Are innovation and new technologies in precision medicine paving a new era in patients centric care?

Abstract

Healthcare is undergoing a transformation, and it is imperative to leverage new technologies to generate new data and support the advent of precision medicine (PM). Recent scientific breakthroughs and technological advancements have improved our understanding of disease pathogenesis and changed the way we diagnose and treat disease leading to more precise, predictable and powerful health care that is customized for the individual patient. Genetic, genomics, and epigenetic alterations appear to be contributing to different diseases. Deep clinical phenotyping, combined with advanced molecular phenotypic profiling, enables the construction of causal network models in which a genomic region is proposed to influence the levels of transcripts, proteins, and metabolites. Phenotypic analysis bears great importance to elucidat the pathophysiology of networks at the molecular and cellular level. Digital biomarkers (BMs) can have several applications beyond clinical trials in diagnostics-to identify patients affected by a disease or to guide treatment. Digital BMs present a big opportunity to measure clinical endpoints in a remote, objective and unbiased manner. However, the use of "omics" technologies and large sample sizes have generated massive amounts of data sets, and their analyses have become a major bottleneck requiring sophisticated computational and statistical methods. With the wealth of information for different diseases and its link to intrinsic biology, the challenge is now to turn the multi-parametric taxonomic classification of a disease into better clinical decision-making by more precisely defining a disease. As a result, the big data revolution has provided an opportunity to apply artificial intelligence (AI) and machine learning algorithms to this vast data set. The advancements in digital health opportunities have also arisen numerous questions and concerns on the future of healthcare practices in particular with what regards the reliability of AI diagnostic tools, the impact on clinical practice and vulnerability of algorithms. AI, machine learning algorithms, computational biology, and digital BMs will offer an opportunity to translate new data into actionable information thus, allowing earlier diagnosis and precise treatment options. A better understanding and cohesiveness of the different components of the knowledge network is a must to fully exploit the potential of it.

Keywords: Artificial intelligence; Autoimmune and inflammatory diseases; Biomarkers; Cancer; Deep phenotyping; Diabetes; Digital biomarkers; Epigenetics; Genetics; Genomics; Immuno-oncology; Machine learning; Modeling and simulation; Personalized medicine; Precision medicine; Proteomics; Transcriptomics; miRNAs; microRNAs.

Fig. 1

Critical checkpoints for host and tumor profiling. A multiplexed biomarker approach is highly integrative and includes both tumor- and immune-related parameters assessed with both molecular and image-based methods for individualized prediction of immunotherapy response. By assessing patient samples continuously one can collect a dynamic data on tissue-based parameters, such as immune cell infiltration and expression of immune checkpoints, and pathology methods. These parameters are equally suited for data integration with molecular parameters. TILs: tumor-infiltrating lymphocytes. PD-L1: programmed cell death-ligand 1. Immunoscore: a prognostic tool for quantification of in situ immune cell infiltrates. Immunocompetence: body’s ability to produce a normal immune response following exposure to an antigen (tumor drawing has been adapted from [42])

Fig. 2

The cancer immunogram. The schema depicts the seven parameters that characterize aspects of cancer-immune interactions for which biomarkers have been identified or are plausible. Italics represent those potential biomarkers for the different parameters (adapted from [4])

Fig. 3

Schematic of an integrated biologic information for a targeted therapeutic intervention. Ag, antigen; BETi, inhibitors of bromodomain and extraterminal proteins; carbo, carboplatin; CSF1, colony stimulating factor 1; CFM, cyclophosphamide; CTLA-4, cytotoxic T-lymphocyte-associated antigen 4; HDAC, histone deacetylase; HMA, hypomethylating agents; IDO, indoleamine 2,3-dioxyenase; IO, immune-oncology; LN, lymph nodes; LAG-3, lymphocyte-activation gene 3; MDSC, myeloid-derived suppressor cells; P13K, phosphoinositide 3-kinase; PD-1, programmed cell death-1; PD-L1, programmed cell death-ligand 1; STING, stimulator of interferon genes; TIM3, T cell immunoglobulin and mucin domain 3; TME, tumor microenvironment; Treg, regulatory T cells; TLR, toll-like receptor; Wnt, wingless, int-1 (adapted from [3, 42])

Fig. 4

Schematic of a comprehensive biomedical knowledge network that supports a new taxonomy of disease. The knowledge network of disease would incorporate multiple parameters rooted in the intrinsic biology and clinical patient data originating from observational studies during normal clinical care feeding into Information Commons which are further linked to various molecular profiling data enabling the formation of a biomedical information network resulting in a new taxonomy of disease. Information Commons contains current disease information linked to individual patients and is continuously updated by a wide set of new data emerging though observational clinical studies during the course of normal health care. The data in the Information Commons and Knowledge Network provide the basis to generate a dynamic, adaptive system that informs taxonomic classification of disease. This data may also lead to novel clinical approaches such as diagnostics, treatments, prognostics, and further provide a resource for new hypotheses and basic discovery. At this intersection, artificial intelligence and machine learning may help to analyze this highly complex large dataset by pattern recognition, feature extraction yielding Digital BMs. Validation of the findings that emerge from the Knowledge Network, such as those which define new diseases or subtypes of diseases that are clinically relevant (e.g. which have implications for patient prognosis or therapy) can then be incorporated into the New Taxonomy of disease to improve diagnosis (i.e. disease classification) and treatment. This multi-parametric taxonomic classification of a disease may enable better clinical decision-making by more precisely defining a disease (adapted from [72])