COGNITIVE-FUNCTIONING IN LONG-TERM SURVIVORS OF HIGH-GRADE GLIOMA

In a pilot study, two groups of patients with malignant glioma underwe nt sequential neuropsychological evaluations after successful tumor tr eatment. Group 1 included nine patients treated from 1981 to 1985; all patients received irradiation and eight underwent chemotherapy. The b aseline neuropsychological assessment was performed 1 to 63 months aft er tumor diagnosis, with follow-up evaluations at irregular intervals over the next 3 to 7 years. Six patients in Group 1 exhibited impairme nt on most measures at baseline; subsequently, two patients developed profound cognitive impairment. Initially, three patients functioned in the average range on most tasks; thereafter, two deteriorated on one measure each. Group 2 was ascertained prospectively and included 16 pa tients treated from 1985 to 1987, all of whom received irradiation and chemotherapy. The first evaluation was performed 18 months after diag nosis, then every 6 months for 2 years, and then yearly. Compared to a control group, those in Group 2 had significant cognitive impairment at baseline. Cognitive performance did not change over the next 12 mon ths in.10 patients who remained free of tumor, but within 2 years of b aseline testing, deterioration on specific tasks was evident in two of seven disease-free survivors. When last tested, five of six disease-f ree survivors had deteriorated on one or more measures. Unlike Group 1 , severe global cognitive impairment was not seen, perhaps because Gro up 2 was followed for a shorter time. Verbal and nonverbal composite s cores derived from intelligence quotient (IQ) tests showed less impair ment at baseline than did other measures and were more likely to remai n stable subsequently. Verbal memory and sustained attention were the most impaired at baseline, and verbal learning and flexibility in thin king showed the greatest tendency to decline over time. Cognitive func tioning in survivors of high-grade glioma is best measured and monitor ed by tests that probe a broader spectrum of abilities than IQ. Neurop sychological measures used in this analysis lacked sensitivity at the lower end of the impaired range. Future studies should use tests bette r able to discern cognitive differences at low performance levels. Bas ed on this experience the authors conclude that most long-term survivo rs of high-grade glioma will have significant cognitive difficulties, usually evident by the first assessment; some patients will develop pr ofound impairment years later, and few are capable of fully independen t living.