Clínica de Enfermedades Crónicas y Procedimientos Especiales CECYPE, Morelia, Mich., Mexico.
J Affect Disord. 2010 Dec;127(1-3):343-51. doi: 10.1016/j.jad.2010.06.002. Epub 2010 Jul 2.
Major depressive disorder (MDD) is treated with antidepressants, but only between 50% and 70% of the patients respond to the initial treatment. Several authors suggested different factors that could predict antidepressant response, including clinical, psychophysiological, neuropsychological, neuroimaging, and genetic variables. However, these different predictors present poor prognostic sensitivity and specificity by themselves. The aim of our work is to study the possible role of clinical variables, neuropsychological performance, and the 5HTTLPR, rs25531, and val108/58Met COMT polymorphisms in the prediction of the response to fluoxetine after 4weeks of treatment in a sample of patient with MDD.
64 patients with MDD were genotyped according to the above-mentioned polymorphisms, and were clinically and neuropsychologically assessed before a 4-week fluoxetine treatment. Fluoxetine response was assessed by using the Hamilton Depression Rating Scale. We carried out a binary logistic regression model for the potential predictive variables.
Out of the clinical variables studied, only the number of anxiety disorders comorbid with MDD have predicted a poor response to the treatment. A combination of a good performance in variables of attention and low performance in planning could predict a good response to fluoxetine in patients with MDD. None of the genetic variables studied had predictive value in our model.
The possible placebo effect has not been controlled. Our study is focused on response prediction but not in remission prediction.
Our work suggests that the combination of the number of comorbid anxiety disorders, an attentional variable, and two planning variables makes it possible to correctly classify 82% of the depressed patients who responded to the treatment with fluoxetine, and 74% of the patients who did not respond to that treatment.
重度抑郁症（MDD）的治疗方法是使用抗抑郁药，但只有 50%到 70%的患者对初始治疗有反应。几位作者提出了不同的可以预测抗抑郁反应的因素，包括临床、心理生理、神经心理学、神经影像学和遗传变量。然而，这些不同的预测因素本身的预后敏感性和特异性都较差。我们的工作旨在研究临床变量、神经心理学表现以及 5HTTLPR、rs25531 和 val108/58Met COMT 多态性在预测 MDD 患者接受氟西汀治疗 4 周后的反应中的可能作用。
根据上述多态性对 64 例 MDD 患者进行基因分型，并在接受氟西汀治疗前 4 周进行临床和神经心理学评估。使用汉密尔顿抑郁评定量表评估氟西汀的反应。我们对潜在预测变量进行了二元逻辑回归模型分析。
在所研究的临床变量中，只有与 MDD 共病的焦虑障碍数量预测了对治疗的不良反应。注意力变量表现良好和计划能力表现不佳的组合可以预测 MDD 患者对氟西汀的良好反应。研究中没有遗传变量具有预测价值。
未控制可能的安慰剂效应。我们的研究侧重于预测反应，而不是预测缓解。
我们的工作表明，共病焦虑障碍的数量、注意力变量和两个计划变量的组合可以正确分类 82%对氟西汀治疗有反应的抑郁患者，以及 74%对该治疗无反应的患者。
