基于影像学参数分析单节段腰椎内镜下融合术后对侧神经根痛发生因素阈值的临床研究

Clinical study of risk factors and thresholds for contralateral radicular pain following single-level endoscopic lumbar fusion based on imaging parameters

中文关键词:  脊柱内镜  腰椎椎间融合术  侧隐窝  椎间孔  神经根痛

英文关键词:Spinal endoscopy  Lumbar interbody fusion  Lateral recess  Intervertebral foramen  Radiculalgia

中文摘要:

  【摘要】 目的：探究内镜下腰椎椎间融合术后发生对侧神经根痛的解剖学预测因素，并分析术前风险预测的临床相关阈值。方法：回顾性分析45例行内镜辅助下腰椎椎间融合术治疗的L4-5单侧下肢症状的腰椎管狭窄症患者，其中男性15例，女性30例，年龄35~74岁（45.0±9.5岁）。依术后是否出现对侧神经根痛分为症状组（15例）和无症状组（30例）。所有患者术前行高分辨率CT扫描，通过Mimics软件测量L4-5节段对侧骨性侧隐窝总体积、对侧骨性侧隐窝分区（2a、2b、2b-E、3、3-E）体积、对侧椎间孔体积。基于传统腰椎侧方椎管华西分型构建多因素二元Logistic回归预测模型一，以及在此基础上进一步限定以椎弓根最内侧缘水平分界（2b-E、3-E）的改良分区方法构建多因素二元Logistic回归预测模型二。对模型内部两组间的体积参数进行比较，采用二元Logistic回归分析确定侧隐窝及椎间孔体积的预测权重。通过受试者工作特征（receiver operating characteristic，ROC）曲线分析确定预测的最佳体积阈值，探讨影像学解剖指标与内镜下腰椎椎间融合术后对侧神经根症状发生的相关性，筛选独立危险因素，并评估其预测价值。结果：症状组与无症状组比较，骨性侧隐窝总体积（593.09±303.91mm3 vs 773.19±246.32mm3）、椎间孔体积（1085.65±314.14mm3 vs 1440.16±384.30mm3）、骨性侧隐窝分区2a体积（48.52±33.24mm3 vs 105.42±63.85mm3）、骨性侧隐窝分区2b体积（222.51±148.22mm3 vs 304.72±112mm3）、骨性侧隐窝分区2b-E体积（100.34±38.34mm3 vs 212.35±124.98mm3）、侧隐窝矢状径（5.86±2.24mm vs 7.65±2.83mm）均有统计学差异（P＜0.05），侧隐窝横径、骨性侧隐窝分区（3、3-E）无统计学差异（P＞0.05）。二元Logistic分析发现，模型一侧隐窝2a区（OR 0.975，95%CI 0.953-0.998，P=0.032），模型二侧隐窝2b-E区（OR 0.973，95%CI 0.952-0.994，P=0.012）、椎间孔体积（OR 0.997，95%CI 0.994-0.999，P=0.019）与内镜下腰椎椎间融合术后对侧神经根症状显著相关，且为独立危险因素。ROC分析结果显示，模型一中侧隐窝2a区体积有统计学意义上的预测价值（AUC=0.789，95%CI=0.651-0.926，P=0.002），对应最佳临界点40.88mm3；模型二中侧隐窝分区2b-E区体积及椎间孔体积对于预测术后神经根症状均具有统计学意义上的预测价值（P<0.05），其中侧隐窝2b-E区体积的预测效能最佳（AUC=0.818，95%CI=0.695-0.941，P=0.001），对应最佳临界点152.27mm3。与模型一的侧隐窝2a区（AUC=0.789）相比，模型二侧隐窝2b-E（AUC=0.818）在预测术后对侧神经根症状方面具有更优的预测效能。结论：术前对侧侧隐窝上部区域（2a、2b-E）及椎间孔体积减少与内镜下腰椎椎间融合术后对侧神经根症状的发生密切相关。基于改良解剖分区构建的Logistic回归预测模型，侧隐窝2b-E区体积具有最佳预测效能，其预测能力优于传统分区模型及椎间孔体积。

英文摘要:

  【Abstract】 Objectives: To investigate the anatomical predictors of contralateral radicular pain after endoscopic lumbar interbody fusion(Endo-LIF) and to analyze the clinically relevant preoperative risk prediction thresholds. Methods: A retrospective analysis was conducted on 45 patients(presenting with unilateral lower-limb symptoms) treated with Endo-LIF for lumbar spinal stenosis at L4-5. The cohort included 15 males and 30 females, aged 35-74 years(45.0±9.5 years). According to the presence or absence of postoperative contralateral leg pain, the patients were divided into a symptomatic group(n=15) and an asymptomatic group(n=30). All patients underwent preoperative high-resolution CT scan. Mimics software was used to measure the total volume of the contralateral bony lateral recess at L4-5, the volumes of its subregions(2a, 2b, 2b-E, 3, and 3-E), and the volume of the contralateral intervertebral foramen. A multivariable binary logistic regression model(Model 1) was established based on the traditional Huaxi classification of the lateral lumbar spinal canal. Subsequently, a modified subregional classification using the medial border of the pedicle as the anatomical boundary to define the 2b-E and 3-E subregions was applied to construct a second multivariable binary logistic regression model(Model 2). Volumetric parameters were compared between the two groups, and binary logistic regression analysis was performed to determine the predictive contributions of lateral recess and foraminal volumes. Receiver operating characteristic(ROC) curve analysis was used to determine the optimal volumetric thresholds for prediction. The associations between imaging-based anatomical parameters and the occurrence of contralateral radicular symptoms after Endo-LIF were investigated to identify independent risk factors and evaluate their predictive values. Results: Compared with the asymptomatic group, the symptomatic group showed significantly smaller total bony lateral recess volume(593.09±303.91mm3 vs 773.19±246.32mm3), intervertebral foramen volume(1085.65±314.14mm3 vs 1440.16±384.30mm3), subregion 2a volume(48.52±33.24mm3 vs 105.42±63.85mm3), subregion 2b volume(222.51±148.22mm3 vs 304.72±112mm3), subregion 2b-E volume(100.34±38.34mm3 vs 212.35±124.98mm3), and sagittal diameter of the lateral recess(5.86±2.24mm vs 7.65±2.83mm), with statistically significant differences(P<0.05). No significant differences were observed in the transverse diameter of the lateral recess or in the volumes of subregions 3 and 3-E(P>0.05). Binary logistic regression analysis revealed that the lateral recess subregion 2a in Model 1(OR=0.975, 95%CI=0.953-0.998, P=0.032), the lateral recess subregion 2b-E in Model 2(OR=0.973, 95%CI=0.952-0.994, P=0.012), and intervertebral foramen volume(OR=0.997, 95%CI=0.994-0.999, P=0.019) were significantly associated with contralateral radicular symptoms after Endo-LIF and were identified as independent risk factors. ROC curve analysis demonstrated that the volume of lateral recess subregion 2a in Model 1 had significant predictive value(AUC=0.789, 95%CI=0.651-0.926, P=0.002), with an optimal cutoff value of 40.88mm3. In Model 2, both the volume of lateral recess subregion 2b-E and the intervertebral foramen volume showed significant predictive value for postoperative radicular symptoms(P<0.05). Among these, the volume of lateral recess subregion 2b-E demonstrated the best predictive performance(AUC=0.818, 95%CI=0.695-0.941, P=0.001), with an optimal cutoff value of 152.27mm3. Compared with the lateral recess subregion 2a in Model 1(AUC=0.789), subregion 2b-E in Model 2(AUC=0.818) demonstrated superior predictive performance for postoperative contralateral radicular symptoms. Conclusions: Preoperative reduction in the volume of the upper lateral recess on the contralateral side(subregions 2a and 2b-E) and the intervertebral foramen was closely associated with the occurrence of contralateral radicular symptoms following Endo-LIF. In logistic regression predictive models based on a modified anatomical subregional classification, the volume of subregion 2b-E demonstrated the highest predictive performance, surpassing that of both the conventional subregional model and the intervertebral foramen volume.