Research Perspectives for Neuroimaging of Schizophrenia Spectrum Disorders

Abstract

Current evidence from neuroimaging studies investigating schizophrenia spectrum disorders (SSDs) has suggested alterations in grey and white matter [1–3], ventricular volume [4, 5], structural and functional connectivity [6, 7] and neurotransmitter levels [8]. Some of these findings have been consistent, for example, in the case of reduced cortical grey matter [1] and increased lateral ventricle volume [4]; however, others have been less clear with findings of both increased and decreased connectivity across several brain regions [6, 7]. Also of interest are regions that have consistently been associated with structural and neurochemical abnormalities, such as the striatum [8, 9] and the growing area of the role of the immune system in the pathology of SSDs [10].

Appendix: Neuroimaging Studies Published in 2017 Investigated Schizophrenia Spectrum Disorders, First-Episode Psychosis or Clinical High-Risk Populations

Authors	Neuroimaging method	Participants	Diagnostic manual	Outcome
Berger [16]	MRI	89 SCZ, 162 FEP, 135 CHR, 87 HC	DSM-III	A linear trend for increasing ventricular volume across progression of illness, reaching significance only for SCZ patients
Bousman [18]	MRI + DTI	290 SCZ, 175 HC	DSM-IV	SCZ patients with risk single-nucleotide polymorphisms had larger lateral ventricle volumes (left and right) compared to schizophrenics without risk genes
Buchy [24]	MRI	128 FEP	DSM-IV	Clinical insight was not associated with cortical thickness at baseline, but worsening of clinical insight over time was linked with thinning in the dorsal postcentral and precentral gyri
Castro-de-Araujo [28]	MRI	100 FEP, 94 HC	DSM-IV	The presence of FEP altered the relationship between grey matter volume and cognition
Chung [15]	MRI	267 CHR, 132 HC	DSM-IV	Ventricular enlargement over time was linked to grey matter reduction in the prefrontal cortex, superior temporal gyrus and parietal cortices in people at CHR
Cropley [31]	MRI + DTI	326 SCZ and schizoaffective, 197 HC		Significant loss of grey matter progressively across illness, reduced FA after age of 35
Crossley [39]	DTI	76 FEP 74 HC		At baseline FEPs who subsequently responded to antipsychotic medication showed better efficiency in structural connectomes, and these group differences were not apparent after 12 weeks of treatment
Cuesta [14]	MRI	50 FEP, 24 HC, 21 unaffected relatives	DSM-IV	Patients had enlarged left lateral and right lateral ventricles compared with family and larger third ventricle than controls
Dempster [20]	MRI	16 FEP	DSM-IV	Reductions in grey matter in the nucleus accumbens, right globus pallidus, left inferior parietal lobe, Brodmann’s areas 40 and 7 and left superior parietal lobule were associated with poorer cognitive performance over time
Forns-Nadal [19]	MRI	31 FEP + 27 HC	DSM-IV	FEPs had increased nucleus accumbens volumes, and this was not correlated with negative symptoms of psychosis
Klauser [36]	DTI	326 SSD, 197 HC	DSM-IV	Patients showed reduced FA and increased MD in all lobes, and differences were pronounced in the thalamus, cingulum, corpus callosum and areas involved in the rich club organisation
Konishi [17]	MRI	19 CHR, 20 HC	DSM-IV	Enlarged temporal horn area of the lateral ventricle but reductions in amygdala and whole-brain volume leading to an abnormal ratio of temporal horn to total brain volume in CHR patients
Koutsouleris [32]	MRI	92 SCZ	ICD-10	Individual responses to transcranial magnetic stimulation were predicted with 85% accuracy using measures of grey matter density
Knöchel [22]	MRI	29 SCZ, 25 BPD, 93 HC	DSM-IV	Increased changes in apolipoprotein levels were associated with cognitive impairments and reduced volume in the right hippocampus of SCZ patients
Kuang [27]	MRI	15 FEP, 15 HC (historical)		Cortical thickness in ventrolateral PFC did not covary with other brain areas in FEPs
Landin-Romero [23]	MRI	45 schizoaffective, 45 HC	DSM-IV	People with schizoaffective disorder showed grey and white matter reductions in the frontal cortices, left insula, bilateral temporal lobes and posterior cingulate cortex and precuneus and fusiform cortex
Li [41]	MRI	34 SCZ, 34 HC		Better connectivity in the default mode network, the temporal lobe, the language network, the corticostriatal network, the frontoparietal network and the cerebellum was predictive of increased response to electroconvulsive therapy in patients
Mallas [34]	DTI	63 SCZ, 124 HC	DSM-IV	People with CACNA1C gene and SCZ had reduced FA compared to those without the gene, and FA was reduced in SCZ in general compared with controls
Mørch-Johnsen [26]	MRI	194 SSD	DSM-IV	Patients with auditory hallucinations had thinner cortex in left Heschl’s gyrus; no differences in planum temporale or superior temporal gyrus compared with those without hallucinations
Nieuwenhuis [30]	MRI	389 FEP	DSM-IV	Gender but not diagnosis or prognosis of psychotic disorders could be accurately predicted
Picchioni [53]	MRI	70 SCZ, 16 MZ discordant twins, 6 DZ discordant twins, 76 HC	DSM-IV	Whole-brain, grey matter and white matter volumes were reduced in SCZ, and there was a correlation between these volume reductions and schizophrenia liability in discordant co-twins
Rae [37]	DWI + MRI	35 FEP, 19 HC		FEP patients had reduced FA in multiple commissural, corticospinal and association tracts; this was associated with abnormalities in fibre number, density and myelination
Ren [38]	DTI	100 FEP, 140 HC	DSM-IV	Reduced FA in left anterior cingulate cortex, right anterior cingulate cortex, left inferior parietal cortex, left posterior cingulate cortex and right posterior cingulate cortex which were associated with eight gene variations and one cell cycle pathway variation
Rhindress [21]	MRI	29 FEP, 29 HC	DSM-IV	Following antipsychotic treatment, patients showed reductions in dentate gyrus/CA4 volume and increases in subiculum, and there were no significant changes in hippocampal volume in healthy controls
Schmidt [40]	DTI	24 CHR, 24 HC	ICD-10	Rich club organisation was impaired in people at risk of psychosis, and greater impairments were correlated with increased severity of negative symptoms
Serpa [35]	MRI + DTI	25 FEP, 1 HC	DSM-IV	Reduced FA in white matter tracts in the fronto-limbic and the associative, projective and commissural fasciculi in FEP; FA increased upon symptom remission following antipsychotic medication
Squarcina [29]	MRI	127 FEP, 127 HC	ICD-10	Fronto-temporal cortical thickness can be used as a potential marker to classify patients with FEP
Zhou [33]	DTI	48 FEP, 37 GHR, 67 HC	DSM-IV	Decreased FA in corpus callosum, anterior cingulum and uncinate fasciculus for both SCZ and GHR groups, decreased FA in fornix and superior longitudinal fasciculus in SCZ
Anderson [45]	fMRI	18 SCZ, 2 HC	DSM-IV	Patients showed a reduction in the population receptive field of neurons and a reduction in the inhibitory surround in V1, V2 and V4
Braeutigam [48]	MEG	15 SCZ, 16 BPD, 14 HC (all adolescents)	DSM-IV	Adolescents with schizophrenia (FEP) displayed reduced amplitude of MEG waves following mismatch negativity task, and connections appeared to be dominated by the right hemisphere
Falkenberg [53]	fMRI	34 CHR, 20 HC		Altered frontal and cuneus/posterior cingulate activation in UHR; amongst those with poor outcome, there was altered activation of frontal temporal and striatal regions
Gong [46]	fMRI	50 FEP, 122 HC	DSM-IV	FEPs showed decreased intranetwork connectivity and increased internetwork connectivity in drug-naive FEP compared with HC, and aberrant internetwork connectivity was particularly associated with psychotic symptoms and not MDD or PTSD
Hager [55]	fMRI	107 SCZ, 156 HC, 125 BPD, 98 schizoaffective, 230 healthy relatives	DSM-IV	People with SCZ showed decreased neural complexity towards a regular signal in hypothalamus, and SCZ and schizoaffective patients showed increased complexity in PFC
Koike [50]	fNIRS	47 CHR, 30 FEP, 34 SCZ, 33 HC	DSM-IV	The sum of signal changes during the task and the timing of the blood response relative to the task had an 80–90% accuracy in classifying people as non-patient, UHR, FEP or SCZ
Martinelli [54]	fMRI	21 SCZ, 26 HCs	DSM-IV	Patients showed greater brain activation when force was self-generated as opposed to externally produced; this was the opposite of HCs
Mason [49]	fMRI	22 SCZ intervention, 16 SCZ controls		Following CBT for psychosis, long-term psychotic symptoms were predicted by alterations in connectivity in the prefrontal cortex; alterations in connectivity between the amygdala and parietal lobe were predictive of long-term affective symptoms
Rikandi [51]	fMRI	46 FEP 32 HC	DSM-IV	Activity in the precuneus during a fantasy film could be used to classify patients as FEP or healthy controls with 79.5% accuracy
Peters [42]	fMRI	21 SCZ, 42 HC	DSM-IV	Decreased in functional connectivity between the putamen and right interior insula, dorsomedial and dorsolateral PFC and ventral striatum and left anterior insula in people with SCZ during a psychotic episode compared with healthy controls
Schmack [47]	fMRI	21 SCZ, 28 HC	ICD-10	Belief-related connectivity between the orbitofrontal cortex and visual cortex was higher in patients compared with HC
Spilka [52]	fMRI	28 SCZ, 27 GHR, 27 HC	DSM-IV	Patients with SCZ showed impairments in both age and emotion discrimination during a task and showed reduced activation of the medial prefrontal cortex
Xu [44]	fMRI	98 SCZ, 102 HC	DSM-IV	Patients showed decreased functional connectivity between the orbitofrontal cortex subregions
Wu [43]	fMRI	45 SCZ, 45 HC	DSM-IV	Patients with impaired working memory capacity and decreased brain activation/deactivation showed decreased functional activation of the dorsolateral prefrontal cortex with the angular cortex compared with controls
Di Biase [59]	PET	10 CHR, 18 FEP, 15 SCZ, 27 HC	DSM-IV	No evidence of altered microglial activation in UHR, FE or SCZ
Artiges [62]	PET + MRI	21 SCZ, 30 HC	DSM-IV	Higher dopamine transporter availability in the midbrain, in striatal and limbic regions and in amygdala/hippocampus positively correlated with positive symptoms
Hafizi [61]	PET + MRI	24 CHR, 23 HC	DSM-IV	No significant activation of microglia in dorsolateral prefrontal cortex or hippocampus in CHR compared with HC
Kim [64]	EEG + MRI	29 SCZ, 40 CHR, 47 HC	DSM-IV	Significant deficits in mismatch negativity and current source density strength found in temporal and frontal cortices in people with SCZ and CHR
Kindler [58]	MRI + fMRI	32 SCZ, 31 HC, 29 CHR, 12 FEP, 31 HC	ICD-10	Increased regional blood flow in the striatum and decreased regional blood flow in the prefrontal cortex in CHR, FEP and SCZ compared with controls
Ranlund [65]	MRI + EEG	703 SCZ, 68 schizophreniform, 60 schizoaffective, 2794 HC	DSM-IV	Polygenic risk scores predicted poorer performance on a cognitive block task for people with SCZ, relatives and controls; SCZ patients had higher polygenic risk scores
Selvaraj [60]	PET + MRI	14 SCZ, 14 CHR, 22 HC	DSM-IV	Microglia activation was associated with cortical grey matter loss in SCZ, and there was a trend for this in UHR
Solé-Padullés [63]	fMRI + DTI	44 CHR, 34 FEP, 35 HC (all adolescents)	DSM-IV	Reduced intrinsic functional connectivity in the right middle/inferior gyrus in patients compared with controls; values for CHR were intermediate between FEP and controls