CTIM-19. MOLECULAR AND GENETIC DETERMINANTS OF RESPONSE TO PD-1 BLOCKADE IN RECURRENT GLIOBLASTOMA PATIENTS. (9th November 2020)

• Record Type: Journal Article
• Authors: Prins, Robert; Galvez, Mildred; Amouzgar, Meelad; Campbell, Katie; Wells, Daniel; Bayless, Nicholas; Everson, Richard; Cloughesy, Timothy
• Is Part Of: Neuro-oncology. Volume 22(2020)Supplement 2
• Page Start: ii37
• Page End: ii37
• Publication Date: 2020-11-09
• Subjects: Brain Neoplasms -- Periodicals; Brain -- Tumors -- Periodicals; Brain -- Cancer -- Periodicals; Nervous system -- Cancer -- Periodicals; 616.99481
• DOI: 10.1093/neuonc/noaa215.153
• Languages: English
• ISSNs: 1522-8517

Abstract

Abstract: Despite immune checkpoint inhibitors having success in several other tumor types, many glioblastoma (GBM) patients fail to respond or maintain a sustained response. Work published by our group (Cloughesy et al, 2019) demonstrated that relative to adjuvant programmed cell death-1 (PD-1) blockade, neoadjuvant treatment doubled the median overall survival (OS) for recurrent GBM patients and resulted in an enhanced interferon-γ signature. This suggests that anti-PD-1 given in the neoadjuvant setting may improve outcomes for recurrent GBM patients. The challenge remains in identifying the molecular and genetic signatures associated with response to immune checkpoint blockade. To address this, we analyzed the tumor sample and clinical response data from the patients treated in this clinical trial (n=31). We stratified patients as stable disease (SD) versus progressive disease (PD) based on their response assessment in neuro-oncology criteria (RANO) scores from cycle 2 of treatment post-surgery. Among the SD patients, 77.8% received neoadjuvant treatment while 22.2% received adjuvant therapy. In this group, a median OS was not reached. Among the PD patients, 40.9% received neoadjuvant treatment and 59.1% received adjuvant therapy, with a median OS of 257 days. Next, we analyzed factors that impact response to immunotherapy, which includes somatic mutational burden and interferon-γ pathway induction. We calculated somatic mutational variants, copy number variants (CNVs), and differential gene expression from the bulk tumor exome and RNA-sequencing data. The total mutation counts were similar between groups and no association was identified with increased mutational burden. In addition, total CNV stability was similar between groups. However, when looking at genes involved in the JAK/STAT signaling pathway, there were notably more copy number losses of JAK2 in the PD group when compared to the SD group (85.0% versus 66.7%). These findings merit further exploration as the JAK/STAT pathway has been implicated in response to immune checkpoint blockade.