The purpose of this retrospective study was to analyze the indications for spinal instrumentation, report the clinical features, operative details and outcome in 16 patients with active pyogenic spinal infection.

Between January 1991 to October 1999, 81 patients with spontaneous pyogenic spinal infection were treated at the authors’ institution. Surgery (other than biopsy) was indicated in 24 patients for neurological deterioration, deformity or instability. Sixteen of these patients were treated with instrumentation in the presence of active spinal infection. Six patients underwent combined anterior and posterior procedures. 10 had a posterior procedure only. Outcomes assessed were control of infection, neurology, fusion, back pain and complications.

At a mean follow up period of 26. 9 months, all surviving patients were free of clinical infection. None of the patients had neurological deterioration. All patients who had neurological deficit preoperatively improved by at least one Frankel grade. A solid fusion was achieved in 15 patients. 12/15 patients remained asymptomatic or had very little pain. The remaining 3 patients had mild to moderate back pain. The mean correction of the kyphotic deformity was 18. 92 degrees. Postoperative complications included bronchopneumonia, nonfatal pulmonary embolism and seizures in 3 patients. One patient developed progressive kyphosis despite instrumentation but eventually fused in kyphus.

Given early recognition of pyogenic spinal infection, most cases can be managed non-operatively. Our results support that instrumented fusion with or without decompression may be used safely when indicated without the risk of recurrence of infection. Instrumentation facilitates nursing care and allows early mobilisation. For biomechanical reasons, a combined procedure is probably indicated for lesions above the conus. For lesions below the conus, we were able to achieve successful results with posterior approach only.

Objective: To assess the validity of the Tokuhashi and Tomita scoring systems in the prediction of prognosis following spinal surgery for skeletal metastases.

Design: A retrospective cohort study of patients treated in a specialist spinal unit

Subjects: All patients undergoing definitive surgery for metastastes of the spine were considered eligible. Time to death or current length of survival was available in 147* of these which was confirmed by the Cancer Registry. Medical and nursing case notes were reviewed and prognostic scores using the methods of Tokuhashi et al, and Tomita et al. were calculated for each patient.

*

death data for further patients currently awaited from Cancer Registry.

Outcome measures: Mean survival period with 95% confidence intervals for patients grouped according to prognostic score.

Results: Thirty-two patients were still alive and 113 had confirmed death dates. Forty-three patients had Tokuhashi scores of 9 or greater with a mean survival of 20.1 months (95% confidence interval 5.8 months) compared to 9.5 months (2.9 months) for scores 6–8 and 3.5 months (1.8 months) for scores below this. Tomita scoring showed a similar trend with those with better prognostic profiles but without the same degree of statistical significance. The overall 30-day mortality was 8.2% with no significant difference between any other groups.

Conclusion: In patients presenting with metastatic disease involving the spine, published prognostic profiles offer some guidance to likely survival of the patient and so the appropriateness of surgical treatment.