Prognostic value of magnetic resonance imaging-guided stereotactic biopsy in the evaluation of recurrent malignant astrocytoma compared with a lesion due to radiation effect.

OBJECT: The prognostic value of differentiating between recurrent malignant glioma and a lesion due to radiation effect by performing stereotactic biopsy has not been assessed. Thus, this study was undertaken to determine such value. METHODS: Between 1995 and 2001, 114 patients underwent magnetic resonance (MR) imaging-guided stereotactic biopsy to differentiate lesions caused by a recurrence of malignant astrocytoma and by radiation effect. All patients had previously undergone tumor resection (World Health Organization Grade III or IV) followed by radiotherapy. Disease diagnosis based on biopsy and patient characteristics were assessed as predictors of survival according to results of a multivariate Cox regression analysis. The diagnosis determined with the aid of biopsy was compared with that established during a subsequent resection in 26 patients. Survival following stereotactic biopsy was markedly increased in patients suffering from radiation effect compared with those harboring recurrent malignant glioma (p < 0.0001). In patients with radiation effect on biopsy, an increasing patient age (p < 0.05), having had two compared with one prior resection (p < 0.05), and a decreasing time from radiotherapy to biopsy (p < 0.001) were factors associated with decreased survival. Nevertheless, in patients with biopsy-defined radiation effect at second progression or with an age younger than 50 years the survival rate remained higher than that in patients with recurrent tumor on biopsy (p < 0.01). A biopsy-based diagnosis of radiation effect obtained less than 5 months after radiotherapy was not associated with an increased rate of patient survival compared with a diagnosis of recurrent malignant glioma on biopsy (p = 0.286). Eighty-six percent of lesions initially determined to be due to radiation effect on biopsy fewer than 5 months after radiotherapy were characterized as recurrent glioma by a mean of 11 months later. In contrast, only 25% of lesions initially diagnosed as attributable to radiation effect on biopsy more than 5 months after radiotherapy were classified as recurrent glioma a mean of 12 months later (p < 0.05). CONCLUSIONS: With the aid of stereotactic biopsy the authors demonstrated prognostic significance in differentiating recurrent malignant astrocytoma from a lesion due to radiation effect in patients presenting more than 5 months after having undergone radiotherapy. In patients who presented earlier than 5 months after radiotherapy, radiation effect on biopsy was not associated with an improved rate of survival compared with that in patients harboring recurrent malignant astrocytoma.

Duke Scholars

Author Thomas John Cummings Pathology

Author Henry Seth Friedman Neurosurgery, Neuro-Oncology

Author Allan Howard Friedman Neurosurgery