Thus far, a handful of highly penetrant mutations conferring risk of psychosis have been discovered. Here we used whole-genome sequencing and long-range phasing to investigate an Icelandic kindred containing ten individuals with psychosis (schizophrenia, schizoaffective disorder or psychotic bipolar disorder). We found that all affected individuals carry RBM12 (RNA-binding-motif protein 12) c.2377G>T (P = 2.2 × 10), a nonsense mutation that results in the production of a truncated protein lacking a predicted RNA-recognition motif. We replicated the association in a Finnish family in which a second RBM12 truncating mutation (c.2532delT) segregates with psychosis (P = 0.020). c.2377G>T is not fully penetrant for psychosis; however, we found that carriers unaffected by psychosis resemble patients with schizophrenia in their non-psychotic psychiatric disorder and neuropsychological test profile (P = 0.0043) as well as in their life outcomes (including an increased chance of receiving disability benefits, P = 0.011). As RBM12 has not previously been linked to psychosis, this work provides new insight into psychiatric disease.

BackgroundData on the haematological outcomes of patients who continue clozapine treatment following neutropenia are very rare as even mild neutropenia results in mandatory discontinuation of clozapine in most countries. However, in Iceland where clozapine monitoring is less stringent allows an observational study to be done on the risk of agranulocytosis and neutropenia during treatment with clozapine compared with other antipsychotics among patients with schizophrenia.MethodsThe present study is a part of a wider ongoing longitudinal study of schizophrenia in Iceland. We identified 201 patients with schizophrenia treated with clozapine and 410 patients with schizophrenia who had never been on clozapine by searching the electronic health records of Landspitali, the National University Hospital. Neutrophil counts were searched in electronic databases to identify patients who developed neutropenia/agranulocytosis and the frequency of neutrophil measurements was examined as well.ResultsThe median number of days between neutrophil measurements during the first 18 weeks of clozapine treatment was 25 days but after the first 18 weeks on the drug the median became 124 days. Thirty four cases of neutropenia were identified during clozapine treatment with an average follow up time of 9.2 years. The majority, 24 individuals developed mild neutropenia (1500–1900 neutrophils/mm3). None of these progressed to agranulocytosis. The remaining 10 patients developed neutropenia in the range 500–1400 /mm3 of whom one developed agranulocytosis, three stopped clozapine use and 6 patients continued on clozapine for at least a year without developing agranulocytosis. Unexpectedly, schizophrenia patients on other antipsychotics had an equal risk of developing neutropenia as those on clozapine.ConclusionsNeutropenia is common both in patients with schizophrenia on clozapine treatment and in those never on clozapine. Therefore a large part of neutropenia during clozapine treatment is probably not caused by clozapine. These findings have implications in assessing the balance between the risk of progression from neutropenia to agranulocytosis against the morbidity resulting from the premature discontinuation of clozapine under the current monitoring regulations in the US and in most of Europe.

Background: Clozapine is the only drug approved for treatment-resistant schizophrenia. There is evidence that clozapine is underutilized. Aims: To evaluate the initiation and discontinuation of clozapine at Landspitali University Hospital in Iceland and the prevalence of antipsychotic polypharmacy in clozapine-treated patients. Methods: The study is a part of an ongoing longitudinal study of schizophrenia in Iceland. We identified 201 patients on clozapine or who have been on clozapine by using a keyword search in the electronic health records and by reviewing their medical records. Results: Mean age at first treatment with clozapine was 37.8 years. Mean follow-up period on clozapine was 11 years. After 20 years of treatment 71.2% of patients were still on clozapine. After one year of treatment 84.4% of patients were still receiving clozapine treatment. We estimate that 11.4% of patients with schizophrenia in Iceland are taking clozapine and that 16% have been treated with clozapine at some point. Polypharmacy is common, since nearly 2/3, 65.6%, of patients taking clozapine use at least one other antipsychotic and 16.9% are also receiving depot injections. Conclusions: We need to increase the awareness of psychiatrists in Iceland with regard to treatment with clozapine, since only about half of the estimated population of patients with treatment-resistant schizophrenia in Iceland have ever been treated with clozapine. Nearly two thirds of patients who are prescribed clozapine in Iceland remain on it long-term.

The atypical antipsychotic clozapine is the only effective medication for treatment-resistant schizophrenia. However, it can also induce serious adverse drug reactions, including agranulocytosis and neutropenia. The mechanism by which it does so is largely unknown, but there is evidence for contributing genetic factors. Several studies identified HLA-DQB1 variants and especially a polymorphism located in HLA-DQB1 (6672G>C, rs113332494) as associated with clozapine-induced agranulocytosis and neutropenia. We analysed the risk allele distribution of SNP rs113332494 in a sample of 1396 controls and 178 neutropenia cases of which 60 developed agranulocytosis. Absolute neutrophil counts of 500/mm3 and 1500/mm3 were used for defining agranulocytosis and neutropenia cases, respectively. We also performed association analyses and analysed local ancestry patterns in individuals of European ancestry, seeking replication and extension of earlier findings. HLA-DQB1 (6672G>C, rs113332494) was associated with neutropenia (OR = 6.20, P = 2.20E−06) and agranulocytosis (OR = 10.49, P = 1.83E−06) in individuals of European ancestry. The association signal strengthened after including local ancestry estimates (neutropenia: OR = 10.38, P = 6.05E−08; agranulocytosis: OR = 16.31, P = 1.39E−06), with effect sizes being considerably larger for agranulocytosis. Using local ancestry estimates for prediction, the sensitivity of rs113332494 increased from 11.28 to 55.64% for neutropenia and from 16.67 to 53.70% for agranulocytosis. Our study further strengthens the evidence implicating HLA-DQB1 in agranulocytosis and neutropenia, suggesting components of the immune system as contributing to this serious adverse drug reaction. Using local ancestry estimates might help in identifying risk variants and improve prediction of haematological adverse effects.