Beyond sequence variation: assessment of copy number variation in adult glioblastoma through targeted tumor somatic profiling.

Glioblastoma is the most common primary malignancy of the adult central nervous system. Gliomagenesis involves a complex range of alterations, including sequence changes, copy number variations (CNVs), and epigenetic modifications, that have clinical implications for disease classification and prognosis. Thus, multiple testing modalities are required to support a complete diagnostic workup. The goal of this study was to streamline the multipart workflow by predicting both sequence changes and CNVs (specifically EGFR amplifications) from a single next-generation sequencing (NGS) test. Eighty-six primary and secondary glioblastomas were submitted for clinical NGS to report sequence variants from a concise panel of cancer-relevant genes. Most specimens underwent concomitant testing by methylation-specific polymerase chain reaction, immunohistochemistry, and fluorescence in situ hybridization. Using data generated during the course of clinical testing, we found that NGS-based variant predictions were concordant with immunohistochemistry and fluorescence in situ hybridization for IDH mutation and EGFR amplification status, respectively. We also noted that EGFR amplifications correlated with polysomy of chromosome 7, 19, and 20, and loss of PTEN and CDKN2A. EGFR-unamplified cases had lower rates of chromosome 7 polysomy, and PTEN and CDKN2A loss, but more CNVs overall. TP53, NF1, ATRX, and PDGFRA mutations were nearly exclusive to specimens without EGFR amplification. EGFR amplification was not associated with longer progression-free survival in this cohort, but amplifications were enriched in a group with slightly longer overall survival despite radiographic evidence of disease progression. Further study is needed to explore the mechanisms responsible for noted patterns of co-occurring variants and to correlate them with specific clinical outcomes.

Analysis of point mutations and copy number variation in Grade II and III meningioma.

Meningiomas are among the most common tumors of the adult central nervous system (CNS). They are classified by the World Health Organization into three pathologic grades with increasing severity: grade I are benign with favorable treatment outcomes and low recurrence rates while grade III display malignant behavior and poor progression-free survival. Previous studies have shown that inactivation of NF-2 is the most common genetic event in high-grade meningioma; however, there is dearth of molecular data to distinguish grade II (AM-II) from the even more aggressive grade III (AM-III). As part of a routine diagnostic workup, 19 AM-II and 5 AM-III were submitted for targeted sequencing on a panel of twenty-four genes relevant to CNS tumors. The data generated during the course of clinical care was collected and re-analyzed with the aim of identifying molecular features to distinguish AM-II and AM-III. Our cases contained several well-characterized, potentially actionable mutations, but we did not find any novel, recurrent sequence variants. Copy number variations were common in both AM-II and AM-III; chr22q loss was the most prevalent followed in decreasing frequency by losses of chr1p, chr14q, and chr10. In particular, chr10 loss was noted in 4 of 5 AM-III cases but none of the AM-II cases. This suggests that chr10 loss may serve as a diagnostic and perhaps a prognostic marker to differentiate AM-II from AM-III. If confirmed in larger studies, our finding could further aid the classification of meningioma.

FGFR2 amplification in colorectal adenocarcinoma.

FGFR2 is recurrently amplified in 5% of gastric cancers and 1%-4% of breast cancers; however, this molecular alteration has never been reported in a primary colorectal cancer specimen. Preclinical studies indicate that several FGFR tyrosine-kinase inhibitors (TKIs), such as AZD4547, have in vitro activity against the FGFR2-amplified colorectal cell line, NCI-H716. The efficacy of these inhibitors is currently under investigation in clinical trials for breast and gastric cancer. Thus, better characterizing colorectal tumors for FGFR2 amplification could identify a subset of patients who may benefit from FGFR TKI therapies. Here, we describe a novel FGFR2 amplification identified by clinical next-generation sequencing in a primary colorectal cancer. Further characterization of the tumor by immunohistochemistry showed neuroendocrine differentiation, similar to the reported properties of the NCI-H716 cell line. These findings demonstrate that the spectrum of potentially clinically actionable mutations detected by targeted clinical sequencing panels is not limited to only single-nucleotide polymorphisms and insertions/deletions but also to copy-number alterations.

Molecular Profiling of Malignant Pleural Effusion in Metastatic Non-Small-Cell Lung Carcinoma. The Effect of Preanalytical Factors.

RATIONALE: Non-small-cell lung cancer (NSCLC)-associated malignant pleural effusions (MPEs) are sometimes the only available specimens for molecular analysis. OBJECTIVES: This study evaluates diagnostic yield of NSCLC-associated MPE, its adequacy for molecular profiling and the potential influence of MPE volume/cellularity on the analytic sensitivity of our assays. METHODS: Molecular results of 50 NSCLC-associated MPE cases during a 5-year period were evaluated. Molecular profiling was performed on cell blocks and consisted of fluorescent in situ hybridization (FISH) for ALK gene rearrangements and the following sequencing platforms: Sanger sequencing (for EGFR) and high-throughput pyrosequencing (for KRAS and BRAF) during the first 4 years of the study period, and targeted next-generation sequencing performed thereafter. RESULTS: A total of 50 NSCLC-associated MPE cases were identified where molecular testing was requested. Of these, 17 cases were excluded: 14 cases (28%) due to inadequate tumor cellularity and 3 cases due to unavailability of the slides to review. A total of 27 out of 50 MPE cases (54%) underwent at least EGFR and KRAS sequencing and FISH for ALK rearrangement. Of the 27 cases with molecular testing results available, a genetic abnormality was detected in 16 cases (59%). The most common genetic aberrations identified involved EGFR ( 9 ) and KRAS ( 7 ). Six cases had ALK FISH only, of which one showed rearrangement. MPE volume was not associated with overall cellularity or tumor cellularity (P = 0.360). CONCLUSIONS: Molecular profiling of MPE is a viable alternative to testing solid tissue in NSCLC. This study shows successful detection of genetic aberrations in 59% of samples with minimal risk of false negative.

Molecular Alterations in Patients with Pulmonary Adenocarcinoma Presenting with Malignant Pleural Effusion at the First Diagnosis.

OBJECTIVES: The aim of this study was to report cytologic and molecular features of pulmonary adenocarcinoma patients presenting with a malignant pleural effusion at the first diagnosis. STUDY DESIGN: Patients who had a cytopathologic diagnosis conclusive for lung adenocarcinoma for the first time on their pleural fluid specimen, and molecular testing done, were studied. The control group consisted of patients with a malignant pleural effusion that developed during disease progression. RESULTS: We identified 18 patients (9 males and 9 females). Micropapillary and/or solid adenocarcinoma type features predominated among cytologic specimens (n = 15), while acinar patterns predominated in controls. Survival was not significantly different from that of the control group (mean 13.8 vs. 13.9 months, respectively; p = 0.61). Ten (55%) cases had mutations in EGFR (n = 6; 60%), KRAS (n = 3; 30%), or ALK translocation (n = 1; 10%). No mutations were identified in BRAF, AKT, ERBB2, NRAS, or PIK3CA (tested in 7 patients). Patients positive for the tested mutations had a better overall survival than patients negative for the mutations (mean survival 16.2 vs. 6.05 months, respectively; p = 0.006, log-rank test). Ten (84%) control patients were positive for mutations in EGFR (n = 5; 42%), KRAS (n = 4; 34%), or ALK translocation (n = 1; 8.4%). CONCLUSION: In our series, a micropapillary-like and solid-like morphology, common in cytologic specimens, and alterations in EGFR were the most frequent identifiable molecular changes.

Targeted Next-Generation Sequencing in Molecular Subtyping of Lower-Grade Diffuse Gliomas: Application of the World Health Organization's 2016 Revised Criteria for Central Nervous System Tumors.

The 2007 World Health Organization Classification of Tumours of the Central Nervous System classifies lower-grade gliomas [LGGs (grades II to III diffuse gliomas)] morphologically as astrocytomas or oligodendrogliomas, and tumors with unclear ambiguous morphology as oligoastrocytomas. The World Health Organization's newly released (2016) classification incorporates molecular data. A single, targeted next-generation sequencing (NGS) panel was used for detecting single-nucleotide variation and copy number variation in 50 LGG cases originally classified using the 2007 criteria, including 36 oligoastrocytomas, 11 oligodendrogliomas, 2 astrocytomas, and 1 LGG not otherwise specified. NGS results were compared with those from IHC analysis and fluorescence in situ hybridization to assess concordance and to categorize the tumors according to the 2016 criteria. NGS results were concordant with those from IHC analysis in all cases. In 3 cases, NGS was superior to fluorescence in situ hybridization in distinguishing segmental chromosomal losses from whole-arm deletions. The NGS approach was effective in reclassifying 36 oligoastrocytomas as 30 astrocytomas (20 IDH1/2 mutant and 10 IDH1/2 wild type) and 6 oligodendrogliomas, and 1 oligodendroglioma as an astrocytoma (IDH1/2 mutant). Here we show that a single, targeted NGS assay can serve as the sole testing modality for categorizing LGG according to the World Health Organization's 2016 diagnostic scheme. This modality affords greater accuracy and efficiency while reducing specimen tissue requirements compared with multimodal approaches.

A single EBUS-TBNA procedure can support a large panel of immunohistochemical stains, specific diagnostic subtyping, and multiple gene analyses in the majority of non-small cell lung cancer cases.

Targeted therapies for pulmonary adenocarcinoma (ACA) necessitate specific subtyping and molecular testing of non-small cell lung carcinomas (NSCLC). However, endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA) has decreased the tissue available for these assessments. While EBUS-TBNA specimens have previously been reported to successfully subtype NSCLC, allow immunohistochemistry (IHC), and support molecular diagnostics, no studies have documented the extent to which all objectives are possible in a single sample. Of 107 consecutive EBUS-TBNA specimens that were eligible for molecular testing, 98.8% had enough tissue for IHC, 80.2% received a definitive subtype, and 71.0% had both sufficient tissue to attempt molecular testing and technical success on multigene next-generation sequencing and ALK fluorescence in situ hybridization assays. Both subtyping and molecular diagnostics were possible in 57.9% of patients. The mean number of immunostains performed did not differ between patients with or without successful molecular testing (4.4 versus 4.6, P = .88). Only 40% of patients with insufficient tissue underwent repeat sampling. These findings indicate that a majority of EBUS-TBNA specimens provide sufficient tissue for subtyping pulmonary NSCLC, performing IHC, and completing multiple gene analyses. Although priorities must be assessed for each case individually, performance of IHC does not detract from completion of molecular diagnostics in general. Because most patients never undergo repeat sampling, the tissue yield of EBUS-TBNA should be improved to maximize evaluation for targeted therapies.

Non-p.V600E BRAF Mutations Are Common Using a More Sensitive and Broad Detection Tool.

OBJECTIVES: To assess the performance of a next-generation sequencing (NGS) platform for the clinical detection of BRAF mutations. METHODS: In this retrospective quality assessment of an NGS assay, we analyzed BRAF mutations within parts of exons 11 and 15 in 835 neoplastic tissues submitted to our molecular diagnostics laboratory. RESULTS: The NGS assays detected a BRAF mutation in 5.9% of lung adenocarcinomas, 13% of colorectal cancers, and 44% of melanomas. Mutant allele frequencies were less than 20% in 28% of 88 BRAF-mutated specimens. Two lymph node specimens with subcapsular or infiltrative metastasis showed 1% to 2% mutant alleles. There were 26 unique BRAF mutations in exons 11 and 15, including three novel mutations. Mutations were located outside codon 600 in 39% of BRAF-mutated tumors. Lung adenocarcinomas showed significantly higher non-p.V600E mutations (86%) than did colorectal cancers (23%) and melanomas (34%). The three most common BRAF mutations in lung cancers accounted for only 41% of the observed BRAF mutations (p.D594G [18%], p.V600E [14%], and p.G469A [9%]). CONCLUSIONS: The NGS assay demonstrated a high analytic sensitivity and a broad reportable range for clinical detection of BRAF mutations. Elucidating the spectrum of non-p. V600E BRAF mutations in different malignancies is a first step toward understanding their clinical significance.

Persistent STAT5 phosphorylation and epigenetic dysregulation of GM-CSF and PGS2/COX2 expression in Type 1 diabetic human monocytes.

STAT5 proteins are adaptor proteins for histone acetylation enzymes. Histone acetylation at promoter and enhancer chromosomal regions opens the chromatin and allows access of transcription enzymes to specific genes in rapid response cell signals, such as in inflammation. Histone acetylation-mediated gene regulation is involved in expression of 2 key inflammatory response genes: CSF2, encoding granulocyte-macrophage colony stimulating factor (GM-CSF), and PTGS2, encoding prostaglandin synthase 2/cyclooxygenase 2 (PGS2/COX2). Prolonged CSF2 expression, high GM-CSF production, and GM-CSF activation of PTGS2 gene expression all are seen in type 1 diabetes (T1D) monocytes. Persistent phosphorylation activation of monocyte STAT5 (STAT5Ptyr) is also found in individuals with or at-risk for T1D. To examine whether elevated T1D monocyte STAT5Ptyr may be associated with aberrant inflammatory gene expression in T1D, blood monocytes from non-autoimmune controls and T1D patients were analyzed by flow cytometry for STAT5Ptyr activation, and by chromatin immuno-precipitation (ChIP) analyses for STAT5Ptyr's ability to bind at CSF2 and PTGS2 regulatory sites in association with histone acetylation. In unstimulated monocytes, STAT5Ptyr was elevated in 59.65% of T1D, but only 2.44% of control subjects (p<0.0001). Increased STAT5Ptyr correlated with T1D disease duration (p = 0.0030, r(2) = 0.0784). Unstimulated (p = 0.140) and GM-CSF-stimulated (p = 0.0485) T1D monocytes, had greater STAT5Ptyr binding to epigenetic regulatory sites upstream of CSF2 than control monocytes. Increased STAT5Ptyr binding in T1D monocytes was concurrent with binding at these sites of STAT6Ptyr (p = 0.0283), CBP/P300 histone acetylase, acetylated histones H3, SMRT/NCoR histone deacetylase (p = 0.0040), and RNA Polymerase II (p = 0.0040). Our study indicates that in T1D monocytes, STAT5Ptyr activation is significantly higher and that STAT5Ptyr is found bound to CSF2 promoter and PTGS2 enhancer regions coincident with histone acetylation and RNA polymerase II. These findings suggest that the persistent activation of STAT5 by GM-CSF may be involved in altering the epigenetic regulation of these inflammatory response genes in T1D monocytes.

Radiographic response of brain metastasis after linear accelerator radiosurgery.

BACKGROUND: Radiographic response of brain metastasis to stereotactic radiosurgery (SRS) over time has not been well characterized. Being able to predict SRS-induced changes in tumor size over time may allow improved counseling of patients and potentially earlier recognition of poor response to SRS. OBJECTIVE: To quantify the rate of change in size of metastatic brain tumors after treatment with a linear accelerator (LINAC) SRS. METHODS: We performed a retrospective analysis of patients with single metastatic brain tumors treated with LINAC SRS at the University of Florida between 1992 and 2009 who had at least one MRI after treatment. A total of 218 patients with 406 follow-up MRI scans were included in the study. Tumor area was calculated by measuring the largest tumor area on axial imaging and using the equation for area of an ellipse. Primary outcome was percent change in tumor size. The contribution of several factors including gender, primary tumor histology, synchronous or asynchronous presentation, prior treatment, primary tumor control, and SRS dose were examined using multivariate analysis. RESULTS: Mean patient age was 58.3 years (range 4-86), and 48.6% of patients were female. Sixty-three percent of patients had primary tumor control and 70.6% had asynchronous presentation of their brain metastases. SRS peripheral dose range was 1,000-2,250 cGy with a median of 1,750 cGy. The mean percent size change was -22.6% with a mean rate of change of -7.0% per month. The median percent change was -49.7% with a median rate of change of -8.8% per month. The median follow-up was 4.8 months (range 0.3-52.5). Female gender and melanoma histology were found to be significant predictors of an increase in tumor size. Lack of previous surgical resection was a significant predictor of a decrease in tumor size after SRS. Other factors tested with multivariate analysis, including age, synchronicity of presentation, dose, dose volume, Karnofsky performance score, and primary tumor control, were not significant in predicting tumor size change after SRS. CONCLUSION: In this study, brain metastases decreased in size by a median of 50% for a median follow-up of 4.8 months after SRS. The overall rate of decrease was 9% per month after treatment with SRS. Melanoma histology was a predictor of poor tumor control.