Personalized Psychiatry

Bernhard Baune, Bernhard Baune

• Editorial: Academic Press
• Año: 2019
• ISBN:
  9780128131770
• Categoría: Biological Sciences
• Categoría: Neuroscience

Personalized Psychiatry presents the first book to explore this novel field of biological psychiatry that covers both basic science research and its translational applications. The book conceptualizes personalized psychiatry and provides state-of-the-art knowledge on biological and neuroscience methodologies, all while integrating clinical phenomenology relevant to personalized psychiatry and discussing important principles and potential models. It is essential reading for advanced students and neuroscience and psychiatry researchers who are investigating the prevention and treatment of mental disorders.

• Combines neurobiology with basic science methodologies in genomics, epigenomics and transcriptomics
• Demonstrates how the statistical modeling of interacting biological and clinical information could transform the future of psychiatry
• Addresses fundamental questions and requirements for personalized psychiatry from a basic research and translational perspective

Chapter 1

What is personalized psychiatry and why is it necessary?

Bernhard T. Baune    Department of Psychiatry, University of Münster, Münster, Germany
Department of Psychiatry, Melbourne Medical School, The University of Melbourne, Melbourne, VIC, Australia
The Florey Institute of Neuroscience and Mental Health, The University of Melbourne, Parkville, VIC, Australia

Abstract

Personalized medicine, often interchangeably referred to as precision or individualized medicine, is an emerging approach for disease classification and treatment stratification as well as for the prevention of illness that takes into account individual variability in genes, environment, and lifestyle. Following the personalized medicine approach, patients are separated or stratified into different groups—with medical decisions, practices, interventions, and/or products being tailored to the individual patient based on their predicted response to treatment or risk of disease. Following these underlying principles, personalized psychiatry is an emerging field within medicine with promising prospects.

Keywords

Environment; Psychiatry; Diagnoses; Treatment; Stratification; Risk of disease

Personalized medicine, often interchangeably referred to as precision or individualized medicine, is an emerging approach for disease classification and treatment stratification as well as for the prevention of illness that takes into account individual variability in genes, environment, and lifestyle. Following the personalized medicine approach, patients are separated or stratified into different groups—with medical decisions, practices, interventions, and/or products being tailored to the individual patient based on their predicted response to treatment or risk of disease.

Most of the current treatments are approved and developed on the basis of their performance in a large population of people and each treatment is prescribed to all patients with a certain diagnosis. However, psychiatry is now developing personalized solutions for a particular patient’s needs that will become more readily available. In case of complex psychiatric disorders, the conventional “one-drug-fits-all” approach involves trial and error before appropriate treatment is found. Clinical trial data for new treatments merely show the average response of a study group. Given the considerable individual clinical and biological variation, some patients show no response whereas others show dramatic response. Although approximately 99.9% of our DNA sequence is identical, the 0.1% difference between any two individuals (except identical twins) is medically significant. Within this small percentage of difference lie the clues to hereditary susceptibility to psychiatric disorders. At the DNA level, this 0.1% difference translates into 3 million sites of genomic variation.

It is obvious that the concept of “one medicine fits all patients with the same disease” does not hold true and a more individualized approach is needed. The aim of personalized psychiatry is to match the right treatment to the right patient at the right time, and in some cases even to design the treatment for a patient according to genotype and other individual biological, psychological, or environmental characteristics.

Personalized medicine in the context of psychiatry seeks to identify environmental, biological, and clinical factors that contribute to disease vulnerability, the onset and course of mental disorders, and the predicted response to pharmacological and nonpharmacological interventions. In psychiatry, this task is complicated by the syndromic nature of diagnoses. These diagnostic categories, as represented in the Diagnostic and Statistical Manual of Mental Disorders (DSM), have been developed without a detailed understanding of underlying biological mechanisms. While they have provided the basis for the development of the science of psychiatry, they are not sufficient or reliable descriptions of an individual’s clinical presentation, response to treatment, illness, or functional trajectories, resulting in uncertainty in psychiatric diagnosis and prediction of treatment outcomes. Clinically, this may lead to a “wait and watch” approach where multiple trials of different medications with unnecessary side effects or nonefficient nonpharmacological interventions become an additional burden due to poor efficacy.

New research and clinical approaches in psychiatry have become necessary. These should aim to combine advances from clinical phenotyping, biology, neuroscience, and bioinformatic methodologies into a single but complex approach to seek better understanding of complex mental illnesses, better translate research findings into clinical practice, and stimulate reverse translation from clinical to basic science research.

By applying a personalized psychiatry approach, it is not proposed that outcomes are completely predetermined at disease onset, but that structured assessment and modeling of multivariate cross-sectional and longitudinal predictors can be used to describe the risk of specific outcome trajectories and to predict response to specific treatments. As a result, such a predictive approach could aid decision-making processes in clinical psychiatry. For this purpose, predictor variables are grouped into key psychiatric systems that can be differentiated into (a) individual clinical characteristics of a patient, including risk factors such as maternal pregnancy complications, early neurodevelopmental history, and socioeconomic status; (b) their neurocognitive, affective, and functional profile; (c) brain structure and neural function; (d) molecular profile; and (e) modulating prognostic factors such as personality, insight, and resilience. In a complex modeling approach, data from these systems are modeled to derive descriptions of illness trajectories and outcomes. In a translational next step, structured assessments feed into computerized prognostic models that are used to determine the best current treatment based on the most likely response and illness trajectory. The basic idea of these exciting developments in the field of psychiatry is conceptualized in this book.

The concept of personalized psychiatry that is presented here brings the four pillars of personalized medicine to the field of psychiatry: Prevention, Prediction, Personalization and Participation. With this approach, the fundamental questions and requirements for personalized psychiatry from a research as well as a clinical perspective will be addressed. The concept includes knowledge and extensive developments in basic science methodologies (e.g., genomics, epigenomics, transcriptomics) and neurobiological and clinical conceptual underpinnings of clinical disease relevant for personalized psychiatry. Importantly, the key principles of translation into clinical practice for the individual patient are embedded into personalized clinical trials and in complex prediction modeling. A personalized approach to mental health requires extensive developments in statistical modeling to decipher the complex interacting biological and clinical information as well as the development of a digitalized measurement-based assessment of diagnosis and treatment progression, response to treatment, and trajectories of illness. Taken together, these could transform clinical practice in psychiatry.

Chapter 2

The modeling of trajectories in psychotic illness

Scott R. Clark^a; Klaus Oliver Schubert^a,b; Bernhard T. Baune^{c,d,e a} Discipline of Psychiatry, University of Adelaide, Adelaide, SA, Australia
^b Northern Adelaide Local Health Network, Mental Health Service, Adelaide, SA, Australia
^c Department of Psychiatry and Psychotherapy, University of Münster, Münster, Germany
^d Department of Psychiatry, Melbourne Medical School, The University of Melbourne, Melbourne, Australia
^e The Florey Institute of Neuroscience and Mental Health, The University of Melbourne, Parkville, VIC, Australia

Abstract

The natural history of psychotic illness is one of fluctuating symptoms and functions, from prodromal illness, to first episode, to episodic relapse, and recovery or chronic illness. Recent evidence suggests that complex biological, symptomatic, and functional trajectory patterns underlie clinical presentations and outcomes. Statistical techniques for the analysis of longitudinal data have evolved to provide better estimates of trends and the determinants of growth and change for individual patients. The analysis of complex multimodal trajectories may be key to the personalization of psychosis treatment. This chapter will discuss the advantages and disadvantages of trajectory modeling techniques for psychosis research.

Keywords

Psychosis; Trajectory; Longitudinal; General linear model; Bayesian; Mixed modeling; Latent class analysis; Growth mixture modeling

1 Introduction

The natural history of psychotic illness is one of fluctuating symptoms and function. Its course may take a number of paths through prodromal illness, from first episode, to recovery; or though episodic relapse, or chronic illness (Clark, Schubert, & Baune, 2015; Schubert, Clark, & Baune, 2015). For some cases, these trajectories may be clinically intuitive; for example, a poor prognosis for the patient with first presentation psychosis, prominent negative symptoms, and cognitive impairment. However, recent studies have shown that complex multiple trajectory patterns layered from physiological to phenomenological, and overlapping across time, may underlie higher order phenotypic presentations. For example, brain indices of gray and white matter volume and connectivity evolve in different patterns across childhood and youth in patients that develop schizophrenia or bipolar disorder in comparison with siblings and controls (Baker, Holmes, Masters, et al., 2014; Katagiri, Pantelis, Nemoto, et al., 2015; Liberg, Rahm, Panayiotou, et al., 2016; Ordonez, Luscher, & Gogtay, 2016; Schmidt, Crossley, Harrisberger, et al., 2016). Consequently, the identification of key illness and functional trajectory signatures from longitudinal data may be critical to improve the prediction of long-term illness prognosis, facilitating personalized treatment of psychosis (Clark et al., 2015; Schubert et al., 2015).

A trajectory can be simply defined as the pattern of change in a measure of interest over time. Trajectories can be easily visualized by plotting mean values or raw data by time, and these plots can be used to grossly determine between subject variability and trend in the measure of interest (Wu, Selig, & Tood, 2013; Xiu, 2015). Simple calculations can provide values for the mean change between baseline and follow up, or the mean slope (change per unit time) where there are two or more time points. These crude analyses are suitable for exploration, but fail to capture extremes and patterns of fluctuation that are likely to be important in the prediction of long-term outcomes in mental illness (Adamis, 2009). For example, Fig. 1A reproduced from Clark et al. (2015) illustrates four possible trajectories of psychotic symptoms for those at clinical high risk (CHR) of psychosis. Patients may oscillate between CHR and normal status, may remain at CHR risk without transition, or may take a standard path from CHR to first episode psychosis. In a large clinical trial involving people at CHR of psychosis, Polari, Lavoie, Yuen, et al. (2018) noted 17 different trajectories in a 12-month period, which were classified as: recovery (35.7%), remission (7.5%), any recurrence (20%), no remi...