OBJECTIVETo evaluate the differences in selecting patients who will undergo operation by comparing the evaluating systems and to discuss the indication of surgery performed on patients with spinal metastases.

METHODSOf the 113 patients presented at Jishuitan Hospital, Beijing, PRC. from February 1994 to February 2002, who were diagnosed as spinal metastases, the clinical, pathological and imaging records of 82 patients were reviewed. Harrington's classification system, Tokuhashi's scoring system, Tomita's scoring system, and Grubb's scoring system were used to evaluate each of these patients at the same time. The classification or score concerning each patient was recorded and compared. SPSS software was used for statistical analysis. McNemar Test was used to compare the proportion of surgery indicated by different evaluating systems.

RESULTS(1) Surgical treatment was indicated for 92.7% of patients by Tomita's scoring system and 69.5% by Harrington's classification system respectively. The difference was statistically significant at the level of P < 0.001. (2) Seventy-five cases with spinal cord palsy and progressive pain could be assessed by Tokuhashi's system. These patients were evaluated by Tomita's system at the same time. The type of surgery determined by Tokuhashi's score was excisional operation for 20% of the patients, palliative operation for 20%. As for the remaining 60%, either excisional or palliative operation might be chosen. The type of surgery determined by Tomita's score was excisional operation for 49.3% of the patients, palliative for 42.7%, and nonsurgical modality for the remaining 8.0%. (3) Twenty-five patients were classified as I, II or III, by Harrington's classification system. These patients were evaluated by the Grubb scoring system to predict the possibility of pathologic fracture and by Tomita's system at the same time to determine if an operation was necessary. By Grubb's system, prophylactic stabilization was necessary for twenty-two patients. By Tomita's system, surgery was chosen for 23 patients. But the treatment options were different in 5 patients assessed by the two systems.

CONCLUSIONSThere was a great difference in the selection of treatment modality for patients with spinal metastases by different evaluating systems. That no systems have incorporated the evaluation of life expectancy and the local lesion causes differences. The patients should be evaluated individually by more factors than those proposed by the established systems.