Search

BACKGROUND Surgery is the mainstay of meningioma treatment, but improvements in meningioma risk stratification are needed and indications for postoperative radiotherapy are controversial. DNA methylation profiling, copy number variants (CNVs), exome sequencing, and RNA sequencing have improved understanding of meningioma biology, but have not superseded histologic grading, or revealed biomarkers for radiotherapy responses. To address these unmet needs, we optimized and validated a targeted gene expression biomarker predicting meningioma outcomes and responses to radiotherapy. METHODS Targeted gene expression profiling was performed on a discovery cohort of 173 meningiomas (median follow-up 8.1 years) and a validation cohort of 331 meningiomas (median follow-up 6.1 years) treated with surgery (n=504) and postoperative radiotherapy (n=73) at independent, international institutions (70% WHO grade 1, 24% WHO grade 2, 6% WHO grade 3). Optimized targeted gene expression models predicting clinical outcomes (34 genes) or radiotherapy responses (12 genes) were developed from the discovery cohort, and compared to histologic and molecular classification systems by performing DNA methylation profiling, CNV analysis, exome sequencing, and RNA sequencing on the same meningiomas. RESULTS Targeted gene expression profiling achieved a concordance-index of 0.75 ± 0.03 (SEM) for local freedom from recurrence (LFFR) and 0.72 ± 0.03 for overall survival (OS) in the validation cohort, outperforming WHO grade (5-year LFFR delta-AUC 0.15, 95% CI 0.076-0.229, p=0.001) and DNA methylation grouping (delta-AUC 0.075, 95% CI 0.006-0.130, p=0.01) for LFFR, disease-specific survival, and OS. The biomarker was independently prognostic after accounting for WHO grade, extent of resection, primary versus recurrent presentation, CNV status, DNA methylation group, and Ki67 labeling index, and identified meningiomas benefiting from radiotherapy (interaction p-value=0.0008), suggesting postoperative radiotherapy could be refined in 30.2% of cases. CONCLUSIONS Targeted gene expression profiling of 504 meningiomas improves discrimination of meningioma local recurrence, disease-specific survival, and overall survival, and predicts radiotherapy responses.

BACKGROUND Meningiomas are the most common primary central nervous system tumor in adults. The majority are clinically indolent and WHO grade I. However, 20-30% are WHO grade II and 1-3% are WHO grade III, which have shorter progression-free (PFS) and overall survival. A subset of grade I meningiomas also follow an aggressive course, requiring serial resection and/or radiotherapy, similar to high-grade meningiomas. The objective of this study was to characterize the genetic alterations in meningiomas with an aggressive clinical course. METHODS Targeted next-generation sequencing (NGS) of 40 aggressive meningiomas was performed using the UCSF500 cancer panel, which assesses ~500 cancer-related genes. Information on patient demographics, tumor histopathology, and treatment history was also collected. RESULTS Meningiomas analyzed included 15 (38%) WHO grade I, 11 (28%) WHO grade II, and 13 (33%) WHO grade III. At the time of genetic profiling, 71% of patients had received prior treatment (68% surgery, 48% fractionated radiotherapy, 23% radiosurgery). The most commonly altered gene was NF2, with 71% of tumors demonstrating biallelic inactivation. Other common alterations included biallelic CDKN2A/B deletion (15%) and TERT amplification or promoter mutation (13%). Other recurrent alterations involved SUFU (8%), and ARID1A, ATM, BAP1, DMD, KDM6A, and PTEN (each 5%). Genes which are commonly altered in indolent WHO grade I meningiomas, such as AKT1, KLF4, PIK3CA, SMO, and TRAF7, were intact in this cohort. CONCLUSIONS The additional genetic alterations beyond NF2 inactivation that drive aggressive meningiomas are diverse and include homozygous deletion of CDKN2A (negative regulator of cyclin-dependent kinases 4/6), inactivating mutations in SUFU (a negative regulator of Hedgehog signaling), homozygous deletion of PTEN (a negative regulator of mTOR signaling), and mutations/deletions in epigenetic regulatory factors (e.g. ARID1A, KDM6A). Clinical trials are needed to assess the efficacy of therapeutics targeting these specific pathways in patients with meningiomas refractory to conventional treatments.

Tumor-treating fields (TTF) increase patient survival in glioblastoma, and are approved for recurrent (2011) and de novotumors (2016). Since 2012, 52 patients with gliomas at Columbia University were treated with TTF. Given preclinical data suggesting synergy between TTF and PD1-targeted therapy, the rates and outcomes of combination TTF and immunotherapy were analyzed. Of 52 patients, 40% were women. Median age at diagnosis was 58 (23-84) years. WHO glioma grade was II (6%), III (10%), or IV (84% of patients). TTF was used at recurrence in 42 (81%), with 0-4 prior recurrences. Prior therapy included bevacizumab (20%), resection (gross-total in 54%), radiation (60Gy in 85%), and temozolomide (100%). 11 (75%) patients met a compliance goal of 75%. Only 2 (4%) experienced skin toxicity requiring interruption. Concurrent systemic therapy included temozolomide (31%), nitrosourea (8%), bevacizumab (37%), or immunotherapy (54% of patients). At interim follow-up, progression free (PFS) and overall (OS) survival were 3.6 and 21.5 months, respectively. Objective response rate (ORR) based on RANO was increased in patients treated with concurrent immunotherapy (26% vs. 13%). Among GBM patients treated with concurrent immunotherapy, median PFS was significantly increased (4.6 vs. 3.5 mo, p=0.04), and survival trended upward (8.3 vs. 6.5 mo, p=0.32). After long-term follow-up, PFS, OS, and impact of compliance rates will be re-analysed. TTF improves survival for glioblastoma patients and immunotherapy shows promise in other cancers. The possible synergy between these modalities in preclinical studies and this retrospective series warrants further study in a randomized controlled trial. We propose a single-arm phase II trial of concurrent PD1 inhibitor and TTF for patients with bevacizumab-naïve glioblastoma, at diagnosis or recurrence. OS is the primary endpoint; secondary endpoints include PFS, ORR, and predictive value of PDL1 expression, and impact of compliance. Results from the EF-14 trial will serve as a historical control.