Kicking off day two of Biomarkers & Precision Medicine US 2025, a diverse group of experts gathered for a roundtable discussion on technological developments guiding the future of precision medicine. The conversation traversed the evolving landscape of clinical trials, the promise and pitfalls of omics technologies, the challenges of technology adoption, and the critical need for validation and cost-effectiveness in clinical practice. The dialogue illuminated both the excitement and scepticism surrounding new approaches, offering a nuanced view of the field’s trajectory.
The Complexity of Patient Stratification
A recurring theme was the complexity of patient stratification in clinical trials. Participants discussed the use of omics-based propensity score matching, which leverages gene expression and other molecular data to match real patients to outcomes, especially in early-phase trials. This approach, while innovative, was met with caution. Several experts noted that gene expression is not yet widely used in clinical settings for patient stratification; instead, mutation status and protein expression remain the primary tools. Only a handful of proteins currently have companion diagnostics, highlighting the gap between research and clinical application.
The group debated the feasibility of relying on gene expression for prospective clinical decision-making. While omics data can offer precision, translating these findings into simple, actionable clinical features often results in a loss of granularity. The consensus was that any new stratification strategy must be validated with robust companion diagnostics and external datasets before it can be widely adopted.
The Role of Artificial Intelligence and Explainability
Artificial intelligence (AI) emerged as both a source of excitement and of scepticism. The roundtable acknowledged the potential of AI models to uncover patterns in complex datasets, but concerns about explainability and trust were prominent. Participants stressed that black-box models, while powerful, are difficult to translate into clinical practice unless their decision-making processes can be clearly understood and validated. Manual annotation and double-checking by a human remain essential, and synergistic evidence from multiple sources is needed to elevate any target or concept.
The discussion highlighted the importance of extracting biological principles from omics data, rather than relying solely on algorithmic outputs. The ultimate goal is to distil findings into pragmatic, reproducible solutions that can be prospectively validated in phase three trials. This requires bridging the gap between sophisticated discovery processes and simple, clinically implementable assays.
Technology Adoption and Cost Barriers
The conversation shifted to the challenges of technology adoption, particularly for smaller organizations and startups. The high cost of collecting tissue samples, generating next-generation sequencing (NGS) data, and conducting exploratory analyses in early-phase trials was identified as a major barrier. Participants lamented missed opportunities when resource constraints prevent comprehensive biomarker analysis, especially in phase one trials where safety and toxicity are the primary focus.
A proposed solution was the creation of third-party infrastructures to support startups in collecting and analysing biomarker data. Academic efforts and collaborative trials were cited as promising models, enabling rapid iteration and learning without placing undue financial risk on individual companies. The group agreed that maximising output and cost-effectiveness is essential, given the expensive nature of emerging technologies.
The Evolution of Omics and Spatial Technologies
Reflecting on the history of next generation sequencing (NGS), participants noted how technology adoption drives down costs over time. What once required millions of dollars can now be accomplished for a fraction of the price, thanks to mainstream market uptake. The same trend is beginning to emerge for single-cell and spatial technologies, though the path to widespread adoption remains uncertain. Manufacturing processes and standardisation are still evolving, and it is unclear whether all technologies will reach the early majority market or fade away.
The group discussed the disruptive potential of spatial profiling, which can reveal diagnostic features not accessible through NGS or single-cell approaches. For example, measuring the distance between immune cells and tumour cells within tissue can provide insights into tumour penetration and immune response. However, demonstrating clinical utility through improved survival outcomes remains a significant hurdle. Large patient cohorts and rigorous validation are required to prove that new technologies offer meaningful benefits.
Standardisation and Repeatability
As technologies mature, the need for standardisation becomes paramount. The roundtable agreed that mainstream market users seek platforms with established standards to ensure repeatability and compatibility. NGS has achieved this level of maturity, but spatial technologies are still in the process of developing shared data formats and processing pipelines. The interaction between technology providers and users is crucial for driving improvements and establishing standards that facilitate broader adoption.
The Search for Clinical Utility
Ultimately, the discussion converged on the search for “killer applications,” areas where new technologies can fill critical gaps and demonstrate clear value for patients. Participants agreed that cost-effectiveness and clinical utility are the keys to adoption. Technologies must not only offer superior diagnostic or prognostic capabilities but also prove their worth in terms of patient outcomes and reimbursement. The journey from research innovation to clinical routine is long and fraught with challenges, but the potential rewards are significant.
Conclusion
The roundtable at Biomarkers & Precision Medicine US 2025 provided a rich tapestry of perspectives on the future of the field. The dialogue underscored the importance of validation, explainability, cost-effectiveness, and standardisation in translating new technologies from bench to bedside. While optimism abounds for the promise of omics, AI, and spatial profiling, the path to clinical impact requires careful navigation of scientific, operational, and economic hurdles. As the field continues to evolve, collaborative efforts and pragmatic solutions will be essential for realising the full potential of precision medicine.
