A Scoring System For Differential Diagnosis Of Spinal Metastases And Spinal Tuberculosis

Objectives:To develop a scoring system for differential diagnosis of spinal metastases(SM)and spinal tuberculosis(STB)to improve clinical diagnostic accuracy.Methods:The clinical records of 82 cases of SM and 59 cases of STB who were diagnosed by pathological examination from 2014 to 2016 in our department were retrospectively analyzed.The following clinical data were recorded:age,gender,past history,erythrocyte sedimentation rate(ESR),C-reactive protein(CRP),number of the vertebral lesions,the distribution of vertebral lesions,the involved vertebral element,the intervertebral disc was involved or not,if there was a sequestra formation and if there was a soft tissue component.Univariate logistic regression analysis was used to screen the main points of differential diagnosis.The independent differential diagnosis factors which were regard as the item of the scoring system were determined by the results of multivariate logistic regression analysis.The score of each item is based on the relative size of the odds ratio(OR).The optimal cut-off value of the scoring system is determined by the receiver operating characteristic(ROC)curve.Results:Univariate logistic regression analysis showed that age>50 years old,malignant tumor history,ESR<65 mm/h,CRP<13 mg/L,one or three more vertebral lesions,isolated or skipped vertebral lesions,vertebral posterior elements destroyed,preserved intervertebral discs,no sequestra formation and no paravertebral soft tissue component were the main distinguishing points between SM and STB.Multivariate logistic regression analysis showed that malignant tumor history,isolated or skipped vertebral lesions,vertebral posterior elements destroyed,preserved intervertebral discs and no sequestra formation were the independent differential factors and were respectively weighted as 2 points,1 point,2 points,3 points and 3 points according to the OR.The total score rank from 0 to 10 points.SM showed a statistic significant higher score than did STB(8.12 vs.2.00;t=-18.598;P<0.001).The ROC curve showed optimal cut-off value was 4 points.Spinal lesions with the score of 5 to 10 would be diagnosed of SM and the score of of 0 to 4 were STB.The sensitivity and specificity were 91.15%and 95.52%,respectively.Conclusion:The scoring system shows satisfactory diagnostic efficiency in the differential diagnosis of SM and STB.This scoring system still needs further revision and improvement.