Implementation of precision medicine in healthcare—A European perspective

The technical development of high‐throughput sequencing technologies and the parallel development of targeted therapies in the last decade have enabled a transition from traditional medicine to personalized treatment and care. In this way, by using comprehensive genomic testing, more effective treatments with fewer side effects are provided to each patient—that is, precision or personalized medicine (PM). In several European countries—such as in England, France, Denmark, and Spain—the governments have adopted national strategies and taken “top‐down” decisions to invest in national infrastructure for PM. In other countries—such as Sweden, Germany, and Italy with regionally organized healthcare systems—the profession has instead taken “bottom‐up” initiatives to build competence networks and infrastructure to enable equal access to PM. In this review, we summarize key learnings at the European level on the implementation process to establish sustainable governance and organization for PM at the regional, national, and EU/international levels. We also discuss critical ethical and legal aspects of implementing PM, and the importance of access to real‐world data and performing clinical trials for evidence generation, as well as the need for improved reimbursement models, increased cross‐disciplinary education and patient involvement. In summary, PM represents a paradigm shift, and modernization of healthcare and all relevant stakeholders—that is, healthcare, academia, policymakers, industry, and patients—must be involved in this system transformation to create a sustainable, non‐siloed ecosystem for precision healthcare that benefits our patients and society at large.

The technical development of high-throughput sequencing technologies and the parallel development of targeted therapies in the last decade have enabled a transition from traditional medicine to personalized treatment and care. In this way, by using comprehensive genomic testing, more effective treatments with fewer side effects are provided to each patient-that is, precision or personalized medicine (PM). In several European countries-such as in England, France, Denmark, and Spain-the governments have adopted national strategies and taken "top-down" decisions to invest in national infrastructure for PM. In other countries-such as Sweden, Germany, and Italy with regionally organized healthcare systems-the profession has instead taken "bottom-up" initiatives to build competence networks and infrastructure to enable equal access to PM. In this review, we summarize key learnings at the European level on the implementation process to establish sustainable governance and organization for PM at the regional, national, and EU/international levels. We also discuss critical ethical and legal aspects of implementing PM, and the importance

Introduction
Over the last two decades, there has been significant progress in developing and implementing the concept of personalized or precision medicine (PM). In this time period, we have witnessed a rapid development of high-throughput genomic technologies, which were first applied in research to unravel disease mechanisms and subsequently in the clinical setting as powerful diagnostic tools [1]. In parallel, new targeted therapies have been developed targeting key pathomechanism(s), particularly in cancer and rare diseases [2]. This development has paved the way for improved diagnostics and more personalized treatment and care, comprising a true paradigm shift from "one-size-fitsall" to precision healthcare [3]. Initially, the main push forward came from research in rare inherited diseases and cancer but now encompasses other large disease areas such as infectious and complex diseases [4]. In more recent years, continuous development of other high-throughput molecular and imaging techniques has added important layers of information that will significantly improve precision diagnostics and care [5,6].
Along this path, PM has developed from being a research endeavor to becoming a clinical concept. The continuous flow of basic biomedical research, drug discovery, clinical application, registry data, and improved patient outcomes has made it clear that PM deserves a central place in healthcare. As a result, PM is today recognized as an indispensable and important part of our future healthcare. This has also resulted in an important change in strategy, from PM being a scientifically driven "bottom-up" development pushing its way forward to also becoming a "top-down" approach, in which decision-makers and healthcare systems are ensuring that healthcare enables PM-based approaches.
In this review, we will discuss key aspects of the implementation process for PM, including strategies for sustainable governance and organization of PM infrastructure at the regional or national level, which incorporate all relevant stakeholders, enable a continuous research-implementation feedbackloop, and at the same time secure equal access to precision healthcare for all citizens. We will highlight the need for innovative PM trial concepts and access to real-world data (RWD) for evidence generation, as well as systematic healtheconomy studies as a basis for sustainable reimbursement models. We will also discuss ethical and legal aspects of implementing PM, the need of new models for cross-disciplinary education and training, improved patient and next of kin involvement and activities raising awareness for society at large. Despite a focus on the European perspective, most aspects discussed are equally relevant beyond Europe, and hopefully, the sharing of our experience will contribute to the acceleration of PM implementation across borders and internationally. Centers for Personalized Medicine (ZPM) at four university hospitals (Fig. 1). It focuses on harmonizing diagnostics, decision-making in molecular tumor boards (MTBs), and equal access to drugs, within one reimbursement framework as well as one common data exchange system [7]. Financed by the Federal Joint Committee (Gemeinsamer Bundesausschuss, GBA), the network has now been extended to 26 German university hospitals to enable a nationwide new healthcare structure consisting of ZPMs certified by the German Cancer Society (DKG). At the same time, the national government funded the genomDE initiative that aims at providing a platform to promote genomic medicine in oncology and rare inherited diseases. The merge of bottom-up and top-down approaches is also visible in the national legislation and in the construction of a national data platform in PM. Starting in January 2024, German legislation will allow large-scale sequencing of patients with progressed oncological diseases within DNPM centers as part of the genomDE initiative. Moreover, the IT infrastructure that was initially designed in Baden-Württemberg will be rolled out to become an important part of the national data platform.
Similarly, owing to the regionalized healthcare system in Sweden, Genomic Medicine Sweden (GMS) was initially a bottom-up initiative (Fig. 1), starting in 2017 and gathering a multiprofessional workforce of diagnosticians, clinicians, researchers, and informaticians to coordinate the implementation of genomic-based PM into Swedish healthcare. A key success factor has been the close collaboration between academia and healthcare, work-ing together with patient organizations, industry, and governmental agencies. GMS has established regional Genomic Medicine Centers at all university hospitals across Sweden, stimulating the uptake of genomic-based diagnostics, developed by GMS together with the Swedish research infrastructure Science for Life Laboratory (SciLifeLab), and supporting equitable access to genomic-based PM across Sweden for patients with rare diseases, cancer, infectious diseases, and complex diseases [9]. A National Genomic Platform has been established that will allow for data storage and sharing of data both nationally and internationally [7,9]. GMS is mainly funded by the government, the universities with medical faculty, and the university healthcare regions. The bottom-up work of GMS has increased awareness about the value of PM at the governmental level, pushing, for example, governmental investigations in secondary use of health data (finalized in September, 2023) and dialogues on national strategy and infrastructure for PM, thereby bridging the bottom-up with a topdown approach. These examples show how localized initiatives, which can rapidly introduce organizational changes, can become a role model for larger or even national initiatives.
A key example of a top-down approach is the NHS Genomic Medicine Service in England (Fig. 1). To support equitable access to genomics across the NHS in England and provide standardized care across the population, NHS England launched the NHS Genomic Medicine Service in 2018, building upon the expertise and results from the 100,000 Genomes Project [10]. In October 2022, NHS England published the first NHS genomics strategy (accelerating genomic medicine in the NHS) to outline the vision for embedding genomics in the NHS over the next 5 years through four priority areas to (1) embed genomics in the NHS through a world-leading innovative service model; (2) deliver equitable and innovative genomic testing for improved outcomes in cancer, rare, inherited, and common diseases and to enable precision treatments and interventions; (3) enable genomics to be at the forefront of the data and digital revolution; and (4) evolve the service through cutting-edge science, research, and innovation [11]. Major national infrastructures were introduced in 2018 to support the implementation of the NHS genomics strategy ( Fig. 1), including Genomic Laboratory Hubs to perform genomic testing; the National genomic test directory, which outlines the range of genomic tests that are available as part of the funded NHS clinical service; integrated clinical genomics services, which provide diagnostic and genetic counseling; the national whole-genome sequencing (WGS) provision delivered in conjunction with Genomics England, which offers WGS as part of routine care; and the National Genomic Research Library, which has been developed in partnership between NHS England and Genomics England to support research and discovery from approved researchers, academia, and industry, in which individual patients and family members have provided their informed consent for their data to be used.
In Denmark, the Danish government and the five healthcare regions published the first strategy for personalized medicine in 2016 [12], followed by an update in 2020. The aim of the strategy was to build and implement a national infrastructure for PM in the Danish healthcare system to enable equal and standardized care across Denmark. In this way, the Danish National Genome Center (DNGC) was established under the Danish Ministry of Health with the responsibility of developing the national infrastructure and to host a national genome database, both in close collaboration with a broad range of relevant stakeholders in the healthcare system and research community. Until now, 17 patient groups within rare diseases and cancer, with 88 underlying clinical indications, have been selected to be offered WGS as a first-line test in Denmark. The patient groups have been selected through a thorough process involving clinicians, Danish medical societies, and other stakeholders. Two clinical sequencing centers provide the WGS data, and specialists from all 17 patient groups have regular, structured multidisciplinary meetings. Furthermore, to support the strategy, a new legislation concerning reporting and storing of genetic data was passed in 2018. The legislation states that all genetic data deriving from advanced genetic analysis performed after 1 July 2019 must be reported to the Danish national genome database hosted by the DNGC. The genetic data stored in the national genome database can be accessed by healthcare professionals and reused if necessary for treating the individual patient. Additionally, the genetic data can be used for secondary purposes in future research projects. Because data might be used for future research, the legislation concerning mandatory reporting of genetic data states that genetic data can only be reported if the patient or research subject has been informed about the mandatory reporting and the possibility of opting out of their data being used in any future research.
In France, the French Genomic Medicine Initiative (Plan France Médecine Génomique 2025-PFMG2025) was launched in 2016 by the prime minister in order to integrate genomic medicine into the French healthcare system (Fig. 1) [8]. Representatives from different health and research institutions were brought together to define the PFMG2025 organizational framework. This was first introduced for patients with cancer and rare diseases and will be expanded to cover common diseases, according to medical and scientific advances. Seventy clinical indications have been selected through successive calls for applications for health professionals. Up to now, two PFMG2025 clinical sequencing laboratories provide comprehensive molecular tests for all patients in France. A generic genomic care pathway has been defined, with multidisciplinary meetings upstream and downstream of genomic testing. Then, it was adapted to the characteristics of each pre-indication by the health professionals concerned, in particular by setting up specific multidisciplinary meetings and organizing the samples circuits. Beyond its primary health vocation, PFMG2025 is part of a continuum between care and research, and secondary use of data for research projects is one of its major objectives. Clinical and genomic data will be accessible through a dedicated national IT and data infrastructure under construction, the central data analyzer (Collecteur Analyseur de Données-CAD).
In Italy, one of the major limits for the implementation of a National Genomic and PM initiative is represented by the regional organization of the health system that leads to significant difference and disparities to biomarker and PM access. In 2019, the Italian National Agency for Health Service (AGENAS) released guidelines for the organization of regional oncological networks aimed to improve the standards of cancer care. These guidelines identified for the first time the need to organize referral molecular biology laboratories connected with the regional oncology networks. Scientific societies and health institutions collaborated in preparing recommendations for the appropriate use of NGS in the tumor gene profiling and for the access to tests and targeted drugs [13]. Nevertheless, recent data have shown an extremely low usage of NGS in routine clinical practice in Italy, with only 2% of cancer patients who required a biomarker test being analyzed with this technology [14]. Furthermore, access to new drugs is limited in Italy, significantly reducing the possibility of patients undergoing comprehensive genomic profiling to receive a treatment matched to genomic alterations [15]. Recently, some legislative initiatives have been taken to build a path for the implementation of precision oncology. The Italian regions have been invited to identify reference centers for carrying out NGS tests, capable of analyzing at least 5000 samples per year. These centers will cover all NGS testing needs, including rare diseases, cancer, infectious diseases, and complex diseases. Furthermore, a law established the formation of MTB in each region, associated with NGS laboratories capable of performing comprehensive genomic profiling tests. Finally, two decrees have identified specific budgets to support the use of NGS for genomic profiling of lung cancer and cholangiocarcinoma. Alleanza Contro il Cancro (ACC), an association in which all the major Italian cancer centers are members, has promoted some genomic profiling studies on a national scale. For example, the GERSOM study is evaluating somatic and germline mutations in selected patients with ovarian, breast, and colon cancers. However, there is no ongoing large-scale national initiative for the introduction of more complex technologies such as WGS or WES in the clinical management of patients aimed at promoting PM in oncology.
The Spanish Government approved in 2020 the call for the precision medicine infrastructure associated with Science and Technology (IMPaCT) pro-gram of the Strategic Health Action 2017-2020. The initiative has been a first step in the implementation of PM in the national health system through a strategy based on science and innovation [16].
IMPaCT consists of three programs: predictive medicine, data science, and genomic medicine. The three programs, closely related to each other, constitute an infrastructure that seeks to generate knowledge and scientific-technical bases to support the deployment of the national PM strategy in the national health system. The major aim of the predictive medicine program is to build up a large population cohort with clinical, epidemiological, and biological data, including ethnic variability and geographic and environmental diversity in order to compute predictive models of disease, identify health inequalities, monitor key indicators, and evaluate the impact of health policies [17].
The data science program seeks to develop a system for the collection, integration, and analysis of clinical and molecular data aimed at improving the health of each individual patient, and that allows the secondary use of existing information focusing on public health. Genomic medicine will develop infrastructures and coordination protocols to carry out genomic analysis and other "omics" data in an effective, efficient, and accessible manner across the country and paying special attention to early molecular diagnosis of rare diseases [17].
Ingenio is another national project that seeks to boost precision oncology through the implementation of predictive biomarkers as well as intelligence artificial cohorts (clinical and genomic) whose main objective is to develop a platform that integrates data on biomarkers and clinical results in lung cancer [18]. In some regions, programs have been set up to make precision oncology a reality in the hospital environment: in   which ensures that decision-making is aligned with current research and real-world medical and clinical care problems. The examples also demonstrate the willingness to invest in PM, both at the regional and national level. Nevertheless, aligning different areas of PM (e.g., cancer and rare diseases), involving health insurances, harmonizing data protection, and securing long-term funding are only possible if bottom-up approaches are aligned with a government-or health-systemsponsored national strategy in PM.

Strategies and partnerships for PM in Europe
The European Commission (EC) plays a key role in the development and implementation of PM [25,26]. The EC started to deploy focused efforts in the field of PM in 2010 by supporting research projects ranging from the basic understanding of diseases to biomarker validation, novel diagnostics, and innovative therapies to applications of patient stratification methods in healthcare settings. Overall, it has been estimated that over 3 billion Euros have been invested so far in this area by the EU itself, with additional investment by governments at the national level. Aside from the regular calls for proposals that have relevance for various aspects of PM, ongoing efforts are focused on the development of long-term strategies and public-private or public-public facilitating structures, such as the ERA PerMed, the Innovative Health Initiative, the International Consortium for Personalised Medicine (ICPerMed), and the future European Partnership for Personalised Medicine (EP PerMed), which is expected to become a major driver in the further development of PM. In addition, the EC has provided important legislative contributions in the support of the development of PM, for example, in regard to data sharing and use with the European Health Data Space (EHDS), the clinical trials regulation, the General Data Protection Regulation (GDPR), the health technology assessment (HTA) regulation, and the in vitro diagnostic medical devices regulation (IVDR) (Fig. 2).
From EC research projects experience, it has emerged that PM requires not only interdisciplinary collaboration, but also cross-border 442 collaboration and above all the involvement of all relevant stakeholders in the healthcare field. Progress is needed in important areas, such as data protection regulation, clinical trial strategies, health economics, HTA, including in AI and clinical decision support software, and regulatory and market access, as well as genomic-drug databases or additions to formularies for medicines use within a healthcare system to ensure that PMbased approaches are implemented and reach the patients within a sustainable healthcare system. Such innovation can only be achieved through partnerships and collaboration, in particular public-private, academic, health system, and industry partnerships. In addition, there is still a significant need for investment in further PMbased research, both within cancer, rare diseases, and other major disease areas.
From an overall European perspective, it can be expected that PM-based approaches will develop further along the full value chain at a rapid pace in the future. However, as outlined in the Strategic Research and Innovation Agenda (SRIA) for Personalised Medicine, which was developed in support of the upcoming European Partnership for Personalised Medicine (https://www.icpermed.eu/ en/strategic-research-and-innovation-agendafor-personalised-medicine-1100.php), it is critical that focus is on research, innovation, and implementation. Particularly an implementation of PM in healthcare systems is far from trivial and will require major changes and developments, both regarding healthcare infrastructures and formal frameworks. A further complication is the diversity of healthcare systems across countries and regions across Europe. In order to ensure equity in access to healthcare, an implementation of PM-based approaches must be adapted to fit the resources and possibilities that the individual healthcare systems offer.
As outlined in the SRIA, PM must be regarded as a "System of Health," where all key players interact (Fig. 2) in order to achieve the ultimate vision to improve health outcomes within sustainable healthcare systems through research, development, innovation, and implementation of PM approaches for the benefit of patients, citizens, and society. Notably, the unique ecosystem of PM networks also has economic implications (e.g., creation of new jobs, development of new enterprises, and industries working on shared goals with other stakeholders).
By pooling EU resources, regional and national resources, and scientific expertise across borders, it will become possible to accelerate developments in key areas and ultimately enable improved PM-based health and care services across Europe and beyond. With the drive and motivation seen today by key players, both topdown and bottom-up, there is reason to believe that progress toward PM-based healthcare systems in Europe will develop rapidly over the next decade.

Ethical debate-topics, best practice, and new challenges
Because of the desired rapid exchange of knowledge and information at the intersection between research and clinical care, relevant and specific ethical questions arise in PM. The main areas are (1) genetic data sharing and the risk of reidentification that is inherent to large genomic data; (2) policies for handling additional or incidental findings originating from both the research and healthcare context; (3) the usability of the data, which can create a tension with the patient's right to control and the informational self-determination of their data from both the research and the healthcare system; (4) new ways of having patients participate in research governance and trust-architecture; and (5) fair and equal access to the benefits of PM [27].
PM is only as good as the research data that needs to be annotated with the clinical history data. Patients are therefore asked in advance to make their molecular and clinical data available for research, although at the time of the consent, the research projects making use of the data cannot be specified [28]. This so-called broad consent is a deviation from specific consent, which has so far been considered the ethical gold standard. It must therefore be complemented by a governance structure that ensures that patient data are (a) secure and not accessible by unauthorized third parties, (b) only used for the purposes to which the patient has consented, and (c) used in a context that secures accountability for usage and trustworthy governance [29]. Best practice solutions for consent modules that assure that patients are informed about potential risks, and benefits of PM have therefore been developed by different large data initiatives such as the Global Alliance for Genomics and Health [30], a global platform to enable genomic data sharing and adapt consent modules to the legal requirement of European member states as, for example, for the German Human Genome-Phenome Archive-the place to store genomic data from genomDE and research projects [31].
Two more recent aspects that are worth highlighting are patient and public involvement as well as equal access to precision diagnostics. We have witnessed in the last years that patients are a driving force in almost every aspect of PM: They largely support secondary research with their data [32] and are advocates for research, especially in underfunded areas [33]. Patients have a specific interest that their data also yield a benefit, that data initiatives are trustworthy, and, hence, that they are involved as patient experts in the governance of larger data initiatives. Patients are surely also a strong partner for the last crucial ethical topic that needs to be tackled, that is, equal access to PM. This is a real area of need according to a recent study from the European Society for Medical Oncology (ESMO) on the availability and accessibility of biomolecular technologies in Europe (presented by A. Bayle at ESMO 2022, ESMO Study on Availability and Accessibility to Biomolecular). [34] The study was set up by the health policy and strategy working group of ESMO and surveyed 48 countries with 2-20 reporters per country. Overall, it showed a vastly heterogenous picture in Europe with limited access to testing for more advanced biomarkers (ROS1, BRAF, NTRK) with clinical actionability, and limited knowledge of variant classification schemes (such as ESCAT [35]) for interpretation; only half of the countries had access to MTBs. In a survey of all European countries plus the United Kingdom, the International Quality Network for Pathology (IQN Path), the European Cancer Patients Coalition (ECPC), and the European Federation of Pharmaceutical Industries and Associations (EFPIA) also reported limited access to biomarker testing in several countries [14]. In particular, on average, only 10% of European cancer patients requiring a biomarker test were analyzed using NGS, with huge disparities in access to NGS among the different European countries. Hence, training and knowledge of the respective medical experts, reimbursement of testing for clinically relevant biomarkers, and access to MTBs need to be ensured as well as patient and public forums associated with clinical genomic medicine services, especially when patients and family members are consenting for their clinical genomic data to be used from routine care for research purposes.

Legal issues: the General Data Protection Regulation and the European Health Data Space
PM scenarios are complex due to multipolar interests of the actors, and conflicting interests associated with the same actors and related to data sharing. This gives rise to complex constellations that require respect for competing interests and rights that need to be carefully balanced when making research-and healthcare-related decisions. PM's stratified and public health perspectives influence the balancing among these various interests. Consideration must be given to the fact that intervening in data protection rights and interests of patients is increasingly justified by the individual and stratified benefits of the intrusion, similar to genomic data sharing but more explicit.
The EU GDPR attempts to balance these opposite poles of interests by taking a risk-based approach (Fig. 3) (cf. Recital 76 GDPR) [36]. Reducing risks to the rights and interests of data subjects by technical and organizational measures (Art. 32 GDPR) influences the weighing of corresponding obligations to protect against those risks and obligations in order to promote the ethical mandate of data sharing and usage. Accordingly, the primary role of data security can be seen as a response to the outcome of the trade-off among the different facets of competing interests: The lawfulness of the processing is technically implemented, and the standard of data security determines how the data can be processed [37]. In this way, trustworthy and secure data processing systems emerge to become a decisive principle of PM.
Although guidance exists, far too frequently, the legislator leaves the weighting of affected rights and the conditions for data sharing to those responsible for data processing (Fig. 4). However, data sensitivity, data processing aims and effects, and their justification related to PM are often comparable across data processing scenarios in a way that calls for better standardization of the governance setup. Additionally, the effects of processing can relate to the fundamental rights of patients; therefore, it is an explicit task of the legislator to define the framework for data processing, salvaging law interpretation from breaches by various actors applying the law but allowing enough space to ensure fairness in deciding individual cases.
The draft regulation on the EHDS sets up an approach that allows for cross-border data exchange/access by means of separate infrastructures, depending on the processing purposes of healthcare and secondary use of data (Fig. 2) [38]. It provides extensive measures to significantly expand the availability of data. Not only are member states obliged to join the infrastructures, but there are also obligations for certain natural and legal persons to provide or to register data in order to increase data availability (cf. Art. 7(1), 4(1) (b), 33 EHDS draft regulation).
In order to protect the rights of natural persons and to comply with the principles of data minimization and purpose limitation, electronic health data must generally be made available in anonymized form for secondary use in research. If this approach is not sufficient to achieve the processing purpose of the data user, access to the data may be provided in pseudonymized form (Art. 44(3) EHDS draft regulation). Given that data processing is to be limited to the technical infrastructure of the EHDS, which does not allow users to download or otherwise duplicate the data in question, making data available anonymously does not seem to lead to better data protection and, moreover, to a loss of the analytical value of the data for research [38].
Provided there is a prior successful compliance check, third countries can be linked to the infrastructures via national contact points (Art. 13(3), 52(5) EHDS draft regulation). These national contact points should be established through implementing acts by the EC (cf. Art. 52(5), (13) EHDS draft regulation). The relation between the EHDS compliance check and the conditions for third country data transfers still need clarification with regard to the rules on international transfers set out under the GDPR; this reconciliation is also intended to remedy a possible duplication of protection.
Additionally, data governance structures created by the draft regulation on the EU level (joint controllership group, EHDS Board) need to be matched with those structures on the national level (digital health authorities, health data access bodies, national contact points, and more generally with data trustee approaches) from various perspectives, including the distribution of

Fig. 4 The EU General Data Protection Regulation (GDPR) and its impact on the precision medicine ecosystem. Although national interpretations of GDPR may differ and different data protection rules may apply in EU/EEA and non-EU/-EEA countries, national/regional precision medicine initiatives as well as international research networks/consortia pave the way for secure data sharing, based on common ethicolegal governance, broad consents, and standards/guidelines, that in turn enables a continuous research and innovation cycle to the benefit of patients and society at large.
regulatory competences between the EU and its member states, and from that of practicability.

Evidence generation-innovative precision oncology trials and real-world data
Successful development of precision cancer medicine relies on evidence generated by prospective clinical trials. These may still follow the traditional trial design ranging from phases 1 to 3 and phase 4 for post-marketing surveillance. However, the portfolio of trial designs has significantly widened and now includes, for example, umbrella, basket and platform trials as well as master protocols, investigating more than one drug per cancer type simultaneously in several substudies within a single protocol [39][40][41][42]. Master protocols may use umbrella or basket trial designs or combine both trial types. Basket trials investigate the efficacy of a drug targeting a specific disease mechanism (e.g., a specific mutation) shared by different tumor types (one target-different tumor types). Umbrella trials, on the other hand, investigate different drugtarget matches in a single tumor type (different targets-one tumor type). Although umbrella trials usually employ control arms receiving standard-ofcare therapy not matching to a specific molecular lesion, basket trials lack control arms by design. Basket trials, which are frequently used to investigate pan-tumor drug-target matches, may be less complex and easier to execute, but the lack of control arms can pose a challenge for HTA bodies in the EU because their evaluation criteria often follow classic trial designs, that is, the concept of a randomized controlled clinical trial.
Platform trials, also known as multiple arm, multiple stage studies, are randomized controlled clinical trials which simultaneously investigate several treatments in different study arms against a single control arm, all embedded in one trial. The trial design is adaptive, meaning that treatment arms may be discontinued, and new study arms investigating a specific drug-target match may be added. Patients may also be reassigned to a new treatment arm based on their response to previous treatment and/or change in biomarker status. This trial approach dissolves the classic stepwise trial design and aims at accelerating drug development.
All trial types require sophisticated and standardized molecular screening protocols to identify the different drug targets or predictive biomarkers investigated in these trials. A multitude of trials analyzing cancer vaccines and cell therapies will likely require broad and highly standardized profiling approaches to reliably determine neoantigens, HLA haplotypes, and other tumor characteristics including the immune cell contexture. Future clinical trial designs will frequently incorporate dynamic measurement of certain biomarker settings to identify resistance mechanisms, relapse, and measurable residual disease, and where possible to identify eligibility from standard-of-care cancer genomic testing. Together with the shift of personalized treatment approaches to earlier disease stages and earlier lines of therapy, these developments will necessitate broader use of repetitive liquid biopsies during the course of disease, often delivered in conjunction with the healthcare system when they are offered as part of routine care. The ecosystem needed for these complex profiling approaches in clinical trials requires standardization, very high levels of multidisciplinary expertise, and tremendous resources, all of which must also be available after an approval of a certain drug-target combination by the regulator (e.g., FDA, European Medicines Agency [EMA]) for standard care. In other words, diagnostic and clinical requirements derived from approved clinical trials, which become increasingly complex and sophisticated, must be transferable to the realworld setting to ensure equal access to therapy. Adequate reimbursement schemes that reflect the diversified resource burden are critical in this context.
The trial ecosystem described before can only be successful if both academia and industry partners leverage their individual assets and strengths. For example, observations made in registry studies, for example, a specific molecular setting associated with long-term response, can inform basic research programs but may also lead to the development of new, possibly smaller and more efficient clinical trials investigating this observation further. Similarly, the reuse of available drugs with diseasespecific approval for new indications is a growing field in which public-private partnerships including the integration of HTA bodies play a crucial role [42][43][44]. The collaboration among academic institutions, pharmaceutical industries, and possibly regulatory agencies is also necessary to favor the access of cancer patients to NGS tests. In this respect, an academic Italian initiative funded by several pharmaceutical industries allowed the access of patients with advanced cancer to comprehensive genomic profiling, thus favoring patients' enrollment in clinical trials of targeted therapies [15].
The term RWD refers to health data that are collected during the course of disease, including treatment and outcome data [45]. These may be derived from individual patient records or registrybased studies and may also include data collected via health apps or by wearables [46,47]. Such data reflect to a certain extent the reality of treatment and associated outcomes outside the context of clinical studies which-by design and intentionally-carefully control numerous parameters and factors.
In fact, the majority of data related to drug efficacy and disease outcomes is generated outside of clinical trials, and the volume of such RWD continues to grow. Real-world datasets potentially harbor a wealth of scientific knowledge and can support preclinical studies, drug development, clinical trial designs, post market surveillance, and regulatory processes by distilling real-world evidence (RWE) from these datasets [48]. However, collection and exploration of these data are still challenging for many reasons, including the standardization of the type, quantity, and quality of RWD items as well as ethical and legal considerations that may differ between countries. To facilitate and foster the generation and use of RWD and RWE, respectively, the concept of research-based healthcare, which continuously generates knowledge through routine clinical practice by systematically collecting different data layers (e.g., clinical including outcome parameters and various layers of molecular data) for each patient, is paramount. Acknowledging this scenario, the FDA has issued guidance documents that aim at supporting the use of RWD and RWE in drug development [49]. Future developments may lead to hybrid trial designs involving concomitant RWE generation [50].
One example where RWE currently plays a significant role is in the context of basket trials in oncology, which lack control arms by design. Particularly low frequency genetic aberrations, such as certain gene fusions, will make the generation of control arms impossible as hundreds of thousands of patients would have to be screened, who-in the case of being randomized into the control arm-would receive standard treatment in a metastatic/advanced disease setting after several lines of therapies-a scenario that has significant references to practical applications of clinical ethics. These considerations are important for HTA bodies in the EU that assess the benefit of the drug in comparison to the current standard treatment as well as the cost-benefit ratio of a certain drug or drug-target match after approval by the regulator EMA [51]. HTA-body decisions in the EU countries may differ because their assessment framework and methodologies are not harmonized and rely on national legislations and procedures (Fig. 2). This may result in divergent, country (HTAbody)-specific access to drugs, although the underlying clinical trial data submitted to and approved by the regulator EMA are the same. For example, some HTA bodies may allow for externally generated control arms, whereas others would only consider internal control arms as in phase III clinical trials. The EU is facing this challenge by trying to harmonize the work of HTA bodies in the EU [52].

Value recognition and financing
The financing of PM should be holistic, that is, it must not only cover the diagnostic genomic test, but also the required wrap-around infrastructure (e.g., IT and data infrastructure, the laboratory operating costs, the building infrastructure, sequencing equipment) and the services of the respective specialist expertise (e.g., case management, molecular reports, MTBs, genetic counseling). This will enable phenotype-genotype correlations to be made and pathogenicity of variants to be established, together with potentially actionable genomic targets that can lead to effective treatment interventions.
Precision therapies (on-label, off-label, clinical trials) should also be considered conceptually because a reimbursement structure based exclusively on the diagnostic process does not take into context how genomically informed treatments need to be made clinically in both cancer and rare diseases. Importantly, the entire process is knowledge-generating care that will continue to evolve as more evidence emerges from further interpretation of the genome. It follows that this requires the integration of molecular and clinical data, which is the essential component of PM, and that this requires a suitable data infrastructure that needs to be financed.
Due to the fact that the success of PM is driven by a complex multidimensional infrastructure and multidisciplinary teams, the reimbursement structure is also multidimensional and requires coordination and integration of the various providers and stakeholders. As the healthcare system of many countries is hugely siloed, this point is essential. Payers who pay for clinical diagnostic services or therapeutic services, for example, must coordinate with payers who pay for infrastructure measures, and funders for research must also be included in this coordination process (definition of interfaces). Full cost calculations are necessary, otherwise misconceptions arise (the frequently cited "$100 genome" does not correspond to a diagnostic clinical-grade genomic test) that do not allow for sustainable financing and end-to-end provision [53].
In some EU countries, such as France and Denmark, governmental or both governmental and private funding have been used to implement WGS as part of PM in the clinical setting where patients are expected to benefit from the test. The objective has been to introduce new diagnostic methods and technology broadly into healthcare equally across the country. It has involved not only diagnostic laboratories, interpretation staff, and IT and data infrastructure but also many healthcare professionals who needed to change the way they work. In the Danish initiative, a process has been initiated on how to develop a framework for evaluating the implementation of WGS in a clinical setting. In France, the clinical benefit of genomic testing for the PFMG2025 pre-indications will be evaluated by the French Health Technology Assessment agency (Haute Autorité de Santé) for long-term reimbursement by the French health services. In England, NHS funding has enabled WGS to be established as part of routine diagnostic care across the whole population and linked to PMs and advanced medical therapeutic products. This is as part of an integrated and contracted infrastructure inclusive of other non-WGS genomic tests, genomic laboratories, clinical, scientific and data and informatics leadership and services, and where this is an explicit relationship and coordinated support for research and clinical trial opportunities. This WGS service is delivered in a partnership with Genomics England that through government funding supports the development of an underpinning ordering, analytical and results platform and when informed consent is provided by patients the ongoing population of the National Genomic Research Library (Fig. 1).

Master's programs
Several universities across Europe offer master's programs in precision medicine, which provide a comprehensive overview of the field and cover topics, such as genomics, bioinformatics, data analysis, and ethical and legal considerations Short courses and workshops Various short courses and workshops are available that provide healthcare professionals with an introduction to precision medicine and cover specific topics, such as biomarker discovery, pharmacogenomics, and trial design In the European initiative 1+Million Genomes, a working group has been established to look at methods to evaluate the socioeconomic effect of using WGS in healthcare. Hopefully, these initiatives-together with experience from the clinical workflow and literature gathered over time-will support decision-making.

Cross-disciplinary education and training of healthcare professionals
As the field of PM continues to grow, healthcare professionals in Europe will need to be trained and educated on how to effectively implement this approach into their practice. Healthcare professionals need to possess a wide range of knowledge and skills from various disciplines, including genetics, bioinformatics, epidemiology, pharmacology, and more [54].
One way to facilitate cross-disciplinary education and training of healthcare professionals in PM is through the development of educational programs/activities that bring together professionals from diverse fields, such as genetics, bioinformatics, and clinical medicine. These should be designed to provide professionals with a comprehensive understanding of the principles of PM, as well as practical training on how to apply these principles in their practice. In addition to crossdisciplinary training, it will also be important to provide ongoing education and training opportu-nities for healthcare professionals to keep up with the rapidly evolving field of PM, which also include necessary skills to work with advanced technologies, such as genomics and large-scale data analytics (Table 1).
Finally, it will be important to ensure that healthcare professionals are able to effectively communicate with patients about PM and the implications of using this approach in their treatment. This will require the development of training programs focused on effective patient communication and engagement. Genomics education and training resources are designed for "just in time" (reactive) or "just in case" (proactive) learning. GeNotes (genomic notes for clinicians) is a two-tier "just in time" resource. Tier 1 (In the Clinic) resources support clinicians to request the right genomic test and return the genomic results [58]. They are specialty-specific and mapped to England's National Genomic Test Directory [59]. Available throughout are links into tier 2 (the "Knowledge Hub"), which provides further learning.

Experiences within NHS England on education and training
A number of "just in case" educational resources and programs has also been developed, from bitesized guides and introductory courses to massive open online courses and a Master's in Genomic Medicine framework.
Recently, a Genomics Training Academy (GTAC) has been established for the specialist laboratory and clinical workforce [60]. The virtual "hub"/inperson "spoke" model will increase training capacity, ensure high-quality, consistent education, and promote interprofessional learning not just in England but with other countries.

Awareness-raising initiatives and success criteria on implementing precision medicine
Awareness-raising initiatives play a crucial role in the successful implementation of PM in healthcare as they increase the understanding of the potential benefits and challenges of PM among the general public, patients, healthcare professionals, and policy makers. To measure the success of PM, various metrics can be used. For instance, PM implementation can be evaluated through various measurements and criteria, including improved patient outcomes, cost-effectiveness, infrastructure and resource utilization, patient engagement and satisfaction, and research and innovation (Table 2). By tracking such metrics, healthcare providers and policymakers can assess the impact of PM and identify areas for improvement. These criteria can also help to guide the development and implementation of PM approaches in healthcare and ensure that they deliver their intended benefits.

Education of patient representatives, next of kin, and the society at large
A shift in international culture is underway in engaging with patients, next of kin, and the general public. The paradigm shift of personalized and data-driven healthcare, however, involves new questions for patients, next of kin, and the society at large, to which they must at least have a basic understanding in order to actively participate in collaborations and the ongoing conversation.
Many initiatives are ongoing, locally, nationally, and internationally, to educate patient representatives. One early initiative is EUPATI [63], an international nonprofit organization providing, for example, patient expert training and toolboxes to empower patients and patient representatives with knowledge and skills to engage and partner with other stakeholders. National platforms to educate patient representatives are also starting-for example, EUPATI Sweden [64] is now being launched by the Swedish Disability Rights Federation.
Engaging the society at large will be one key part to continue developing the field of PM in an ethical and responsible way. Here, we need to better understand the general public and patients' attitudes toward sharing health data. Surveys among patient groups-for example, patients with rare diseases-show a general positive attitude toward sharing their health data [65]. Similarly, this has been shown in an opinion poll in Sweden in 2022 in which 87% of the public were in favor of sharing their health data for research and health purposes. When questions were asked about DNA, genetics, and genomics, however, the global "Your DNA, Your Say" survey showed that two in three respondents were unfamiliar with DNA, genetics, and genomics, and only 52% of people said they would donate anonymous DNA and medical Table 2. Assessing the impact of implementation of precision medicine.

Improved patient outcomes
The ultimate goal of precision medicine is to improve patient outcomes, such as survival rates, disease remission, and quality of life. Therefore, the success of precision medicine implementation can be measured by evaluating patient outcomes before and after the implementation of precision medicine approaches. These outcomes can be assessed through clinical trials, real-world evidence, and patient-reported outcomes Cost-effectiveness The cost-effectiveness of precision medicine can be measured by evaluating the cost of implementing precision medicine approaches relative to the benefits they provide. The success of precision medicine implementation can be assessed by measuring the reduction in costs associated with ineffective treatments and the overall cost savings associated with precision medicine approaches Infrastructure and resource utilization Precision medicine requires a significant investment in infrastructure, including data management systems, genetic testing, and data analysis tools. The success of precision medicine implementation can be measured by evaluating the adoption and utilization of these infrastructure resources

Patient engagement and satisfaction
Precision medicine requires a collaborative approach between healthcare providers and patients. Therefore, the success of precision medicine implementation can be measured by evaluating patient engagement, satisfaction, and experience with precision medicine approaches. This can be done through surveys and patient-reported outcomes

Research and innovation
Precision medicine requires continuous research and innovation to advance the field and improve patient outcomes. The success of precision medicine implementation can be measured by evaluating the level of investment in precision medicine research, the number of publications and patents related to precision medicine, and the adoption of new precision medicine technologies information for use by medical doctors [66]. The need to understand the public's knowledge and perception of genetic data has become increasingly urgent as we enter a new era of genomics/PM in which unique ethical and moral issues arise, both on a personal and political level. The work to increase societal awareness around these questions has started-for example, through public dialogues, community projects, and debate articles-and needs to be increased and continued.

Concluding remarks
Europe is at the forefront of implementing PM in healthcare, and many initiatives and networks have been established aimed at advancing the field. Such initiatives-whether top-down, bottomup, or a mixture of both-involve a wide range of stakeholders, including researchers, clinicians, patients, and policy makers as well as the soci-ety at large. Only consensus among all these players can create and establish a solid and financially sustainable ecosystem including an ethicolegal framework that supports and advances crucial developments in comprehensive molecular diagnostics, data sharing, MTBs, genetic counseling, drug access, RWE generation, and novel trial designs. Breaking down silos and developing a comprehensive infrastructure for multidisciplinary life-long learning and training will be as important as the active engagement of patients to successfully leverage a deeper understanding of disease biology for improving patient outcomes.

Author contributions
Conceptualization; visualization; writing-original draft; writing-review and editing: Albrecht Stenzinger and Richard Rosenquist. Writing-original draft; writing-review and editing: Eva Winkler,