Neurocognitive Benefits of Second-Generation Antipsychotics Versus Placebo: Insufficient Evidence Based on a Systematic Review
There have been a number of meta-analyses of randomized controlled trials (RCTs) examining the effect of second-generation antipsychotics (SGAs) versus first-generation antipsychotics (FGAs) on neurocognition in schizophrenia. For instance, Woodward et al1 found advantageous effects of SGAs on global neurocognition over FGAs; more recently, Désaméricq et al2 also showed superiority of risperidone, olanzapine, and quetiapine over haloperidol, whereas Nielsen et al3 reported no overall differences between SGAs and FGAs. Given that Mishara and Goldberg4 revealed that FGAs have more favorable effects on neurocognition than placebo in their meta-analysis, one could postulate that SGAs hold at least a similar magnitude of superiority for neurocognition over placebo. To our knowledge, however, there have been no systematic reviews and/or meta-analyses of RCTs comparing neurocognitive effects between SGAs and placebo in patients with schizophrenia.
Two authors (HT and ST) independently conducted a systematic literature search using MEDLINE, Embase, and Cochrane Central Register of Controlled Trials (CENTRAL) with the following keywords: (schizophr* OR schizoaff*) AND (antipsychotic* OR risperidone OR olanzapine OR quetiapine OR aripiprazole OR clozapine OR paliperidone OR ziprasidone OR amisulpride OR asenapine OR iloperidone OR lurasidone OR sertindole OR zotepine OR perospirone OR blonanserin OR brexpiprazole OR cariprazine) AND placebo AND (cogniti* OR neurocogniti* OR neuropsycholog*), with a limitation of RCT and English language (last search: May 18, 2016). We selected studies that met the following eligibility criteria: (1) RCTs, (2) patients with schizophrenia and/or schizoaffective disorder, (3) study arms including at least 1 SGA as monotherapy and placebo, and (4) reported neurocognitive outcomes. The 2 authors also independently collected the information on study design and neurocognitive findings from the identified studies and extracted or calculated effect sizes (Cohen d) of the differences in changes from baseline to endpoint in global neurocognition (composite scores) between SGAs and placebo, if available.
We screened 319 records after removing duplications and excluded 307 records at a title or abstract level. We assessed 12 full-text articles and identified 4 studies that met eligibility criteria.5–8 The characteristics and main findings of these studies are summarized in Table 1. All studies were parallel-group RCTs examining one or more of risperidone, olanzapine, quetiapine, and lurasidone as antipsychotic monotherapy, published after 2012. All studies were conducted in a double-blind fashion with study duration of up to 12 weeks. Two studies included new compounds (ie, BL-1020 and AZD2624) as the primary tested drug.5,8 As shown in Table 1, of the 4 studies, 1 demonstrated superior effects of SGAs over placebo in memory,7 whereas 3 did not.5,6,8 Only 2 studies used a cognitive battery that can compute global neurocognitive scores (ie, composite scores) and reported sufficient data to calculate effect sizes (Cohen d) for global neurocognition5,6; all effect sizes were as small as less than 0.1 (Table 1).