皮質下缺血性中風患者之大腦皮質脊髓徑結構完整性與側下肢動作功能之相關性

Abstract

背景與目的：本研究主要在探討皮質下缺血性中風患者兩側大腦皮質脊髓徑(corticospinal tract, CST)之中期及長期(中風後20至30天內、第90天及第180天)纖維結構完整性與健康受試者之差異及其隨時間改變之情形；並探討此纖維結構完整性與患者患側下肢動作功能之相關性。法：本研究中五位單側皮質下缺血性之中風患者(平均63.1 ± 5.6歲)，皆分別於中風後20至30天內、第90天及第180天，接受一次臨床評估與擴散頻譜造影(diffusion spectrum imaging, DSI)檢查。以傅格-梅爾評估量表(Fugl-Meyer Assessment Scale, FMA)中下肢動作功能部分評估患者三測試時間之患側下肢動作功能障礙程度。另收入五位與中風患者年齡、性別及慣用腳配對之健康受試者(平均64.3 ± 4.2歲)為對照組，僅接受一次臨床評估與擴散頻譜造影檢查。擴散頻譜造影資料分析採特定神經束定量分析(tract-specific quantitative analysis)計算CST不同區段(大腦腳、內囊、內囊以上、大腦腳(含)以上等四區段之分析用於所有受試者；損傷區以下、損傷區段、及損傷區以上等三區段之分析僅用於中風受試者)之普擴散不等向性分數(generalized fractional anisotropy, GFA)與相對普擴散不等向性分數(relative GFA, rGFA)，以量測CST之纖維結構完整性。亦以興趣區域(region of interest, ROI)分析方式計算中風患者健、患側及健康受試者兩側內囊後肢(posterior limb of internal capsule)控制下肢動作區域之GFA值及其rGFA值，作為探討特定神經束定量分析之GFA值及rGFA值效度之參考。以獨立t檢定探討三時間點患者健側、患側CST之4個區段(大腦腳、內囊、內囊以上及大腦腳以上)特定神經束定量及內囊後肢興趣區域分析定量之GFA值與rGFA值，與健康受試者之值的差異。以雙因子(大腦側x中風後天數)重複測量變異數分析(two-way repeated measures of ANOVA)分析中風患者CST之7個區段(大腦腳、內囊、內囊以上、大腦腳以上、損傷區以下、損傷區及損傷區以上)特定神經束定量分析及依興趣區域分析所得之GFA值於三測試時間之差異。以單因子(中風後天數)重複測量變異數分析探討中風患者CST之7個區段特定神經束定量分析及依興趣區域分析所得之rGFA值隨時間之差異。以斯皮爾曼相關係數(Spearman correlation coefficient)探討三時間點之GFA值及rGFA值與同時期傅格-梅爾下肢動作功能評估量表結果之相關性，並探討在中風後30天內或中風後第90天時測得之GFA值與rGFA值是否可預測中風後第90天或中風後第180天測得之傅格-梅爾下肢動作功能評估結果。α值皆設為0.05。果：患者於三時間依特定神經束定量分析所得之GFA值及rGFA值與依興趣區域分析之相對應值皆具有中高度之相關(r= 0.6-0.9)。除中風後第180天時患側CST內囊及大腦腳區段外，其餘區段患側CST之GFA值皆較健康受試者之值顯著為低(p< 0.05)；患者健側CST之值則皆與健康受試者之值無顯著差異(p> 0.05)。中風後30天內及第90天患者內囊區段之rGFA值則較健康受試者之值顯著為高(p< 0.05)。中風患者患側CST大腦腳以上及損傷區以上等區段特定神經束定量之GFA值及其rGFA值隨時間呈現顯著變化(p< 0.05)，但在其他區段則無隨時間變化之趨勢。此外，在與臨床測試結果之相關性方面，主要發現在中風後第90天患側CST損傷區段特定神經束定量之GFA值與rGFA值及其內囊區段GFA值，皆與同時期測得之傅格-梅爾下肢動作功能評估量表分數呈現高度相關(r值依序為0.894、-0.894及0.894，p= 0.041)；於中風後第180天測得患側CST之特定神經束定量之損傷區段、大腦腳以上區段GFA值與其內囊區段GFA值及其rGFA值，皆與同時期測得之傅格-梅爾下肢動作功能評估量表分數呈現高度相關(r值依序為 0.894、0.918、0.918及-0.894，p= 0.041)。此外，中風後30天內測得CST之內囊區段特定神經束定量及內囊後肢興趣區域分析之rGFA值，皆與中風後第90天及第180天測得之傅格-梅爾下肢動作功能評估量表分數呈現高度負相關(r= -0.894，p= 0.041)；於中風後第90天測得患側CST之特定神經束定量損傷區段之GFA值及其rGFA值與其內囊區段GFA值，皆與中風後第180天測得之傅格-梅爾下肢動作功能評估量表分數呈現高度相關(r依序為0.894、-0.894及0.894，p= 0.041)。論與結論： 本研究結果顯示特定神經束定量分析內囊區段之GFA值及rGFA值與傳統興趣區域分析所得之值具有高度相關性，代表特定神經束定量分析之GFA值與rGFA值具有高效度(validity)。結果亦支持特定神經束定量分析損傷區及內囊區段之GFA值及rGFA值可反應皮質下缺血性中風患者患側CST之結構完整性，並與其患側下肢動作功能呈現高度同時期及預測之相關性。結果並未發現損傷區段結構參數值隨時間呈現顯著改變，可能因功能恢復較好與功能恢復較差之患者具有相反趨勢所致。未來將針對健康受試者進行DSI影像之重複測量信度檢測、中風患者患側相關上肢控制之CST結構完整性與其患側上肢動作功能之相關性、以及不同嚴重程度患者CST結構完整性隨時間之變化是否不同等議題再作進一步之探討。

Backgrond and Purposes：This study aimed to find the differences of the structural integrity of bilateral corticospinal tracts (CST) between persons with subcortical infarction at the mid- and long-term durations after stroke (< 20-30 days (D30), 90th day (D90) and 180th day (D180)) and healthy adults, the changes of structural integrity in stroke group between the three times, and the relationships between the integrity of the affected CST and the motor function of the affected lower extremity.ethods：Five hemiplegic patients (63.1 ± 5.6 yrs) received clinical assessments and the diffusion spectrum imaging (DSI) scan at D30, D90 and D180, respectively. The motor function of the affected lower extremity was evaluated for each patient by the lower-extremity motor part of the Fugl-Meyer Assessment scale (FMA-LE). Five healthy adults who matched stroke with age, gender and footedness (64.3 ± 4.2 yrs) received clinical assessments and the DSI scan one time. For DSI data analysis, the generalized fractional anisotropy (GFA) and the relative GFA (rGFA) values of different segments (cerebral peduncle (CP), internal capsule (IC), above IC and above CP for both groups; below lesion, lesion and above lesion for stroke group only) of bilateral CSTs were calculated by tract-specific (TS) quantitative analysis for each subject. We also used the traditional region of interest (ROI) analysis to calculate the GFA and the rGFA values of bilateral posterior limb of the internal capsule (PLIC) for each subject so that we could understand the validity of the GFA and rGFA values by TS analysis. Independent t test was applied for comparions of the GFA and rGFA values between stroke and healthy groups. Two-way (hemisphere x post-stroke day) repeated measures of ANOVA was used to investigate differences in the GFA values of different segments of bilateral CSTs by TS analysis and those by ROI analysis between the three testing times. One-way (post-stroke day) repeated measures of ANOVA was used to investigate differences in the rGFA values of the CST by TS analysis and those by ROI analysis between the three testing times. Spearman correlation coefficient was used for the correlations between these GFA and rGFA values with the FMA-LE at three testing times concurrently and the correlations between these GFA and rGFA values at D30 and D90 with the FMA-LE at D90 and D180, respectively. esults：The GFA and rGFA values by TS analysis highly related to those by ROI analysis in stroke at all three testing times (r= 0.6-0.9). The GFA values of the affected CST in stroke at all three testing time, except the CP and IC segments at D180 were significantly lower than those of bilateral CSTs in healthy adults (p< 0.05), but the GFA values of the unaffected CST were the same as those of healthy adults (p> 0.05). Only the rGFA values of the IC segment at D30 and D90 in stroke were higher than those in healthy adults (p< 0.05). The GFA and the rGFA values of above lesion and above CP segments of the affected CST of the stroke changed between different testing times (p< 0.05). The rGFA value of the IC segment of the CST by TS analysis and by ROI analysis at D30 highly related to the FMA-LE at D90 and D180, respectively (r= -0.894, p= 0.041). The GFA and rGFA values of the lesion segment and the GFA value of the IC segment of the affected CST by TS analysis at D90 highly related to the FMA-LE at D90 and D180, respectively (r= 0.894, -0.894 and 0.894, p= 0.041). iscussion and Conclusions：The results indicate that the GFA and rGFA values by TS analysis have high validity due to high correlations between those values by TS analysis and by ROI analysis. Structural integrity of the affected CST highly relates to the motor function of the affected lower extremity concurrently and predictively. No significant change of structural integrity of the lesion segment of the affected CST was found due to different trends between patients with better motor recovery and those with poor recovery. In future, we can measure the test-retest reliability of the integrity in healthy adults, and find relationships between integrity of the affected CST and motor functions of the affected upper extremity in stroke and the differences of the changes between stroke with different severity.