Biotech Review of the Year: Six months on

Authors: Vik Khurana and Will Hewitt

It has only been six months since we last wrote about how organisations in the life sciences sector can successfully leverage AI but since then, a lot has changed. In addition to the mounting number of use cases and organisations entering this space, we have seen the text of the EU AI Act agreed and on 21 May 2024, the Act passed its final vote by the Council of Ministers, meaning it is set to enter into force in the coming weeks (although most of the key provisions won’t apply as law for a further two years). The Act introduces significant regulatory requirements for AI providers, deployers and distributors, adopting a risk-based approach for the regulation of AI systems. Compliance will be driven by the introduction of official standards, with a focus on data governance, safety, transparency and systematic enforcement of these standards.

What will be a relief for many is the fact that the Act will not regulate all uses of AI systems in the life sciences sector. Instead, it classifies certain practices as either “prohibited” or “high risk”, with most other AI systems permitted without additional restrictions (general purpose AI systems a notable exception). It seems unlikely that an organisation in the life sciences sector will typically be carrying out any uses that will be prohibited under the Act, but there is certainly a range of internal uses and types of systems that will fall into the high risk category. This includes uses in recruitment (e.g. to efficiently screen candidates), in performance evaluation (e.g. using tools that automatically assess employee performance) and in monitoring access controls using biometric identification systems. When looking at the application of the Act to a life science organisation’s products and services, the most obvious high risk use case will be the use of AI systems that also qualify as a medical device (or where an AI system is intended to be used as a safety component of a medical device). Developers of medical devices will therefore have to manage and address the complex interplay between existing medical device regulations and the new AI Act.

For organisations providing, deploying or distributing high risk AI systems, there is a suite of new restrictions to get to grips with. The systems must have an appropriate level of accuracy, robustness and cybersecurity. This is something we had already recommended in our previous article, but this obligation has now become a regulatory obligation. There are also data and data governance obligations, including requirements to analyse and mitigate against bias, record keeping obligations for traceability purposes, obligations to document the AI system’s technical specification and - crucially for users of AI systems - transparency obligations designed to ensure that users can interpret and use AI system outputs appropriately.

Interestingly for life sciences organisations, the Act contains a scientific research exemption which exempts AI systems specifically developed for the sole purpose of scientific research from being subject to these obligations and so we anticipate many organisations in this space looking to rely on this exemption as far as possible. It will be interesting to see where the boundaries lie on what is considered ‘scientific research’ and how closely this will align to the scientific research exemption set out in the GDPR.

We will be unpicking these requirements in more detail over the coming months and no doubt this will be a topic that features in the 12th edition of the Biotech Review later this year.

Original article: “Integrating generative AI into the life science ecosystem – five tips for success” co-authored by Vik Khurana and Will Hewitt. Published on 18 January 2024.

Author: Xisca Borrás

Further to the European Commission’s proposed reform of the EU pharmaceutical law regime, 2024 is witness to the European Parliament’s proposal, which, in many aspects, it is miles apart from the Commission’s proposal.  Importantly, the Parliament proposes to increase most of the incentives for medicinal products compared with the Commission, especially those for unmet medical needs, and is therefore more sensitive to the reality that biotech companies face when they attempt to bring their products onto the market. 

Whereas the Commission proposes to bring down the baseline for regulatory data protection from 8 to 6 years, the Parliament proposes a baseline of 7.5 years, which is welcome news. Both the Commission and the Parliament propose the baseline to be topped up, but the Parliament caps this exclusivity at 8.5 years. What are the different ways to top it up according to the Parliament’s proposal? In an interesting attempt to foster the development of medicines for unmet medical needs, the Parliament proposes an additional year of regulatory data protection; if the R&D of a product takes place in the EU territory in collaboration with a “public entity”, it proposes 6 more months; and, in line with the Commission, it proposes 6 more months for products approved including data of comparative trials. More attuned with the reality of the commercialisation of medicines, the Parliament proposes to get rid of the possibility of extending regulatory data protection by 2 years for launch in all Member States within 2 (or, in some cases, 3) years of the granting of the marketing authorisation. Importantly, the Parliament proposes to leave the “+2(+1)” period of market exclusivity as it currently stands, which is important for predictability. 

While there are some improvements that  Parliament proposes to incentivise the development of medicines for rare diseases, there are also some less-obvious cuts for these products. For “normal” orphans, the Parliament proposes 9 years of orphan market exclusivity, whereas for orphan medicines addressing high unmet medical needs it proposes to extend the exclusivity to 11 years. However, in line with the Commission, it proposes the introduction of the concept of “global orphan marketing authorisation”. Similarly to the concept of “global marketing authorisation”, it would mean that separate periods of orphan market exclusivity would not be available for the same product even if it was approved for an indication for a different orphan condition. Instead, a new indication for a different orphan condition would only extend the orphan marketing exclusivity of the product by an additional year, and this could happen only twice for any given product. 

The Parliament disappoints when it comes to the transferable exclusivity voucher that the Commission had proposed to improve the development of medicines to tackle antimicrobial resistance.  It proposes to dilute the incentive and introduces stricter conditions.  Indeed, the Parliament’s proposal subjects the duration of the exclusivity on the WHO’s priority pathogens list: 1 year for medicines ranked “critical”, 9 months for those ranked “high” and 6 months for those ranked “medium”.

There is still more to come and the future of the EU pharma law regime is far from settled. The European Council’s position will be key before the Commission, the Parliament and the Council can start their interinstitutional negotiations. 

Authors: Gregory Bacon, Charlie French and Jonathan Ross

1 June 2024 marked the anniversary of the UPC opening for business last year.  The 11th edition of Biotech Review reflected on the first 7 months of the new court’s operation, up to 31 December 2023. How have things changed since? 

Participating states

The number of EU member states participating in the UPC remains at 17. However, Romania has now ratified the UPC Agreement and will become the 18th UPC contracting member state on 1 September 2024. Meanwhile, the UPC referendum in Ireland , previously expected to take place in June 2024, has been delayed so the timing of Ireland’s participation in the UPC is currently uncertain. 

Statistics

Since our last report from 31 December 2023:

• The number of opt-outs of classic European Patents (EPs) from the UPC has remained relatively static, standing at 597,840 as of 1 June 2024 (as compared to 579,928 at the end of last year). 
• The number of requests for unitary patents (UPs) has increased from 17,788 to 28,452 by 1 June 2024. Overall, the uptake rate for UPs has gone up from 17.5% to 19.3%.
• The number of UPC cases has also increased. The number of infringement actions, on 1 June, stood at 134 and the number of revocation actions at 39 (up from 67 and 24, respectively).
• The German local divisions continue to be the preferred venue for infringement actions, with Munich taking the lead with 40.3% of all UPC infringement actions, followed by Düsseldorf with 20.1% and Mannheim with 11.9%.
• While the German local divisions are still taking the lead in terms of venue, there has been a notable shift from German to English as the preferred language for infringement proceedings. At the end of 2023, 59% of infringement actions were in German and 30% were in English, whereas the percentages now stand at 44% in German and 50% in English. This reflects more parties opting to file proceedings in English and also a number of cases in which the language of proceedings has switched to English under Article 49(5) of the UPC Agreement and Rule 323 of the Rules of Procedure. In April 2024, the Court of Appeal set out some guiding principles on changing the language of proceedings in Curio v 10x Genomics (UPC_CoA_101/2024), stating that if the outcome of balancing of interests of the parties is equal, the position of the defendant should be the decisive factor.
• The statistics for revocation actions are less informative, since the language of revocation proceedings and division to which they are assigned are pre-determined by the language and subject matter of the patent. In addition, only two of the three eventual seats of the central division are currently operational (Paris and Munich). The third seat of the central division in Milan will open on 26 June 2024.

What do these statistics tell us overall? Usage of the UPC system continues to grow and the German local divisions remain the most popular for infringement actions, although there has been a shift towards English language proceedings.

However, with popularity comes increased case load pressure. In an attempt to address the growing backlog of cases before the German divisions, the Presidium of the UPC recently appointed two additional judges to the Munich local division (Ms Ulrike Voß and Mr Daniel Voß) and one to the Mannheim local division (Mr Dirk Böttcher). The Presidium also increased the working time of existing judges in the Munich and Düsseldorf local divisions, the Munich central division and Paris local and central divisions. Notwithstanding this, it may be that we start to see a shift towards other local divisions by parties looking for more rapid decisions and seeking to rely on the promise of unified decision-making by the new court regardless of the division.

Further decisions

The number of UPC decisions has increased substantially since the end of 2023. Perhaps unsurprisingly for a new court spread across multiple divisions and bringing together judges from numerous jurisdictions, the decision-making has not always been consistent between different divisions. However, we are seeing increased guidance from the Court of Appeal on key topics including the approach to preliminary injunctions (10x Genomics v NanoString (UPC_CoA_335/2023); Vusion Group SA (formerly SES-imagotag SA) v Hanshow (UPC_CoA_1/2024)) and access to documents (Ocado v AutoStore (UPC_CoA_404/2023)).

Please refer to our dedicated UPC microsite for summaries of the key decisions and other updates.

Looking ahead

With the opening of the Milan seat of the Central Division on 26 June, the Court of Appeal finding its feet and the promise of full decisions on the merits later this year, there is clearly a lot to look forward to in 2024!

Original article: “The first seven months of the Unified Patent Court” co-authored by Gregory Bacon, Charlie French and Jonathan Ross. Published on 18 January 2024.

Authors: Adrian Chew and Jonathan Ross

The postscript to our article on the application of G2/21 by the Court of Appeal in the apixaban proceedings noted that the law on plausibility remains far from settled, but that the UK Supreme Court was set to hear the appeal between FibroGen and Akebia in March 2024.

Unfortunately for those hoping for guidance from the UK’s highest court, FibroGen and Akebia settled before the hearing. However, a number of other actions are ongoing where plausibility is a key issue. These include BioNTech & Pfizer v CureVac and the proceedings brought by Teva, Mylan and Glenmark against AstraZeneca in relation to its dapagliflozin (Forxiga®) product. It remains to be seen whether, and if so when, these proceedings progress to the Supreme Court.

Meanwhile, the apixaban proceedings have rumbled on throughout Europe. The Irish and Spanish courts have found the patent invalid at first instance, in contrast to the Swiss court which considered it valid. Judgments are awaited from the Norwegian Court of Appeal and the Dutch Supreme Court.

At the EPO, although the TBA in T 116/18 considered the decision in G2/21 to be akin to ab initio implausibility, other boards have applied an ab initio plausibility standard. 

Six months on, the meaning of G2/21 is no clearer, at both a UK and European level.

Original article: “The application of G2/21 by the Court of Appeal” co-authored by Adrian Chew and Jonathan Ross. Published on 18 January 2024.

Authors: Claire Smith and Erik Müürsepp

Since publishing our article “Strategic considerations for partnering with CDMOs for cell and gene therapies” in the 11th edition of the Bristows Biotech Review, we have continued to see some of the trends which we highlighted in that piece play out in the pharma M&A landscape. As mentioned in the article, pharma and biotech companies face a tricky balance of deciding how much of their cell and gene therapy manufacturing activities to keep in-house vs outsourcing these to a specialist CDMO with expertise in the area. As strategic considerations around a company’s in-house manufacturing capacity for advanced therapeutics change, the most efficient solution may be to enter into talks with existing CDMOs to acquire or divest some specialised manufacturing facilities and the teams working there.

One such deal was recently wrapped up between Ascend Advanced Therapies, an end-to-end gene therapy CDMO, and Beacon Therapeutics, a gene therapy company focused on ophthalmic (eye-related) conditions. Ascend acquired the chemistry, manufacturing and controls (CMC) facility and team of Beacon’s site in Alachua, Florida. The good manufacturing practice (GMP) facility boosts Ascend’s adeno-associated virus (AAV) platform offering to its clients seeking advanced gene therapy development and manufacturing solutions. Crucially, the acquisition also involved establishing a long-term partnership between Ascend and Beacon for Ascend to continue manufacturing Beacon’s products for clinical and commercial use.

This strategic alignment between biotech and CDMO partners demonstrates how specialised manufacturing expertise can be used to its full potential beyond the needs of a single biotech whilst guaranteeing product supply for the biotech and enabling it to focus on clinical development activities.

Original article: “Strategic considerations for partnering with CDMOs for cell and gene therapies” co-authored by Claire Smith and Erik Müürsepp. Published on 18 January 2024.

Authors: Sophie Lawrance and Sean-Paul Brankin

Shortly after we published our piece “Biotech settlement agreements – a competition law checklist”, the European Commission (EC) published its Update on Competition Enforcement in the Pharmaceutical Sector (2018-2022) report (the Report) (available here). The Report emphasises that ensuring access to affordable treatments – including generics and biosimilars – is a competition law enforcement priority for the EC and the other, national competition enforcement agencies across Europe. It includes an entire section on the issues we addressed in our piece – pay-for-delay settlements in general and the Cephalon judgment in particular. As such, the Report is another reminder that the issues potentially raised by patent settlement agreements remain a focus of antitrust enforcement action.

In that regard, it is also worth noting that what should be the final step in the Servier case – the judgment of the EU Court of Justice – is due later this year. That judgment should provide further clarity on the legal rules in this area. If, as seems likely, it upholds the EC infringement decision – including certain aspects of that decision originally overturned by the EU General Court – it may encourage further action by the EC and national competition authorities. Overall, we feel, this may be a good time to invest 10 minutes reviewing our practical summary of the potential pitfalls in this area and how to avoid them.

Original article: “Biotech settlement agreements – a competition law checklist” co-authored by Sophie Lawrance and Sean-Paul Brankin. Published on 18 January 2024.