Stand-alone posterior surgery posterior VCR is the recommended technique, as this offers more advantages compared with combined approaches [6]. With VCR, there is no thoracotomy and postoperative thorax wall scarring leading to thoracic deformities, which might impair the development of the lungs. For older children (above 10 years of age), the recommendations for treating deformities become increasingly similar to those for adults. The posterior approach allows for all the necessary steps for safe vertebral resection and correction. It allows continuous visual control of the neural structures during resection and correction. A costotransversectomy on one or on both sides in the thoracic area, or removal of the transverse processes in the lumbar area, provides good visualization and a capacious working area. This approach also allows for safe circumferential preparation around the vertebral body and control of the segmental vessels. One drawback is the difficulty in controlling the major vessels in the prevertebral area. In some cases, a combined anteroposterior or postero-anteroposterior approach cannot be avoided [5, 16, 17].

A conventional radiolucent orthopaedic surgery table is used, with the patient being placed in a prone position on specially made Maquet-like adjustable modular foam pillows that act as a frame. Children under halo-gravity traction are positioned with a traction weight of 1–3 kg.

Technically, there are small differences compared with the vertebral resections described elsewhere [5, 6]. In contrast to the usual subperiosteal preparation, preservation of the periosteal layer is attempted to avoid unnecessary spontaneous fusion and allow for unaltered or minimally altered growth. This also allows for conversion to a growing rod system or even for implant removal to regain mobility at a later stage (Figs. 33.3 and 33.4). This is followed by verification of the correct levels using fluoroscopy, usually in the PA projection. The next step is to identify the entry points of all planned pedicle screws by using anatomical landmarks. With the standard technique, a pedicle finder or probe is used to identify and prepare the pedicle followed by tapping or using a self-tapping screw. Alternatively, according to our practice, and especially in small pedicles, a 22 Gauge needle is first inserted followed by fluoroscopic verification in anteroposterior and lateral views. Then the pedicle is prepared using a 1.5 mm drill. Pre-bent K-wires are inserted into the holes, and fluoroscopic imaging is performed. Fine adjustment of the screw trajectories can be made in the next step using a 2.0 or 2.5 mm drill for final preparation. In most cases, pedicle screws with a diameter of 3.5 mm are inserted, though depending on the diameter of the pedicles, 2.7, 4.0, 4.75 or 5.00 mm diameter screws may also be used. Screws are preferable to hooks or wires because of their better stability and thus superior potential to achieve and maintain the correction (primary stability) [18, 19]. It also seems that pedicle screws in the growing spine do not result in clinically significant alteration in the development of the spinal canal [20]. The placement of pedicle screws at strategically important sites prior to vertebral resection is of major importance, since it allows spinal alignment to be controlled and the neural structures to be protected during the unstable phase of the resection and correction.

It is helpful to mark the exiting nerve roots using a vessel loop. A single nerve root may be sacrificed in the thoracic area between T3 and T9. However, a deficiency of T4 and/or T5 root function may lead to sensory loss in the mammary area, which may cause problems later in life (e.g. with breastfeeding).