Newton001: Biomarkers of treatment naive psychosis.

Lead Research Organisation: King's College London
Department Name: Social Genetic and Dev Psychiatry Centre


The First Episode of Psychosis is perhaps the most important stage of schizophrenia, unless effective treatment is established during this phase many patients develop an unfavourable clinical outcome. If we can understand the biological markers associated with FEP and subsequent response to treatment, we maybe able to develop ways of predicting which treatment patients will respond best to. However, to understand this we need to examine samples from patients before they undergo drug treatment, as otherwise it is unclear what changes are due to the disorder/episode and what are due to the prescribed medication. We have collected blood samples (RNA, DNA and serum) from psychosis patients who are treatment naïve (before they received drug treatment for the disorder), and who have been followed for 8 weeks (N=80) and one year (N=30) after drug treatment. This group of patients is unique and being able to study biomarkers unaffected by the drugs given as treatments may allow us greater understanding of the causes and potential underlying risks for psychosis.

Technical Summary

First Episode Psychosis: recruitment, psychiatric evaluation and blood sampling (team responsibility)
Current status of recruitment (Brazil):
We have collected and isolated the mRNA and the DNA from 150 antipsychotic-naïve FEP patients (anFEP), 75 FEP patients treated with Risperidone for eight weeks (FEP-8w), and 20 FEP followed by one year of treatment (FEP-1y).

Psychiatric Evaluation (Brazil)
We selected individuals between the ages of 18-35 fulfilling criteria for psychotic diagnoses according to the DSM-IV. Assessments are made made at three points: 1) First Episode of Psychosis before the treatment (antipsychotic-naïve), 2) 8 weeks after risperidone treatment and 3) one year after treatment. For all three points, we evaluate symptom severity (Positive and Negative Symptoms Scale - PANSS), drug abuse (cannabis, alcohol and Tabaco), clinical sociodemographic information, Clinical Global Impression (CGI) and Global Assessment of Function (GAF).

Analysis (UK and Brazil)
Under the hypothesis that gene expression and methylation differences for FEP individuals and for treatment response are determined by genetic variance, polygenic scores and Matrix-eQTL ( will be used for the eQTL analyses. For our network analyses Whole Genome Coexpression Network Analyses ( Network/Rpackages/WGCNA/) will be used to identify modules of genes differing between cases and controls for both expression and methylation data, and causal anchoring analyses will be used to analyse the effect of mutations found. We will adjust for relevant covariate including treatment, smoker/non-smoker (Eysenck scores).
Integration Module
Advanced Bioinformatics Course (UK and Brazil)
In September 2015, we expect to organize a 1-week advanced course in Bioinformatics at UNIFESP.

Planned Impact

The First Episode of Psychosis is a key stage of schizophrenia, after this phase many patients has an unfavourable clinical development (25). One reason for this development is the delay in the establishment of effective therapeutic interventions. Thus far, we have collected (RNA, DNA and serum) from FEP patients who are all antipsychotic naïve, received then the same treatment protocol (all treated with risperidone) and have been followed for 8 weeks (N=80) and one year (N=30). To our knowledge, this sample is unique in the world and other international groups who investigated baseline and followed FEP patients did not collected RNA and DNA samples concomitantly.

In the last few years UNIFESP group emerged as a leading centre in the psychiatric genetics field in Brazil, working mainly with expression of target genes in blood of schizophrenia patients (Ota et al., 2014) or in the brain of animal models (Santoro et al., 2014). Although they have been reaching relevant results with those studies, the studies are now moving towards hypothesis-free approaches, verifying the whole genome and not just a few target genes. This hypothesis-free or neutral approach has been highlighted by journal reviewers as a means of increasing the competitiveness of our work and to enable publication in outstanding journals. Besides, RNA samples are delicate and are constantly subjected to degradation, thus, successive freeze-thaw events (e.g. for different projects focusing on target genes) are not recommended and enhances the chance of accidental RNase contamination. In this way and because of the rarity of the sample we have collected, we must now focus on genome wide approaches.

King's College London (KCL) MRC SGDP Centre is today one of the most respected centres in the psychiatric genetics field. Dr Gerome Breen and his group has an extensive experience in GWAS and pathway bioinformatics (Pedroso et al., 2012; Breen et al., 2011) and is involved in important ongoing consortiums (specifically the Psychiatric Genomics Consortium subgroups on Schizophrenia, Bipolar, Depression (a major contributor), Anorexia (co-chair), Pathway Analysis (chair) and Alzheimer groups). Dr. Breen is one of the principal investigators in an ongoing project entitled "Methylomic profiling in schizophrenia: Towards an integrated genetic-epigenetic approach.",. The analysis of this cohort of anFEP patients will complement this project by providing an unique resource to help understand the biological processes of schizophrenia before any treatment intervention.

In the last few years, "Big data" generated by Transcriptome sequencing and Methylome profiling demanded a revolution in Bioinformatics field and new approaches of analysis were created. Since bioinformatics is still a growing field in Brazil, with this grant we expect to organize an advanced course taught by Dr Breen and his group for young investigators of the State of São Paulo.

In this way, this proposal will benefits both sides. It will allow us to verify the transcriptome expression and the methylation profile in a rare group of patients followed by 8 weeks and one year. Above all, we expect to establish a stable partnership between UNIFESP and King's College London, and to organize an advanced course on bioinformatics that will train young researchers in genomic analysis.
Description We found that patients admitted to hospital with schizophrenia were more likely to respond better to treatments if they had a high genetic risk for the disorder. This was despite the fact the patients with a higher genetic risk or load present with more severe symptoms when they are admitted. We have now published our second and third papers. We plan to apply for further funding in the next year.
Exploitation Route We are currently trying to replicate the work in Brazil, where we can uniquely recruit treatment-naive patients, and trying to develop an algorithm to predict response to treatment in first episode psychosis for patients in the UK.

We are considering further funding applications to MRC.
Sectors Healthcare

Pharmaceuticals and Medical Biotechnology

Description We are currently replicating the findings and seeking to design a clinically useful prediction tool.
First Year Of Impact 2017
Sector Healthcare,Pharmaceuticals and Medical Biotechnology
Description Treatment Naive Schizophrenia 
Organisation Federal University of Sao Carlos
Country Brazil 
Sector Academic/University 
PI Contribution We have begun a longer term collaboration continuing the work started in our Newton grant. The main goal is to extend the collection of treatment naive schizophrenia psychosis patients to >500 patients. Our work is already showing that such phenotyping and sampling yields very different results for polygenic risk scores compared to patients who are undergoing treatment, enabling us to under the phenotypic domains through polygenic risk exerts its influence on risk and response to treatment. We are applying for further funding.
Collaborator Contribution Internal Brazilian funding and institutional support via a network of hospitals, enabling a stream of collection.
Impact We are submitting multiple papers this year.
Start Year 2015