The first cases of osteotomy reported by Smith-Petersen et al. in 1945 were for post-ankylosing spondylitis deformity. This technique (Fig. 49.3) consists of posterior element subtraction, taking away the spinous process, lamina, and facet joints on one or more levels. Correction is obtained through external manipulations resulting in compression of the posterior elements, opening of the disk space, and tearing of anterior longitudinal ligament. The major drawback of this technique is that it lengthens the anterior column, potentially injuring the spinal cord and tearing the great vessels. Lengthening of the anterior column has been measured by Scudese and Calabro who noted 1.7 cm of lengthening for a 40° correction [11]. Moreover, closure of the posterior elements may compress the foramen, causing radiculopathy. The average correction that can be expected is 5–10° per osteotomy level [12].

The Ponte-type osteotomy was first described by Ponte et al. in 1984 for Scheuermann kyphosis although the terminology only entered the US literature in 2007 [13]. The Ponte osteotomy was described as wide segmental osteotomies followed by posterior compression along unfused regions of the deformity in Scheuermann patients.

The authors are mostly speaking of multilevel facet and lamina local osteotomy that by their multilevel situation corrects the sagittal deformity. The surgical technique is a multilevel reshaping of the canal associate to an inferior and medial facetectomy, with ligamentum flavum resection (Fig. 49.4).

Although today the terms Smith-Petersen and Ponte are often used interchangeably, the technique currently used is Alberto Ponte’s. In addition, SPOs have become a mainstay in the correction of coronal deformities, such as adolescent idiopathic scoliosis; however, they were not originally described for this indication.

Corporeal subtraction osteotomy consists of a wedge osteotomy of the vertebral body performed through a posterior approach (Fig. 49.5). It allows shortening of the posterior column without lengthening of its anterior aspect, thus sparing the neural elements. Moreover, it gives a larger surface of compressive bone contact than the Smith-Petersen technique, thereby enhancing stability and fusion.

Comparison between Smith-Petersen osteotomy (SPO) and pedicle subtraction osteotomy (PSO) underscores major differences. The SPO group tends to produce more coronal decompensation and requires multiple levels of osteotomy for substantial correction [13]. There is a higher risk of vascular injury as well as a higher rate of nonunion because the opening is through the disk space. The bulk of axial loads in the standing position passes through the anterior column, so opening of the disk space predisposes to nonunion [14]. A PSO on the other hand is associated with a significantly higher blood loss and a higher union rate [15, 16].

These results and complications explain our surgical evolution and desire to minimize bleeding and to simplify the osteotomy procedure. The development of the posterior wedge intracorporeal impaction osteotomy is to us a good way to obtain such results.

The main principles of this technique are to accomplish the posterior closing osteotomy by intracorporeally impacting the cancellous bone of the posterior vertebral body wedge. The apex of the osteotomy must be as close as possible to the anterior cortex while preserving a hinge.

The patient is placed in the prone position on an Allen® radiolucent carbon surgical table (Fig. 49.6a, b). The chest is supported by two thoracic mobile stands (up and down) while the iliac crest and pelvis are positioned on a fixed one; the fixed lower limbs are positioned along the body’s axis. This setup leaves the abdomen free from compression, thereby decreasing bleeding, and favors the natural restoration of the lumbar lordosis by hyperextending the lumbar spine.

A posterior incision is made. Muscle must be handled with care because of its importance in spinal stability and injuries caused by previous surgery. Dissection is gently done with a knife. Knife dissection is preferred because it decreases the risk of dural tears, especially if a laminectomy was performed on previous procedures, and of muscle necrosis in multi-operated cases. Monopolar cautery may be used to dissect soft tissue off any spinal instrumentation. The surface anatomy is thoroughly cleaned of all soft tissue in order to identify essential anatomical landmarks before starting the osteotomy. Intraoperative radiographic control is used to verify the level of the selected vertebra. Anatomical limits of bone resection, depending on the amount of correction, are represented by two lines drawn with monopolar cautery on either side, perpendicular to the spinal axis of pedicles screw holes easily visible after instrumentation removal. Instrumentation length is tailored to each individual case, taking various factors into account (quality of bone, adjacent pathologies, topping of syndrome, nonunion, etc.). A minimum of two levels over and under the osteotomy level is recommended. Bone resection consists of a laminectomy at the selected level and cephalad vertebral level including the target level pedicles, facet joints, and transverse processes. The neural elements, i.e., the nerve roots above and below the pedicle of interest, are protected by cotton pads. The pedicle subtraction is performed using a high-speed diamond drill. This results in a large foramen whose contents include the two exiting nerve roots, maintained in distraction using a distraction clamp (Fig. 49.6a, b). Bleeding from epidural veins can be controlled by haemostatic agents (Surgicel®, Surgiflo®, or others) or bipolar cautery if necessary. Corporeal osteotomy starts on the posterior wall, using two osteotomes. The extent of posterior body resection is determined by the obliquity required to affix the osteotomy margins just behind the anterior vertebral cortex, leaving an anterior bony hinge that prevents forward translation. Osteotome progression inside the vertebral body is controlled by fluoroscopy (Fig. 49.7).

There are two major differences between the PIO and PSO techniques at this step of the procedure. With the PSO technique, it is recommended to progressively remove the cancellous bone from the vertebral body, creating a wedge-shaped void giving place for future closure and correction. With the PIO technique, the cancellous bone, as well as the posterior wall of the vertebral body, is impacted inside the vertebral body. Impacting the bone instead of removing it decreases intraoperative bleeding. It also simplifies the procedure. Bone impactors have been designed specifically for this purpose. They are long enough to deal with obese patients and narrow enough to be easily handled up to the anterior extent of the osteotomy (Fig. 49.8). The cancellous vertebral body bone is progressively impacted out of the osteotomy site, alternating on both sides. This is the first step of the osteotomy procedure.

The second step of wedge creation is the cutting of the lateral vertebral body wall. In the PSO technique, you elevate the lateral soft tissues and then cut the lateral wall with an osteotome. Control of the lumbar segmental vessels is difficult and local bleeding can be voluminous. In PIO you weaken the lateral edges of the vertebral body with the bone impactor. The tip of the impactor is manipulated from inside the vertebral body, weakening the lateral wall without injuring the lumbar vessels.

The last step of the decancellation is the resection of the vertebral median posterior wall just in front of the dural sac. It is the last hurdle to closing the wedge. The resistance of this median posterior cortical wall depends on the patient’s bone quality. In most cases the same weakening technique used to deal with the lateral wall is employed. A smaller bone impactor is used. The instrument is inserted obliquely from both sides in front of the dural sac and a progressive impaction maneuver will break it down.

The wedge is now complete, bleeding is controlled by cancellous bone impaction, and posterior closure is obtained by progressively releasing the distraction clamp (Fig. 49.9a, b). Satisfying lumbar lordosis is obtained in most cases at this point. In stiffer patients, in order to complete the posterior wedge closure, it is necessary to perform an upward translation of the thoracic platform and a hyperextension at the site of the osteotomy by raising the lower limbs until the upper and lower osteotomy limits contact bone on bone. A thorough evaluation of the posterior decompression is assessed to make sure there is no dural impingement or foraminal compression. Careful inspection of each foramina and nerve root is mandatory all along the closure. Root entrapment or impingement is the most frequent possible complication. The procedure is completed by local decortication and posterolateral graft application using locally harvested morsellized bone chips. Mechanical stability is obtained by strict patient fixation to the operative table, thus avoiding instrumentation at this time of the procedure. Spinal fixation is then performed spanning at least two levels above and two levels below the osteotomy. This is possible due to the stability of the patient on the operative table. We use rigid instrumentation, favoring highly rigid chromium cobalt alloy rods to titanium ones. Finally, lumbar lordosis is obtained using the operative table. The desired posture is maintained by placing screws in previous pedicle trajectories and using pre-contoured adapted rods (Fig. 49.9a, b). There is no need to perform in situ bending. Two suction drains are placed into the epidural space and the wound is closed in layers. The patient is kept supine for the first day postoperatively and then mobilized on the second post-op day with a rigid Boehler-type brace which is used for walking and standing for the first 3 months. Sitting is not allowed for the first 45 days for osteotomies lower than L3.

The complete resection of the pedicle is the standpoint of PSO or IPO techniques; in some cases where the needed correction is lower than 20°, it is possible to only remove the lower two-third of the pedicle. It avoids violating the upper foramina, thus decreasing morbidity, and shortens the surgical time (Fig. 49.10).

The choice Ponte osteotomy and PSO/PIO mostly depends on deformity reducibility. Correction of the sagittal imbalance is assessed on lateral full spine X-rays (EOS® system if possible) and flexion-extension films (Fig. 49.11a, b).

Multilevel deformities are more often addressed by a Ponte osteotomy. Stiff deformities will need PSO or PIO; reducible ones will be addressed with Ponte osteotomy.

The axial extension of the deformity is also considered. A localized deformity is more efficiently dealt with by a PSO or PIO (Fig. 49.12).

Multiple factors (type and level of pathology, local modifications) weigh on the choice of the osteotomy level. The indication is tailored to each pathology and deformity.