Patient Selection and Outcome in Low-Grade Glioma Surgery

Margret Jensdottir^1,2*Ole Solheim³Alba Corell⁴Eddie de Dios⁴Tora Dunås⁵Alexander Fletcher-Sandersjöö⁶Sasha Gulati⁷Klas Holmgren⁸Francesco Latini⁹Ruby Mahesparan¹⁰Peter Milos¹¹Alice Neimantaite⁵Henrietta Nittby Redebrandt¹²Lars Kjelsberg Pedersen¹³Rickard L. Sjöberg¹⁴Björn Sjögren¹¹Gregor Tomasevic¹²Øystein Vesterli Tveiten¹⁰Tomás Gómez Vecchio⁵Maria Zetterling⁹Jiri Bartek Jr⁶Asgeir S Jakola⁵

• ¹Department of Neurosurgery, Karolinska University Hospital, Stockholm, Sweden
• ²Karolinska Institutet Institutionen for klinisk neurovetenskap, Stockholm, Sweden
• ³St Olavs Hospital Universitetssykehuset i Trondheim Nevrokirurgisk avdeling, Trondheim, Norway
• ⁴Sahlgrenska universitetssjukhuset, Gothenburg, Sweden
• ⁵Goteborgs universitet Sahlgrenska Akademin, Gothenburg, Sweden
• ⁶Karolinska Universitetssjukhuset, Stockholm, Sweden
• ⁷St Olav's Hospital HF, Trondheim, Norway
• ⁸Umea universitet Neurovetenskaper, Umeå, Sweden
• ⁹Akademiska sjukhuset, Uppsala, Sweden
• ¹⁰Haukeland Universitetssjukehus, Bergen, Norway
• ¹¹Universitetssjukhuset i Linkoping, Linköping, Sweden
• ¹²Skanes universitetssjukhus Lund, Lund, Sweden
• ¹³Universitetssykehuset Nord-Norge HF, Tromsø, Norway
• ¹⁴Norrlands universitetssjukhus, Umeå, Sweden

Background and Objectives: Maximal safe resection is cornerstone in diffuse low-grade glioma (dLGG) management although epidemiological data are limited. We aim to explore surgical selection, techniques and outcome in a population-based cohort. Materials and methods: Multi-center case series (9 out of 10 neurosurgical departments in Norway and Sweden) of all adults (≥18 years) with histopathologically verified supratentorial dLGG, WHO grade 2, undergoing primary surgery 2012-2017. Complications within 30 days and neurological outcomes at 3 months assessed. Pre-and postoperative MRIs were reviewed centrally, blinded to patient outcome and center. Results: Of 517 included patients, 217 (41.7%) were female, and the mean (SD) age was 44.5 (15.0) years. Biopsy only was performed in 119 (23.0%) patients (13.8-38.9% across centers), and 398 (77.0%) underwent resection (61.1-86.2%). Intraoperative neurophysiological monitoring (IONM) was used in 142 (35.7%, 0-58.7%) resections. The biopsy only patients were older (52.7 years vs. 42.1 years, P<.001), had larger tumors (56.6 ml vs. 31.9 ml, P<.001) more often eloquently located (86.6 % vs. 56.5%, P<.001). The median (IQR) extent of resection (EOR) was 82.9% (63.3-97.7%), 69.7-100.0% across centers. The median (IQR) residual tumor was 4.6 ml (0.5-19.9 ml), 0.0-14.1 ml across centers. Age and histopathology were the most important predictors of EOR. New/worsened neurological deficits occurred in 165 patients (41.5%), 23.1-66.7% across centers, and persisting in 19 (4.8%, 0-22.7%) at 3 months after surgery. A complication was observed in 87 patients (21.4%, 0-31.7%) with 12 patients (3.1%, 0-9.8%) having a grade III-IV complications. Conclusions: We find surgical selection associated with age, tumor size and location. The median EOR in a population-based cohort is 83% with age and tumor biology being significant predictors. EOR did not correlate with higher risks or worse neurological outcome. We provide an epidemiological perspective demonstrating a variation in surgical selection and techniques reflecting persistent controversy in dLGG management.