SURG-14. PATTERNS OF PROGRESSION OF TEMPORAL LOBE GLIOBLASTOMAS: EVIDENCE FOR THE ROLE OF TEMPORAL LOBECTOMY

Abstract Glioblastomas extend beyond the contrast-enhancing tumor, and supramaximal resections into non-enhancing (NE) tumor improves survival. Yet, supramaximal resection is difficult to achieve in most cases. For temporal lobe tumors, supramaximal anterior temporal lobe resection is feasible and safe. Case series have shown a survival advantage by resecting the anterior temporal lobe. This study aims to explore patterns of tumor progression in temporal lobe glioblastomas and whether the presence of NE tumor in the temporal tip could be used as a biomarker to predict the site of progression. We searched a single institution tumor database looking for patients who had glioblastomas in the temporal lobe, who underwent complete resection of enhancing tumor, and went on to have chemoradiotherapy. Patients needed to have imaging demonstrating first progression. We determined pre-operatively if NE tumor had extended into the temporal tip. At progression, the direction of tumor progression was recorded and assessed if it would have been resected should a lobectomy have been performed. We identified 55 patients that fulfilled our entry criteria (mean age 57.7 years, range 19.6-72.9; 37 male). In 16 cases, the anterior temporal lobe was resected; these patients were not used in further assessments. In 25 cases (64.1%), the tumor progressed into the residual anterior temporal lobe. In 16 cases (41.0%), a lobectomy would have encompassed the area of progression. NE tumor was seen in the tip in 28 cases (71.8%). This could predict progression into the anterior temporal lobe with a sensitivity of 76.2% (95%CI: 54.9-90.6%) and specificity of 33.1% (95%CI: 11.8-61.6%). The likelihood ratio of tumor progression was 1.14 (95%CI: 0.75-1.74). We conclude that temporal lobe glioblastomas most commonly progress into the anterior temporal lobe, but NE tumor is not a suitable predictive biomarker. Developing suitable biomarkers and biomarker-driven trials of temporal lobectomy for glioblastomas are needed.