Purpose: We examined our preliminary results in a series of nine patients treated for thoracolumbar callus deformitis using a technique associating simultaneous anterior and posterior approaches and in situ contourning.

Material and methods: The series included seven women and two men, mean age 42 years operated on after January 2001. The patients had deformed callus after fractures (n=8) or spondylodiscitis (n=1). Surgical treatment was used initially for five of the fracture patients. The deformed callus involved the thoracolumbar junction in 56% of the patients. Mean follow-up was 14 months (6–22). The same surgical technique was used in all nine patients by two surgery teams. The patient was positioned in lateral decubitus. After posterior arthrectomy and anterior osteotomy, the correction was obtained by combined anterior distraction and lordosis contourning of the posterior material. An intercorporeal graft was encastrated anteriorly.

Results: Preoperative regional kyphosis was 30°. It was 4° postoperatively and 5° at last follow-up. Kyphosis improved in 87% of patients. There was no neurological aggravation. The main complication was posterior infection with aggravation of the regional kyphosis to 10° in one patient.

Discussion: Posterior or anterior spinal approach, alone or in combination have been proposed for callus deformitis of the spine. Results in the literature have shown moderate and incomplete correction of the kyphosis.

Conclusion: The proposed technique allows good reduction of the deformed callus with results that appear to persist with time.

Purpose: There is growing concern about the effect of anterior fusion of the cervical spine on the adjacent levels. Long-term assessment is indispensable to understand the mechanisms involved in the degradation observed and to support the development of materials preserving discal mobility.

Material: Three hundred patients who underwent cervical arthrodesis were reviewed in 1996 forty months after the procedure for physical examination and an x-ray work-up including stress views. Cervical spine and radicular pain were assessed on a visual analogue scale.

Methods: A complete data set was available for 136 patients and a partial set for 34. Twenty-two patients only accepted a phone interview. The clinical outcomes in these three groups were not significantly different and the mean scores for these three groups were in the general average in 1996. Eight patients had died.

Results: Patients were divided into three groups by type of disease diagnosed preoperatively (trauma, degenerative spine, myelopathy). Mean follow-up was 102.5 months (range 84 – 180 months).

Trauma: Among the 42 patient reviewed again in 2001, mean deterioration in the subjacent segment increased from 21% in 1996 to 69% in 2001. Deterioration of the supraja-cent segment increased from 26% to 47.6%. Cervical pain remained moderate (20/100 in 1996 and 27/100 in 2001). Degenerative spine (root compression requiring simple discectomy or with arthrodesis or single-level corporectomy): Among the 42 patients reviewed again in 2001, deterioration of the subjacent segment increased from 57% in 1996 to 89% in 2001. Deterioration of the suprajacent segment increased from 22% to 41%. Cervical pain increased from 14/100 in 1996 to 41/100 in 2001.

Myelopathy: Among the 52 patients reviewed again in 2001, deterioration of the subjacent segment increased from 54% in 1996 to 81% in 2001 when there had been one or two corporectomies and from 40% to 70% beyond two. Deterioration of the suprajacent segment increased from 26% to 50%. Cervical pain remained moderate (18/100 in 1996 and 23/100 in 2001).

Conclusion: Although a statistical analysis was not possible because of the small number of patients and the large percentage lost to follow-up, these results confirm that fusion of the cervical spine accelerates the degradation of adjacent levels. Longer follow-up demonstrates that the trauma group “catches up” with the degenerative group.

Use of mobile materials should enable differentiating between effects related to the degenerative process and those induced by the arthrodesis.

Purpose: Access to the cervicothoracic junction is difficult both via a posterior and via an anterior approach. Tumour localisations or more rarely trauma however require access. Using the posterior approach, anterior decompression is limited by the narrow access and the vulnerability of the cord. Anterior reconstruction is impossible. Using the pure anterior approach, fixation and decompression of the caudal component is limited. Preoperative MRI shows the respective position of the manubrium sternal and the diseased vertebra, allowing a clear surgical strategy. To avoid sternotomy or even partial cleidectomy, both causes of postoperative pain and complications, we developed a medial sternal resection maintaining the stability of the sternoclavicular joints and allowing spinal decompression by corporectomy to T3 and fixation to T4.

Material and methods: A left anterolateral cervical approach was used to avoid injury to the recurrent nerve. This is a classical cervical approach generally used for access to C7-T1. It is prolonged caudally a few centimetres on the mid line to reach the anterior aspect of the sternum. After section of the sternohyoid, sternothyroid and scapulohyoid muscles, the three upper centimetres of the sternum are resected with a microdrill over a width of two centimetres. This give direct access to the anterior walls of T3 and T4. The lower limit of the exposure is described by the aortic arch (except in patients with severe kyphosis). The left brachiocephalic venous trunk is the crucial element situated just horizontally behind the sternum and protected by fat and fibrous tissue. It is important to release this trunk precautiously because injury at this level is difficult to suture and would require ligature (this is still possible if necessary but would lead to oedema of the left arm by defective drainage). After releasing the vein, the resection of the posterior wall of the sternum is completed with a Kerrison gouge. This gives a U-shaped groove that does not destabilise the sternoclavicular articulations and allows retraction of the vessels to expose the vertebral bodies. Screw fixation of T4 is possible, generally with slightly descending screws. The classical closure method is used.

Results: We have operated 13 patients with tumours or fractures using this approach (five T4, seven T3, one T2). Corporectomy was performed above T4. This approach did not lead to any direct complications. Postoperative pain was considered to be less than with sternotomy or cleidectomy, approaches we have now abandoned. Use of the endoscope improves visibility but the incision cannot be smaller because of the axe required for screwing. The important features of this method are the correct analysis of the preoperative relation between the target vertebra and the manubrium sternal and the dissection of the left brachiocephalic venous trunk.