Department of Neuropsychiatry, Keio University School of Medicine, Tokyo, Japan.
Department of Neuropsychiatry and Clinical Ethics, Faculty of Medicine, University of Yamanashi, Yamanashi, Japan.
J Clin Psychiatry. 2018 Dec 4;80(1):18m12144. doi: 10.4088/JCP.18m12144.
Placebo effects remain largely unexplored in clinical trials of long-acting injectable (LAI) antipsychotics for schizophrenia. This study aims to characterize patients showing improvements after placebo injections and to search for criteria for the prediction of subsequent response based on the magnitude of score changes after the first week of treatment.
Data from 450 patients with schizophrenia (DSM-IV) who received placebo injections in 4 double-blind randomized controlled trials evaluating efficacy of LAI paliperidone palmitate obtained through the Yale University Open Data Access (YODA) project were analyzed. These 4 studies were conducted from October 2003 to March 2008. Multiple logistic regression analyses were conducted to examine associations between placebo response and demographic and clinical characteristics. The predictive power of improvement at week 1 for response at week 9 was investigated; sensitivity and specificity of incremental 5% cutoff points between a 5% and 25% reduction in Positive and Negative Syndrome Scale (PANSS) total score at week 1 were calculated.
Percent reduction in the PANSS total score at week 1 and a lower PANSS G12 item score (ie, better in judgment and insight) at baseline were significantly associated with placebo response at week 9 (odds ratio [OR] = 1.063; 95% CI, 1.040-1.087, P < .001; and OR= 0.739; 95% CI, 0.553-0.986, P = .040, respectively, in the per-protocol analysis). Cutoffs of a 10% (accuracy = 0.724 in the per-protocol analysis) and 15% (accuracy = 0.722 in the last-observation-carried-forward analysis) reduction in the PANSS total score at week 1 showed the highest predictive power.
The appreciation that longer-term response following placebo injections can be predicted by a 10%-15% PANSS total score reduction at week 1 could guide the design of future clinical trials of LAI antipsychotics in schizophrenia to identify and exclude potential placebo responders early during the course of the study.
长效注射用（LAI）抗精神病药治疗精神分裂症的临床试验中，安慰剂效应在很大程度上仍未得到探索。本研究旨在描述接受安慰剂注射后病情改善的患者特征，并根据治疗第一周后评分变化的幅度，寻找基于预测后续反应的标准。
通过耶鲁大学开放数据访问（YODA）项目获得的 450 例接受 LAI 棕榈酸帕利哌酮安慰剂注射的精神分裂症患者（DSM-IV）的 4 项双盲随机对照试验数据进行分析。这 4 项研究于 2003 年 10 月至 2008 年 3 月进行。进行多变量逻辑回归分析以检查安慰剂反应与人口统计学和临床特征之间的关联。研究了第 1 周的改善程度对第 9 周反应的预测能力；计算了第 1 周阳性和阴性综合征量表（PANSS）总分降低 5%（从 5%到 25%）时第 1 周 PANSS 总评分增量 5%的截断值的灵敏度和特异性。
第 1 周 PANSS 总评分降低和基线时 PANSS G12 项目评分较低（即判断和洞察力较好）与第 9 周的安慰剂反应显著相关（比值比[OR] = 1.063；95%置信区间[CI]，1.040-1.087，P <.001；和 OR= 0.739；95% CI，0.553-0.986，P =.040，在方案分析中）。第 1 周 PANSS 总评分降低 10%（方案分析中的准确率为 0.724）和 15%（最后观察结转分析中的准确率为 0.722）的截断值显示出最高的预测能力。
通过第 1 周 PANSS 总评分降低 10%-15%可以预测安慰剂注射后的长期反应，这可以指导 LAI 抗精神病药治疗精神分裂症的未来临床试验的设计，以便在研究过程早期识别并排除潜在的安慰剂反应者。
