Abstract
Chronic low-grade peripheral and central nervous system inflammation may have a role in the pathogenesis of schizophrenia (SCZ). Inhibition of cyclooxygenase-2 (COX2), the arachidonic acid pathway, may inhibit cytokine responses and minimize inflammation. In this study, we added the COX2 inhibitor celecoxib to risperidone monotherapy to examine its efficacy on clinical symptoms and cognitive deficits in drug-naïve first episode (DNFE) SCZ patients. First, we genotyped two polymorphisms (rs5275 and rs689466) in the COX-2 gene in a case-control study of 353 SCZ patients and 422 healthy controls. Ninety patients participated in a 12-week, double-blind, randomized, placebo-controlled trial of celecoxib 400 mg/day. We used the Positive and Negative Syndrome Scale (PANSS) and the Repeatable Battery for the Assessment of Neuropsychological Status (RBANS) to assess clinical symptoms and cognition. Our results show that the COX2 rs5275 polymorphism was significantly correlated with SCZ and positive symptoms. After 12-week treatment, celecoxib significantly improved the PANSS total and three subscale scores of SCZ patients. Furthermore, patients with the rs5275 TT genotype had greater improvement in PANSS total score than patients carrying the C allele. However, no significant difference in RBANS total and subscale scores existed between the celecoxib and placebo groups at week 12. Our findings suggest that COX2 inhibitors may be promising therapeutics for clinical symptoms rather than cognitive impairment in first episode SCZ patients. COX2 rs5275 gene polymorphism may be implicated in the development and the efficacy of treating clinical symptoms in SCZ.
Trial Registration Number: The trial was registered with www.clinicaltrials.gov (NCT00686140).
Subject terms: Pharmacology, Psychology, Outcomes research
Introduction
Schizophrenia (SCZ) is a severe psychiatric disorder that affects roughly 1% of the world’s population [1]. Although many antipsychotics have been developed and used widely in clinical practice, approximately 50% of SCZ patients do not obtain adequate relief with antipsychotics, with some 30% still exhibiting markedly disturbed behavior [2]. In addition, all the antipsychotic drugs cause side effects to a greater or lesser extent, which often leads to patient noncompliance with treatment [3]. Clinically, cognitive impairment in SCZ patients is receiving increasing attention because it is a core symptom of SCZ and it has a significant impact on occupational functioning, social functioning, and independent living of SCZ patients, which is detrimental to their long-term recovery [4]. Several neurocognitive domains including attention, memory, executive function, working memory and social cognition have been shown to be impaired in SCZ patients [5]. Currently, most antipsychotics are effective in improving clinical symptoms of SCZ, especially positive symptoms, but not cognitive impairment [6]. Therefore, cognitive impairment in patients with SCZ is a potential therapeutic target.
In the past decade, there has been renewed interest in immune and inflammatory abnormalities associated with the pathophysiology of SCZ [7–12], and at least a subgroup of SCZ patients have increased proinflammatory markers such as cytokines in their blood and cerebrospinal fluid [13, 14]. Moreover, these increased proinflammatory markers have been reported to be associated with clinical symptoms and cognitive abnormalities in SCZ patients [15]. Furthermore, in SCZ patients, neuroimaging studies demonstrate central nervous system volume loss and altered microglial cell activity [16]. These findings suggest that chronic low-level neuroinflammatory processes, occurring in the peripheral and central nervous systems of a subset of SCZ individuals, are accompanied by alterations in dopaminergic, serotonergic, glutamatergic, and noradrenergic neurotransmission [17]. In addition, the strongest genetic signals including GWAS results for SCZ occur in several immune-related genes and on the 6p22.1 chromosome, a region linked to the major histocompatibility complex and other immunological processes [18]. Together, these findings indicate that immune system abnormalities are implicated in the pathogenesis and psychopathology of SCZ [19].
Based on this hypothesis of immune dysfunction and low-grade inflammatory processes in SCZ, anti-inflammatory agents may be theoretically re-purposed to treat SCZ patients, and several clinical trials have tested this hypothesis [20–22]. A double-blind, randomized, placebo-controlled study using 1000 mg/day aspirin as an adjuvant to regular antipsychotic treatment reduced the symptoms of SCZ spectrum disorders [23]. One recent review paper indicated that aspirin has both protective and therapeutic effects on SCZ [24], but another meta-analysis found no statistically significant improvement in the symptoms of SCZ from adjunctive aspirin therapy [25]. Another anti-inflammatory agent, minocycline as an add-on agent showed improvement in negative symptoms and executive functioning [26–28], but a more recent study found very little promise for minocycline as an adjunctive treatment [29].
Another anti-inflammatory agent targets cyclooxygenase 2 (COX2), a crucial enzyme in the arachidonic acid biosynthetic pathway that regulates pain and inflammation, leading to the release of pro-inflammatory cytokines [30]. COX-2 activation exacerbates inflammation and causes widespread apoptosis and neuronal death, while inhibition of COX-2 dampens cytokine response and minimizes inflammation [31]. Celecoxib is a COX2 inhibitor that has been studied in 4 randomized, double-blind, placebo-controlled clinical trials as an antipsychotic add-on [32–35]. Three of these studies showed benefits in PANSS total score and all subscale scores in patients with chronic, early and acutely deteriorating SCZ [32, 33, 35]. However, one study showed no benefit in continuously ill outpatients with SCZ [34, 36]. In addition, studies have found that COX-2 inhibitors have shown beneficial effects in diseases related to cognitive impairment, such as Alzheimer’s disease [37]. A previous study has shown that celecoxib has been used as a adjunctive agent to risperidone and has shown beneficial effects in the treatment of cognitive impairment in SCZ patients [38], suggesting that the add-on treatment with celecoxib may be an effective strategy for treating cognitive impairment in SCZ patients. A recent animal study has shown that celecoxib can improve the cognitive dysfunction of ischemic rats by inhibiting inflammation because it can inhibit COX-2 [39].
The COX-2 gene, which spans 8.3 kp and has 10 exons, is found at 1q25.2-q25.3, and several COX-2 single nucleotide polymorphisms (SNPs) are functional, including 8473C/T (rs5275) in the 3’untranslated region (3’UTR) and −1195A>G (rs689466) in the promoter region [40–42]. The rs5275 influences COX-2 protein levels and mRNA stability [43], and rs689466 G allele transcriptionally activates COX-2 [42, 43]. These polymorphisms are correlated with a variety of diseases involving COX-2, including bronchial asthma, cerebral infarction, stroke and several cancers [40, 42]. However, there have been no studies to explore the link between COX-2 gene polymorphisms and psychiatric disorders.
To our best knowledge, no studies have examined the efficacy of celecoxib as an add-on treatment for drug-naïve first episode (DNFE) patients with SCZ on clinical symptoms and cognitive impairment, nor have they explored whether COX-2 gene polymorphisms (rs5275 and rs689466) would be associated with susceptibility to SCZ or influence response to celecoxib. Studies of DNFE patients have been particularly beneficial in understanding the psychopharmacological mechanisms of SCZ because of minimal confounding factors such as disease duration, the effects of previous medication, and medical comorbidity due to disease chronicity. Therefore, the current study investigated the potential effects and safety of celecoxib on patients with DNFE SCZ in a randomized, double-blind, placebo-controlled trial. We hypothesized that SCZ patients treated with celecoxib would have significantly improved clinical symptoms and cognitive impairment compared to patients treated with placebo, and that COX-2 gene polymorphisms (rs5275 and rs689466) would be pharmacogenetically associated with this treatment response.
Methods
Subjects and setting
We recruited 353 DNFE patients (111 men and 129 women) from Beijing Hui-Long-Guan Hospital. To establish a DSM-IV diagnosis of SCZ, two experienced psychiatrists employed the Chinese version of the Structured Clinical Interview for DSM-IV (SCID) to assess each patient at baseline and at the end of treatment. The patients had to meet the following criteria: (1) between the ages of 18 and 45 years, (2) having acute episode at admission, (3) a duration of symptoms ≤60 months, and (4) not receiving any previous antipsychotic treatment. The patients were 27.0 ± 8.8 years on average and had a mean education of 11.1 ± 6.6 years.
We recruited 422 healthy controls (169 men and 253 women) by placing advertisements in the local community. All healthy controls were interviewed by trained investigators supervised a research psychiatrist. Using SCID, a research psychiatrist assessed their mental health and asked if they had a personal or family history of mental illness. This study excluded participants with Axis I disorders or a personal or family history of psychiatric disorders.
Participants were first recruited in June 2008. All of the participants were Han Chinese from Beijing who underwent a full physical examination and laboratory tests. Participants with clinically significant illness or laboratory abnormalities were ruled out. Also, the participants were excluded if they 1) had persistent infections, allergies, or a history of autoimmune disease; 2) used any immune related agents including corticosteroids and non-steroidal anti-inflammatory drugs within the past 12 weeks before the start of the study; 3) abused or were dependent on illicit drugs or alcohol; 4) were pregnant or breastfeeding, and (5) were unable to provide signed consent forms.
Each individual signed a written informed consent form prior to participating in the trial and after they had been provided with a detailed explanation of the study. The trial was registered with www.clinicaltrials.gov (NCT00686140). The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national and institutional committees on human experimentation and with the Helsinki Declaration of 1975, as revised in 2008. All procedures involving human subjects/patients were approved by the Ethical Board of Beijing Hui-Long-Guan hospital (Approval number: SCH-A01).
Clinical treatment and clinical ratings
Of the 353 patients, we randomly selected 100 patients, 90 of whom were included in the celecoxib clinical trial by signing informed consent for both the cross-sectional study and the clinical trial, whereas those who only signed informed consent for the cross-sectional study were excluded. These 90 patients were all inpatients. They were randomly and double-blindly assigned to either capsulized celecoxib (400 mg/day) or an identically capsulized placebo as an add-on medication to a fixed dose of risperidone for 12 weeks. The flow chart of the study is shown in Figure S1. Patients reached the prescribed risperidone dose of 4–6 mg/day by the end of the first week. Celecoxib was given at a dose of 200 mg per day for the first week, then 400 mg per day (2 times daily) from week 2 until the end of the trial. Both celecoxib and placebo were provided by the Pfizer Pharmaceuticals LLC, China. Concurrent use of psychotropic drugs was not permitted during the trial, with the exception of chloral hydrate or lorazepam for insomnia, and benzhexol hydrochloride for extrapyramidal parkinsonian symptoms, when blinded psychiatrists determined that need.
Randomization
Patients were randomly assigned to the active or placebo groups by an independent third party using basic computer-generated randomization. The treatment assignment was kept blind from both the investigators and the subjects.
Assessment and outcomes
The baseline assessments included demographics, comprehensive medical histories and laboratory examinations, including liver and renal function, routine bloods and ECG. The primary outcome measures were symptom severity on the Positive and Negative Syndrome Scale (PANSS) scores, which includes sub-scales consisting of the PANSS positive symptom subscale, negative symptom subscale, general psychopathology subscale, as well as the percentage of responders calculated based on the PANSS total score [44]. The secondary outcome measure was cognitive function on the Repeatable Battery for the Assessment of Neuropsychological Status (RBANS) [45]. It consists of 12 subtests from which 5 age-adjusted index scores and a total score were generated [45], including attention, language, visuospatial/ constructional, immediate memory, and delayed memory. Our research team translated RBANS into Chinese, and demonstrated clinical validity and tested test—retest reliability in SCZ patients and controls [46]. The other measure included the Treatment Emergent Symptom Scale (TESS) for side events, the Simpson Scale for Extrapyramidal Side Effects (SEPS) for extrapyramidal side effects and the Abnormal Involuntary Movement Scale (AIMS) for tardive dyskinesia.
Prior to the study, 4 research psychiatrists who had been in clinical practice for at least 5 years underwent concurrent training courses on the use of these rating scales to ensure consistency and reliability of ratings. After repeated assessments, their inter-rater correlation coefficients (ICC) for all of these rating scales remained over 0.8.
The PANSS scores were obtained 7 times: at baseline and every two weeks. The TESS, SEPS and AIMS scales were obtained at baseline and weeks 4, 8 and 12. The RBANS was measured at baseline and at week 12, totaling 2 times.
Genotyping
The anticoagulant ethylenediaminetetraacetic acid (EDTA) was used to store genomic DNA, which was then extracted using salting-out extraction method and kept at −80 degrees. Based on prior research [40–42], we chose 2 tag SNPs in the COX-2 gene, −8473 T>C (rs5275) and 1195A>G (rs689466).
According to the usual protocol outlined in our previous study [47], these two polymorphisms were genotyped using the Matrix-Assisted Laser Desorption/Ionization Time of Flight Mass Spectrometry (MALDI-TOF MS) (Sequenom Inc., San Diego, CA, USA). The rs5275 primers were sense: 5′-TATGTGATGGACTCACCAGGT-3′, antisense: 5′-CCTCTACATGGCCCTGTCTT-3′. The rs689466 primers were sense: 5′-TGTGACCACAGCAATGG GTAGGAGA-3′, antisense: 5′-CCCAGTGTGTGGCCA TATCTTCTTA-3′.
In terms of quality control, 5% of the samples were chosen at random for repeated genotyping, with a repeat genotyping error rate of less than 0.1 percent.
Statistical analyses
Initial analysis included all subjects. Differences between groups were assessed by using χ2 tests for categorical variables and analysis of variance (ANOVA) for continuous variables. The χ2 test was used to compare allele and genotype frequencies of COX-2 gene polymorphisms between healthy controls and SCZ patients. A χ2 goodness-of-fit test was used to assess deviations from Hardy–Weinberg equilibrium (HWE) in healthy controls and patients.
The changes in the PANSS total and its subscale scores (baseline minus week 12) were calculated for each patient in the celecoxib and placebo groups. Further, we used ANOVA to analyze the mean reduction in total and subscale PANSS scores between the two groups. A patient was considered to have a clinical response if the PANSS total score reduced by 50% to quantify the difference between the two groups. The χ2 test was then used to compare the proportions of patients in each group who had a clinical response. Further, multiple linear regression was utilized to examine potential predictors of response linked to PANSS total score changes during the 12 weeks of treatment.
We utilized RM ANCOVA to analyze the significant interactions of genotype grouping with changes in PANSS total and subscale scores, using baseline and biweekly values as dependent variables and demographic and clinical variables as covariates, to explore the impact of COX-2 genotype on clinical symptom improvement. Treatment response rates were compared using logistic regression analysis across treatment groups and genotypes. In the logistic analysis at week 12, we entered age, sex and PANSS total and subscale scores at baseline in the first step, treatment group in the second step, and COX-2 genotype and genotype x treatment group interaction terms in the third step. All statistical tests were declared statistically significant if the p value was less than 0.05.
Power calculations and effect size determinations were performed using the G*Power 3.1.9.2 program with a significance level of 0.05, and a power of 0.80. The first study focused on the effect of genetic polymorphisms, and the sample size required for a medium effect size (ρ = 0.3) was108. In this study, we recruited 353 patients and 422 healthy controls, which met the requirement for an appropriate sample size. Study 2 focused on a randomized, double-blind, placebo-controlled trial and the sample size required for a medium effect size (ρ = 0.25) was 66. In this study, we recruited 42 patients in the celecoxib group and 38 patients in the placebo group, which met the requirement for an appropriate sample size.
Results
Demographic data and cognition of healthy controls and SCZ patients
Table 1 exhibits the demographic data comparing SCZ patients to healthy controls. We found significant differences between the two groups in terms of age (F = 407.48, df = 1, 773, p < 0.001), sex (x2 = 8.83, df = 1, p < 0.01), education (F = 12.13, df = 1,770, p < 0.01), smoking status (x2 = 5.81, df = 1, p < 0.05) and BMI (F = 146.87, df = 1,758, p < 0.001), which were controlled for in the following analysis.
Table 1.
Demographic characteristics, cognition and allele and genotype distributions of two polymorphisms in the COX-2 between patients with schizophrenia and healthy controls.
| Characteristics | Schizophrenia (n = 353) | Healthy Controls (n = 422) | F/χ2 | p |
|---|---|---|---|---|
| Males/Females | 179/174 | 169/253 | 8.83 | <0.01 |
| Age (years) | 27.0 ± 8.8 | 44.6 ± 13.6 | 407.48 | <0.001 |
| Education (years) | 11.1 ± 6.6 | 9.7 ± 5.5 | 12.13 | <0.01 |
| Smoking (%) | 27.3% | 36.5% | 5.81 | <0.05 |
| BMI (kg/m2) | 21.6 ± 3.6 | 25.1 ± 4.2 | 146.87 | <0.001 |
| RBANS | Schizophrenia (n = 178) | Healthy Controls (n = 392) | ||
|---|---|---|---|---|
| Immediate memory | 70.5 ± 16.9 | 75.8 ± 17.6 | 12.28 | <0.001 |
| Attention | 82.7 ± 15.6 | 87.0 ± 18.8 | 7.87 | <0.005 |
| Language | 87.8 ± 13.1 | 93.9 ± 13.1 | 26.70 | <0.001 |
| Visuospatial/Constructional | 85.4 ± 19.0 | 87.5 ± 20.4 | 1.31 | >0.05 |
| Delayed memory | 73.7 ± 20.3 | 86.3 ± 15.3 | 66.65 | <0.001 |
| Total | 75.1 ± 16.3 | 80.1 ± 15.2 | 12.89 | <0.001 |
| COX-2 rs5275 | ||||
| Allele frequency (%) | 5.22 | 0.022 | ||
| T | 584 (83.7%) | 666 (79.1%) | ||
| C | 114 (16.3%) | 176 (20.9%) | ||
| Genotype frequency (%) | 7.16 | 0.028 | ||
| TT | 252 (72.2%) | 267 (63.4%) | ||
| TC | 80 (22.9%) | 132 (31.4%) | ||
| CC | 17 (4.9%) | 22 (5.2%) | ||
| COX-2 rs689466 | ||||
| Allele frequency (%) | 1.46 | 0.23 | ||
| A | 361 (51.4%) | 460 (54.5%) | ||
| G | 341 (48.6%) | 384 (45.5%) | ||
| Genotype frequency (%) | 2.08 | 0.35 | ||
| AA | 90 (25.6%) | 128 (30.3%) | ||
| GA | 181 (51.6%) | 204 (48.3%) | ||
| GG | 80 (22.8%) | 90 (21.3%) | ||
Values are presented as mean ± SD.
BMI body mass index, RBANS Repeatable Battery for the Assessment of Neuropsychological Status, COX-2 cyclooxygenase-2.
The RBANS total and index scores for 178 SCZ patients and 392 healthy controls are shown in Table 1. Except for the visuospatial/constructional index (p > 0.05), patients performed considerably worse on the RBANS total score and the four subscales when compared to healthy controls (all p < 0.001; all Bonferroni-corrected p < 0.01). After accounting for gender, age, education, smoking and BMI, the significant differences persisted (all p < 0.01).
Allele and genotype frequencies of COX-2 rs5275 and COX-2 rs689466 in healthy controls and SCZ patients
Table 1 displays allele and genotype frequencies, and statistical analysis. The genotypic distributions of COX-2 rs5275 or COX-2 rs689466 in control and SCZ patient groups were compatible with Hardy–Weinberg equilibrium (all p > 0.05).
In the case of rs5275, we found substantial variations in genotype and allele frequencies between healthy controls and SCZ patients (allele X2 = 5.22, p = 0.022; genotype X2 = 7.16, p = 0.028), showing that the frequency of the T allele (major allele) of rs5275 was higher in SCZ than in healthy controls (Table 1). After accounting for age, gender, education, smoking and BMI, a substantial difference in rs5275 genotype frequency between healthy controls and SCZ patients persisted (X2 = 3.91, p = 0.048; OR = 1.37, 95% CI: 1.003–1.857). In the case of rs689466, we did not find significant differences in genotype and allele frequencies between healthy controls and SCZ patients (Table 1).
In addition, 345 SCZ patients completed PANSS test. We observed a significant effect of rs5275 genotype on PANSS-positive symptom scores (TT = 23.6 ± 6.8 vs TC = 22.5 ± 6.2 vs CC = 27.3 ± 9.5; F = 3.24, df = 2, p = 0.04; Bonferroni corrected p > 0.05), but no effect of rs689466.
Clinical trial of celecoxib
Demographic and basic descriptive data
Of the 353 DNFE patients, 90 SCZ patients in this trial and randomly assigned to either the celecoxib (n = 46) or placebo (n = 44) treatment group. These selected 90 patients were similar to the larger group of 353 patients in terms of sociodemographic and clinical characteristics (all p > 0.05).
Further, the two treatment groups showed no statistical differences at baseline in sociodemographic or clinical variables (Table 2). The 10 study dropouts included 4 on celecoxib and 6 on placebo, who were considered non-responders.
Table 2.
Comparison of PANSS and RBANS total and subscale scores before and after treatment in the celecoxib and placebo group.
| Variables | Baseline | Week 2 | Week 4 | Week 6 | Week 8 | Week 10 | Week 12 | Group F (p) | Time F(p) | Group * Time F(p) |
|---|---|---|---|---|---|---|---|---|---|---|
| PANSS total score | ||||||||||
| Celecoxib (n = 42) | 84.3 ± 18.2 | 73.1 ± 18.3 | 61.5 ± 17.8 | 55.2 ± 16.8 | 48.9 ± 13.0 | 43.0 ± 10.9 | 40.2 ± 11.0 | 0.93 | 242.78 | 7.38 |
| Placebo (n = 38) | 82.6 ± 20.2 | 71.7 ± 19.2 | 60.2 ± 18.3 | 54.8 ± 18.2 | 52.3 ± 15.5 | 50.7 ± 14.7a** | 50.3 ± 14.7a*** | (0.34) | (<0.0001) | (0.008) |
| Positive symptom | ||||||||||
| Celecoxib (n = 42) | 26.2 ± 6.5 | 20.6 ± 7.2 | 16.3 ± 6.9 | 14.1 ± 6.4 | 11.9 ± 5.3 | 10.1 ± 4.9 | 9.3 ± 5.1 | 0.11 | 157.95 | 2.45 |
| Placebo (n = 38) | 24.8 ± 5.8 | 20.1 ± 6.4 | 15.2 ± 7.1 | 12.9 ± 6.0 | 11.7 ± 5.2 | 11.5 ± 5.0a* | 11.5 ± 5.2a** | (0.75) | (<0.0001) | (0.024) |
| Negative symptom | ||||||||||
| Celecoxib (n = 42) | 18.6 ± 6.5 | 17.0 ± 6.9 | 15.2 ± 6.0 | 13.6 ± 4.8 | 13.2 ± 4.3 | 11.0 ± 4.6 | 10.5 ± 3.8 | 1.93 | 49.90 | 6.68 |
| Placebo (n = 38) | 18.5 ± 8.2 | 17.4 ± 7.2 | 15.6 ± 6.7 | 15.6 ± 7.3 | 15.2 ± 6.4a* | 14.4 ± 6.0a** | 14.0 ± 5.7a** | (0.17) | (<0.0001) | (0.012) |
| General psychopathology | ||||||||||
| Celecoxib (n = 42) | 39.6 ± 11.3 | 35.5 ± 10.2 | 30.0 ± 10.4 | 27.7 ± 8.9 | 24.9 ± 7.5 | 21.9 ± 4.8 | 20.5 ± 4.8 | 0.34 | 157.77 | 3.67 |
| Placebo (n = 38) | 39.4 ± 11.7 | 34.3 ± 10.7 | 29.3 ± 8.6 | 26.7 ± 7.3 | 25.3 ± 7.2 | 24.9 ± 6.5a* | 24.9 ± 6.7a** | (0.56) | (<0.0001) | (0.059) |
| RBANS total score | ||||||||||
| Celecoxib (n = 40) | 74.6 ± 16.6 | – | – | – | – | – | 82.4 ± 13.6 | 0.23 | 27.56 | 1.85 |
| Placebo (n = 35) | 77.7 ± 17.7 | – | – | – | – | – | 82.4 ± 16.0 | (0.64) | (<0.0001) | (0.18) |
| Immediate memory | ||||||||||
| Celecoxib (n = 40) | 69.5 ± 18.5 | – | – | – | – | – | 76.9 ± 15.7 | 1.11 | 10.55 | 1.51 |
| Placebo (n = 35) | 74.7 ± 19.1 | – | – | – | – | – | 78.0 ± 19.6 | (0.30) | (0.002) | (0.22) |
| Attention | ||||||||||
| Celecoxib (n = 40) | 85.0 ± 17.3 | – | – | – | – | – | 91.0 ± 15.8 | 0.15 | 12.67 | 0.11 |
| Placebo (n = 35) | 84.1 ± 22.3 | – | – | – | – | – | 89.3 ± 16.9 | (0.70) | (0.001) | (0.74) |
| Language | ||||||||||
| Celecoxib (n = 40) | 89.3 ± 12.6 | – | – | – | – | – | 93.0 ± 12.5 | 2.80 | 2.77 | 0.47 |
| Placebo (n = 35) | 86.0 ± 14.7 | – | – | – | – | – | 87.6 ± 12.4 | (0.09) | (<0.10) | (0.49) |
| Visuospatial/Constructional | ||||||||||
| Celecoxib (n = 40) | 86.7 ± 17.1 | – | – | – | – | – | 88.7 ± 15.4 | 0.07 | 1.83 | 0.24 |
| Placebo (n = 35) | 88.4 ± 22.5 | – | – | – | – | – | 90.0 ± 19.6 | (0.80) | (0.18) | (0.63) |
| Delayed memory | ||||||||||
| Celecoxib (n = 40) | 69.9 ± 23.1 | – | – | – | – | – | 83.4 ± 18.4 | 1.24 | 39.57 | 1.13 |
| Placebo (n = 35) | 76.4 ± 18.9 | – | – | – | – | – | 85.9 ± 19.2 | (0.27) | (<0.0001) | (0.29) |
PANSS Positive and Negative Syndrome Scale, RBANS Repeatable Battery for the Assessment of Neuropsychological Status.
*P < 0.05; **P < 0.01; ***P < 0.001.
aComparison between pretreatment and post-treatment.
Primary outcome
Figure 1A shows that celecoxib had a highly significant therapeutic effect at 12 weeks of treatment based on PANSS scores. As shown in Table 2, RM MANOVA on PANSS total score revealed a significant group × time effect (F = 7.38, p = 0.008), and a significant time effect (F = 242.78, p < 0.0001). ANCOVA revealed that the celecoxib group had a significantly lower PANSS total score than the placebo group at weeks 10 (F = 9.92, p = 0.002, effect size = 0.602) and week 12 (F = 15.54, p < 0.0001, effect size = 0.786).
Fig. 1. Effect of 12 weeks of celecoxib treatment on clinical symptoms and cognitive function.
A Comparison of PANSS scores in the celecoxib group (n = 42) and placebo group (n = 38). At the end of 12-week treatment, celecoxib significantly improved the PANSS total and three subscale scores of SCZ patients compared to placebo. B Comparison of RBANS scores in the celecoxib group (n = 40) and placebo group (n = 35). There were no significant difference in RBANS total and subscale scores between the celecoxib and placebo groups at week 12. * indicates comparison between celecoxib and placebo groups at different time point. *p < 0.05; **p < 0.01; ***p < 0.001.
Similarly, RM MANOVA on the PANSS positive symptom subscore showed a significant group-by-time effect (F = 2.45, p = 0.024) and a significant time effect (F = 157.95, p < 0.0001). ANCOVA demonstrated that the celecoxib group had a significantly lower PANSS positive symptom score than the placebo group at week 10 (F = 4.81, p = 0.032, effect size = 0.283) and week 12 (F = 8.99, p = 0.004, effect size = 0.427).
RM MANOVA on the PANSS negative symptom subscore showed a significant group-by-time effect (F = 6.68, p = 0.012) and a significant time effect (F = 49.9, p < 0.0001). ANCOVA revealed that the celecoxib group had a significantly lower PANSS negative symptom score than the placebo group at week 8 (F = 7.80, p = 0.006, effect size = 0.374), week 10 (F = 9.05, p = 0.004, effect size = 0.630) and week 12 (F = 11.65, p = 0.001, effect size = 0.737).
RM MANOVA on the PANSS general psychopathology subscore showed a significant trend towards group-by-time effect (F = 3.67, p = 0.059) and a significant time effect (F = 157.77, p < 0.0001). ANCOVA demonstrated that the celecoxib group had a significantly lower PANSS general psychopathology score than the placebo group at week 10 (F = 6.92, p = 0.01, effect size = 0.531) and week 12 (F = 13.78, p < 0.0001, effect size = 0.765).
In addition, there was also a significant difference in the decrease in PANSS total score from baseline to week 12 between the celecoxib and placebo groups (45.81 ± 20.4 versus 32.75 ± 21.02; F = 7.72, p = 0.007, effect size = 0.631).
Further, to evaluate the treatment effect, we calculated the percentage of responders. At week 12, 28 of 42 patients (66.7%) treated with celecoxib and 10 of 38 patients (26.3%) treated with placebo had more than 50% improvement in their total PANSS score (Χ2 = 13.03, df = 1, p < 0.001).
Secondary outcome
The changes in RBANS scores at baseline and during the 12-week treatment are shown in Fig. 1B. As shown in Table 2, RM ANOVA revealed significant time effects for RBANS total score and domain scores (all p < 0.01) between baseline and follow-up, except for language and visuospatial/constructional scores (both p > 0.05). However, there was no significant group-by-time effect or group effect on RBANS total score and its domain scores (all p > 0.05). Moreover, ANCOVA revealed no any significant difference in RBANS scores at week 12 (all p > 0.05). These findings indicated that celecoxib treatment may not improve cognitive deficits occurring in SCZ patients.
Side effects
Systemic adverse effects on routine physical examinations and laboratory tests were minor in both the celecoxib and placebo groups, as Supplementary Table S1 shows. Neither group reported serious adverse events, and no patient was withdrawn due to serious side effects. During the 12-week study, 3 celecoxib patients and 2 placebo patients were withdrawn due to symptom recovery and hospital discharge. One celecoxib patient and 4 placebo patients were withdrawn because they asked to switch from risperidone to another treatment. In addition, Table 3 shows no significant differences in the total scores of TESS, AIMS and SEPS at weeks 4, 8 and 12 (all p > 0.05). However, bodyweight gain was significantly greater on celecoxib than on placebo, which related to a lower baseline BMI and weight of 61 ± 12 kg for celecoxib versus 64 ± 12 kg for placebo patients. At week 12, the two groups increased to 65 ± 11 kg versus 65 ± 10 kg, respectively (F = 6.67, p = 0.012).
Table 3.
Comparison of mean scores at baseline, week 4, and week 12 on side effects measured by TESS, AIMS and SEPS total scores in the Celecoxib group and placebo group.
| Variables | Week 4 | Week 8 | Week 12 |
|---|---|---|---|
| TESS total score | |||
| Celecoxib (n = 42) | 4.8 ± 3.9 | 4.3 ± 3.8 | 4.1 ± 3.9 |
| Placebo (n = 38) | 4.7 ± 3.8 | 4.8 ± 4.2 | 5.9 ± 4.9 |
| AIMS total score | |||
| Celecoxib (n = 42) | 0.2 ± 0.9 | 0.1 ± 0.3 | 0.1 ± 0.3 |
| Placebo (n = 38) | 0.1 ± 0.9 | 0.1 ± 0.5 | 0.1 ± 0.4 |
| SEPS total score | |||
| Celecoxib (n = 42) | 1.5 ± 2.9 | 1.1 ± 2.2 | 0.6 ± 1.3 |
| Placebo (n = 38) | 1.2 ± 1.7 | 1.3 ± 2.4 | 1.3 ± 2.3 |
There were no significant differences in TESS, AIMS or SEPT total score at week 4, week 8 and week 12 between celecoxib group and placebo group (all p > 0.05).
TESS Treatment Emergent Symptom Scale, SEPS Simpson Scale for Extrapyramidal Side Effects, AIMS Abnormal Involuntary Movement Scale.
In addition, among the demographic and clinical variables multivariate linear regression revealed that only the PANSS at baseline (beta = 0.76, t = 6.08, p < 0.001) and education (beta = 0.23, t = 3.03, p = 0.003) were positively correlated with the PANSS improvement on celecoxib (n = 42).
Effect of COX-2 genotypes on treatment outcomes
As mentioned earlier, the number of celecoxib responders far exceeded the number of placebo responders (66.7% vs. 26.3%). Further, allele frequencies and genotypes of the COX-2 rs5275 or rs689466 polymorphisms were balanced between the celecoxib and placebo groups.
A significant interaction between rs5275 genotype and treatment improvement was found in a logistic regression model of response, with an adjusted odds ratio of 2.31 (95% confidence interval 1.252 ~ 4.264; x2 = 7.17; p = 0.007). As shown in Table 4, the significant rs5275 genotype* treatment interaction indicated that the COX-2 rs5275 TT group (21/30 = 70% vs. 5/25 = 20%) had a greater likelihood of clinical response to celecoxib treatment compared to the C allele carrier group. In addition, COX-2 rs689466 had no genotype effects on clinical response.
Table 4.
Logistic regression models of treatment response at 12-week follow-up.
| Predictor | OR (95% CI) | p |
|---|---|---|
| Sex | 0.873 (0.327–2.332) | 0.79 |
| Age | 0.967 (0.915–1.022) | 0.24 |
| Education | 1.01 (0.882–1.160) | 0.84 |
| COX-2 rs5257 genotype | 1.84 (0.562–6.019) | 0.31 |
| Treatment | 6.11 (2.187–17.090) | 0.001 |
| rs5257* Treatment | 2.31 (1.252–4.262) | 0.007 |
COX-2 cyclooxygenase-2, CI confidence interval, OR odds ratio.
Discussion
We had several main findings in this study. First, the COX-2 gene rs5275 was linked to genetic predisposition to SCZ in a Chinese population. Second, celecoxib (400 mg/day) added on risperidone treatment significantly improved the clinical symptoms of first-episode SCZ patients, without having severe side effects, suggesting that celecoxib is a safe and efficient medication for SCZ patients. However, celecoxib did not appear to have a beneficial effect on cognitive impairment in SCZ patients. Third, we observed a better clinical response to celecoxib treatment in the COX-2 rs5275 TT group than in the C allele carrier group. This is the first study that we are aware of that shows a possible COX-2 gene pharmacogenetic effect on clinical symptom improvement in patients with DNFE SCZ treated with celecoxib added to risperidone, although three previous studies have shown the efficacy of celecoxib for significantly augmenting the antipsychotic efficacy [32, 33, 35].
COX-2 gene polymorphism and SCZ susceptibility
In the Chinese population, the wild-type allele T of the COX-2 rs5275 polymorphism, but not rs689466 polymorphism, may be associated with SCZ susceptibility. This COX-2 rs5275 (8473T>C) polymorphism may alter COX-2 protein levels and mRNA through the binding affinity of transcriptional factors or through mRNA stability and/ or translational efficiency [41, 43]. Therefore, we speculated that the COX-2 rs5275 polymorphism may be associated with elevated levels of mRNA and protein expression in SCZ patients [40]. Overexpression of COX-2 is frequently associated with a variety of chronic inflammatory conditions, which may be implicated in an increased risk of SCZ [17, 18]. However, more recent studies have found that this COX-2 rs5275 polymorphism does not affect COX-2 mRNA, protein levels, or enzymatic activity [48]. Therefore, the relationship between the COX-2 rs5275 polymorphism, its mRNA and expression levels and the pathogenesis of SCZ patients still deserves further investigation.
We found no association between the rs689466 polymorphism and SCZ susceptibility. Previous studies in patients with major depressive disorder have found similar results [49]. As suggested by Serretti et al., the rs689466 polymorphism was not associated with antidepressant response, remission or treatment-resistance [49]. Since there is an overlap between depressive symptoms and SCZ negative symptoms [50], it could be argued that the rs689466 polymorphism also fails to account for an association with response to celecoxib treatment, particularly with remission of SCZ negative symptoms.
Efficacy and safety of celecoxib treatment for SCZ
We found a significant reduction in overall clinical symptoms on the PANSS total score with celecoxib. This significant reduction emerged at week 10 for positive symptoms, but earlier at week 8 for negative symptoms. However, after 12 weeks of treatment, there was no significant difference in RBANS ratings between the celecoxib and placebo groups.
To date, efficacy of celecoxib in SCZ patients has been explored in several clinical trials, with mixed results. The first prospective, double-blind and randomized trial using celecoxib plus risperidone showed that celecoxib treatment improved PANSS total scores in patients with acute onset SCZ [32]. However, subsequent trials did not replicate the clinical benefit of celecoxib in chronic SCZ patients [34]. Another trial revealed that the combination of celecoxib and risperidone did show significant superiority in PANSS positive symptom, general psychopathology and total scores in the active phase of SCZ patients compared to risperidone alone [35]. More interestingly, a later trial demonstrated that celecoxib reduced negative symptoms during the early stage of SCZ [33], which our current study confirmed. Improvement in negative symptoms is critical for patients with first-episode SCZ, because negative symptoms prolong the course of the illness and limit the recovery of patients’ social functioning [33]. Compared to healthy controls, SCZ patients have elevated levels of several cytokines, suggesting neuroinflammatory dysfunction in SCZ patients [29, 51]. For example, TNF-α is a multifunctional proinflammatory cytokine with modulatory effects on neuronal cell growth, differentiation, apoptosis, and synaptic function, and several studies have shown a positive correlation between TNF-α and negative symptoms in people at risk for SCZ and in SCZ patients [29, 52, 53], suggesting that excessive TNF-α is associated with the pathomechanism of negative symptoms in SCZ patients. It has also shown that celecoxib downregulates TNF-α [54] and IL-12 [55]. Since celecoxib is a COX-2 inhibitor that modulates immune function by improving the inflammatory microenvironment [54], we hypothesize that celecoxib ameliorates negative symptoms in SCZ patients by suppressing their immune function. Taken together, adjunctive treatment with celecoxib may have a greater therapeutic effect for clinical symptoms in the early stages of SCZ. However, the therapeutic efficacy of celecoxib is controversial in patients with chronic SCZ [36].
Studies have found that COX-2 inhibitors show beneficial effects in disorders associated with cognitive impairment, such as Alzheimer’s disease [37]. However, the present study demonstrated that celecoxib did not have a better effect on cognitive impairment in patients with DNFE SCZ compared to placebo, as measured by RBANS, which is inconsistent with a previous study [38]. This may be due to the insufficient sensitivity of the cognitive function test we used to identify subtle changes in this function. In addition, the patients recruited for this study were first-episode cases, which is significantly different from the previous study [38]. These patients in this study may have suffered from relatively mild cognitive impairment with limited scope for improvement. In the present study, we found significant improvements in cognitive impairment in both the risperidone alone group and the risperidone combined with celecoxib group, implying that there may be a ceiling effect.
Critically, the side effects of celecoxib treatment in this study were unremarkable, which is consistent with previous studies [32–35], and even the increase in bodyweight in the celecoxib group was attributable to a baseline weight and BMI difference from placebo patients. Overall, no patient dropped out from the trial due to side effects.
Association between treatment outcome and COX-2 genotypes
We found that patients with the COX-2 rs5275 TT genotype displayed a greater likelihood of clinical response to celecoxib treatment compared to patients carrying the C allele. While the underlying mechanism for the pharmacogenetic effects of celecoxib in treating SCZ is unclear, the rs5275 polymorphism is located where COX-2 mRNA is targeted for microRNA-mediated degradation. miR-542–3p binds to transcripts originated from the rs5275 T allele and inhibits mRNA translation, while the rs5275 C allele variant interferes with the binding of miR-542-3p, thereby stabilizing this mRNA [56]. Hence, COX-2 mRNA and protein levels are expected to be lower in T homozygous carriers than in their C allele counterparts. Those SCZ TT genotype patients who have these comparatively lower COX-2 levels at baseline will have their COX-2 enzyme more readily inhibited through celecoxib treatment. However, we did not measure the COX-2 mRNA or peripheral COX-2 levels in this study to test this particular hypothesis. Therefore, the mechanism by which celecoxib treatment affects COX-2 levels and how COX-2 rs5275 T/C genotype is associated with clinical outcomes of celecoxib treatment in SCZ deserve further exploration.
In addition to lacking this biochemical assessment for COX-2 activity, this study has several limitations. First, this study had a relatively small sample size and clearly needs to be replicated. In addition, significant differences in demographic factors did exist between schizophrenia patients and healthy population in this study. Although we have incorporated demographic factors as covariates in subsequent statistical analyses, the validity and reliability of the results may be deteriorated. Therefore, in future studies, we need larger samples with matching demographics characteristics to address this issue. Second, the duration of treatment is relatively short, and the anti-inflammatory mechanism of celecoxib in the brain may not be fully demonstrated. Therefore, it would be necessary to investigate the effects of celecoxib on clinical symptoms of SCZ over a longer duration of treatment. Third, only one dose of celecoxib was administered in this study based on previous studies [32, 33, 35], although the therapeutic recommendations vary from 100 mg/day to 800 mg/day according to the different diseases. Greater doses may produce more effects on the clinical symptoms of patients, and the side effect profile provides no contraindication to examining up to 800 mg/day, twice the dose used in this and previous studies. Fourth, there was no specific scale to assess celecoxib-related side effects. In this study, we used TESS, SEPS and AIMS to assess the general side effects that may occur in the use of other medications. In addition, some common celecoxib-related side effects such as heartburn, dysgeusia, bloating and sore throat were not considered, and the association between psychotropic medications and common side effects of celecoxib was not investigated. A comprehensive assessment should be considered in future studies to ensure the validity of the study. Finally, we focused on the two COX-2 polymorphisms because they are functional and have been widely investigated in physical diseases, but other COX-2 genetic variants deserve further investigation. Taken together, additional studies examining the therapeutic benefit of add-on treatment with celecoxib may require larger sample sizes, longer treatment duration, and further examination of the dose-response effect and the role of this specific polymorphism and perhaps other COX-2 polymorphisms.
Supplementary information
Author contributions
DMW and X-YZ were responsible for study design, statistical analysis, and manuscript preparation. DC and MX was responsible for recruiting the patients, performing the clinical rating and collecting the samples. LW and TRK were involved in evolving the ideas and editing the manuscript. DMW and X-YZ were responsible for providing the funding for the study. All authors contributed to and have approved the final manuscript.
Funding
Funding for this study was provided by grants from CAS International Cooperation Research Program (153111KYSB20190004).
Data availability
The data that support the findings of this study are available from the corresponding author, [ZXY], upon reasonable request.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary information
The online version contains supplementary material available at 10.1038/s41386-023-01760-8.
References
- 1.Hogerzeil SJ, Hoek HW, van Hemert AM. The impact of study design on schizophrenia incidence estimates: a systematic review of Northern European studies 2008-2019. Schizophr Res. 2021;231:134–41. doi: 10.1016/j.schres.2021.03.017. [DOI] [PubMed] [Google Scholar]
- 2.Ceraso A, Lin JJ, Schneider-Thoma J, Siafis S, Tardy M, Komossa K, et al. Maintenance treatment with antipsychotic drugs for schizophrenia. Cochrane Database Syst Rev. 2020;8:Cd008016. doi: 10.1002/14651858.CD008016.pub3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Pillinger T, McCutcheon RA, Vano L, Mizuno Y, Arumuham A, Hindley G, et al. Comparative effects of 18 antipsychotics on metabolic function in patients with schizophrenia, predictors of metabolic dysregulation, and association with psychopathology: a systematic review and network meta-analysis. Lancet Psychiatry. 2020;7:64–77. doi: 10.1016/s2215-0366(19)30416-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Zanelli J, Mollon J, Sandin S, Morgan C, Dazzan P, Pilecka I, et al. Cognitive change in schizophrenia and other psychoses in the decade following the first episode. Am J Psychiatry. 2019;176:811–9. doi: 10.1176/appi.ajp.2019.18091088. [DOI] [PubMed] [Google Scholar]
- 5.Harvey PD, Isner EC. Cognition, social cognition, and functional capacity in early-onset schizophrenia. Child Adolesc Psychiatr Clin N Am. 2020;29:171–82. doi: 10.1016/j.chc.2019.08.008. [DOI] [PubMed] [Google Scholar]
- 6.Nielsen RE, Levander S, Kjaersdam Telléus G, Jensen SO, Østergaard Christensen T, Leucht S. Second-generation antipsychotic effect on cognition in patients with schizophrenia-a meta-analysis of randomized clinical trials. Acta Psychiatr Scand. 2015;131:185–96. doi: 10.1111/acps.12374. [DOI] [PubMed] [Google Scholar]
- 7.Müller N. Neuroprogression in schizophrenia and psychotic disorders: the possible role of inflammation. Mod Trends Pharmacopsychiatry. 2017;31:1–9. doi: 10.1159/000470802. [DOI] [PubMed] [Google Scholar]
- 8.Müller N. Inflammation in schizophrenia: pathogenetic aspects and therapeutic considerations. Schizophr Bull. 2018;44:973–82. doi: 10.1093/schbul/sby024. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Comer AL, Carrier M, Tremblay M, Cruz-Martín A. The inflamed brain in schizophrenia: the convergence of genetic and environmental risk factors that lead to uncontrolled neuroinflammation. Front Cell Neurosci. 2020;14:274. doi: 10.3389/fncel.2020.00274. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Fond G, Lançon C, Korchia T, Auquier P, Boyer L. The Role of Inflammation in the Treatment of Schizophrenia. Front Psychiatry. 2020;11:160. doi: 10.3389/fpsyt.2020.00160. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Miller BJ, Goldsmith DR. Evaluating the hypothesis that schizophrenia is an inflammatory disorder. Focus. 2020;18:391–401. doi: 10.1176/appi.focus.20200015. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.van Mierlo HC, Schot A, Boks MPM, de Witte LD. The association between schizophrenia and the immune system: review of the evidence from unbiased ‘omic-studies. Schizophr Res. 2020;217:114–23. doi: 10.1016/j.schres.2019.05.028. [DOI] [PubMed] [Google Scholar]
- 13.Momtazmanesh S, Zare-Shahabadi A, Rezaei N. Cytokine alterations in schizophrenia: an updated review. Front Psychiatry. 2019;10:892. doi: 10.3389/fpsyt.2019.00892. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Reale M, Costantini E, Greig NH. Cytokine imbalance in schizophrenia. From research to clinic: potential implications for treatment. Front Psychiatry. 2021;12:536257. doi: 10.3389/fpsyt.2021.536257. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Wu ZW, Yu HH, Wang X, Guan HY, Xiu MH, Zhang XY. Interrelationships between oxidative stress, cytokines, and psychotic symptoms and executive functions in patients with chronic schizophrenia. Psychosomatic Med. 2021;83:485–91. doi: 10.1097/psy.0000000000000931. [DOI] [PubMed] [Google Scholar]
- 16.Vidal PM, Pacheco R. The cross-talk between the dopaminergic and the immune system involved in schizophrenia. Front Pharmacol. 2020;11:394. doi: 10.3389/fphar.2020.00394. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Trossbach SV, Hecher L, Schafflick D, Deenen R, Popa O, Lautwein T, et al. Dysregulation of a specific immune-related network of genes biologically defines a subset of schizophrenia. Translational psychiatry. 2019;9:156. doi: 10.1038/s41398-019-0486-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Müller N. Immunology of schizophrenia. Neuroimmunomodulation. 2014;21:109–16. doi: 10.1159/000356538. [DOI] [PubMed] [Google Scholar]
- 19.Sperner-Unterweger B, Fuchs D. Schizophrenia and psychoneuroimmunology: an integrative view. Curr Opin Psychiatry. 2015;28:201–6. doi: 10.1097/yco.0000000000000153. [DOI] [PubMed] [Google Scholar]
- 20.Çakici N, van Beveren NJM, Judge-Hundal G, Koola MM, Sommer IEC. An update on the efficacy of anti-inflammatory agents for patients with schizophrenia: a meta-analysis. Psychol Med. 2019;49:2307–19. doi: 10.1017/s0033291719001995. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Cho M, Lee TY, Kwak YB, Yoon YB, Kim M, Kwon JS. Adjunctive use of anti-inflammatory drugs for schizophrenia: a meta-analytic investigation of randomized controlled trials. Aust N Z J Psychiatry. 2019;53:742–59. doi: 10.1177/0004867419835028. [DOI] [PubMed] [Google Scholar]
- 22.Jeppesen R, Christensen RHB, Pedersen EMJ, Nordentoft M, Hjorthøj C, Köhler-Forsberg O, et al. Efficacy and safety of anti-inflammatory agents in treatment of psychotic disorders - a comprehensive systematic review and meta-analysis. Brain Behav Immun. 2020;90:364–80. doi: 10.1016/j.bbi.2020.08.028. [DOI] [PubMed] [Google Scholar]
- 23.Laan W, Grobbee DE, Selten JP, Heijnen CJ, Kahn RS, Burger H. Adjuvant aspirin therapy reduces symptoms of schizophrenia spectrum disorders: results from a randomized, double-blind, placebo-controlled trial. J Clin Psychiatry. 2010;71:520–7. doi: 10.4088/JCP.09m05117yel. [DOI] [PubMed] [Google Scholar]
- 24.Müller N. COX-2 inhibitors, aspirin, and other potential anti-inflammatory treatments for psychiatric disorders. Front Psychiatry. 2019;10:375. doi: 10.3389/fpsyt.2019.00375. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Schmidt L, Phelps E, Friedel J, Shokraneh F. Acetylsalicylic acid (aspirin) for schizophrenia. Cochrane Database Syst Rev. 2019;8:Cd012116. doi: 10.1002/14651858.CD012116.pub2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Levkovitz Y, Mendlovich S, Riwkes S, Braw Y, Levkovitch-Verbin H, Gal G, et al. A double-blind, randomized study of minocycline for the treatment of negative and cognitive symptoms in early-phase schizophrenia. J Clin Psychiatry. 2010;71:138–49. doi: 10.4088/JCP.08m04666yel. [DOI] [PubMed] [Google Scholar]
- 27.Chaudhry IB, Hallak J, Husain N, Minhas F, Stirling J, Richardson P, et al. Minocycline benefits negative symptoms in early schizophrenia: a randomised double-blind placebo-controlled clinical trial in patients on standard treatment. J Psychopharmacol. 2012;26:1185–93. doi: 10.1177/0269881112444941. [DOI] [PubMed] [Google Scholar]
- 28.Zhang L, Zheng H, Wu R, Kosten TR, Zhang XY, Zhao J. The effect of minocycline on amelioration of cognitive deficits and pro-inflammatory cytokines levels in patients with schizophrenia. Schizophr Res. 2019;212:92–8. doi: 10.1016/j.schres.2019.08.005x. [DOI] [PubMed] [Google Scholar]
- 29.Krynicki CR, Dazzan P, Pariante CM, Barnes NM, Vincent RC, Roberts A, et al. Deconstructing depression and negative symptoms of schizophrenia; differential and longitudinal immune correlates, and response to minocycline treatment. Brain Behav Immunity. 2021;91:498–504. doi: 10.1016/j.bbi.2020.10.026. [DOI] [PubMed] [Google Scholar]
- 30.Sommer IE, van Westrhenen R, Begemann MJ, de Witte LD, Leucht S, Kahn RS. Efficacy of anti-inflammatory agents to improve symptoms in patients with schizophrenia: an update. Schizophr Bull. 2014;40:181–91. doi: 10.1093/schbul/sbt139. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Watkins CC, Andrews SR. Clinical studies of neuroinflammatory mechanisms in schizophrenia. Schizophr Res. 2016;176:14–22. doi: 10.1016/j.schres.2015.07.018. [DOI] [PubMed] [Google Scholar]
- 32.Müller N, Riedel M, Scheppach C, Brandstätter B, Sokullu S, Krampe K, et al. Beneficial antipsychotic effects of celecoxib add-on therapy compared to risperidone alone in schizophrenia. Am J Psychiatry. 2002;159:1029–34. doi: 10.1176/appi.ajp.159.6.1029. [DOI] [PubMed] [Google Scholar]
- 33.Muller N, Krause D, Dehning S, Musil R, Schennach-Wolff R, Obermeier M, et al. Celecoxib treatment in an early stage of schizophrenia: results of a randomized, double-blind, placebo-controlled trial of celecoxib augmentation of amisulpride treatment. Schizophr Res. 2010;121:118–24. doi: 10.1016/j.schres.2010.04.015. [DOI] [PubMed] [Google Scholar]
- 34.Rapaport MH, Delrahim KK, Bresee CJ, Maddux RE, Ahmadpour O, Dolnak D. Celecoxib augmentation of continuously ill patients with schizophrenia. Biol Psychiatry. 2005;57:1594–6. doi: 10.1016/j.biopsych.2005.02.024. [DOI] [PubMed] [Google Scholar]
- 35.Akhondzadeh S, Tabatabaee M, Amini H, Ahmadi Abhari SA, Abbasi SH, Behnam B. Celecoxib as adjunctive therapy in schizophrenia: a double-blind, randomized and placebo-controlled trial. Schizophr Res. 2007;90:179–85. doi: 10.1016/j.schres.2006.11.016. [DOI] [PubMed] [Google Scholar]
- 36.Zheng W, Cai DB, Yang XH, Ungvari GS, Ng CH, Müller N, et al. Adjunctive celecoxib for schizophrenia: a meta-analysis of randomized, double-blind, placebo-controlled trials. J Psychiatr Res. 2017;92:139–46. doi: 10.1016/j.jpsychires.2017.04.004. [DOI] [PubMed] [Google Scholar]
- 37.Jordan F, Quinn TJ, McGuinness B, Passmore P, Kelly JP, Tudur Smith C, et al. Aspirin and other non-steroidal anti-inflammatory drugs for the prevention of dementia. Cochrane Database Syst Rev. 2020;4:Cd011459. doi: 10.1002/14651858.CD011459.pub2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Müller N, Riedel M, Schwarz MJ, Engel RR. Clinical effects of COX–2 inhibitors on cognition in schizophrenia. Eur Arch Psychiatry Clin Neurosci. 2005;255:149–51. doi: 10.1007/s00406-004-0548-4. [DOI] [PubMed] [Google Scholar]
- 39.Gou J, Liang S, Cheng W, Wu S, Ye Z, Ma Y, et al. Neuroprotective effect of combined use of nicotine and celecoxib by inhibiting neuroinflammation in ischemic rats. Brain Res Bull. 2021;175:234–43. doi: 10.1016/j.brainresbull.2021.07.022. [DOI] [PubMed] [Google Scholar]
- 40.Li D, Hao S-H, Sun Y, Hu C-M, Ma Z-H, Wang ZM, et al. Functional polymorphisms in COX-2 gene are correlated with the risk of oral cancer. BioMed Res Int. 2015;2015:580652. doi: 10.1155/2015/580652. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Guan X, Dong C, Zhu P, Chen C, Wang T, Wu M, et al. Association of the cyclooxygenase-2 1759A allele with migraine in Chinese Han individuals. PloS One. 2020;15:e0239856. doi: 10.1371/journal.pone.0239856. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Zhang YC, Zhao H, Chen C, Ali MA. COX-2 gene rs689466 polymorphism is associated with increased risk of colorectal cancer among Caucasians: a meta-analysis. World J Surg Oncol. 2020;18:192. doi: 10.1186/s12957-020-01957-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Pereira C, Sousa H, Silva J, Brandão C, Elgueta-Karstegl C, Farrell PJ, et al. The -1195G allele increases the transcriptional activity of cyclooxygenase-2 gene (COX-2) in colon cancer cell lines. Mol Carcinog. 2014;53:E92–5. doi: 10.1002/mc.22049. [DOI] [PubMed] [Google Scholar]
- 44.Kay SR, Fiszbein A, Opler LA. The positive and negative syndrome scale (PANSS) for schizophrenia. Schizophr Bull. 1987;13:261–76. doi: 10.1093/schbul/13.2.261. [DOI] [PubMed] [Google Scholar]
- 45.Randolph C, Tierney MC, Mohr E, Chase TN. The repeatable battery for the assessment of neuropsychological status (RBANS): preliminary clinical validity. J Clin Exp Neuropsychol. 1998;20:310–9. doi: 10.1076/jcen.20.3.310.823. [DOI] [PubMed] [Google Scholar]
- 46.Zhang B-H, Tan YL, Zhang W-F, Wang Z-R, Yang G-G, Shi C, et al. Repeatable battery for the assessment of neuropsychological status as a screening test in Chinese: reliability and validity. Chin Mental Health J. 2008;22:865–9.
- 47.Zhang XY, Tan YL, Chen DC, Tan SP, Malouta MZ, Bernard JD, et al. Serum IL-18 level, clinical symptoms and IL-18-607A/C polymorphism among chronic patients with schizophrenia in a Chinese Han population. Psychoneuroendocrinology. 2016;68:140–7. doi: 10.1016/j.psyneuen.2016.03.002. [DOI] [PubMed] [Google Scholar]
- 48.Vogel LK, Sæbø M, Høyer H, Kopp TI, Vogel U, Godiksen S, et al. Intestinal PTGS2 mRNA levels, PTGS2 gene polymorphisms, and colorectal carcinogenesis. PLoS One. 2014;9:e105254. doi: 10.1371/journal.pone.0105254. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49.Serretti A, Chiesa A, Calati R, Massat I, Linotte S, Kasper S, et al. No influence of PTGS2 polymorphisms on response and remission to antidepressants in major depression. Psychiatry Res. 2011;188:166–9. doi: 10.1016/j.psychres.2010.11.023. [DOI] [PubMed] [Google Scholar]
- 50.Kulhara P, Avasthi A, Chadda R, Chandiramani K, Mattoo SK, Kota SK, et al. Negative and depressive symptoms in schizophrenia. Br J Psychiatry. 1989;154:207–11. doi: 10.1192/bjp.154.2.207. [DOI] [PubMed] [Google Scholar]
- 51.Malashenkova IK, Ushakov VL, Zakharova NV, Krynskiy SA, Ogurtsov DP, Hailov NA, et al. Neuro-immune aspects of schizophrenia with severe negative symptoms: new diagnostic markers of disease phenotype. Sovrem Tekhnologii Med. 2021;13:24–33. doi: 10.17691/stm2021.13.6.03. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 52.Ye J, Wei Y, Zeng J, Gao Y, Tang X, Xu L, et al. Serum levels of tumor necrosis factor-α and vascular endothelial growth factor in the subtypes of clinical high risk individuals: a prospective cohort study. Neuropsychiatr Dis Treat. 2023;19:1711–23. doi: 10.2147/ndt.S418381. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 53.Mullapudi T, Debnath M, Govindaraj R, Raj P, Banerjee M, Varambally S. Effects of a six-month yoga intervention on the immune-inflammatory pathway in antipsychotic-stabilized schizophrenia patients: a randomized controlled trial. Asian J Psychiatry. 2023;86:103636. doi: 10.1016/j.ajp.2023.103636. [DOI] [PubMed] [Google Scholar]
- 54.Cao P, Zhang H, Meng H, Cheng Y, Xu H, Zang S, et al. Celecoxib exerts a therapeutic effect against demyelination by improving the immune and inflammatory microenvironments. J Inflam Res. 2020;13:1043–55. doi: 10.2147/jir.S282128. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 55.McLaughlin M, Alloza I, Quoc HP, Scott CJ, Hirabayashi Y, Vandenbroeck K. Inhibition of secretion of interleukin (IL)-12/IL-23 family cytokines by 4-trifluoromethyl-celecoxib is coupled to degradation via the endoplasmic reticulum stress protein HERP. J Biol Chem. 2010;285:6960–9. doi: 10.1074/jbc.M109.056614. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 56.Moore AE, Young LE, Dixon DA. A common single-nucleotide polymorphism in cyclooxygenase-2 disrupts microRNA-mediated regulation. Oncogene. 2012;31:1592–8. doi: 10.1038/onc.2011.349. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data that support the findings of this study are available from the corresponding author, [ZXY], upon reasonable request.