Spinal deformity in adults is one of the most frequently treated pathologies in spinal surgery. Their surgical treatment remains a challenge. In particular, the choice of the borderline vertebrae for thoracolumbar fusion correction remains controversial, including the choice between an instrumented last vertebra at L5 and an arthrodesis extending to the pelvis. Studies and meta-analyses show that an ilio-sacral socket provides better correction of pelvic parameters and avoids the development of distal junctional syndrome. On the other hand, there is more proximal junctional syndrome in patients fixed to the sacrum, and the sacral socket exposes patients to the risk of pseudarthrosis of the L5-S1 level and failure of the fixation hardware. Moreover, although this factor is not described in the studies, it would seem that an ilio-sacral socket is more stiffening. Overall, no study has shown a statistically significant difference in clinical scores between patients whose last instrumented vertebra is L5 and those with pelvic fixation.

One of the theoretical risks of an extended fixation with a proximal thoracic grasp is to observe a premature disc degeneration of the L5-S1 disc since it would be subjected to more stress. This degeneration could be the cause of distal junctional syndrome and lumbo-radiculalgia due to inflammatory disc disease and disco-radicular impingement. However, to Investigator's knowledge, no clinical study has yet investigated the medium- and long-term fate of the L5-S1 disc in patients with extended fusion correction of a thoracic vertebra to L5.

Investigator's team has sometimes favored a final instrumented vertebra at L5 when possible, i.e., when the L5-S1 disc was considered to be minimally or not degenerative and when its horizontality could be obtained after correction of the underlying curvature. Investigators therefore have a cohort of patients who have undergone this technique. The usual follow-up of these patients includes a consultation appointment every 2 to 5 years and a control MRI at a distance of more than 5 years: investigators thus have a control MRI and the clinical score of the patients.

Investigator's objective is to determine whether there is degeneration of the L5-S1 disc, using MRI imaging data, which has never been published to investigator's knowledge.

Inclusion Criteria:

• Patient whose age is ≥ 18 years
• Spinal fixation extended from the thorax to L5,
• Performed between 2007 (15 years of recoil and beginning of systematic preoperative MRI) and 2017 (5 years of recoil minimum)
• French speaking patient

Exclusion Criteria:

• Neuromuscular scoliosis
• Indication of fixation for fracture of ankylosed spine (impossibility of analyzing the L5-S1 disc)
• Contraindication to the realization of an MRI except if the patient has been operated on for a complementary fixation of L5-S1 in the aftermath
• Patient under guardianship or curatorship (unless consent is provided for this purpose)
• Patient deprived of liberty
• Patients who oppose participation in the study

Patient whose age is ≥ 18 years, with spinal fixation extended from the thorax to L5, performed between 2007 (15 years of recoil and beginning of systematic preoperative MRI) and 2017 (5 years of recoil minimum).