Influence and interaction of genetic, cognitive, neuroendocrine and personalistic markers to antidepressant response in Chinese patients with major depression.

ABSTRACT

OBJECTIVE: Despite there is a wide range of antidepressants available, with various mechanisms of actions, the efficacy of current therapeutic options is yet satisfactory. Previous shreds of evidence have indicated that genetics, cognitive, neuroendocrine, as well as personality factors, are all intrinsically linked and contribute to the diversity of treatment outcomes. We, therefore, sought to investigate this hypothesis in this study. METHOD: Based on 610 samples treated with a selection of serotonin reuptake inhibitor (SSRI), serotonin-norepinephrine reuptake inhibitors (SNRI), noradrenergic and specific serotonergic antidepressant (NaSSA) or tricyclic antidepressant (TCA), we compared the therapeutic effects of these four classes of drugs by survival analyses. Pharmacogenomic and survival analyses were carried out to explore the hereditary factors for curative effect and the accumulation of genetic factors was further discussed through pathway analysis and the global test. We built a machine learning-based prediction model that integrates genetic and non-genetic factors (including cognition, endocrinology, personality intelligence) to distinguish drug efficacy in single class drug situations. The values of the non-genetic makers after 6 weeks' treatment were collected to evaluate the efficacy of the model. RESULTS: Our results from the 6-week antidepressant therapeutic study indicated that SSRI and SNRI are better treatments than those of TCA and NaSSA in the Chinese population. Among all possible paired single-agent survival analyses, citalopram and venlafaxine were more effective than mirtazapine. Allele C carriers at rs6354 (SLC6A4) and allele G carriers at rs12150214 (SLC6A4) were significantly prone to poorer treatment response to fluoxetine. Besides, the combination of three loci (rs929377-rs6191-rs32897) located in HPA pathway was significantly associated with the treatment outcome of fluoxetine. In female MDD patients, the minor allele of rs6323 and rs1137070 on the MAOA gene likely lead to a worse response to venlafaxine. Furthermore, genetic variants linked to drug efficacy tended to concentrate on the neurotrophin pathway in depressed patients comorbid with anxiety. From multivariate models, more severe cognitive deficits, psychopathic personality and lower levels of operational intelligence, and higher levels of cortisol predicted worse response status with SSRI or SNRI after 6-week treatment. Notably, genetic factors in the multi-dimensional prediction model for both classes of drugs include loci in HTR2A and CRHBP genes. CONCLUSION: SSRI and SNRI are more suitable for the treatment of Chinese people with depression. SLC6A4 genetic variants, as well as HPA pathway, play an important role in the fluoxetine antidepressant therapeutic response while the polymorphism of MAOA gene involved in the pharmacological action of venlafaxine among female MDD patients. The presence of anxiety in MDD patients was related to the neurotrophin pathway. Genetic, cognitive, neuroendocrine, and personality intelligence factors combined have an ensemble impact on the medication effect of patients with major depression, leading to more precise and personalized medicine for specific groups of people.