Cerebral oligodendroglioma: prognostic factors and life history.

The records of 137 patients with supratentorial oligodendroglioma treated surgically between 1953 and 1986 were reviewed. The tumors were rated histologically benign or malignant. In the 105 patients followed up with a minimum observation time of 5 years to December 1991, the mean postoperative survival was 90.2 months (standard error, 9), the median 64 months (standard error, 9.6), the 5-year survival rate 52.4%, and the 10-year survival rate 24%. Sixteen possible prognostic factors, broken down into two or more variables each, were considered in the survival study on univariate methods (5-year survival rate, survival curves, and Cox's hazard function) and on multivariate analysis according to Cox's stepwise proportional hazards model. The latter showed that variables correlated positively with survival were benign histological findings (P, 0.000), postoperative radiation therapy (P, 0.004), and time of operation from 1977 to 1986 (P, 0.044) in 105 patients of the whole series, and period of surgery from 1977 to 1986 (P, 0.000), subtotal or total surgical resection of the tumor (P, 0.001), and radiation therapy (P, 0.005) in the subgroup of 79 patients operated on for benign tumors. However, the most interesting point to emerge from the study was the relevance of admission clinical status to the survival of patients who did not receive radiation therapy and to the prognostic response of those who did. Of the 40 patients with seizures and negative neurological status--Clinical Syndrome A--the 10 who did not receive radiation therapy had survived as long as the 30 who did (5-year survival rate, 80 versus 67%; P, not significant; median survival, 122 versus 85 months; Breslow and Mantel-Cox P, not significant), whereas of the 65 patients with intracranial hypertension and/or neurological deficits--Clinical Syndrome non-A--the 18 who did not receive radiation therapy had short survival times, and the 47 who did fared significantly better (5-year survival rate, 11 versus 53%; P, 0.002; median survival, 32 versus 64 months; Breslow and Mantel-Cox P, 0.000). These findings were not significantly affected by the exclusion of malignant neoplasms and in the group of benign tumors, in which the histological characteristics have not been found to be significantly different between those with A and those with non-A clinical syndrome, did not depend on different frequencies of subtotal or total tumor removal.(ABSTRACT TRUNCATED AT 400 WORDS)