Abstract
Introduction
Adolescent Idiopathic Scoliosis (AIS) has been thought to be a relatively painless three dimensional (3D) deformity of the spine, but some studies has shown that the incidence of mild to moderate pain in AIS ranges from 25 to 50%. With the arrival of the new low dose radiation EOS technology, it is now possible to quantify intersegmental changes of the scoliotic spine in an upright position that could be related to pain.
Hypothesis
EOS 3D morphological analysis of AIS patients will better correlate with presence of pain than the Lenke classification.
Methods
Fifty-nine patients (7 male/52 female, mean age of 14.3 years old) who were scheduled for elective posterior spinal fusion with diagnosis of AIS. Preoperative clinical pain data recorded consisted of: 1) numerical visual analogue (VAS) pain scores quantifying patients average and worst pain over the preceding month; 2) SRS22 scores. Preoperative PA and lateral imaging were utilized to reconstruct and generate 3D models using the EOS software. Global and segmental intervertebral orientation in all three planes including the Da Vinci diagram identifying maximal deformity orientation were generated and correlated with patients' clinical presentation. In addition, standard curve magnitude, curve classification (Lenke) and pelvic parameters were also analyzed to observe their association to pain. Statistical analysis was performed with GraphPad Prism 6.
Results
Lenke classification and its subtypes did not correlate with pain, nor did any of the classic curve parameters, with the exception of the presence of hyper lumbar lordosis (>60 degrees) (r = 0.32, p = 0.06). Hyperlordotic patients reported greater pain intensity than normal lordotic patients (U = 81.50, p = 0.04). Additional new 3D parameters from both global and intersegmental vertebral orientations in space were investigated. With the exception of high intervertebral frontal, lateral tilt of L4 over L5, no 3D correlations with pain patterns were observed.
Conclusion
Despite the additional 3D morphological analysis generated by the EOS imaging, we were not able to identify anatomical characteristics associated to the pain experience reported by patients, with the exception of hyperlumbar lordosis.