Antipsychotic agents deteriorate brain and retinal function in schizophrenia patients with combined auditory and visual hallucinations: A pilot study and secondary follow‐up study

Abstract Introduction Schizophrenia patients often experience auditory hallucinations (AHs) and visual hallucinations (VHs). However, the degree and type of brain and retinal alterations associated with combined AHs and VHs in schizophrenia patients remain unknown. There is an urgent need for a study that investigates the trajectory of brain and retinal alterations in patients with first‐episode untreated schizophrenia accompanied by combined AHs and VHs (FUSCHAV). Methods FUSCHAV patients (n = 120), divided into four groups according to AH and VH symptom severity (severe AHs combined with severe VHs [FUSCHSASV, 20 patients]; middle‐to‐moderate AHs combined with severe VHs [FUSCHMASV, 23 patients]; severe AHs combined with middle‐to‐moderate VHs [FUSCHSAMV, 28 patients]; and middle‐to‐moderate AHs combined with middle‐to‐moderate VHs [FUSCHMAMV, 26 patients]), were compared to healthy controls (n = 30). Gray matter volume (GMV) was adopted for brain structural alteration assessment. Total retinal thickness was adopted as a measure of retinal thickness impairment. Results In the pilot study, the rate of GMV reduction showed an inverted U‐shaped pattern across the different FUSCHAV patient groups according to AH and VH severity. The degree of retinal impairment remained stable across the groups. More notably, in the secondary follow‐up study, we observed that, after 6 months of treatment with antipsychotic agents, all the GMV reduction‐related differences across the different patient groups disappeared, and both GMV and retinal thickness demonstrated a tendency to deteriorate. Conclusions These findings indicate the need for heightened alertness on brain and retinal impairments in patients with FUSCHAV. Further deteriorations induced by antipsychotic agent treatment should be monitored in clinical practice.

Visual disturbances in patients with schizophrenia usually present as visual hallucinations (VHs) or vision distortions (Green, Hellemann, Horan, Lee, & Wynn, 2012). The prevalence of VHs in schizophrenia patients is 25%-30% (Waters et al., 2014). Many previous studies have reported that VHs can be observed in any stage of schizophrenia (including the prodromal stage, acute stage, and chronic stage).
In addition, VHs have also been observed in children, adolescents, and young adults at a high risk for schizophrenia (Garcia-Portilla et al., 2019;Grano et al., 2015;Guidotti & Grayson, 2011;Hebert et al., 2010;Mittal, Gupta, Keane, & Silverstein, 2015). Some studies that adopted neuroimaging and neuro-electrophysiology techniques reported that VHs are usually accompanied by brain cortex-related functional and structural alterations (Henke, Robinson, Drysdale, & Loxley, 2014;Pajani, Kok, Kouider, & Lange, 2015;Stephan-Otto et al., 2017). Most of those studies reported that functional disturbances in the visual cortex and parietal cortex, and structural alterations in the brain gray matter volume (GMV) were characteristic of VHs (De Haan, Nys, Zandvoort, & Ramsey, 2007;Garcia-Portilla et al., 2019;Grano et al., 2015;Guidotti & Grayson, 2011;Hebert et al., 2010;Henke et al., 2014;Mittal et al., 2015;Oertel et al., 2007;Pajani et al., 2015;Stephan-Otto et al., 2017). Several previous studies have reported that VHs are also accompanied by visual perception organization-related impairments, especially retinal thickness impairments (Lee et al., 2014;Schönfeldt-Lecuona et al., 2019). In recent years, an increasing number of studies have been using optical coherence tomography (OCT) for the investigation of the retinal thickness-related alterations in schizophrenia patients (Calderone et al., 2013;Lencer, Nagel, Sprenger, Heide, & Binkofski, 2005;Nagel et al., 2007;Onitsuka et al., 2007;Silverstein et al., 2009Silverstein et al., , 2010; however, few studies have focused on the retinal thickness-related impairments in schizophrenia patients with VHs. There is an urgent need for studies to be conducted to bridge this gap and explore the mechanisms of VHs in schizophrenia patients. AHs and VHs can co-occur in patients with schizophrenia (Andreasen & Flaum, 1991;van Os & Kapur, 2009), and such a co-occurrence is usually associated with an increased difficulty in treatment administration and prognostic deterioration (Elkis & Buckley, 2016;Potkin et al., 2020). However, to the best of our knowledge, few studies have reported on the brain-related features of schizophrenia patients with VH and AH co-occurrence. A previous study demonstrated that the brain functional connectivity-related disturbances in the mesolimbic pathway, especially in the nucleus accumbens, were characteristic of AH and VH co-occurrence in schizophrenia patients (Rolland et al., 2015). The findings of that study provide important clues for the re-consideration of the brain-related features in such settings. Auditory hallucinations are usually accompanied by functional disturbances in the components of the speech network or salience network (Jardri, Pouchet, Pins, & Thomas, 2011). Visual hallucinations are accompanied by functional disturbances mainly in the visual cortex (Behrendt, 2010) and hippocampus (Amad et al., 2014). Schizophrenia patients with the co-occurrence of VHs and AHs demonstrate functional connectivity-related disturbances in the mesolimbic pathway (Rolland et al., 2015). These findings indicate that, in schizophrenia patients, AH and VH co-occurrence may be accompanied by more complex in-brain functional and structural alterations. However, although the aforementioned studies provide information on the brain-related pathological features of schizophrenia patients with AH and VH co-occurrence, several questions remain unanswered. For example, (a) Are the different categories of AH and VH co-occurrence severity in schizophrenia patients associated with different brain-related structural and functional alterations? (b) Do the retinal thickness-related impairments differ across the different VH and AH combination categories in such patients? (c) Will the VHs and AHs deteriorate reciprocally? and (d) Will the antipsychotic agents used for treatment influence the brain and retinal alterations in such settings?
Based on the findings of previous studies (Amad et al., 2014;Rolland et al., 2015;Silverstein, 2016), especially those obtained by Silverstein (Silverstein, 2016), we conducted a study to investigate the pathological features associated with first-episode unmedicated schizophrenia with combined auditory and visual hallucinations (FUSCHAV). Unlike prior studies that categorized patients by their experience of either AHs or VHs, in the current study, we investigated all the brain and retinal aberrations presenting in those with a combination of AHs and VHs. We hypothesized that: (a) FUSCHAV would be associated with brain structural/functional and retinal alterations, (b) The brain and retinal alteration patterns may differ across different AH-VH combination severity categories, and (c) AHs and VHs may show reciprocal deterioration in these patients. Questionnaire (Dragovic & Hammond, 2007); (e) ophthalmic disease;

| ME THODS
(f) high myopia; (g) any MRI contraindication, including claustrophobia; and (h) intelligence quotient <80. The healthy controls did not have a positive family history, as assessed by the Family Interview for Genetic Studies (1992).

| MRI data acquisition
The 3.0-Tesla MR system (Discovery MR750, General Electric) was used for the collection of MRI data. Functional MRI (fMRI) was performed using a GE Healthcare Discovery MR750 3 T MRI system (General Electric) with an eight-channel phased-array head coil.
Participants lay in a supine position and were asked to restrict their thoughts and head movements during imaging. The imaging parameters were as follows: 2,000 ms repetition time (TR), 45 ms echo time (TE), 32 slices, 4 mm slice thickness, 0.5 mm gap, field of view, 64 × 64 acquisition matrix, and 90° flip angle. SENSitivity Encoding (SENSE) with a SENSE factor of 2 and parallel imaging were used for all scans. Images were obtained with a high-resolution, three-dimensional turbo-fast echo T1-weighted sequence with the following parameters: 8.2/3.2-ms TR/TE, 188 slices, 1 mm thickness, no gap, 256 × 256 FOV, 256 × 256 acquisition matrix, and 12° FA.

| OCT data acquisition
OCT data were acquired with the OCT 4000 system (Zeiss [premierop.com/zeiss-cirrus-hd-oct-4000-used]) that collects retinal scans of both eyes. The hand-held probe was mounted, and participants were positioned on a chin-head rest and asked to focus on a fixed target. A 5-s volumetric 10 mm × 5 mm scan of the foveal center, marked by outer segment layer thickening, was captured. Each scan comprised 500 A-scan/B-scans, 50 B-scans, and five frames/B-scan, with acceptable scans containing ≥ five consecutive B-scan frames of the foveal center with no movement artefacts.

| Psychotic, auditory, and visual hallucination symptom severity assessment
The total severity of schizophrenia was assessed by the Positive and Negative Symptoms Scale (PANSS) (Fleischhacker et al., 2019). The Scale for the Assessment of Positive Symptoms (SAPS) (Kumari et al., 2017) was used to measure the severity of AHs and VHs. Cognitive ability was assessed by the MATRICS Consensus Cognitive Battery (MCCB) (Lees et al., 2015). The Global Assessment of Functioning (GAF) was used to assess global function (Gspandl, Peirson, Nahhas, Skale, & Lehrer, 2018).
Patients were categorized into four groups according to AH and VH severity, as measured by the SAPS items of VH and AH.

| Structural MRI data processing
The analysis of brain volume differences was performed by voxelbased morphometry, using SPM8 (Statistical Parametric Mapping; v5; Institute of Neurology, London, UK). Bias correction, spatial normalization, segmentation into gray and white matter, imaging of cerebrospinal fluid, and intensity modulation were performed on 3D-FSPGR images in native space using SPM8. The Diffeomorphic Anatomical Registration Through Exponential Lie Algebra toolbox, as proposed by Ashburner, was used following a high-dimensional normalization protocol. Intensity modulation was performed by multiplying the voxel values of the segmented images by the measure of warped and unwarped structures derived from the nonlinear step of the spatial normalization. During this step, relative regional gray matter density was converted into absolute gray matter density and expressed as the amount of gray matter per unit volume of brain tissue before spatial normalization. The resulting modulated gray and white matter images were smoothed with a 6-mm Gaussian kernel. Multiple pattern recognition analysis was used to regress out the covariates' influence on gray matter. The significance threshold was set at an FWE-corrected p value <.05.
Covariates included age, sex, education level, symptoms severity, GAF score, and MCCB score.
The first 10 scan volumes were discarded to allow stabilization of the scanner and to allow patients to get acclimatized to the testing situation. The remaining volumes were corrected for slice timing and motion artifacts. Allowable motion thresholds (translational and rotational motion <2 mm and 2°, respectively) were checked for all fMRI data. Six of the motion parameters and the average blood oxygen level-dependent signals of the ventricles and white matter were removed. Data with specificvolume framewise displacement values >0.5 were excluded from the analysis. Bandpass frequencies ranging from 0.01 to 0.08 Hz were used for data filtration. Individual structural images were co-registered to the mean functional image, and the transformed structural images were co-registered to the Montreal Neurological Institute (MNI) space using linear registration. The motion-corrected functional volumes were spatially normalized to the MNI using parameters estimated during linear co-registration. Finally, the functional images were re-sampled into 3-mm cubic voxels for further analysis.

| Global Functional Connectivity Density (gFCD) calculation
The gFCD was calculated for each voxel using a customized Linux script (Zhuo et al., 2014). Pearson's linear correlation was used to explore the functional connectivity between voxels, with a correlation coefficient threshold of r > .6. Only voxels within the cerebral gray matter mask were used in the calculation of gFCD, and the gFCD for any given voxel (x0) was calculated as the total number of functional connections [k(x0)] between x0 and all other voxels using a growth algorithm. This procedure was repeated for all voxels. To increase the normality of the distribution, each gFCD value was divided by the mean value of all the included voxels. A 6 × 6 × 6 mm 3 Gaussian kernel was used to spatially smooth the gFCD maps to minimize the impact of anatomical differences between the participants (Zhuo et al., 2014). temporal parafoveal region = 750-1,500 µm.

| Statistical analyses
One-way analyses of variance were used to analyze the participants' demographic and clinical characteristics (McHugh, 2011). The Mann-Whitney U test was used to compare the GMV and gFCD values between the groups (Matsouaka & Betensky, 2015). Sex-related differences in terms of the GMV and gFCD corrected were explored using a chi-square test between the two groups (Thomas, 1990). A p-value <.05 was considered significant.

| Demographic and clinical characteristics of the participants
MRI data from five patients and OCT data from two patients were excluded from the analysis due to poor acquisition quality. Of the remaining 113 patients, 109 demonstrated brain gray matter reductions, and 97 showed brain and retinal co-structural impairment.
Data from these 97 patients were used for the analysis. Patients were divided into four groups according to their SAPS-AH and SAPS-VH scores. Score greater than 4 denoted greater severity. Patients with scores between 2 and 3 were defined as having middle-to-moderately severe AHs and VHs. According to the above-mentioned

| GMV reduction
Compared to the healthy control group, all four patient groups demonstrated GMV reductions, which were predominantly observed in the temporal, frontal, parietal, and occipital lobes, with the greatest extent of GMV reduction observed in the frontalparietal lobe across all four groups ( Figure 1a). The four groups, ranked by the extent of GMV reduction according to the peak value of reduction in the frontal-parietal lobe, were as follows: FUSCHSASV, FUSCHMASV, FUSCHSAMV, and FUSCHMAMV.

TA B L E 1 Participants' demographic and clinical characteristics
The extent of GMV reduction demonstrated an inverted U-shaped pattern ( Figure 1b).

| gFCD alterations
Compared to the healthy control group, all four patient groups dem-

| GMV and gFCD differences among the patient groups
Alterations in the GMV were compared across the patient groups ( Figure 1d). Although the extent of the GMV and gFCD decreases in some brain regions differed between patient groups, and the GMV and gFCD even increased to a greater degree in some patient groups than others, the largest decrease in the scope of GMV and gFCD alterations in comparison with that in the control group remained in the frontal and parietal lobes. This indicates that GMV/gFCD alterations in these regions are uniform pathological features of FUSCHAV.

| Retinal alterations
While all four patient groups showed significant retinal thinning compared to the healthy control group, there were no significant differences among the four patient groups in retinal thickness ( Figure 2 and Table 2).

| AIM OF THE S ECONDARY FOLLOW-UP S TU DY
In the pilot study, we observed that the GMV and total retinal thickness reduced in cases with FUSCHAV. We also noted a compensatory gFCD increase in these patients. Compared to the healthy controls, we observed serious GMV and retinal thickness reductions in the FUSCHAV patients. Previous studies have reported that antipsychotic agent treatment deteriorates the rate of GMV reduction within the 2-3 years after acceptance. However, few studies have focused on whether retinal thickness deteriorates with antipsychotic agent administration; this question is also the research hot spot proposed by Silverstein (2016). Based on our pilot study and the findings of previous studies, pertaining to the effect of antipsychotic agents on the brain and retina, we attempted to conduct a secondary follow-up study to clarify the relationship between antipsychotic agent treatment and brain and retinal alterations in the FUSCHAV patients. In the secondary follow-up study, we hypothesized that (a) brain and retinal alterations will be accom-

| S ECONDARY FOLLOW-UP S TUDY PROCEDURE
In the secondary follow-up study, the follow-up time was set at 6 months after the administration of regular antipsychotic agents. The PANSS and SAPS scales were used to assess treatment effect. GMV, gFCD, and total retinal thickness were adopted to assess the effect of antipsychotic agent treatment on the brain and retina in the FUSCHAV patients. The MRI scanner, OCT system, and the parameters and analysis methods employed were the same as those used in the pilot study.  (Table 3).

| RE SULTS OF THE S ECONDARY FOLLOW-UP S TUDY
In the secondary follow-up study, we observed that the GMV and gFCD sharply decreased (Figures 3 and 4). The maximum GMV reduction scope was observed in the FUSCHSASV group, followed by the FUSCHSAMV, FUSCHMASV, and FUSCHMAMV groups ( Figure 3). The maximum gFCD reduction scope was observed in the FUSCHSAMV group, followed by the FUSCHMASV, FUSCHMAMV, TA B L E 2 Total retinal thickness alterations among different groups on OCT and FUSCHSASV groups. More notably, no significant differences were observed in the degree of GMV reduction between the four groups (a t test was used to compare two groups). However, the gFCD did not show significant differences across the groups (a t test was used to compare two groups) (Figure 4). The inverted U-shaped pattern of the GMV and gFCD increase we observed at the baseline disappeared. The maximum GMV reduction rates compared to those at the baseline were 1.37% (self-comparison) and 1.93 (compared to that in the healthy controls at the baseline); this value was observed in the FUSCHSASV group.
More interestingly, although we observed that the GMV reduction rate was correlated with the total dosage of chlorpromazine equivalent in the FUSCHSASV and FUSCHSAMV groups, the correlation coefficient was very small (r = .0023, p < .001; r = .0017, p < .001, respectively). We believe that this correlation was nonsignificant, as, according to statistical theory, r < .2 usually denotes that the correlation is delicate (Burger & Missov, 2016;Mbuagbaw et al., 2013).
In this secondary follow-up study, the total retinal thickness deteriorated after 6 months of treatment with antipsychotic agents. The maximum total thickness decrease was observed in the FUSCHSAMV group, followed by the FUSCHSASV, FUSCHMASV, and FUSCHMAMV groups. The maximum reduction rate was 14.8% (self-compared, compared to that at the baseline) and 20.0% (compared to that in the healthy controls at the baseline) ( F I G U R E 3 GMV reductions after antipsychotic agent use. GMV, gray matter volume was observed between the total thickness deterioration rate and dosage of chlorpromazine equivalent.

| D ISCUSS I ON
The present pilot and secondary follow-up study investigated the brain and retinal alterations that occurred in FUSCHAV patients. In the pilot study, all four FUSCHAV groups demonstrated GMV reductions, primarily in the temporal, occipital, frontal, and parietal lobes. These findings indicate that the primary auditory and visual cortexes are impaired in FUSCHAV patients (Bernardin et al., 2019;Csaszar, Kapocs, & Bokkon, 2019;Moseley, Mitrenga, Ellison, & Fernyhough, 2018). First-episode schizophrenia patients with congenital deficiencies or serious brain damage generally have severe AHs and VHs. More notably, the GMV reduction in the frontal and parietal lobes suggests that the cortex involvement affects higher integration. The patients in the current study were experiencing their first episode of schizophrenia and were unmedicated; therefore, our findings suggest that GMV reductions were present prior to the psychotic episode (Bordier, Nicolini, Forcellini, & Bifone, 2018;Zmigrod et al., 2016) and that there may be a brain-related pathology associated with the onset of AHs and VHs. Moreover, the FUSCHAV patients demonstrated increased gFCD in the same brain regions showing GMV reductions, suggesting that functional hyperactivity may compensate for structural impairment (Cao et al., 2018;Collier et al., 2014;Csaszar et al., 2019;Kremlacek et al., 2016;Xu, Qin, Liu, Jiang, & Yu, 2016). While some brain regions demonstrated an increase in the GMV and gFCD between the different FUSCHAV groups, these alterations may be distinct to each patient group. In addition to alterations in the brain gray matter and connectivity, the total retinal thickness was decreased in the FUSCHAV patients compared to that in the healthy controls, suggesting the presence of brain and eye co-impairment.

| Limitations
There are several limitations to the current pilot and secondary follow-up study that must be considered in result interpretation. First, the secondary follow-up study was conducted based on the findings of the pilot study; the validity of this method has not been confirmed.
Further studies must be performed to test the validity of this method.
Second, in the secondary follow-up study, we observed brain and retinal deteriorations after 6 months; it is of concern that 6 months may not be sufficient in confirming a patient's brain and retinal deterioration tendency; therefore, long-term ( Sixth, the small sample size of each group limits the significance of the findings. Seventh, the patients were artificially divided into four categories, and this method requires further study for validity testing. Eighth, total retinal thickness was used to assess eye impairment which was limited by the use of OCT-4000. The retina was not divided into ten layers for the precise clarification of retinal impairment, and, as such, may not provide the full picture of retinal impairment in FUSCHAV. Ninth, the correlations between the GMV, we observed the gFCD, retinal thickness, antipsychotic agent dosage, and clinical symptoms were not significantly in our pilot and secondary followup study. These findings weaken our conclusions, which state that antipsychotic agent treatment is accompanied by deteriorated GMV and retinal thickness. Future large-sample, long-term cohort studies should be conducted to enhance the strength of our conclusions.

| CON CLUS ION
Despite the above-described limitations, our findings indicate the need for heightened alertness on brain and retinal impairments in FUSCHAV, and monitoring for the further deterioration induced by antipsychotic agent treatment in clinical practice.

ACK N OWLED G M ENTS
This work was supported by grants from the National Natural Science

CO N FLI C T O F I NTE R E S T
None declared.

AUTH O R S CO NTR I B UTI O N S
CZhuo, XB, CC and XL conceived and designed research; CC, XB, DJ, and GL collected data and conducted research; XM, RL, and LW analyzed and interpreted data; YX and CZhou wrote the initial paper; CZhuo and CZhou revised the paper; XL had primary responsibility for final content. All authors read and approved the final manuscript.

DATA AVA I L A B I L I T Y S TAT E M E N T
The datasets generated and analyzed during the present study are available from the corresponding author on reasonable request.