The Preoperative Radiological Risk Factors For Predicting Recurrent Herniation In Petd And Construction Of The Predictive Models

Objective: To investigate the value of preoperative radiological characteristic factors in predicting recurrent lumbar disc herniation(r LDH)within 1 year after percutaneous endoscopic transforaminal lumbar discectomy(PETD),and establish a series of prediction models to provide a reliable reference for clinical.Methods: This study was a retrospective case-control study.After preliminary screening and deletion of lost follow-up,a total of 1210 patients with single-segment L4/5 lumbar disc herniation who received PETD surgery from January 2013 to November 2019 were collected.A total of 62 patients were diagnosed with r LDH based on the clinical and imaging data obtained during the 1 year follow-up.Due to the differences in the local structure of the lumbar spine and the degree of intervertebral disc degeneration,the mechanisms of r LDH in young patients and elderly patients may be very different.We therefore performed subgroup analyses by age and the receiver operating characteristic(ROC)was used to determine the age threshold.Considering the large number of non-recurrent patients and the huge internal variability among them,a propensity score matching(PSM)was performed to minimize the influence of selection bias and to control potential confounding factors.The PSM analysis matched three variables of sex,BMI,and preoperative symptom duration.Of the 1148 non-r LDH patients,224 patients were matched and included.The preoperative imaging parameters of the two groups of patients were measured and recorded,including disc herniation type,Modic changes,Pfirrmann grade of disc degeneration,lumbosacral transition vertebra(LSTV),endplate concave angle(ECA),lumbar lordosis(LL),sacral slope(SS),disc height index(DHI),retrolisthesis(posterior spondylolisthesis),flexion/extension intervertebral angle(IVA),and sagittal range of motion(s ROM)),the degree of upper/nether vertebral compression,etc.The relationship between risk factors and r LDH was assessed in two subgroups by univariate and multivariate analysis,and predictive models were established based on the logistic regression analysis.The evaluation of the models was done with ROC,calibration plots,and decision curve analysis(DCA).Results: The logistic regression analysis identified six important predictors associated with r LDH in the young group: superior endplate concave angle,sacral slope,Modic change,sagittal range of motion(s ROM),hyperextension intervertebral angle(IVA),and lumbar lordosis.Four significant predictors were identified in the elderly group: disc height index,retrolisthesis,s ROM,and hyperextension IVA.Validation of both models demonstrated strong discrimination(AUC= 0.940 and 0.931,respectively),and both were better than the ungrouped overall analysis(AUC=0.895).Decision curve analysis confirmed the excellent clinical utility of this set of models.Conclusion: The predictive models we constructed can accurately predict the recurrence risk of individual patients.For the young patients,the inflammatory reaction of the intervertebral discs and the disturbance of the local microenvironment are the main threats to r LDH,while for the elderly patients,poor maintenance of the disc height and biomechanical-related changes,such as retrolisthesis,increase the risk of r LDH.Excessive intervertebral activity is a high risk factor for recurrence in patients of all ages.Therefore,individualized measures should be taken for patients of different ages with the above risk factors,and a tailored postoperative rehabilitation plan should be developed for patients undergoing PETD.