Hello everyone, I hope everyone is staying safe. My name is Ashish Shah and I'm a PGOA 6 neurosurgery resident at the University of Miami. I am honored by the AANS tumor section to give this presentation on some work we have conducted on lobectomies for newly diagnosed GBM and our perspective on supermaximal resection. The concept of lobectomies for glioma were first introduced in the early 20th century as a salvage treatment for an otherwise dismal prognosis. It was said that solid gliomas present a fairly hopeless situation, unless they are so situated that a lobectomy can be performed. However, for lobectomies in the mid-20th century, operable morbidity remained incredibly high with a mortality rate of almost 50%. Given the high morbidity of lobectomies in this era and the popularity of the microscope in the 1980s, microsurgery for gliomas also became more feasible, permitting an improved extent of resection and survival compared to traditional macrosurgical techniques. The initial series with the microscope demonstrated that there was no difference in survival for performing a lobectomy versus a traditional lesionectomy with an overall survival of 12 months in both groups. Consequently, over the last decade, we have established a new threshold for extent of resection for glioma, ranging from 80% to 98% of the contrast enhancing border to confer a survival benefit. All of these studies validate that tumor resection should be maximized without increasing neurological morbidity, which is essentially a maximal safe resection. However, was resection of the contrast enhancing rim enough? Pat Kelly's work in the 1960s suggested otherwise. His work demonstrated that tumor infiltration extended well beyond the contrast enhancement on CT scans. Now we can fast forward 50 years to the concept of modern day supermaximal resection. The first reports of modern supermaximal resection were described in France by Hugh Defoe for low grade gliomas using awake surgery to maximize the extent of resection. This concept was then translated to high grade gliomas by Dr. Defoe and another group from MD Anderson that demonstrated that resection of 50% of the peripheral flare added a survivable benefit. To extend this concept even further, lobectomy has now been re-introduced as a strategy to improve survival for GBM, with improved survival noted in several small series. So we sought to understand if lobectomy for non-eloquent GBM would confer a similar survival benefit. Our study design was split into two main arms given the different strategies employed by treating neurosurgeons at our institution. One a lobectomy cohort and two a lesionectomy or traditional GTR group, which was essentially a resection of the contrast enhancing lesion. Of note, we only included non-eloquent regions in our study, which included right frontal, right temporal, right occipital, or left occipital GBMs. The study was designed as a retrospective propensity match study to eliminate selection bias between cohorts. All relevant clinical pathological data was collected, including KPS, MGMT status, IDH, and tumor size. We utilized a type of propensity score matching called IPTW, which is inverse probability of treatment weighting, to simulate a randomized control trial and reduce founding effects in our retrospective study. We controlled for KPS, tumor size, location, age, and size of flare. Overall our study was the largest lobectomy series to date, with a total of 69 patients, 32 lobectomies, and 37 lesionectomies. Of note, there was significantly more IDH wild type and preoperative flare volumes in our lobectomy group. Here are a few selected examples of our lobectomy and GTR cohort. Our results indicated that after propensity matching, overall survival was higher in the lobectomy group compared to the GTR group, without any significant difference in post-operative functional status. Overall survival remained 30.7 months in the lobectomy group compared to 14.1 months in the lesionectomy group for newly diagnosed GBM. This trend also held for the IDH wild type and MGMT unmethylated GBMs as well, as you can see by the green curves. Nevertheless, we must acknowledge several limitations of these types of retrospective studies. One, we included many surges in our analysis, which may confound the operative technique and outcomes. Additionally, we could get benefit from longer follow-up and the inclusion of neuropsychological testing to assess for any cognitive and functional impairments beyond the standard KPS score. In conclusion, our propensity match study suggests that lobectomy for non-eloquent GBM confers an added survival benefit compared to GTR alone. For patients with non-eloquent GBM, a supermaximal resection by means of an anatomic lobectomy should be considered as a primary surgical treatment and select patients if feasible. Thank you. I want to give thanks to my mentors Dr. Komotar and Dr. Ivan for their support in scientific design for this project, as well as several prospective neurosurgeons including Anil Mavadi, Alexis Samache, and Long Di.

lobectomies

newly diagnosed glioblastoma

supermaximal resection

microsurgery

tumor infiltration