McIntyre Roger S, Gorwood Philip, Thase Michael E, Liss Charlie, Desai Dhaval, Chen Ji, Bauer Michael
From the *Department of Psychiatry, University Health Network, University of Toronto, Toronto, Ontario, Canada; †Centre de Psychiatrie et Neurosciences (INSERM U894) Paris-Descartes University, Sainte-Anne Hospital (CMME), Paris, France; ‡Department of Psychiatry, University of Pennsylvania School of Medicine, Philadelphia, PA; §AstraZeneca Pharmaceuticals, Wilmington, DE; and ∥University Hospital Carl Gustav Carus, Technische Universität Dresden, Dresden, Germany.
J Clin Psychopharmacol. 2015 Dec;35(6):706-10. doi: 10.1097/JCP.0000000000000416.
The aim of this post-hoc analysis was to determine whether early symptom improvement with extended release quetiapine (quetiapine XR) may predict treatment outcome in patients with major depressive disorder. Data were from 6, double-blind, placebo-controlled studies of quetiapine XR (2 fixed-dose and 2 flexible-dose monotherapy and 2 adjunct studies) in adult patients with major depressive disorder. Montgomery-Åsberg Depression Rating Scale (MADRS) and Clinical Global Impression-Severity Score (CGI-S) were assessed at baseline, weeks 2, 4, and 6. Hamilton Rating Scale for Depression (HAM-D) was assessed at baseline and week 6. The MADRS improvement at week 2 (15%, 20%, 25%, 30%) was used to predict response and remission, based on MADRS (50% improvement; total score ≤ 12) or HAM-D (50% improvement; total score ≤ 7). The CGI-S improvement (1 point) at week 2 was used to predict final outcome (CGI-S score ≤ 2). The predictive value for early improvement with quetiapine XR was found to be "very strong" (Yule's Q coefficient, a combined measure of sensitivity and specificity) using 30% MADRS improvement as the threshold. This was relatively comparable for response and remission and for fixed-dose, flexible-dose, and adjunct studies. This was also observed for placebo. Exceptions were: adjunct studies (where predictivity was lower for ongoing antidepressant/placebo), and for remission (predictivity for remission appeared lower than for response with placebo). In conclusion, outcome at week 6 with quetiapine XR for a major depressive episode could be predicted by 30% improvement after 2 weeks, a finding that could give doctors confidence to continue treatment and may facilitate adherence in patients.
这项事后分析的目的是确定使用缓释喹硫平（quetiapine XR）早期症状改善是否可预测重度抑郁症患者的治疗结果。数据来自6项关于quetiapine XR（2项固定剂量和2项灵活剂量单药治疗以及2项辅助治疗研究）在成年重度抑郁症患者中的双盲、安慰剂对照研究。在基线、第2周、第4周和第6周评估蒙哥马利-奥斯伯格抑郁评定量表（MADRS）和临床总体印象-严重程度量表（CGI-S）。在基线和第6周评估汉密尔顿抑郁量表（HAM-D）。基于MADRS（改善50%；总分≤12）或HAM-D（改善50%；总分≤7），使用第2周时MADRS的改善情况（15%、20%、25%、30%）来预测缓解和痊愈。使用第2周时CGI-S的改善情况（1分）来预测最终结果（CGI-S评分≤2）。以30%的MADRS改善作为阈值，发现quetiapine XR早期改善的预测价值为“非常强”（Yule's Q系数，一种敏感性和特异性的综合指标）。对于缓解和痊愈以及固定剂量、灵活剂量和辅助治疗研究，这一情况相对类似。安慰剂组也观察到了这一情况。例外情况为：辅助治疗研究（正在使用抗抑郁药/安慰剂时预测性较低），以及痊愈情况（痊愈的预测性似乎低于安慰剂组的缓解预测性）。总之，重度抑郁发作患者使用quetiapine XR治疗第6周时的结果可通过2周后30%的改善情况进行预测，这一发现可让医生有信心继续治疗，并可能促进患者的依从性。
