巴金森氏病相關疼痛之腦白質微結構改變

Abstract

研究目的: 疼痛是巴金森氏病常見的非動作症狀，影響病患生活品質甚鉅。目前引起病患疼痛的機轉未明。磁振掃描擴散頻譜影像可偵測大腦白質細微結構的改變。我們藉由比較健康控制組、沒有疼痛的巴金森患者、有疼痛的巴金森患者三組受測者的腦部微結構改變，進而探討巴金森疼痛之致病機轉。 方法：這是一個橫斷型病例對照研究。研究共有三組受試者(健康控制組、沒有疼痛的巴金森患者、有疼痛的巴金森患者)。受測者接受巴金森動作障礙嚴重度、疼痛程度、憂鬱、焦慮等神經臨床評估，及擴散頻譜影像掃描。獲得影像後以全腦徑束自動分析(tract-based automatic analysis, TBAA)來重建全腦神經徑束，以比較三組間白質全束平均和段落的不等向性分率 (generalized Fractional anisotropy, GFA)。白質的GFA結果亦和臨床資料進行關聯性分析。 結果：三組受測者 (依序為健康對照組、無疼痛的巴金森患者、有疼痛的巴金森患者)各15人，三組在平均年齡 (53.9±5.5, 56.3±4.2, 55.5±5.8)；男女比 (6/9, 5/10, 6/9);得病時間(-, 3.7±2.0, 4.1±2.7)，動作嚴重度之Hoehn-Yahr分級(-, 1.53±0.51, 1.87±0.74) 等均無顯著差異。而在疼痛程度(1.9±5.5, 3.7±7.3, 45.0±29.6*)、憂鬱分數 (3.1±3.2, 4.7±3.9, 16.0±15.2*)、和焦慮分數 (1.5±3.1, 3.6±3.1, 15.4±10.5*)中，有疼痛的巴金森組都顯著高於另外兩組。 在白質微結構部分影像共4整束平均及20個段落有組間差異。將結果依白質相關腦區及功能分組後，顯示兩組巴金森受測者在體感覺皮質相關神經束都有GFA下降。疼痛的巴金森組在背外側前額葉 (dorsolateral prefrontal cortex)相關的神經束，及連接中央前迴的額葉紋狀體神經路徑 (frontostriatal tract of precentral gyrus)有GFA下降。而沒有疼痛的巴金森組在連接海馬迴的扣帶迴 (Cingulum_hippocampal part)有GFA下降。 討論：由上述結果提出巴金森疼痛的可能機轉如下。所有巴金森氏病患者都有體感覺處理系統的退化，使得感覺訊號傳入大腦時混雜許多雜訊。在無疼痛的巴金森患者中，出現連接海馬迴的扣帶迴的退化，可能使得感覺處理系統對雜訊不會過度敏感。且健康的背外側前額葉發揮感覺及疼痛調節的作用，使得患者不會感到自發的疼痛；而在有疼痛症狀的巴金森患者中，並沒有相對應的連接海馬迴的扣帶迴的退化，使得感覺處理系統對雜訊過度敏感，加上背外側前額葉失去功能無法發揮疼痛調節的作用，使得患者產生自發的疼痛。 結論：利用全腦徑束自動分析，我們找到在巴金森氏病的感覺系統、以及巴金森相關疼痛的白質微結構變化。我們藉此提出巴金森氏病可能造成的感覺及疼痛處理系統的改變，若缺少因應的保護性應結構變化，以及出現疼痛調節腦區的功能異常，可能導致巴金森相關疼痛。奠基於此研究，未來可對巴金森疼痛的相關機轉能有更深入的探討，也有助發展對應的治療。 關鍵字: 巴金森氏病、疼痛、擴散頻譜影像、全腦徑束自動分析

Objective: Pain is a frequent non-motor symptom in Parkinson’s disease (PD), and largely affects the quality of life of PD patients. The mechanism of pain in PD is still unclear. Diffusion spectrum imaging (DSI) is an MRI technique to exam the microstructural integrity of the cerebral white matter tract bundles. By comparing the microstructural change in healthy controls(HC), PD patients without pain (PD-noP) and PD subjects with pain (PD-P), we could clarify the mechanism of pain in PD. Method: This is a cross sectional case-control study. Three groups of subjects were recruited (HC, PD-noP, and PD-P). The subjects received neurologic examinations including the parkinsonism motor disability, the severity of pain, and the intensity of depression/ anxiety, and also received MRI scanning. After acquisition of the DSI data, whole brain tractography was reconstructed by a novel tract-based automatic analysis (TBAA). Mean GFAs and segmental GFAs were then derived for further intergroup comparison. Correlation analyses between GFAs and clinical parameters including were then conducted. Results: Each group included 15 subjects. There was no statistical difference among the three groups (HC, PD-noP, PD-P sequentially) in age (53.9±5.5, 56.3±4.2, 55.5±5.8), male/female ratio (6/9, 5/10, 6/9), disease duration (-, 3.7±2.0, 4.1±2.7), and Hoehn-and-Yahr stage (-, 1.53±0.51, 1.87±0.74). PD-P group had significantly higher score than the other 2 groups in the pain intensity (1.9±5.5, 3.7±7.3, 45.0±29.6*), depression score (3.1±3.2, 4.7±3.9, 16.0±15.2*), and anxiety score (1.5±3.1, 3.6±3.1, 15.4±10.5*). The microstructural study showed significant intergroup difference in the whole-tract mean GFAs in 4 tracts and cluster GFAs in 20 segments. After grouping the data according to the connecting brain regions or subgroup function, it showed that both PD-P and PD-noP had decreased GFAs in the somatosensory cortices related tracts; the PD-P group had decreased GFA in dorsolateral prefrontal cortex (DLPFC) related tracts and frontostriatal tracts of precentral gyrus; the PD-noP group had decreased GFAs in the cingulum of hippocampal part. Discussion: We proposed a possible mechanism of PD-related pain. All PD subjects have degeneration in primary sensory processing system, and which leads to noise in sensory inputs. In PD subjects without pain, corresponding degeneration in cingulum_hippocampus occur, and may avoid the sensitization of the noisy sensory disturbance. Preserved DLPFC function also works on pain modulation, and all above-mentioned together keeps the subjects away from having spontaneous pain. On the other hand, in the PD subjects with pain, there is no corresponding alteration in cingulum_hippocampus, and which results in the sensitization of noisy sensory inputs. At the same time, weakened DLPFC failed in pain modulation. Pain is then developed after serial of structural and functional change. Conclusions: Using the whole brain TBAA method, we found altered white matter microstructure property in sensory processing system in PD and PD related pain. We proposed possible changes in sensory and nociception processing system in PD, and lack of protective structural change with subsequent dysfunction in pain modulatory areas further leads to PD related pain. Based on our study, the future investigation of the mechanism of PD related pain could be better specified, and could be helpful in developing effective treatments.