An analysis of clinical risk factors for adolescent scoliosis caused by spinal cord abnormalities in China: proposal for a selective whole-spine MRI examination scheme.

Background: Approximately 80% of adolescent scoliosis cases are idiopathic, and some non-idiopathic scoliosis cases caused by spinal cord abnormalities are misdiagnosed as idiopathic scoliosis. This study examined the risk factors for non-idiopathic scoliosis with intramedullary abnormalities, explored the feasibility of whole-spine MRI, and provided a theoretical basis for the routine diagnosis and treatment of adolescent idiopathic scoliosis.

Method: The clinical data of adolescent scoliosis patients who were admitted to Shanghai Tongren Hospital and Shanghai Changhai Hospital between July 1, 2013, and December 31, 2018, were reviewed. According to the whole-spine MRI results, the patients were divided into either the idiopathic group or the intramedullary abnormality group. Sex, age, main curvature angle, main curvature direction, kyphosis angle, scoliosis type, coronal plane balance, sagittal plane balance, abdominal wall reflex, sensory abnormality, ankle clonus and tendon reflexes were compared between the two groups. Student's t test was used to evaluate the differences in the continuous variables, and the chi-square test was used to evaluate the differences in the categorical variables. Fisher's exact test was applied to detect the difference in the rate of intraspinal anomalies between the groups. Logistic regression was used to evaluate the correlation between the multivariate risk factors and intramedullary abnormalities.

Result: A total of 714 adolescent scoliosis patients with a mean age of 13.5 (10-18 years) were included in the study, and intramedullary abnormalities were found in 68 (9.5%) patients. There were statistically significant differences in the incidence rates of intramedullary abnormalities between males and females, left and right thoracic curvatures, angular scoliosis and smooth scoliosis, and abnormal abdominal wall reflex and ankle clonus (P < 0.01). Logistic regression showed that the ratios for sex, scoliosis direction, scoliosis type, abdominal wall reflex and ankle clonus were 2.987, 3.493, 4.823, 3.94 and 8.083, respectively. The ROC curve showed a sensitivity of 66.18% and a specificity of 89.01%, and the Youden index corresponding to the optimal critical point was 0.5519.

Conclusion: Risk factors associated with adolescent scoliosis caused by abnormal intramedullary abnormalities included male sex, thoracic scoliosis on the left side, sharp curvature of the spine, abnormal abdominal wall reflex and ankle clonus. In adolescent scoliosis patients, the incidence of scoliosis caused by intramedullary abnormalities was approximately 9.5%. These clinical indicators suggest that there is a high-risk adolescent scoliosis population who should undergo whole-spinal MRI preoperatively to rule out intramedullary abnormalities.