In Silico World D9.2 In-depth analysis of legal and ethical requirements

Executive Summary The present Deliverable, ‘D9.2 – In-depth analysis of legal and ethical requirements’, part of Work Package 9 ‘Ethical and Legal Framework’ (WP9) of the In Silico World project, assesses the core pieces of legislation and ethical principles identified in Deliverable ‘D9.1 – Legal and Ethical Inventory’. The report analyses the following areas of legislation – and identifies some key issues – relevant to the In Silico World project and, thus, for in silico trials: Privacy and Data Protection: Data protection is a long-established legislation in the European Union (EU), which has evolved throughout the last decades, and the GDPR is the fundamental EU law to consider for in silico trials. Hence, the report introduces the principles of data processing, the notion of personal data and health data and the legal bases for data processing. In silico trials may highlight some of the common challenges for the healthcare sector in data protection. These include the legal bases and the further processing of health data. In silico trials may also raise interpretative questions. These include doctrinal discussions about the nature of synthetic data in their relationship with anonymization and pseudonymisation, and the concept of data ownership. Data Governance: The legal landscape concerning health data sharing is changing due to a series of new EU legislative initiatives. These initiatives encompass the Data Governance Act, the European Health Data Space (EHDS) proposal, and the Data Act proposal. The Data Governance Act sets rules for re-using certain categories of personal data and introduces the concept of data altruism in healthcare. The EHDS proposal regulates, inter alia, the primary and secondary use of health data. The Data Act proposal proposes business-to-consumer, business-to-business and business-to-government data sharing rules. These three pieces of legislation are expected to apply simultaneously once all are approved. Nevertheless, some challenges may arise in the future, which may concern the appropriate legal basis for data sharing and data altruism, and the interaction of these with the GDPR and national legislation. Clinical Trials, Medicinal Products and Medical Devices: Clinical trials, medicinal products and medical device legislation are relevant to the very essence of in silico trials. For clinical trials and medicinal products, Regulation 726/2004 provides the legal basis for the EMA to deal with novel methodologies for drug development. The Medical Device Regulation (MDR) and In Vitro Medical Device Regulation (IVDR) are the applicable laws for medical device legislation and explicitly mention modelling and simulation. However, while these laws do not prohibit in silico trials, they do not extensively address them either. There exist challenges and barriers that need to be addressed both on a regulatory and legislative perspective. On the regulatory side, guidance and standardisation efforts are needed in general (to tackle, for example, Artificial Intelligence (AI)) and in particular (on the verification and validation of in silico models). On the legislative side, there is common agreement that the current pharmaceutical legal framework lacks behind digital innovation processes. The EU Pharmaceutical Strategy promises the reform of the existing pharmaceutical framework. It is desirable that the reform will address more comprehensively the innovative aspects of the medicinal product’s lifecycle, including in silico trials. Artificial Intelligence: The legal framework of AI is in the process of being established in the EU. The AI Act proposal, expected to be approved in 2023, will introduce new requirements for providers, users and all the actors involved in AI systems. These include, inter alia, adopting a risk management system, a quality management system, documentation duties, ensuring transparency, human oversight, accuracy, robustness, and cybersecurity. The AI Act will be relevant for medical devices and in vitro diagnostic medical devices, as – according to the latest available version of the proposal – they are explicitly included in the scope of the regulation. The regulation sparked several discussions that this report cannot summarise comprehensively. Therefore, the report chooses a new item currently negotiated in the latest proposal’s version may be of crucial relevance for medicine and healthcare: the potential non-application of AI rules in the context of scientific research. The preliminary conclusion is that these rules, as currently formulated, may generate legal uncertainties in the future. The report concludes with a section on the ‘Ethics Principles’. The section is based on the Biomedical Ethics principles that were illustrated in D9.1, i.e. autonomy, justice, beneficence, and non-maleficence. The report offers some examples (patients’ self-determination for autonomy; incidental findings for beneficence; safety and security risks for non-maleficence; patients’ representativeness for justice) to show how these principles could guide stakeholders to ensure the protection and advancement of human values in the context of in silico trials