MANAGEMENT OF GLIOBLASTOMA-MULTIFORME WITH SPECIAL REFERENCE TO RECURRENCE

Between 1985 and 1995, 46 patients underwent craniotomy for glioblasto ma multiforme. The mean age was 47, varying from 9 to 71 years. The in fluence of such prognostic factors as age, preoperative Karnofsky scor e, extent of resection, tumour site, tumour size, radiotherapy, reoper ation as well as initial symptoms upon survival were studied. Of these , gross complete removal, radiotherapy, preoperative Karnofsky score, and reoperation were shown to be statistically significant to the surv ival time according to logrank and univariate tests. However, age, pre operative Karnofsky score, tumour size and temporal localisation remai ned as significant factors in multivariate analysis. The overall media n survival was 53 weeks, with no patients surviving more than 3 years. Of the patients, 41% survived over a year and 8.6% lived over two yea rs. Twenty-six patients developed a recurrent mass after an interval o f 32 weeks. The median interval time from operation to recurrence was longer in those patients who underwent gross removal than in those who had a subtotal resection, 28.2 against 20 weeks (P < 0.05). Of patien ts who had a recurrent mass, 16 were reoperated on, with a subsequent median survival time of 26.5 weeks. Our experience suggests that the s urvival of patients with glioblastoma depends on many factors, includi ng radical surgery as an initial step. In addition, the gross total re moval of the tumour also delays the development of recurrence. (C) 199 7 Elsevier Science B.V.