Risk Factors and Scoring System of Cage Retropulsion after Posterior Lumbar Interbody Fusion: A Retrospective Observational Study.

OBJECTIVE: To investigate risk factors of cage retropulsion after posterior lumbar interbody fusion (PLIF) in China and to establish a scoring system of cage retropulsion. METHODS: The retrospective analysis was based on two hospital databases. The medical data records of posterior lumbar interbody fusion with cage retropulsion were selected from August 2009 to August 2019. Inclusion and exclusion criteria were set in advance. Risk factors including patients' baseline demographics (age, gender, operation diagnosis time difference), preoperative neurological symptoms, whether the fusion involves single or double segments, screw type, intraoperative compression, preoperative bone mineral density, whether there are neurological symptoms before surgery, whether there is urine dysfunction before surgery, disease type, complete removal of the endplate, and patient's education level. The research endpoint was the retropulsion of fusion cages. The Kaplan-Meier (K-M) method was used to analyze potential risk factors, and multivariate Cox regression was used to identify independent risk factors (P < 0.05). The Statistical Package for the Social Sciences (version 22.0; SPSS, IBM, Chicago, IL, USA) software was used for statistical analysis, and univariate analysis was used to screen out the factors related to cage retropulsion. All independent risk factors were included to predict the survival time of the retropulsion of cage. RESULTS: This study included a total of 32 patients with PLIF between 2009 to 2019. All patients were residents of China. Univariate analysis showed that there were 13 patients over 60 years old and 19 patients under 60 years old. There were 20 male patients and 12 female patients. The surgical diagnosis time was seven patients within 1 month, 17 patients within 1 to 3 months, and eight patients over 3 months. The disease type was 18 cases of lumbar disc herniation, 10 cases of lumbar spinal stenosis, four cases of lumbar spondylolisthesis. The fusion segment was 18 cases of single segment, 14 cases of double segment. The intraoperative compression was seven cases of compression, 25 cases of no compression. The preoperative bone mineral density was 10 cases of low density, 18 cases of normal, four cases of osteoporosis. The screw type was 27 cases of universal screw, five cases of one-way screw. Preoperative neurological symptoms were found in 25 cases and not in seven cases. Preoperative urination dysfunction occurred in 8 cases, whereas 24 cases did not have this dysfunction. The endplate was completely removed in 10 cases and not in 22 cases. Education level was nine cases of primary school education, 10 cases of secondary school, 13 cases of university level. Cox regression analysis showed that intraoperative pressure (hazard ratio [HR] = 4.604, P = 0.015) and complete removal of the endplate (HR = 0.205, P = 0.027) are associated with the time of cage retropulsion. According to the HR of each factor, the scoring rules were formulated, and the patients were divided into the low-risk group, moderate-risk group, and high-risk group according to the final score. The three median survival times of the three groups were 66 days in the low-risk group, 55 days in the moderate-risk group, and 45 days in the high-risk group, with statistical significance (P < 0.05). CONCLUSION: Intraoperative pressure and complete removal of the intraoperative endplate can be helpful to evaluating the expected time of cage retropulsion in patients with PLIF, and this clinical model guided the selection of postoperative prevention and follow-up treatment.