Complications of the Luque-Galveston scoliosis correction technique in paediatric cerebral palsy

Purpose of the study: Severe scoliosis is a very frequent condition in cerebral palsy children (CP). It is surgically managed, with unit rod instrumentation being the gold standard in English-speaking countries. The purpose of this work was to report on a small, homogeneous series of non-ambulatory, quadriplegic, spastic patients treated by the Luque-Galveston technique in Strasbourg, France. We present the radiographic outcome of the technique along with a full description of any post-operative complications encountered.

Materials and methods: Twenty-eight children were operated on according to the Luque-Galveston technique between January 1997 and January 2006. This instrumentation, with fusion, included the whole spine from the sacrum to level T2. All procedures were performed as a one-stage posterior arthrodesis. The spinal deformities were single thoraco-lumbar curvatures, except in one patient. Both curve magnitude and pelvic obliquity were measured by X-ray pre-operatively, post-operatively and after longest follow-up (over 24 months). Our study focused on the rate of complications of this treatment. Only 16 patients out of 28 were tracked since the remaining 12 were lost to follow-up.

Results: Mean curve magnitude was corrected from 80 degrees to 34.8 degrees (mean correction, 56.5%), and pelvic obliquity, from 20.9 degrees to 4.2 degrees (mean correction, 79.6%). Loss of correction at average 3.46-year follow-up was 3.9 degrees of curve magnitude and 2.7 degrees of pelvic obliquity. Mean operating time was 301.5 minutes, and average blood loss was 861.9 ml. Patients were discharged from hospital after an average 19.5-day stay, including mean 8.4-day intensive care unit stay. A single major complication, monocular blindness, occurred during the procedure, probably resulting from air embolism. Post-operative complications (totaling 57.1% of our 28 patients) were: one death, three pneumothoraxes, six segmental atelectasias, seven pneumonias and one superficial wound infection. Late-onset complications (totaling 56.2% of our 16 patients at latest follow-up) were: seven broken sublaminar wires, one iliac perforation by the rod, one skin irritation from extreme malnutrition needing hardware removal, and three superficial sacral decubitus ulcers.

Discussion: Our correction rate in children affected by CP and manifesting severe scoliosis is similar to that reported in the literature by different surgical teams. Moreover, we did not observe any deep wound infection, haematoma, septicaemia, neurological and digestive complications. Late-onset complications mainly involved asymptomatic sublaminar wire breakage at the two uppermost levels, but no major complication was due to hardware failure, and vertebral fracture did not occur. There was no need for re-intervention because of the hardware, except for one case in which extreme malnutrition provoked skin conflict with the rod. We encountered 10 "windshield wiper" effects in the iliac bone, but we believe they cannot be considered as complications since they seemed to disappear after fusion was fully obtained. Last but not least, unit rod instrumentation is not very expensive compared to more modern techniques.

Conclusion: Correction of scoliosis and pelvic obliquity, attributed to CP in non-ambulatory children, by the Luque-Galveston technique is both an effective and safe choice in such an indication. Moreover, it is far less expensive than most other techniques, an aspect which should be taken into consideration.

Level of evidence: Level IV retrospective therapeutic study.