that the data was favourable to Jyseleca and that was not necessarily so in relation to the 52 week data referred to. The Panel noted Galapagos’ submission that specifying the dose was not required as 200mg was the licensed dose and the 100mg dose was only appropriate for specialist populations which was explained in detail on the ‘Mechanism of Action & Dosing’ page. In this regard, the Panel noted that within the website, the dose was not stated at the top of pages which discussed clinical data and considered that it was relevant to the clinical outcomes and in this regard, considered that it would be helpful and relevant to state the dose in relation to the claims in question. The Panel noted that Jyseleca was indicated in combination and in monotherapy and in this regard, particularly given that the efficacy section discussed certain monotherapy data, considered that the claims in question, on a page that aimed to summarise the efficacy section, should be clear whether they related to combination or monotherapy treatment. In relation to the bullet point ‘Zero radiographic progression* at Week 52 in 88% of patients (n=475) vs. 82.4% for adalimumab patients (n=325)’, the relevant section within the efficacy page, Radiographic progression, stated as a footnote that 52 week data were not controlled for multiplicity, therefore treatment differences could represent chance findings. In the Panel’s view, readers would assume that the bullet point on the page in question referred to statistically significant data at 52 weeks unless they were told clearly otherwise and that was not so. The Panel did not have a copy of the relevant studies but on the basis of the limited information on the website considered that the page in question on the ‘At a glance webpage’ was incomplete and misleading such that readers did not have sufficient information to form their own therapeutic value of the medicine and a breach of the Code was ruled. The Panel considered that high standards had not been maintained in this regard and a breach of the Code was ruled.


In summary, AbbVie alleged breaches of several clauses of the Code. Furthermore, due to the nature of the breaches, ie primarily focused around patient safety, combined with Galapagos not adhering to the agreements and spirit of the inter-company dialogue, AbbVie took the view that the PMCPA should also consider a breach of Clause 2.

The detailed response from Galapagos is given below.

The Panel noted the particular circumstances of this case and the evidence before it. The Panel did not consider that AbbVie had established, on balance of probabilities, that Galapagos did not adhere to the spirit of inter-company dialogue, as alleged.

The Panel further noted its comments and rulings above, including its rulings of a breach of the Code to reflect that high standards had not been maintained. The Panel noted that despite its concerns detailed at Point 1.2 above, it noted that the adverse event rates were given in the table in question at Point 1.2. The Panel considered that its ruling of a breach for failing to maintain high standards adequately covered its concerns on this point. On balance, the Panel did not consider that the complainant had established that patient safety had been prejudiced as referred to in the supplementary information to Clause 2 of the Code as alleged. In the particular circumstances of this case, the Panel therefore, and on balance, ruled no breach of Clause 2 which was used as a sign of particular censure and reserved for such use.