Faculty Bibliography

PURPOSE/OBJECTIVE:Radiomics is the extraction of multidimensional imaging-features which when correlated with genomics is termed radiogenomics. However, radiogenomic biological validation is not sufficiently described in the literature. We seek to establish causality between differential gene expression status and MRI-extracted radiomic-features in glioblastoma. METHODS:Radiogenomic predictions and validation were done using the Cancer Genome Atlas and Repository of Molecular Brain Neoplasia Data glioblastoma patients (N=93) and orthotopic xenografts (OX)(N=40). Tumor phenotypes were segmented, and radiomic-features extracted using the developed radiome-sequencing pipeline. Patients and animals were dichotomized based on Periostin (POSTN) expression levels. RNA and protein levels confirmed RNAi-mediated POSTN knockdown in OX. Total RNA of tumor cells isolated from mouse brains (knockdown and control) was used for microarray-based expression profiling. Radiomic-features were utilized to predict POSTN expression status in patient, mouse, and inter-species. RESULTS:Our robust pipeline consists of segmentation, radiomic-feature extraction, feature normalization/selection, and predictive-modeling. The combination of skull stripping, brain-tissue focused normalization and patient-specific normalization are unique to this study, providing comparable cross-platform, cross-institution radiomic-features. POSTN expression status was not associated with qualitative or volumetric MRI parameters. Radiomic-features significantly predicted POSTN expression status in patients (AUC 76.56%, sensitivity/specificity: 73.91/78.26%) and OX (AUC 92.26%, sensitivity/specificity: 92.86%/91.67%). Furthermore, radiomic-features in OX were significantly associated with patients with similar POSTN expression levels (AUC 93.36%, sensitivity/specificity: 82.61%/95.74%; p-value 02.021E-15). CONCLUSION/CONCLUSIONS:We determined causality between radiomic texture features and POSTN expression levels in a pre-clinical model with clinical validation. Our biologically validated radiomic pipeline also showed the potential application for human-mouse matched co-clinical trials.

OBJECTIVE Dexamethasone, a known regulator of mesenchymal programming in glioblastoma (GBM), is routinely used to manage edema in GBM patients. Dexamethasone also activates the expression of genes, such as CEBPB, in GBM stem cells (GSCs). However, the drug's impact on invasion, proliferation, and angiogenesis in GBM remains unclear. To determine whether dexamethasone induces invasion, proliferation, and angiogenesis in GBM, the authors investigated the drug's impact in vitro, in vivo, and in clinical information derived from The Cancer Genome Atlas (TCGA) cohort. METHODS Expression profiles of patients from the TCGA cohort with mesenchymal GBM (n = 155) were compared with patients with proneural GBM by comparative marker selection. To obtain robust data, GSCs with IDH1 wild-type (GSC3) and with IDH1 mutant (GSC6) status were exposed to dexamethasone in vitro and in vivo and analyzed for invasion (Boyden chamber, human-specific nucleolin), proliferation (Ki-67), and angiogenesis (CD31). Ex vivo tumor cells from dexamethasone-treated and control mice were isolated by fluorescence activated cell sorting and profiled using Affymetrix chips for mRNA (HTA 2.0) and microRNAs (miRNA 4.0). A pathway analysis was performed to identify a dexamethasone-regulated gene signature, and its relationship with overall survival (OS) was assessed using Kaplan-Meier analysis in the entire GBM TCGA cohort (n = 520). RESULTS The mesenchymal subgroup, when compared with the proneural subgroup, had significant upregulation of a dexamethasone-regulated gene network, as well as canonical pathways of proliferation, invasion, and angiogenesis. Dexamethasone-treated GSC3 demonstrated a significant increase in invasion, both in vitro and in vivo, whereas GSC6 demonstrated a modest increase. Furthermore, dexamethasone treatment of both GSC3 and GSC6 lines resulted in significantly elevated cell proliferation and angiogenesis in vivo. Patients with mesenchymal GBM had significant upregulation of dexamethasone-regulated pathways when compared with patients with proneural GBM. A prognostic (p = 0.0007) 33-gene signature was derived from the ex vivo expression profile analyses and used to dichotomize the entire TCGA cohort by high (median OS 12.65 months) or low (median OS 14.91 months) dexamethasone signature. CONCLUSIONS The authors present evidence that furthers the understanding of the complex effects of dexamethasone on biological characteristics of GBM. The results suggest that the drug increases invasion, proliferation, and angiogenesis in human GSC-derived orthotopic tumors, potentially worsening GBM patients' prognoses. The authors believe that careful investigation is needed to determine how to minimize these deleterious dexamethasone-associated side effects in GBM.

Mutational inactivation of ATRX (a-thalassemia mental retardation X-linked) represents a defining molecular alteration in large subsets of malignant glioma. ATRX encodes a chromatin binding protein widely implicated in epigenetic regulation and remodeling. However, multiple studies have also associated its loss in cancer with replication stress, DNA damage, and copy number alterations (CNAs). The mechanisms by which ATRX deficiency drives this global genomic instability remain unclear. Here we report that ATRX inactivation in isogenic glioma model systems induces replication stress and DNA damage by promoting the formation of deleterious G-quadruplex (G4) secondary structure in DNA. Moreover, these effects are associated with the acquisition of disease-relevant CNAs over time. Prior work has linked G4s with genomic instability as well as CNAs in cancer. We then demonstrate, both in vitro and in vivo, that chemical G4 stabilization with CX-3543 (Quarfloxin) selectively enhances cell death in ATRX deficient isogenic cell lines as well as ATRX-mutant primary glioma stem cells derived from patients. Finally, we show that G4 stabilization synergizes with other DNA-damaging therapies, including ionizing radiation, in the ATRX-deficient context. Our findings clarify distinct mechanisms by which DNA secondary structure influences ATRX-deficient glioma pathogenesis and indicate that G4-stabilization may represent and attractive therapeutic strategy for the selective targeting of ATRX-mutant cancers. Opportunities for the development of radiosensitization approaches based on G4-stabilization are particularly intriguing, as ionizing radiation remains among the most effective non-surgical treatments for malignant glioma

Purpose Both temozolomide (TMZ) and poly (ADP-ribose) polymerase (PARP) inhibitors are active in small-cell lung cancer (SCLC). This phase II, randomized, double-blind study evaluated whether addition of the PARP inhibitor veliparib to TMZ improves 4-month progression-free survival (PFS). Patients and Methods A total of 104 patients with recurrent SCLC were randomly assigned 1:1 to oral veliparib or placebo 40 mg twice daily, days 1 to 7, and oral TMZ 150 to 200 mg/m²/day, days 1 to 5, of a 28-day cycle until disease progression, unacceptable toxicity, or withdrawal of consent. Response was determined by imaging at weeks 4 and 8, and every 8 weeks thereafter. Improvement in PFS at 4 months was the primary end point. Secondary objectives included overall response rate (ORR), overall survival (OS), and safety and tolerability of veliparib with TMZ. Exploratory objectives included PARP-1 and SLFN11 immunohistochemical expression, MGMT promoter methylation, and circulating tumor cell quantification. Results No significant difference in 4-month PFS was noted between TMZ/veliparib (36%) and TMZ/placebo (27%; P = .19); median OS was also not improved significantly with TMZ/veliparib (8.2 months; 95% CI, 6.4 to 12.2 months; v 7.0 months; 95% CI, 5.3 to 9.5 months; P = .50). However, ORR was significantly higher in patients receiving TMZ/veliparib compared with TMZ/placebo (39% v 14%; P = .016). Grade 3/4 thrombocytopenia and neutropenia more commonly occurred with TMZ/veliparib: 50% versus 9% and 31% versus 7%, respectively. Significantly prolonged PFS (5.7 v 3.6 months; P = .009) and OS (12.2 v 7.5 months; P = .014) were observed in patients with SLFN11-positive tumors treated with TMZ/veliparib. Conclusion Four-month PFS and median OS did not differ between the two arms, whereas a significant improvement in ORR was observed with TMZ/veliparib. SLFN11 expression was associated with improved PFS and OS in patients receiving TMZ/veliparib, suggesting a promising biomarker of PARP-inhibitor sensitivity in SCLC.

Adult diffuse gliomas are a diverse group of brain neoplasms that inflict a high emotional toll on patients and their families. The Cancer Genome Atlas (TCGA) and similar projects have provided a comprehensive understanding of the somatic alterations and molecular subtypes of glioma at diagnosis. However, gliomas undergo significant cellular and molecular evolution during disease progression. We review the current knowledge on the genomic and epigenetic abnormalities in primary tumors and after disease recurrence, highlight the gaps in the literature, and elaborate on the need for a new multi-institutional effort to bridge these knowledge gaps and how the Glioma Longitudinal AnalySiS Consortium (GLASS) aims to systemically catalog the longitudinal changes in gliomas. The GLASS initiative will provide essential insights into the evolution of glioma toward a lethal phenotype, with the potential to reveal targetable vulnerabilities, and ultimately, improved outcomes for a patient population in need.

BACKGROUND:Meningioma is the most common primary brain tumor and has a variable risk of local recurrence. While World Health Organization (WHO) grade generally correlates with recurrence, there is substantial within-grade variation of recurrence risk. Current risk stratification does not accurately predict which patients are likely to benefit from adjuvant radiation therapy (RT). We hypothesized that tumors at risk for recurrence have unique gene expression profiles (GEP) that could better select patients for adjuvant RT. METHODS:We developed a recurrence predictor by machine learning modeling using a training/validation approach. RESULTS:Three publicly available AffymetrixU133 gene expression datasets (GSE9438, GSE16581, GSE43290) combining 127 primary, non-treated meningiomas of all grades served as the training set. Unsupervised variable selection was used to identify an 18-gene GEP model (18-GEP) that separated recurrences. This model was validated on 62 primary, non-treated cases with similar grade and clinical variable distribution as the training set. When applied to the validation set, 18-GEP separated recurrences with a misclassification error rate of 0.25 (log-rank p=0.0003). 18-GEP was predictive for tumor recurrence [p=0.0008, HR=4.61, 95%CI=1.89-11.23)] and was predictive after adjustment for WHO grade, mitotic index, sex, tumor location, and Simpson grade [p=0.0311, HR=9.28, 95%CI=(1.22-70.29)]. The expression signature included genes encoding proteins involved in normal embryonic development, cell proliferation, tumor growth and invasion (FGF9, SEMA3C, EDNRA), angiogenesis (angiopoietin-2), cell cycle regulation (CDKN1A), membrane signaling (tetraspanin-7, caveolin-2), WNT-pathway inhibitors (DKK3), complement system (C1QA) and neurotransmitter regulation (SLC1A3, Secretogranin-II). CONCLUSIONS:18-GEP accurately stratifies patients with meningioma by recurrence risk having the potential to guide the use of adjuvant RT.

Mutational inactivation of the SWI/SNF chromatin regulator ATRX occurs frequently in gliomas, the most common primary brain tumors. Whether and how ATRX deficiency promotes oncogenesis by epigenomic dysregulation remains unclear, despite its recent implication in both genomic instability and telomere dysfunction. Here we report that Atrx loss recapitulates characteristic disease phenotypes and molecular features in putative glioma cells of origin, inducing cellular motility although also shifting differentiation state and potential toward an astrocytic rather than neuronal histiogenic profile. Moreover, Atrx deficiency drives widespread shifts in chromatin accessibility, histone composition, and transcription in a distribution almost entirely restricted to genomic sites normally bound by the protein. Finally, direct gene targets of Atrx that mediate specific Atrx-deficient phenotypes in vitro exhibit similarly selective misexpression in ATRX-mutant human gliomas. These findings demonstrate that ATRX deficiency and its epigenomic sequelae are sufficient to induce disease-defining oncogenic phenotypes in appropriate cellular and molecular contexts.

Studies on melanoma brain metastases (MBM) with regard to mutational status are lacking. We investigated the outcomes of MBM in molecularly characterized patients for BRAF and NRAS mutations receiving conventional treatment. We investigated associations between outcomes [competing risk of local and distant brain failure (LF, DF) and overall survival (OS)] and clinical/pathological features of patients with known mutation status following initial treatment of MBM. Competing risk analysis was performed using the methods of Fine and Gray. We identified 235 patients with MBM diagnosed from 2005 to 2011. Mutation prevalence was BRAF ^non-V600K 98 (42%), BRAF ^V600K 34 (14%), NRAS 43 (18%), and wild-type for both genes (WT) 60 (26%) patients. Six month cumulative incidence LF rates were 3% for combined SRS or surgery with adjuvant radiation, 18% for surgery, 18% for SRS, 60% for WBRT, and 67% for systemic therapy. On multivariate analysis, only mutation status and initial treatment type were found to be independent predictors of local control. As compared to WT, NRAS (HR 2.58, 95% CI 1.18-5.67, p = 0.02) and BRAF ^V600K (HR 2.83, 95% CI 1.23-6.47, p = 0.01) mutational status were statistically significant while BRAF ^non-V600K status was not statistically significant (p = 0.23). Mutation status was not associated with DF or OS. BRAF ^V600K and NRAS mutation status predict increased LF following conventional treatments for MBM. These data can inform the design and interpretation of future MBM trials.