Jesper Frank Bastlund, Senior Director in Biological and Translational Research, and Enchi Liu, Vice President, Head of Experimental Medicine at Lundbeck, examine how biomarkers can help advance the field of neuroscience. 

Navigating the challenges of CNS Drug development

Neuroscience research and development has historically been one of the most challenging areas in pharmaceutical innovation, with a disproportionately high number of failures in Phase III clinical trials. These late-stage failures are not only costly but also devastating for patients awaiting breakthrough treatments for neurological and psychiatric conditions with high unmet need. However, through the strategic implementation of biomarkers, particularly in early-stage development, drug developers can significantly de-risk their programmes and address many of the historical challenges that have plagued the field. 

The landscape of CNS biomarkers has evolved dramatically over the past decade. Historically, validated biomarkers in central nervous system disorders were virtually non-existent, forcing researchers to progress drug candidates into late-stage trials without sufficient evidence of their efficacy or biological activity. This approach inevitably led to high-profile failures as companies lacked the tools to make data-driven decisions about asset progression. 

Today, the picture is changing, albeit unevenly across different disease areas. Alzheimer’s disease represents perhaps the most dramatic transformation, with validated biomarkers now playing a central role in drug development and even gaining regulatory acceptance. However, the majority of CNS indications still lack robust, validated biomarkers, leaving significant gaps in our ability to make informed development decisions.

The critical role of patient selection

One of the most crucial applications of biomarkers in neuroscience R&D is patient selection: ensuring that the right populations are enrolled in clinical trials to maximise the probability of success. This approach involves using biomarkers or other selection criteria like genetics to identify patients who are more likely to respond to a particular treatment based on their biological profile. 

Patient selection is particularly important in conditions where different underlying pathologies may present with similar clinical symptoms. By stratifying patients according to relevant biomarkers, it is possible to design more targeted trials that are better positioned to demonstrate efficacy, even with smaller sample sizes.

A structured framework for biomarker integration 

Successful approaches to biomarkers require a structured framework that breaks down biological hypotheses into different levels of biological engagement. This typically includes linking target engagement, mechanistic effects, pathway engagement and symptom effects or disease modification to specific unmet medical needs. 

In Alzheimer’s disease, mechanistic biomarkers have allowed researchers to interrogate their hypotheses directly in patients. This represents a remarkable evolution from just a decade ago, when definitive diagnosis was only possible through postmortem examination. The approval of amyloid PET tracers around 2014 transformed the field, moving the gold standard from neuropathology to in-life imaging. This breakthrough has subsequently enabled the development and validation of fluid biomarkers in cerebrospinal fluid and plasma that can help diagnose and track disease progression. 

By contrast, in certain disease areas like epilepsy, genetic knowledge provide causal evidence for a disease biology hypothesis, providing valid cross-species translational opportunities and making it easier to identify and monitor relevant biomarkers, such as seizure frequency, early in clinical development. This enables faster decision-making and more efficient resource allocation with a fail-fast mindset.

Addressing Phase II attrition

A major challenge in CNS trials is patient attrition, particularly in Phase II studies. To address this challenge, a “let the molecule speak” approach to guide drug development is adopted. This strategy uses knowledge of target mechanism and biology disease hypotheses to explore different opportunities where a molecule could work, testing these possibilities in early development and using the resulting data to make decisions about optimal positioning. 

Disease biology biomarkers help select the right patients for clinical studies, avoiding large, resource-intensive trials with high attrition rates due to lack of efficacy in a heterogenous population. This targeted approach not only improves success rates but also reduces the burden on patients who might otherwise participate in trials unlikely to benefit them.

Early integration of biomarker strategy

Determining the right biomarkers to deploy in early-stage development involves examining the interplay between the molecular target, its biological impact and the disease biology in a selected target clinical indication starting already in the early discovery phase of a R&D programme. In essence, this approach combines a holistic “bottom-up” proximal biomarkers (dictated by the molecular target) with “top-down” disease-related pathway biomarkers. This comprehensive strategy helps de-risk the development process, enables faster decision making, and supports a more targeted approach to late-stage development. 

A crucial aspect of any biomarker strategy is ensuring its implementation early in the development process. This means establishing a clear, structured and deliberate plan for each programme that encompasses biomarkers being investigated, preclinically, and clinically. In early-stage development, these biomarkers don’t necessarily need to be regulatory-critical but should help build an understanding of how the molecule affects certain mechanisms.

Connecting biology to patient needs

When focusing on biologically-based biomarkers, it is essential to ensure they link to unmet medical needs that matter to patients. A key part of this is finding connections between outcome measures and patient value to avoid developing molecules that target the right biology but fail to address meaningful patient needs. 

This approach requires close collaboration with patients and careful consideration of what endpoints will translate to real-world benefits. A biomarker that shows impressive biological activity but doesn’t correlate with symptom improvement or quality of life enhancement may have limited clinical utility.

The future: technology and AI in biomarker development

Looking ahead, harnessing technology and artificial intelligence will expand our ability to use biomarkers to de-risk drug development. AI tools can help validate more complex biomarkers that aren’t based on just one biological hypothesis, potentially uncovering patterns and relationships that might not be apparent through traditional analysis. 

Emerging technologies capable of monitoring movement, such as motor control and typing speed analysis, show promise for tracking disease progression in various neurological conditions. Additional, psychological markers like sleep/wake homeostasis – an important factor in almost all CNS diseases – are increasingly measurable through digital tools, including smartphones. Implementing these tools in drug development could provide invaluable insights into a molecule’s efficacy in real-world settings.

Learning from competitive intelligence

Beyond biomarkers, another key strategy involves learning from competitors’ experiences. The neuroscience field offers many lessons from past mistakes, such as the pursuit of incorrect pathways or targets. Being able to observe these experiences, understand failures, and make corresponding adjustments to your own programme is crucial for avoiding similar pitfalls. By learning from others’ setbacks, researchers can refine hypotheses, adjust biomarker strategies, and ultimately increase a programme’s chances of success which ultimately benefit patients by ensuring relevant clinical trials and faster development of next generation medicines.

Conclusion

As neuroscience R&D continues to evolve, biomarkers will play an increasingly central role in derisking development programs and improving success rates. By implementing structured biomarker strategies already in the research phase enabling exploration in early in development, selecting the right patients, connecting biological activity to meaningful outcomes, and leveraging new technologies, drug developers can address many of the historical challenges that have limited progress in this field. 

The transformation witnessed in Alzheimer’s disease biomarkers over the past decade offers hope that similar advances can be achieved across other CNS indications. With continued investment in biomarker research and development, the future of neuroscience R&D looks increasingly promising for patients awaiting breakthrough treatments for these devastating conditions. 

About the authors

Enchi Liu is Vice President, Head of Experimental Medicine at Lundbeck, overseeing clinical pharmacology, clinical biomarkers, and PKPD Modeling and Simulation. Her therapeutic experience spans CNS, neurodegeneration, and rare diseases development. She has extensive experience in implementation of biomarkers to support candidate product registration (CSF, blood, imaging, pharmacogenomics). 

Jesper Frank Bastlund is passionate for advancing mechanistic understanding and cross-species translational strategies to enable clinically meaningful therapeutic innovation. He has demonstrated leadership in shaping scientific strategy and fostering productive external collaborations with academia and biotech to accelerate pipeline progression.