OS5.6 Initial treatment strategy for presumed low-grade glioma: a preoperative perspective

Abstract

INTRODUCTION: Early resection has been put forward as standard of care for presumed diffuse low-grade gliomas (LGGs) that are eligible for extensive resection. This is based on studies that selected for histologically confirmed LGG and therefore those studies included enhancing lesions and discarded non-enhancing high grade lesions. This selection does not represent the actual clinical situation in which a clinician is confronted with a LGG-like lesion on MRI that is eligible for extensive resection. Therefore a study with patient selection based on preoperative characteristics is more clinically relevant. We conducted a retrospective study to examine if early resection improved overall survival (OS) compared to wait-and-scan or a biopsy for LGG-like lesions that were eligible for extensive resection. MATERIALS AND METHODS: We searched the records of all glioma patients (both low and high-grade) in three large neurosurgical institutions in the Netherlands between 1990-2010. From this set of 1115 patients, images in 498 patients were available. To identify patients with a LGG-like lesion that was eligible for extensive resection, we screened for well-known prognostic favorable characteristics; i.e. supratentorial, non-enhancing lesion with <6 cm maximal diameter and only epilepsy or minimal neurological deficits. Based on these criteria, a total of 150 patients were identified. OS was used as primary outcome measure. Median follow-up was. 7.1 years. RESULTS: As initial treatment strategy, patients underwent either an early resection (n=83), a wait-and-scan approach (n=38) or a biopsy (n=29). Clinical and tumor characteristics were evenly distributed, except for histology and grade: the biopsy group consisted of more astrocytomas (75.9% vs 48.2% in resection and 42.1% in wait-and-scan; P=0.01) and the wait-andscan group consisted of more higher grade gliomas (24.3% vs 10.8% in resection and 3.4% in biopsy; P=0.04). The latter can be expected, since histological diagnosis in the wait-and-scan group was obtained after a median of 2.95 years from initial imaging diagnosis. Median OS was not reached in the resection group, 11.9 years in the wait-and-scan group and 9.1 years in the biopsy group. There was no difference in OS for early resection versus wait-and-scan with a hazard ratio (HR) of 0.92 (95% CI 0.43-2.01; P=0.85). However, the biopsy group showed a significant shorter OS compared to early resection and wait-and-scan (HR 2.69; 95% CI 1.19-6.06; P=0.02). CONCLUSIONS: We observed no difference in OS for early resection versus an initial wait-and-scan approach. This suggests that wait-and-scan is safe and as effective as early resection for patients with minimal neurological deficits and small LGG-like lesions eligible for extensive resection. Our data argue against biopsy as initial strategy for LGG patients. This, however, is difficult to explain and needs further investigation.