P08.01 Tumor cell subpopulation changes during intracranial glioblastoma patient derived xenograft recurrence following radiation, temozolomide, or combination therapy

Abstract

AbstractWe’ve established a model for recurrent GBM by treating mice bearing previously untreated, luciferase-modified, intracranial GBM PDX, with radiation (RT) (2 Gy/day, M-F), temozolomide (TMZ) only (10 mg/kg/day, M-F), or concurrent RT/TMZ. PDX respond to treatment by showing a reduced rate of increase in tumor bioluminescence (RT only), or decreased tumor bioluminescence (TMZ only or RT/TMZ), but with eventual tumor growth requiring animal euthanasia at 5.5 (RT only), 26.5 (TMZ only), or 38 (RT/TMZ) days beyond the median survival of untreated control animal subjects. All survival extensions are significant relative to untreated control mice (p<0.05). To examine the effects of treatment on tumor cell subpopulations in intracranial PDX, we employed flow cytometric analyses of human leukocyte antigen (HLA) positive GBM cells, using a panel of markers (CD133, CD15, ABCG2, SOX2, EZH2, HIF1A, MGMT, and H3K27me3). RT alone did not cause a significant enrichment for any specific subpopulation, as identified using the markers indicated above. In contrast, PDX exposed to TMZ monotherapy or RT/TMZ combination therapy showed enrichment for CD133+HIF1A+ cells (p<0.005). PDX treated with RT/TMZ were also enriched for CD133+CD15+, and additionally showed enrichment for CD133+SOX2+ and CD133+EZH2+ cells (p<0.005 for each). Perhaps not so surprisingly, recurrent tumors that had been treated with TMZ or RT/TMZ showed increased cellular positivity for MGMT (Control 0.78% vs. TMZ 3.4% p<0.005 and RT/TMZ 4.3% p<0.0005), though, paradoxically, these tumors also showed increased cellular positivity for EZH2 (Control 0.38% vs. TMZ 3.0% p=ns and RT/TMZ 7.8% p<0.0005) and a product of EZH2 activity, H3K27me3 (Control 0.55% vs. TMZ 18.2% p<0.0005 and RT/TMZ 4.3% p=ns), which is associated with transcriptional suppression. Corresponding immunohistochemical analysis is ongoing, with results to be presented at the meeting. In total, our data indicate that, in recurrent tumors, cell subpopulation selection/enrichment depends on the type of therapy administered, which should, in turn, influence the selection of salvage therapy for treating recurrent GBM. Supported by NIH grants NS095642 and NS096376.