Purpose of the study: Malignant tumours of the thoracic wall are rare. Treatment protocols include extensive surgical resection. In children, these resections can generate very severely progressive scoliosis. We studied the local conditions favouring the development of scoliosis in a consecutive series of eight patients.
Material and methods: From November 2004 to December 2007, six boys and two girls, mean age 7.6 years (range 4 months – 15 years) underwent extensive thoracic wall resection for a malignant tumour: Ewing sarcoma (n=5), spindle-cell sarcoma (n=2), neuroblastoma (n=1). All patients received adjuvant chemotherapy followed by resection removing on average 3.1 rigs (range 1 – 4 ribs). Six resections involved the costovertebral angle and three were associated with partial pneumonectomy. The number of ribs resected was noted in Roman numerals. The thoracic wall was divided into three sectors in the horizontal plane according to the level of the resection (A, B, and C, posteriorly to anteriorly). Reconstruction was achieved with a Gortex plaque (n=1), Borreli staples (n=2). Posterior spinal instrumentation was performed in one patient. Despite in sano resection, six patients underwent postoperative radiotherapy.
Results: At mean 2.9 years follow-up (range 1 – 5 years), four patients of the eight developed scoliosis convex on the operated side. These patients had IVA resections for three patients (50, 50 and 32° scoliosis) and type IA-IIB for one patients (13° scoliosis). These deformities occurred despite a corset and instrumentation in one patient. None of the patients had a humpback. Patients who did not develop scoliosis had resections IA-IIB, IA-IB, IVC and, IC.
Discussion: The type of reconstruction, the histological type, use of complementary postoperative radiotherapy or pulmonary resection did not appear to be factors favouring the development of scoliosis. Conversely, resection in zone A (posterior) over more than two adjacent levels, might be a predictive element of the risk of scoliosis.
Conclusion: In the event of a resection of ribs in a posterior zone over more than two levels for resection of a malignant tumour, posterior instrumentation should be discussed. This preliminary work offers evidence for thought but needs to be strengthened with a multicentric study to enable a statistical analysis.