Antipsychotics for negative and positive symptoms of schizophrenia: dose-response meta-analysis of randomized controlled acute phase trials

Determining the optimal antipsychotic target dose in acute phase treatment is of high clinical relevance. The effect of antipsychotics on negative symptoms should be taken into account because patients will often continue on the treatment received in the acute phase. Therefore, we conducted a formal dose-response meta-analysis of negative symptoms and positive symptoms based on a systematic review of fixed-dose randomized controlled trials (RCTs) that examined the effectiveness of antipsychotics for the acute exacerbation of schizophrenia. Forty RCTs included a total of 15,689 patients. The 95% effective doses per day for the 13 antipsychotics included and 3 long acting were mostly different for negative and positive symptoms: amisulpride (481 mg, 690.6 mg); aripiprazole (11.9 mg, 11 mg); asenapine (7.61 mg, 5.66 mg); brexpiprazole (2.1 mg, 4 mg); cariprazine (4 mg, 6.51 mg); haloperidol (6.34 mg, 7.36 mg); lurasidone (58.2 mg, 86.3 mg); olanzapine (15.5 mg, 9.52 mg); olanzapine long-acting injection (15.7 mg, 13.5 mg); paliperidone (7.2 mg, 7 mg); paliperidone long-acting injection (7.5 mg, 5.9 mg); quetiapine instant-release (264.2 mg, 316.5 mg); quetiapine extended-release (774 mg, 707.2 mg); risperidone (7.5 mg, 7.7 mg); risperidone long-acting injection (5.13 mg, 6.7 mg); sertindole (13.5 mg, 16.3 mg); and ziprasidone (71.6 mg, 152.6 mg). The shape of the dose-response curves varied across different drugs with most drugs showing a plateau at higher doses. Most dose-response curves suggested that the near-maximum effective doses could be in the lower-to-medium range of the licensed dose. Additional RCTs are necessary to establish the optimal dose.


Rationale
3 Describe the rationale for the review in the context of what is already known.

3-4
Objectives 4 Provide an explicit statement of questions being addressed with reference to participants, interventions, comparisons, outcomes, and study design (PICOS).

METHODS
Protocol and registration 5 Indicate if a review protocol exists, if and where it can be accessed (e.g., Web address), and, if available, provide registration information including registration number.
14 Supplementary notes Eligibility criteria 6 Specify study characteristics (e.g., PICOS, length of follow-up) and report characteristics (e.g., years considered, language, publication status) used as criteria for eligibility, giving rationale. 16 Information sources 7 Describe all information sources (e.g., databases with dates of coverage, contact with study authors to identify additional studies) in the search and date last searched.

15-16
Search 8 Present full electronic search strategy for at least one database, including any limits used, such that it could be repeated. 16.

Supplementary notes
Study selection 9 State the process for selecting studies (i.e., screening, eligibility, included in systematic review, and, if applicable, included in the meta-analysis).

SM Fig 1
Study characteristics 18 For each study, present characteristics for which data were extracted (e.g., study size, PICOS, follow-up period) and provide the citations.
SM Table 1 Risk of bias within studies 19 Present data on risk of bias of each study and, if available, any outcome level assessment (see item 12).  Table S2 Additional analysis 23 Give results of additional analyses, if done (e.g., sensitivity or subgroup analyses, meta-regression [see Item 16]

DISCUSSION
Summary of evidence 24 Summarize the main findings including the strength of evidence for each main outcome; consider their relevance to key groups (e.g., healthcare providers, users, and policy makers).

5/34
Stefan Kaiser has received royalties on an institutional account for education and research for cognitive test and training software from Schufried. All other authors declare no conflicts of interest.

Review question
The primary aim of the present review is to examine the dose-response relationship between first-and second-generation antipsychotic drugs for negative and positive symptoms of schizophrenia in acute schizophrenia by applying a dose-response meta-analysis approach.

Search terms
The systematic review shall be conducted using the following databases: MEDLINE, EMBASE, PubMed, PsyARTICLES, PsyINFO, Cochrane Database of Systematic Reviews and different trail registries (ClinicalTrials.gov. and clinicaltrialsregister.eu).
We will use combinations of the following terms: The full list of search term used for one database can be found below.
(benperidol OR chlorpromazine OR clopenthixol OR flupenthixol OR fluphenazine OR fluspirilene OR haloperidol OR levomepromazine OR methotrimeprazine OR molindone OR penfluridol OR perazine OR perphenazine OR pimozide OR thioridazine OR thiothixene OR trifluoperazine OR zuclopenthixol OR amisulpride OR aripiprazole OR asenapine OR brexpiprazole OR cariprazine OR clozapine OR iloperidone OR lurasidone OR loxapine OR olanzapine OR paliperidone OR quetiapine OR risperidone OR sertindole OR ziprasidone OR zotepine) AND (schizo* OR psychosis OR psychotic).
The following search limits were applied: English language, human studies, adult population (aged 18-65-year-old). Search results will be limited to clinical trials and randomized controlled trials.
Excluded studies: studies from mainland China were excluded to avoid a systematic bias.
We will inspect titles, abstracts, and methods of all papers or clinical trials identified in the electronic searches.

6/34
We will search previous reviews investigating antipsychotics in general schizophrenia to identify additional studies.
In order to reduce the exclusion of unpublished papers, additional sources including Food and Drug administration website, from data from Cochrane reviews, and previous published meta-analysis on the use on antipsychotics for acute schizophrenia will be screened for additional studies.

Condition or domain being studied
Negative symptoms of schizophrenia contribute heavily to functional disability and burden of the disease and are still an unmet clinical need.
The treatment of the negative symptoms during an acute schizophrenia episode is still rarely addressed and remain a considerable clinical challenge.
The issue of prescribing the best antipsychotics medications for patients with schizophrenia presenting predominant negative symptoms have been addresses in a recent meta-analysis by Krause and colleagues (Krause et al., 2018). In addition, Leucht and colleagues have focused on the dose-response profile of antipsychotics medication from all published randomized controlled trials (RCTs) for patients with schizophrenia (Leucht et al., 2020). Considering the small number of trials available, the authors have focused on acute exacerbation. However, these authors have mostly focused on the score change from baseline on the Positive and Negative Syndrome Scale (PANSS) and only optimal doses for treatment with amisulpride in a specific population of patients affected by predominant negative symptoms.
To the best of our knowledge, when considering all antipsychotics, uncertainty persists about the dose dependency and optimal target dose for negative symptoms antipsychotic medications. Hence, there is a clear need to identify with available studies the near-maximum effective doses for the treatment of negative symptoms of schizophrenia. This information would be important for decision making by clinicians. Therefore, we decided to conduct a systematic review and dose-response meta-analysis of double-blind RCTs that used fixed doses (or fixed doses ranges) of antipsychotic drugs for the treatment of negative symptoms in acute schizophrenia.

Participants/ population
We will include adult people (18-65 years, no restriction in setting, gender, ethnicity) with schizophrenia or related disorders (such as schizophreniform, or schizoaffective disorders).
We will include studies in which a small proportion of the participants presents other psychiatric disorders than schizophrenia (<20%).
We will not include patients presenting a first episode of schizophrenia.

Comparator(s)/control
In order to estimate a flexible dose-response model defined by 2 coefficients to compare at least three fixed dose levels of treatments (with consideration of the placebo dose of 0 mg) in order to estimate model parameters (Crippa & Orsini, 2016a). No restriction will be applied regarding dose of treatment.

Context
We will include adult people with schizophrenia, schizophreniform or schizoaffective disorders with an acute exacerbation of their primary disease.
We will include studies in which less than 20% of participants are suffering from other psychiatric disorders.
We will include outpatient and inpatient settings.
Included studies must be of at least 3-week duration trials.
Included studies must report at least three fixed dose (or fixed doses range) levels of treatments (with consideration of the placebo dose of 0 mg) in order to estimate model parameters according to the statistical model used (Crippa & Orsini, 2016a).

Main outcome
The main outcomes are the intention to treat score change from baseline (mean change) negative and positive symptoms.

Additional outcomes
-Estimation of effective doses 9/34 We will estimate for each drug the 50% (ED50) and 95%(ED95) effective doses, as it is commonly performed in dose-response analysis (Pinheiro et al. 2016).
The ED50 is the mean dose that produces half of the maximum reduction of the patient's symptoms.

-Estimation of dose equivalence
The ED95 of each drug will be used to obtain the risperidone dose equivalence ratios.

Data extraction
Two review authors (MS and NZ) will independently screen the titles and abstracts yielded by the search against the inclusion criteria. We will obtain full reports for all titles that appear to meet the inclusion criteria or for which there is any uncertainty. Both review authors will then screen the full-text reports and decide whether these meet the inclusion criteria.
We will seek additional information from study authors where necessary to resolve questions about eligibility. We will resolve disagreement through discussion.
We will record the reasons for excluding trials.

Risk of bias assessment
Two reviewers (MS and SK) will independently assess the risk of bias employing the Cochrane Collaboration bias assessment tool: 1) Random sequence generation; was the allocation sequence adequately generated?
2) Allocation concealment; was the allocated treatment adequately concealed from the study participants and clinicians and other healthcare or research staff at the enrollment stage?

10/34
3) Blinding of participants and personnel; were the personnel assessing outcomes and analyzing data sufficiently blinded to the intervention allocation throughout the trial? 4) Blinding of outcome assessment; were all measures used, if any, to blind outcome assessors from the knowledge of which intervention a participant received, was there a report on the intended blinding effectiveness? 5) Incomplete outcome data; were participant exclusions, attrition, and incomplete outcome data adequately addressed in the published report? 6) Selective reporting; is there evidence of selective outcome reporting and might this have affected the study results?
Unpublished studies will be searched by checking pre-registration of trials.
If possible, publication bias will be assessed using a funnel plot. Disagreements between the reviews authors over the risk of bias in particular studies will be resolved by discussion.

Strategy for data synthesis
We plan to conduct a dose-response meta-analysis of aggregated data following the method proposed by Crippa and Orsini (Crippa & Orsini, 2016).
Our statistical approach is entirely based on the two-stage approach dose-response meta-analysis of differences in means described in the Crippa and Orsini paper (Crippa & Orsini, 2016).
We plan to pool dose-response relations from aggregated data where the changes in the distribution of the quantitative outcome are expressed in terms of differences in means.
The effect size was the standardized mean difference (cohen's d).
The Crippa and Orsini model is deployed using a two-stage approach:

11/34
First, a flexible dose-response model is estimated within each study taking into account the covariance of the data points (mean differences, standardized mean differences). Regression splines will be used to flexibly model the dose of interest. Splines represent a family of functions that describe a wide range of curves.
These curves consist of piecewise polynomials over consecutive intervals defined by k knots. We will use knots located at the 25th, 50th and 75th percentiles.
And second, parameters describing the study-specific curves are then combined using a multivariate random-effects model to address heterogeneity across studies. Dose-response curves will be estimated.
These dose-response curves characterize the relative efficacy of the dose under investigation using the placebo effect as referent.

Analysis of subgroups or subsets
Subgroup analyses will be conducted by restricting included studies to those defining change in negative symptoms as primary outcome. If appropriate, additional sensitivity analysis will be conducted.

12/34
Supplementary  1 All patients had a primary diagnosis of DSM-III-R or DSM-IV schizophrenia, with no other primary psychiatric diagnoses. A combined score of at least eight on any two of the positive symptoms of the BPRS a (hallucinatory behavior, conceptual disorganization, unusual thought content, or suspiciousness) was required in order to enroll patients with active psychosis of at least moderate severity. Scores of less than three on each item of the AIMS were also required to restrict individuals with tardive dyskinesia from entry, which ensured the exclusion of patients with even mild tardive dyskinesia. Inpatients were included (18 to 65-year-old). All patients had a primary diagnosis of chronic or subchronic schizophrenia, with acute exacerbation, according to DSM-III-R criteria; a minimum score of 4 (moderate) on at least two of four core positive symptoms on the BPRS; and a minimum total score of 12 for four core BPRS items. Inpatients only were included (18 to 60-year-old). Patients had a diagnosis of schizophrenia DSM-IV and were experiencing an acute exacerbation of symptoms that required inpatient hospitalization. In addition, patients were required to have PANSS Total score of 60 or more (1-7 scale) and a score of at least 4 on two or more of the following PANSS items at the baseline assessment: delusions, hallucinatory behavior, conceptual disorganization or suspiciousness/persecution. Inpatients were included (≥18-year-old). Patients had a DSM-IV-TR diagnosis of schizophrenia with an acute exacerbation of psychotic symptoms at study enrollment. For inclusion, patients had to present PANSS total score of 60 or higher, with scores of 4 or higher on at least 2 of 5 predefined PANSS positive subscale items (delusions, conceptual disorganization, hallucinatory behavior, grandiosity, and suspiciousness/persecution) at the initial screening assessment and at baseline for enrolled patients, and a CGI-S f score of 4 or higher (moderately ill) at baseline. Inpatients were included. These patients could continue the study as outpatients (≥18year-old). Patients had a DSM-IV-TR diagnosis of schizophrenia with an acute exacerbation of psychotic symptoms at study enrollment. The current acute exacerbation of schizophrenia had to be of ≤2 months duration. Other key inclusion criteria were a PANSS total score ≥60, with scores of ≥4 in two or more of five items on the PANSS positive subscale (delusions, conceptual disorganization, hallucinatory behavior, grandiosity, suspiciousness/persecution) at the initial screening assessment and at baseline, and a score of ≥4 on the CGI-S scale at baseline. Inpatients were included (20 to 64-year-old Patients were diagnosed with DSM-IV-TR for schizophrenia and confirmed by the Mini International Neuropsychiatric Interview assessment for experiencing acute exacerbation of psychotic symptoms, psychotic disorders, and marked deterioration of normal function by meeting the following criteria at screening and baseline: CGI-S score of ≥4, BPRS score of ≥40, and score of ≥4 for two or more of the BPRS items (conceptual disorganization, hallucinatory behavior, suspiciousness, and unusual thought content). Inpatients were included (18 to 65-year-old).  11 Included patients met the DSM-IV-TR criteria for schizophrenia. Patients had the diagnosis for at least one year, a current exacerbation less than 2 weeks' duration, and at least one psychotic episode requiring hospitalization/antipsychotic medication change/intervention during the preceding year. PANSS total score between 80 and 120, a score≥4 (moderate) on at least 2 of 4 PANSS positive symptoms (delusions, hallucinatory behavior, conceptual disorganization, suspiciousness/persecution), and CGI-S rating ≥4 were required. Body mass index between 18 and 35 was also required.  12 Included patients had a DSM-IV-TR criteria for schizophrenia, present for more than one year and with at least one psychotic episode that required hospitalization or change of antipsychotic medication during the past year. To ensure that participants' current psychotic episode was acute, duration of the current episode must be inferior to two weeks. A CGI-S score ≥4, a PANSS total score ≥80 and ≤120, and a score ≥4 on at least 2 of the PANSS positive symptoms of delusions, hallucinatory behavior, conceptual disorganization or suspiciousness/persecution was also required. Inpatients were included. These patients could continue the study as outpatients (18 to 60-year-old).  13 Include patients had a schizophrenia diagnosis for 1 year or longer based on the DSM-IV-TR, with a current psychotic episode less than 4 weeks in duration and at least one other psychotic episode in the past year that required hospitalization or change in antipsychotic medication. At both screening and randomization, all patients had a PANSS total score of 80-120 (inclusive), a score of 4 or higher on either the PANSS delusions item or the hallucinatory behavior item, a score of 4 or higher on either the PANSS conceptual disorganization item or the suspiciousness/persecution item, and a CGI-S score of 4 or higher. Inpatients were included. These patients could continue the study as outpatients ( 14 Patients with a current diagnosis of schizophrenia defined by criteria from the DSM-IV-TR were included. Participants had been diagnosed for 1 year or more and had a current psychotic episode of less than 2 weeks' duration with at least 1 psychotic episode that required hospitalization or change in antipsychotic therapy during the previous year. A CGI-S score of 4 or higher, PANSS total score of 80 or more and 120 or less, and a score of 4 or higher on at least 2 of the following 4 PANSS items: delusions, hallucinatory behavior, conceptual disorganization, or suspiciousness/persecution were required. Inpatients were included. These patients could continue the study as outpatients ( 15 Included patients had DSM-IV-TR criteria for schizophrenia with disorganized, paranoid, or undifferentiated subtypes. Patients were required to have an exacerbation of psychotic symptoms within 60 days before screening, with a PANSS total score of ≥80, including a score of ≥4 (moderate) on two or more of the following PANSS items: delusions (P1), conceptual disorganization (P2), hallucinations (P3), suspiciousness (P6), and unusual thought content (G9) at screening and baseline visits. Inpatients were included (18 to 74-year-old).  16 Included patients had a DSM-IV-TR criteria for a primary diagnosis of schizophrenia as determined by clinical interview using the Mini International Neuropsychiatric Interview. Subjects were also required to have an illness duration greater than 1 year with the current acute exacerbation of psychotic symptoms no longer than 2 months and, at the Screening and Baseline visits, to have a CGI-S score ≥4 (moderate or greater) and a PANSS total score ≥80, including a score ≥4 (moderate) on two or more of the following PANSS items: delusions, conceptual disorganization, hallucinations, unusual thought content, and suspiciousness. Inpatients were included (18 to 75-year-old). The study enrolled patients with a DSM-IV criteria for primary diagnosis of schizophrenia who were hospitalized for an acute exacerbation of symptoms. Patients were also required to have illness duration of at least 1 year, no psychiatric hospitalization within the 3 months prior to study entry, a BPRS derived from the PANSS of ≥42, a score of ≥4 on two or more items of the positive symptoms subscale on the PANSS, and a CGI-S score of ≥4 (moderate). Inpatients were included. These patients could continue the study as outpatients ( 20 All patients enrolled met the DSM-IILR criteria for schizophrenia (295.1-295.3, 295.9) as established by clinical interview and chart review. Residual type 295.6 was excluded. Patients were required to have a minimum BPRS, total score (BPRS items scored 0-6) extracted from the PANSS of at least 24. Also, patients were required to have a CGI-S score >4. Patients with a diagnosis of a DSM-III-R organic mental disorder or substance-use disorder active within 3 months of study entry were excluded as were patients at serious suicidal risk. Inpatients and outpatients were included. Inpatients could continue the study as outpatients ( the last injection must have been received at least 2 weeks or 1 injection interval, which was longer before double-blind treatment. Inpatients and outpatients were included. Patients were initially all hospitalized, and could then continue the study as outpatients (18 to 75-year-old). Paliperidone ER (Canuso et al., 2010) 23 Included patients met the DSM-IV criteria for an acute exacerbation of a schizoaffective disorder. Patients were required to have a PANSS total score of at least 60 and a score ≥4 on at least 2 of the following PANSS items (Pt, P4, G4, G8, G14 Included patients presented a diagnosis of schizophrenia for at least 1 year before screening, a Positive and Negative Syndrome Scale (PANSS) total score at screening and baseline between 70 and 120 (inclusive), and with a body mass index (BMI) >17.0 kg/m2. Patients were initially all hospitalized for a minimum duration of 8 days, and could then continue the study as outpatients (≥18-year-old). Enrolled patients had a diagnosis of schizophrenia according to DSM-IV criteria for at least 1 year had a PANSS total score of 70-120, inclusive, at screening, and 60-120 inclusive, on day 1 before the start of double-blind study drug, and had a body mass index (BMI) range of 15-35 kg/m2. Patients were initially all hospitalized for a minimum duration of 14 days, and could then continue the study as outpatients (18-65-year-old). Included patients were affected by an acute exacerbation of an established diagnosis of schizophrenia defined by the DSM-IV, whose disease diagnosis was able to be documented as present for at least 1 year before study screening, who demonstrated PANSS total score between 70 and 120 at screening and between 60 and 120 at baseline. Patients were initially all hospitalized for a minimum duration of 8 days, and could then continue the study as outpatients (≥18-year-old). Eligible patients who met the diagnostic criteria for schizophrenia according to the DSM-IV-TR for at least 1 year before screening. Patients had a PANSS total score at screening and baseline of 70-120 and a body mass index (BMI)>15.0 kg/m2. Patients were initially all hospitalized for a minimum duration of 8 days, and could then continue the study as outpatients (≥18-year-old). Patients with a DSM-IV diagnosis of schizophrenia-catatonic (DSM-IV diagnostic code 295.20); disorganized (295.10); paranoid (295.30); or undifferentiated (295.90) were eligible to participate. To be included in the study, patients had to meet the following criteria: a PANSS total score ≥60; a score of ≥4 for at least one of the PANSS items of delusions, conceptual disorganization, hallucinatory behavior, and suspiciousness/persecution; a Clinical Global Impressions-Severity of Illness (CGI-S) score ≥4 (at least moderately ill) and, in the opinion of the investigator, a worsening of the patient's condition in the previous 3 weeks. Patients who were screened as outpatients were hospitalized when enrolled and could be discharged from the hospital on day 10 at the investigator's discretion (18-65-year-old Included patient met the DSM-III-R criteria for schizophrenia. To be included in the study, each patient was also required to have a total score on the PANSS scale of no less than 60 and no greater than 120. The mean length of hospitalization was different among each inclusion center (see Table 2  Eligible patients met a DSM-IV diagnosis of schizophrenia (any subtype). Subjects were required at baseline to be inpatients and to have a PANSS score of 80-120 and a score of at least 8 on any two of the BPRS psychosis cluster items combined. Subjects were required at baseline to be inpatients, and had to remain at least 2 weeks at hospital (18- 37 Patients with a DSM-IV criteria of schizophrenia were enrolled. Inclusion criteria included a PANSS total score of 60 to 120 and good general health. Inpatients and outpatients were included (18- 1 All patients had a primary diagnosis of DSM-III-R or DSM-IV schizophrenia, with no other primary psychiatric diagnoses. A combined score of at least eight on any two of the positive symptoms (hallucinatory behavior, conceptual disorganization, unusual thought content, or suspiciousness) of the BPRS was required in order to enroll patients with active psychosis of at least moderate severity. Scores of less than three on each item of the AIMS were also required to restrict individuals with tardive dyskinesia from entry, which ensured the exclusion of patients with even mild tardive dyskinesia. Inpatients were included (18 to 65-year-old). Included patients presented an acute exacerbation of chronic or subchronic schizophrenia (295.x3) or schizoaffective disorder (295.x4) as defined in DSM-III-R were eligible to enter. They were to have been hospitalized within the previous 4 weeks and been diagnosed at least 6 months before the study. The patients were required to have a total score > 60 on the PANSS and a score of at least 4 on two or more core items in the PANSS (conceptual disorganization, hallucinatory behavior, suspiciousness, and unusual thought content) in the 24 hours before study treatment was started. In addition, the patients were required to have a score of 3 (minimally improved) or greater (worse) on the CGI-I at baseline as compared with screening (≥18-year-old). Included patients presented an acute exacerbation of chronic or subchronic schizophrenia or schizoaffective disorder as defined in DSM-III-R who had been hospitalized within the previous 3 weeks were allowed to enter the study. Patients were required to have a minimum duration of illness of at least 1 year. At screening and 24 h before study treatment was started, patients were required to have a total score of greater than 37 on the BPRS (anchored version, 1-7 rating system) and a score of at least four (moderate) on two or more of the BPRS core items (suspiciousness, conceptual disorganization, hallucinatory behavior, unusual thought content). All patients were inpatients (18 to 64-year-old).  Figure 2. Dose-response curves for quetiapine ER and ziprasidone after exclusion of failed studies. The figures represent pooled doseresponse association between one antipsychotic and the mean change in the negative or the positive subscale scores of the PANSS (solid line). The antipsychotic dosage is modeled with restricted cubic splines in a random-effects model. Dash lines represent the 95 % confidence intervals for the spline model. The placebo group (dose= 0) served as the referent group. Circles indicate observed mean differences in individual studies; size of bubbles is proportional to precision (inverse of variance) of the standardized mean differences. Right axis represents percentage of the maximum predicted effect