To image or not to image? Researchers performed a systematic review and meta-analysis of neuroradiological abnormalities in first-episode psychosis.
“Mr House” is 22-year-old male with no psychiatric history who presents to the emergency department with his family. He is a senior in college. Per his sister, the patient has exhibited delusional thinking that he is a superhero and that his roommates at college are talking about him. He has also experienced a paranoid ideation that his room at college was being wiretapped. He tells the psychiatrist that he will be meeting the president the next day. The patient also has periods in which speech is mostly incoherent.
His laboratory studies, including a urine drug screen, are unremarkable, and he does not have any focal neurologic findings on physical exam. He is diagnosed with a first episode of psychosis (FEP). Mr House’s family asks if neuroimaging is indicated. As his psychiatrist, how would you respond?
There is no present consensus regarding the role of magnetic resonance imaging (MRI) scanning in the clinical assessment of patients presenting with FEP. Some guidelines recommend scanning all patients,1 whereas others recommend restricting to cases in which a secondary cause of psychosis is suspected.2
Although most radiological abnormalities in patients with FEP are incidental, a minority of abnormalities require a change in clinical care. Estimates of clinically relevant radiological abnormalities in FEP range from 0%3 to 10%.4 It is important to note that in otherwise healthy individuals, white matter hyperintensities are associated with incident cognitive decline, cerebrovascular risk, and mortality.5,6
The Current Study
Blackman and colleagues7 performed a systematic review and meta-analysis of neuroimaging abnormalities in FEP. The investigators searched Ovid, MEDLINE, PubMed, Embase, PsycINFO, and Global Health through July 21. Inclusion criteria were patients with FEP and the frequency of intracranial abnormalities. Included studies were assessed for risk of bias. Exclusion criteria were absence of data on prevalence estimates, studies restricted to pre-specified neuroanatomical regions, and patients with clinical signs and/or symptoms suggestive of a neurological disorder.
Radiological abnormalities were defined as any intracranial findings and were categorized by clinical relevance (leads to a change in management or diagnosis). Abnormalities were sub-grouped into white matter, vascular, ventricular, cyst, pituitary, tumor, cerebral atrophy, and other.
For each study, the (transformed) proportion of patients with FEP and any of subtypes of radiological abnormality was calculated. Random-effects inverse variance method meta-analysis was used. The number of patients needed to be scanned to detect 1 abnormality (NNA) was calculated. For studies with a healthy control group, relative risks of neuroradiological abnormalities were also calculated.
Possible publication bias was assessed with funnel plots and Egger’s test. Sensitivity analyses were performed to investigate effects of studies 1) with a mean age >35 years, 2) assessment performed by a non-radiologist, and 3) research samples.
The investigators reviewed 240 publications in full. Approximately 12 studies (comprising 13 samples), including 1618 patients with FEP, met the inclusion criteria.; 9 studies, comprising 1318 patients, reported clinically relevant abnormalities; and 8 studies included a healthy control group (total controls =1866).
For individual studies, the mean age ranged from 20 to 60 years, and the proportion of females ranged from 27% to 70%. There were 5 studies that reported data from routine clinical practice, and 6 studies that reported data from clinical research studies. In 9 samples, MRI scans were interpreted by a neuroradiologist. Overall, 10 samples were at medium risk and 3 samples were at low risk of bias.
The pooled prevalence of any abnormality was 26% (95% confidence interval [CI] 16% to 38%), corresponding to NNA=4, and between-study heterogeneity was high. The pooled prevalence of clinically relevant abnormalities was 5.9% (95% CI 3.2% to 9.0%), corresponding to NNA=18, and between-study heterogeneity was moderate. Among any abnormalities, white matter (7.9%) and ventricular (typically enlargement, 5.0%) were the most prevalent. The most prevalent clinically relevant abnormalities were white matter (0.9%) and cysts (0.5%).
Compared with healthy controls, patients with FEP were almost 3 times more likely to have any radiological abnormality (RR=2.8, 95% CI 1.3 to 5.9) and 1.5 times more likely to have a clinically relevant abnormality (RR=1.5, 95% CI 0.8 to 2.8). Publication year, age, study sample type, and scanner field strength did not moderate the prevalence of clinically relevant abnormalities. The pattern of findings was unchanged in subgroup analyses. There was no evidence of publication bias.
The investigators found the prevalence of any radiologically MRI abnormality in FEP was 26%, and 6% for a clinically relevant abnormality. Patients with FEP had a significant, 2-fold higher prevalence of radiological abnormalities compared to controls. The most common finding was white matter abnormalities, predominantly small hyperintensities. The investigators determined that 18 patients needed to be scanned to detect 1 clinically relevant abnormality.
Study strengths include exploration of potential moderating factors through subgroup meta-analysis and meta-regression and the comparison of the prevalence of abnormalities between FEP and controls. Study limitations are the inclusion of studies with MRI as part of research rather than routine clinical care and the exclusion of patients with clinical evidence of a potential secondary cause of psychosis. Longitudinal follow-up data were not available.
The Bottom Line
About 6% of patients presenting with psychosis have a clinically relevant MRI abnormality. Increasing availability and decreasing cost of MRI supports offer MRI to patients with FEP.
Dr Miller is a professor in the Department of Psychiatry and Health Behavior at Augusta University in Georgia. He is on the Editorial Board and serves as the schizophrenia section chief for Psychiatric Times®. The author reports that he receives research support from Augusta University, the National Institute of Mental Health, and the Stanley Medical Research Institute.
1. Galletly C, Castle D, Dark F, et al. Royal Australian and New Zealand College of Psychiatrists clinical practice guidelines for the management of schizophrenia and related disorders. Aust N Z J Psychiatry. 2016;50(5):410-472.
2. Structural neuroimaging in first-episode psychosis: technology appraisal guidance [TA136]. National Institute for Health and Care Excellence. February 27, 2008. Accessed October 1, 2023. https://www.nice.org.uk/guidance/ta136
3. Robert Williams S, Yukio Koyanagi C, Shigemi Hishinuma E. On the usefulness of structural brain imaging for young first episode inpatients with psychosis. Psychiatry Res. 2014;224(2):104-106.
4. Sommer IE, de Kort GA, Meijering AL, et al. How frequent are radiological abnormalities in patients with psychosis? a review of 1379 MRI scans. Schizophr Bull. 2013;39(4):815-819.
5. Debette S, Markus HS. The clinical importance of white matter hyperintensities on brain magnetic resonance imaging: systematic review and meta-analysis. BMJ. 2010;341:c3666.
6. Kloppenborg RP, Nederkoorn PJ, Geerlings MI, van den Berg E. Presence and progression of white matter hyperintensities and cognition: a meta-analysis. Neurology. 2014;82(23):2127-2138.
7. Blackman G, Neri G, Al-Doori O, et al. Prevalence of neuroradiological abnormalities in first-episode psychosis: a systematic review and meta-analysis. JAMA Psychiatry. 2023;80(10):1047-1054.