NEW TREATMENTS FOR REFRACTORY AFFECTIVE ILLNESS

在锂或卡马西平单一疗法的双盲、随机、为期一年的试验中，只有四分之一的双相情感障碍门诊患者显示出足够的临床预防反应。即使联合使用锂和卡马西平，这些门诊患者中也只有50%有反应。在额外使用丙戊酸盐的情况下，40%的患者仍然没有反应。正是从这一治疗难治的患者群体中，该处寻求更好地了解复发性单相和双相情感障碍的不同病理生理机制，并开发新的治疗方式。该分部的主要新治疗举措是使用重复的经颅磁刺激(RTMS)脑。在一项先导性研究中，前6名患者中有2名在左侧额叶皮质运动阈值的80%处进行了20赫兹刺激(George等人，1995年，NeuroReport)。一项双盲、随机、交叉试验表明，与假治疗组相比，活性rTMS治疗两周有显著的抗抑郁作用(George等人，1998年，Am J精神病学)。下一项研究现已完成，评估了在80%的运动阈值下，左侧额叶皮质对低频(1赫兹)和高频(20赫兹)rTMS刺激与假刺激的不同反应性。15名受试者的这些数据表明同一患者对这些不同的频率有不同的反应。此外，那些具有基线低代谢模式的人倾向于对20赫兹的刺激做出反应，而那些具有基线高代谢模式的人更有可能对1赫兹的刺激做出反应(Kimbrell等人，1998)。由于临床反应性的发生率和大小不足以满足许多患者的需要，已经启动了第四项使用更高强度(100%运动阈值)的研究(18名患者参加)(Speer等人，1998年)。这项研究重复了个体患者不同响应性的发现，发现20赫兹刺激增加了0-15个血流量，而1赫兹rTMS减少了血流量。一项在正常志愿者中进行的进一步对照研究证实，额叶皮质上的1赫兹rTMS可诱导双侧额叶和纹状体代谢的相对减少。一项重要的住院药理学研究涉及一项双盲随机试验，该试验使用增强抑制性GABA能功能的药物(加巴喷丁[GPN])与任何降低兴奋性谷氨酸能功能的药物(拉莫三嗪[LTG])与安慰剂进行为期六周的治疗，患者交叉使用其他药物治疗，以确定不同的临床反应。这项涉及37名患者的研究得到了《普通精神病学文献》的好评，发现LTG(53%的应答率)明显优于GPN(27%)和安慰剂(22%)(Frye等人，1998年)。初步证据表明，0-15PET的基线灌注模式与临床对这两种药物的反应相互作用。应答者在基线水平较低，并随着治疗的进行而增加，而无应答者在正常范围内，并显示下降。一项关于T3、TRH和文拉法辛安慰剂强化的双盲随机试验正在检验TRH是否可以加速抗抑郁药的起效速度。在与预测因素的关系上，初步证据表明，在PET上表现为整体高代谢的抑郁症患者，特别是在左侧脑岛，更有可能对卡马西平有反应(N=26)，而那些额叶和左侧胰岛代谢较典型的患者更有可能对二氢吡啶L型钙通道阻滞剂尼莫地平有反应。我们发现，尼莫地平可增加脑脊液(CSF)中生长抑素的水平，而基线水平脑脊液中生长抑素水平较低的患者更有可能对尼莫地平诱导的这些升高产生临床反应。因此，该分支机构已经开创了一些有前途的、机械上新颖的治疗方法，并将进一步努力确定rTMS反应的最佳参数，并阐明这种反应的临床和神经生物学标志物。这项工作以及为难治性双相患者开发新治疗方法的相关工作也正在更广泛的基础上进行，建立了第一个由NIMH-Stanley基金会支持的双相治疗成果网络，在美国和欧洲分别有多个地点和一个地点。这个临床试验网络解决了NIMH 1989和1994年关于双相情感障碍的会议的大部分建议。

Only one-quarter of our bipolar outpatients show an adequate clinical prophylactic response in double-blind, randomized, one-year trials of lithium or carbamazepine monotherapy. Even with the combination of lithium and carbamazepine, only 50% of these outpatients respond. With the additional use of valproate, 40% of patients still remain unresponsive. It is from this pool of treatment-refractory patients that the Branch seeks to better understand the differential pathophysiological mechanisms in recurrent unipolar and bipolar affective disorders and develop new therapeutic modalities. The major new treatment initiative in the Branch is the use of repeated transcranial magnetic stimulation (rTMS) of the brain. Two of the first six patients responded in a pilot study of 20 Hz stimulation at 80% of motor threshold of left frontal cortex (George et al, 1995, NeuroReport). A double-blind, randomized, crossover trial indicated significant antidepressant effects of active rTMS for two weeks compared with the sham (George et al, 1998, Am J Psychiatry). The next study, now completed, assessed the differential responsivity to low-frequency (1 Hz) versus higher frequency (20 Hz) rTMS vs. sham stimulation over left frontal cortex at 80% of motor threshold. These data in 15 subjects suggest differential responses within the same patient to these different frequencies. Moreover, those with a pattern of baseline hypometabolism tend to respond to the 20 Hz stimulation, while those with baseline patterns of hypermetabolism are more likely to respond to the 1 Hz stimulation (Kimbrell et al, 1998). As the incidence and magnitude of clinical responsivity was not adequate for many patients, a fourth study using higher intensities (100% of motor threshold) has been initiated (18 patients enrolled) (Speer et al, 1998). This study replicated the findings of differential responsivity within individual patients and revealed that 20 Hz stimulation increases 0-15 blood flow while 1 Hz rTMS decreases it. A further controlled study in normal volunteers has confirmed that 1 Hz rTMS over frontal cortex induces relative decrements in bilateral frontal and striatal metabolism on PET.A major inpatient pharmacological study involves a double-blind, randomized trial of six weeks of treatment with an agent that enhances inhibitory GABAergic function (gabapentin [GPN]), versus any that decreases excitatory glutamatergic function (lamotrigine [LTG]), versus placebo, with patients crossing over to the other drug treatments in order to ascertain differential clinical response. This study, involving 37 patients, has been favorably reviewed by the Archives of General Psychiatry and found significant benefit of LTG (53% response rate) over GPN (27%) and placebo (22%) (Frye et al, 1998). Preliminary evidence suggests that baseline patterns of perfusion on 0-15 PET interacted with clinical response to both of these agents. Responders were low at baseline and increased with treatment, while nonresponders were in the normal range and showed decreases. A double-blind, randomized trial of T3 versus TRH versus placebo augmentation of venlafaxine is examining whether TRH can accelerate the rapidity of antidepressant onset. In relationship to predictors, preliminary evidence indicates that depressed patients with global hypermetabolism on PET, especially in the left insula, are more likely to be responsive to carbamazepine (N = 26), while those with the more classic pattern of frontal and left insular hypometabolism are more likely to be responsive to the dihydropyridine L-type calcium channel blocker nimodipine. We have found that nimodipine increases somatostatin in cerebrospinal fluid (CSF) and that those with lower CSF somatostatin at baseline tend to be more likely to respond clinically to these nimodipine-induced increases.Thus, a number of promising and mechanistically novel treatment approaches have been pioneered in the Branch and additional effort will be aimed at defining optimal parameters for rTMS response and the elucidation of clinical and neurobiological markers of such response. This and related work on the development of new treatment approaches for refractory bipolar patients is also being pursued on a wider basis with the establishment of the first NIMH-Stanley Foundation-supported Bipolar Treatment Outcome Network with multiple sites in the U.S. and one in Europe. This clinical trials Network addresses most of the recommendations of the NIMH 1989 and 1994 meetings on bipolar illness.