目的:基于过伸性颈脊髓损伤的影像学特征建立新的临床分型并验证其临床指导意义。方法:采用回顾性病例系列研究分析2009年4月至2016年12月同济大学附属同济医院收治的88例过伸性颈脊髓损伤患者临床资料，其中男68例，女20例；年龄18～87岁[(52.9±14.8)岁]。根据颈椎椎管退变状态将过伸性颈脊髓损伤进行分型(Cheng氏分型)：无退变性椎管狭窄为A型，单纯椎间盘突出为B型，椎间盘突出合并骨赘增生为C型，椎管内韧带骨化为D型。根据脊髓信号改变和椎间盘韧带复合体(DLC)损伤情况各型又分为3个亚型：无脊髓信号改变或DLC损伤为A1、B1、C1、D1型，伴脊髓信号改变但无DLC损伤为A2、B2、C2、D2，伴脊髓信号改变及DLC损伤为A3、B3、C3、D3型。分析不同分型的基本特征及脊髓信号改变与DLC损伤节段的一致率。A1、A2型采用非手术治疗(非手术组，22例)，A3型与B、C、D型采用手术治疗(手术组，66例)。比较两组治疗前和末次随访时美国脊髓损伤协会(ASIA)评分、ASIA损伤分级(AIS)和日本骨科学会(JOA)评分。根据是否有颈椎椎管退变，将患者分为无退变类(A型)和退变类(B、C、D型)，比较两类患者的年龄及治疗前和末次随访时ASIA评分和JOA评分。结果:本组A型24例，B型22例，C型34例，D型8例，其中退变性患者占73%(64/88)，以B、C、D型为主。B、C型占退变患者的88%(56/64)，并以中老年患者为主。脊髓信号改变与颈椎退变状态有关，而与DLC损伤节段的一致率为40%。所有患者随访1～9.1年[4(1，6)年]。非手术组治疗前与末次随访时ASIA运动评分为90(88，96)分、100(100，100)分，感觉评分为216(212，221)分、224(224，224)分；AIS治疗前22例均为D级，末次随访时均为E级；治疗前与末次随访时JOA评分分别为13(12，14)分、17(17，17)分。手术组治疗前与末次随访时ASIA运动评分为76(62，86)分、98(94，100)分，感觉评分为204(191，212)分、220(212，224)分；AIS治疗前A级4例、B级3例、C级18例、D级41例，末次随访时C级3例、D级51例、E级12例；治疗前与末次随访JOA评分分别为10(7，11)分、16(14，17)分。两组末次随访时ASIA评分、AIS、JOA评分均较术前明显改善(n P0.05)。退变类患者年龄[(58.4±11.7)岁]显著高于无退变类患者[(38.1±11.9)岁](n P<0.01)。无退变类患者治疗前ASIA感觉及运动评分、JOA评分显著高于退变类患者(n P<0.01)。末次随访时无退变类及退变类患者ASIA感觉及运动评分均较治疗前显著改善(n P<0.05)。退变类患者末次随访时C型JOA评分优于D型(n P0.05)。n 结论:根据颈椎椎管退变状态、DLC损伤、脊髓信号改变建立了过伸性颈脊髓损伤影像学分型。中老年患者颈椎退变程度越重，DLC损伤发生率越高，脊髓损伤程度也越重，更容易残留神经功能障碍。根据该分型选择治疗方法，患者脊髓功能改善明显，提示该分型可有效用于指导临床治疗。“,”Objective:To innovate a classification of hyperextension cervical spinal cord injury according to the radiological characteristics and to verify its clinical significance.Methods:A retrospective case series study was performed to analyze the clinical data of 88 patients with hyperextension cervical spinal cord injury admitted to Tongji Hospital Affiliated to Tongji University between April 2009 and December 2016. The patients included 68 males and 20 females, aged 18-87 years [(52.9±14.8)years]. Hyperextension cervical spinal cord injury was classified by the degree of degeneration (Cheng's classification): type A (non-degenerative spinal stenosis), type B (disc herniation alone), type C (disc herniation along with osteophyte) and type D (ossification of posterior longitudinal ligament). Further, all types were divided into 3 subtypes based on disc-ligmentous complex (DLC) injury and intramedullary signal intensity change: without high-intensity intramedullary signal or DLC injury as subtypes A1, B1, C1 and D1), with high-intensity intramedullary signal but without DLC injury as subtypes A2, B2, C2 and D2, with high-intensity intramedullary signal and DLC injury as subtypes A3, B3, C3 and D3. The concordance rate between intramedullary signal intensity change and DLC injury segment as well as the basic characteristics of different types were analyzed. Types A1 and A2 were treated with non-surgical treatment (22 patients, non-surgical group). Types A3, B, C and D were treated operatively (66 patients, surgical group). The American Spinal Injury Association (ASIA) score, ASIA impairment scale (AIS) and Japanese Orthopedic Association (JOA) score were assessed before treatment and at final follow-up. According to status of cervical spinal canal degeneration, the patients were further divided into two kinds including non-degeneration (type A) and degeneration (types B, C, D). The age and ASIA score and JOA score before treatment and at final follow-up were compared between two kinds of patients.Results:In all, 4 patients were classified as type A, 22 as type B, 34 as type C and 8 as type D. The majority was patients with degeneration (type B, C, D), accounting for 73% (66/88), of which type B and C accounted for 88% (56/64). Most patients were middle-aged and aged. Intramedullary signal intensity changes were related to the degeneration of the cervical spine, with concordance rate with DLC injury segment for only 40%. The average follow-up duration was 4(1, 6)years (range, 1-9.1 years). In non-surgical group, the ASIA motor score before treatment and at final follow-up were 90(88, 96)points and 100(100, 100)points respectively, the ASIA sensory score before treatment and at final follow-up were 216(212, 221)points and 224(224, 224)points respectively, the AIS in all patients was grade D before treatment and became grade E at final follow-up, the JOA score before treatment and at final follow-up were 13(12, 14)points and 17(17, 17)points respectively. In surgical group, the ASIA motor score before treatment and at final follow-up were 76(62, 86)points and 98(94, 100)points respectively; the ASIA sensory score before treatment and at final follow-up were 204(191, 212)points and 220(212, 224)points respectively; the AIS was grade D in 4 patients, grade B in 3, grade C in 18 and grade D in 41 before treatment and was grade C in 3 patients, grade D in 51 and grade E in 12 at final follow-up; the JOA score before treatment and at final follow-up were 10(7, 11)points and 16(14, 17)points respectively. The ASIA score, AIS and JOA score at final follow-up in both groups were significantly better than that before treatment (n P0.05). The average age of patients with degeneration [(58.4±11.7)years] was significantly higher than that of patients without degeneration [(38.1±11.9)years] (n P<0.01). The ASIA sensory and motor scores and JOA score of patients without degeneration before treatment were significantly higher than those with degeneration (n P<0.01). At final follow-up, the ASIA sensory and motor scores of patients with or without degeneration were all significantly improved compared with those before treatment (n P<0.05). At final follow-up, the patients with degeneration showed that the JOA score of type C was better than that of type D (n P0.05).n Conclusions:The radiological classification of hyperextension cervical spinal cord injury is established based on the degeneration of the cervical spinal canal, DLC injury and intramedullary signal intensity change. For middle-aged and elderly patients, the more severe the degeneration of the cervical spine, the higher the incidence of DLC injury, the more severe the degree of spinal cord injury, and the more likely to have residual neurological dysfunction. The treatment method is selected according to this classification and the patients' spinal cord function is improved significantly, suggesting that this classification can be effectively used to guide clinical treatment.