Duncan Whitney

A Bronchial Genomic Classifier for the Diagnostic Evaluation of Lung Cancer

BACKGROUND Bronchoscopy is frequently nondiagnostic in patients with pulmonary lesions suspected to be lung cancer. This often results in additional invasive testing, although many lesions are benign. We sought to validate a bronchial-airway gene-expression classifier that could improve the diagnostic performance of bronchoscopy. METHODS Current or former smokers undergoing bronchoscopy for suspected lung cancer were enrolled at 28 centers in two multicenter prospective studies (AEGIS-1 and AEGIS-2). A gene-expression classifier was measured in epithelial cells collected from the normal-appearing mainstem bronchus to assess the probability of lung cancer. RESULTS A total of 639 patients in AEGIS-1 (298 patients) and AEGIS-2 (341 patients) met the criteria for inclusion. A total of 43% of bronchoscopic examinations were nondiagnostic for lung cancer, and invasive procedures were performed after bronchoscopy in 35% of patients with benign lesions. In AEGIS-1, the classifier had an area under the receiver-operating-characteristic curve (AUC) of 0.78 (95% confidence interval [CI], 0.73 to 0.83), a sensitivity of 88% (95% CI, 83 to 92), and a specificity of 47% (95% CI, 37 to 58). In AEGIS-2, the classifier had an AUC of 0.74 (95% CI, 0.68 to 0.80), a sensitivity of 89% (95% CI, 84 to 92), and a specificity of 47% (95% CI, 36 to 59). The combination of the classifier plus bronchoscopy had a sensitivity of 96% (95% CI, 93 to 98) in AEGIS-1 and 98% (95% CI, 96 to 99) in AEGIS-2, independent of lesion size and location. In 101 patients with an intermediate pretest probability of cancer, the negative predictive value of the classifier was 91% (95% CI, 75 to 98) among patients with a nondiagnostic bronchoscopic examination. CONCLUSIONS The gene-expression classifier improved the diagnostic performance of bronchoscopy for the detection of lung cancer. In intermediate-risk patients with a nondiagnostic bronchoscopic examination, a negative classifier score provides support for a more conservative diagnostic approach. (Funded by Allegro Diagnostics and others; AEGIS-1 and AEGIS-2 ClinicalTrials.gov numbers, NCT01309087 and NCT00746759.).

Derivation of a bronchial genomic classifier for lung cancer in a prospective study of patients undergoing diagnostic bronchoscopy

BackgroundThe gene expression profile of cytologically-normal bronchial airway epithelial cells has previously been shown to be altered in patients with lung cancer. Although bronchoscopy is often used for the diagnosis of lung cancer, its sensitivity is imperfect, especially for small and peripheral suspicious lesions. In this study, we derived a gene expression classifier from airway epithelial cells that detects the presence of cancer in current and former smokers undergoing bronchoscopy for suspect lung cancer and evaluated its sensitivity to detect lung cancer among patients from an independent cohort.MethodsWe collected bronchial epithelial cells (BECs) from the mainstem bronchus of 299 current or former smokers (223 cancer-positive and 76 cancer-free subjects) undergoing bronchoscopy for suspected lung cancer in a prospective, multi-center study. RNA from these samples was run on gene expression microarrays for training a gene-expression classifier. A logistic regression model was built to predict cancer status, and the finalized classifier was validated in an independent cohort from a previous study.ResultsWe found 232 genes whose expression levels in the bronchial airway are associated with lung cancer. We then built a classifier based on the combination of 17 cancer genes, gene expression predictors of smoking status, smoking history, and gender, plus patient age. This classifier had a ROC curve AUC of 0.78 (95% CI, 0.70-0.86) in patients whose bronchoscopy did not lead to a diagnosis of lung cancer (n = 134). In the validation cohort, the classifier had a similar AUC of 0.81 (95% CI, 0.73-0.88) in this same subgroup (n = 118). The classifier performed similarly across a range of mass sizes, cancer histologies and stages. The negative predictive value was 94% (95% CI, 83-99%) in subjects with a non-diagnostic bronchoscopy.ConclusionWe developed a gene expression classifier measured in bronchial airway epithelial cells that is able to detect lung cancer in current and former smokers who have undergone bronchoscopy for suspicion of lung cancer. Due to the high NPV of the classifier, it could potentially inform clinical decisions regarding the need for further invasive testing in patients whose bronchoscopy is non diagnostic.

Shared Gene Expression Alterations in Nasal and Bronchial Epithelium for Lung Cancer Detection

Background: We previously derived and validated a bronchial epithelial gene expression biomarker to detect lung cancer in current and former smokers. Given that bronchial and nasal epithelial gene expression are similarly altered by cigarette smoke exposure, we sought to determine if cancer-associated gene expression might also be detectable in the more readily accessible nasal epithelium. Methods: Nasal epithelial brushings were prospectively collected from current and former smokers undergoing diagnostic evaluation for pulmonary lesions suspicious for lung cancer in the AEGIS-1 (n = 375) and AEGIS-2 (n = 130) clinical trials and gene expression profiled using microarrays. All statistical tests were two-sided. Results: We identified 535 genes that were differentially expressed in the nasal epithelium of AEGIS-1 patients diagnosed with lung cancer vs those with benign disease after one year of follow-up (P < .001). Using bronchial gene expression data from the AEGIS-1 patients, we found statistically significant concordant cancer-associated gene expression alterations between the two airway sites (P < .001). Differentially expressed genes in the nose were enriched for genes associated with the regulation of apoptosis and immune system signaling. A nasal lung cancer classifier derived in the AEGIS-1 cohort that combined clinical factors (age, smoking status, time since quit, mass size) and nasal gene expression (30 genes) had statistically significantly higher area under the curve (0.81; 95% confidence interval [CI] = 0.74 to 0.89, P = .01) and sensitivity (0.91; 95% CI = 0.81 to 0.97, P = .03) than a clinical-factor only model in independent samples from the AEGIS-2 cohort. Conclusions: These results support that the airway epithelial field of lung cancer–associated injury in ever smokers extends to the nose and demonstrates the potential of using nasal gene expression as a noninvasive biomarker for lung cancer detection.

Alterations in Bronchial Airway miRNA Expression for Lung Cancer Detection

We have previously shown that gene expression alterations in normal-appearing bronchial epithelial cells can serve as a lung cancer detection biomarker in smokers. Given that miRNAs regulate airway gene expression responses to smoking, we evaluated whether miRNA expression is also altered in the bronchial epithelium of smokers with lung cancer. Using epithelial brushings from the mainstem bronchus of patients undergoing bronchoscopy for suspected lung cancer (as part of the AEGIS-1/2 clinical trials), we profiled miRNA expression via small-RNA sequencing from 347 current and former smokers for which gene expression data were also available. Patients were followed for one year postbronchoscopy until a final diagnosis of lung cancer (n = 194) or benign disease (n = 153) was made. Following removal of 6 low-quality samples, we used 138 patients (AEGIS-1) as a discovery set to identify four miRNAs (miR-146a-5p, miR-324-5p, miR-223-3p, and miR-223-5p) that were downregulated in the bronchial airway of lung cancer patients (ANOVA P < 0.002, FDR < 0.2). The expression of these miRNAs is significantly more negatively correlated with the expression of their mRNA targets than with the expression of other nontarget genes (K-S P < 0.05). Furthermore, these mRNA targets are enriched among genes whose expression is elevated in cancer patients (GSEA FDR < 0.001). Finally, we found that the addition of miR-146a-5p to an existing mRNA biomarker for lung cancer significantly improves its performance (AUC) in the 203 samples (AEGIS-1/2) serving an independent test set (DeLong P < 0.05). Our findings suggest that there are miRNAs whose expression is altered in the cytologically normal bronchial epithelium of smokers with lung cancer, and that they may regulate cancer-associated gene expression differences. Cancer Prev Res; 10(11); 651–9. ©2017 AACR.

Abstract 1954: microRNA expression in bronchial epithelium for lung cancer detection

Introduction We have previously shown that gene expression alterations in the cytologically-normal mainstem bronchus can be leveraged as a biomarker for lung cancer detection (Silvestri et al. NEJM 2015), a test that is now used clinically. Extending this approach, we hypothesized that bronchial microRNA (miRNA) expression is altered in patients with lung cancer and that incorporating miRNA expression into the mRNA classifier may improve its performance. Methods Using bronchial brushes collected prospectively from current and former smokers undergoing bronchoscopy for suspect lung cancer across 28 medical centers as part of the AEGIS 1 and 2 clinical trials, we profiled miRNA expression via small RNA sequencing of 341 patients for which gene expression data was also available on the same bronchial brush sample. Patients were followed for up to one year post-bronchoscopy until a final diagnosis was established. 138 patients from AEGIS 1 (88 cancer-positive and 50 cancer-free) served as a discovery set, while other 203 patients from AEGIS 1 and 2 (103 cancer-positive and 100 cancer-free) were used as an independent test set. First, we identified miRNAs whose expression is associated with cancer by linear modeling in the discovery set. We next explored the relationships between the expression of these miRNAs and their predicted mRNA targets. Lastly, using logistic regression, we incorporated a cancer miRNA feature into our bronchial gene-expression classifier (Silvestri et al., NEJM 2015) and validated its performance in the test set. Results We found that expression profiles of 42 miRNAs were associated with cancer status in the discovery set (p Conclusions We have established that there are alterations in miRNA expression in the cytologically normal mainstem bronchus of smokers with lung cancer. Importantly, we demonstrated the potential of these miRNA alterations to improve the performance of an existing bronchial gene expression biomarker for lung cancer detection. Citation Format: Ana Brandusa Pavel, Joshua Campbell, Gang Liu, Sherry Zhang, Hanqiao Liu, Ji Xiao, Kate Porta, Duncan Whitney, Steven Dubinett, David Elashoff, Marc Lenburg, Avrum Spira. microRNA expression in bronchial epithelium for lung cancer detection. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 1954.

Abstract 3077: Dysregulation of microRNA-mRNA regulatory networks in the bronchial airway epithelium of smokers with lung cancer

We have previously shown that gene expression alterations in cytologically normal epithelial cells from the bronchial airway can be used as an early detection biomarker for lung cancer in smokers. We hypothesize that bronchial epithelial expression of microRNAs, as regulators of gene expression, may also be affected by the presence of cancer and may regulate some of these gene expression differences. We propose a novel method to identify microRNAs functionally associated with disease that leverages the relationship between microRNA and mRNA expression by determining the differential connectivity (DC) of microRNA-mRNA association networks between disease and normal states. Bronchial epithelial brushes were collected from 220 former and current smokers who underwent bronchoscopy for suspicion of lung cancer (120 lung cancer patients and 100 healthy controls). For these subjects, we profiled microRNA expression via small RNA sequencing and gene expression via microarray. Each microRNA node is assigned a DC score, which captures the overall difference in the pairwise microRNA-gene correlation strengths between lung cancer and control subjects. We quantify the change in both the directionality and strength of the correlations between a microRNA and the gene nodes. Then, the observed DC scores are compared to the DC scores obtained with permuted class labels to identify microRNAs with signinficant disease-specific differences in microRNA-mRNA connectivity. The proposed DC method identifies 54 microRNAs which are significantly differentially connected in lung cancer cases compared to controls (FDR We propose a novel approach for integrating microRNA and gene expression data to identify disease-associated changes in gene regulation by microRNAs and show that the microRNA-mRNA networks are significantly different between disease and normal states. These data suggest that changes in microRNA expression may drive some of the gene expression alterations observed in the cytologically normal epithelium from the proximal airway of patients with lung cancer and that airway microRNA-mRNA expression changes may ultimately serve as a biomarker for lung cancer detection. Citation Format: Ana Brandusa Pavel, Joshua D. Campbell, Gang Liu, Sherry Zhang, Hanqiao Liu, Lingqi Luo, Ji Xiao, Kate Porta, Duncan Whitney, Steven Dubinett, David Elashoff, Marc E. Lenburg, Avrum Spira. Dysregulation of microRNA-mRNA regulatory networks in the bronchial airway epithelium of smokers with lung cancer. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 3077. doi:10.1158/1538-7445.AM2015-3077

Abstract A1-66: Leveraging Gene Expression in the Bronchial Airway to Develop a Nasal Biomarker for Early Detection of Lung Cancer

Rationale: Lung cancer results in five times more deaths per year than car accidents in the United States. Approximately 57% of lung cancers diagnosed this year will be diagnosed at a late stage and these patients will exhibit a 5-year survival rate of only 4%. Annual screening of high-risk current and former smokers by chest-CT can reduce cancer mortality, however this procedure has a 95% false positive rate. It is therefore critical to develop methods to rapidly and accurately determine which patients with nodules on chest CT have lung cancer and potentially spare those with benign disease an unnecessary invasive procedure. We have previously demonstrated that specific gene expression alterations in cytologically normal bronchial epithelial cells from patients with lung cancer can be leveraged to form a clinically informative lung cancer biomarker in the population of patients undergoing bronchoscopy for suspect lung cancer. We hypothesized that there might be similar expression differences in nasal epithelium and that these could form the basis of a less invasive test that could be applied more broadly to individuals with screen detected nodules on chest CT. Methods: Bronchial (n=676) and nasal (n=280) epithelial brushings were collected from current and former smokers undergoing bronchoscopy for clinical suspicion of lung cancer within the AEGIS clinical trial. 271 subjects had matched bronchial and nasal samples. RNA was extracted and hybridized to Affymetrix Human Gene ST 1.0 Arrays. To establish a connection between bronchial and nasal epithelial gene expression signal for cancer, we first applied the bronchial gene expression-based diagnostic test, BronchoGen, directly to our nasal cohort. Gene Set Enrichment Analysis was then used to determine the concordance of cancer signal between the bronchial and nasal epithelium. To develop the nasal gene expression biomarker for lung cancer detection, we examined the correlation of each gene between the bronchial and nasal epithelium as well as the significance of each gene9s association with cancer in each tissue. Genes passing our selection criteria were passed to a biomarker discovery pipeline in which we examined the performance of different biomarker algorithm configurations (e.g. feature-selection algorithms, classification algorithms, and other biomarker parameters) using cross-validation. Results: Direct application of BronchoGen to our nasal cohort resulted in an AUC of 0.64 on a set of NE samples (n=110) with a matched bronchial sample in the training set used to develop the test. On an independent set of nasal samples (n=109), BronchoGen achieved an AUC of 0.67. Gene Set Enrichment Analysis revealed high levels of concordance between cancer-associated nasal and bronchial gene expression. Using a cross-validation approach, we found that nasal biomarkers built from sets of genes showing significant correlation (p Conclusions: Given the larger sample size, more isolated location in the airway, and higher RIN scores that characterize the bronchial cohort, we sought to leverage bronchial airway epithelial gene-expression to inform which genes in the nasal epithelium should be indicative of the presence of cancer. We have shown that gene expression in the nasal epithelium reflects the presence of lung cancer and can serve as a diagnostic biomarker. We have further demonstrated concordance between bronchial and nasal airway gene expression differences associated with lung cancer. These results suggest the potential to develop a robust nasal gene expression biomarker for lung cancer diagnosis that leverages cancer-associated gene expression differences occurring at other airway sites. Citation Format: Joseph F. Perez-Rogers, Joseph Gerrein, Christina Anderlind, Rebecca L. Kusko, Joshua D. Campbell, Teresa W. Wang, Kate Porta, Duncan Whitney, Avrum Spira, Marc Lenburg. Leveraging Gene Expression in the Bronchial Airway to Develop a Nasal Biomarker for Early Detection of Lung Cancer. [abstract]. In: Proceedings of the AACR Special Conference on Translation of the Cancer Genome; Feb 7-9, 2015; San Francisco, CA. Philadelphia (PA): AACR; Cancer Res 2015;75(22 Suppl 1):Abstract nr A1-66.

Abstract 1574: Leveraging bronchial airway gene expression to develop a nasal biomarker for lung cancer detection

Proceedings: AACR 106th Annual Meeting 2015; April 18-22, 2015; Philadelphia, PA Rationale: Using nasal gene expression to predict the presence of lung cancer would offer a less invasive alternative to diagnostic approaches we have pioneered using bronchial airway epithelial (BE) gene expression. We have previously demonstrated that cytologically normal BE and nasal epithelial (NE) cells harbor gene expression differences that reflect tobacco-related lung disease and that these changes in the BE form the basis of a clinically informative lung cancer biomarker. Given the concordance of BE and NE gene-expression, we hypothesized that gene signatures associated with the presence of lung cancer extend from the airway to the nose and that lung cancer associated BE gene-expression could be leveraged to develop more accurate nasal lung cancer biomarkers. Methods: BE (n = 676) and NE (n = 280) brushings were collected from current and former smokers undergoing bronchoscopy for clinical suspicion of lung cancer. We leveraged two methods to determine the concordance between BE and NE gene-expression signal for cancer. First we applied the bronchial gene expression-based diagnostic test directly to our nasal cohort. Second, we used Gene Set Enrichment Analysis (GSEA) to quantify the relationship between the BE and NE. To develop the nasal gene expression biomarker, we examined the correlation of each gene between the BE and NE. Genes passing our selection criteria were passed to a biomarker discovery pipeline in which we examined the performance of different biomarker algorithm configurations using cross-validation. Results: Direct application of the bronchial airway gene-expression classifier to an independent set of nasal samples (n = 109) resulted in an AUC of 0.67. GSEA revealed high concordance (p<0.001) between cancer-associated nasal and bronchial gene expression profiles from the same patients. Using a cross-validation approach, we found that nasal biomarkers built from sets of genes showing significant correlation (p<0.05) between the BE and NE as well as significance for cancer in both tissues (p<0.05) perform better, on average, than biomarkers built from genes significant for cancer (p<0.05) in the NE alone. Conclusions. We have demonstrated concordance between BE and NE gene expression differences associated with lung cancer. We have further shown that gene expression in the NE reflects the presence of lung cancer and can serve as a diagnostic biomarker. These results demonstrate the feasibility of leveraging cancer-associated gene expression changes throughout the airway to develop a minimally invasive and robust nasal gene expression biomarker for lung cancer diagnosis. Citation Format: Joseph Perez-Rogers, Joseph Gerrein, Christina Anderlind, Xiaohui Xiao, Hanqiao Liu, Rebecca Kusko, Joshua Campbell, Teresa Wang, Yuriy Alekseyev, Gang Liu, Kate Porta, Duncan Whitney, Avrum Spira, Marc Lenburg. Leveraging bronchial airway gene expression to develop a nasal biomarker for lung cancer detection. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 1574. doi:10.1158/1538-7445.AM2015-1574

Abstract 1485: Biomarker development for lung cancer diagnosis using integrative microRNA and gene expression networks

Proceedings: AACR Annual Meeting 2014; April 5-9, 2014; San Diego, CA Introduction: We have previously shown that smoking creates a “molecular field of injury” throughout the epithelial cells that line the respiratory tract and that gene-expression alterations in the cytologically-normal mainstem bronchus epithelium can serve as an early detection biomarker for lung cancer. We hypothesize that microRNAs (miRNAs) regulate these airway gene expression changes and that miRNA expression differences in this tissue can be used as biomarkers for lung cancer diagnosis. Methods: We profiled miRNA expression via small RNA sequencing in bronchial epithelial brushes collected from the mainstem bronchus of 230 subjects undergoing bronchoscopy for suspect lung cancer and gene expression (mRNA) via microarray for 201 matched samples. Bilal et al. (2013) have shown that incorporating biological knowledge into model building improves prediction. Therefore, we sought to test the hypothesis that including information about the expression levels of the predicted mRNA targets of miRNA may improve miRNA feature selection and aid in interpretation of signatures. We used mirConnX which combines miRNA with mRNA data to create disease-specific, genome-wide regulatory networks. Results: First, we show that there is a miRNA expression signal for cancer: across many combinations of feature selection methods, predictive models and different parameters within 100 bootstraps to predict cancer phenotype, we find the AUC values obtained are consistently higher compared to the random control procedure where we randomly shuffle the class labels (p < 0.001). Second, using a training set of 106 samples, we built networks separately for cancer and non-cancer samples, using 10-fold cross validation in order to determine robust cancer and non-cancer specific features. The disease-state specific networks are then aggregated by taking the overlapping features across the ten folds. Next, we select the non-overlapping features (miRNAs and genes) between cancer and non-cancer as those that capture the difference between the two phenotypes. The selected genes are enriched for relevant cancer related pathways, including KEGG pathways in cancer (p = 0.0003), cell cycle (p = 0.008), WNT signaling (p = 0.0001), basal cell carcinoma (p = 0.023), MAPK signaling pathway (p = 0.024), TGF-beta signaling pathway (p = 0.00014), p53 signaling pathway (p = 0.087) etc. Most importantly, the miRNA features from these disease specific networks are found to have higher predictive power (highest AUC 0.72) compared to all miRNA features (highest AUC 0.57), on a second set of 68 samples in cross-validation. Conclusion: Using novel integrative analysis, we improved miRNA biomarker prediction. This is the first report of cancer-associated miRNA expression differences in cytologically normal bronchial epithelium, for lung cancer diagnosis, and it extends our previous work focused on mRNA biomarkers from this tissue. Citation Format: Ana Pavel, Joshua Campbell, Gang Liu, Sherry Zhang, Hanqiao Liu, Steven Dubinett, David Elashoff, Kate Porta, Duncan Whitney, Marc Lenburg, Avrum Spira. Biomarker development for lung cancer diagnosis using integrative microRNA and gene expression networks. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 1485. doi:10.1158/1538-7445.AM2014-1485

Abstract 1414: Race-associated variation in the airway transcriptome response to cigarette smoke

INTRODUCTION: Tobacco-induced lung cancer is the leading cause of cancer death. Lung cancer incidence is 15% higher in African Americans (AFA) than Caucasians (CAU) even after taking into account variations in smoking behavior and other environmental variables, and the reasons for this difference are unclear. As 90% of patients with lung cancer have a smoking history, we hypothesize that race-related differences in the response to tobacco smoke might contribute to the elevated lung cancer risk observed in AFA. To begin to test this hypothesis, we have used gene expression profiling to identify differences in the biologic response of the bronchial airway epithelium of AFA and CAU to tobacco smoke. METHODS: Cytologically normal bronchial epithelial cells obtained during bronchoscopy from current (C) and former (F) smokers with and without lung cancer were hybridized to Affymetrix Human Gene 1.0 ST microarrays (n=885). After quality control, n=810 samples (from n=710 subjects) were selected for subsequent analysis. RMA-normalized gene expression levels were batch-corrected using ComBat. Genes whose expression is altered by smoking differently in the airway epithelium of AFA (n=155; nC=76, nF=79) as compared to CAU (n=655; nC=283, nF=372), were identified via ANOVA using a linear model including terms for self-reported race, smoking status, and the interaction between race and smoking status, as well as other covariates including gender, cancer status, COPD status, age, cumulative cigarette smoke exposure, and RIN. Smoking-associated race-dependent pathways were identified using Enrichr. An independent dataset of cytologically normal bronchial epithelial cells hybridized to the Affymetrix HGU133A microarrays (nCAU=154, nAFA=58) was analyzed as detailed above and relationships between the two datasets were established using GSEA for the genes associated with race, and the interaction between race and smoking status. RESULTS: We identified 361 genes that displayed smoking-associated race-dependent patterns of expression (FDR CONCLUSIONS: These findings suggest that race influences the airway epithelial response to cigarette smoke at the level of individual genes and pathways. Our findings also suggest that gene expression profiling of the airway epithelium might be useful to understand the molecular basis of the elevated risk of lung cancer among African Americans. Citation Format: Anna Tassinari, Duncan Whitney, Kate Porta, Marc Lenburg, Avrum Spira, Jennifer Beane. Race-associated variation in the airway transcriptome response to cigarette smoke. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 1414. doi:10.1158/1538-7445.AM2014-1414