心率指標用於重度憂鬱症患者接受重複經顱磁刺激治療療效預測生物指標之潛在角色

Abstract

背景/目的：重複經顱磁刺激（repetitive transcranial magnetic stimulation, rTMS）是一種非侵入性神經刺激技術，針對大腦背外側前額葉皮質（dorsolateral prefrontal cortex, DLPFC）進行刺激，已被美國食品藥物管理局（US FDA）核准用於治療難治型憂鬱症（treatment-resistant depression, TRD）。憂鬱症患者若無心臟節律相關疾病，其典型的生理症狀之一，就是心跳頻率（簡稱心率）（heart rate, HR）上升以及心率變異度（heart rate variability, HRV）下降。初步的人體研究指出，以TMS刺激大腦的DLPFC部位，會透過自律神經系統（autonomic nervous system, ANS）調節的機制，降低心率並提升心率變異度。然而，目前尚不清楚心率相關指標是否具備預測 rTMS 治療成效的潛力，仍需進一步研究探討。本研究的第一目標為評估 rTMS 治療對重度憂鬱症（major depressive disorder, MDD）或是難治型憂鬱症（treatment-resistant depression, TRD）患者HR與HRV變化的影響；第二目標為找出最佳預測治療成效（efficacy）的心率變化之指標。 方法：本研究納入隨機對照試驗之介入組及常規神經刺激治療接受rTMS或間歇性西塔波叢集磁波刺激（intermittent theta burst stimulation, iTBS）治療的重度憂鬱症或難治型憂鬱症患者。從刺激開始前一分鐘起至當次治療結束期間，以指尖型光體積描記法之穿戴裝置（photoplethysmography, PPG）連續記錄心率資料。於治療前、後以 MADRS (Montgomery-Åsberg Depression Rating Scale)、HAM-D-17 (17-item Hamilton Depression Rating Scale) 或BDI-II (Beck Depression Inventory-II) 等三種評估憂鬱症狀嚴重程度，並蒐集人口學與臨床基線資料。為探討 HR 與 HRV 指標與治療成效之關聯，研究採用邏輯斯迴歸分析，建立未調整模型與控制年齡、性別後的調整模型，評估各指標與治療反應（response）或緩解（remission）之相關性。進一步以接受者操作特徵曲線（receiver operating characteristic, ROC）方法分析，評估各指標前後變化對治療反應的預測能力，並鑑別出最具預測潛力之指標，做為未來臨床應用與進一步研究的基礎。 結果：本研究共納入47位受試者，在2023年十二月到2025年五月期間於台北的兩家醫院，其中女性34位、男性13位，平均年齡41.51歲（標準差：13.28）。結果顯示治療後憂鬱量表分數顯著下降，完成rTMS或iTBS治療的整個療程（sessions）時，整體治療反應率為57.45%。其中，在最早一次的開始治療前15秒至開始治療後60秒區間中HRV的低頻與高頻功率比值（LF/HF ratio）於年齡性別調整模型下達到最高曲線下面積（area under curve, AUC）值（0.69，95%信賴區間為0.52–0.85，p < 0.05），敏感度為0.56，特異度為0.89，陽性預測值（positive predictive value, PPV）為0.88，為最佳指標。 結論：以心率資料計算出來的指標作為生物指標（biomarker）可能有助於預測rTMS的治療反應，反映出磁刺激治療目標區域是否有成功被活化，未來可望成為發展出個別化定位與優化臨床實用性rTMS治療方案之依據。

Background/Aims: Repetitive transcranial magnetic stimulation (rTMS), a non-invasive neurostimulation technique targeting the dorsolateral prefrontal cortex (DLPFC) in the brain, has emerged in recent decades as a US FDA-approved treatment for treatment-resistant depression (TRD). One of the typical physiological symptoms of depression is an increase in heart rate (HR) and a decrease in heart rate variability (HRV) for patients without heart rhythm-related diseases. Preliminary evidence given by human research has suggested that DLPFC stimulation may modulate the autonomic nervous system (ANS), potentially reducing HR and increasing HRV. However, the predictive value of heart rate profile (HR and HRV indicators) in rTMS treatment outcomes for depression remains unclear and warrants further investigation. The first aim is to evaluate the impact of rTMS on changes of HR and HRV indicators before and after the start of the earliest TMS treatment in patients with MDD/TRD involved in the intervention group of ongoing RCTs or TMS treatment in regular practice. The second aim is to identify optimal cut-off points for HR and HRV changes that best predict therapeutic outcomes of depression. Materials and Methods: Participants were recruited from the intervention arms of randomized controlled trials or regular neurostimulation treatments (rTMS or iTBS) in regular practice of MDD or TRD treatment to assemble a clinical cohort between December 2023 and May 2025 at two medical centers in Taipei. Heart rate data were collected using fingertip photoplethysmography (PPG) during the earliest treatment session, with consecutive recording from one minute before stimulation start to session end. Depression severity was assessed at baseline and post-treatment using the Montgomery-Åsberg Depression Rating Scale (MADRS), 17-item Hamilton Depression Rating Scale (HAM-D-17), or Beck Depression Inventory-II (BDI-II). Baseline demographics and clinical features were also recorded for each participant. Logistic regression (crude and adjusted for age and sex) was used to evaluate these heart rate profile indicators measured in various period combinations in relation to response and remission, defined as meeting criteria on any of the above three depression severity scales. Receiver operating characteristic (ROC) analysis was conducted to evaluate the predictive performance of physiological changes and to identify candidate indicators of target engagement. Results: A total of 47 participants, including 34 females and 13males with a mean age of 41.51 years old (SD = 13.28), were enrolled in the study. The results showed that depression rating scales significantly decreased after the completion of stimulation treatment sessions, with an overall response rate of 57.45%. The low-frequency to high-frequency power ratio of HRV (LF/HF ratio) at 15–60s combination periods in age- and sex-adjusted model achieved the highest AUC (0.69, 95% CI: 0.52–0.85, p < 0.05), with a sensitivity of 0.56, specificity of 0.89, and positive predictive value (PPV) of 0.88. Conclusions: Heart rate profile indicators calculated by heart rate data may aid as biomarkers in predicting rTMS treatment response, offering potential for individualized target engagement and refining the development of more efficient and clinically feasible rTMS protocols.