Neural Correlates of Positive and Negative Formal Thought Disorder in Individuals with Schizophrenia: An ENIGMA Schizophrenia Working Group Study

Formal thought disorder (FTD) is a key clinical factor in schizophrenia, but the neurobiological underpinnings remain unclear. In particular, relationship between FTD symptom dimensions and patterns of regional brain volume deficiencies in schizophrenia remain to be established in large cohorts. Even less is known about the cellular basis of FTD. Our study addresses these major obstacles based on a large multi-site cohort through the ENIGMA Schizophrenia Working Group (752 individuals with schizophrenia and 1256 controls), to unravel the neuroanatomy of positive, negative and total FTD in schizophrenia and their cellular bases. We used virtual histology tools to relate brain structural changes associated with FTD to cellular distributions in cortical regions. We identified distinct neural networks for positive and negative FTD. Both networks encompassed fronto-occipito-amygdalar brain regions, but negative FTD showed a relative sparing of orbitofrontal cortical thickness, while positive FTD also affected lateral temporal cortices. Virtual histology identified distinct transcriptomic fingerprints associated for both symptom dimensions. Negative FTD was linked to neuronal and astrocyte fingerprints, while positive FTD was also linked to microglial cell types. These findings relate different dimensions of FTD to distinct brain structural changes and their cellular underpinnings, improve our mechanistic understanding of these key psychotic symptoms.


Introduction
Formal thought disorder (FTD) is a syndrome characterized by disorganized and incoherent speech [1], [2].While FTD is a cross-diagnostic syndrome, it constitutes a key clinical factor of schizophrenia and a core component of the diagnostic criteria for this disorder in the DSM-5.Multiple lines of evidence point towards a key role for FTD in the pathophysiology of schizophrenia.It predicts transition into psychosis in clinical high-risk samples [3]- [5] and is closely correlated with long-term outcome in chronic disease states [6]- [11].Clinically, FTD is highly heterogeneous, with impairments ranging from impoverished thought to disorganized thinking to pressured speech [12].These different facets differentially impact clinical outcomes [7], [9], [13].To reduce this heterogeneity, researchers have suggested grouping FTD in positive and negative symptoms [14]- [16].While positive formal thought disorder symptoms are characterized by disorganized or unusual forms of thought or language, negative formal thought disorder includes paucity or slowing of thought or speech [12].
Brain volume reductions in fronto-temporo-basal ganglia-thalamic networks are a hallmark of schizophrenia pathology [17], [18] and might provide a major avenue towards unraveling the neurobiological basis of FTD.These brain structural abnormalities are progressive over the disease course [17] and are, at least retrospectively, linked to symptom patterns [19], [20].Although there is an established relationship between volume de cits and overall symptom patterns in schizophrenia, major questions remain, speci cally about the neuroanatomical basis of FTD in schizophrenia [18], [19], [21].Several pioneering publications reported brain structural correlates of FTD, but these studies enrolled comparatively small numbers of schizophrenia patients (20-30 patients) [22]- [24].Studies on much larger samples of thousands of patients, on the other hand, have pooled across patients with various diagnoses [10], [25], [26].This approach reduces the risk of spurious correlations and increases the power compared to smaller samples; it also has the advantage that it can identify transdiagnostic markers of FTD.However, it might be less sensitive towards pathologies associated speci cally with schizophrenia, as their progressive nature, their location and extent might set them apart from brain structural changes observed in other disorders [17], [27], [28].
Finally, a major limitation of conventional structural neuroimaging is that its ndings can re ect various histological correlates and are, necessarily, neurobiologically nonspeci c. Prior work in post-mortem samples has suggested that multiple neuronal and glial cell types are likely involved in the pathophysiology of schizophrenia, making histological speci city especially important [29]- [32].The lack of speci city in structural neuroimaging is a major obstacle with regard to the identi cation of cellular mechanisms and neuronal circuits associated with FTD, which is di cult to model in animals.However, while neuroimaging even at ultra-high magnet strengths cannot provide a su cient resolution to capture this level of detail, modern computational approaches based on gene expression data allow inference on the cellular composition of brain regions [33], [34].These approaches have been subsumed under the umbrella term "virtual histology".
This study assesses the neuroanatomical basis of FTD based on a large multi-site cohort and uses virtual histology tools on implicated brain regions to assess its cellular underpinnings.It presents the rst study using a large, multi-site ENIGMA Schizophrenia Working Group dataset to identify the structural correlates of FTD.As previous neuroimaging studies [35], [36] have suggested at least at distinct correlates of positive and negative FTD, we decided to separately explore different symptom dimensions, namely positive, negative and total formal thought disorder.We then use virtual histology tools to relate each of the structural ndings to cellular distributions in the cortex.To our knowledge, this is the largest study on the structural correlates of FTD in schizophrenia to date, and the rst to apply virtual histology to the structural correlates FTD.

Structural Associations with Formal Thought Disorder
We employed a general linear model to associate FTD scores with regional surface area or cortical thickness measurements of each Desikan-Killiany brain atlas region [37], controlling for age, age squared to capture potential non-linear relationships, gender and intracranial volume with FDR correction.Separate models were used for total (PANSS items: P2, N5, N6, N7), positive (P2) and negative (N5, N6, N7) FTD, and also for each of the three PANSS items contributing to negative FTD separately.These de nitions of formal thought disorder have been used previously for investigating the neural basis of formal thought disorder [35].

Total Formal Thought Disorder
Total FTD was associated with lower volumes of the bilateral pallidum and left amygdala, and thicker cortex in frontooccipital regions.In addition, association with thicker cortex in rostral middle frontal and postcentral pole regions, as well as right caudal anterior cingulate cortex and lateral occipital and pericalcarine regions were implicated (Fig. 1A, D, G; Table 1).Positive Formal Thought Disorder Positive FTD was associated with frontal, temporal and occipital cortical regions.These include associations with smaller surface area in the bilateral temporal, bilateral lingual, and right medial orbitofrontal cortex, and associations with thinner cortex in in the left lateral orbitofrontal and right medial orbitofrontal cortex, the rostral anterior cingulate, left caudal anterior and posterior cingulate, and temporal cortex (including the superior and middle temporal gyri and temporal poles), the parahippocampal, entorhinal and fusiform cortex, and the insula.There were also associations with thicker cortex in the right cuneus, pericalcarine and lateral occipital cortex, left lingual, and pericalcarine (Fig. 1B, E, H; Table 2).Negative Formal Thought Disorder Negative FTD was associated with smaller surface area in the lateral occipital cortex, lower pallidum and left amygdala volume, and thicker cortex in the right caudal anterior cingulate and rostral middle frontal gyri.Examining the three PANSS scores contributing to the negative FTD score separately, show that the signi cant ndings were predominantly driven by associations with item N5 (di culty with abstract thinking) ratings.Item N6 ratings were not signi cantly associated with cortical volume, and item N7 ratings were only associated with thicker medial orbitofrontal cortex (Fig. 1C, F, I; Table 3; Supplementary Table 3).While a fronto-occipital network was implicated in all three symptom dimensions-total, positive and negative FTD-frontal changes, in particular those in the medial orbitofrontal cortex, showed opposite patterns for cortical thickness: positive FTD was associated with thinner cortex, while total and negative FTD was associated with thicker cortex relative to other patients.Lateral temporal regions were only implicated in positive FTD.These ndings suggest a neurobiological divergence of positive and negative FTD and are in line with recent functional neuroimaging ndings [21], [35].

The in uence of site on the results
To assess possible site-speci c effects on the results, we conducted a leave-one-out analysis.Results were consistent for positive and negative FTD, while ndings for total FTD did not reproduce (Fig. 2, Tables 5 and 6).These results suggest robust ndings for both FTD subscales, while structural correlates of total FTD may at least partially vary by site.

Patient vs. Control Comparisons
In order to compare results related to FTD with those more generally associated with schizophrenia, we used a two-sample ttest comparing patients with schizophrenia with controls.We identi ed wide-spread smaller cortical surface area and volume, smaller subcortical volumes, and larger lateral ventricle volumes in patients compared to healthy controls, but also larger left caudate volumes in schizophrenia.The regions that showed positive associations with FTD were smaller in schizophrenia compared to the control group, indicating a process of relative sparing rather than enlargement (Supplementary Table 2).

Auditory Hallucination Associations
Prior research has led to the hypothesis that FTD primarily emerges from disturbances in language processing networks (dyssemantic hypothesis of FTD) [38], [39].To test whether the identi ed structural changes were speci c for FTD or rather indicative of unspeci c changes in language processing networks in schizophrenia [2], we explored potential correlations between structural variation in FTD-related brain regions and auditory hallucinations, another major language-associated pathology in schizophrenia [40].The regions which were signi cantly associated with FTD were also examined for a relationship with auditory hallucination symptoms using similar general linear models.There were no signi cant relationships with auditory hallucinations, suggesting the observed network is speci c to FTD, not general to auditory and language dysfunction in schizophrenia.

Cellular Genetic Fingerprint Associations
MRI imaging has provided valuable insight into in vivo pathologies in schizophrenia, but is unable to provide information about underlying histological changes.Novel virtual histology approaches based on gene expression databases such as the Allen Human Brain Atlas [41] utilize complex gene expression patterns to identify the cellular composition of a given brain region [33], [34].Capitalizing on this approach, we correlated the distribution of these gene expression ngerprints with the patterns of cortical thickness change identi ed in our data as described previously [33].All FTD dimensions were bilaterally associated with transcriptomic ngerprints for astrocytes ("astrocyte") and dendritic spine maintenance ("CA1 pyramidal").
The "CA1 pyramidal" label here refers to the original source for the genetic ngerprint, but the respective gene expression patterns have been con rmed to be distributed throughout the cortex [33].Positive FTD was also associated with the ngerprint for microglial function ("microglia"), but negative FTD was associated with microglial function in the right hemisphere only.Hence, the brain structural dissociation between positive and negative FTD appears to be somewhat accompanied by a dissociation also on a cellular level (Table 4).Positive FTD, which was associated with greater cortical atrophy was also more strongly associated with microglia which have been previously associated with excessive synaptic pruning in schizophrenia [42] .

Discussion
This study identi ed novel neural networks associated with different symptom dimensions of FTD in a large cohort of individuals with schizophrenia (Fig. 1).While prior studies of FTD, including meta-analyses have identi ed abnormalities in superior temporal gyrus activation [43] and connectivity associated with FTD-the latter speci cally for positive FTD [35].
This study suggests that these earlier ndings are only one component of a more extensive network.The temporal lobe component of the structural network identi ed in this study overlaps with previously identi ed functional ndings but connects to a wider range of regions, especially in the occipital lobe which is rarely studied in the context of schizophrenia.
Both positive and negative FTD were found to be related to fronto-occipital brain regions, namely the medial orbitofrontal cortex, anterior cingulate, lateral occipital cortex and negative FTD was also found to be related to the left amygdala.
However, anatomical measures related differentially between the two FTD dimensions.The associations of our FTD core network provide further insight into a long-standing controversy in the eld: whether FTD emerges from dysfunction in language processing networks ("dyssemantic hypothesis") [44] or rather from de cits in higher-order cognitive processes ("dysexecutive hypothesis") [38].Of note, these core regions were outside canonical language-related circuits [45], but rather associated with cognitive and behavioral control (medial orbitofrontal [46]-[48] and anterior cingulate [49], [50]), affective processing (amygdala [51], [52]) which have been associated with schizophrenia in previous studies [18], [22], [53], and abstract thinking and imagination (lateral occipital cortex [54]- [59]), which has been less commonly associated with schizophrenia [57], [58], [60].Our ndings suggest a potential role for dysfunctional executive processing as a common feature shared across FTD domains.However, it should be noted that positive FTD was also associated with classical language-related regions, suggesting a role for impaired semantic functions especially in the case of positive FTD (Fig. 1B,   1E; Tables 1-3).
A closer look at differences between ndings for distinct FTD symptom dimensions demonstrates a dissociation between the structural features of positive and negative FTD.In particular, positive and negative FTD showed opposing patterns of associations with cortical thickness in the orbitofrontal cortex and the rostral anterior cingulate (Fig. 1B-C, 1E-F).Negative FTD showed a positive correlation with cortical thickness in these frontal brain regions, however, it should be noted that these ndings were indicative of a relative sparing from a fronto-temporal pattern of atrophy [18], [22] in individuals with schizophrenia, rather than an absolute increase, when compared to healthy controls (Fig. 1).When schizophrenia subjects were compared to healthy controls, widespread smaller cortical surface area, thinner cortex, and smaller volumes were observed (Supplementary Table 2).Different de ciency patterns in two central hubs involved in cognitive control may indicate that differential biological mechanisms or different cellular populations play a role in the emergence of these types of FTD.Additionally, positive FTD was the only symptom dimension that implicated brain regions in the temporal cortex, particularly in language-related areas (Fig. 1B, 1E; Table 2).A previous meta-analysis from our lab highlighted functional changes of the superior and medial temporal gyrus in FTD [43]; central hubs of the human language processing network [61].
The temporal pole, in turn, has been linked to a semantic network involved in creative thinking [62].Importantly, connectomebased modeling with seeds in the superior and medial temporal cortex edpredicted positive FTD symptom severity, but not any other FTD dimension [21], which is well in line with our own ndings.Together, these ndings paint the picture of a role for language-related networks exclusively in positive FTD.The results of our study further support the idea of a fundamental neurobiological divergence between positive and negative symptom dimensions [1], [63] which has both been shown for general schizophrenia psychopathology and its neural correlates [21].
Due to its limited spatial resolution, MR imaging does not allow a direct link between macroscopic changes and underlying molecular or cellular pathologies, a key requirement for the development of new therapeutic approaches.Novel methods, however, allow at least indirect inference on these molecular processes.Our virtual histology approach [33], based on gene expression patterns provided by the Allen Human Brain Atlas [41], identi ed distinct transcriptomic ngerprints associated with each of the three symptom dimensions (Table 4).Common to both the positive and negative FTD dimensions was a transcriptomic signature associated with dendritic spine maintenance and astrocytes.Consistent with this nding, post mortem studies have reported lower dendritic spine density and impaired dendritic plasticity in the brains of individuals with schizophrenia [32], [64].Mechanistically, loss of dendritic spines has been linked to altered function in human complex 4 (C4) [65].Complex genetic variation in the C4 gene, in turn, has been linked to schizophrenia risk [66].Our own nding that neuroanatomical variation associated with both FTD dimensions is situated in brain regions with a high demand for dendritic spine maintenance appears plausible in light of these prior ndings.Besides their role in synapse formation during development [67], astrocytes are known to modulate glutamatergic signaling [29], [31].Pharmacological antagonization of glutamate signaling, in turn, has been shown to induce both positive and negative FTD in healthy subjects [68]- [70].Beyond these signatures, positive FTD was also found to be associated with brain regions enriched for another non-neuronal cellular ngerprint: microglia.As resident immune cells of the central nervous system, microglia cells are involved in synaptic pruning during development [42], [71].
Our study has several limitations.The large sample size of our study made possible by the ENIGMA consortium has enabled us to identify novel neural networks associated with FTD that earlier, smaller studies were unable to identify.However, pooling over multiple sites limited us to operationalization of FTD through the use of several PANSS items, as rating scales tailored more precisely for formal thought disorder were not available across all cohorts.Additionally, neuroanatomical abnormalities in schizophrenia have been shown to be progressive.Hence, a longitudinal approach might provide an even better link between brain structural alterations and FTD than a cross-sectional approach and will be an important aim for future studies.Finally, we used a recently established virtual histology approach to identify potential cellular contributions to FTD [33].While we regard this approach as a unique option to generate observer-independent data-driven hypotheses about the cellular underpinnings of brain structural changes associated with FTD, it does not provide direct evidence of cellular pathologies in the brains of individuals with schizophrenia.Postmortem histological studies with ante-mortem FTD ratings are warranted to verify or falsify the virtual histology ndings.
In sum, this study demonstrates a convergence between neuroimaging and cellular endophenotypes and is, to the best of our knowledge, the rst to associate glial function with formal thought disorder speci cally.The identi cation of a multi-scale associations between structural and transcriptomic networks associated with cellular function is of speci c interest clinically, because it provides the basis for linking neuroimaging ndings and clinically relevant molecular targets in a way that is not possible with either method in isolation.

ENIGMA Schizophrenia Working Group
This study included the data from 752 individuals with schizophrenia and 1256 healthy controls pooled by the ENIGMA Consortium Schizophrenia Working Group.The data comprised cortical thickness and cortical surface area, and subcortical volume for each region in the Desikan-Killiany atlas, subcortical volumes, demographic information, and PANSS symptom scores (see Supplementary Table 1 for sample details and image processing).

Formal Thought Disorder Scores
We calculated formal thought disorder and auditory hallucination symptom score for each patient from the PANSS.Four symptom scores were calculated for each patient: total formal thought disorder, positive formal thought disorder, negative formal thought disorder and hallucinations.
The total formal thought disorder scale was derived from the sum of items P2, N5, N6 and N7, as has been used in previous studies [35].Of this subset, P2 was used as a measure of positive formal thought disorder, while negative formal thought disorder was measured as the sum of N5, N6 and N7.

General Linear Models
Each of the major comparisons in this study used a general linear model examining associations between a variable of interest and the regional cortical thickness, surface area and subcortical volume values in a mass univariate manner which was then controlled for multiple comparisons to false discovery rate q = 0.001.The same model was applied to the relationship between Auditory Hallucinations for regions that showed signi cant correlations with the formal thought disorder scores.Similar statistical models have been used in multiple ENIGMA studies [72], [73].
Leave-one-out Validation for Site Variability Matlab scripts were to for ndings predominantly driven by results from single sites.This analysis repeated the statistical analysis 10 times, each time removing the data from one site.Using the output of these analyses, we summarized any regions that appeared as signi cant in one or more of the leave one out analyses, taking note of the directionality of the change (see Fig. 3).

Schizophrenia vs. Control Comparison
A two-sample t-test was performed comparing the schizophrenia and control groups, without controlling for the additional variables of no interest.This was included along with the GLM for diagnosis effects to model effects which include the global reduction in ICV which has been previously demonstrated in schizophrenia.
Cellular Genetic Fingerprint Associations R scripts released by the Paus Lab were used to associate our regional cortical thickness ndings with the regional distributions of different cellular ngerprints (see Shin et al 2018 for details on the method) [33]. Figures

Figure 1 Regions
Figure 1

Table 2 T
statistics of regions signi cantly associated with positive FTD (FDR corrected p = 0.001).

Table 3 T
statistics of regions signi cantly associated with negative FTD (FDR corrected p = 0.001).

Table 4 FDR
Corrected P-Values for associations between FTD related cortical thickness changes and distributions of cellular transcriptomic signatures.

Table 5 T
statistics of regions signi cantly associated with positive FTD after "leave one out" analysis

Table 6 T
statistics of regions signi cantly associated with negative FTD after "leave one out" analysis Deutscherelationship between formal thought disorder related ndings and auditory hallucinations were carried out to test the speci city of the results to formal thought disorder.