A primer on the use of machine learning to distil knowledge from data in biological psychiatry

Applications of machine learning in the biomedical sciences are growing rapidly. This growth has been spurred by diverse cross-institutional and interdisciplinary collaborations, public availability of large datasets, an increase in the accessibility of analytic routines, and the availability of powerful computing resources. With this increased access and exposure to machine learning comes a responsibility for education and a deeper understanding of its bases and bounds, borne equally by data scientists seeking to ply their analytic wares in medical research and by biomedical scientists seeking to harness such methods to glean knowledge from data. This article provides an accessible and critical review of machine learning for a biomedically informed audience, as well as its applications in psychiatry. The review covers definitions and expositions of commonly used machine learning methods, and historical trends of their use in psychiatry. We also provide a set of standards, namely Guidelines for REporting Machine Learning Investigations in Neuropsychiatry (GREMLIN), for designing and reporting studies that use machine learning as a primary data-analysis approach. Lastly, we propose the establishment of the Machine Learning in Psychiatry (MLPsych) Consortium, enumerate its objectives, and identify areas of opportunity for future applications of machine learning in biological psychiatry. This review serves as a cautiously optimistic primer on machine learning for those on the precipice as they prepare to dive into the field, either as methodological practitioners or well-informed consumers.

BRD3 Regulates the Inflammatory and Stress Response in Rheumatoid Arthritis Synovial Fibroblasts

BACKGROUND

Individual functions of members of the bromodomain (BRD) and extra-terminal (BET) protein family underlying the anti-inflammatory effects of BET inhibitors in rheumatoid arthritis (RA) are incompletely understood. Here, we aimed to analyze the regulatory functions of BRD3, an understudied member of the BET protein family, in RA synovial fibroblasts (FLS).

METHODS

BRD3 was silenced in FLS prior to stimulation with TNF. Alternatively, FLS were treated with I-BET. Transcriptomes were analyzed by RNA sequencing (RNAseq), followed by pathway enrichment analysis. We confirmed results for selective target genes by real-time PCR, ELISA, and Western blotting.

RESULTS

BRD3 regulates the expression of several cytokines and chemokines in FLS, and positively correlates with inflammatory scores in the RA synovium. In addition, RNAseq pointed to a profound role of BRD3 in regulating FLS proliferation, metabolic adaption, and response to stress, including oxidative stress, and autophagy.

CONCLUSIONS

BRD3 acts as an upstream regulatory factor that integrates the response to inflammatory stimuli and stress conditions in FLS and executes many functions of BET proteins that have previously been identified using pan-BET inhibitors.

The histone acetyl transferases CBP and p300 regulate stress response pathways in synovial fibroblasts at transcriptional and functional levels

The activation of stress response pathways in synovial fibroblasts (SF) is a hallmark of rheumatoid arthritis (RA). CBP and p300 are two highly homologous histone acetyl transferases and writers of activating histone 3 lysine 27 acetylation (H3K27ac) marks. Furthermore, they serve as co-factors for transcription factors and acetylate many non-histone proteins. Here we showed that p300 but not CBP protein expression was down regulated by TNF and 4-hydroxynonenal, two factors that mimic inflammation and oxidative stress in the synovial microenvironment. We used existing RNA-sequencing data sets as a basis for a further in-depth investigation of individual functions of CBP and p300 in regulating different stress response pathways in SF. Pathway enrichment analysis pointed to a profound role of CBP and/ or p300 in regulating stress response-related gene expression, with an enrichment of pathways associated with oxidative stress, hypoxia, autophagy and proteasome function. We silenced CBP or p300, and performed confirmatory experiments on transcriptome, protein and functional levels. We have identified some overlap of CBP and p300 target genes in the oxidative stress response pathway, however, with several genes being regulated in opposite directions. The majority of stress response genes was regulated by p300, with a specific function of p300 in regulating hypoxia response genes and genes encoding proteasome subunits. Silencing of p300 suppressed proteasome enzymatic activities. CBP and p300 regulated autophagy on transcriptome and functional levels. Whereas CBP was indispensable for autophagy synthesis, silencing of p300 affected late-stage autophagy. In line with impaired autophagy and proteasome function, poly-ubiquitinated proteins accumulated after silencing of p300.

Polygenic risk scores analyses of psychiatric and metabolic traits with antipsychotic-induced weight gain in schizophrenia: an exploratory study

Given the polygenic nature of antipsychotic-induced weight gain (AIWG), we investigated whether polygenic risk scores (PRS) for various psychiatric and metabolic traits were associated with AIWG. We included individuals with schizophrenia (SCZ) of European ancestry from two cohorts (N = 151, age = 40.3 ± 11.8 and N = 138, age = 36.5 ± 10.8). We investigated associations of AIWG defined as binary and continuous variables with PRS calculated from genome-wide association studies of body mass index (BMI), coronary artery disease (CAD), fasting glucose, fasting insulin, high-density lipoprotein cholesterol, low-density lipoprotein cholesterol (LDL-C), triglycerides, type 1 and 2 diabetes mellitus, and SCZ, using regression models. We observed nominal associations (uncorrected p < 0.05) between PRSs for BMI, CAD, and LDL-C, type 1 diabetes, and SCZ with AIWG. While results became non-significant after correction for multiple testing, these preliminary results suggest that PRS analyses might contribute to identifying risk factors of AIWG and might help to elucidate mechanisms at play in AIWG.

OP0199 RADIOMIC SIGNATURES REFLECT TREATMENT RESPONSE TO NINTEDANIB IN PRECLINICAL LUNG FIBROSIS MODEL

Background

Responses to anti-fibrotic drugs in preclinical disease models are difficult to quantify by histological analysis of single tissue sections. Quantitative in-depth analysis of imaging data, termed “radiomics”, may represent a more reliable and accurate measure of treatment response since the pathology of the whole organ is captured.

Objectives

To study the potential of µCT-derived radiomic features to reflect response to Nintedanib in the bleomycin (BLM)-induced murine model of fibrosing interstitial lung disease.

Methods

All C57BL/6J mice from both study groups were intratracheally instilled with 2 U/kg BLM on day 0 to induce lung fibrosis. Nintedanib was administered daily by gavage at 60 mg/kg for two weeks starting from day 7 (n=15). Controls received equivalent treatment with vehicle-only (n=19). Whole lung µCT scans (SkyScan 1176, Bruker) of each animal were acquired at baseline (day 0), pre-treatment (day 7), and post-treatment (day 21). The Ashcroft score was assessed on Sirius Red stained lung sections post-treatment. Lung volumes in µCTs were defined semi-automatically in MIM Software (6.9.2), followed by extraction of radiomic features with our in-house developed software Z-Rad (7.3.1). Each data set contained 1’386 features, describing image characteristics with histogram, texture, and wavelet functions. Data pre-processing involved removal of features sensitive to intra- and interobserver delineation variability (ICC<0.75), highly correlated features (Pearson’s r>0.95), and features not significantly changing between days 0 and 7 (p>0.05). Agglomerative clustering of radiomic temporal trajectories was performed on the Nintedanib group to identify distinct feature clusters. The identified feature sets were then used to plot average feature value trajectories for both study groups in each cluster. To identify features significantly different between a) Nintedanib vs. control, and b) pre- vs. post-treatment, Mann-Whitney U and Wilcoxon signed-rank tests were used, respectively. Samples were pooled from two independent experiments.

Results

Evaluation of tissue sections did not show a significant treatment-induced reduction of fibrosis with average Ashcroft scores of 3.7 (±1.2 s.d.) and 3.4 (±1.6 s.d.) in Nintedanib and control samples, respectively (p>0.05). Radiomics data analysis revealed two feature clusters in Nintedanib samples, composed of 52 features (cluster 1) and 96 features (cluster 2), the trajectories of which were then plotted for both study groups. In cluster 1, feature value trajectories significantly decreased in both Nintedanib and control group between pre-and post-treatment (p<0.001), whereas feature values in cluster 2 remained flat (p>0.05). Importantly, Nintedanib-treated mice displayed a much more pronounced feature value decrease post-treatment in cluster 1 compared to the control group (p<0.05). Here, feature values post-treatment resembled pre-disease baseline conditions in the Nintedanib group (p>0.05), whereas the control group remained significantly different from baseline (p<0.01). Cluster 1 was composed of 6 histogram, 11 texture, and 35 wavelet features, emphasizing the role of high-dimensional metrics for the detection of differences.

Conclusion

Histological quantification of lung fibrosis accounts only for a small fraction of the whole pathology in a spatially heterogeneous disease. We demonstrated that µCT-derived radiomic features identified significant differences on imaging level following Nintedanib treatment, which we could not reliably detect on tissue level using Ashcroft scoring. These findings hold great potential for the development of novel readouts for improved stratification of anti-fibrotic treatment effects in preclinical models.

Acknowledgements

This study received funding support from the Swiss Lung Association.

Disclosure of Interests

David Lauer Shareholder of: Roche (no relation to project), Employee of: Former employee of Roche (no relation to project), Janine Schniering: None declared, Hubert Gabrys: None declared, Malgorzata Maciukiewicz: None declared, Matthias Brunner: None declared, Oliver Distler Speakers bureau: Speaker fees in the area of systemic sclerosis and related complications from Bayer, Boehringer Ingelheim, Janssen, Medscape, Consultant of: Consultancies in the area of systemic sclerosis and its complications with Abbvie, Acceleron, Alcimed, Amgen, AnaMar, Arxx, AstraZeneca, Baecon, Blade, Bayer, Boehringer Ingelheim, Corbus, CSL Behring, 4P Science, Galapagos, Glenmark, Horizon, Inventiva, Kymera, Lupin, Miltenyi Biotec, Mitsubishi Tanabe, MSD, Novartis, Prometheus, Roivant, Sanofi and Topadur, Grant/research support from: Grant/research support from Kymera, Mitsubishi Tanabe, Boehringer Ingelheim, Thomas Frauenfelder: None declared, Stephanie Tanadini-Lang: None declared, Britta Maurer Speakers bureau: Received speaker fees from Boehringer-Ingelheim as well as congress support from Medtalk, Pfizer, Roche, Actelion, Mepha, and MSD, Consultant of: Consultancies with Novartis, Boehringer Ingelheim, Janssen-Cilag. Has a patent mir-29 for the treatment of systemic sclerosis issued (US8247389, EP2331143), Grant/research support from: Had grant/research support from AbbVie, Protagen, Novartis Biomedical Research.

POS0426 BRD3 REGULATES THE INFLAMMATORY AND STRESS RESPONSE IN RHEUMATOID ARTHRITIS SYNOVIAL FIBROBLASTS

Background

Small molecule inhibitors targeting members of the bromodomain and extra-terminal (BET) protein family (BRD2, BRD3, BRD4) have anti-inflammatory properties in rheumatoid arthritis (RA). BET proteins are readers of acetylated histone side chains and activators of transcription. BRD3 is an understudied member of BET proteins.

Objectives

To analyse individual functions of BET proteins and mechanisms underlying BET inhibition in RA synovial fibroblasts (SF).

Methods

The expression of BRD2, BRD3, and BRD4 was silenced by lenti-viral transduction followed by TNF stimulation (10 ng/µl, 24h). Silencing was confirmed by Western blotting. Transcriptomes were determined by RNA-seq (Illumina NovaSeq 6000, n=3). Pathway enrichment analysis for KEGG and Reactome databases was conducted with significantly affected genes (± fold change > 1.5, FDR < 0.05). SF were treated with I-BET (1 µM) and TNF (10 ng/µl, 24h). Autophagy was evaluated by Western blotting using the conversion of LC3B as a marker (n=9). I-BET-induced global changes on post-translational histone modifications were analysed by mass spectrometry (Mod Spec, Active Motif; n=2; 120h protocol) and Western blotting (H3K27ac, H3K18ac, total acH3; n=7; 24h and 120h protocol). For this purpose, SF were stimulated with I-BET (1 µM) for 24h, and either co-stimulated with TNF (24h protocol), or washed with PBS, followed by a 24h stimulation with TNF 120h after the I-BET treatment (120h protocol).

Results

Silencing of BRD2 and BRD4 in SF was, in contrast to silencing of BRD3, associated with high levels of cell death, and therefore not analyzed further. We detected 257 and 324 differentially expressed genes (DEG) that were affected by BRD3 silencing in unstimulated and TNF-stimulated SF, respectively. 105 DEG overlapped between the two groups. DEG were enriched in inflammatory pathways such as “TNF signaling pathway”, “rheumatoid arthritis”, “Toll-like receptor cascades”, “MAPK signaling pathway”, “IL-17 signaling pathway” and “signaling by interleukins”. Furthermore, pathway enrichment analysis suggested a role for BRD3 in different stress-associated pathways, including “DNA repair”, “chaperone mediated autophagy”, “cellular responses to stress”, and “autophagy”. In line with the pathway enrichment analysis, I-BET induced levels of LC3B-II in unstimulated (4.3 fold, p=0.07) and TNF-stimulated (2.9 fold, p=0.07) SF, indicating a role of BET proteins in the regulation of autophagy. To further study the mechanisms underlying I-BET-mediated suppression of gene expression, we analyzed potential effects of I-BET on histone modifications. Mod Spec analysis indicated that I-BET induced profound changes in chromatin modifications, with a global reduction of acetylation on different histone side chains. We confirmed some of these differences in independent samples. I-BET treatment reduced mean TNF-induced levels of total acH3 by 25.2% (120h; p=0.0303), of H3K18ac by 35.3% (24h; p=0.0288) and by 29.3% (120h; p=0.0373) and of H3K27ac by 41.7% (120h; p=0.0587).

Conclusion

BRD3 acts as an upstream regulatory factor that integrates the response to inflammatory stimuli and stress conditions in SF. Our data suggest that BET inhibitors do not only prevent the reading of acetylated histone side chains, but also directly affect the chromatin structure, in particular by downregulating global levels of histone acetylation.

Disclosure of Interests

Malgorzata Maciukiewicz: None declared, Larissa Moser: None declared, Monika Krosel: None declared, Tanja Seifritz: None declared, Matija Tomsic: None declared, Britta Maurer Speakers bureau: Boehringer-Ingelheim, Consultant of: Novartis, Boehringer Ingelheim, Janssen-Cilag, Grant/research support from: AbbVie, Protagen, Novartis Biomedical Research, Oliver Distler Consultant of: Abbvie, Caroline Ospelt: None declared, Kerstin Klein Grant/research support from: Novartis Foundation for biomedical research (2019)

Computed tomography-based radiomics decodes prognostic and molecular differences in interstitial lung disease related to systemic sclerosis

Background

Radiomic features calculated from routine medical images show great potential for personalised medicine in cancer. Patients with systemic sclerosis (SSc), a rare, multiorgan autoimmune disorder, have a similarly poor prognosis due to interstitial lung disease (ILD). Here, our objectives were to explore computed tomography (CT)-based high-dimensional image analysis (“radiomics”) for disease characterisation, risk stratification and relaying information on lung pathophysiology in SSc-ILD.

Methods

We investigated two independent, prospectively followed SSc-ILD cohorts (Zurich, derivation cohort, n=90; Oslo, validation cohort, n=66). For every subject, we defined 1355 robust radiomic features from standard-of-care CT images. We performed unsupervised clustering to identify and characterise imaging-based patient clusters. A clinically applicable prognostic quantitative radiomic risk score (qRISSc) for progression-free survival (PFS) was derived from radiomic profiles using supervised analysis. The biological basis of qRISSc was assessed in a cross-species approach by correlation with lung proteomic, histological and gene expression data derived from mice with bleomycin-induced lung fibrosis.

Results

Radiomic profiling identified two clinically and prognostically distinct SSc-ILD patient clusters. To evaluate the clinical applicability, we derived and externally validated a binary, quantitative radiomic risk score (qRISSc) composed of 26 features that accurately predicted PFS and significantly improved upon clinical risk stratification parameters in multivariable Cox regression analyses in the pooled cohorts. A high qRISSc score, which identifies patients at risk for progression, was reverse translatable from human to experimental ILD and correlated with fibrotic pathway activation.

Conclusions

Radiomics-based risk stratification using routine CT images provides complementary phenotypic, clinical and prognostic information significantly impacting clinical decision making in SSc-ILD.

CT-based radiomics decodes phenotypic, prognostic and molecular differences in SSc-ILD, and predicts progression-free survival with a significant impact on future clinical decision making in SSc-ILDhttps://bit.ly/3zPaMOn

Cognition and Its Association with Psychosocial and Occupational Functioning during Treatment with Escitalopram in Patients with Major Depressive Disorder: A CAN-BIND-1 Report: La Cognition Et Son Association Avec Le Fonctionnement Psychosocial Et Professionnel Durant Le Traitement Par Escitalopram Chez Des Patients Souffrant De Trouble Dépressif Majeur: Une Étude Can-Bind-1

OBJECTIVES

Major depressive disorder (MDD) is associated with impairments in both cognition and functioning. However, whether cognitive deficits significantly contribute to impaired psychosocial and occupational functioning, independent of other depressive symptoms, is not well established. We examined the relationship between cognitive performance and functioning in depressed patients before and after antidepressant treatment using secondary data from the first Canadian Biomarker Integration Network in Depression-1 study.

METHODS

Cognition was assessed at baseline in unmedicated, depressed participants with MDD (n = 207) using the Central Nervous System Vital Signs computerized battery, psychosocial functioning with the Sheehan Disability Scale (SDS), and occupational functioning with the Lam Employment Absence and Productivity Scale (LEAPS). Cognition (n = 181), SDS (n = 175), and LEAPS (n = 118) were reassessed after participants received 8 weeks of open-label escitalopram monotherapy. A series of linear regressions were conducted to determine (1) whether cognitive functioning was associated with psychosocial and occupational functioning prior to treatment, after adjusting for overall depressive symptom severity and (2) whether changes in cognitive functioning after an 8-week treatment phase were associated with changes in psychosocial and occupational functioning, after adjusting for changes in overall symptom severity.

RESULTS

Baseline global cognitive functioning, after adjusting for depression symptom severity and demographic variables, was associated with the SDS work/study subscale (β = -0.17; P = 0.03) and LEAPS productivity subscale (β = -0.17; P = 0.05), but not SDS total (β = 0.19; P = 0.12) or LEAPS total (β = 0.41; P = 0.17) scores. Although LEAPS and SDS scores showed significant improvements after 8 weeks of treatment (P < 0.001), there were no significant associations between changes in cognitive domain scores and functional improvements.

CONCLUSION

Cognition was associated with occupational functioning at baseline, but changes in cognition were not associated with psychosocial or occupational functional improvements following escitalopram treatment. We recommend the use of more comprehensive functional assessments to determine the impact of cognitive change on functional outcomes in future research.

Individual functions of the histone acetyl transferases CBP and p300 in regulating the inflammatory response of synovial fibroblasts

Chromatin remodeling, and a persistent histone 3 lysine 27 acetylation (H3K27ac) in particular, are associated with a sustained inflammatory response of synovial fibroblasts (SF) in rheumatoid arthritis (RA). Here we investigated individual functions of the writers of H3K27ac marks, the homologues histone acetyl transferases (HAT) CBP and p300, in controlling the constitutive and inflammatory gene expression in RA SF. We applied a silencing strategy, followed by RNA-sequencing and pathway analysis, complemented with the treatment of SF with inhibitors targeting the HAT (C646) or bromo domains (I-CBP) of CBP and p300. We showed that CBP and p300 undertook overlapping and, in particular at gene levels, distinct regulatory functions in SF. p300 is the major HAT for H3K27ac in SF and regulated more diverse pathways than CBP. Whereas both factors regulated genes associated with extracellular matrix remodeling, adhesion and proliferation, p300 specifically controlled developmental genes associated with limb development. Silencing of CBP specifically down regulated the TNF-induced expression of interferon-signature genes. In contrast, silencing of p300 resulted in anti- and pro-inflammatory effects. Integration of data sets derived from RNA-sequencing and chromatin immunoprecipitation sequencing for H3K27ac revealed that changes in gene expression after CBP or p300 silencing could be only partially explained by changes in levels of H3K27ac. Inhibition of CBP/p300 using HAT and bromo domain inhibitors strongly mirrored effects obtained by silencing of p300, including anti- and pro-inflammatory effects, indicating that such inhibitors are not sufficient to be used as anti-inflammatory drugs.

OP0150 MACHINE LEARNING APPROACHES FOR RISK MODELLING IN INTERSTITIAL LUNG DISEASE ASSOCIATED WITH SYSTEMIC SCLEROSIS USING HIGH DIMENSIONAL IMAGE ANALYSIS

Background:

The interstitial lung disease (ILD) associated with connective tissue diseases including systemic sclerosis (SSc) is heterogenous disease characterized by reduced survival of approximately 3 years (1). “Radiomics’’ is a field of research which describes the in-depth analysis of tissues by computational retrieval of high-dimensional quantitative features from medical images (2). Our previous study suggested capacity of radiomics features to differentiate between “high” and “low” risk groups for lung function decline in two independent cohorts (3).

Objectives:

  •bTo develop robust, machine learning (ML) workflow for “radiomics” data in SSc-ILD to select optimal methods for prediction.

  •oTo predict the time to individual lung function decline defined as defined by the time to a relative decline of ≥ 15% in Forced Vital Capacity (FVC)% as previously (3), using workflow.

Methods:

We investigated two cohorts of SSc-ILD: 90 patients (76.7% female, median age 57.5 years) from the University Hospital Zurich and 66 patients (75.8% female, median age 61.0 years) from Oslo University Hospital’s. Patients were retrospectively selected if (3): a) diagnosed with early/mild SSc according to the Very Early Diagnosis of Systemic Sclerosis (VEDOSS) criteria, b) presence of ILD on HRCT as determined by a senior radiologist. For every subject, we defined 1,355 robust radiomic features from HRCT images. The follow-up period was defined as the time interval between baseline visit and the last available follow-up visit.

We have developed a systematic computational workflow to build predictive ML models. To reduce the number of redundant radiomic features, we applied correlation thresholds. We applied distinct methods including 1) Lasso Penalized Regression for feature selection, and 2) Random Forest (RF) for modeling using the R package ‘caret’. To select the optimal ML model, we randomly divided derivation cohort into Training (70%) and Holdout (30%) sets and applied fivefold cross-validation (5kCV) for feature and classifier selection on Training set only.

Results:

We have investigated various methods to select the optimal set of predictive radiomic features. Since the ML model performance is affected by both, feature, and classifier selection, we assessed these factors first.

Results from feature filtering and selection, suggested that the combination of correlation threshold of 0.9 with Lasso regression proved best. As we perform feature selection in 5k CV workflow, features present in at least 2 sets entered model optimization step.

During model selection, we selected RF classifier. We detected positive correlation between actual and predicted values with Spearman’s rho = 0.313, p = 0.167 and Spearman’s rho = 0.341, p = 0.015 in Oslo and Holdout sets respectively, as shown on Figure 1. The percentage of variance remained modest for both Holdout (Rsq = 0.104) and Oslo (Rsq = 0.126) datasets.

Figure 1.

Performance of the best, RF classifier shown as scatterplot between actual and predicted values of individual time to lung decline.

Conclusion:

In summary, we: (1) developed ML workflow that allowed to select o optimal methodology for modeling (i.e., feature and classifier selection), and (2) provide models that predicted time to individual lung function decline, characterized by significant correlation between predicted and actual values.

References:

[1]Hansell DM, Goldin JG, King TE, Jr., Lynch DA, Richeldi L, Wells AU. CT staging and monitoring of fibrotic interstitial lung diseases in clinical practice and treatment trials: a position paper from the Fleischner Society. Lancet Respir Med. 2015;3(6):483-96.

[2]Lambin, P. et al. Radiomics: extracting more information from medical images using advanced feature analysis. Eur. J. Cancer 48, 441–446 (2012).

[3]Schniering J. et al. Resolving phenotypic and prognostic differences in interstitial lung disease related to systemic sclerosis by computed tomography-based radiomics. https://www.medrxiv.org/content/10.1101/2020.06.09.20124800v1

Disclosure of Interests:

None declared

POS0866 TWO-DIMENSIONAL HRCT-BASED RADIOMIC FEATURES IN SSC-ILD DISTINGUISH DRUG RESPONDERS FROM NON-RESPONDERS

Background:

Management of patients with systemic sclerosis-associated interstitial lung disease (SSc-ILD) is complicated by high inter-patient variability. To date, no validated predictors of treatment response are available for routine use. High resolution computed tomography (HRCT)-based radiomics, i.e. the high-dimensional, quantitative analysis of imaging metadata, have previously been shown to be successful in discriminating (SSc-)ILD phenotypes in preclinical and clinical studies1. Since HRCT is an integral part of the routine work-up in SSc, HRCT-based radiomic features may hold potential as non-invasive biomarkers.

Objectives:

To predict treatment response using two-dimensional (2D) HRCT-based radiomics in SSc-ILD patients from a prospectively followed cohort.

Methods:

Inclusion criteria were diagnosis of SSc-ILD in HRCT, availability of a suitable chest HRCT scan within 12 months prior to initiation of a new treatment, and availability of clinical baseline and follow-up information. Treatment response was defined as the absence of all of the following over a follow-up period of 12-24 months: relative decrease in forced vital capacity (FVC) ≥5%, increase of ILD in HRCT as assessed by a radiologist, change in treatment regimen due to insufficient response, ILD-related death or lung transplantation. Of each pre-treatment HRCT, 6 slices (15±5 mm apart, starting from the basal lung margin) were manually segmented and 1513 2D radiomic features were extracted using the in-house software Z-Rad (Python 2.7). Features were Z-score transformed and pre-filtered for inter- and intra-reader robustness (intraclass correlation coefficient >0.85) and inter-feature correlation (Spearman’s rho <0.9). A categorical linear regression model was created using 3-fold cross-validated elastic nets for feature selection. Features were then summarized and divided by their number. For generation of a score cut-off, Youden’s score was used. For two-group analyses of continuous variables, Wilcoxon’s test was performed, whereas categorical data was assessed using Fisher’s exact test.

Results:

A total of 64 pre-treatment HRCTs from 54 patients were analyzed. In 9 patients, >1 asynchronous treatments were assessed, while 45 patients had only 1 eligible treatment approach. The response rate within the assessed follow-up period was 45.3% (n=29). For score generation, 13 radiomic features were selected and an optimal cut-off value of -0.1589 was determined. Univariate linear regression showed significant association between our categorical radiomics-based score and treatment response (p=0.007, area under the curve = 0.65 (0.51-0.79), sensitivity=0.90, specificity=0.43), whereby a high score was predictive for treatment response.

No differences between patients with high (n=46) or low (n=18) scores were detected for baseline age (mean±SD=55.5±12.0 and 55.5±13.6 years, p=0.84), duration of SSc (mean±SD=6.2±8.4 and 4.7±4.4 years, p=0.79), time since ILD diagnosis (2.7±2.9 and 2.4±3.1 years, p=0.59), FVC (77.6±20.6 and 80.1±17.9, p=0.41) or DLco (54.4±21.0 and 57.6±18.9, p=0.40). Distribution of anti-Scl-70 positivity (45.7% vs. 55.6%, p=0.58) and diffuse cutaneous disease (47.7% vs. 61.1%, p=0.41) was not significantly different between patients with high and low scores, respectively, although a trend towards higher percentages in the high score group was observed.

Conclusion:

Our results indicate that, following validation in external cohorts, radiomics may be a promising tool for future pre-treatment patient stratification. Moreover, our radiomics-based score seems not to be associated with commonly studied clinical predictors such as anti-Scl-70 positivity or lung function, underlining a possible additive value to ‘traditional’ clinical parameters.

References:

[1]Schniering, J., et al. Resolving phenotypic and prognostic differences in interstitial lung disease related to systemic sclerosis by computed tomography-based radiomics. medRxiv [Preprint] doi:10.1101/2020.06.09.20124800 (2020).

Disclosure of Interests:

Chantal Meier: None declared, Malgorzata Maciukiewicz: None declared, Matthias Brunner: None declared, Janine Schniering: None declared, Hubert Gabrys: None declared, Anja Kühnis: None declared, Oliver Distler Speakers bureau: Speaker fee on Scleroderma and related complications: Bayer, Boehringer Ingelheim, Medscape, Novartis, Roche. Speaker fee on rheumatology topic other than Scleroderma: MSD, iQone, Novartis, Pfizer, Roche, Consultant of: Consultancy fee for Scleroderma and its complications: Abbvie, Acceleron Pharma, Amgen, AnaMar, Arxx Therapeutics, Bayer, Baecon Discovery, Boehringer, CSL Behring, ChemomAb, Corbus Pharmaceuticals, Horizon Pharmaceuticals, Galapagos NV, GSK, Glenmark Pharmaceuticals, Inventiva, Italfarmaco, iQvia, Kymera, Medac, Medscape, Mitsubishi Tanabe Pharma, MSD, Roche, Roivant Sciences, Sanofi, UCB. Consultancy fee for rheumatology topic other than Scleroderma: Abbvie, Amgen, Lilly, Pfizer, Grant/research support from: Research Grants to investigate the pathophysiology and potential treatment of Scleroderma and its complications: Kymera Therapeutics, Mitsubishi Tanabe, Thomas Frauenfelder: None declared, Stephanie Tanadini-Lang: None declared, Britta Maurer Speakers bureau: Speaker fees from Boehringer-Ingelheim, Grant/research support from: Grant/research support from AbbVie, Protagen, Novartis Biomedical Research, congress support from Pfizer, Roche, Actelion, mepha, and MSD

OP0020 HOMEOBOX D TRANSCRIPTION FACTORS SHAPE DIFFERENTIAL JOINT ENVIRONMENT BETWEEN ANTERIOR FINGER JOINTS AND THUMB

Background:

The expression of embryonic Homeobox (HOX) genes is tightly regulated based on anatomic location in human adult dermal and synovial fibroblasts. Previously, we showed that HOX-D10,-D11 and -D13 are higher expressed in synovial fibroblasts from small distal joints from the hands and feet, in particular in digits II-V and wrists compared to thumb and that this expression pattern is epigenetically imprinted1. The consequences of the tightly restricted expression of these transcription factors are largely unknown.

Objectives:

To elucidate the function of HOXD10, -D11 and -D13 in synovial fibroblasts.

Methods:

Synovial tissues were isolated from paws of naïve C57BL/6 mice (n=8), from patients with rheumatoid arthritis (RA), osteoarthritis (OA) and from healthy controls. Synovial fibroblasts were cultured and transfected with GapmeR to silence HOXD10, -D11, and -D13, respectively or with control GapmeR. RNA sequencing was performed on the NovaSeq platform and pathway analysis was done using R packages and web-based tools (GSEA, EnrichR, Cytoscape). HOXD target gene expression was measured by qPCR (n=3-6).

Results:

To confirm and further analyze the distinctive expression pattern of HOXD genes, we measured their expression in healthy synovial tissues of different joints of human feet and mouse paws. Similar to what we had found in hands, HOXD10, -D11 and -D13 were less abundant in the joints of the first digit of human feet compared to digits II-V (n=3-4 in each joint). Measurements in joints of mouse paws showed lower expression of HoxD10, -D11 and -D13 in distal interphalangeal joints compared to proximal interphalangeal joints and metacarpophalangeal (MCP) joints, respectively. Silencing of HOXD10, -D11 and –D13, affected the expression of 5333, 2217 and 7347 genes, respectively, in cultured RA synovial fibroblasts from human wrists (n=3).

There were more transcripts equally regulated by HOXD10 and -D13 (40% of all HOXD10 and 31% of all HOXD13 regulated transcripts), than by HOXD11 and either -D10 or -D13 (18% of all HOXD10 regulated genes and 16% of all HOXD13 regulated genes), suggesting most redundancy between HOXD10 and -D13 (Figure 1). Among genes differentially expressed in SF isolated from MCP II-V versus thumb joints, 19%, 4% and 33% were regulated by HOXD10, -D11 or –D13, respectively, supporting a role for HOXD13 in particular in shaping the joint specific environment. All three HOXD transcription factors regulated genes involved in cell cycle progression, demonstrating dependence of synovial fibroblasts on these HOX genes for cell division. Other enriched pathways were Toll-like receptor and integrin signaling pathways, regulation of unsaturated fatty acid synthesis and autophagy and extra-cellular matrix protein organization. We could confirm several targets of HOXD10, -D11, and –D13 by qPCR, e.g. NR4A1, ROR2, LIF, ATF3.

Figure 1.

Comparison of the genes which were differentially expressed after HOXD10-11-13 silencing

Conclusion:

The expression of HOXD10, -D11 and –D13 in synovial fibroblasts and tissues strikinglyoverlaps with predilection sites for RA. Silencing experiments suggested that these embryonic HOX transcription factors have a crucial role in regulating fibroblast functions and might shape a joint specific environment that modulates the development and course of RA in specific joints.

References:

[1]Klein K, et al. Ann Rheum Dis 2018;77(Suppl 1):P126

Disclosure of Interests:

Masoumehalsadat Mirrahimi: None declared, Kerstin Klein: None declared, Miranda Houtman: None declared, Malgorzata Maciukiewicz: None declared, Mojca Frank Bertoncelj: None declared, Astrid Juengel: None declared, Martin Berli: None declared, Miriam Marks: None declared, Oliver Distler Speakers bureau:1-Speaker fee on Scleroderma and related complications: Bayer, Boehringer Ingelheim, Medscape, Novartis, Roche

• Speaker fee on rheumatology topic other than Scleroderma: MSD, iQone, Novartis, Pfizer, Roche, Consultant of: • Consultancy fee for Scleroderma and its complications: Abbvie, Acceleron Pharma, Amgen, AnaMar, Arxx Therapeutics, Bayer, Baecon Discovery, Boehringer, CSL Behring, ChemomAb, Corbus Pharmaceuticals, Horizon Pharmaceuticals, Galapagos NV, GSK, Glenmark Pharmaceuticals, Inventiva, Italfarmaco, iQvia, Kymera, Medac, Medscape, Mitsubishi Tanabe Pharma, MSD, Roche, Roivant Sciences, Sanofi, UCB

• Consultancy fee for rheumatology topic other than Scleroderma: Abbvie, Amgen, Lilly, Pfizer, Grant/research support from:. OD has/had consultancy relationship and/or has received research funding in the area of potential treatments for systemic sclerosis and its complications from (last three years):

• Abbvie, Acceleron Pharma, Amgen, AnaMar, Arxx Therapeutics, Baecon Discovery, Blade Therapeutics, Bayer, Boehringer Ingelheim, ChemomAb, Corbus Pharmaceuticals, CSL Behring, Galapagos NV, Glenmark Pharmaceuticals, GSK, Horizon (Curzion) Pharmaceuticals, Inventiva, iQvia, Italfarmaco, iQone, Kymera Therapeutics, Lilly, Medac, Medscape, Mitsubishi Tanabe Pharma, MSD, Novartis, Pfizer, Roche, Sanofi, Serodapharm, Topadur, Target Bioscience and UCB. Patent issued “mir-29 for the treatment of systemic sclerosis” (US8247389, EP2331143).

• Research Grants to investigate the pathophysiology and potential treatment of Scleroderma and its complications: Kymera Therapeutics, Mitsubishi Tanabe, Caroline Ospelt: None declared

Contributions of cholinergic receptor muscarinic 1 and CYP1A2 gene variants on the effects of plasma ratio of clozapine/N-desmethylclozapine on working memory in schizophrenia

BACKGROUND

Clozapine has heterogenous efficacy in enhancing working memory in schizophrenia. We have previously hypothesized that this is due to opposing effects of clozapine and its metabolite, N-desmethylclozapine, at the muscarinic M1 receptor and demonstrated that a lower clozapine/N-desmethylclozapine ratio is associated with better working memory than clozapine or N-desmethylclozapine levels alone.

AIMS

In this study, we expanded the above hypothesis to explore whether genetic variation in the cholinergic receptor muscarinic 1 gene, encoding the M1 receptor, affects the relationship between clozapine/N-desmethylclozapine and working memory. Further, we explored whether CYP1A2 gene variants affect the ratio of clozapine/N-desmethylclozapine and by this, working memory performance.

METHODS

We evaluated two functionally significant single nucleotide polymorphisms, rs1942499 and rs2075748, in cholinergic receptor muscarinic 1, with the haplotype T-A associated with lower transcriptional activity than the haplotype C-G. Further, we examined CYP1A2 *1F, with *1F/*1F conferring high inducibility in the presence of smoking.

RESULTS

In a sample of 30 patients with schizophrenia on clozapine monotherapy, clozapine/N-desmethylclozapine was correlated with working memory only in non-carriers of the haplotype T-A of the cholinergic receptor muscarinic 1 gene. Interaction of CYP1A2 genotype and smoking status significantly affected clozapine concentrations, but there were no significant effects of CYP1A2 genotype and smoking status on the relationship between clozapine/N-desmethylclozapine on working memory.

CONCLUSIONS

Our finding that the relationship between clozapine/N-desmethylclozapine and working memory is specific to patients with potentially higher transcription of M1 receptor (i.e. non-carriers of the haplotype T-A of cholinergic receptor muscarinic 1) supports a cholinergic mechanism underlying this relationship.

Association between the -2548G/A polymorphism of the leptin gene and antipsychotic-induced weight gain: Analysis of the CATIE sample and meta-analysis

BACKGROUND

Antipsychotics, especially most of the second-generation antipsychotics, have a high risk for metabolic syndrome and antipsychotic-induced weight gain (AIWG). A promoter variant of the leptin (LEP) gene, -2548G/A (rs7799039), has been associated with AIWG in several studies. The aim of this study was to evaluate this association in the Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE) sample, followed by meta-analysis.

METHODS

We investigated the association between rs7799039 and AIWG in a sub-sample of European (N = 164) individuals from the CATIE study. Body mass index (BMI) change and weight gain (presence or absence) was analyzed using ANCOVA and logistic regression, respectively. For the meta-analysis, a literature search was conducted using MEDLINE, Embase, and PsycINFO up to October 2019. The pooled odds ratio was calculated for presence or absence of weight gain (≥7% weight change) using a random effects model.

RESULTS

We did not detect an association between rs7799039 and BMI change or weight gain (presence or absence) in the CATIE sample. As for the meta-analysis, we included 12 studies. No significant associations between the LEP rs7799039 polymorphism and AIWG were observed under the allelic genetic model (allele A vs. allele G) (OR = 1.10 [0.71, 1.70], p = .68). In the subgroup analyses of first-episode schizophrenia patients, a significant association between the A-allele and weight gain was observed, respectively (OR = 2.32 [1.41, 3.82], p = .0009).

CONCLUSIONS

The present meta-analysis showed no significant effect of rs7799039 on AIWG. However, this variant may influence AIWG in first-episode schizophrenia patients. Further investigation of a larger and more homogenous sample is required to elucidate the role of the LEP gene in AIWG.

Validation study of microRNAs previously associated with antidepressant response in older adults treated for late-life depression with venlafaxine

BACKGROUND

MicroRNAs (miRNAs) are small 22 nucleotides long, non-coding RNAs that are potential biomarkers for antidepressant treatment response. We aimed to replicate previous associations of miRNAs with antidepressant treatment response in a sample of older adults diagnosed with late-life depression.

METHODS

Our sample included 184 older adults diagnosed with moderately severe depression that received open-label venlafaxine (up to 300 mg/day) for approximately 12 weeks. We quantified miRNA expression levels at baseline and week 12 for miRNAs miR-1202, miR-135a-5p, miR-16-5p, miR-146a-5p, miR-146b-5p, miR-425-3p, and miR-24-3p to explore their association with remission status, response trajectories, and time-to-remission.

RESULTS

At T0 and T12, there were no differences in miRNA expression levels between remitters and non-remitters. However, remitters showed a trend toward higher baseline miR-135a-5p (Median = 11.3 [9.9, 15.7], p = .083). Prior to correction, baseline miR-135a-5p expression levels showed an association with remission status (OR = 1.8 [1.0, 3.3], p = .037). Individuals with higher baseline miR-135a-5p showed better response trajectories (F = 4.5, FDR-corrected p = 4.4 × 10^-4), particularly at weeks 10 and 12 (p < .05). In addition, individuals with higher miR-135a-5p expression reached remission faster than those with lower expression (HR = 0.6 [0.4, 0.9], FDR-corrected p = .055).

LIMITATIONS

Although the sample size was relatively modest, our findings are consistent with the literature suggesting that higher miR-135a-5p levels may be associated with better antidepressant treatment response.

CONCLUSIONS

However, the miRNA signature of antidepressant response in older adults may be different as compared to younger adults.

Pharmacogenetic Implications for Antidepressant Pharmacotherapy in Late-Life Depression: A Systematic Review of the Literature for Response, Pharmacokinetics and Adverse Drug Reactions

Affecting up to 15% of older adults, late-life depression (LLD) is characterized by the occurrence of depressive symptoms after the age of 50-65 years and maybe pathophysiologically distinct from depression in younger adults. Therefore, LLD is challenging to treat, and predictive genetic testing might be essential to improve treatment in this vulnerable population. The current review aims to provide a summary of the literature exploring genetic associations with antidepressant treatment outcomes in late-life. We conducted a systematic search of three integrated electronic databases. We identified 29 articles investigating genetic associations with antidepressant treatment outcomes, pharmacokinetic parameters, and adverse drug reactions in older adults. Given the small number of investigations conducted in older adults, it is difficult to conclude the presence or absence of genetic associations with the outcomes of interest. In sum, the most substantial amount of evidence exists for the CYP2D6 metabolizer status, SLC6A4 5-HTTLPR, and BDNF rs6265. These findings are consistent in the literature when not restricting to older adults, suggesting that similar treatment recommendations may be provided for older adults regarding genetic variation, such as those outlined for CYP2D6 by the Clinical Pharmacogenetics Implementation Consortium. Nonetheless, further studies are required in well-characterized samples, including genome-wide data, to validate if similar treatment adjustments are appropriate in older adults, given that there appear to be significant effects of genetic variation on antidepressant treatment factors.

SAT0569 “IMAGES ARE MORE THAN PICTURES, THEY ARE DATA” [1] – EXPLORATION OF RADIOMICS ANALYSIS FOR SYSTEMIC SCLEROSIS-ASSOCIATED INTERSTITIAL LUNG DISEASE

Background:

Interstitial lung disease (ILD) affects 60% of patients with systemic sclerosis (SSc) and is the primary cause of death. Medical imaging is an integral part of the routine work-up for diagnosis and monitoring of SSc-ILD and includes high-resolution computed tomography (HRCT). Radiomics is a novel research area that describes the in-depth analysis of tissue phenotypes in medical images with computational retrieval of quantitative, mineable metadata appropriate for statistical analyses.

Objectives:

To explore the performance of HRCT-derived radiomic features for the assessment of SSc-associated ILD (i.e. diagnosis, staging, and lung function).

Methods:

Radiomics analysis was performed on HRCT scans from 98 SSc patients, including n=33 SSc patients without ILD, n=33 with limited and n=32 with extensive ILD as defined by 0%, <20% and ≥20% visual extent of fibrosis on HRCT, respectively. Following semi-automated segmentation of lung tissue on 3D reconstructed HRCT scans, 1386 radiomic features, including 17 intensity, 137 texture, and 1232 wavelet features were extracted using the in-house developed software Z-Rad (Python 2.7). In order to identify robust features, we conducted intra- and inter-reader correlation analysis (ICC) in a subgroup of patients. Only features with good reproducibility (ICC ≥ 0.75) entered subsequent analyses. We applied the Wilcoxon test, followed by Receiver Operating Characteristic ROC) curve analyses, to identify features significantly different between a) ILD and non-ILD and b) limited vs. extensive ILD patients. Spearman rank correlation was performed to reveal significant associations of radiomic features from a) and b) with lung function as measured by percentage of predicted forced vital capacity (FVC% predicted).

Results:

In total, 1355/1386 radiomic features passed the test of robustness and were eligible for further, exploratory analyses. Radiomic features with good performance (area under the ROC curve (AUC) ≥ 0.7 and p-value ≤ 0.05) were considered as potential candidate discriminators. Under this criterion, we identified 288/1355 (21.3%) radiomic features that were significantly different between ILD and non-ILD patients and 409/1355 (30.2%) features that significantly discriminated between limited and extensive ILD (Fig. 1). For diagnosis, the texture featuredependence count entropywas the top parameter to distinguish ILD patients from healthy controls (AUC = 0.89, p = 1.83x10-10), whereas for staging the wavelet featureHHH long run high grey level emphasisproved to be best suited to separate limited from extensive ILD (AUC = 0.88, p = 7.76x10-9).

Fig 1.

Correlation analysis of the most significant (best performing) discriminative radiomic features with lung function revealed a significant negative correlation ofdependence count entropy(rho = -0.51, p = 9.89x10-8) andHHH long run high grey level emphasis(rho = -0.51, p = 1.73x10-5) with FVC% predicted.

Conclusion:

Our study adds novelty to the field of SSc-ILD showing that radiomic features have great potential as quantitative imaging biomarkers for diagnosis and staging of SSc-ILD and that they may reflect lung function. As the next step, we are planning to build predictive models, using machine learning, for diagnosis, staging, and lung function and validate them in external patient cohorts. If validated such models will pave the way for computer-aided management in SSc-ILD and thus improve patients’ outcome.

References:

[1]Gillies, R. J., Kinahan, P. E. & Hricak, H. Radiomics: Images Are More than Pictures, They Are Data. Radiology 278, 563-577, doi:10.1148/radiol.2015151169 (2016).

Disclosure of Interests:

Janine Schniering: None declared, Malgorzata Maciukiewicz: None declared, Hubert Gabrys: None declared, Matthias Brunner: None declared, Christian Blüthgen: None declared, Oliver Distler Grant/research support from: Grants/Research support from Actelion, Bayer, Boehringer Ingelheim, Competitive Drug Development International Ltd. and Mitsubishi Tanabe; he also holds the issued Patent on mir-29 for the treatment of systemic sclerosis (US8247389, EP2331143)., Consultant of: Consultancy fees from Actelion, Acceleron Pharma, AnaMar, Bayer, Baecon Discovery, Blade Therapeutics, Boehringer, CSL Behring, Catenion, ChemomAb, Curzion Pharmaceuticals, Ergonex, Galapagos NV, GSK, Glenmark Pharmaceuticals, Inventiva, Italfarmaco, iQvia, medac, Medscape, Mitsubishi Tanabe Pharma, MSD, Roche, Sanofi and UCB, Speakers bureau: Speaker fees from Actelion, Bayer, Boehringer Ingelheim, Medscape, Pfizer and Roche, Matthias Guckenberger: None declared, Thomas Frauenfelder: None declared, Stephanie Tanadini-Lang: None declared, Britta Maurer Grant/research support from: AbbVie, Protagen, Novartis, congress support from Pfizer, Roche, Actelion, and MSD, Speakers bureau: Novartis

Genome-wide association study on antipsychotic-induced weight gain in Europeans and African-Americans

BACKGROUND

Antipsychotic (AP) medications are the first line of treatment for schizophrenia. However, most conferr a risk of antipsychotic-induced weight gain (AIWG). The objective of this investigation was to conduct a genome-wide association study (GWAS) of AIWG, followed by comprehensive, post-GWAS approaches.

METHODS

We investigated n = 201 schizophrenia or schizoaffective disorder patients of European and African American ancestry who were treated primarily with clozapine or olanzapine. We conducted a genome-wide association analysis for AIWG, defined primarily as a percentage of weight change from baseline.

RESULTS

When examining Europeans (n = 147), we noticed an association between rs62097526 (β = 0.39, p = 3.59 × 10-6, CADD = 2.213) variant, located downstream of the CIDEA gene, which is considered a risk factor for AIWG. In the entire sample, we observed a significant association between rs1525085 (β = 0.411, p = 3.15 × 10-9) variant of the DGKB gene and AIWG. The association was nominally significant in Europeans (β = 0.271, p = 0.002) and African Americans (β = 0.579, p = 5.73 × 10-5) with the same risk allele. Our top genes (p < 5 × 10-5) were enriched in the GWAS catalog for the risk of obesity and interacted with the known risk factors for obesity (G6PD) and diabetes (IRS1). In addition, these genes are targeted by miRNAs related to schizophrenia (mir-34a) and obesity (mir-19b). However, our polygenic risk score analyses did not provide support for major genetic overlap between obesity and the risk of AIWG.

CONCLUSIONS

In summary, we propose that the CIDEA and DGKB genes are risk factors for AIWG in transethnic populations. Additionally, our evidence suggests that the G6PD and IRS1 gene-related pathways might be involved in AIWG.

GWAS-based machine learning approach to predict duloxetine response in major depressive disorder

Major depressive disorder (MDD) is one of the most prevalent psychiatric disorders and is commonly treated with antidepressant drugs. However, large variability is observed in terms of response to antidepressants. Machine learning (ML) models may be useful to predict treatment outcomes. A sample of 186 MDD patients received treatment with duloxetine for up to 8 weeks were categorized as "responders" based on a MADRS change >50% from baseline; or "remitters" based on a MADRS score ≤10 at end point. The initial dataset (N = 186) was randomly divided into training and test sets in a nested 5-fold cross-validation, where 80% was used as a training set and 20% made up five independent test sets. We performed genome-wide logistic regression to identify potentially significant variants related to duloxetine response/remission and extracted the most promising predictors using LASSO regression. Subsequently, classification-regression trees (CRT) and support vector machines (SVM) were applied to construct models, using ten-fold cross-validation. With regards to response, none of the pairs performed significantly better than chance (accuracy p > .1). For remission, SVM achieved moderate performance with an accuracy = 0.52, a sensitivity = 0.58, and a specificity = 0.46, and 0.51 for all coefficients for CRT. The best performing SVM fold was characterized by an accuracy = 0.66 (p = .071), sensitivity = 0.70 and a sensitivity = 0.61. In this study, the potential of using GWAS data to predict duloxetine outcomes was examined using ML models. The models were characterized by a promising sensitivity, but specificity remained moderate at best. The inclusion of additional non-genetic variables to create integrated models may improve prediction.

Genome-wide association studies of placebo and duloxetine response in major depressive disorder

We investigated variants associated with treatment response in depressed patients treated with either the antidepressant duloxetine or placebo using a genome-wide approach. Our sample (N=391) included individuals aged 18-75 years, diagnosed with major depressive disorder and treated with either duloxetine or placebo for up to 8 weeks. We conducted genome-wide associations for treatment response as operationalized by percentage change in Montgomery-Åsberg Depression Rating Scale score from baseline, as well as mixed models analyses across five time points. In the placebo-treated subsample (N=205), we observed a genome-wide association with rs76767803 (β=0.69, P=1.25 × 10^-8) upstream of STAC1. STAC1 rs76767803 was also associated with response using mixed model analysis (χ²=3.95; P=0.001). In the duloxetine-treated subsample (N=186), we observed suggestive associations with ZNF385D (rs4261893; β=-0.46, P=1.55 × 10^-5), NCAM1 (rs2303377; β=0.45, P=1.76 × 10^-5) and MLL5 (rs117986340; β=0.91, P=3.04 × 10^-5). Our findings suggest that a variant upstream of STAC1 is associated with placebo response, which might have implications for treatment optimization, clinical trial design and drug development.

Norepinephrine Transporter Gene Variants and Remission From Depression With Venlafaxine Treatment in Older Adults

OBJECTIVE

The primary objective of this study was to investigate five putatively functional variants of the norepinephrine transporter (SLC6A2, NET) and serotonin transporter (SLC6A4, SERT) genes and remission in depressed older adults treated with venlafaxine. A secondary objective was to analyze 17 other variants in serotonergic system genes (HTR1A, HTR2A, HTR1B, HTR2C, TPH1, TPH2) potentially involved in the mechanism of action of venlafaxine.

METHOD

The sample included 350 adults age 60 or older with DSM-IV-defined major depressive disorder and a score of at least 15 on the Montgomery-Åsberg Depression Rating Scale (MADRS). Participants received protocolized treatment with open-label venlafaxine, up to 300 mg/day for approximately 12 weeks, as part of a three-site clinical trial. Each individual was genotyped for 22 polymorphisms in eight genes, which were tested for association with venlafaxine remission (a MADRS score ≤10) and changes in MADRS score during treatment.

RESULTS

After adjusting for multiple comparisons, NET variant rs2242446 (T-182C) was significantly associated with remission (odds ratio=1.66, 95% CI=1.13, 2.42). Individuals with the rs2242446 C/C genotype were more likely to remit (73.1%) than those with either the C/T (51.8%) or the T/T genotype (47.3%). Individuals with the C/C genotype also had a shorter time to remission than those with the C/T or T/T genotypes and had a greater percentage change in MADRS score from baseline to end of treatment (up to week 12).

CONCLUSIONS

NET rs2242446/T-182C may serve as a biomarker to predict the likelihood of remission with venlafaxine in older adults with major depression. These findings may help to optimize antidepressant outcomes in older adults.

An Association Between Functional Polymorphisms of the Interleukin 1 Gene Complex and Schizophrenia Using Transmission Disequilibrium Test

IL1 gene complex has been implicated in the etiology of schizophrenia. To assess whether IL1 gene complex is associated with susceptibility to schizophrenia in Polish population we conducted family-based study. Functional polymorphisms from IL1A (rs1800587, rs17561, rs11677416), IL1B (rs1143634, rs1143643, rs16944, rs4848306, rs1143623, rs1143633, rs1143627) and IL1RN (rs419598, rs315952, rs9005, rs4251961) genes were genotyped in 143 trio with schizophrenia. Statistical analysis was performed using transmission disequilibrium test. We have found a trend toward an association of rs1143627, rs16944, rs1143623 in IL1B gene with the risk of schizophrenia. Our results show a protective effect of allele T of rs4251961 in IL1RN against schizophrenia. We also performed haplotype analysis of IL1 gene complex and found a trend toward an association with schizophrenia of GAGG haplotype (rs1143627, rs16944, rs1143623, rs4848306) in IL1B gene, haplotypes: TG (rs315952, rs9005) and TT (rs4251961, rs419598) in IL1RN. Haplotype CT (rs4251961, rs419598) in IL1RN was found to be associated with schizophrenia. After correction for multiple testing associations did not reach significance level. Our results might support theory that polymorphisms of interleukin 1 complex genes (rs1143627, rs16944, rs1143623, rs4848306 in IL1B gene and rs4251961, rs419598, rs315952, rs9005 in IL1RN gene) are involved in the pathogenesis of schizophrenia, however, none of the results reach significance level after correction for multiple testing.

Catechol-O-Methyltransferase Val158Met Polymorphism and Clinical Response to Antipsychotic Treatment in Schizophrenia and Schizo-Affective Disorder Patients: a Meta-Analysis

BACKGROUND

The catechol-O-methyltransferase (COMT) enzyme plays a crucial role in dopamine degradation, and the COMT Val158Met polymorphism (rs4680) is associated with significant differences in enzymatic activity and consequently dopamine concentrations in the prefrontal cortex. Multiple studies have analyzed the COMT Val158Met variant in relation to antipsychotic response. Here, we conducted a meta-analysis examining the relationship between COMT Val158Met and antipsychotic response.

METHODS

Searches using PubMed, Web of Science, and PsycInfo databases (03/01/2015) yielded 23 studies investigating COMT Val158Met variation and antipsychotic response in schizophrenia and schizo-affective disorder. Responders/nonresponders were defined using each study's original criteria. If no binary response definition was used, authors were asked to define response according to at least 30% Positive and Negative Syndrome Scale score reduction (or equivalent in other scales). Analysis was conducted under a fixed-effects model.

RESULTS

Ten studies met inclusion criteria for the meta-analysis. Five additional antipsychotic-treated samples were analyzed for Val158Met and response and included in the meta-analysis (ntotal=1416). Met/Met individuals were significantly more likely to respond than Val-carriers (P=.039, ORMet/Met=1.37, 95% CI: 1.02-1.85). Met/Met patients also experienced significantly greater improvement in positive symptoms relative to Val-carriers (P=.030, SMD=0.24, 95% CI: 0.024-0.46). Posthoc analyses on patients treated with atypical antipsychotics (n=1207) showed that Met/Met patients were significantly more likely to respond relative to Val-carriers (P=.0098, ORMet/Met=1.54, 95% CI: 1.11-2.14), while no difference was observed for typical-antipsychotic-treated patients (n=155) (P=.65).

CONCLUSIONS

Our findings suggest that the COMT Val158Met polymorphism is associated with response to antipsychotics in schizophrenia and schizo-affective disorder patients. This effect may be more pronounced for atypical antipsychotics.

Association of the glucocorticoid receptor gene polymorphisms and their interaction with stressful life events in Polish adolescent girls with anorexia nervosa

BACKGROUND

Disturbances in stress response mechanisms and hypothalamic-pituitary-adrenal axis (HPA) functioning are considered important factors involved in the pathophysiology of anorexia nervosa (AN). Thus, genetic variations in the end effector of HPA - glucocorticoid receptor gene and relationships to stressful life events (SLE) may be connected to a higher risk of illness. The aim of the study was examining the association between glucocorticoid receptor gene (NR3C1) polymorphisms and risk factors among stressful life events in AN patients.

SUBJECTS AND METHODS

This study comprised 256 patients with AN and 167 control subjects. The questionnaires examining brief history of the mother's pregnancy and long-acting stress factors, as well as life events checklist to assess stressful life events during the 6 months prior to hospitalization were used. The eight common SNPs (rs6198, rs6191, rs6196, rs258813, rs33388, rs41423247, rs56149945 and rs10052957) of NR3C1 gene were genotyped.

RESULTS

The association of five polymorphisms (rs6191, rs258813, rs33388, rs41423247 and rs10052957) and one complex allele (TCAGT) of NR3C1 gene with increased risk of AN were found. However, no significant correlations between early, long-acting and predicting hospitalization SLE and any of the analyzed polymorphisms were observed.

CONCLUSIONS

The results confirm that the NR3C1 gene is associated with AN risk regardless of the type of stressful triggering factors.

Preliminary evidence for association of genome-wide significant DRD2 schizophrenia risk variant with clozapine response

AIM

The recent Psychiatric Genomics Consortium genome-wide association study identified an SNP, rs2514218, located 47kb upstream of the DRD2 gene to be associated with risk for schizophrenia (p = 2.75e-11). Since all antipsychotics bind to dopamine D2 receptors, we examined rs2514218 in relation to response to antipsychotic treatment.

PATIENTS & METHODS

We investigated the SNP in relation to treatment response in a prospective study consisting of 208 patients (151 Caucasians, 42 African-Americans and 15 others) treated with clozapine for 6 months.

RESULTS

rs2514218 was associated with total score change in the brief psychiatric rating scale under an additive model (pcorr= 0.033).

CONCLUSION

Our finding provides evidence for rs2514218 association with antipsychotic response, but further replication is required before firm conclusions can be drawn.

Association study of functional polymorphisms in interleukins and interleukin receptors genes: IL1A, IL1B, IL1RN, IL6, IL6R, IL10, IL10RA and TGFB1 in schizophrenia in Polish population

Schizophrenia has been associated with a large range of autoimmune diseases, with a history of any autoimmune disease being associated with a 45% increase in risk for the illness. The inflammatory system may trigger or modulate the course of schizophrenia through complex mechanisms influencing neurodevelopment, neuroplasticity and neurotransmission. In particular, increases or imbalance in cytokine before birth or during the early stages of life may affect neurodevelopment and produce vulnerability to the disease. A total of 27 polymorphisms of IL1N gene: rs1800587, rs17561; IL1B gene: rs1143634, rs1143643, rs16944, rs4848306, rs1143623, rs1143633, rs1143627; IL1RN gene: rs419598, rs315952, rs9005, rs4251961; IL6 gene: rs1800795, rs1800797; IL6R gene: rs4537545, rs4845617, rs2228145, IL10 gene: rs1800896, rs1800871, rs1800872, rs1800890, rs6676671; IL10RA gene: rs2229113, rs3135932; TGF1B gene: rs1800469, rs1800470; each selected on the basis of molecular evidence for functionality, were investigated in this study. Analysis was performed on a group of 621 patients with diagnosis of schizophrenia and 531 healthy controls in Polish population. An association of rs4848306 in IL1B gene, rs4251961 in IL1RN gene, rs2228145 and rs4537545 in IL6R with schizophrenia have been observed. rs6676671 in IL10 was associated with early age of onset. Strong linkage disequilibrium was observed between analyzed polymorphisms in each gene, except of IL10RA. We observed that haplotypes composed of rs4537545 and rs2228145 in IL6R gene were associated with schizophrenia. Analyses with family history of schizophrenia, other psychiatric disorders and alcohol abuse/dependence did not show any positive findings. Further studies on larger groups along with correlation with circulating protein levels are needed.

Genetic variation in IL-1β, IL-2, IL-6, TSPO and BDNF and response to duloxetine or placebo treatment in major depressive disorder

AIM

This study investigated polymorphisms of five inflammatory-related genes for association with duloxetine and placebo response in patients with major depression.

PATIENTS & METHODS

Twenty SNPs in IL-1β, IL-2, IL-6, TSPO and BDNF were genotyped in major depressive disorder patients treated with either duloxetine (n = 215) or placebo (n = 235) for up to 8 weeks. Treatment response was measured with the Montgomery-Åsberg Depression Rating Scale.

RESULTS

IL-6 variants rs2066992 and rs10242595 were nominally associated with response to duloxetine (p = 0.047 and p = 0.028, respectively). Notably, the variant rs2066992 was also associated with placebo response (p = 0.026). However, none of our results remained significant after correction for multiple testing.

CONCLUSION

Our findings tentatively suggest that IL-6 variants play a role in duloxetine and placebo response, which warrants further investigation.

FKBP5 polymorphism is associated with major depression but not with bipolar disorder

BACKGROUND

Altered activity of hypothalamus-pituitary-adrenal glands (HPA) axis in response to stress underlies the pathogenesis of mood disorders such as depression and bipolar disorder. Chaperone proteins regulate sensitivity of glucocorticoid receptor (GR) to steroids. We hypothesized that genetic variants within the FKBP5 - gene encoding co-chaperone protein essential in GR signaling - may influence the susceptibility to major depressive disorder and bipolar disorder.

METHODS

In the study participated 528 bipolar patients, 218 patients with major depressive disorder and 742 subjects from control group. Genotypes for eight FKBP5 polymorphisms (rs1360780, rs755658, rs9470080, rs4713916, rs7748266, rs9296158, rs9394309, rs3800373) were established by TagMan SNP Genotyping Assays (Applied Biosystems). Linkage disequilibrium analysis for FKBP5 gene was done in Haploview. Gene-gene interactions between FKBP5 and NR3C1 polymorphisms (reported previously) were analyzed using the multidimensionality-reduction method (MDR).

RESULTS

We have observed an association between five FKBP5 polymorphisms (rs1360780, rs9470080, rs4713916, rs9296158 and rs9394309) and major depressive disorder (p=0.011; p=0.007, p=0.038; p=0.030; p=0.018, respectively), but not bipolar disorder. In linkage disequilibrium analysis we found that seven FKBP5 polymorphisms build haplotype block (rs3800373, rs755658, rs9296158, rs7748266, rs1360780, rs9394309, rs9470080, respectively). We observed that two haplotype combinations (ACATTGT and CCACTAT) were significantly more frequent in the MDD patients than in controls (p=0.014 and p=0.043). We have not observed such an association for BD patients. We have found that interaction between rs9470080 of FKBP5 and rs6198 of NR3C1 influences MDD risk.

LIMITATIONS

The main limitations of this study include low power and limited sample size of MDD patients.

CONCLUSIONS

Single markers and haplotypes of FKBP5 gene and the interaction with glucocorticoid receptor gene (NR3C1) may influence MDD predisposition.

Analysis of genetic association and epistasis interactions between circadian clock genes and symptom dimensions of bipolar affective disorder

Bipolar affective disorder (BD) is a severe psychiatric disorder characterized by periodic changes in mood from depression to mania. Disruptions of biological rhythms increase risk of mood disorders. Because clinical representation of disease is heterogeneous, homogenous sets of patients are suggested to use in the association analyses. In our study, we aimed to apply previously computed structure of bipolar disorder symptom dimension for analyses of genetic association. We based quantitative trait on: main depression, sleep disturbances, appetite disturbances, excitement and psychotic dimensions consisted of OPCRIT checklist items. We genotyped 42 polymorphisms from circadian clock genes: PER3, ARNTL, CLOCK and TIMELSSS from 511 patients BD (n = 292 women and n = 219 men). As quantitative trait we used clinical dimensions, described above. Genetic associations between alleles and quantitative trait were performed using applied regression models applied in PLINK. In addition, we used the Kruskal-Wallis test to look for associations between genotypes and quantitative trait. During second stage of our analyses, we used multidimensional scaling (multifactor dimensionality reduction) for quantitative trait to compute pairwise epistatic interactions between circadian gene variants. We found association between ARNTL variant rs11022778 main depression (p = 0.00047) and appetite disturbances (p = 0.004). In epistatic interaction analyses, we observed two locus interactions between sleep disturbances (p = 0.007; rs11824092 of ARNTL and rs11932595 of CLOCK) as well as interactions of subdimension in main depression and ARNTL variants (p = 0.0011; rs3789327, rs10766075) and appetite disturbances in depression and ARNTL polymorphism (p = 7 × 10(-4); rs11022778, rs156243).