O8.01 * PRE-OPERATIVE CHEMOTHERAPY AS A NEW STRATEGY IN TREATING LOW GRADE GLIOMAS IN ELOQUENT AREAS

Abstract

BACKGROUND: Recent studies suggest total/near total resection to be significantly associated with increased survival in low grade gliomas (LGGs) and reduced risk of malignant transformation. This study is aimed to prospectively evaluate whether preoperative chemotherapy with temozolomide can reduce tumor infiltration and improve the surgical resectability of LGGs in eloquent areas. METHODS: Patients with LGG requiring treatment after biopsy or partial resection at first surgery, received dose dense temozolomide (150 mg/m2/die 1 week on/1 week off). Response on MRI-FLAIR images was evaluated every 3 months based on RANO and volumetric criteria. Changes on diffusion tensor imaging (DTI) were assessed by histogram analysis and functional diffusion maps and compared with RANO and volumetric criteria and response of seizures. After a minimum of 6 cycles of temozolomide, the feasibility of reoperation according to MR-DTI findings was evaluated with the neurosurgeon. RESULTS: 26 patients with an histological diagnosis of grade II glioma were enrolled from 2008 to 2013: 19/26 males, median age 34 years (range 23-56) and median KPS 90. Tumor location was as follows: left-frontal in 10/26, fronto-temporo-insular in 9/26 (5 left, 4 right), left temporal in 4/26, right fronto-temporal in 1/26 and fronto-parietal in 2/26 (1 left, 1 right). Reasons for chemotherapy were either a large residual tumor in 15/26 (seizures in 13/15) or tumor progression on MRI in 11/26 (seizures in 5/11). 19 of 26 patients are evaluable for response, while 7 patients are still under treatment. Median number of cycles was 6 (range 4-18). Best response according to RANO was: MR in 3/19 (16%), SD in 14/19 (74%) and PD in 2/19 (10%). Among patients with response or stable disease according to RANO tumor volume reduction on FLAIR images ranged between 4.3% and 42.7% (median: 26.6%). In all patients significant changes in diffusion patterns, suggestive of a reduced tumor infiltration, were observed after 3 cycles (p < 0.05) and became more significant (p < 0.01) after 6 cycles. Twelve out of 15 patients had a significant reduction of seizure frequency concurrent with significant changes on DTI despite a volumetric stability. 17 out of 19 patients underwent reoperation, and most of them after 6 cycles of temozolomide: a near-total resection (≥ 95%) was achieved in 9/19 (47%). In 4 patients total resection was confirmed by the absence of IDH1 mutated cells in tumor margins. In two patients reoperation was not feasible because MR-DTI failed to demonstrate significant changes. CONCLUSIONS: Preoperative chemotherapy in patients with LGG in “eloquent” area can allow a total/near-total resection at reoperation in a significant subset of patients. DTI changes may be an early signature for response and correlate with seizure reduction better than conventional MRI. Moreover, DTI could improve the evaluation of feasibility and timing of reoperation.