Systems theory and systems biology paradigms
In order to achieve PM, innovative theoretical concepts and strategies need to be embraced. The traditional reductionis- tic approach in AD research aims to characterize single pathophysiological pathways affecting specific components of particular systems in a linear, non-dynamic and over-sim- plified manner. This myopic view has resulted in a limited representation of complex pathophysiological processes and their interactions. For instance, the prevailing amyloid cas- cade hypothesis speculates that the Ab peptide is the cause of AD and that, as a result, targeting Ab should lead to substantial disease modification at advanced clinical stages in LOAD. This assumption has been challenged by numerous failures of phase-III clinical trials aimed at modu- lating Ab production40 or increasing clearance from the brain41,42. Additionally, as stated previously, in no complex model of LOAD, Ab seems to occur in isolation of bio- logical dysfunction related to other systems. An agnostic, hypothesis-free, unbiased systems theory approach seems better suited to explain the complex and heterogeneous origin and time course of the pathophysiological failure underlying different forms of AD4. For multifactorial dis- eases like AD, comprehensive holistic systems-level approaches are necessary; this is the case of the SB model, which aims at understanding the genotype–phenotype rela- tionships and the mechanisms at the level of genome/epi- genome, transcriptome, microRNome, proteome/peptidome, metabolome/lipidome, microbiome, lifestyle, and environ- mental factors participating in complex cellular net- works3,4,43. Correspondingly, SB is based on: (1) advanced molecular and high-throughput ‘omics’ methods disclosing and characterizing biomarkers associated with disease mechanisms, and (2) computational and integrative network biology tools for assimilating multimodal information to comprehensively understand the systems-level dysfunc- tion44. Longitudinal investigations using the above-men- tioned SB-based methodologies can provide a full characterization of the complex molecular pathophysiology of both single gene and sporadic forms of AD. The working hypothesis is that most, if not all, AD subforms evolve through non-linear dynamic convergence of alterations and/ or failures in several ‘systems’, networks, signaling path- ways, or pathophysiological processes3. As a result, the spe- cific intervention needed for a particular individual would depend on the specific system-level alteration and/or dys- function at a given time point, which may change as a function of time and progression with (i.e. the specifically effective treatment for a given patient may vary over time).
CLIMACTERIC 5 ‘Omics’-based technologies for biomarker
PM is biomarker-guided medicine. According to the Food and Drug Administration (FDA) and the NIH Biomarkers, Endpoints, and other Tools (BEST) Resource, biomarker cate- gories can be divided into the following categories: (1) susceptibility/risk biomarker, (2) diagnostic biomarker, (3) monitoring biomarker, (4) prognostic biomarker, (5) pre- dictive biomarker, (6) pharmacodynamic/response biomarker, and (7) safety biomarker45. In the AD field, however, such fine-grained separation between different types of bio- markers is largely absent. For example, it is assumed that amyloid positivity is both a diagnostic and predictive bio- marker, which may or may not be the case for given patients. However, the fine-grained specifications of the exact function of each biomarker (or biomarker profile) is required to advance PM in AD45. When combining this fine-grained cat- egorization of specific types of biomarkers with the evolution of the ‘omic’ technologies currently available under the SB methods, there now exists the foundation for building the PM-based paradigm for treating and preventing AD across the spectrum of disease progression45. In genomics, the development of less expensive and comprehensive genome- wide arrays paved the way to the genome-wide association studies (GWAS). However, although initial results were prom- ising, numerous GWAS were disappointing due to inadequate sample size, limitation of arrays for certain genetic variations (genetic markers), and/or heterogeneity in phenotype46,47, as well as the focus on finding the gene(s) responsible for AD rather than looking for subsets of AD cases. Large collabora- tions, such as the International Genomics of Alzheimer’s Project (IGAP), and advanced genomic imputation techniques (in silico) generated highly consistent GWAS results48,49, which replicate and provide insights to underlying biological pathways. Notably, the introduction of next-generation sequencing (NGS)-based methods led to significantly improvement in the genomic analyses. Particularly, unbiased whole-genome sequencing (WGS) and whole-exome sequencing (WES) support the identification of many genetic variants, including SNPs, single nucleotide variants, small insertions/deletions, and structural and genomic variants50,51. Besides genomics, high-throughput screening methods led to substantial AD-related discoveries in other ‘omic’ areas, espe- cially proteomics52,53 and metabolomics/lipidomics54–56 that may change over time by contrast to the genome.
The ‘omics’-based screening of disease states is supposed to result in improved personalized, mechanistically based interventions (therapeutic and/or preventive) by revealing precise patterns of biomarkers and molecular signatures underlying the exact molecular pathophysiological mecha- nisms active in specific disease states and in individual patients57. Substantial attempts are ongoing to explicate key pathways functions, signaling network organization, and organism-level responses via high-throughput biological data (for instance, global gene expression, comprehensive prote- omic data)58.
Notably, applying SB to blood-based ‘omic’ technologies to promote the PM paradigm for AD will enable two primary
6 H. HAMPEL ET AL.
advances for improved patient outcomes13,59: (1) generation
and validation of enhanced multi-stage neurodiagnostic proc- esses, and (2) identification of targeted therapeutic interven- tion strategies for specific patients or subgroups of patients59. As with the PM paradigm successfully imple- mented in oncology, a primary key to success is the gener- ation of early detection biomarkers identifying patients before significant pathological accumulation. As with other frontline detection strategies, blood-based tools detecting patients within primary-care settings in the earliest stages of disease progress will foster a multi-stage diagnostic process for appropriate referrals to cerebrospinal fluid (CSF) and posi- tron emission tomography (PET) biomarker methods. Additionally, once such a multi-stage process is established, it would provide support to the global AD clinical trials com- munity. The second advancement will be the identification of which specific patients are most likely to benefit from precise and definite interventions. Applying SB for the analysis of multi-level blood-based ‘omic’ data will facilitate the segrega- tion of patient populations into biologically based subgroups that can be further scrutinized for targeted interventions. Using SB methods to create diagnostic biomarkers of specific subsets of AD patients will have a tremendous impact on the advancement of the PM paradigm in AD59.