Rare copy number variants in treatment-resistant psychosis


Vignette’s case

“Mr. Hunter” is a 34-year-old African-American man with a history of chronic schizophrenia and intellectual disability. He also has comorbid type 2 diabetes and hypothyroidism. The onset of his psychosis was in his early teens. He has been clinically stable on clozapine for about 14 years, and has not been admitted to a psychiatric hospital during that time.

Mr. Hunter was adopted, and nothing is known of his family’s psychobiological history. His adoptive mother died 7 years ago, and his adoptive aunt became his legal guardian. His short stature (height 154 cm) and a distinctly unsatisfactory face are noted for any common causes of intellectual disability. On a recent outpatient visit, his aunt inquired if his schizophrenia was caused by “genetics”.

1 in 3 patients with psychosis develops treatment-resistant psychotic symptoms (TRS).1 These patients have greater cognitive deficits, impaired functioning, and suicide rates.2 Clinical predictors of TRS are limited.3 Recent studies of TRS patients have found an increased burden of rare and deleterious copy number variants (CNVs).4,5 CNVs may inform about the biological mechanisms underlying treatment resistance.

The current study

Farrell and colleagues6 investigated rare CNVs in a sample of 509 patients with clinically confirmed TRS and compared the prevalence of schizophrenia-related CNVs in this sample with a cohort not selected for TRS.7 They recruited participants from 5 Pennsylvania hospitals and their affiliated long-term structured residences. They included participants under 18 years of age who were able and willing to give informed consent and who had been diagnosed with schizophrenia, schizoaffective disorder, a mood disorder with psychotic features, or a psychotic disorder NOS. Participants had continuous psychotherapy for 5 years and no clinical improvement despite ≥3 antipsychotic trials.

Exclusion criteria were psychosis related to substance dependence, medical conditions known to cause psychosis, and any sustained treatment response. Participants were clinically assessed using the Positive and Negative Syndrome Scale (PANSS). Of the initial 690 participants, 509 were included in the final sample.

DNA was extracted from the blood sample and genome-wide SNP genotypes obtained from the Illumina Infinium Global Screening Array. CNVs were identified from array density data using 3 calling algorithms. Quality control steps were used to remove low confidence CNV determinations. Exome sequencing data, using the Agilent SureSelectXT Clinical Research Exome, was available for 478 participants and used as another approach to identify CNV. CNVs are referenced with List of CNVs Associated with Neurodevelopmental Disorders.

The authors also compared the prevalence of (total) schizophrenia CNVs in their sample to a sample of 21,094 participants with schizophrenia who were not selected for treatment resistance, using a chi-square test. Loci-based comparisons of schizophrenia CNVs were performed between the two samples using Fisher’s exact test, correcting for multiple comparisons.

The average age of the participants was 52 years, 66% male and 75% white. The most common diagnoses were schizophrenia (47%) and schizoaffective disorder (46%). Among the participants, 51% were exposed to clozapine. Forty-seven of the 509 participants (9.2%) had at least 1 CNV potentially relevant to clinical presentation; 24 (4.7%) 1 of the CNVs had neurodevelopmental risk, most commonly a 16p11.2 repeat (n = 6), a 15q11.2–13.1 repeat (n = 4), and a 22q11.21 deletion (n = 4); And 21 of these 24 cases were also carriers of other schizophrenia CNVs.

Eleven patients had a large number of CNVs (>1 Mb) not overlapping with a neurodevelopmental CNV or schizophrenia, and 12 patients had variants of uncertain significance, most commonly a 15q11.2 duplication (n = 4) and a 15q13.3 duplication (n = 4). n = 3). Participants with CNVs had higher positive PANSS scores than non-carriers (21.0 vs 19.1). The prevalence of schizophrenia CNVs in the current study (4.1%) was approximately twice as high as that in the sample of participants not selected for treatment resistance (2.2%).

Study conclusions

The authors found a 9.2% prevalence of point variations of neuropsychiatric disease risk in TRS cases. There was also an increased prevalence of CNVs associated with schizophrenia in this sample. CNVs may contribute to general or specific risks and outcomes in TRS. The genomic region 15q11.2-13.1, in particular, requires further investigation. Whether more widespread genetic testing in schizophrenia is warranted is an unresolved question. Strengths of this study include the use of a well-defined TRS kit and a comprehensive approach to CNV detection.

bottom line

Rare CNVs may influence clinical phenotypes and may serve as biological entry points for the study of treatment-resistant schizophrenia.

Dr. Miller Professor in the Department of Psychiatry and Health Behavior at Augusta University in Augusta, Georgia. He is a member of the editorial board and serves as chair of the schizophrenia department psychic timesTM. The author reports that he is receiving research support from Augusta University, the National Institute of Mental Health, and the Stanley Institute of Medical Research.


1. Howes OD, McCutcheon R, Agid O, et al. Treatment-resistant schizophrenia: response to treatment and resistance to psychosis (TRRIP) Working Group Consensus Guidelines on Diagnosis and Terminology. I am a psychiatrist. 2017; 174 (3): 216-229.

2. de Bartolomeis A, Balletta R, Giordano S et al. Differential cognitive performance between responders to schizophrenia and nonresponders to antipsychotics: associations with disease course, psychopathology, and attitude to treatment and antipsychotic doses.. Res Psychiatry. 2013; 210 (2): 387-395.

3. Smart SE, Kępińska AP, Murray RM, MacCabe JH. Predictors of treatment-resistant schizophrenia: a systematic review of prospective observational studies. Psychol Med. 2021; 51 (1): 44-53.

4. Zoghbi AW, Dhindsa RS, Goldberg TE, et al. Rare high-impact genetic variants in severe schizophrenia. Proc Natl Acad Sci USA. 2021; 118 (51): e2112560118.

5. Ruderfer DM, Charney AW, Readhead B, et al. Polygenic overlap between risk of schizophrenia and response to antipsychotics: a genomic medicine approach. Lancet Psychiatry. 2016; 3 (4): 350-357.

6. Farrell M, Dietterich TE, Harner MK et al. Increased prevalence of rare copy number variants in treatment-resistant psychosis [published online ahead of print, 2022 Dec 1]. Schizophrenia urine. 2022; sbac175.

7. Marshall Cr, Hourigan DB, Merico D, et al. Contribution of copy number variants to schizophrenia from a genome-wide study of 41,321 subjects. [published correction appears in Nat Genet. 2017 Mar 30;49(4):651] [published correction appears in Nat Genet. 2017 Sep 27;49(10):1558]. Nat Genet. 2017; 49 (1): 27-35.

Leave a Comment