Your password has been changed

Enter your email address below and we will send you reset instructions

If the address matches an existing account, you will receive an email with instructions to reset your password

Enter your email address below and we will send you your username

If the address matches an existing account, you will receive an email with instructions for retrieving the username

The American Psychiatric Association (APA) has updated its privacy policy and terms of use to include new information specifically for individuals in the European Economic Area. As stated in the privacy policy and terms of use, this website uses cookies, including in order to provide the best online experience and services tailored to your preferences.

Please read the entire privacy policy and terms of use. By closing this message, browsing this website, continuing to navigate or otherwise continuing to use APA's website, you confirm that you understand and accept the terms of the privacy policy and terms of use, including the use of cookies.

Search for more papers by this author

Search for more papers by this author

Search for more papers by this author

Search for more papers by this author

Search for more papers by this author

Search for more papers by this author

Search for more papers by this author

Search for more papers by this author

Search for more papers by this author

Search for more papers by this author

Search for more papers by this author

Search for more papers by this author

Search for more papers by this author

Search for more papers by this author

Search for more papers by this author

Search for more papers by this author

Search for more papers by this author

Search for more papers by this author

Search for more papers by this author

Search for more papers by this author

Search for more papers by this author

Search for more papers by this author

Search for more papers by this author

The inability of antipsychotic drugs to solve the symptoms of patients with schizophrenia creates a clinical challenge called treatment resistance. The reason for treatment resistance is unclear, but it is related to an earlier age of onset and more severe cognitive deficits. The authors tested the hypothesis that white matter defects involved in neurodevelopment and the severity of cognitive deficits in schizophrenia are associated with a higher risk of treatment resistance.

The sample of the study (N=122; average age, 38.2 years) included patients with schizophrenia at the start of treatment (N=45), patients whose symptoms were responsive to treatment (N=40), and patients whose symptoms were resistant to treatment ( N=37), and healthy control subjects (N=78; mean age, 39.2 years). The White Matter Regional Vulnerability Index (RVI) is tested as a predictor of treatment resistance and cognitive deficits. A higher RVI indicates a better agreement between the anisotropy of the diffusion tensor imaging score of the individual's brain and the pattern determined by the largest meta-analysis of white matter defects in schizophrenia to date.

Patients with treatment resistance symptoms showed the highest white matter RVI (mean = 0.38 [SD = 0.2]), which was significantly higher than the RVI of patients with treatment response symptoms (mean = 0.30 [SD = 0.02]). At the beginning of treatment, the RVI of patients with schizophrenia was significantly higher than that of healthy controls (mean = 0.18 [SD = 0.03] and mean = 0.13 [SD = 0.02], respectively). RVI is significantly related to processing speed and the performance of negative symptoms.

Schizophrenia affects the white matter microstructure in specific regional patterns. Susceptibility to defects in white matter areas is related to an increased likelihood of treatment resistance. Overcoming the development of treatment resistance in schizophrenia should consider white matter as an important goal.

Modern antipsychotic drugs cannot solve the clinical symptoms of approximately one-third of patients with schizophrenia, and this disease is called treatment resistance (1). For decades, this clinical challenge has prompted attempts to develop more effective next-generation antipsychotic drugs (2-6). However, due to insufficient understanding of the neurobiological mechanisms of treatment resistance and the lack of effective brain biomarkers, research has been hindered. Current antipsychotic drugs share their neurotransmitter targets in the dopaminergic system. Their ineffectiveness in patients with treatment-resistant symptoms suggests that other mechanisms and neurotransmitter systems may be involved, although the evidence is still in its preliminary stages (7-10). Some structural neuroimaging studies have shown that treatment resistance is related to a decrease in gray matter and cortical thickness (7, 11, 12). However, a meta-analysis found that when comparing patients with refractory schizophrenia with patients whose clinical symptoms responded to treatment, there were no repeated neuroimaging findings (13).

Patients with refractory schizophrenia have two consistent characteristics: the age of onset is earlier and the cognitive deficits are more severe (14-17); this indicates the contribution of neurodevelopment and cognition to treatment resistance. Therefore, brain measurements that track neurodevelopmental and cognitive dysfunction in schizophrenia may help identify biomarkers of treatment resistance. White matter shows the promise of being the focus of schizophrenia research. The largest meta-analysis study of white matter defects in schizophrenia so far was carried out by the Enhanced Neuroimaging Genetics through Meta-Analysis (ENIGMA) Consortium, which identified the pattern of white matter defects in specific regions (18 ). The regional difference patterns found in the ENIGMA study confirm the results of previous diffusion tensor imaging (DTI) studies, which showed significant defects in the frontal lobe-related white matter areas. This includes the genera and bodies of the anterior corona and corpus callosum (19-23). The pattern shown in the ENIGMA study is also consistent with the histological findings of decreased glial cell density and myelination in the frontal lobe of patients with schizophrenia (24-26). The pattern of regional white matter defects identified in the ENIGMA study leads to core cognitive dysfunction, especially the processing speed deficit in schizophrenia (27-30). Interestingly, of all cognitive abnormalities related to treatment resistance, the deficit in processing speed is the largest (16). The development of white matter and processing speed all follow the inverted U-shaped neurodevelopmental trajectory, and the peak myelination of the related white matter overlaps with the peak processing speed capability (31, 32). The origin of the white matter defect pattern in schizophrenia is not fully understood, but developmental changes related to developmental disorders prevent the normal development of late myelinated areas, which may prevent the establishment of normal interneuronal communication patterns (33). This may lead to an observational comparison of regional defects between the late and early myelinated white matter regions in schizophrenia, as seen in ENIGMA results and other studies (18, 34, 35). Therefore, we hypothesized that the regional white matter defect pattern may represent the neurobiological mechanism leading to treatment resistance in schizophrenia. To test this hypothesis, we compared patients with refractory schizophrenia with patients whose symptoms responded to treatment to assess whether white matter area defects indicate treatment resistance. At the same time, we compared the regional defects of patients who underwent imaging examinations within 2 weeks of starting antipsychotic treatment with healthy control subjects. We use this comparison to control the effects of potential chronic antipsychotics on white matter defects, and to determine whether the identified white matter defects are also related to schizophrenia, not to chronic diseases and treatment processes. The evidence supporting this hypothesis suggests that white matter defects are a contributing factor to treatment resistance.

Participants were 122 schizophrenic patients (among them 57 men; mean age 38.2 years [SD=13.3]) and 78 healthy control subjects (among them 37 men; mean age 39.2 years [SD=14.0]) ( Table 1). Schizophrenia patients include individuals with treatment-resistant symptoms (including 17 males; average age, 47.8 years [SD = 8.9]), as the main control group, patients with treatment-responsive symptoms (including 18 males; average age, 46.3 years old [SD=11.5]), their age and gender frequency match (all p-values> 0.2). In addition, a third patient cohort was recruited, which included individuals (22 males; mean age 28.6 years [SD = 10.1]) within 2 weeks before the start of treatment to determine whether any identified white matter defects were also related to mental illness. Schizophrenia is independent of chronic diseases and treatment processes. This treatment start group is necessarily younger; however, the age and gender distribution frequencies of the three patient groups and the healthy control group match (all p-values> 0.2). The data was collected in 2017 and 2018. The patient came from Beijing Huilongguan Hospital. Recruit healthy control subjects through local advertising. All patients met the DSM-IV criteria for schizophrenia. Participants have a homogenous Chinese ethnic background, which is believed to facilitate the identification of treatment resistance biomarkers, because race may have a significant influence on treatment resistance (36). According to the Declaration of Helsinki, written informed consent was obtained from all study participants, and the study protocol was approved by the local ethics committee.

Table 1. Demographic and clinical characteristics of patients with schizophrenia and healthy controls​​

Table 1. Demographic and clinical characteristics of patients with schizophrenia and healthy controls​​

The treatment resistance and treatment response groups are defined according to consensus guidelines (1). Patients with refractory schizophrenia meet the following criteria: 1) Little response to at least two different antipsychotic drugs, equivalent dose of chlorpromazine ≥600 mg/day for ≥12 weeks, 2) Brief psychiatric assessment Scale (BPRS) score ≥45, and 3) Clinical Global Impression Severity Scale (CGI-S) score ≥4 during the current evaluation period. The treatment response group is defined as the period of good clinical response to antipsychotic drugs, as shown by the CGI-S score <3 for a duration of ≥12 weeks. The treatment resistance group and the treatment response group were frequency matched in terms of age, gender, years of education, and treatment duration. Patients who did not meet the criteria of either group were excluded.

Participants in the treatment initiation group had not been exposed to antipsychotics prior to study registration and were included to determine whether treatment resistance biomarkers (if found) were present at the onset of the disease with minimal antipsychotic exposure. These patients received treatment immediately upon entering the study, and the treatment helped stabilize them for MRI scans. The imaging data was collected within 2 weeks after the start of treatment.

All participants had no current or past neurological diseases, unstable major medical conditions, or current or previous substance dependence (although nicotine dependence is allowed). Table 1 summarizes the demographic and clinical characteristics of the study participants.

6 patients did not take drugs (N=4, N=2, and N=0 in the treatment initiation group, treatment response group, and treatment resistance group, respectively). Seven patients are taking first-generation antipsychotic drugs (N=4, N=3, and N=0 in the three groups). The remaining patients took the following second-generation antipsychotic drugs: risperidone (N=17, N=11, and N=7 in the three groups), clozapine (N=0, N=9, and N=22, for three groups) Group), Olanzapine (N=9, N=11, N=7 in the three groups), Aripiprazole (N=8, N=5, N=2 in the three groups), Paliperidone (N=2, N=0 and N=3 in the three groups), or amisulpride, iloperidone, lurasidone or quetiapine (N=1, N=4 in the three patient groups) And N=4), respectively); Among the patients taking second-generation antipsychotics, 45 took more than one antipsychotic (the three groups were N=4, N=7, and N=34). Calculating the chlorpromazine equivalent (37), the average chlorpromazine equivalent dose of patients in the treatment resistance group was almost twice that of patients in the treatment response group (p<0.001) (Table 1). Patients in the treatment start group took much smaller doses per day, with an average taking time of 4.2 days (SD=2.3; range, 0-12 days).

One of the three attending psychiatrists used the Positive and Negative Syndrome Scale (PANSS), BPRS and CGI​​-S to evaluate the patients; the reliability among the interactors was maintained at >0.80. BPRS and CGI​​-S are only used for group definitions. PANSS is used for symptom assessment. Use the MATRICS Consensus Cognitive Battery (MCCB) to assess cognitive function, covering seven cognitive domains and a comprehensive score (38-40). The original score is converted to the Chinese standard T score (40).

A 3-T Prisma MRI scanner (Siemens Medical Solutions, Erlangen, Germany) equipped with a 64-channel radio frequency head coil was used to collect image data at the Imaging Research Center of Beijing Huilongguan Hospital. DTI data is a spin echo and echo plane imaging sequence with a spatial resolution of 1.7×1.7×1.7 mm (TE=87 ms, TR=8,000 ms, field of view=200 mm, axial slice orientation, 82 slices) Collected without gaps, 98 isotropically distributed diffusion weighting directions, two diffusion weighting values ​​[b=0 and 1,000 seconds/mm2] and five b=0 images). Participants' head movement is minimized by restrictive fillers. DTI data is processed using the ENIGMA-DTI analysis pipeline (https://www.nitrc.org/projects/enigma_dti) (41). All data included in the analysis passed ENIGMA-DTI quality assurance and quality control. Based on the ENIGMA-DTI map, regional white matter anisotropy scores (FA) were generated for 21 main regions and averaged across the entire hemisphere.

This cohort is independent of the samples used in the ENIGMA study and our previous studies (18, 27). The auxiliary purpose is to use samples to replicate the association between the pattern of regional defects and cognitive deficits identified in the ENIGMA study. Schizophrenia (27).

The ENIGMA study provides a meta-analysis of case/control effects related to schizophrenia in 21 major white matter areas (Cohen's d) (for more details, see Table S1 in the online supplement). We developed an individual-level regional injury index, the Regional Vulnerability Index (RVI), as a simple measure of the consistency between the individual FA patterns in these 21 regions and the expected schizophrenia patterns in these regions, as described by ENIGMA Show the result. The FA of each white matter area is calculated by 1) calculating the residual value after the regression of the effects of age and gender, and 2) for each individual, subtracting the average value of one area and dividing it by the deviation of the standard from the healthy control group. This generates a normalized z-value vector (one for each region) for each person in the research sample. Then the RVI is calculated as the correlation coefficient (standardized dot product) between the area z-value vector of the research object and the magnitude vector of the regional schizophrenia health control effect in the ENIGMA study. We use the term “vulnerability” here to express a narrow definition—whether the regional white matter defect pattern identified in the ENIGMA study is related to the increased likelihood of treatment resistance. A higher RVI value means that the white matter regional value pattern follows the regional vulnerability pattern of schizophrenia determined by the ENIGMA meta-analysis. While controlling for age and gender, a general linear model was used to compare all groups of imaging and cognitive measures. Bonferroni correction is used to correct the number of inspection areas. While controlling for age and gender, the correlation between RVI and clinical measurement was checked by bivariate correlation analysis, and Bonferroni correction was applied.

Compared with healthy control subjects, the average FA of the whole brain of patients with schizophrenia was significantly reduced (Cohen's d=0.69, t=5.1, p=1×10−6), and it was corrected in four of the 21 regions After the area, the area effect size was significantly multiple (N=21) comparisons (see Table S1 in the online supplement). The average effect size in this sample is not significantly different from the effect size in the ENIGMA study (t=1.2, p=0.2). The impact of the regional schizophrenia-health control difference is related to the impact in the ENIGMA study (r=0.85, p=1×10-5) (Figure 1A). When the three patient groups were compared with the healthy control group, this pattern was also observed (Figure 1B-1D), and the effect was the strongest in the anti-treatment group (r=0.92, p=1×10-8) ( Figure 1B-1D) 1B). In the ENIGMA study, the anterior radiating crown and the frontal lobe tracts of the corpus callosum and knees showed the greatest health control effect in schizophrenia, and they were also the area where FA decreased the most in the treatment resistance group compared with the healthy control group (Figure 1B). However, there were no significant differences in the whole brain average or regional FA measurements between the treatment resistance group and the treatment response group (Table 2) or the treatment start group and the healthy control group (Table 2).

Figure 1. The size of the impact in the 21 white matter regions (schizophrenia group compared to the healthy control group), plotted with the size of the impact in the ENIGMA studya

a The average effect size of the group of 21 white matter regions (Cohen's d) is plotted against the effect size of this region in the meta-analysis enhanced neuroimaging genetics (ENIGMA) meta-analysis of all subjects (Figure A) and three Schizophrenia subgroup (Figure BD). The significant correlation indicates that the pattern of fractional anisotropy defects in the Chinese sample is consistent with the pattern in the ENIGMA study. All effect size data (mean and standard deviation) are listed in Table S1 supplemented online.

Table 2. Matrix Consensus Cognitive Battery (MCCB) Cognitive Domain and Total Score, and the Association with Treatment Stage and Regional Vulnerability Index (RVI)a

a All analyses use age and gender as covariates.

b Bonferroni-corrected p-values ​​are associated with treatment status (14 comparisons; p<0.05/14=0.0036) or MCCB and RVI (21 correlation analyses; p<0.05/21=0.0024).

c is nominally significant at p<0.05.

Table 2. Matrix Consensus Cognitive Battery (MCCB) Cognitive Domain and Total Score, and the Association with Treatment Stage and Regional Vulnerability Index (RVI)a

Compared with the healthy control group, the average RVI of the combined patient group (treatment start, treatment response, and treatment resistance) was significantly higher (mean = 0.29, SD = 0.01, compared to mean = 0.13, SD = 0.02; t = 6.4, df=198, p=1×10−9). The treatment resistance group had the highest RVI (mean=0.38, SD=0.02), followed by the treatment response group (mean=0.30, SD=0.03) and the treatment initiation group (mean=0.18, SD=0.02), and healthy Compared with the control group, all three patient groups showed significant differences (mean = 0.13, SD = 0.02; all t-values ​​≥ 2.7, p-values ​​≤ 0.01) (Figure 2A).

Figure 2. Regional Vulnerability Index (RVI)a of healthy control subjects and schizophrenia patients

a Figure A shows the average coefficient of RVI plotted by group. Figure B shows a significant sequential trend in the increase in RVI relative to group allocation (p<0.001).

The RVI of the treatment resistance group was significantly higher than that of the treatment response group (t=2.7, df=75, p=0.01), indicating that higher RVI is related to treatment resistance. The RVI of patients in the treatment initiation group was significantly higher than that of individuals in the healthy control group (t=2.8, df=121, p=0.007), indicating that higher RVI is unlikely to be the result of chronic diseases or chronic diseases. Antipsychotic drug exposure. Compared with patients in the treatment start group, chronic patients (combined treatment resistance and treatment response cohort) have significantly higher RVI (t=4.2, df=153, p=4×10-5), which indicates that there may be Other disease progression or chronic drug effects (Figure 2).

These relationships can be further illustrated by examining the sequential trends of the group RVI in the following order: health control, treatment initiation, treatment response, and treatment resistance. We applied intermediate scores to assign the scores of group variables, and performed general linear regression analysis on the ordinal trend (42), and found that the trend was statistically significant (F=57.2, df=1, 198, p<0.001) (Figure 2B) .

The total score of MCCB decreased the most in the treatment resistance group (t=11.6, df=113, p=3×10-17), followed by the treatment response group (t=8.4, df=116, p= 7×10-13) ) And the treatment start group (t=8.2, df=121, p=2×10-12), compared with healthy control subjects (Table 2). Compared with patients in the treatment response group, patients with refractory schizophrenia after Bonferroni correction showed no significant damage to the MCCB total score or domain score; however, there were nominally significant differences in working memory between the two groups (Table 2) . There were significant differences in cognitive measures between the treatment start group and the healthy control group (Table 2). After adjusting for multiple comparisons, RVI was only significantly correlated with the treatment speed of the treatment response group (r=-0.53, p=7×10-4) (Table 1). We also explored the correlation between MCCB and the average FA and 21 regional values ​​of the whole brain. However, after correction for multiple comparisons, there was no significant regional FA-MCCB correlation in the combined or single patient group, except for the significant correlation between the average whole brain FA and the processing speed score (r = 0.31, r = 5× 10 −4) In a pooled patient sample after correction for multiple comparisons.

There was no significant correlation between RVI and PANSS total score, positive or general symptom score (all r-values ​​<0.1, all p-values>0.2), but there was a significant correlation with negative symptoms in the combined patient group (r=0.26, p=0.002). However, after Bonferroni correction, no regional FA values ​​were significantly associated with negative symptoms (all r values ​​<0.27, all p values> 0.003). There was no significant regional FA value in any patient subgroup (all p values> 0.1). In any combination or single patient group, RVI was not significantly related to disease duration (all p-values> 0.20), chlorpromazine equivalent dose (all p-values> 0.41), or smoking status (all p-values> 0.15) sex. More than half of the patients in the treatment resistance group are taking clozapine (N=22, compared with N=15 for patients not taking clozapine), but between patients taking clozapine and patients not taking clozapine There is no difference in the RVI value (t=0.2, df=35, p=0.8).

The magnitude of the regional FA effect in the ENIGMA study is significantly correlated with the correlation coefficient of regional FA processing speed (r=0.90, p<0.001) (Figure 3A) and the correlation coefficient of regional FA working memory (r=0.84, p<0.001) (Figure 3A) 3C), copy the findings of another sample (27). When controlling working memory, ENIGMA-based regional effect size is still significant in explaining the correlation coefficient of regional FA processing speed (part r=0.88, p<0.001) (Figure 3B), but the opposite is not significant (part r=0.25 , p=0.4) (Figure 3D). Permutation analysis showed significant differences in partial correlation coefficients (p=0.01), indicating that the relationship between cognitive measurement and white matter is mainly driven by processing speed rather than working memory. Similar trends were observed in patients (Figure 3E-H) and control subjects (Figure 3I-L). Our findings in this Chinese patient sample (Figure 3) largely replicate the findings in the US sample (see Figure S1 in the online supplement).

Figure 3. Reproduction of the relationship between cognition and white matter susceptibility to schizophrenia identified in the ENIGMA studya

a ENIGMA=enhanced neuroimaging genetics through meta-analysis. These graphs show a reproduction of the relationship between cognition and white matter susceptibility to schizophrenia previously discovered by our research team (27) (see the original discovery graph for the US sample in Supplementary Figure S1 online). It also shows the impact size of the 21 main white matter areas in the ENIGMA study. Compared with healthy control subjects in the ENIGMA meta-analysis (x-axis), higher values ​​indicate that patients with schizophrenia have more severe damage. In addition, the correlation coefficient (y-axis) between the score anisotropy (FA) of each region in the Chinese sample in this study and the cognitive measure of processing speed or working memory is also shown. Figure A shows the relationship between the correlation coefficient of regional FA processing speed (y-axis) and the size of the FA effect (x-axis, Cohen's d) in schizophrenia from the ENIGMA project. Panel B displays the partial correlation coefficient in panel A after the working memory is calibrated. Figure C shows the relationship between the correlation coefficient between the regional FA value and working memory (y-axis) and the size of the regional FA effect (x-axis) in schizophrenia. Panel D displays the partial correlation coefficient in panel C after correcting the processing speed. This was checked in the complete sample (figure AD) and then in the three patient groups (figure EH) and the healthy control group (figure IL).

We found that the white matter area defect pattern of schizophrenia is significantly related to treatment resistance. White matter RVI can significantly distinguish between patients with refractory schizophrenia and patients with treatment-responsive schizophrenia, even if the treatment duration matches and there is no overall FA difference. The RVI measurement further significantly distinguished patients at the beginning of treatment from healthy control subjects. These results indicate that the extent of white matter vulnerability in the region defined by RVI can be observed in the initial diagnosis and treatment of schizophrenia, and may indicate a tendency to resist treatment with currently available antipsychotic drugs. Follow-up longitudinal studies are needed to test whether higher RVI will change with the development of treatment resistance.

Treatment resistance is associated with cognitive deficits, especially in terms of processing speed (15, 16) and negative symptoms (14), although the underlying mechanism is still unknown (28-30). The data from this study suggests that white matter may represent a common underlying neurobiology, because the pattern of regional fragility seems to be related to treatment resistance and more severe negative symptoms. We also replicated the results of previous studies in which the schizophrenia pattern in the ENIGMA meta-analysis predicted the association between FA and processing speed and working memory (27) (Figure 3; see also Figure S1 in the online supplement). The study used only two cognitive tasks, which is considered a limitation (27). In this study, we used MCCB to replicate these findings, and the similar patterns of two different (US and Chinese) samples further validated the vulnerability of white matter regions in indexing the core cognitive deficits of schizophrenia.

The results of white matter volume reduction initially suggested that white matter is related to treatment resistance (43), but the opposite result was also reported (11). Several DTI studies have further shown that there is a white matter effect on patients with refractory schizophrenia (44, 45). However, it is difficult to explain whether these previous findings are related to general schizophrenia or are specific to treatment if a sufficient sample size is not used to directly compare patients who are matched with age, gender, and duration of treatment and who have responded to treatment. Fight against schizophrenia. We found that compared with healthy control subjects, FA in multiple white matter regions of schizophrenia patients was significantly reduced (see Table S1 in the online supplement), which is consistent with many previous studies (19-23, 27, 46, 47). After correction for multiple comparisons, no single white matter area can consistently distinguish patients with refractory schizophrenia from patients with treatment-responsive schizophrenia. Why RVI rather than individual regional measures capture treatment resistance (compared to treatment response) is not clear. RVI is an easily available index that correlates the normalized area FA of each person with the size of the schizophrenia effect in the ENIGMA study, and assumes that the associated white matter areas that reflect the late development are highly vulnerable to schizophrenia and early The contrast between the lower vulnerabilities. -Developing regions (34, 48, 49). We speculate that by considering the white matter of the entire brain, RVI may reduce the non-specific effects that affect all white matter regions, and further aggravate the regional effects specific to schizophrenia (Figure 1A-1C) and treatment resistance (Figure 1D). Therefore, the higher RVI in the patient group may have identified individuals with more severe neurodevelopmental white matter damage patterns, and these individuals may be more susceptible to treatment resistance.

Compared with patients with treatment-responsive schizophrenia, patients with treatment-resistant schizophrenia always take higher doses of drugs, including multiple drugs, and are more likely to receive a prescription for clozapine. These factors may hinder the separation of therapeutic resistance biomarkers and drug effects, thereby confounding the neurobiological findings of therapeutic resistance. To alleviate this concern, we analyzed a group of patients who were evaluated within 2 weeks of starting antipsychotic medication. We observed significant heterogeneity in the RVI value within this patient group. However, candidate anti-therapeutic biomarkers exist in patients with the least exposure to antipsychotics, excluding the possibility that higher RVI is a chronic disease effect, and even associated with schizophrenia at the beginning of treatment. However, this study is limited by its cross-sectional nature, and follow-up studies are needed to test whether the RVI observed at this stage can predict treatment resistance compared to the treatment response of patients who started treatment.

White matter pathways are most closely related to treatment resistance—for example, fornix (the main white matter tracts of axon fibers inside and outside the hippocampus) and precoronary radiation (the main white matter connecting the ipsilateral prefrontal cortex) (effect size approximately 1.0)— -It is a well-known associative pathway that supports cognitive function (50, 51). In patients with treatment-responsive schizophrenia, the effect size of the same beam is weak (for example, the range of fornix and anterior coronary radiation, 0.5-0.6) (Figure 1C), which indicates that impairment of the associated beam supporting cognition may play a role Treatment resistance is greater. The agreement (r = 0.92) between the white matter FA regional effect size of patients with refractory schizophrenia in this study and the patients in the ENIGMA study also provides a new perspective for the results of ENIGMA. Currently available antipsychotic drugs fail to relieve the symptoms of refractory schizophrenia and have limited impact on cognitive deficits (52, 53). ENIGMA-DTI samples are global, meta-analysis aggregation may eliminate specific local drugs and environmental variables, resulting in untreated and shared cross-site schizophrenia-related neurobiology. The significant agreement between the patients in the ENIGMA study and the treatment resistance group in this study indicates that the white matter area effect size pattern identified in the ENIGMA study may be an unmet key treatment goal (ie treatment resistance, and cognitive deficits) The basis of) (27) In schizophrenia. This may also indicate that the white matter area effect size model identified in the ENIGMA meta-analysis is limited because it has a low correlation with schizophrenia during the initiation of treatment (r = 0.58), but is associated with treatment resistance in schizophrenia Stronger.

Another potential limitation is that patients with refractory schizophrenia choose clozapine for treatment. Post-hoc analysis did not show an association between RVI and clozapine, and there was already an elevated RVI in the treatment start group. The focus of this study is the difference between age- and gender-matched treatment resistance groups and treatment response groups. The average age of the treatment initiation group is significantly lower than that of the treatment resistance and treatment response groups, which is a potential limitation. However, RVI was developed to be independent of age and gender, and all analyses were performed while controlling for age and gender. Finally, only FA is used to indicate white matter abnormalities, and other diffusion parameters (axial, radial, and average diffusion rate) are not explored. We chose FA because it exhibits a higher sensitivity to schizophrenia defects than these other parameters (18). Future research should re-examine these findings using more advanced DTI parameters (54).

Part of the cause of treatment resistance in schizophrenia may be white matter defects. Patients who exhibit a pattern of regional white matter damage in the late-developing frontal lobe-related fibers and with no or limited damage in the early-developing sensory and motor fibers are more likely to have symptoms that are resistant to contemporary antipsychotic drugs. The development of new therapies and therapies to overcome the treatment resistance of schizophrenia should more strongly consider strategies for white matter-related mechanisms.

Dr. Rowland has served as a consultant for the development and commercialization of Otsuka Pharmaceutical, specializing in the education platform PsychU. Dr. Jahanshad and Dr. Thompson received research funding from BioGen. Dr. Hong has received or plans to receive research grant support or consulting fees from Heptares, Mitsubishi, Neuralstem, Pfizer, Regeneron, Sound Pharma, Taisho, Takeda, and Your Energy Systems LLC. All other authors report no economic relationship with commercial interests.

Supported by China's National Key R&D Program (Approved 2016YFC1307000), China's National Natural Science Foundation of China (funded 8107086 and 8141113016) and the National Institutes of Health (Grants R01MH112180, R01MH116948, S10OD023696, R01EB015611, U54 EB020403, T32MH067533 and U01MH108148 ).

1 Howes OD, McCutcheon R, Agid O, etc. : Anti-therapeutic schizophrenia: Psychiatric Treatment Response and Resistance (TRRIP) Working Group Diagnostics and Terminology Consensus Guidelines. Am J Psychiatry 2017; 174:216–229Link, Google Scholar

2 Samara MT, Dold M, Gianatsi M, etc. : Efficacy, acceptability and tolerability of antipsychotic drugs in refractory schizophrenia: a network meta-analysis. JAMA Psychiatry 2016; 73:199–210 Crossref, Medline, Google Scholar

3 Kulkarni J, Gavrilidis E, Wang W, etc. : Estradiol in the treatment of refractory schizophrenia: a large-scale randomized controlled trial for women of childbearing age. Moore Psychiatry 2015; 20:695–702 Crossref, Medline, Google Scholar

4 Hoffman RE, Hawkins KA, Gueorguieva R, etc. : Transcranial magnetic stimulation and drug-resistant auditory hallucinations in the left temporal parietal cortex. Arch Gen Psychiatry 2003; 60:49–56 Crossref, Medline, Google Scholar

5 Heresco-Levy U, Ermilov M, Shimoni J, etc. : Placebo-controlled trials of adding d-cycloserine to conventional neuroleptics, olanzapine or risperidone for schizophrenia. Am J Psychiatry 2002; 159:480–482Link, Google Scholar

6 Kane J, Honigfeld G, Singer J, etc. : Clozapine for the treatment of resistant schizophrenia: a double-blind comparison with chlorpromazine. Arch Gen Psychiatry 1988; 45:789–796 Crossref, Medline, Google Scholar

7 Mouchlianitis E, Bloomfield MA, Law V, etc. : Compared with patients with schizophrenia who respond to treatment, patients with treatment-resistant schizophrenia have higher levels of glutamate in the anterior cingulate cortex. Schizophrenia Bull Market in 2016; 42:744–752 Crossref, Medline, Google Scholar

8 Goldstein ME, Anderson VM, Pillai A, etc. : Glutamatergic neurometabolite in clozapine-responsive and drug-resistant schizophrenia. Int J Neuropsychopharmacol 2015; 18:18(6) pii: pyu117. doi: 10.1093/ijnp/pyu117Crossref, Google Scholar

9 Demjaha A, Egerton A, Murray RM, etc. : Resistance to antipsychotic therapy in schizophrenia is associated with elevated glutamate levels but normal dopamine function. Biological Psychiatry 2014; 75:e11–e13 Crossref, Medline, Google Scholar

10 Berretta S, Heckers S, Benes FM: Searching for the mechanism of mental illness in the human brain. Enlightenment for schizophrenia research. Schizophrenia Research 2015; 167:91–97 Crossref, Medline, Google Scholar

11 Anderson VM, Goldstein ME, Kydd RR, etc. : Extensive gray matter volume reduction in refractory schizophrenia. Int J Neuropsychopharmacol 2015 ;18:pyv016Crossref, Medline, Google Scholar

12 White TP, Wigton R, Joyce DW, etc. : Dysfunctional striatal system in refractory schizophrenia. Neuropsychopharmacology 2016; 41:1274–1285 Crossref, Medline, Google Scholar

13 Nakajima S, Takeuchi H, Plitman E, etc. : Neuroimaging findings of refractory schizophrenia: systematic review: lack of neuroimaging related to refractory schizophrenia. Schizophrenia Research 2015; 164:164–175 Crossref, Medline, Google Scholar

14 Demjaha A, Weinstein S, Stahl D, etc. : Formal thinking disorder in people at very high risk of schizophrenia. 2017 BJPsych Open; 3:165–170 Crossref, Medline, Google Scholar

15 de Bartolomeis A, Balletta R, Giordano S, etc. : Differences between schizophrenic responders and antipsychotic non-responders. Cognitive performance: correlation with disease course, psychopathology, treatment attitude and antipsychotic dose. Psychiatry Research 2013; 210:387–395 Crossref, Medline, Google Scholar

16 Frydecka D, Beszłej JA, Gościmski P, etc. : Analysis of cognitive impairment in patients with refractory schizophrenia. Psychiatry Research 2016; 235:133–138 Crossref, Medline, Google Scholar

17 Elkis H, Buckley PF: Refractory schizophrenia. Psychiatr Clin North Am 2016; 39:239–265 Crossref, Medline, Google Scholar

18 Kelly S, Jahanshad N, Zalesky A, etc. : 4322 Extensive white matter microstructure differences in personal schizophrenia: results from the ENIGMA DTI working group on schizophrenia. Moore Psychiatry 2018; 23:1261–1269 Crossref, Medline, Google Scholar

19 Nazeri A, Chakravarty MM, Felsky D, etc. : Changes in surface white matter in schizophrenia and its relationship with cognitive ability. Neuropsychopharmacology 2013; 38:1954–1962 Crossref, Medline, Google Scholar

20 Friedman JI, Tang C, Carpenter D, etc. : Diffusion tensor imaging results of first-episode and chronic schizophrenia patients. Am J Psychiatry 2008; 165:1024–1032Link, Google Scholar

21 Alba-Ferrara LM, de Erausquin GA: What is anisotropy measured? Insights into the increase and decrease of selective fiber bundle anisotropy in schizophrenia. Front Integr Nuerosci 2013; 7:9 Crossref, Medline, Google Scholar

22 Kubicki M, McCarley R, Westin CF, etc. : A review of diffusion tensor imaging research in schizophrenia. J Psychiatr Res 2007; 41:15–30 Crossref, Medline, Google Scholar

23 Ellison-Wright I, Bullmore E: A meta-analysis of diffusion tensor imaging studies in schizophrenia. Schizophrenia Research 2009; 108:3–10 Crossref, Medline, Google Scholar

24 Davis KL, Stewart DG, Friedman JI, etc. : White matter changes in schizophrenia: evidence of myelin-related dysfunction. Arch Gen Psychiatry 2003; 60:443–456 Crossref, Medline, Google Scholar

25 Hof PR, Haroutunian V, Friedrich VL, Jr. etc. : Loss and spatial distribution of oligodendrocytes in the superior frontal gyrus of patients with schizophrenia in 2003; 53:1075–1085 Google Scholar

26 Uranova NA, Vostrikov VM, Orlovskaya DD, etc. : Oligodendrocyte density in the prefrontal cortex in schizophrenia and mood disorders: a study by the Stanley Association for Neuropathology. Schizophrenia Research 2004; 67:269–275 Crossref, Medline, Google Scholar

27 Kochunov P, Coyle TR, Rowland LM, etc. : The association between white matter and core cognitive deficits in patients with schizophrenia. JAMA Psychiatry 2017; 74:958–966 Crossref, Medline, Google Scholar

28 Knowles EE, David AS, Reichenberg A: Defects in the processing speed of schizophrenia: re-examining the evidence. Am J Psychiatry 2010; 167:828–835Link, Google Scholar

29 Dickinson D, Ramsey ME, Gold JM: Ignore the obvious: a meta-analytical comparison of the numerical symbol coding task and other cognitive measures in schizophrenia. Arch Gen Psychiatry 2007; 64:532–542 Crossref, Medline, Google Scholar

30 Keefe RS, Eesley CE, Poe MP: Defining cognitive decline in schizophrenia. Biological Psychiatry 2005; 57:688–691 Crossref, Medline, Google Scholar

31 Bartzokis G: Glial Pharmacology: Myelination is a common mechanism of psychotropic drug therapy. Neuropharmacology 2012; 62:2137–2153 Crossref, Medline, Google Scholar

32 Bartzokis G, Lu PH, Tingus K, etc. : Life trajectory of myelin integrity and maximum speed of movement. Neurobiol Aging 2010; 31:1554–1562 Crossref, Medline, Google Scholar

33 Gibson EM, Purger D, Mount CW, etc. : Neuronal activity promotes oligodendrocyte production and adaptive myelination in the mammalian brain. Science 2014; 344:1252304 Crossref, Medline, Google Scholar

34 Kochunov P, Ganjgahi H, Winkler A, etc. : The heterochronism of white matter development and aging explains the difference in regional control in patients with schizophrenia. Om Brain Map 2016; 37:4673–4688 Crossref, Medline, Google Scholar

35 Tishler TA, Bartzokis G, Lu PH, etc. : The abnormal trajectory of intracortical myelination in schizophrenia indicates that white matter is related to the pathophysiology of the disease and the therapeutic mechanism of antipsychotic drugs. Biol Psychiatry Cogn Neurosci Neuroimaging 2018; 3:454–462 Crossref, Medline, Google Scholar

36 Lally J, Ajnakina O, Di Forti M, etc. : Two different models of treatment resistance: clinical predictors of treatment resistance in first-episode schizophrenia spectrum psychosis. Psychological Medicine 2016; 46:3231–3240 Crossref, Medline, Google Scholar

37 Gardner KN, Bostwick JR: Response to antipsychotic medication in schizophrenia. Am J Health Syst Pharm 2012; 69:1872–1879 Crossref, Medline, Google Scholar

38 Kern RS, Nuechterlein KH, Green MF, etc. : MATRICS Consensus Cognitive Battery, Part 2: Common Specification and Standardization. Am J Psychiatry 2008; 165:214–220Link, Google Scholar

39 Nuechterlein KH, Green MF, Kern RS, etc. : MATRICS Consensus Cognitive Battery, Part 1: Test Selection, Reliability, and Effectiveness. Am J Psychiatry 2008; 165:203–213Link, Google Scholar

40 Zou Yuzhongjun, Wang Jie: The clinical reliability and validity of the "Schizophrenia Consensus Cognitive Battery Improved Cognitive Measurement and Treatment Research" version. Chinese Journal of Psychiatry. 2009; 42:29–33 Google Scholar

41 Jahanshad N, Kochunov PV, Sprooten E, etc. : Multi-site genetic analysis and voxel heritability analysis of diffusion images: a pilot project of the ENIGMA-DTI working group. Neuroimaging 2013; 81:455–469. doi: 10.1016/j.neuroimage.2013.04.061 Crossref, Medline, Google Scholar

42 Agresti A: Classification data analysis, 3rd edition, Hoboken, New Jersey, Wiley and Sons, Google Scholar 2013

43 Molina V, Reig S, Sanz J and others. : During treatment with atypical antipsychotics in schizophrenia, the volume of gray matter in the cortex increases and the volume of white matter decreases. Schizophrenia Research 2005; 80:61–71 Crossref, Medline, Google Scholar

44 Blade LD, Mouchlianitis E, Wood TC, etc. : White matter changes in the treatment of refractory schizophrenia: are cognitive control and myelination important? Neuroimaging Clinic 2018; 18:186–191 Crossref, Medline, Google Scholar

45 Holleran L, Ahmed M, Anderson-Schmidt H, etc. : Changes in the microstructure and organization of the white matter between the hemispheres and the temporal lobe in severe chronic schizophrenia. Neuropsychopharmacology 2014; 39:944–954 Crossref, Medline, Google Scholar

46 Pérez-Iglesias R, Tordesillas-Gutiérrez D, McGuire PK, etc. : White matter integrity and cognitive impairment in first episode psychosis. Am J Psychiatry 2010; 167:451–458Link, Google Scholar

47 Phillips KA, Rogers J, Barrett EA, etc. : The genetic contribution of the mid-sagittal region of the corpus callosum. Twin Res Hum Genet 2012; 15:315–323 Crossref, Medline, Google Scholar

48 Weinberger DR: On the rationality of the "neurodevelopmental hypothesis" in schizophrenia. Neuropsychopharmacology 1996; 14(Suppl):1S-11SCrossref, Medline, Google Scholar

49 Weinberger DR, Lipska BK: Cortical dysplasia, antipsychotics, and schizophrenia: looking for common ground. Schizophrenia Research 1995; 16:87–110 Crossref, Medline, Google Scholar

50 Shamy JL, Carpenter DM, Fong SG, etc. : Changes in white matter tracts after neurotoxic hippocampal lesions in rhesus monkeys: a study of diffusion tensor imaging. Haima 2010; 20:906–910 Medline, Google Scholar

51 Yu J, Lee TMC, Frontotemporal Degeneration Neuroimaging I; Frontotemporal Degeneration Neuroimaging Initiative: Behavioral Variation: Longitudinal decline in white matter microstructure integrity in frontotemporal dementia and its association with executive function. Neurobiol Aging 2019; 76:62–70 Crossref, Medline, Google Scholar

52 Manschreck TC, Boshes RA: CATIE Schizophrenia Trial: Results, Impact, Controversy. Harv Rev Psychiatry 2007; 15:245–258 Crossref, Medline, Google Scholar

53 Mishara AL, Goldberg TE: A meta-analysis and critical review of the effects of traditional antipsychotic treatment on cognition in schizophrenia: opening a closed book. Biological Psychiatry 2004; 55:1013–1022 Crossref, Medline, Google Scholar

54 Kochunov P, Rowland LM, Fieremans E, etc. : Diffusion-weighted imaging reveals possible sources of schizophrenia processing speed defects. Proc Natl Acad Sci USA 2016; 113:13504–13509 Crossref, Medline, Google Scholar