One of the main obstacles in providing effective treatments for major depressive disorder (MDD) is clinical heterogeneity, whose neurobiological correlates are not clearly defined. A biologically meaningful stratification of depressed patients is needed to promote tailored diagnostic procedures.

Using structural data, we performed an unsupervised clustering to define clinically meaningful clusters of depressed patients.

T1-weighted and diffusion tensor images were obtained from 102 MDD patients. In 64 patients, clinical symptoms, number of stressful life events, severity and exposure to adverse childhood experiences were evaluated using the Beck Depression Inventory (BDI), Schedule of Recent Experiences (SRE), Risky Family Questionnaire (RFQ), and Childhood Trauma Questionnaire (CTQ). Clustering analyses were performed with extracted tract-based fractional anisotropy (TBSS, FSL), cortical thickness, surface area, and regional measures of grey matter volumes (CAT12). Gaussian mixture model was implemented for clustering, considering Support Vector Machine (SVM) as classifier. A 10x2 repeated cross-validation with grid search was performed for hyperparameters tuning and clusters’ stability. The optimal number of clusters was determined by normalized stability, Akaike and Bayesian information criterion. Analyses were adjusted for total intracranial volume, age, and sex. The clinical relevance of the identified clusters was assessed through MANOVA, considering domains of clinical scales as dependent variables and clusters’ labels as fixed factors. Discriminant analysis was subsequently performed to assess the discriminative power of these variables.

Cross-validated clustering approach identified 2 highly stable clusters (normalized stability=0.316, AIC=-80292.48, BIC=351329.16). MANOVA showed a significant between-clusters difference in clinical scales scores (p=0.038). Discriminant analysis distinguished the two clusters with an accuracy of 78.1%, with BDI behavioural and CTQ minimisation/denial domains showing the highest discriminant values (0.325 and 0.313).

Our results defined two biologically informed clusters of MDD patients associated with childhood trauma and specific clinical profiles, which may assist in targeting effective interventions and treatments.

None Declared

Depression is the predominant mood alteration in bipolar disorder (BD), leading to overlapping symptomatology with major depressive disorder (MDD). Consequently, in clinical assessment, almost 60% of BD patients are misdiagnosed as affected by MDD. This calls for the creation of a framework for the differentiation of BD and MDD patients based on reliable biomarkers. Since machine learning (ML) enables to make predictions at the single-subject level, it appears to be particularly suitable for this task.

We implemented a ML pipeline for the differentiation between depressed BD and MDD patients based on structural neuroimaging features.

Diffusion tensor imaging (DTI) and T1-weighted magnetic resonance imaging (MRI) data were acquired for 282 depressed BD (n=180) and MDD (n=102) patients. Axial (AD), radial (RD), mean (MD) diffusivity, and fractional anisotropy (FA) maps were extracted from DTI images, and voxel-based morphometry (VBM) measures were obtained from T1-weighted images. Each feature was entered separately into a 5-fold nested cross-validated ML pipeline differentiating between BD and MDD patients, comprising: confound regression for nuisance variables removal (i.e., age and sex), feature standardization, principal component analysis, and an elastic-net penalized regression. The models underwent 5000 random permutations as a test for significance, and the McNemar’s test was used to assess whether there was any significant difference between the models (significance threshold was set to p<0.05).

The performance of the models and the results of the permutation tests are summarized in Table 1. McNemar’s test showed that the AD-, RD-, MD-, and FA-based models did not differ between each other and were significantly different from the VBM.Table 1.Models’ performance and p-value at 5000 permutation test.

In conclusion, our models differentiated between BD and MDD patients at the single-subject level with good accuracy using structural MRI data. Notably, the models based on white matter integrity measures relying on true information, rather than chance.

None Declared

Bipolar patients (BP) frequently have cognitive deficits, that impact on prognosis and quality of life. Finding biomarkers for this condition is essential to improve patients’ healthcare. Given the association between cognitive dysfunctions and structural brain abnormalities, we used a machine learning approach to identify patients with cognitive deficits.

The aim of this study was to assess if structural neuroimaging data could identify patients with cognitive impairments in several domains using a machine learning framework.

Diffusion tensor imaging and T1-weighted images of 150 BP were acquired and both grey matter voxel-based morphometry (VBM) and tract-based white matter fractional anisotropy (FA) measures were extracted. Support vector machine (SVM) models were trained through a 10-fold nested cross-validation with subsampling. VBM and FA maps were entered separately and in combination as input features to discriminate BP with and without deficits in six cognitive domains, assessed through the Brief Assessment of Cognition in Schizophrenia.

The best classification performance for each cognitive domain is illustrated in Table 1. FA was the most relevant neuroimaging modality for the prediction of verbal memory, verbal fluency, and executive functions deficits, whereas VBM was more predictive for working memory and motor speed domains.Table 1.Performance of best classification models.

Overall, the tested SVM models showed a good predictive performance. Although only partially, our results suggest that different structural neuroimaging data can predict cognitive deficits in BP with accuracy higher than chance level. Unexpectedly, only for the attention and processing speed domain the best model was obtained combining the structural features. Future research may promote data fusion methods to develop better predictive models.

None Declared

Choroid plexus (CP) is a physiological barrier, producing cerebrospinal fluid (CSF), neurotrophic, and inflammatory factors. It’s also involved in the neuro-immune axis, facilitating the interplay between central and peripheral inflammation, allowing trafficking of immune cells. Coherently, CP enlargement has been found in psychiatric diseases characterized by inflammatory signature. Although CP volume correlates with central microglia activation in major depressive disorder (MDD), it’s never been directly associated with peripheral markers in mood disorders.

Examine CP volume in mood disorders and healthy controls (HC) in relation to clinical features and peripheral inflammatory markers.

CP volume was extracted with FreeSurfer in 72 HC and 152 age- and sex-matched depressed patients: 79 BD and 73 MDD. Plasma analytes in patients were collected through immunoassay technology (Bioplex). We tested for the effect of age by group on CP volume. Then we focused on the interaction between illness duration and diagnosis in predicting CP volume. After testing the effect of specific analytes by diagnosis, we calculated moderated moderation models (SPSS, PROCESS) setting each analyte as independent variable, CP volume as predicted variable and illness duration and diagnosis as moderators. We get the effects’ significance with the likelihood ratio statistic, always controlling for age, sex, and intracranial volume.

Patients were comparable in illness duration and severity. CP volume is differentially distributed through groups (right: p=0.04; left: p<0.01), with higher volumes in the clinical groups. Age by group significantly predict right CP volume (p=0.01). Also, duration of illness differently predicts right CP volume in MDD and BD (p=0.03) (Figure1). Then, given the significant interaction effect of IL13 (p=0.02) and IL1ra (p=0.01) in predicting right CP, we run the moderated moderation model. Longer illness duration has an effect in strengthening the opposite predicting value of IL1ra (ΔR2=0.03, p<0.01) on right CP volume in MDD and BD (Figure2).

Image:

Image 2:

Our findings propose CP as a proxy of inflammation in depression, being significantly predicted by peripheral immune markers in MDD and BD. In particular, the signature of inflammation in depression, could represent the neurotoxic load of the disease over the illness, with a worse effect in BD, with possible disruption of brain barriers permeability and an opposite effect of tightening and central segregation in MDD. Further analyses are needed to better elucidate this neurobiological mechanisms across mood disorders.

None Declared

Every year at least one million people die by suicide, with major depressive disorder (MDD) being one of the major causes of suicide deaths. Current suicide risk assessments rely on subjective information, are time consuming, low predictive, and poorly reliable. Thus, finding objective biomarkers of suicidality is crucial to move clinical practice towards a precision psychiatry framework, enhancing suicide risk detection and prevention for MDD.

The present study aimed at applying machine learning (ML) algorithms on both grey matter and white-matter voxel-wise data to discriminate MDD suicide attempters (SA) from non-attempters (nSA).

91 currently depressed MDD patients (24 SA, 67 nSA) underwent a structural MRI session. T1-weighted images and diffusion tensor imaging scans were respectively pre-processed using Computational Atlas Toolbox 12 (CAT12) and spatial tract-based statistics (TBSS) on FSL, to obtain both voxel-based morphometry (VBM) and fractional anisotropy (FA) measures. Three classification models were built, entering whole-brain VBM and FA maps alone into a Support Vector Machine (SVM) and combining both modalities into a Multiple Kernel Learning (MKL) algorithm. All models were trained through a 5-fold nested cross-validation with subsampling to calculate reliable estimates of balanced accuracy, specificity, sensitivity, and area under the receiver operator curve (AUC).

Models’ performances are summarized in Table 1.Table 1.

Models’ performances.

Overall, although overcoming the random classification accuracy (i.e., 50%), performances of all models classifying SA and nSA MDD patients were moderate, possibly due to the imbalanced numerosity of classes, with SVM on FA reaching the highest accuracy. Thus, future studies may enlarge the sample and add different features (e.g., functional neuroimaging data) to develop an objective and reliable predictive model to assess and hence prevent suicide risk among MDD patients.

None Declared

About 60% of bipolar disorder (BD) cases are initially misdiagnosed as major depressive disorder (MDD), preventing BD patients from receiving appropriate treatment. An urgency exists to identify reliable biomarkers for improving differential diagnosis (DD). Machine learning methods may help translate current knowledge on biomarkers of mood disorders into clinical practice by providing individual-level classification. No study so far has combined biological data with clinical data to provide a multifactorial predictive model for DD.

Define a predictive algorithm for BD and MDD by integrating structural neuroimaging and inflammatory data with neuropsychological measures (NM). Two different algorithms were compared: multiple kernel learning (MKL) and elastic net regularized logistic regression (EN).

In a sample of 141 subjects (70 MDD; 71 BD), two different models were implemented for each algorithm: 1) structural neuroimaging measures only (i.e. voxel-based morphometry (VBM), white matter fractional anisotropy (FA), and mean diffusivity (MD)); 2) VBM, FA, and MD combined with NM. In a subsample of 71 subjects (36 BD; 38 MDD), two similar models were implemented: 1) VBM, FA, and, MD combined with only NM; 2) VBM, FA, and MD combined with NM and peripheral inflammatory markers. Finally, the best model was selected for comparison with healthy controls (HC).

Overall, the EN model based on all the modalities achieved the highest accuracy (AUC = 90.2%), outperforming MKL (AUC=85%). EN correctly classified BD and MDD with a diagnostic accuracy of 78.3%, sensitivity of 75%, and specificity of 81.6%. The most significant predictors of BD (variable inclusion probability (VIP) > 80%) were the parahippocampal cingulate, interleukin 9, chemokine CCL5, posterior thalamic radiation, and internal capsule, whereas MDD was best predicted by chemokine CCL23, the anterior cerebellum, and the sagittal stratum. In contrast, NM did not help to differentiate between MDD and BD. However, they help to distinguish patients from HC. Psychomotor coordination and speed of information processing discriminated between MDD and HC (VIP>90%), whereas fluency, working memory, and executive functions differentiated between BD and HC (VIP>80%).

In summary, BD was predicted by a strong proinflammatory profile, whereas MDD was identified by structural neuroimaging data. A multimodal approach offers additional instruments to improve personalized diagnosis in clinical practice and enhance the ability to make DD.

None Declared

Many studies have searched for an association between violence and psychiatric diagnoses, without providing a confirmative result.

We have sought to deepen this topic, analysing different aspects of aggressivity, focusing on a specific diagnosis and its particular phases of illness, and looking for a correlation between psychiatric co-diagnoses and outpatients’ visits adherence.

We studied 151 bipolar type I inpatients presenting complaint, past medical and family history; we collected information about lifetime hetero/self-aggressive behaviours, irritability, agitation, suicide attempts, alcohol, or substance abuse.

The overall aggressivity in our sample resulted in 11.92% of cases, while the number of aggressive episodes during euthymia decreased to 2.64%, close to the population without psychiatric disorders. Personality disorders and alcohol abuse appeared to be the main risk factors for irritability [Fig. 1]; substance abuse for both irritability and hetero-aggressive behaviour [Fig. 2]. We observed that subjects who displayed better compliance to follow-up visits exhibited a significant lower aggressive behaviour than less adherent subjects. Moreover, our data disconfirm the common conception that correlates the presence of psychotic features to violence.

Studying aggressive in a bipolar population, we observed that the rare episodes of aggressiveness were condensed in active phases of illness and mainly related to alcohol or substance abuse, while violent acts during long periods of wellbeing appear in line with those of the general population. We are confident our data might be helpful in deconstructing the stigma that a psychiatric diagnosis equals to violent behaviour.

No significant relationships.

Attention Deficit / Hyperactivity Disorder (ADHD) is a highly prevalent neurodevelopmental condition characterized by inattention, motor hyperactivity and impulsivity. ADHD cognitive and behavioral presentation is characterized by a high heterogeneity (APA, 2013). Indeed, a complex diagnostic process, that considers several validated tools, is, to date, necessary.

The main aim is to develop supervised machine learning (ML) algorithms that could be used to support the diagnostic process for ADHD, by identifying the most relevant features in discriminating between the presence or absence of the ADHD diagnosis in children.

We analyzed data from 342 children (Mean age: 8y 8m ± 1y; 61 F) referred for possible ADHD symptomatology. Assessments were performed by an expert clinician and through questionnaires: Social Responsiveness Scale (SRS), Child Behavior Checklist (CBCL), Conners Rating Scale for Parents (CPRS) and for Teachers (CTRS). Data were analyzed using a decision tree classifier and random forest algorithms.

The decision tree model performed an accuracy of 0.71. The random forest model that was identified as the best tested, performed an accuracy of 0.77 (Figure 1) and it identified as most informative parent- and teacher-rated DSM-oriented ADHD symptoms (Figure 2).

Figure 1: Random forest confusion matrix and statistics.

Figure 2: Ranking of variables importance.

A random forest classifier could represent an effective algorithm to support the identification of ADHD children and to simplify the diagnostic process as an initial step. The use of supervised machine learning algorithms could be useful in helping the diagnostic process, highlighting the importance of a personalized medicine approach.

No significant relationships.

Cariprazine (CAR) is a D2, D3, 5HT1A receptor partial agonist and a 5HT2A, 5HT2B antagonist, used to treat Schizophrenia and Bipolar disorder. Interindividual variability in therapeutic and side effects of antipsychotics is difficult to predict, due to non-genetic and genetic factors. Single nucleotide polymorphisms (SNPs) are the main source of genetic variability, the ones in dopamine and serotonin receptors to which CAR binds are indeed likely to determine response to treatment.

The aim of the study is to define a relationship between CAR clinical efficacy and SNPs in dopamine and serotonin receptors genes of patients affected by schizophrenia and bipolar disorder.

We recruited 16 patients starting a monotherapy with CAR, evaluated at baseline and after 2, 4 and 8 weeks through BPRS rating scale. We selected a panel of SNPs in DR2, DR3, 5HT1A and 5HT2A receptors, with a frequency higher that 10% in Caucasians and functionally characterized. Cut-off for response to treatment was a 50% reduction of BPRS score. Statistical analysis was performed with one-way ANOVA followed by the test for linear trend between columns.

All subjects achieved response after 8 weeks of treatment, but 6 patients after 4 weeks. Early responders have a genetic profile associated with increased dopamine and serotonin receptor expression and/or binding affinity for their specific ligands. The association don’t reach statistical significance, probably due to low number of patients.

Preliminary results suggest that an array of dopamine and serotonin receptors SNPs could predict time to respond to CAR in schizophrenia and bipolar disorder.

The study is founded by Recordati AG, that commercialize the drug under study (Cariprazine) in Switzerland. Funding covers the costs for genetic analysis and other procedures of the study, no financial compensation is planned for investigators/authors.

Prostate and breast cancer share many similarities: high lifetime prevalence, increasing frequency, role of environmental factors, long survival also in metastatic disease and possibility of screening. The aim of this work is to evaluate the characteristics related to the patients, disease and treatment which can affect HRQoL at the beginning and after radiotherapy.

since June 2009, we have recruited patients, providing informed consent, before radiotherapy (T0). We assess demographic characteristic (age, qualification, work, marital status…); neoplastic staging and grading; radiation dose and other antineoplastic treatment (hormonal/chemio-therapy or surgery); concomitant medical disease and pharmacological therapy. We evaluate HRQoL by EORTC-QLQ-C30 and EORTC-QLQ-PR25 (prostate-specific) or EORTC-QLQ-BR23 (breast-specific). The protocol also includes HADS, Paykel Life Events Scale and EPQ-R. The work is ongoing and implies a follow-up at 6 and 12 months (T1/T2).

The majority of men have a localized disease with Gleason score between 6 and 8 and the median pretreatment PSA is 10.52 ng/mL; 70% will undergo adjuvant-RT; median age is 69.30 years. Women have a median age of 58.46 years, all underwent surgery and all have a localized disease and positive receptorial status. Global QoL is lightly higher in the man sample; both groups report a major deficit at Emotional Function and high levels of Fatigue. The personological characteristic more represented is “Extravertion”.

The results show an association between worse QoL, “Nevroticism” and high Anxiety levels only in the men sample at T0. At the moment, there is no significant relation in the women sample.

To evaluate the subjective well-being of a group of patients who were hospitalized at the Institute of Psychiatry (Novara), compared to the severity of illness.

Patients are evaluated at admission and discharge through self-administration of the SWN (Subjective Well-being under Neuroleptics) scale, which contains five subscales (emotional regulation; self-control; mental functioning; social integration and physical functioning) assessing patients’ psychophysical and emotional well-being, calculating a value for each subscale and a total score. The clinician fills in the CGI (Clinical Global Impression) for each patient, which provides a global judgement in three areas: severity of illness, global improvement and therapeutic effectiveness.

From June 2009, 51 patients were evaluated at admission and discharge: 26 diagnosed with psychosis and 25 diagnosed with personality disorders. Preliminary data suggest a meaningful improvement of the physical functioning in the psychotic group, a tendency to improvement of the social integration area in the personality disorders group. Among the psychotic group, the schizophrenic patients (n°=14) have shown an improvement in the self-control subscale.

Literature suggests that a high SWN score is associated with a better compliance and an early improvement of subjective well-being is a major predictor of the chance of remission. This study will allow to compare the subjective well-being evaluated by SWN with the clinical judgment of the CGI and above all if this can represent a predictor index for the compliance and the chance of remission.

To assess the use of SWN in the acute phase of psychiatric disease as a predictor of clinical outcome.

This study started in June 2009 and at the moment we have recruited 150 patients. The patients were divided into 4 groups according to their psychiatric diagnosis (schizophrenic psychosis, mood disorders, personality disorders, acute stress reaction) and each diagnostic group into three subgroups according to length of stay (T1< 7 days, T2 = 7–14 days, T3> 14 days). The subjective well-being indicators (subscales SWN: emotional regulation; self-control; mental functioning; social integration and physical functioning) and the severity of illness (CGI-S) were evaluated at admission and discharge.

At discharge there is a statistically significant difference in the SWN subgroups among the four diagnostic groups except for social integration and total score with equal CGI-S scores. Schizophrenic patients and personality disorders show a subjective improvement at T2; mood disorders at T3; acute stress reactions T1 = T2. CGI shows a statistically improvement regardless of the length of stay.

Preliminary data suggest that SWN represents a predictor of clinical outcome and remission and together with the clinical evaluation it can help clinician to settle therapeutic programs.

The catechol-O-methyltransferase (COMT) enzyme inactivates catecholamines, and the COMT Val(108/158)Met polymorphism (rs4680) influences the enzyme activity. Recent clinical studies found a significant effect of rs4680 on antidepressant response to fluoxetine and paroxetine, but several other studies were negative. No study considered drug plasma levels as possible nuisance covariate.

We studied the effect of rs4680 on response to fluvoxamine antidepressant monotherapy.

Forty-one consecutively admitted inpatients affected by a major depressive episode in course of major depressive disorder were administered fluvoxamine for 6 weeks. Changes in severity of depression were assessed with weekly Hamilton Depression ratings and analyzed with repeated measures ANOVA in the context of General Linear Model, with rs4680 and fluvoxamine plasma levels as factors.

rs4680 significantly interacted with time in affecting antidepressant response to fluvoxamine, with outcome being inversely proportional to the enzyme activity: better effects in Met-carriers, worse effects in Val/Val homozygotes. The effect became significant at the fourth week of treatment, and influence final response rates. Fluvoxamine plasma levels had marginal effects on outcome.

This is the first study that reports a positive effect of rs4680 on response to fluvoxamine, and the third independent report of its influence on response to selective 5-HT reuptake inhibitors (SSRIs). Our findings support the hypothesis that factors affecting catecholaminergic neurotransmission might contribute to shape the individual response to antidepressants irrespective of their primary molecular target.

To examine the perceived needs by patients and radiotherapists using a modified by us version of the Camberwell Assessment of Need (CAN).

We eliminated 4/22 areas of the CAN scale -ideated for psychotic patients- in order to adapt it to oncological subjects (naming it CANo). Each of the scale areas values: the existence of a specific need; the help received from care-givers; the help coming from social services; the completeness of the help received. CANo was administrated to 30 solid cancer subjects consecutively admitted in 2007 to the Radiotherapy Department of Novara Hospital (Italy), and to their respective treating radiotherapists. Patients with cognitive impairment were excluded. Patients were also administrated the following protocol: HADS (Hospital anxiety and depression scale); Paykel's list of stressful events; MBTI (Mayer-Briggs Type Indicator); EORTC QLQ-C30.

Anxiety and depression occurred at any level in 15/30 of cases. There was a significant correlation (Spearman coefficient: SC) between the numbers of needs on CANo scale and anxiety (SC:0.4; p=0.002) or depression (SC:0.48; p=0.006) levels. Higher scores in all functional EORTC scales corresponded to lower needs detected by CANo. Patient needs were perceived less important by patients themselves than their physicians (mean satisfied need scores: 1.87 vs. 3; unsatisfied need scores: 0.63 vs. 1.03). The staff overestimated patient physical health needs (7/30 vs 3/30), psychological distress (20/30 vs 5/30), relationship difficulties (9/30 vs 2/30), received information correctness (7/30 vs 2/30).

The CANo scale may be useful to detect oncological patient needs and to improve the quality of care.

Trazodone is a 5-HT(2) antagonist and 5-HT reuptake inhibitor (SARI), and an antidepressant with therapeutic effects on its target symptoms depressed mood, anxiety and insomnia [1,2]. The aim of our study is to present a possible line of treatment of a depressive episode in bipolar disorder type II with three case report.

BT is a 35-year-old Caucasian lady affected by bipolar disorder type II with a depressive episode. CA is a 40-year-old Caucasian lady affected by bipolar disorder type II with recurrent depressive episode with atypical symptoms. FR is a 38-year-old Caucasian gentleman affected by bipolar disorder type II with recurrent depressive episode. All patients are treated with trazodone immediate release subsequently increased up to 50-75 mg/day. After 6 months of treatment the patients showed a good outcome.

These case reports underscore the possibility of tailoring therapeutic strategies for the treatment of depressive episode in bipolar disorder type II. Our interest in trazodone lies in the possibility of treatment of depressive episode with the added benefit of resolution of affective symptoms, with fewer adverse effects and a well done effect in this patients. Moreover our opinions is that a therapy like trazodone with this particular profile of action, could represent a new strategy of tratment this patients. Further research is warranted to confirm the efficacy of this treatment.

Bipolar disorder (BD) is a severe, disabling and life-threatening illness. Disturbances in emotion and affective processing are core features of the disorder with affective instability being paralleled by mood-congruent biases in information processing that influence evaluative processes and social judgment. Several lines of evidence, coming from neuropsychological and imaging studies, suggest that disrupted neural connectivity could play a role in the mechanistic explanation of these cognitive and emotional symptoms. The aim of the present study is to investigate the effective connectivity in a sample of bipolar patients.

Dynamic causal modeling (DCM) technique was used to study 52 inpatients affected by bipolar disorders consecutively admitted to San Raffaele hospital in Milano and forty healthy subjects. A face-matching task was used as activation paradigm.

Patients with BD showed a significantly reduced endogenous connectivity in the DLPFC to Amy connection. There was no significant group effect upon the endogenous connection from Amy to ACC, from ACC to Amy and from DLPFC to ACC.

Both DLPFC and ACC are part of a network implicated in emotion regulation and share strong reciprocal connections with the amygdale. The pattern of abnormal or reduced connectivity between DLPFC and amygdala may reflect abnormal modulation of mood and emotion typical of bipolar patients.