心境障碍的下一代精准医学:可重复血液生物标志物实现客观诊断、亚型分型与靶向治疗
该研究通过跨诊断全转录组分析,采用微阵列和RNA测序两个平台及多个独立队列设计,寻找可重复的血液基因表达生物标志物以预测抑郁和躁狂状态及未来住院。研究发现抑郁、双相和躁狂的顶级生物标志物分别为FKBP1A/FOSL2、CTSB和ATP6V1C2,并基于性别特异性标志物面板生成个体化报告,在569例临床诊断为抑郁或双相障碍的患者中区分出8种亚型,且高达半数患者被重新分类,临床诊断与生物标志物评估不匹配与更差的未来住院结局相关。
要点
- 血液基因表达生物标志物可预测抑郁、躁狂状态及未来相关住院。
- 顶级抑郁标志物为FKBP1A和FOSL2,双相为CTSB,躁狂为ATP6V1C2。
- 生物标志物报告在569例中区分8亚型,半数患者分类与临床诊断不同。
核心结论
- 血液基因表达生物标志物可重复地预测心境状态和未来住院,并支持客观诊断、亚型分型和靶向治疗匹配。
- 临床诊断与生物标志物评估不匹配(尤其抑郁被误判为双相谱系)与更差的未来住院结局相关。
研究设计与样本
- 跨诊断精神病患者研究,从全转录组开始,分别研究低心境/抑郁和高心境/躁狂两个表型。
- 每个表型进行两项独立研究,每项研究使用微阵列和RNA测序两个平台,采用多个独立队列设计。
- 聚焦于每个表型在各研究内跨平台收敛且可重复的生物标志物,最终比较整合两个表型发现以识别双相标志物。
- 使用多证据评分(polyevidence score)排序顶级生物标志物,并在n=569例临床诊断为抑郁或双相障碍的患者中应用个体化报告。
主要发现
- 发现可预测抑郁和躁狂状态及未来相关住院的新及已知生物标志物。
- 多证据评分下,抑郁顶级血液基因表达标志物为FKBP1A和FOSL2,双相为CTSB,躁狂为ATP6V1C2。
- 抑郁顶级生物学通路与免疫应答相关,双相为凋亡,躁狂为坏死性凋亡。
- 在n=569例临床诊断为抑郁或双相障碍的患者中,报告区分8种亚型:抑郁障碍、双相2型、双相1型、双相3型、躁狂障碍,以及较 mild 形式心境恶劣、环性心境、情感高涨和无情境障碍。
- 高达半数患者被报告分类与临床诊断不同;临床诊断为抑郁但生物标志物提示双相谱系的患者未来因心境障碍加重、自杀倾向和酒精滥用住院增加,而反向不成立。
关键数值与推荐
- 抑郁顶级生物标志物:FKBP1A和FOSL2(基于多证据评分)。
- 双相顶级生物标志物:CTSB;躁狂顶级生物标志物:ATP6V1C2。
- 抑郁顶级通路:免疫应答;双相:凋亡;躁狂:坏死性凋亡。
- 抑郁顶级治疗匹配:omega-3脂肪酸、锂盐、沃替西汀;双相抑郁:锂盐、丙戊酸盐、氯氮平;躁狂:锂盐、丙戊酸盐、omega-3脂肪酸。
- 药物重定位识别天然化合物姜黄素和小檗碱为抑郁潜在治疗。
临床怎么用
- 基于性别特异性顶级血液生物标志物面板生成个体化患者报告,辅助诊断和个性化匹配潜在治疗。
- 在临床诊断为抑郁或双相障碍的患者中使用生物标志物报告进行亚型分型,识别临床诊断与生物标志物评估不匹配者。
- 对于临床诊断为抑郁但生物标志物提示双相谱系的患者,需警惕未来心境障碍加重、自杀倾向和酒精滥用相关住院风险增加。
- 结果提示心境稳定剂(如低剂量锂盐)可考虑更常作为心境障碍患者经验性一线选择,无论是否联用抗抑郁药,尤其考虑到其抗自杀特性。
局限性
- 原文未报告研究样本的总样本量、各队列具体样本量、随访时长、置信区间或P值等统计细节。
- 原文未报告生物标志物面板的具体基因列表、检测阈值或诊断准确性指标(如敏感度、特异度)。
- 原文未报告研究的具体局限性讨论。
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
原文
Mood disorders are highly prevalent in society, often disabling, and can lead to a reduced healthspan and lifespan, including by suicide. Depression and bipolar disorders are currently deemed to be chronic and/or treatment-refractory in up to half the patients treated. Additionally, depression and bipolar disorders need to be better distinguished, especially upon initial depressive presentations, as treatments can be quite different. Lack of widespread use of objective and/or quantitative information has hampered treatment and prevention efforts. We sought to uncover the underlying genomic and biological basis of mood disorders in a way that is reliable, comprehensive, and actionable. Blood biomarkers that track mood state can provide a window into the biology of mood, as well as could help with assessment and treatment of mood disorders. Previous studies by us were encouraging. Here we describe new studies we conducted trans-diagnostically in psychiatric patients, starting with the whole transcriptome, to expand the identification, prioritization, validation and testing of blood gene expression biomarkers for mood disorders, and their practical application. We first studied separately the two diametrical phenotypes, low mood/depression and high mood/mania. For each of the phenotypes, we carried out two separate studies, each of them on two different platforms, microarrays and RNA sequencing, using for each platform and study a multiple independent cohorts design. This was done to ensure both the biological and technical reproducibility of the final findings. We focused on biomarkers that were convergent and reproducible between the platforms in each study, for each phenotype. We found new as well as previously known biomarkers that were predictive of depression and of mania states, and of future hospitalizations related to them. We then compared and integrated the findings from the studies on the two phenotypes at the end, identifying bipolar biomarkers, that were changed in expression in opposite direction in the two phenotypes. Using a polyevidence score, the overall top gene expression blood biomarkers for depression were FKBP1A and FOSL2, for bipolar CTSB, and for mania ATP6V1C2. The top biological pathways for depression were related to immune response, for bipolar apoptosis, and for mania necroptosis. Top therapeutic matches for depression were omega-3 fatty acids, lithium, and vortioxetine; for bipolar depression lithium, valproate, and clozapine; and for mania lithium, valproate, and omega-3 fatty acids. Drug repurposing also identified the natural compounds curcumin and berberine as potential treatments for depression. We also illustrate how personalized patient reports for doctors based on gender- specific panels of top blood biomarkers would look like, which can aid with diagnosis and match patients in a personalized way to potential suggested treatments. Using such reports in n = 569 patients clinically diagnosed with depression or bipolar disorder, we distinguish 8 subtypes (depressive disorder, bipolar 2, bipolar 1, bipolar 3, and manic disorder, as well as the milder forms- dysthymia, cyclothymia, hyperthymia, and no mood disorder). Up to half of the patients were classified differently by our reports compared to their clinical diagnosis. Moreover, we demonstrate the clinical utility of our reports by showing that a mismatch between clinical diagnosis and biomarker- based assessment can lead to worse future hospitalizations outcomes. Specifically, patients who were (mis)diagnosed clinically as depression and were in fact on the bipolar spectrum based on biomarkers had increased future hospitalizations, not only for mood disorders exacerbation but also for suicidality and alcoholism. The converse was not true, patients who were (mis)diagnosed clinically as bipolar and were in fact on the depressive spectrum based on biomarkers fared well. Taken together, our results suggest that mood stabilizers, such as low-dose lithium, could be considered more often empirically first line in mood disorders patients, without or with the use of antidepressants, especially given the anti-suicidal properties. As mood disorders are highly prevalent, can severely affect quality of life, and lead to shortened lifespans, there is an urgent need for such insights, and for the added precision and personalization that biomarker tests can offer, to be applied to and improve clinical diagnosis, treatment, and prevention.
引用
- Balducci J, Gobel E, Schmitz M, et al. Next-generation precision medicine for mood disorders: reproducible blood biomarkers enable objective diagnostics, subtyping, and targeted therapeutics. Mol Psychiatry. 2026. PMID: 42547801. doi:10.1038/s41380-026-03736-8
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。