Your career integrates fundamental research, clinical practice, teaching and active participation in the definition of health policies. What motivated you to build such a transversal career within medicine, and which moments do you consider decisive in consolidating your career?
My motivation to build a transversal path probably began when I chose Anatomical Pathology as my specialty while still studying Medicine. That was when the curiosity was “born”: I had exemplary professors who opened my eyes to the potential of Pathology as a central discipline in medicine, the area where understanding the mechanisms of disease can, in a direct way, improve diagnosis and guide clinical decision-making. After that first milestone, the post-doctorate at IPATIMUP/FMUP was decisive in structuring a research trajectory with clinical relevance and in reinforcing the idea that science must return to the patient through better methods and better diagnostic quality. The leadership of the Molecular Pathology Unit and its consolidation as an internationally recognised infrastructure represented another determining step: transforming innovation into routine and reducing variability in practice. International experience was equally structuring. At the University of Toronto, as Full Professor and pathologist at Toronto General Hospital (UHN), I was able to integrate, within a highly demanding academic-hospital environment, diagnostic practice, translational research and teaching, and to understand how these dimensions reinforce one another when the system is designed to accelerate the passage of knowledge into practice. In Luxembourg, as Director of the Department of Medicine and Pathology at the Laboratoire National de Santé, the connection between organisation, technology and implementation policy became particularly evident; the telepathology project within the Luxembourg Plan Cancer is a concrete example of innovation with impact on diagnostic time and quality at system scale.
As for active participation in the definition of health policies, I view it as a natural consequence of the path: first to learn, to research and to teach; then to assume the responsibility of helping to “change the system” so that students, doctors and researchers find better conditions and more sustainable pathways. This intervention occurs both through institutional and regulatory channels and through scientific-technical and consensus work via international committees (International Academy of Cytology) and through editorial and standardisation participation in systems such as those of the World Health Organization, which transform evidence into a common, comparable and applicable language. In Portugal, this same logic is also expressed in an institutional commitment to creating conditions for clinical and translational research in network and with impact: the leadership of RISE-Health, another fundamental milestone in the construction of my career, reinforces this ambition by structuring capacities, promoting collaboration and accelerating the transfer of knowledge into practice, contributing to a stronger ecosystem for the next generation of students, doctors and researchers.

You lead RISE-Health and the Associate Laboratory RISE, both firmly committed to translational research. What do you consider to be today’s greatest challenges and opportunities for translational research to translate increasingly into real impact in healthcare?
The greatest challenge of translational research today is to ensure that scientific discovery does not become “trapped” between the laboratory and publication, but travels the entire path to implementation. In practice, this comes up against several critical points: fragmentation of data and infrastructures (often without interoperability and with difficult access to high-quality clinical data), bureaucratic and contractual complexity (ethics, data protection, procurement, partnerships, trials), and the difficulty in securing sustainable funding for the most “expensive” phase of translation; multicentric validation, proof of clinical utility and cost-effectiveness evaluation, without which the NHS and providers are unable to scale solutions. This is a theme very present in the construction of RISE-Health as a network, precisely because clinical research requires common processes, coordination and critical mass to overcome these barriers. In parallel, there are human and systemic challenges: careers still poorly aligned with the real time of translation, academic incentives not always centred on impact and implementation, and the need for hybrid teams (clinicians, researchers, bioinformaticians, project managers, technology transfer professionals, etc.) working with shared objectives and metrics. As for opportunities, I see three that are particularly strong. First, the network model: both the Associate Laboratory RISE and the RISE-Health Research Unit bring together units with complementary competences, which facilitates multicentric studies, sharing of resources and the creation of projects of sufficient scale to generate robust evidence and attract competitive funding. Second, digital transformation and the integration of bioinformatics/AI with clinical and molecular data, which accelerate biomarker discovery, stratification and precision medicine. Third, proximity to society and training: literacy initiatives and links with the community, and institutional investment in Biomedical Innovation create a talent pipeline and a culture of translation from an early stage, making it more likely that innovation is designed with the final user (patient, professional and system) in mind. Ultimately, real impact happens when research begins to be planned “backwards”: starting from the clinical problem and the context of the system, designing the evidence needed to change practice, and only afterwards choosing technology and method. It is this logic that we seek to reinforce at RISE-Health and at the Associate Laboratory RISE.

Pathology is an absolutely central specialty in modern medicine and has become even more crucial with the advance of precision medicine. How would you describe the current role of pathology in this new ecosystem that is so demanding and oriented towards the personalisation of diagnosis and treatment?
Today, pathology is, simultaneously, the starting point and the integrative “hub” of precision medicine: it continues to be grounded in morphology, but has come to incorporate, in an inseparable way, additional layers of information – molecular, computational and, increasingly, digital – which directly condition therapeutic decisions. Diagnosis is no longer merely “giving a name” to the disease; it is producing a clinically actionable, robust and reproducible phenotype. In this new context, the pathologist assumes a particularly critical role as “manager of biological material” (tissue stewardship) and guarantor of quality: deciding what to test, when to test and in which sample, ensuring tumour adequacy, controlling pre-analytical variables and avoiding waste of tissue/time, in a scenario in which biomarkers are continually increasing and multigene panels (NGS) replace, advantageously, sequential strategies. Finally, pathology is also a pillar of clinical governance and equity in precision medicine: heterogeneity between laboratories, the need for accreditation/standardisation, the incorporation of bioinformatics and AI (with clear ethical and regulatory requirements) and the reduction of disparities in access to high-quality testing are, today, as decisive as the technology itself.

You have accompanied several paradigm shifts in cytopathology and breast pathology, both your specialties. In which tools or approaches do you observe the greatest potential to optimise the diagnosis and treatment of breast cancer?
Over these years, what has changed most is the transition from a diagnosis of “morphology + few markers” to an integrated ecosystem, in which sample quality, correlation with imaging, rigorous evaluation of biomarkers and molecular characterisation directly condition therapeutic decision-making. In this context, I see five areas with the greatest potential to optimise the diagnosis and treatment of breast cancer. The first is the optimisation of sampling and sample representativeness, with an increasingly image-guided approach and the need to reduce insufficient samples and false negatives. A very concrete example is the use of core needle biopsy, which performs better than fine-needle aspiration, with greater sensitivity and fewer insufficient samples. The same logic applies to precursor lesions, where minimally invasive techniques (including vacuum-assisted biopsy/excision) help to avoid unnecessary surgery without losing diagnostic safety. The second is the standardisation and “robustness” of classical and emerging biomarkers, because this is where therapeutic precision is gained (or lost). In current practice, pathology continues to be central in determining prognostic and predictive factors, such as – histological grade, biomarkers, margins and nodal stage. At the same time, molecular classification itself is being refined (for example, the need to identify HER2-low when there is no amplification, with later therapeutic implications). The third is the systematic integration of molecular medicine into routine practice, with “reflex” strategies and selection of methodologies according to the clinical question. Today, it makes sense to think by subtype and by context: prognostic genomic panels in early luminal disease; PIK3CA and ESR1 in metastatic luminal disease; rigorous confirmation of HER2 in HER2+; and, in triple-negative disease, germline BRCA and PD-L1 as tests that have become decisive for access to innovative therapies. The fourth is liquid biopsy as a dynamic complement to tissue biopsy: monitoring response, early identification of resistance and, above all, the possibility of detecting minimal residual disease before imaging evidence. At the same time, it is crucial to maintain a clear message: liquid biopsy does not replace tissue and requires caution in interpretation (for example, interference from clonal haematopoiesis). The fifth, and perhaps the most transformative in terms of scale, is digital pathology and artificial intelligence, not as “futurism”, but as a practical tool to reduce variability and increase reproducibility. Digitalisation opens the way to learning models capable of supporting demanding tasks such as histological grading and biomarker determination/quantification. The next step will be multimodal AI (histology + imaging + genomics + clinical data) and, with prudence, the use of generative AI for well-defined tasks (e.g., structured documentation), always with validation, traceability and regulatory framing. Ultimately, more than “one” predominant technology, the greatest potential lies in integrating these approaches into a coherent pathway: better sampling, better pathology (more reproducible), better molecular stratification and better monitoring over time.

Your collaboration with the World Health Organization leverages standardisation and quality in diagnosis and treatment. How can harmonisation of criteria contribute to greater global quality and equity in access to healthcare?
Harmonisation of criteria is, in practice, a way of transforming scientific knowledge into clinical safety. When different countries, hospitals and laboratories use the same language, the same diagnostic categories and the same decision thresholds, ambiguity and interobserver variability are reduced, communication with clinical teams is improved and therapeutic decision-making and follow-up become more consistent. That is precisely the logic of the WHO nomenclature systems: to standardise diagnostic criteria, define categories with estimated risk of malignancy and support more predictable and comparable clinical decisions. From a quality perspective, this harmonisation is not limited to the “name” of the diagnosis. It also includes recommendations of good practice for sampling, handling and processing of specimens, because an important part of errors and discrepancies arises even before microscopic interpretation (pre-analytical phase and material triage). By standardising minimum report components, an objective basis is created for audit, benchmarking, accreditation and continuous improvement. As regards equity, the gain is particularly clear: WHO systems explicitly recognise that infrastructure and resources vary greatly between countries and contexts and, therefore, are designed to be applicable at different resource levels, valuing key morphological criteria and proposing management pathways that can be locally adapted. This helps to reduce “two-speed medicine”: it does not eliminate resource differences, but reduces the impact of those differences on diagnostic consistency and decision-making, and creates an international reference to justify investment in laboratory capacity and training. It is within this framework that I see the value of collaboration with WHO: creating a global “common language”, supported by criteria, risk and recommendations, which improves quality where there are already many resources and, at the same time, raises the minimum standard of diagnostic safety and predictability where resources are more limited.

How do you see the future of medicine in the era of digital, artificial intelligence and molecular pathology? In your opinion, are we witnessing a redefinition of clinical practice through new tools or are we facing a deeper transformation?
I believe we are facing a deeper transformation. If it were merely a “redefinition through tools”, we would mainly be speaking about efficiency: digitising processes, automating tasks, accelerating response times, reducing costs and decreasing variability. That is important, but it is only doing better what we were already doing. What is happening goes beyond that, because it alters three pillars of medicine:

1. The concept of disease is changing
Disease ceases to be defined only by a morphological or anatomical label and comes to be defined by biological mechanisms and trajectories. In oncology, for example, clinical decision-making is no longer organised only by “organ” or “histological type”, but by actionable pathways, molecular profiles, clonal heterogeneity and mechanisms of resistance. This changes the way we classify, stratify risk and choose therapy.

2. The method of clinical decision-making is changing
Decision-making ceases to be only interpretative and becomes, in part, computational and probabilistic, constructed from models that integrate multiple sources (clinical, imaging, digital pathology, genomics, longitudinal data). The clinician no longer decides only on the basis of isolated “signals”, but decides on the basis of quantitative syntheses, with explicit levels of uncertainty and with continuous updating as new data arrive (for example, molecular monitoring over time). This is a paradigm shift: from episodic medicine to continuously informed medicine.

3. The system of evidence production is changing
The classic model (“clinical trial → guideline → practice”) is being complemented by a faster cycle: real-world data, structured registries, continuous validation, and system learning. Instead of evidence being an event, it becomes a process. This creates enormous opportunities, but also requires new rules of quality, governance and transparency.

Therefore, yes: clinical practice will be redefined – but because the conceptual and operational foundations of medicine are being transformed. The tools are the visible engine; the transformation is the change in the “how” and the “what” it means to diagnose, prognosticate and treat. It should be emphasised that this transformation will only be positive if it is guided by three principles: rigour (validation, quality control), integration (teams and data speaking the same language) and equity (so that digital and molecular advances do not create two-speed medicine). Otherwise, we will merely have sophisticated technology amplifying inequalities and uncertainty.

What do you consider essential to convey to the new generations of doctors and researchers, in an increasingly competitive, accelerated and multidisciplinary scientific world?
The central message I would convey is simple: in an accelerated world, what sustains a long and useful career is rigour, purpose and the ability to collaborate. First, learn to ask good questions, design solid studies, treat data seriously and accept that reproducibility and quality are worth more than volume. Reputation is built slowly and lost quickly; therefore, ethics, transparency and quality control are not “extras”, they are the core. Second, choose a real problem and maintain focus. As a colleague of mine used to say, dispersion is the first step towards incompetence. Working on clinically relevant problems, with measurable impact, is the best way to give meaning to effort and to get through difficult phases. Third, learn to work in a team and to speak several “languages”. Medicine and research today are inevitably multidisciplinary: clinicians, pathologists, biologists, bioinformaticians, engineers, statisticians, project managers. The essential point is knowing how to collaborate, communicate clearly, share credit and build trustworthy networks. Fourth, digital and critical literacy. It is mandatory to understand data, molecular biology and digital/AI tools sufficiently to use, question and validate them, but without being dazzled. Useful technology is that which improves decisions and outcomes; everything else is noise. Finally, do not confuse speed with direction. A career is a marathon: it is necessary to protect time to think, read, teach and train others. When we reach a certain maturity, the next step is to try to improve the system – to create conditions and opportunities for those who come next, and not merely to accumulate individual achievements.

What continues to motivate you as a researcher and doctor? Are there fundamental questions that have accompanied you since the beginning of your career and that you continue to ask yourself?
What continues to motivate me is, essentially, the same idea that led me to choose Anatomical Pathology: to be at the point where disease can be understood in depth and, at the same time, where that understanding can change real decisions for the patient. With each difficult case, each discrepancy, each unexpected result, I once again feel that there is still much to learn. I am also motivated by the human and institutional dimension: training people, creating teams and conditions so that others can do better than we did. There comes a moment in one’s career when impact ceases to be only what we publish or diagnose; it becomes the capacity to build structures that raise standards in care, research and training. And it is precisely here that, today, perhaps my greatest motivation lies: the commitment to build and consolidate RISE-Health. Because RISE represents the possibility of creating real conditions for translational research with scale, quality and impact, generating collaborative networks, training talent and accelerating the passage of knowledge into clinical practice, in a sustainable manner and with ambition for equity. I therefore continue to ask myself one final, very practical question: does what we are building concretely improve the life of the patient and the future of the next generations of doctors and researchers? When the answer is “yes”, motivation renews itself every day.

You can find more information on the professor and researcher here.