Abstract
Major depressive disorder (MDD) implicates a huge burden for patients and
society. Although currently available antidepressants are effective treatment
options, more than 50% of the patients do not respond to the first
administered antidepressant. In addition, in more than 25% with
antidepressants-treated patients, adverse effects occur. Currently, the
selection of treatment does not reflect objectively measurable data from
neurobiological and behavioral systems. However, in the last decades, the
understanding of the impact of genetic variants on clinical features such as
drug metabolism has grown and can be used to develop tests that enable a
patient-tailored individual treatment. In fact, robust evidence was found that
genetic variants of CYP450 enzymes such as CYP2D6 and CYP2C19 can be surrogate
markers for the metabolism of certain drugs. This article describes a pilot
study design aimed to combine clinical variables such as therapeutic drug
monitoring, inflammatory and stress markers with static and variable genetic
information of depressed patients to develop an algorithm that predicts
treatment response, and tolerability using machine learning algorithms.
Psychometric evaluation covers the Hamilton Depression Rating Scale, the
Childhood Trauma Questionnaire, and adverse drug reactions. An in-depth
(epi-)genetic assessment combines genome-wide gene association data with DNA
methylation patterns of genes coding CYP enzymes along with a pharmacogenetic
battery focusing on CYP enzymes. Using these measures to stratify depressed
patients, this approach should contribute to a data-driven assessment and
management of MDD, which can be referred to as precision medicine or
high-definition medicine.
Key words
pharmacogenetics - precision medicine - personalized medicine - biomarker - depression - antidepressants - machine learning
