Decker, Paul A.

1509 Background: Lack of MGMT methylation in GBM tumors is associated with a worse outcome with combined radiation (RT) and temozolomide (TMZ) therapy, and it is unclear whether non-methylated tumors benefit from TMZ treatment. Methods: The potential benefits of concomitant or adjuvant TMZ relative to methylation status were evaluated in 10 GBM tumors established from patient samples and maintained by serial passage in mice. For each tumor line, 32 mice with established orthotopic xenografts were randomized to A) placebo, B) TMZ PO daily × 5 (120 mg/kg or 66 mg/kg, as indicated) C) RT, 2 Gy bid × 10 fx., D) combined TMZ and RT, and mice were followed until moribund. Efficacy was evaluated by regression modeling and Kaplan-Meier statistical analyses. Results: Six lines with non-methylated MGMT (GBM10, 14, 26, 28, 36, 43) and 4 with methylated MGMT (GBM8, 12, 16, 34) were evaluated, and complete results are available in 6 lines. In all tumor lines and at both dose levels, there was a significant benefit for TMZ relative to placebo (p<0.003). Both methylated tumors (GBM8 and 12) and one non-methylated tumor (GBM14) were markedly sensitive to TMZ with greater than a 50 day extension in survival with monotherapy. Individually and in a pooled analysis, there was no evidence for a synergistic interaction between TMZ and RT in the non-methylated tumors (p=0.91 for pooled analysis). In contrast, there was a striking degree of synergy in GBM12 (p<0.001) and at least additive effects of combined treatment in GBM8 (5 of 8 mice still surviving in the RT/TMZ arm). In a pooled analysis of these 2 MGMT methylated tumors, there was a trend towards synergy (p=0.11). Conclusions: TMZ treatment did not sensitize MGMT non-methylated tumors to radiation, and with one exception, these non-methylated tumors were only modestly sensitive to TMZ alone. In contrast, treatment with TMZ appears to sensitize MGMT-methylated tumors to radiation and monotherapy is associated with marked and significant benefit. [Table: see text] No significant financial relationships to disclose.

7555 Background: Lymph node status is a major determinant of stage and survival in patients with lung cancer; however, little information is available about the yield of a mediastinal lymphadenectomy done at the time of pulmonary resection. Methods: The ACOSOG Z0030 trial is a prospective, randomized trial of mediastinal lymph node sampling versus complete mediastinal lymphadenectomy during an operation for early stage lung cancer. Total enrollment from July 1999 to February 2004 was 1,111 patients, of which 1,023 were eligible and/or evaluable. There were 524 patients who underwent complete mediastinal lymph node resection after randomization to this arm that were declared eligible and/or evaluable with lymph node data available. The number of lymph nodes examined from each station was collected beginning in January 2002. Prospectively collected data from these patients was analyzed to determine the number of lymph nodes obtained. Results: Median age was 67 (range 37–87) and 267 (52%) were men. Histology was squamous cell in 141 (27%), adenocarcinoma in 227 (44%), large cell in 22 (4%), bronchoavelolar in 32 (6%) and other non-small cell in 99 (19%). There were 317 right sided cancers and 207 left sided cancers. For lymphadenectomy for cancers in the right lung the yield from station 2R was a median of 2 lymph nodes (range 1 to 15), station 4R was 2 (1 –17), station 7 was 2 (1–24), station 8 was 1 (1–5), station 9 was 1 (1–6) and station 10R was 1 (1–10). For lymphadenectomy for cancers on the left side the yield from station 2L was 2 (1–4), station 4L was 1 (1–12), station 5 was 2 (1–18), station 6 was 2 (1–11), station 7 was 2 (1–16), station 8 was 1 (1–3), station 9 was 1 (1–8) and 10L was 2 (1–12). The total number of lymph nodes or fragments obtained for right sided cancers was a median of 13.5 (range 1 to 56) and for left sided tumors 15 (range 4 to 81). Conclusions: Although high variability exists in the actual number of lymph nodes obtained from various nodal stations, a complete mediastinal lymphadenectomy should obtain one or more lymph nodes from each mediastinal station. Adequate mediastinal lymphadenectomy should include exploration and remove of lymph nodes from stations 2R, 4R, 7, 8, and 9 for right sided cancers and stations 4L, 5, 6, 7, 8 and 9 for left sided cancers. No significant financial relationships to disclose.

Background: Current cardiovascular disease (CVD) risk scores are derived from research cohorts and are particularly inaccurate in women, older adults, and those with missing data. To overcome these limitations, we aimed to develop a cohort to capitalize on the depth and breadth of clinical data within electronic health record (EHR) systems in order to develop next-generation sex-specific risk prediction scores for incident CVD. Methods: All individuals 30 years of age or older residing in Olmsted County, Minnesota on 1/1/2006 were identified. We developed and validated algorithms to define a variety of risk factors, thus building a comprehensive risk profile for each patient. Outcomes including myocardial infarction (MI), percutaneous intervention (PCI), coronary artery bypass graft (CABG), and CVD death were ascertained through 9/30/2017. Results: We identified 73,069 individuals without CVD (Table). We retrieved a total of 14,962,762 lab results; 14,534,466 diagnoses; 17,062,601 services/procedures; 1,236,998 outpatient prescriptions; 1,079,065 heart rate measurements; and 1,320,115 blood pressure measurements. The median number of blood pressure and heart rate measurements ascertained per individuals were 11 and 9, respectively. The five most prevalent conditions were: hypertension, hyperlipidemia, arthritis, depression, and cardiac arrhythmias. During follow-up 1,455 MIs, 1,581 PCI, 652 CABG, and 2,161 CVD-related deaths occurred. Conclusions: We developed a cohort with comprehensive risk profiles and follow-up for each patient. Using sophisticated machine learning approaches, this electronic cohort will be utilized to develop next-generation sex-specific CVD risk prediction scores. These approaches will allow us to address several challenges with use of EHR data including the ability to 1) deal with missing values, 2) assess and utilize a large number of variables without over-fitting, 3) allow non-linear relationships, and 4) use time-to-event data.

Background: The focus of most existing phenotyping algorithms based on electronic health record (EHR) data has been to accurately identify cases and non-cases of specific diseases. However, a more challenging task is to accurately identify disease incidence, as identifying the first occurrence of disease is more important for efficient and valid clinical and epidemiological research. Moreover, stroke is a challenging phenotype due to diagnosis difficulty and common miscoding. This task generally requires utilization of multiple types of EHR data (e.g., diagnoses and procedure codes, unstructured clinical notes) and a more robust algorithm integrating both natural language processing and machine learning. In this study, we developed and validated an EHR-based classifier to accurately identify stroke incidence among a cohort of atrial fibrillation (AF) patients Methods: We developed a stroke phenotyping algorithm using International Classification of Diseases, Ninth Revision (ICD-9) codes, Current Procedural Terminology (CPT) codes, and expert-provided keywords as model features. Structured data was extracted from Rochester Epidemiology Project (REP) database. Natural Language Processing (NLP) was used to extract and validate keyword occurrence in clinical notes. A window of ±30 days was considered when including/excluding keywords/codes into the input vector. Frequencies of keywords/codes were used as input feature sets for model training. Multiple competing models were trained using various combinations of feature sets and two machine learning algorithms: logistic regression and random forest. Training data were provided by two nurse abstractors and included validated stroke incidences from a previously established atrial fibrillation cohort. Precision, recall, and F-score of the algorithm were calculated to assess and compare model performances. Results: Among 4,914 patients with atrial fibrillation, 1,773 patients were screened. 3,141 patients had no stroke-related codes or keywords and were presumed to be free of stroke during follow-up. Among the screened patients, 740 had validated strokes and 1,033 did not have a stroke based on review of the EHR by trained nurse abstractors. The best performing stroke incidence phenotyping classifier utilized Keywords+ICD-9+CPT features using a random forest classifier, achieving a precision of 0.942, recall of 0.943, and F-score of 0.943. Conclusion: In conclusion, we developed and validated a stroke algorithm that performed well for identifying stroke incidence in an enriched population (AF cohort), which extends beyond the typical binary case/non-case stroke identification problem. Future work will involve testing the generalizability of this algorithm in a general population.

Purpose: Develop and evaluate a mammography intervention that provides hope about cancer prevention and treatment. Design: Pilot randomized controlled trial. Setting: Two communities on the Navajo Nation. Participants: Navajo women and support persons. Intervention: Both groups received standard care: one home visit discussing mammography pros/cons and barriers. The treatment group received an intervention based on Navajo language via an additional home visit with health education materials (written and oral) in English and Navajo, including a Navajo Cancer Glossary with a new descriptive phrase for cancer. Analysis: Between control and intervention conditions, we compared baseline sociodemographics; changes from baseline to 3 months on mammography completion and breast cancer literacy scores. Measures: (1) intervention feasibility; (2) self- and clinic-reported mammography screening completion; (3) breast cancer literacy. Results: A total of 25 participants were randomly assigned (13 treatment, 12 control), with 7 support persons in each arm. Mean age was 53 years, 90% had a high school degree or higher, 86% spoke Navajo and English. At 3 months, 44% had a clinically verified mammogram. Mammography completion was 57% among those with a support person and 27% among those without ( P = .14). Intervention women reported more breast cancer beliefs consistent with mammography ( P = .015). Conclusions: Increases in breast cancer beliefs consistent with mammography show promise. Findings highlight a need to tailor education materials to Navajo culture/language and focus on enhancing support.

Background: Bronchial carcinoid often appears hypervascular on bronchoscopic visualization and may be associated with hemoptysis. The diagnostic yield and bleeding complications associated with bronchoscopic biopsy of bronchial carcinoid tumors remain unclear. Materials and Methods: Patients with bronchial carcinoid tumors that were bronchoscopically visualized and biopsied at our tertiary referral medical center, over an 8-year period from 2010 to 2017, were retrospectively identified and reviewed to assess diagnostic yield and bleeding complications. Correlations with patient characteristics and carcinoid tumor features were analyzed. Results: Forty-nine patients were included (57% female). Tumors were predominantly (71%) located in proximal airways (mainstem and lobar bronchi). Bronchoscopic biopsy was diagnostic in 45 patients (92%). Thirteen patients (27%) experienced moderate (n=12, 25%) or severe (n=1, 2%) bleeding. Among these, 6 tumors (46%) had a vascular appearance and 4 patients (31%) had experienced recent hemoptysis. However, neither vascularity nor hemoptysis was associated with bleeding at biopsy (P=0.68 and 0.73, respectively). Carcinoid tumors were classified as typical in 79% and atypical in 21% with no difference in diagnostic yield or bleeding risk (P=0.28 and 0.92, respectively). Tumor size was also not associated with increased diagnostic yield or bleeding risk (P=0.54 and 0.39, respectively). Conclusion: Bronchoscopic biopsy of endobronchial carcinoid is associated with a high diagnostic yield and severe bleeding is rarely encountered. Diagnostic yield and bleeding seemed independent of vascular tumor appearance or history of recent hemoptysis.

Abstract Purpose: Glioblastoma is the most frequent and lethal primary brain tumor. Development of novel therapies relies on the availability of relevant preclinical models. We have established a panel of 96 glioblastoma patient-derived xenografts (PDX) and undertaken its genomic and phenotypic characterization. Experimental Design: PDXs were established from glioblastoma, IDH-wildtype (n = 93), glioblastoma, IDH-mutant (n = 2), diffuse midline glioma, H3 K27M-mutant (n = 1), and both primary (n = 60) and recurrent (n = 34) tumors. Tumor growth rates, histopathology, and treatment response were characterized. Integrated molecular profiling was performed by whole-exome sequencing (WES, n = 83), RNA-sequencing (n = 68), and genome-wide methylation profiling (n = 76). WES data from 24 patient tumors was compared with derivative models. Results: PDXs recapitulate many key phenotypic and molecular features of patient tumors. Orthotopic PDXs show characteristic tumor morphology and invasion patterns, but largely lack microvascular proliferation and necrosis. PDXs capture common and rare molecular drivers, including alterations of TERT, EGFR, PTEN, TP53, BRAF, and IDH1, most at frequencies comparable with human glioblastoma. However, PDGFRA amplification was absent. RNA-sequencing and genome-wide methylation profiling demonstrated broad representation of glioblastoma molecular subtypes. MGMT promoter methylation correlated with increased survival in response to temozolomide. WES of 24 matched patient tumors showed preservation of most genetic driver alterations, including EGFR amplification. However, in four patient–PDX pairs, driver alterations were gained or lost on engraftment, consistent with clonal selection. Conclusions: Our PDX panel captures the molecular heterogeneity of glioblastoma and recapitulates many salient genetic and phenotypic features. All models and genomic data are openly available to investigators.

Abstract Genome-wide association studies (GWAS) have revealed that 25 regions in 24 genes are associated with adult diffuse glioma development. These regions were identified by performing GWAS of glioma overall and GWAS by pathology (GBM and nonGBM). Subsequently, these regions have been evaluated for associations with specific molecular subtypes. The 2016 WHO Classification of Tumors of the Central Nervous System utilizes two somatic alterations to molecularly-classify adult diffuse glioma: IDH mutation and 1p/19q codeletion. TERT promoter mutation has also been shown to be associated with age at diagnosis and patient outcome. We hypothesized that germline variants may increase susceptibility to, or interact with, these somatic alterations to accelerate the development of specific molecular subtypes of glioma. To test our hypothesis, we performed a GWAS by glioma molecular subtype – as defined by presence or absence of IDH and TERT somatic mutation and 1p/19q codeletion – utilizing a two-stage design and subsequent meta analysis that included 3001 total glioma cases and 2697 total controls. Data were imputed using the Michigan Server and logistic regression was used, adjusting for age and sex. The Cancer Genome Atlas (TCGA) data were used to perform an expression quantitative trait loci (eQTL) analysis on candidate germline variants. Variants in 2q37 and 7p22 were associated with IDH-mutated glioma (meta analysis p< 5x10-8). The eQTL analyses demonstrated significant associations between 2q37 variants and expression of nearby genes as well as associations between 7p22 variants and nearby genes (p< 0.0001). In conclusion, we identified and validated novel germline variants in two genes that are associated with etiology of IDH-mutated adult diffuse glioma.

Abstract BACKGROUND Oligodendrogliomas are classified as either WHO grade II or III depending on histologic features. Grade often influences treatment decisions. However, there is variability in patient outcome within tumors of similar grade. We hypothesized that copy number burden (CNB) and specific copy number variants (CNV) might be associated with oligodendroglioma grade and prognosis. METHODS Copy number array analyses were performed on 285 molecular oligodendrogliomas (IDH-mutant, 1p/19q-whole arm-codeleted) from the Mayo Clinic internal and consult neuropathology practice and 167 TCGA molecular oligodendrogliomas. CNB was defined as the total number of copy number alterations. The association of CNB and CNV with grade and overall survival (when available) was assessed. All Mayo and TCGA data were evaluated using the ChAS software suite (Thermo- Fisher) and blindly reviewed by a clinical cytogeneticist (RBJ). RESULTS The mean CNB was 5.0 and 10.4 in the Mayo WHO grade II and III oligodendrogliomas, respectively (p = 5.4 x 10–17). Among the TCGA WHO grade II and III oligodendrogliomas the mean CNB was 4.4 and 5.3, respectively (p = 0.034). Common CNVs (occurring in at least 5% of cases) were -4/4q-, +8/8q+, -9/9p-/cnLOH 9p, +11/11q+, -14/14q-, -15 and -18/18q-. Of these, -9/9p-/cnLOH 9p was significantly associated with higher grade in both the Mayo and TCGA cohorts (p = 8.3 x 10-10 and 0.018, respectively). In the TCGA cohort the presence of >10 CNVs or +11/11q+ was associated with a poorer survival (p = 0.016 and 0.006, respectively). CONCLUSIONS CNB is significantly associated with WHO grade in IDH-mutant, 1p/19q co-deleted oligodendrogliomas. The presence of a significantly elevated CNB in some WHO grade II tumors may be suspicious for the presence of unappreciated grade III histologic features. Longer follow-up will be necessary to determine if CNB or single CNVs are associated with survival.