Objectives:To investigate the relationship between the improvement of lumbar symptoms and imaging parameters in patients with tandem spinal stenosis(TSS)primarily manifesting as cervical spondylotic myelopathy(CSM)after initial cervical decompression surgery,and to explore related imaging predictive factors.Methods:A retrospective analysis was conducted on 69 TSS patients who underwent primary cervical decom-pression surgery,with an average age of 64.3±10.5 years(ranging from 41 to 86 years old)and a follow-up period of 33.8±5.5 months(ranging from 24 to 48 months).Preoperative symptoms and signs,Nurick gait clas-sification,and preoperative and final follow-up Japanese Orthopaedic Association(JOA)scores for both the cer-vical and lumbar spine were recorded and analyzed.The patients were divided into an improvement group(n=37)and a non-improvement group(n=32)based on the improvement conditions of lumbar spine JOA scores at the final follow-up.Imaging parameters were measured including spinal cord compression ratio,cross-sectional area of the dural sac at the narrowest point of the cervical spinal cord,the ratio of the vertebral canal to the vertebral body and the actual spinal canal width from C3 to C7,and the actual width of the vertebral canal from L1 to L5.The conditions of spinal canal stenosis were evaluated according to the grading system of spinal canal stenosis of cervical and lumbar spine proposed by Lee et al,and the number of cervical verte-brae with a stenosis score ≥1,score at the narrowest part of cervical spinal canal,and the total score of cervical stenosis,as well as the number of lumbar vertebrae with a stenosis score ≥1,score at the narrowest part of lumbar spinal canal,and the total score of lumbar stenosis were calculated.Intergroup comparisons were performed using t tests,chi-square tests,and Mann-Whitney U tests.For the statistic data with statisti-cal differences between the two groups,receiver operating characteristic(ROC)curve was used to determine op-timal thresholds for each parameter,and the area under the ROC curve(AUC)and its corresponding 95％ con-fidence interval(CI)were calculated.Multivariate logistic regression analysis was conducted to identify radiolog-ical predictive factors for non-improvement of lumbar symptoms in TSS patients.Results:The non-improvement group was significantly higher than the improvement group in the total score of lumbar stenosis(5.00±1.68 vs 2.68±1.23,P＜0.001),the number of lumbar vertebrae with a stenosis score≥1(2.47±0.84 vs 1.86±0.95,P=0.004),and the prevalence of redundant nerve roots(14/18 vs 6/13,P=0.017).The AUC of the total lumbar stenosis score was 0.864(P＜0.001,95％CI 0.779-0.950),with an optimal threshold of 3.5(sensitivity:81.3％;specificity:75.7％).The AUC of the number of lumbar vertebrae with a stenosis score ≥1 was 0.691(P=0.007;95％CI 0.565-0.817),with an optimal threshold of 1.5(sensitivity:87.5％;specificity:56.8％).Multivariate logistic regression showed that the number of lumbar spinal stenosis＞1.5(OR=1.493;95％CI 0.392-5.686;P=0.557)and presence of redundant nerve roots(OR=2.815;95％CI 0.740-10.711;P=0.129)had no significant relationship with improvement of lumbar symptoms.The total lumbar stenosis score＞3.5 was significantly related with improvement of lumbar symptoms(OR=10.983;95％CI 3.261-36.994;P＜0.001),which was an independent risk factor for non-improvement in lumbar symtoms after initial cervical decompression in TSS patients.Conclusions:When the total score of lumbar spinal stenosis exceeds 3.5 in TSS patients,the possibility of improvement in lumbar symptoms after initial cervical decompression is small.

Objectives:To investigate the relationship between the improvement of lumbar symptoms and imaging parameters in patients with tandem spinal stenosis(TSS)primarily manifesting as cervical spondylotic myelopathy(CSM)after initial cervical decompression surgery,and to explore related imaging predictive factors.Methods:A retrospective analysis was conducted on 69 TSS patients who underwent primary cervical decom-pression surgery,with an average age of 64.3±10.5 years(ranging from 41 to 86 years old)and a follow-up period of 33.8±5.5 months(ranging from 24 to 48 months).Preoperative symptoms and signs,Nurick gait clas-sification,and preoperative and final follow-up Japanese Orthopaedic Association(JOA)scores for both the cer-vical and lumbar spine were recorded and analyzed.The patients were divided into an improvement group(n=37)and a non-improvement group(n=32)based on the improvement conditions of lumbar spine JOA scores at the final follow-up.Imaging parameters were measured including spinal cord compression ratio,cross-sectional area of the dural sac at the narrowest point of the cervical spinal cord,the ratio of the vertebral canal to the vertebral body and the actual spinal canal width from C3 to C7,and the actual width of the vertebral canal from L1 to L5.The conditions of spinal canal stenosis were evaluated according to the grading system of spinal canal stenosis of cervical and lumbar spine proposed by Lee et al,and the number of cervical verte-brae with a stenosis score ≥1,score at the narrowest part of cervical spinal canal,and the total score of cervical stenosis,as well as the number of lumbar vertebrae with a stenosis score ≥1,score at the narrowest part of lumbar spinal canal,and the total score of lumbar stenosis were calculated.Intergroup comparisons were performed using t tests,chi-square tests,and Mann-Whitney U tests.For the statistic data with statisti-cal differences between the two groups,receiver operating characteristic(ROC)curve was used to determine op-timal thresholds for each parameter,and the area under the ROC curve(AUC)and its corresponding 95％ con-fidence interval(CI)were calculated.Multivariate logistic regression analysis was conducted to identify radiolog-ical predictive factors for non-improvement of lumbar symptoms in TSS patients.Results:The non-improvement group was significantly higher than the improvement group in the total score of lumbar stenosis(5.00±1.68 vs 2.68±1.23,P＜0.001),the number of lumbar vertebrae with a stenosis score≥1(2.47±0.84 vs 1.86±0.95,P=0.004),and the prevalence of redundant nerve roots(14/18 vs 6/13,P=0.017).The AUC of the total lumbar stenosis score was 0.864(P＜0.001,95％CI 0.779-0.950),with an optimal threshold of 3.5(sensitivity:81.3％;specificity:75.7％).The AUC of the number of lumbar vertebrae with a stenosis score ≥1 was 0.691(P=0.007;95％CI 0.565-0.817),with an optimal threshold of 1.5(sensitivity:87.5％;specificity:56.8％).Multivariate logistic regression showed that the number of lumbar spinal stenosis＞1.5(OR=1.493;95％CI 0.392-5.686;P=0.557)and presence of redundant nerve roots(OR=2.815;95％CI 0.740-10.711;P=0.129)had no significant relationship with improvement of lumbar symptoms.The total lumbar stenosis score＞3.5 was significantly related with improvement of lumbar symptoms(OR=10.983;95％CI 3.261-36.994;P＜0.001),which was an independent risk factor for non-improvement in lumbar symtoms after initial cervical decompression in TSS patients.Conclusions:When the total score of lumbar spinal stenosis exceeds 3.5 in TSS patients,the possibility of improvement in lumbar symptoms after initial cervical decompression is small.