Abstract
Glioblastoma, IDH-wildtype is a clinically and molecularly heterogeneous disease. To investigate how longitudinal molecular evolution drives tumor progression and treatment resistance, we performed comprehensive histopathologic, genomic, and epigenomic evaluation of paired initial and recurrent IDH-wildtype glioblastomas from 106 patients. The vast majority (96/106, 91%) demonstrated clonal genetic evolution and/or epigenetic subclass shifting between primary and recurrent tumors. This included oncogenic alterations initially present in a minute subclone that expanded to predominate at time of recurrence. 11% (12/106) developed hypermutation with a predominance of C > T transitions at time of recurrence secondary to the alkylating effects of temozolomide. In contrast, the remaining 9% (10/106) demonstrated nearly identical molecular signatures between the primary and recurrent tumors without additional acquired genetic drivers. We found that TERT promoter mutation and CDKN2A homozygous deletion were uniformly shared between initial and recurrent tumors, indicating that these are fundamental early events in gliomagenesis, whereas alterations involving EGFR, PDGFRA, PTEN, NF1, and TP53 were commonly among the events private to initial or recurrent tumors, indicating acquisition later during clonal evolution. Furthermore, we identified that 17% (18/106) of glioblastomas underwent sarcomatous transformation (i.e. “gliosarcoma”) at recurrence, which were highly enriched for NF1 inactivation and mesenchymal epigenetic subclass. Cell type deconvolution revealed an increased abundance of T-lymphocytes, tumor-associated macrophages, and neutrophils in NF1 inactivated glioblastomas. Together, our analyses demonstrate that most glioblastomas undergo longitudinal genomic and/or epigenomic evolution. While there is no predominant genetic event enriched in recurrent tumors, personalized genomic analysis at time of recurrence can reveal acquired treatment resistance mechanisms (e.g. EGFR variant switching, temozolomide-induced hypermutation, novel MGMT gene amplification) that may impact therapeutic decision making. As such, we believe that precision medicine treatment regimens for glioblastoma should be based on molecular studies performed on the most recent surgical specimen.