STAR*D Revising Conventional Wisdom
STAR*D Revising Conventional Wisdom
复制标题
DOI：
10.2165/00023210-200923080-00001
复制
发表时间：
2009-01-01
期刊：
影响因子：
6
通讯作者：
Nierenberg, Andrew A.
中科院分区：
文献类型：
--
作者：
Rush, A. John;Warden, Diane;Nierenberg, Andrew A.
星星 *D（缓解抑郁症的序贯治疗替代方案）研究使用了一系列序贯随机治疗试验，在第一个治疗步骤之后，如果需要，随后的治疗步骤，以确定各种选择在急性和长期治疗中的耐受性和有效性。成人门诊患者（n=4041）与非精神病性重度抑郁症，大量的慢性和复发性抑郁症，共病的精神和一般的医疗条件，在41个代表性的初级和专科护理设置。大约三分之一的参与者在西酞普兰的第一步治疗中缓解，其中50%在6周内缓解。可能的结果与少数民族地位、社会经济劣势、更多轴I和III共病障碍、较低的功能和生活质量以及焦虑和抑郁特征相关。在第2步药物转换中，类内、类外或双作用药物之间的缓解率无显著差异：舍曲林（27%）、安非他酮缓释剂（26%）和文拉法辛缓释剂（25%）。在西酞普兰的第2步药物增强中，安非他酮缓释剂（39%）和丁螺环酮（33%）之间的缓解没有显著差异，尽管使用安非他酮缓释剂的参与者症状减轻程度更大，耐受性更好。无论是转换（31% vs 27%）还是增强（31% vs 33%）策略，第2步认知治疗和药物治疗之间的缓解没有显著差异，尽管认知治疗增强的参与者比药物增强的参与者有更长的缓解时间（55 vs 40天）。在第3步中，转换为米氮平（8%）或去甲替林（12%）之间，或锂（13%）或T3（三碘甲状腺原氨酸，碘甲状腺原氨酸）[25%]之间的缓解没有差异，尽管由于不良反应而停用锂的参与者多于停用T-3的参与者。在第四步中，反苯环丙胺（14%）或文拉法辛缓释加米氮平（16%）之间的缓解率没有差异，尽管联合治疗的不良反应较少，并且具有不需要洗脱期或饮食限制的优势。需要两次以上良好递送治疗的参与者可以被表征为治疗抗性，因为在该点之后缓解率显著降低。治疗抵抗与更多并发的轴I或III共病状况、社会经济劣势、慢性和焦虑或焦虑特征相关。然而，如果参与者继续接受长达四个步骤的治疗，约67%的人达到缓解。对于后续治疗步骤，至缓解的时间并没有显著延长。与缓解但未缓解的受试者相比，进入缓解期的受试者在自然随访中的复发率较低，突出了达到缓解的重要性（第1步：34% vs 59%;第2步：47% vs 68%;第3步：42% vs 76%;第4步：50% vs 83%）。基于每次治疗访视时症状和不良反应的逐项测量的临床决策在星星 *D的真实的世界环境中是可行的，并且导致通常超过实践环境中通常发现的适当剂量和治疗持续时间。尽管由于均衡分层随机设计，转换和加强策略不能直接比较，但第2步药物加强的较高缓解率令人感兴趣，值得进一步研究。
The STAR*D (Sequenced Treatment Alternatives to Relieve Depression) study used a series of sequenced, randomized treatment trials following a first and, if needed, subsequent treatment steps to define the tolerability and effectiveness of various options in both acute and longer term treatment. Adult outpatients (n=4041) with nonpsychotic major depressive disorder, substantial chronic and recurrent depression, and co-morbid psychiatric and general medical conditions were enrolled in 41 representative primary and specialty care settings. About one-third of participants remitted in first step treatment with citalopram, 50% of these within 6 weeks. Poorer outcomes were associated with minority status, socioeconomic disadvantage, more axis I and III co-morbid disorders, lower function and quality of life, and anxious and melancholic features. In step 2 medication switch, there were no significant differences in remission among within-class, out-of-class or dual-action agents: sertraline (27%), bupropion-sustained release (26%) and venlafaxine-extended release (25%). In step 2 medication augmentation of citalopram, there was no significant difference in remission between bupropion-sustained release (39%) and buspirone (33%), although participants using bupropion-sustained release had greater symptom reduction and better tolerability. There were no significant differences in remission in step 2 between cognitive therapy and medication treatment in either the switch (31% vs 27%) or augmentation (31% vs 33%) strategies, although participants in cognitive therapy augmentation had a longer time to remission than those in medication augmentation (55 vs 40 days). In step 3, there were no differences in remission between a switch to mirtazapine (8%) or nortriptyline (12%), or between augmentation with lithium (13%) or T3 (triiodothyronine, liothyronine) [25%], although more participants discontinued lithium due to adverse effects than discontinued T-3. In the fourth step, there was no difference in remission between tranylcypromine (14%) or venlafaxine-extended release plus mirtazapine (16%), although the combination treatment had fewer adverse effects and had the advantages of not requiring a washout period or diet restrictions. Participants requiring more than two well delivered treatments may be characterized as treatment resistant given the substantially lower remission rates after that point. Treatment resistance was associated with more concurrent axis I or III co-morbid conditions, socioeconomic disadvantage, chronicity and melancholic or anxious features. However, if participants remained in treatment for up to four steps, about 67% reached remission. Times to remission were not substantially longer for later treatment steps. The importance of reaching remission is highlighted by the lower relapse rates in naturalistic follow-up for participants entering in remission compared with those entering with response but not remission (step 1: 34% vs 59%; step 2: 47% vs 68%; step 3: 42% vs 76%; step 4: 50% vs 83%). Clinical decision making based on the itemized measurement of symptoms and adverse effects at each treatment visit was feasible in STAR*D's real world settings and resulted in adequate dosages and durations of treatment that generally exceeded those typically found in practice settings. Although switch and augmentation strategies could not be directly compared due to the equipoise stratified randomized design, the higher remission rates at step 2 with medication augmentation are intriguing and merit further study.