We performed posterior fixation with a Hartshill-Ransford contoured loop in 43 patients with instability at the craniocervical junction. No external bracing was used. Fifteen patients had congenital malformations, ten had tumours, seven had 'bone-softening' conditions (such as osteogenesis imperfecta), five had suffered complicated fractures, three had occipito-C1-C2 hypermobility due to lax ligaments and three had severe degenerative spondylosis with pseudotumours of the transverse ligament. Twenty-nine patients had transoral decompression of the cord before fixation. In most cases, cancellous bone grafts taken from the iliac crest were used to induce fusion; in nine very ill patients, no bone graft was used. In the whole series there was no instance of construct failure, broken wire or laminar fracture. The best results were achieved in patients with tumours or bone-softening conditions. No patient with normal neurology deteriorated after surgery but seven had worse neurological deficits after operation than before. Neck stiffness caused half the patients to change their lifestyle.

Costoplasty can reduce the important cosmetic deformity of rib prominence in scoliosis but there are few objective reports of correction. We recorded the results of three objective methods of assessing back shape before and after short-segment costoplasty in 55 patients. We showed that significant improvement was maintained over a two-year follow-up period. Primary costoplasty at the time of scoliosis surgery (n = 35) achieved greater proportional correction than secondary costoplasty performed after fusion of the spine (n = 20). The rib segments removed at primary surgery provided enough bone for the autogenous graft; harvesting from the pelvis was unnecessary. We report a new classification of rib morphology which helps in planning the site and extent of costoplasty, and in predicting the possible correction.

Fractures of the atlas constitute 4% to 12% of all bony injuries of the cervical spine; most are treated successfully by a cervical orthosis. Nonunion may be associated with neck or scalp pain on movement and is treated conventionally by some form of craniocervical fusion, which restricts head movement. The authors describe a case in which direct repair of the bony ring with a titanium plate and screws allowed bone healing, relieved the symptoms and maintained a full range of neck movements. The titanium plate interfered little with postoperative MR and CT imaging.

High definition computed cervical myelograms have been made in flexion and extension in 13 patients with Morquio-Brailsford's disease. We observed that: 1) odontoid dysplasia was present in every case, with a hypoplastic dens and a detached distal portion which was not always ossified; 2) atlanto-axial instability was mild, and anterior atlanto-axial subluxation was absent in most cases; 3) severe spinal cord compression, when present, was due to anterior extradural soft-tissue thickening; 4) this compression was not relieved by flexing or extending the neck and was manifested early in life; 5) posterior occipitocervical fusion resulted in disappearance of the soft-tissue thickening and normalisation of subsequent development of the dens. We conclude that the severity of neurological involvement at the craniovertebral junction was determined by soft-tissue changes, not by the type of odontoid dysplasia nor by subluxation. Posterior occipitocervical fusion proved to be an effective treatment.

Since 1981, during operations for spinal deformity, we have routinely used electrophysiological monitoring of the spinal cord by the epidural measurement of somatosensory evoked potentials (SEPs) in response to stimulation of the posterior tibial nerve. We present the results in 1168 consecutive cases. Decreases in SEP amplitude of more than 50% occurred in 119 patients, of whom 32 had clinically detectable neurological changes postoperatively. In 35 cases the SEP amplitude was rapidly restored, either spontaneously or by repositioning of the recording electrode; they had no postoperative neurological changes. One patient had delayed onset of postoperative symptoms referrable to nerve root lesions without evidence of spinal cord involvement, but there were no false negative cases of intra-operative spinal cord damage. In 52 patients persistent, significant, SEP changes were noted without clinically detectable neurological sequelae. None of the many cases which showed falls in SEP amplitude of less than 50% experienced neurological problems. Neuromuscular scoliosis, the use of sublaminar wires, the magnitude of SEP decrement, and a limited or absent intra-operative recovery of SEP amplitude were identified as factors which increased the risk of postoperative neurological deficit.

An operation which combined anterior transoral decompression with posterior occipitocervical fixation was used in 68 rheumatoid patients with irreducible anterior neuraxial compression at the craniocervical junction. Fibre-optic laryngoscopy with nasotracheal intubation was less hazardous than tracheostomy. The patients underwent surgery in the lateral position to allow access both to the mouth and to the back of the neck without moving the head. Specially designed instruments allowed visualisation from the front without dividing the soft palate. Posterior stabilisation was achieved by a preformed contoured loop fixed to the occiput, the atlas and the axis by sublaminar wires. The procedure allowed immediate mobilisation and had a very low morbidity in such ill patients.

The surgical correction of scoliosis in adolescents involves considerable trauma to bone and muscle which, together with hypotensive anaesthesia, might be expected to compromise renal function. Our recent observation of acute renal failure in two such patients prompted a prospective study of renal function following 52 operations in 43 patients. Despite hypotension, blood loss, muscle damage and evidence of fat embolism, renal function was unaltered in all patients, and there was no impairment of spinal cord function. Careful attention was paid to the maintenance of circulating volume which is essential to protect renal perfusion.

Computerised tomography of the lumbar spine was performed on 22 patients with clinical evidence of prolapse of an intervertebral disc and normal or equivocal radiculograms. Of 11 patients with positive scans who had an operation the presence of pathology was confirmed in 10. Although CT scanning is always helpful in diagnosing disc disorders, where facilities are scarce (as in Great Britain) it is best employed in patients with negative or non-contributory radiculography.

Cervical myelopathy is an uncommon but potentially fatal complication of rheumatoid atlanto-axial subluxation. Computerised myelotomography with three-dimensional reconstruction shows that rheumatoid pannus, together with the odontoid peg, contributes significantly to anterior cervico-medullary compression. These findings were the basis for treatment by transoral anterior decompression and posterior occipitocervical fusion, which removes both bony and soft-tissue causes of compression and allows early mobilisation without major external fixation. We report encouraging results from this combined approach in 14 patients who had progressive neurological deterioration.

Rigid posterior fixation of the skull to the third, fourth and fifth cervical vertebrae was achieved in three patients who, as a result of operation, had gross instability of the craniocervical junction. An anatomically contoured steel loop was secured to the occiput via small burr holes and to the vertebrae by sublaminar wiring. This technique has the advantage over bone grafting, either alone or with cement, in that it affords rigid stabilisation, allows early mobilisation and may contribute to eventual bony fusion.

Basilar impression is a well-recognised though rare complication of osteogenesis imperfecta. Three patients, all members of the same family, with advanced basilar impression complicating osteogenesis imperfecta tarda, are described. The clinical features in these cases illustrate the natural history of this condition: from asymptomatic ventricular dilatation, through the foramen magnum compression syndrome, to death from brain-stem compression. The radiological criteria on which the diagnosis is based, are defined. Review of the literature reveals only seven previously documented cases, all in patients with mild forms of osteogenesis imperfecta. The unusually low incidence of basilar impression in osteogenesis imperfecta and its apparent restriction to patients with mild forms of the disease is discussed. The examination of close relatives of patients with basilar impression and osteogenesis imperfecta is emphasised in order to anticipate the onset of severe neurological complications.

The clinical presentation and treatment of 18 cases of osteoid osteoma or osteoblastoma of the spine are described, with an average follow-up of 4.2 years (range three months to 11.5 years). The average delay between the onset of symptoms and definitive diagnosis was 19 months. All patients presented with marked spinal stiffness and a painful scoliosis. The lesion was situated in the pedicle in the 15 patients with involvement of the thoracolumbar spine. A surgical approach allowing direct access to the pedicle without entering the spinal canal or jeopardising spinal stability is described. Surgical treatment afforded immediate relief of pain and an early return of full spinal mobility.

The behaviour pattern of the scoliosis associated with osteoid osteoma or osteoblastoma of the spine is described. In patients presenting with symptoms at or around skeletal maturity, the scoliosis is postural. Excision of the lesion ensures complete resolution of the curve. In the growing child, however, an initial postural scoliosis may develop vertebral rotation with structural characteristics. The magnitude of the curve and the associated vertebral rotation is dependent on the time interval between the onset of symptoms and the surgical treatment. Although removal of the lesion usually results in regression of the curve, a prolonged delay in treatment may result in a progressive structural scoliosis. A possible mechanism for the behaviour of the scoliosis is discussed.

An electrophysiological system for monitoring the spinal cord during operations for scoliosis is described. During the development of the technique the recording of cortical somatosensory evoked potentials from the scalp and spinal somatosensory evoked potentials from the laminae or spines was superseded by the positioning of recording electrodes in the epidural space cephalad to the area to be fused. All recordings were made in response to stimulation of the posterior tibial nerve at the knee. Results in 138 patients are presented and the findings in three patients who exhibited neurological deficits after operation are described. It is concluded that spinal somatosensory evoked potentials are sensitive to minor spinal cord impairment, possible due to ischaemia, and that these changes may be reversed when the cause is quickly remedied. The monitoring system interferes minimally with anaesthetic and surgical procedures and is now performed as a routine.

In a series of over 250 patients, four suffered peritoneal penetration by one of the pelvic pins while being put into a halo-pelvic apparatus. In each case a point on the iliac crest two to three centimetres posterior to the anterior superior iliac crest had been used as the anterior landmark for the positioning of the pelvic pin guide. To avoid this complication we advise that the anterior landmark for the guide should be the "tubercle of the crest", five centimetres or more posterior to the anterior superior iliac spine.

Twenty patients with complete brachial plexus lesions were reviewed approximately nine and a half years after injury. Thirteen were amputees and seven had received no surgical treatment. Amputation did not alleviate pain and a prosthesis was frequently of no greater use of the patient than the useless limb it replaced: only two of the thirteen amputees were true prosthetic users and they both had dominant limb involvement, the rest adapting easily to being one-handed. Initial treatment should therefore be conservative, with intensive rehabilitation and retraining. It is recommended that amputation should not be considered until a year after injury and only if the flail limb causes repulsion, prevents sporting activities or if the patient has difficulty in converting to the non-dominant limb. In no instance should smputation be done for relief of pain.

We have considered the indications for and results of transplanting pronator teres to extensor carpi radialis brevis in cerebral palsy. The operation has some virtue but a very limited application. We achieved satisfactory functional results in six of nine patients and some improvement in one. Two operations failed because of poor selection. In all patients the appearance of the limb was improved.