Global spending on health continues to rise, such that it represents an ever increasing and significant percentage of any country’s gross domestic product. From this perspective, investment in the life sciences sector remains attractive. However, the sector also faces some significant downward pressures, including pressure on the pricing of therapies; the presence in many disease areas of established and effective therapies; and sociopolitical pressures (see the recent US congressional hearing before the Senate Finance Committee). There is, therefore, an ever increasing need in the sector to demonstrate the real value of therapies and healthcare services to payers.
In some instances it will be possible to demonstrate a universal increase in the value of a new therapy compared to existing therapies and services, but in many instances the position will be more nuanced. If the particular benefit is limited to a specific class of patients, then payers may only approve its use for this class by adopting a value-based approach. Over time, this type of approach will lead to the ever increasing personalisation of healthcare to the needs of each patient.
In addition to life sciences companies developing and commercialising medicines tailored for specific individuals (so-called precision medicines, such as gene or adaptive cell therapies), any personalisation of healthcare will require further stratification of patients into classes depending on which therapies are likely to work for them. This has two immediate consequences. First, there is need to undertake more diagnostic testing on patients, with any data collected being held in an appropriately structured database, to allow such stratification. Second, payers (and regulators) will need sufficient evidence to support the correlation between the identified class of patient and the additional therapeutic value which that class will obtain by treatment with a particular personalised therapy. Diagnostic testing and the collection, aggregation and analysis of the resultant patient data will only continue to become more important. From an intellectual property perspective, a value-based approach may result in some significant changes and challenges concerning the protection of innovation.
For traditional therapies, whether these are biologics or small molecules, the core intellectual property rights protecting these innovations are patent rights and associated term extensions (in Europe, supplementary protection certificates (SPCs)). While these rights will also remain available for newly developed therapies and diagnostics, new difficulties may arise when considering patentability. For example, patents for personalised therapies that are found to be effective in only certain classes of patients will need to have well-defined claims to ensure that they are not vulnerable to insufficiency attacks, particularly if this limited efficacy only became apparent following clinical trials. For gene therapies, the principal patent may cover the platform technique or process that facilitates the amendment of a defective gene. If the specific therapeutic application for such therapies had not been identified and relevant research undertaken by the time of the patent filing, then this may preclude the option of obtaining an SPC covering the medicinal product ultimately produced. This would result in a shorter period of exclusivity than might be expected for a traditional therapy, such that there may be a shorter period of time to recoup investment.
As a final example from a patenting perspective, if machine learning or artificial intelligence (AI) is used to interrogate a large data set, and this establishes a new therapeutic indication or dosage regimen, then several other considerations arise. First, who is the inventor of that invention and how is ownership determined? Second, when considering patentability issues, such as whether that invention was obvious, is it appropriate to apply the current human standards to determining this question or should a different standard be applied? Sensing the increasing prevalence of AI and its impact on the patenting system, in May 2018 the European Patent Office held, for the first time, a conference (called “Patenting Artificial Intelligence”) to discuss the framework that would enable inventors to obtain patent protection for AI inventions. In keeping with the message of that conference, in June 2018, the heads of the five largest patent offices (those of the USA, Europe, Japan, Korea and China) met and re-emphasised that the impact of AI on the patent system was one of their “main strategic priorities” and confirmed their intention to continue to explore the legal issues surrounding patentability of AI-derived innovation.
Considering next the collection, aggregation and analysis of the resultant patient data, several intellectual property issues arise in this context. Vast quantities of data may be collected by a variety of parties (such as national health authorities, insurers and technology companies, to name but a few) in several ways, including via smartphones and other connected devices. The diversification of data sources is likely to give rise to complex questions around ownership and access, in addition to privacy and data protection issues. Database rights arising from the aggregation of data are likely to offer only limited practical protection since data is not easily protected – traditionally rights protect the database structure or prohibit persistent extraction from a database. Even if data has been directly obtained by a third party using another party’s patented diagnostic process, questions remain as to whether it is possible to prevent the further use of that data as a “product” of the patented process. No legal consensus has yet emerged and there is a dearth of law on this issue.
Given this, the key protection may relate not to the data itself, but to the algorithms developed to interrogate it. Here, again, there are uncertainties concerning the effectiveness of intellectual property rights because in many jurisdictions mathematical formulae have not traditionally been patentable, although software can be protected by copyright. When considering the patentability of an algorithm, if the algorithms in question merely classify data without any indication of a technical use being made of the resulting classification, then there are likely to be patentability issues. Considering the protection afforded to an algorithm by copyright, unless there has been deliberate copying of the original code of the relevant algorithms, then copyright may offer little in the way of protection too. As a consequence, maintaining the confidentiality in complex and proprietary algorithms and treating them as trade secrets may be the most effective way to protect this type of innovation, at least until new legal rights or mechanisms are developed by which they can be protected. Once trade secrets are revealed or confidence is breached, injunctions can be obtained to restrict the use of the data obtained unlawfully, including the sale or use of any infringing products or services that use those trade secrets. However, where the information or data has been lawfully obtained (which includes discovery or access via normal business practices), then its use cannot be restricted. Although possibly mitigated by robust confidentiality agreements and other practical steps, it may be difficult to completely eliminate any risk that confidentiality in the data or algorithm is lost. Increased reliance on confidential information and trade secrets is likely to further increase the amount of litigation seen in the US and elsewhere surrounding restrictive covenants and the movement of key employees (ie, those who were at management level or otherwise had access to confidential, proprietary information, research and development).
Finally, a more tech- and data-driven R&D process is likely to give rise to a need for life sciences companies to acquire new technologies, skills or platforms that they have not historically had any need for, or experience with. Here, the initial strategic question is whether to invest in this area by acquisition, organic growth or by collaboration. Significant thought will need to be given to how any technology acquisition should be integrated into a company’s broader strategic objectives. Given the high demand for skilled employees and relevant technologies, acquisitions may not always be possible (or affordable) and so collaboration with technology partners is also likely to be on the increase. In any such collaboration, appropriate corporate structures or suites of agreements to manage such ongoing relationships will need to be put in place. This will include defining the limits on the use of existing intellectual property and data, and on the way in which intellectual property or data generated in the collaboration should be handled and owned. It will also be necessary to ensure that such limits are consistent with antitrust and competition laws. Clear and well-defined arrangements covering the entire life cycle of any project (or separate arrangements for different stages of the project) will be particularly critical when the collaboration is based on the principles of “open innovation” (which may include sourcing of ideas from the workforce, the public or external-facing innovation invitations/competitions), with ownership of (or access to) any resulting intellectual property more likely to be shared. This would represent a significant shift from the traditional approach of life sciences companies in terms of their handling and ownership of intellectual property. It requires careful consideration of how to balance the free sharing of ideas and resources, while giving certainty on how the fruits of this labour will be owned. Without such clarity, there is a significant risk of disputes, particularly if the collaborating partners come from industries outside life sciences (eg, technology companies) or, have different cultures (eg, corporate, public sector or charity) or approaches to dispute resolution.
Life sciences companies will continue to need to engage with intellectual property issues at a very early stage of the R&D process, whether this is undertaken in-house or through collaboration. As many of the technologies are nascent in their development, the application of the law to these technologies is at a similarly early stage. An effective intellectual property strategy is therefore likely to make use of a range of intellectual property rights covering core technologies. Given the costs of developing such new diagnostic platforms and therapies, it is also likely that litigation concerning the intellectual property will be commenced at an earlier stage, as interested parties will wish to ensure that they have freedom to operate. This would continue the trend for early-stage litigation already seen in relation to biosimilars.
The sector is entering an exciting and uncertain time. Many aspects of the law and relevant regulation will need to be reassessed as the industry explores the opportunities presented by the move towards a more value-based approach and the personalisation of healthcare. Developing, at an early stage, a clear commercial, legal and technological strategy in light of the different legal and regulatory approaches across jurisdictions will be key to success.