Integrating Genetic Variants and Expression Profiles of Pharmacogenes in Treatment-Resistant Depression

ClaudiaPisanu1

AlessioSquassina1,2,20✉Emailsquassina@unica.it

JúliaPerera-Bel3

RosanaCarvalhoSilva4

LisaBuson5

AnnaMartinezSires3

MarcoBortolomasi6

ValentinaMenesello4,5

GiuliaPerusi7

BernardoCarpiniello8

EwaFerensztaj-Rochowiak9

FilipRybakowski9

FerranSanz3,10

MirkoManchia8,11

MarieClaudePotier12

MaraDierssen13,14,15

BernhardT.Baune16,17,18

MassimoGennarelli4,5

AlessandraMinelli4,5,19✉Phone+39 030 3717255Emailalessandra.minelli@unibs.it

1Section of Neuroscience and Clinical Pharmacology, Department of Biomedical SciencesUniversity of CagliariCagliariItaly

2Department of PsychiatryDalhousie UniversityHalifaxNSCanada

3Hospital del Mar Medical Research Institute (HMRIB)BarcelonaSpain

4Department of Molecular and Translational MedicineUniversity of BresciaBresciaItaly

5Genetics UnitIRCCS Istituto Centro San Giovanni di Dio FatebenefratelliBresciaItaly

6Psychiatric Hospital “Villa Santa Chiara”VeronaItaly

7Department of Mental Health and Addiction ServicesASST Spedali Civili of BresciaBresciaItaly

8Section of Psychiatry, Department of Medical Sciences and Public HealthUniversity of CagliariCagliariItaly

9Department of Adult PsychiatryPoznan University of Medical SciencesPoznanPoland

10Department of Medicine and Life SciencesUniversitat Pompeu FabraBarcelonaSpain

11Department of PharmacologyDalhousie UniversityHalifaxNSCanada

12Paris Brain Institute (ICM)National Centre for Scientific Research (CNRS)ParisFrance

13Centre for Genomic Regulation (CRG)The Barcelona Institute for Science and Technology (BIST)BarcelonaSpain

14Universitat Pompeu Fabra (UPF)BarcelonaSpain

15Biomedical Research Networking Center for Rare Diseases (CIBERER)08003BarcelonaSpain

16Department of PsychiatryUniversity of MünsterMünsterGermany

17Florey Institute of Neuroscience and Mental HealthParkvilleVICAustralia

18Department of PsychiatryUniversity of MelbourneParkvilleVICAustralia

19Department of Molecular and Translational Medicine, Biology and Genetic DivisionUniversity of BresciaViale Europa, 1125123BresciaItaly

20Department of Biomedical SciencesUniversity of Cagliarisp 809042Monserrato, CagliariItaly

Claudia Pisanu^1,δ, Alessio Squassina^1,2,δ*, Júlia Perera-Bel³, Rosana Carvalho Silva⁴, Lisa Buson⁵, Anna Martinez Sires³, Marco Bortolomasi⁶, Valentina Menesello^4,5, Giulia Perusi⁷, Bernardo Carpiniello⁸, Ewa Ferensztaj-Rochowiak⁹, Filip Rybakowski⁹, Ferran Sanz^3,10, Mirko Manchia^8,11, Marie Claude Potier¹², Mara Dierssen^13,14,15, PROMPT Study group, Bernhard T. Baune^16,17,18, Massimo Gennarelli^4,5, Alessandra Minelli^4,5*

^δ Shared first authorship

¹ Section of Neuroscience and Clinical Pharmacology, Department of Biomedical Sciences, University of Cagliari, Cagliari, Italy

² Department of Psychiatry, Dalhousie University, Halifax, NS, Canada

³ Hospital del Mar Medical Research Institute (HMRIB), Barcelona, Spain

⁴ Department of Molecular and Translational Medicine, University of Brescia, Brescia, Italy

⁵ Genetics Unit, IRCCS Istituto Centro San Giovanni di Dio Fatebenefratelli, Brescia, Italy

⁶ Psychiatric Hospital “Villa Santa Chiara”, Verona, Italy

⁷Department of Mental Health and Addiction Services, ASST Spedali Civili of Brescia, Brescia, Italy

⁸Section of Psychiatry, Department of Medical Sciences and Public Health, University of Cagliari, Cagliari, Italy

⁹Department of Adult Psychiatry, Poznan University of Medical Sciences, Poznan, Poland

¹⁰Department of Medicine and Life Sciences, Universitat Pompeu Fabra, Barcelona, Spain

¹¹Department of Pharmacology, Dalhousie University, Halifax, NS, Canada

¹²Paris Brain Institute (ICM), National Centre for Scientific Research (CNRS), Paris, France

¹³Centre for Genomic Regulation (CRG), The Barcelona Institute for Science and Technology (BIST), Barcelona, Spain

¹⁴Universitat Pompeu Fabra (UPF), Barcelona, Spain

¹⁵Biomedical Research Networking Center for Rare Diseases (CIBERER), Barcelona 08003, Spain

¹⁶Department of Psychiatry, University of Münster, Münster, Germany

¹⁷Florey Institute of Neuroscience and Mental Health, Parkville, VIC, Australia

¹⁸Department of Psychiatry, University of Melbourne, Parkville, VIC, Australia

• Corresponding authors:

Minelli Alessandra – Department of Molecular and Translational Medicine, Biology and Genetic Division, University of Brescia, Viale Europa, 11–25123 Brescia, Italy. Tel.: +39 030 3717255 E-mail address: alessandra.minelli@unibs.it

Alessio Squassina – Department of Biomedical Sciences - University of Cagliari, sp 8, 09042 Monserrato, Cagliari Italy E-mail address: squassina@unica.it

Abstract

Background. Treatment-resistant depression (TRD) is a major clinical challenge in the management of major depressive disorder (MDD). While pharmacogenetics has been suggested to have some clinical utility in guiding antidepressant treatment, only few studies have explored if and how pharmacogenes can be involved in TRD pathophysiology and its clinical outcomes. In this study we explored the role of differences in metabolizer phenotypes, gene expression levels, and microRNAs of three key pharmacogenes (CYP2D6, CYP2C19, and CYP2B6) in TRD pathophysiology and antidepressant response in a cohort of 300 patients with MDD from the PROMPT consortium. Results. CYP2D6 phenotype distribution did not differ significantly between TRD and non-TRD groups, but mRNA expression was significantly upregulated in TRD patients (log fold change [logFC] = 0.35, adj p = 0.0002). Additionally, hsa-miR-26b-5p, a microRNA predicted to regulate CYP2D6, was significantly downregulated in TRD (logFC = -0.62, adj p = 2.5E-05). For CYP2C19, intermediate metabolizers (IM) were underrepresented in TRD compared to non-TRD individuals (IM vs normal metabolizers (NM): χ2 = 6.07, p = 0.019). microRNA hsa-let-7d-5p and hsa-miR-27a-3p, predicted to regulate CYP2C19, were significantly downregulated in TRD (logFC = -0.31, adj p = 0.002; logFC = -0.38, adj p = 0.0008). No significant differences were found for CYP2B6. In the analysis of remission to the last antidepressant trial, only nominally significant differences were reported, which did not survive multiple testing corrections. Conclusions. By investigating high levels of evidence pharmacogenes, this study contributes valuable insights to the PROMPT project on how pharmacokinetic gene variants, their expression and regulatory mechanisms may influence antidepressant response and resistance in MDD.

Keywords:

pharmacogenetics

precision psychiatry

CYP2D6

CYP2C19

antidepressants

treatment-resistant depression

mRNA

miRNA

1. Background

Major depressive disorder (MDD) is the most common psychiatric disorder, associated with significant functional impairment, reduced quality of life, and socioeconomic impact. It is characterized by persistent low mood, loss of interest in previously enjoyable activities, and often includes low self-esteem, low energy, and feelings of sadness or hopelessness without a clear cause (1). Diagnosing and managing MDD is challenging due to its clinical heterogeneity, variable clinical course, and inconsistent treatment response. Pharmacological treatment is typically the first-line approach, involving various classes of antidepressants. However, only about 30% of patients achieve remission with the first pharmacological trial, and approximately 15–30% develop treatment-resistant depression (TRD) (2, 3). Treatment failure is further complicated by side effects and tolerability issues, which lead many patients to discontinue medication (4, 5). Finding an effective treatment often requires multiple trials, with the probability of success decreasing with each unsuccessful attempt, contributing to prolonged symptoms, worse long-term prognosis, and increased personal and societal costs (6).

This high variability in antidepressant effectiveness is influenced by both genetic and non-genetic factors, and while the role of genomic variations has yet to be elucidated, in the last two decades the pharmacogenetics of response to antidepressants has been increasingly investigated. Indeed pharmacogenetics can help improve our understanding of how patent-specific genetic-driven variability in pharmacokinetics and pharmacokinetics of drugs can influence efficacy and safety, leading to personalized treatments and reducing the risk of treatment discontinuation, especially in patients with TRD. Despite the role of genetics has been consolidated over the years, only few genes involved in drug metabolism have been reported to significantly impact on treatment outcomes and adverse reactions in antidepressant treatment (7). Indeed, it has been shown that a small group of genes, particularly those coding for cytochromes CYP2D6, CYP2C19, and CYP2B6, which play a central role in the metabolism of antidepressants, significantly influence response and side effects to these medications, but their association with TRD has been inadequately studied (7, 8).

Traditionally, pharmacogenetics has relied on the “diplotype-inferred metabolizer phenotype”, where the presence of specific genetic variants is used to predict enzyme activity and guide dosing. Indeed, the majority of previous studies investigating these pharmacogenes in psychiatric disorders have focused on genetically determined metabolizer phenotypes, without incorporating other molecular measures possibly impacting on cytochromes (CYP) metabolizing properties, such as the expression of the CYP genes or epigenetic mechanisms involved in modulating their transcription (7). Differences in the expression and splicing of genes encoding drug-metabolizing enzymes, transporters, and targets, such as receptors and ion channels, have been associated with interindividual variability in optimal drug dosing, therapeutic effectiveness, and side effects (9). Nevertheless, this has been only limitedly explored in relation to antidepressant treatment in MDD. Recent findings suggest that the transcriptional regulation of pharmacogenes, including variability in mRNA expression and modulation by microRNAs (miRNA), may significantly influence drug metabolism independently of genetic diplotypes (9–11). This suggests that the expression of pharmacogenes, such as CYP2D6 and CYP2C19, can vary within the same genotype-predicted metabolizer groups, potentially altering antidepressant efficacy. Moreover, it has been shown that miRNAs may affect both pharmacokinetics and dynamics of antidepressant as well as play a role in the pathophysiology of depression itself (12). Nonetheless, studies integrating genotyping and expression data in the pharmacogenetics of antidepressants remain scarce, especially in patients with TRD.

Here, we explored the hypothesis that patients with TRD might exhibit unique genotypic and gene expression profiles which could affect drug metabolism. Specifically, the safety and effectiveness of antidepressants may be influenced by the expression levels of genes involved in their pharmacokinetics, which, in the case of CYP enzymes, may not necessarily correspond to genotype status. The expression of these genes can be modulated by specific miRNAs, and studying these regulators will provide deeper insights into the pharmacogenetics of antidepressants, moving beyond the traditional pharmacogenomic model based on the genotype-based metabolizer status.

In this study, explored the role of the pharmacogenes CYP2D6, CYP2C19, and CYP2B6, which present the highest level of evidence of clinical utility for antidepressants as reported in ClinPGX, in in TRD and in remission to antidepressants, in a cohort of 300 patients with MDD collected within the PROMPT consortium (13). Specifically, we investigated: 1) whether patients with TRD show a different distribution of diplotypes and phenotypes at the studied pharmacogenes compared to non-TRD individuals; 2) whether the expression of these pharmacogenes varies between TRD and non-TRD groups; 3) whether the expression levels of miRNAs regulating these pharmacogenes significantly differ between TRD and non-TRD patients. By examining the most relevant pharmacogenes in a large sample of patients and incorporating gene and miRNA expression analyses, this study ultimately aims to improve the understanding of how pharmacogenes contribute to TRD pathophysiology and its clinical outcomes.

2. Methods

2.1 Study participants and clinical assessment

This study included 300 patients with MDD from a retrospective cohort (IRCCS Fatebenefratelli, Brescia, Italy) participating in the phase 1 PROMPT project (Toward PrecisiOn Medicine for the Prediction of Treatment response in major depressive disorder through stratification of combined clinical and -omics signatures; (13)). The diagnostic criterion for inclusion was a diagnosis of moderate to severe MDD according to the Diagnostic and Statistical Manual of Mental Disorders-IV (DSM-IV) classification system. The diagnosis of MDD was confirmed for all participants using the Italian version of the SCID-I diagnostic scale according to DSM-IV criteria. Personality disorders were diagnosed based on clinical symptom evaluation consistent with DSM-IV standards. Exclusion criteria included: (a) a lifetime history of schizophrenic, schizoaffective, or bipolar disorder; (b) primary diagnosis of personality disorder, substance abuse, alcohol abuse or dependency, obsessive-compulsive disorder (OCD), or post-traumatic stress disorder (PTSD); (c) comorbidity with eating disorders; (d) comorbidity with alcohol and substance dependence; (e) intellectual disability and cognitive impairment; (f) neurological disorders such as Parkinson’s disease, multiple sclerosis, Alzheimer’s disease and other dementias, epilepsy, strokes, brain tumors, and traumatic conditions of the nervous system; (g) comorbidity with other severe medical illnesses and severe autoimmune diseases such as cancers, Crohn’s disease, rheumatoid arthritis, scleroderma, psoriasis, myasthenia gravis, Sjögren syndrome, and systemic lupus erythematosus; (h) pregnancy.

Half of the patients (150) were classified as having TRD, while the other 150 were classified as non-TRD. Based on clinical judgment by the treating psychiatrists, TRD was defined as a failure of treatment to produce response or remission for patients after two or more treatment attempts of adequate and recommended dose and duration. Most TRD patients had a long-standing history of the condition and had undergone ECT or other intensive treatments, including drug combinations, augmentation with antipsychotics and/or mood stabilizers, intensive psychotherapy, repetitive transcranial magnetic stimulation, and sleep deprivation. Whereas MDD patients were classified as non-TRD when they achieved response or remission in terms of a reduction in symptomatology with the current antidepressant treatment attempt of adequate dose and duration. All non-TRD patients were either at their first MDD episode, or they had other episode/s in which they responded to the treatment received or had only one failure in their pharmacological anamnesis history.

For a subgroup of patients (N = 228), response to treatment in the current episode was available as assessed with the Montgomery-Åsberg Depression Rating Scale (MADRS) after 8 weeks. This data coming from naturalistic observational cohorts in real-world conditions, allowed us to classify TRD and non-TRD patients as remitters and non-remitters. Remission was defined as a score of ≤ 9 for MADRS at 8 weeks.

For all patients, information such as, age, sex body mass index (BMI), smoking, status age of onset, severity, and psychiatric comorbidities were collected (Table 1). Symptoms evaluations were made using MADRS at the presentation of the patients to psychiatric services or hospital, in concomitance with the blood collection. Clinical features of the sample stratified based on remission status are reported in Supplementary Table 1.

Table 1
Socio-demographic and clinical features of the sample
Socio-demographic and clinical features	Non-TRD N = 150	TRD N = 150	p-value*
Age in years, mean (SD)	48.7 (15.9)	55.8 (11.19)	< 0.001
Females, n (% F)	109 (72.7)	106 (70.7)	0.798
Smokers, n (%)¹	53 (40.2)	54 (37.5)	0.711
Body Mass Index (BMI), mean (SD)²	24.3 (4.9)	25.2 (4.9)	0.097
Age of onset in years, mean (SD) ³	39.1 (14.9)	33.4 (13.7)	< 0.001
MADRS at recruitment, mean (SD)	26.1 (5.8)	31.9 (7.3)	< 0.001
Comorbidity with anxiety disorders, n (%)	53 (35.3)	38 (25.3)	0.078
* nonTRD vs TRD patients with Mann-Whitney U test or Pearson’s chi-squared test ¹Available for 149 non-TRD and 150 TRD; ²Available for 132 non-TRD and 144 TRD; ³Available for 126 non-TRD and 140 TRD.

The study was approved by the Local Ethics Committee (CEIOC IRCCS Istituto Centro San Giovanni di Dio Fatebenefratelli, Brescia, Italy; registration number: 62/2021) and from the Coordinator Ethics Committee (Ethik-Kommission Westfalen-Lippe der Ärztekammer Westfalen-Lippe, Münster, Germany; registration number: 2021-103-f-S).

The participants provided their written informed consent to participate in this study.

2.2 RNA-sequencing and genotyping

Peripheral venous blood samples were collected in the morning between 8 and 9 a.m., after an overnight fast, in EDTA tubes for DNA extraction and in PAXGene Blood RNA Tubes (Qiagen) for RNA extraction and mRNA/miRNA sequencing. For mRNAs, libraries were prepared using the Illumina Stranded Total RNA Prep with Ribo-Zero Plus kit and sequenced on an Illumina NovaSeq 6000 platform (2x30 million 100-base pair reads per sample) (14), while for miRNAs libraries were prepared using the NEBNext® Small RNA Library Prep Set for Illumina® kit (ref. E7330) and sequenced 1* 75 + 8 bp on an Illumina's NextSeq500 (Sirés et al, unpublished). For mRNA and miRNA sequencing analyses, quality control and statistical analyses procedures included RNA-seq preprocessing conducted using STAR 2.7.8a (15) and featureCounts from Subread v2.0.3 (16), normalization of gene expression counts in edgeR v.3.40.2 and removal of batch effect using the limma R package (Sirés et al, unpublished). After quality control, mRNA analyses included 293 participants and 21,564 genes, while mRNA analyses included 299 participants and 1,045 miRNAs.

For genotyping, genomic DNA was extracted from whole blood using the QIAamp DNA Blood Midi Kit (Qiagen), while the quantity and the quality of the DNA were evaluated through spectrophotometric analysis (NanoDrop 2000, Thermo Scientific). Twenty-nine genetic variants in CYP2D6 (n = 15), CYP2C19 (n = 10), and CYP2B6 (n = 4), selected based on the list published by the ClinPGX Association for Molecular Pathology (AMP) PGx Working Group (https://www.ClinPGX.org/ampAllelesToTest) were genotyped with customized TaqMan OpenArray plates on QuantStudio 12K Flex Real-Time PCR System (Applied Biosystems, Foster City, US), according to the manufacturer's instruction. Copy number variation (CNV) of the CYP2D6 gene was evaluated using the TaqMan Copy Number Assay mix specific for CYP2D6 exon 9 (Assay ID: Hs00010001_cn) according to the manufacturer's instructions. Using the AlleleTyper Software (Life Technologies, California, USA) Single nucleotide polymorphisms (SNPs) genotyping results were integrated with the copy number information for the CYP2D6 gene. If determination was uncertain, the diplotype with the higher allele frequency in the European population was selected. Genotype-inferred phenotyping was carried out as per CPIC guidelines (https://cpicpgx.org). Details about haplotype-inferred phenotype are reported in Supplementary Table 2.

2.3 Statistical analysis

The distribution of continuous variables was tested with Kolmogorov-Smirnov test. Differences in socio-demographic and clinical variables among the studied groups were tested with the Mann-Whitney U test or Pearson’s chi-square test. We tested the association between the metabolizer phenotypes of pharmacogenes and three clinical outcomes: 1) TRD vs non TRD, 2) extreme phenotypes of treatment response (TRD non-remitters vs non-TRD remitters), and 3) remission vs non-remission in patients stratified for TRD. The association between the metabolizer phenotypes and the three clinical outcomes was tested with Pearson’s chi-square test. The association between the status of being a carrier of one or more altered function phenotypes and clinical outcomes was also tested. To identify potential confounders, the association between metabolizer phenotypes and socioeconomic or clinical features was tested with the Kruskall-Wallis test or Pearson’s chi square test (Supplementary Table 1). In case of significant association between being a carrier of altered function phenotypes and clinical outcomes, a binary logistic regression model adjusted for confounders associated with pharmacogenes was also constructed. We also conducted analyses with TRD or extreme phenotypes as the outcomes, restricted to the subsample of patients treated with antidepressants for which clinical annotations at level 1a or 2a are available in ClinPGX for CYP2D6 or CYP2C19. No analyses were conducted for CYP2B6, since only 29 participants were treated with sertraline, the only antidepressant for which a clinical annotation is available for this enzyme or with remission as the outcome, and therefore a too limited number of patients who would have been included. A total of 139 patients (97 with TRD and 42 without TRD) were treated with antidepressants for which clinical annotations at level 1a or 2a are available for CYP2D6 (amitriptyline, clomipramine, fluvoxamine, mirtazapine, nortriptyline, paroxetine, venlafaxine, and vortioxetine), while 125 (56 with TRD and 69 without TRD) were treated with antidepressants for which clinical annotations at level 1a or 2a are available for CYP2C19 (amitriptyline, citalopram, clomipramine, escitalopram, and sertraline). For the analyses on the metabolizer phenotypes of the three pharmacogenes, a p-value adjusted based on Bonferroni correction for three tests was significant (i.e., 0.05 / 3 = 0.016).

For mRNA and miRNA analyses, differential expression analysis adjusted for age and sex was conducted using the limma-voom pipeline (limma version 3.54.1). P-values were adjusted for multiple testing using the Benjamini-Hochberg method, with a threshold of false discovery rate (FDR) < 0.05. For the present study, only log2-fold-change (log-FC), p-values and FDR adjusted (adj) p-values of selected genes (CYP2D6, CYP2C19 and CYP2B6) and regulating-miRNAs were extracted. However, mRNA association statistics were available only for CYP2D6, while CYP2C19 and CYP2B6 were too lowly expressed to be analyzed. Experimentally-supported and predicted miRNAs regulating the three pharmacogenes were identified with TarBase v. 9.0 (17). We identified one miRNA for CYP2D6 (hsa-miR-26b-5p), seven for CYP2C19 (hsa-miR-30d-3p, hsa-let-7d-5p, hsa-miR-139-5p, hsa-miR-210-3p, hsa-miR-27a-3p, hsa-miR-3662 and hsa-miR-423-5p) and none for CYP2B6. Analyses were performed with SPSS (v. 29).

3. Results

3.1 Association between CYP2D6 and TRD

CYP2D6 showed a significant association with comorbidity with anxiety disorders in the whole sample (Supplementary Table 3). Frequency of CYP2D6 phenotypes did not show significant differences between patients with TRD and those without TRD or when considering extreme phenotypes of response (Table 2 and Supplementary Table 4). Similarly, no difference was observed when restricting the analysis to patients treated with antidepressants for which high-level of evidence for clinical annotations for CYP2D6 are available on ClinPGX (Supplementary Table 5). Conversely, CYP2D6 mRNA was upregulated in patients with TRD compared to those without TRD (logFC = 0.35, t = 4.54, p = 8.3E-06, adj p = 0.0002; Table 3). In the extreme-phenotype comparison, CYP2D6 showed nominal significance, with the same direction of effect (logFC = 0.29, t = 2.47, p = 0.014, adj p = 0.11; Table 3). hsa-miR-26b-5p was the only miRNA predicted to regulate CYP2D6 levels based on TarBase and was significantly downregulated in patients with TRD compared to those without TRD (logFC = -0.62, t = -5.27, p = 2.6E-07, adj p = 2.5E-05; Table 3). Consistent results were observed in the extreme phenotype comparison, with hsa-miR-26b-5p found to be nominally downregulated in TRD non-remitters patients compared with non-TRD remitters (logFC = -0.44, t = -2.33, p = 0.02, adj p = 0.21; Table 3).

Table 2
Association between pharmacogene metabolizer phenotypes and TRD or remission
Pharmacogene metabolizer phenotype	Non-TRD N = 150	TRD N = 150	χ²	p-value	Non-TRD remitters N = 75	TRD non-remitters N = 57	χ²	p-value
CYP2D6 phenotypes (%)¹	PM = 4.7 IM = 39.6 NM = 51.7 UR = 4.0	PM = 4.1 IM = 35.1 NM = 58.1 UR = 2.7	1.41	0.703	PM = 4.0 IM = 40.0 NM = 53.3 UR = 2.7	PM = 3.5 IM = 31.6 NM = 64.9 UR = 0.0	2.92	0.405
CYP2C19 phenotypes (%)²	PM = 4.7 IM = 32.0 NM = 33.3 RM = 27.3 UM = 2.7	PM = 2.0 IM = 20.1 NM = 43.6 RM = 28.2 UM = 6.0	9.64	0.047	PM = 6.7 IM = 44.0 NM = 25.3 RM = 22.7 UM = 1.3	PM = 5.3 IM = 19.3 NM = 45.6 RM = 21.1 UM = 8.8	13.92	0.008
CYP2B6 phenotypes (%)³	PM = 10.7 IM = 36.7 NM = 51.3 RM = 1.3	PM = 6.8 IM = 47.6 NM = 42.2 RM = 3.4	6.06	0.109	PM = 10.7 IM = 37.3 NM = 52.0 RM = 0.0	PM = 1.8 IM = 52.7 NM = 40.0 RM = 5.5	10.37	0.016
¹Available for 149 non-TRD and 148 TRD; ²Available for 150 non-TRD and 149 TRD; ³Available for 150 non-TRD and 147 TRD and for 74 non-TRD remitters and 55 TRD non-remitters.
Abbreviations: IM, intermediate metabolizers; NM, normal metabolizers; PM, poor metabolizers; RM, rapid metabolizers; UM, ultrarapid metabolizers.

Table 3
Association between pharmacogene mRNA, miRNa regulating pharmacogenes and TRD or remission
Pharmacogene or miRNA	logFC	t	p-value	adj p-value*
Association with TRD vs non-TRD
CYP2D6	0.35	4.54	8.3E-06	0.0002
miRNAs regulating CYP2D6
hsa-miR-26b-5p	-0.62	-5.27	2.6E-07	2.5E-05
miRNAs regulating CYP2C19
hsa-miR-30d-3p	-0.11	-1.12	0.26	0.51
hsa-let-7d-5p	-0.31	-4.18	3.9E-05	0.002
hsa-miR-139-5p	-0.23	-2.46	0.01	0.10
hsa-miR-210-3p	0.20	2.55	0.01	0.09
hsa-miR-27a-3p	-0.38	-4.35	1.9E-05	0.0008
hsa-miR-423-5p	0.08	1.39	0.17	0.40
Association with non-remission in TRD vs remission in non-TRD
CYP2D6	0.29	2.47	0.01	0.11
miRNAs regulating CYP2D6
hsa-miR-26b-5p	-0.44	-2.33	0.02	0.21
miRNAs regulating CYP2C19
hsa-miR-30d-3p	-0.01	-0.05	0.96	0.99
hsa-let-7d-5p	-0.19	-1.68	0.09	0.42
hsa-miR-139-5p	-0.23	-1.64	0.10	0.44
hsa-miR-210-3p	0.19	1.64	0.10	0.44
hsa-miR-27a-3p	-0.44	-3.14	0.002	0.07
hsa-miR-423-5p	0.00	0.04	0.97	0.99
*adjusted p-value based on the original RNA and miRNA sequencing studies (Sirés et al., unpublished)

3.2 Association between CYP2C19 and TRD

CYP2C19 showed a significant association with age at recruitment in the whole sample and in patients with TRD (Supplementary Table 2). CYP2C19 phenotypes showed a nominally significant association with TRD (χ² = 9.64, p = 0.047, Table 2 and Supplementary Table 6) in the whole sample but not when limiting the analysis to patients treated with antidepressants with clinical annotations for CYP2C19 as reported in ClinPGX (Supplementary Table 7). Intermediate metabolizers (IM) (TRD: 20%, non-TRD: 32%) were significantly less represented in TRD than in non-TRD patients (IM vs normal metabolizers (NM): χ² = 6.07, p = 0.019). Similar results were also found when grouping the impaired function phenotypes (poor metabolizers (PM) + IM vs NM: χ² = 9.48, p = 0.004, Supplementary Table 6). The association remained nominally significant in a binary logistic regression model adjusted for age at recruitment (IM vs NM: odds ratio (OR) = 0.54, standard error = 0.31, p = 0.043; PM + IM vs NM: OR = 0.72, standard error = 0.15, p = 0.024).

In the comparison of extreme phenotypes of response, CYP2C19 phenotypes showed a significant association with non-remission in TRD (χ² = 13.92, p = 0.008, Table 2). In details, IM was less represented (IM: 19.3% vs 44.0%) in TRD non-remitters than non-TRD remitters (IM vs NM: χ² = 9.84, p = 0.002; PM + IM vs NM: χ² = 9.48, p = 0.004, Supplementary Table 6). The association remained significant in a binary logistic regression model adjusted for age at recruitment (IM vs NM: OR = 0.28, standard error = 0.49, p = 0.008; PM + IM vs NM: OR = 0.53, standard error = 0.23, p = 0.006).

mRNA counts for CYP2C19 were not available due to low expression. Of seven miRNAs found to regulate CYP2C19 levels based on TarBase, data were available for six miRNAs, two of which showed a significant association with TRD. Specifically, hsa-let-7d-5p (logFC = -0.31, t = -4.18, p = 3.9E-05, adj p = 0.002) and hsa-miR-27a-3p (logFC = -0.38, t = -4.35, p = 1.9E-05, adj p = 0.0008) were significantly downregulated in patients with TRD compared with patients without TRD (Table 3). Two other miRNAs, hsa-miR-139-5p and hsa-miR-210-3p, were nominally down- and upregulated, respectively, in patients with TRD compared with non-TRD (Table 3). Among these, only hsa-miR-27a-3p showed a nominal association in the extreme phenotype analysis (logFC = -0.43, t = -3.14, p = 0.002, adj p = 0.07).

3.3 Association between CYP2B6 and TRD

CYP2B6 showed a significant association with smoking in patients with non-TRD and with recurrence in patients with MDD (Supplementary Table 2). The frequency of CYP2B6 metabolizer phenotypes was not associated with TRD (Table 2). While we identified a significant association between CYP2B6 metabolizer phenotypes and the extreme phenotype of response (Table 2), we found no significant association when evaluating the status of being a carrier of one or more specific metabolizer phenotypes compared to NMs (Supplementary Table 8). The statistics for the mRNA of this gene were not available due to low expression and no miRNA was identified to regulate CYP2B6 levels on TarBase.

3.4 Association between pharmacogenes and remission

We observed no significant association between CYP2D6 phenotypes and remission in either patients with or without TRD (Table 4).

Table 4
Association between pharmacogene metabolizer phenotypes and remission
	Non-TRD				TRD
Pharmacogene metabolizer phenotype	Remitters N = 75	Non-remitters N = 40	χ²	p-value	Remitters N = 55	Non-remitters N = 57	χ²	p-value
CYP2D6 phenotypes (%)¹	PM = 4.0 IM = 40.0 NM = 53.3 UR = 2.7	PM = 2.5 IM = 50.0 NM = 45.0 UR = 2.5	1.38	0.709	PM = 3.7 IM = 35.2 NM = 53.7 UR = 7.4	PM = 3.5 IM = 31.6 NM = 64.9 UR = 0.0	4.92	0.178
CYP2C19 phenotypes (%)	PM = 6.7 IM = 44.0 NM = 25.3 RM = 22.7 UM = 1.3	PM = 2.4 IM = 19.5 NM = 34.1 RM = 41.5 UM = 2.4	10.66	0.031	PM = 0.0 IM = 14.5 NM = 43.6 RM = 36.4 UM = 5.5	PM = 5.3 IM = 19.3 NM = 45.6 RM = 21.1 UM = 8.8	6.02	0.198
CYP2B6 phenotypes (%)²	PM = 10.7 IM = 37.3 NM = 52.0 RM = 0.0	PM = 14.6 IM = 29.3 NM = 51.2 RM = 4.9	4.90	0.179	PM = 3.6 IM = 50.9 NM = 43.6 RM = 1.8	PM = 1.8 IM = 52.7 NM = 40.0 RM = 5.5	1.44	0.697
¹Available for 39 non-TRD remitters, 75 TRD non remitters, 54 TRD remitters and 57 TRD non remitters; ²Available for 40 non-TRD remitters, 75 TRD non remitters, 55 TRD remitters and 55 TRD non remitters.
Abbreviations: IM, intermediate metabolizers; NM, normal metabolizers; PM, poor metabolizers; RM, rapid metabolizers; UM, ultrarapid metabolizers.

Analyses on specific metabolizer phenotypes showed a trend for higher frequency of CYP2D6 ultrarapid metabolizers (UM) in remitters compared with non-remitters in patients with TRD (χ² = 4.76, p = 0.045) but not in patients without TRD (χ² = 0.02, p = 1.00, Supplementary Table 9). In addition, lower levels of CYP2D6 were nominally associated with non-remission in patients with TRD (logFC = -0.25, t = -2.11, p = 0.04, adj p = 0.39) but not in patients without TRD (logFC = -0.06, t = -0.45, p = 0.65, adj p = 0.99). Consistently, hsa-miR-26b-5p was nominally upregulated in non-remitters compared with remitters in patients with TRD (logFC = 0.40, t = 2.08, p = 0.04, adj p = 0.51) but not in those without TRD (logFC = -0.05, t = -0.26, p = 0.80, adj p = 0.99).

Frequency of CYP2C19 phenotypes showed a nominally significant trend of association with remission among patients without TRD, but not in patients with TRD (Table 4). Analyses on specific metabolizer phenotypes showed a nominally significant higher frequency of reduced function phenotypes (IM vs NM: χ² = 5.48, p = 0.036; PM + IM vs NM: χ² = 5.99, p = 0.022) in remitters compared with non-remitters in TRD (Supplementary Table 10). Among miRNAs regulating CYP2C19, only hsa-let-7d-5p was nominally upregulated in non-remitters compared with remitters in patients with TRD (logFC = 0.27, t = 2.31, p = 0.02, adj p = 0.46) but not in those without TRD (logFC = 0.03, t = 0.28, p = 0.78, adj p = 0.99). Finally, CYP2B6 phenotypes were not significantly associated with remission in either TRD or non-TRD patients (Table 4 and Supplementary Table 11).

4. Discussion

In this study we analyzed metabolizer phenotypes, peripheral blood mRNA and miRNA levels of pharmacogenes for antidepressants presenting with the highest level of evidence (1A) for clinical utility in ClinPGX (CYP2D6, CYP2C19, and CYP2B6), to evaluate their association with TRD or remission in the PROMPT project phase 1 cohort. Most of the previous studies exploring the same pharmacogenes in psychiatric disorders restricted the analyses to the genetically-inferred metabolizer phenotypes, not including mRNA or miRNA levels. However, regarding CYP2D6, a recent study showed that different metabolizer phenotype groups are characterized by substantial variability in CYP2D6 liver mRNA levels (11). As regarding psychiatric medications, the importance of differences in mRNA has been recently supported by a study showing that peripheral blood CYP1A2 mRNA expression, rather than genotype, was associated with olanzapine concentration in schizophrenia or bipolar disorder patients (18). While in our study we did not identify a significant association between CYP2D6 metabolizer phenotypes and clinical outcomes, we found that patients with TRD had higher expression of CYP2D6 mRNA (logFC = 0.35, adj p = 0.0002) and lower expression of the CYP2D6 regulating miRNA miR-26b-5p (logFC = -0.62, adj p = 2.5E-05). It could be speculated that higher expression of CYP2D6 could be responsible for an increased elimination rate of the substrates drugs, thus possibly making the “standard” dose inefficacious. Nevertheless, hepatic expression of CYP2D6, as well as information on drug dose and dose adjustments is not available for the studied sample, and, as such, this assumption cannot be confirmed or confuted. Regarding the miRNA, interestingly, in a study on post-mortem brain samples, lower levels of miR-26b-5p in the dorsolateral prefrontal cortex were associated with a higher depressive symptom score measured with the Center for Epidemiological Studies Depression scale in participants from two longitudinal clinical–pathologic cohort studies of aging (AD-ROS and MAP) (19). miR-26b-5p was also recently shown to be downregulated in peripheral blood in patients with depression (20) and in extracellular vesicles in patients with depression and childhood trauma (21), compared with controls, in two pilot studies including a small number of participants and conducted by the same group. The authors of these studies suggested miR-26b-5p to be a potential mediator of the association between depression and bone health, based on the observed correlation between miRNA levels and bone turnover markers (20, 21). Together with the previously published works, our findings suggest that the pharmacokinetics of psychiatric drugs could also be influenced by epigenetically-determined mechanisms rather than the more widely recognized genetically-driven metabolizing properties of cytochromes. In the case of CYP2D6, this could be determined by a miRNA possibly implicated in the disorder, thus highlighting the complex interplay between the genetic/epigenetic makeup of the patient and the efficacy and safety of medications.

As regards CYP2C19, we observed a significant association between CYP2C19 metabolizer phenotypes and TRD. Namely, non-TRD, compared with TRD patients, showed a higher frequency of IMs or IM + PMs phenotypes (p = 0.019 and p = 0.007, respectively, Supplementary Table 5). In addition, among patients with TRD, higher frequencies of IMs or IM + PMs phenotypes were nominally associated with remission (p = 0.045 and p = 0.044, respectively, Supplementary Table 9). These results are in accordance with a recent study including 1,239 patients with MDD reporting that CYP2C19 PMs had higher rates of response and symptom improvement compared to NMs, but also a higher risk of autonomic and neurological side effects (10). Accordingly, a meta-analysis of 13 studies including 5,843 patients with MDD from European and East Asian ancestry populations reported nominal significance for the association between CYP2C19 PM phenotype and a higher remission rate (22), while a meta-analysis including 2,558 patients with MDD from the GENDEP, STAR ∗D, GenPod, and PGRN-AMPS cohorts reported higher symptom improvement and remission rates in CYP2C19 PMs compared with NMs (23). However, these results are in contrast with two studies recently conducted in the UK Biobank cohort. Namely, Kamp and colleagues found lower rates of self-reported response to selective serotonin reuptake inhibitors in CYP2C19 PMs compared with NMs in 19,516 participants from UK Biobank (24), while Wong and colleagues analyzed data from 3,012 individuals prescribed escitalopram, reporting that CYP2C19 PMs were more likely to switch antidepressants, have side effects following first prescription, and be on escitalopram for a shorter duration compared to NMs (25).

While CYP2C19 mRNA levels could not be analysed due to low expression of the gene, two CYP2C19-regulating miRNAs were significantly downregulated in patients with TRD compared with non-TRD: let-7d-5p (logFC = -0.31, adj p = 0.002) and miR-27a-3p (logFC = -0.38, adj p = 0.0008). The latter findings are difficult to interpret in light of the lack of mRNA levels, but, at the same time, they support the hypothesis of a possible role of non-genetic factors involved in CYP regulation in response and resistance to psychiatric drugs.

To our knowledge, this study is the first to conduct a comprehensive analysis of metabolizer phenotypes, mRNA, and regulating miRNA levels of pharmacogenes in patients with MDD characterized for TRD and remission.

Nonetheless, our results have to be interpreted in light of some limitations. Firstly, the retrospective nature of the study did not allow us to explore the causative effect of genetic variants and transcripts levels in modulating remission and resistance to treatments. Moreover, information about blood drug levels was not available, likewise the details on dosage adjustment during the treatment course.

Secondly, the relatively small sample size limited the number of participants with extreme phenotypes that could be included in analyses with remission as the outcome. In addition, all participants were of European origin, thus potentially limiting the application of these results to other populations. Finally, since leukocyte gene expression is not a direct measurement of hepatocyte expression, but rather a biomarker, it might not be able to fully reflect the hepatic activity of cytochromes (26).

Overall, by delving into the exploration of pharmacogenes with a high level of evidence for clinical utility, this work represents an essential integration into the PROMPT project, possibly adding an important piece of knowledge on the role of pharmacokinetic genes and their expression regulation dynamics in the response and resistance to antidepressant treatment in MDD.

Declarations

Ethics approval and consent to participate

Consent for publication

Not applicable

Data Availability

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Competing interests

The authors declare that they have no competing interests.

Funding

The project “Toward PrecisiOn Medicine for the Prediction of Treatment response in major depressive disorder through stratification of combined clinical and -omics signatures” is supported by the German Federal Ministry of Health (BMG) [2521FSB004_PROMPT], the National Centre for Research and Development Poland (NCBR) [PerMed/III/2/PROMPT/2021], the Italian Ministry of Health (IT-MoH) [ERP-2020-23671059], The French National Research Agency (ANR) [ANR-20-PERM-0003] and the Investissement d’Avenir [ANR-10-AIHU-06]. the Health Department – Generalitat de Catalunya (DS-CAT) [SLD044/20/000001] and the Instituto de Salud Carlos III (ISCIII) [IHMC22/00026] under the frame of ERA PerMed.

The assistant research positions of Giulia Perusi and Lisa Buson were funded by ERA-PerMed PROMPT project [IT-MoH ERP-2020-23671059]. The Ph.D. student position of Valentina Menesello is partly funded by the PNRR – DM 117/2023 Grant. The post-doc position of Dr. Rosana Carvalho Silva was partly funded by the Psychiatric Hospital ‘Villa Santa Chiara’, Verona, Italy. Massimo Gennarelli and Alessandra Minelli are supported by the Italian Ministry of Health (IT-MoH) under Grant Ricerca Corrente 2025 (RC-2025).

The funders had no further role in study design; in the collection, analysis and interpretation of data; in the writing of the report; and in the decision to submit the paper for publication.

Author Contribution

Claudia Pisanu: Conceptualization, Formal analysis, Writing – original draft preparation, Writing – review and editing; Alessio Squassina: Conceptualization, Writing – review and editing; Júlia Perera-Bel: Formal analysis, Writing – review and editing; Rosana Carvalho Silva: Writing – original draft preparation; Lisa Buson: Methodology; Anna Martinez Sires: Methodology; Marie Claude Potier: Methodology, Writing – review and editing; Marco Bortolomasi: Investigation; Valentina Menesello: Investigation; Bernardo Carpiniello: Investigation; Ewa Ferensztaj-Rochowiak: Investigation; Giulia Perusi: Investigation; Filip Rybakowski: Funding acquisition, Writing – review and editing; Ferran Sanz: Funding acquisition, Writing – review and editing; Mirko Manchia: Writing – review and editing; Mara Dierssen: Methodology, Funding acquisition, Writing – review and editing; Bernhard T. Baune: Funding acquisition, Writing – review and editing; Massimo Gennarelli: Investigation; Alessandra Minelli: Conceptualization, Funding acquisition, Writing – original draft preparation. All authors contributed to and have approved the final manuscript.

Acknowledgements

We would like to express our sincere gratitude to all the volunteers who participated in the study. We thank all the staff of the Psychiatric Hospital ‘Villa Santa Chiara’. We thank the iGENSEQ platform of the Paris Brain Institute for sequencing all samples.

Electronic Supplementary Material

Below is the link to the electronic supplementary material

Supplementary Material 1

References

Hasin DS, Sarvet AL, Meyers JL, Saha TD, Ruan WJ, Stohl M, et al. Epidemiology of Adult DSM-5 Major Depressive Disorder and Its Specifiers in the United States. JAMA Psychiatry. 2018;75(4):336–46.

Gaynes BN, Rush AJ, Trivedi MH, Wisniewski SR, Spencer D, Fava M. The STAR*D study: treating depression in the real world. Cleve Clin J Med. 2008;75(1):57–66.

Souery D, Serretti A, Calati R, Oswald P, Massat I, Konstantinidis A, et al. Switching antidepressant class does not improve response or remission in treatment-resistant depression. J Clin Psychopharmacol. 2011;31(4):512–6.

Cipriani A, Furukawa TA, Salanti G, Chaimani A, Atkinson LZ, Ogawa Y, et al. Comparative Efficacy and Acceptability of 21 Antidepressant Drugs for the Acute Treatment of Adults With Major Depressive Disorder: A Systematic Review and Network Meta-Analysis. Focus (Am Psychiatr Publ). 2018;16(4):420–9.

Sharma T, Guski LS, Freund N, Meng DM, Gotzsche PC. Drop-out rates in placebo-controlled trials of antidepressant drugs: A systematic review and meta-analysis based on clinical study reports. Int J Risk Saf Med. 2019;30(4):217–32.

Mrazek DA, Biernacka JM, McAlpine DE, Benitez J, Karpyak VM, Williams MD, et al. Treatment outcomes of depression: the pharmacogenomic research network antidepressant medication pharmacogenomic study. J Clin Psychopharmacol. 2014;34(3):313–7.

Jukic M, Milosavljevic F, Molden E, Ingelman-Sundberg M. Pharmacogenomics in treatment of depression and psychosis: an update. Trends Pharmacol Sci. 2022;43(12):1055–69.

Tansey KE, Guipponi M, Hu X, Domenici E, Lewis G, Malafosse A, et al. Contribution of common genetic variants to antidepressant response. Biol Psychiatry. 2013;73(7):679–82.

Zheng N, Niu M, Zang Y, Zhuang H, Jia F, Bao S, et al. Which can Predict the Outcome of Antidepressants: Metabolic Genes or Pharmacodynamic Genes? Curr Drug Metab. 2023;24(7):525–35.

10.

Calabro M, Fabbri C, Kasper S, Zohar J, Souery D, Montgomery S, et al. Metabolizing status of CYP2C19 in response and side effects to medications for depression: Results from a naturalistic study. Eur Neuropsychopharmacol. 2022;56:100–11.

11.

Meaddough EL, Sarasua SM, Kunkel D, Boccuto L, Ganakammal SR, Moersen M, et al. Assessment of CYP2D6 gene expression in liver tissue: Variability in CYP2D6 mRNA levels within genotype-predicted metabolizer phenotype groups. Chem Biol Interact. 2025;416:111526.

12.

Zastrozhin MS, Skryabin V, Smirnov V, Zastrozhina AK, Kaverina EV, Klepikov DA, et al. Impact of the Omics-Based Biomarkers on the Mirtazapine's Steady-State Concentration, Efficacy and Safety in Patients with Affective Disorders Comorbid with Alcohol Use Disorder. Psychopharmacol Bull. 2021;51(2):31–42.

13.

Baune BT, Minelli A, Carpiniello B, Contu M, Dominguez Barragan J, Donlo C, et al. An integrated precision medicine approach in major depressive disorder: a study protocol to create a new algorithm for the prediction of treatment response. Front Psychiatry. 2023;14:1279688.

14.

Sires A, Dominguez-Barragan J, Lorenzon N, Meskini O, Buson L, Silva RC et al. Blood transcriptomic analysis reveals a distinct molecular subtype of treatment resistant depression compared to non-treatment resistant depression. Brain Behav Immun. 2025:106103.

15.

Dobin A, Davis CA, Schlesinger F, Drenkow J, Zaleski C, Jha S, et al. STAR: ultrafast universal RNA-seq aligner. Bioinformatics. 2013;29(1):15–21.

16.

Liao Y, Smyth GK, Shi W. featureCounts: an efficient general purpose program for assigning sequence reads to genomic features. Bioinformatics. 2014;30(7):923–30.

17.

Skoufos G, Kakoulidis P, Tastsoglou S, Zacharopoulou E, Kotsira V, Miliotis M, et al. TarBase-v9.0 extends experimentally supported miRNA-gene interactions to cell-types and virally encoded miRNAs. Nucleic Acids Res. 2024;52(D1):D304–10.

18.

Fekete F, Menus A, Toth K, Kiss AF, Minus A, Sirok D, et al. CYP1A2 expression rather than genotype is associated with olanzapine concentration in psychiatric patients. Sci Rep. 2023;13(1):18507.

19.

Wingo TS, Yang J, Fan W, Min Canon S, Gerasimov ES, Lori A, et al. Brain microRNAs associated with late-life depressive symptoms are also associated with cognitive trajectory and dementia. NPJ Genom Med. 2020;5:6.

20.

Houtenbos SP, He Y, Cazzanelli P, Soultoukis G, Wuertz-Kozak K, Schulz TJ, et al. The underlying mechanisms of the association of bone health with depression - an experimental study. Mol Biol Rep. 2025;52(1):163.

21.

He Y, Wuertz-Kozak K, Cazzanelli P, Houtenbos S, Garcia-Carrizo F, Schulz TJ, et al. Differential expression of plasma extracellular vesicles microRNAs and exploration of their association with bone metabolism in childhood trauma participants treated in a psychosomatic clinic. Front Endocrinol (Lausanne). 2025;16:1515910.

22.

Li D, Pain O, Fabbri C, Wong WLE, Lo CWH, Ripke S, et al. Metabolic activity of CYP2C19 and CYP2D6 on antidepressant response from 13 clinical studies using genotype imputation: a meta-analysis. Transl Psychiatry. 2024;14(1):296.

23.

Fabbri C, Tansey KE, Perlis RH, Hauser J, Henigsberg N, Maier W, et al. Effect of cytochrome CYP2C19 metabolizing activity on antidepressant response and side effects: Meta-analysis of data from genome-wide association studies. Eur Neuropsychopharmacol. 2018;28(8):945–54.

24.

Kamp M, Lo CWH, Kokkinidis G, Chauhan M, Gillett AC, Team AR, et al. Sociodemographic, clinical, and genetic factors associated with self-reported antidepressant response outcomes in the UK Biobank. Psychol Med. 2025;55:e80.

25.

Wong WLE, Fabbri C, Laplace B, Li D, van Westrhenen R, Lewis CM et al. The Effects of CYP2C19 Genotype on Proxies of SSRI Antidepressant Response in the UK Biobank. Pharmaceuticals (Basel). 2023;16(9).

26.

Temesvari M, Kobori L, Paulik J, Sarvary E, Belic A, Monostory K. Estimation of drug-metabolizing capacity by cytochrome P450 genotyping and expression. J Pharmacol Exp Ther. 2012;341(1):294–305.

Abbreviations:

females

number

standard deviation.

Yes

Abstract

Background. Treatment-resistant depression (TRD) is a major clinical challenge in the management of major depressive disorder (MDD). While pharmacogenetics has been suggested to have some clinical utility in guiding antidepressant treatment, only few studies have explored if and how pharmacogenes can be involved in TRD pathophysiology and its clinical outcomes. In this study we explored the role of differences in metabolizer phenotypes, gene expression levels, and microRNAs of three key pharmacogenes (CYP2D6, CYP2C19, and CYP2B6) in TRD pathophysiology and antidepressant response in a cohort of 300 patients with MDD from the PROMPT consortium. Results. CYP2D6 phenotype distribution did not differ significantly between TRD and non-TRD groups, but mRNA expression was significantly upregulated in TRD patients (log fold change [logFC] = 0.35, adj p = 0.0002). Additionally, hsa-miR-26b-5p, a microRNA predicted to regulate CYP2D6, was significantly downregulated in TRD (logFC = -0.62, adj p = 2.5E-05). For CYP2C19, intermediate metabolizers (IM) were underrepresented in TRD compared to non-TRD individuals (IM vs normal metabolizers (NM): 2 = 6.07, p = 0.019). microRNA hsa-let-7d-5p and hsa-miR-27a-3p, predicted to regulate CYP2C19, were significantly downregulated in TRD (logFC = -0.31, adj p = 0.002; logFC = -0.38, adj p = 0.0008). No significant differences were found for CYP2B6. In the analysis of remission to the last antidepressant trial, only nominally significant differences were reported, which did not survive multiple testing corrections. Conclusions. By investigating high levels of evidence pharmacogenes, this study contributes valuable insights to the PROMPT project on how pharmacokinetic gene variants, their expression and regulatory mechanisms may influence antidepressant response and resistance in MDD.