THE GENOMICS OF HIGHLY TREATMENT RESISTANT SCHIZOPHRENIA

Background Understanding the molecular basis of Schizophrenia (SCZ) is one of the most significant problems in psychiatry. The minimal identification of single genes greatly limits the actionability of genomic findings - we need single gene targets for clinical utility, etiological investigation, and therapeutic development. Modern SCZ association studies analyze ~105 subjects to achieve statistical power. To achieve these sample sizes, a range of SCZ subjects – from mild to severe – are included. A time-honored and complementary strategy in genetics is to study an extreme phenotype– here, highly-treatment resistant SCZ (HTRS). It is known that rare Mendelian disorders (e.g., Wilson's Disease) can be misdiagnosed as SCZ because they initially present with psychosis in some individuals. In some of these misdiagnosed cases, antipsychotics are not effective. Therefore, genetically screening individuals with HTRS may uncover Mendelian disorders that present as psychosis and are misdiagnosed as SCZ. If ideal HTRS subjects can be identified, we can screen their genomes for rare Mendelian disorders that mimic SCZ. Methods The Pennsylvania state psychiatric hospitals contain an ideal HTRS sample, and we are recruiting participants for genomic assays (n=1000). Our formal inclusion criteria are: provision of written informed consent; age ≥18 years; DSM-IV diagnosis of SCZ, schizoaffective disorder, or psychosis NOS; ≥5 years of continuous inpatient hospitalization; ≥5 years of persistent psychotic symptoms with GAF scores ≤40 over the course of documented hospitalization; a history of poor treatment response to adequate trials of ≥3 different classes of antipsychotic drugs at recommended maximum dose in trials lasting ≥6 weeks. Medications must include a first-generation antipsychotic and two second-generation antipsychotics (e.g., HTRS with adequate trials of haloperidol, clozapine, and quetiapine). Our exclusion criteria are as follows: DSM-IV psychotic disorder consequent to licit or illicit drug dependence; sustained treatment response (GAF score >40 for any 3-month period); or sustained refusal to take prescribed medications (>5% of the 5-year screening window). For our pilot study, we selected 3 extreme cases for whole genome sequencing and array genotyping. Results In our small pilot study, we identified 22q11DS and pathologically confirmed Huntington Disease in an individual with HTRS, the first reported instance of this combination. This rare combination of genetic variants created a clinical portrait initially indistinguishable from SCZ – the individual was originally treatment responsive. Discussion We are collecting a sample of HTRS that will be screened for Mendelian disorders that mimic schizophrenia. Our preliminary findings suggest that our sample may contain a novel source of genotype-phenotype relationships that can be used to identify actionable variation for etiological research, clinical utility, and therapeutic development.