The foundation and architecture of precision medicine in neurology and psychiatry

Harald Hampel, Peng Gao, Jeffrey L. Cummings, Nicola Toschi, Paul M. Thompson, Yan Hu, Min Cho, Andrea VergalloTrends 2023 · DOI

TL;DR

This paper proposes a new framework for precision medicine (PM) in neurology and psychiatry, moving beyond traditional symptom-based approaches to focus on early, preclinical stages of disease. It outlines a four-pillar architecture—biomarkers, systems medicine, digital health, and data science—to enable personalized, predictive, preventive, and participatory healthcare for complex brain disorders, using Alzheimer's disease as a prime example.

What Problem Does This Paper Address?

Neurological and psychiatric diseases are incredibly complex, characterized by significant genetic and biological variability even when symptoms appear similar. Current medical approaches often focus on late-stage symptoms, missing crucial opportunities for early intervention. This delay, coupled with the inherent heterogeneity of these conditions, leads to limited treatment efficacy and a substantial burden on healthcare systems. The paper aims to address this by proposing a paradigm shift towards precision medicine, enabling earlier detection, more targeted therapies, and ultimately, better patient outcomes by understanding the individual's unique biological and clinical profile.

Methods

The paper doesn't present a single experimental method but rather a conceptual framework. It advocates for an integrative approach combining:

  • Multimodal Biomarkers: Utilizing genetics, epigenomics, transcriptomics, proteomics, metabolomics, and neuroimaging (PET, SPECT, MRI, fMRI, EEG, MEG) to create a comprehensive biological profile.
  • Systems Medicine: Applying systems biology and neurophysiology to understand the complex interactions within biological networks and how their perturbation leads to disease. This involves mapping gene-to-phenotype and protein-to-phenotype connections.
  • Digital Health Technologies: Leveraging wearable devices, smartphones, and other digital tools for continuous, unobtrusive data collection on physiological and behavioral parameters, enabling real-time monitoring and early detection.
  • Data Science: Employing advanced computational methods, including artificial intelligence (AI) and machine learning (ML), to analyze vast, multidimensional datasets, identify patterns, stratify patient populations, and predict disease trajectories.

The paper emphasizes the integration of these pillars to create a "P4" medicine paradigm: predictive, participatory, preventive, and personalized.

Datasets

The paper discusses the use of various types of data, but does not detail specific datasets used for training or evaluation. It refers to:

  • Large-scale population genomic analyses (e.g., GWAS).
  • Multimodal neuroimaging datasets (e.g., from the ENIGMA Consortium).
  • Electronic Health Records (EHRs).
  • Data from digital health devices (wearables, smartphones).
  • Biobank data (e.g., UK Biobank, "All of Us" research program).

Key Insights and Ideas

  • Paradigm Shift to Preclinical Stages: The most critical insight is the call to shift focus from late-stage symptomatic treatment to targeting preclinical and early prodromal stages of neurological and psychiatric diseases, where interventions are likely to be most effective.
  • Four Pillars of PM Architecture: The proposed framework of biomarkers, systems medicine, digital health, and data science provides a structured, actionable approach to implementing PM in these complex fields.
  • Alzheimer's Disease as a Model: Highlighting AD as a case study demonstrates the feasibility and potential of this PM approach, given its advanced research in biomarkers and the urgent need for better treatments.
  • Heterogeneity is Key: The paper strongly emphasizes that understanding and addressing the inherent heterogeneity of brain disorders is paramount for developing effective therapies.
  • Integration is Crucial: The success of PM hinges on the seamless integration of diverse data types and analytical approaches, moving beyond siloed disciplines.

Key Results

As this is a conceptual paper proposing a framework, it does not present quantitative experimental results. However, it highlights the potential of the proposed approach:

  • Improved Disease Characterization: The framework aims to enable a more precise understanding of disease mechanisms by integrating genetic, molecular, and functional data.
  • Early Detection and Prevention: Digital health and advanced biomarkers are expected to facilitate earlier detection and the development of preventive strategies.
  • Personalized Treatment: By stratifying individuals based on their unique biological profiles, PM can lead to tailored therapeutic interventions.
  • Accelerated Drug R&D: The integration of AI and systems biology can streamline drug discovery and development by identifying novel targets and predicting treatment responses.

Limitations and Future Work

The authors acknowledge several limitations and challenges:

  • Complexity of CNS Physiology: Mapping clinical-biological trajectories in asymptomatic or early-stage individuals is challenging due to the intricate nature of CNS physiology and the simultaneous disruption of processes across multiple biological levels.
  • Data Standardization and Curation: The need for consistent curation and harmonization of large, diverse datasets is critical for the successful application of AI and systems biology approaches.
  • Explainability of AI Models: The "black box" nature of some AI models limits understanding of how they arrive at their conclusions, hindering clinical trust and adoption.
  • Cost and Complexity of Techniques: Advanced neuroimaging techniques, while powerful, are often costly and complex, limiting widespread clinical implementation.
  • Accessibility and Equity: Algorithms must consider socioeconomic factors and healthcare access to ensure equitable application of PM.
  • Transferability from Oncology: Direct tissue accessibility in oncology is not always mirrored in neurology and psychiatry, requiring adapted strategies.

Future work should focus on:

  • Validating and standardizing systems-level integration approaches.
  • Developing effective AI-driven drug R&D strategies for pathway-based, symptoms-agnostic therapies.
  • Establishing implementation strategies for PM preventive and diagnostic-therapeutic approaches to ensure healthcare system preparedness.
  • Embracing human diversity and ensuring inclusion in PM development.

Summary updated

Abstract

Neurological and psychiatric diseases have high degrees of genetic and pathophysiological heterogeneity, irrespective of clinical manifestations. Traditional medical paradigms have focused on late-stage syndromic aspects of these diseases, with little consideration of the underlying biology. Advances in disease modeling and methodological design have paved the way for the development of precision medicine (PM), an established concept in oncology with growing attention from other medical specialties. We propose a PM architecture for central nervous system diseases built on four converging pillars: multimodal biomarkers, systems medicine, digital health technologies, and data science. We discuss Alzheimer's disease (AD), an area of significant unmet medical need, as a case-in-point for the proposed framework. AD can be seen as one of the most advanced PM-oriented disease models and as a compelling catalyzer towards PM-oriented neuroscience drug development and advanced healthcare practice.
BibTeX
@inproceedings{hampel2023foundation,
  author = {Harald Hampel and Peng Gao and Jeffrey L. Cummings and Nicola Toschi and Paul M. Thompson and Yan Hu and Min Cho and Andrea Vergallo},
  title = {{The foundation and architecture of precision medicine in neurology and psychiatry}},
  booktitle = {Trends},
  year = {2023},
  doi = {10.1016/j.tins.2022.12.004},
}
Figure