Abstract
Introduction: Some authors (Suk, Barr, Hamill ...) showed that lumbar and thoracic pedicle screws provided adequate reduction of scoliosis. Quality of reduction depends on primary stability of the vertebral anchors. If the anchor has a good primary stability, reduction forces are entirely transferred to the vertebra, which results in reduction of the deformity, whereas, if the anchor has a poor primary stability, it will move when subjected to reduction forces, and this will result in inadequate reduction. Lumbar screws which are advocated by many authors, are extensively used. Thoracic screws are only used by a limited number of surgeons, as most surgeons favour hooks. Polyaxiality facilitates rod positioning; it eliminates the orthogonal stresses that are generated during tightening and which are known to be responsible for screw fracture. The drawback manoeuvre consists in applying forces directly to the vertebra via the anchor; the deformity is reduced by gently translating the vertebra towards the rod. The polyaxial vertebral claw that we are presenting here is a self-stabilising implant that provides the same primary stability as the screw and allows application of multidirectional drawback forces.
Materials and methods: The system consists of self-stabilising vertebral anchors, either screws or claws. Each anchor is polyaxial and features a threaded extension that allows translation of the vertebra towards the rod. Connection of the screw or claw to the rod is provided by connecting clamps. The first operative step consists of inserting the vertebral anchors, favouring the apex of the deformity. The insertion technique is described in detail. The claw is locked independently, prior to securing the rod on to the claw. The second operative step consists of positioning the rods which are bent to the ideal sagittal curve. Polyaxiality and threaded extensions make rod positioning an easy step. Progressive tightening of the nuts results in correction of the deformity as it slowly moves the vertebrae towards the rods. The translation force is distributed over all the anchors, ensuring a gentle reduction manoeuvre with no risk of back out of the implants. Approaching vertebrae at the end of the reduction manoeuvre results in vertebral derotation. It is not necessary to use distraction which is considered hazardous.
Results: 35 such instrumentations have been used in patients with idiopathic scoliosis over the previous 12 months. We have used an average of nine screws and four claws per patient, mainly thoracic pedicle/transverse claws. Main curve correction was 71% (average curve was 59° preoperatively and 17° postoperatively). Average correction of the uninstrumented lumbar curve was 73%. The upper curve improved from 34° to 15°. The slope of the first uninstrumented vertebra was 14° pre-operatively and 6° postoperatively. In the sagittal plane, the average angle of thoracic kyphosis in hollow backs (kyphosis less than 15°) was 9°, increasing up to 27° postoperatively.
Discussion: This instrumentation is characterised by stable implants which provide a quality of reduction similar to that achieved with pedicle screws. Vertebral claws are easy to insert and have a better primary stability than screws.
Poly-axiality is a common feature to all the implants of this system; it greatly facilitates placement of the implants and allows to apply traction simultaneously to all the anchors, which results in progressive, gentle reduction. Simultaneous traction application ensures adequate correction of the thoracic kyphosis (gain of 18°). As a matter of fact, severe kyphosis can be bent into the rods, and translation of the vertebrae towards the rods is very easy. Adequate reduction of the main curve results in correction of the underlying lumbar curve and shifting of the first uninstrumented vertebra into a more horizontal position.
Conclusion: This instrumentation based on stable poly-axial implants, should allow to improve the quality of reduction of scoliosis.
The abstracts were prepared by Dr Robert J. Moore. Correspondence should be addressed to him at The Spine Society of Australia, Institute of Medical and Veterinary Science, The Adelaide Centre for Spinal Research, Frome Road, Adelaide, South Australia 5000
