BACKGROUND:EarlyR gene signature in estrogen receptor-positive (ER+) breast cancer is computed from the expression values of ESPL1, SPAG5, MKI67, PLK1, and PGR. EarlyR has been validated in multiple cohorts profiled using microarrays. This study sought to verify the prognostic features of EarlyR in a case-cohort sample from BIG 1-98, a randomized clinical trial of ER+ postmenopausal breast cancer patients treated with adjuvant endocrine therapy (letrozole or tamoxifen).METHODS:Expression of EarlyR gene signature was estimated by Illumina cDNA-mediated Annealing, Selection, and Ligation assay of RNA from formalin-fixed, paraffin-embedded primary breast cancer tissues in a case-cohort subset of ER+ women (N = 1174; 216 cases of recurrence within 8 years) from BIG 1-98. EarlyR score and prespecified risk strata (≤25 = low, 26-75 = intermediate, >75 = high) were "blindly" computed. Analysis endpoints included distant recurrence-free interval and breast cancer-free interval at 8 years after randomization. Hazard ratios (HRs) and test statistics were estimated with weighted analysis methods.RESULTS:The distribution of the EarlyR risk groups was 67% low, 19% intermediate, and 14% high risk in this ER+ cohort. EarlyR was prognostic for distant recurrence-free interval; EarlyR high-risk patients had statistically increased risk of distant recurrence within 8 years (HR = 1.73, 95% confidence interval = 1.14 to 2.64) compared with EarlyR low-risk patients. EarlyR was also prognostic of breast cancer-free interval (HR = 1.74, 95% confidence interval = 1.21 to 2.62).CONCLUSIONS:This study confirmed the prognostic significance of EarlyR using RNA from formalin-fixed, paraffin-embedded tissues from a case-cohort sample of BIG 1-98. EarlyR identifies a set of high-risk patients with relatively poor prognosis who may be considered for additional treatment. Further studies will focus on analyzing the predictive value of EarlyR signature.

Cancer research is moving at a startling pace, most particularly with the identification and characterization of molecular signatures and biomarkers that will be used for prevention, early detection, diagnosis, treatment, prediction, and prognostication. The goal of cancer therapy is to match the right treatment with the right patient. To this end, several clinical trials are now employing patient stratification using “Massive Parallel Sequencing” or informally called “Next -Generation Sequencing” (NGS) to identify clinically actionable targets in real time. The omics revolution is yielding important new insights into the causes and mechanisms of diseases and drug responses and in understanding the effects of genes and environment in disease predisposition and acquired resistance. It is paving the way for precision medicine (PM) a.k.a. personalized medicine, which focuses its attention on factors specific to an …

Open-source software encourages computer programmers to reuse software components written by others. In evolutionary bioinformatics, open-source software comes in a broad range of programming languages, including C/C++, Perl, Python, Ruby, Java, and R. To avoid writing the same functionality multiple times for different languages, it is possible to share components by bridging computer languages and Bio* projects, such as BioPerl, Biopython, BioRuby, BioJava, and R/Bioconductor. In this chapter, we compare the three principal approaches for sharing software between different programming languages: by remote procedure call (RPC), by sharing a local "call stack," and by calling program to programs. RPC provides a language-independent protocol over a network interface; examples are SOAP and Rserve. The local call stack provides a between-language mapping, not over the network interface but directly in computer memory; examples are R bindings, RPy, and languages sharing the Java virtual machine stack. This functionality provides strategies for sharing of software between Bio* projects, which can be exploited more often. Here, we present cross-language examples for sequence translation and measure throughput of the different options. We compare calling into R through native R, RSOAP, Rserve, and RPy interfaces, with the performance of native BioPerl, Biopython, BioJava, and BioRuby implementations and with call stack bindings to BioJava and the European Molecular Biology Open Software Suite (EMBOSS). In general, call stack approaches outperform native Bio* implementations, and these, in turn, outperform "RPC"-based approaches. To test and compare strategies, we provide a downloadable Docker container with all examples, tools, and libraries included.

The precision medicine (PM) revolution is well underway, and next-generation sequencing (NGS) is helping take cancer treatment to new levels. Finding actionable targets in real time is now a reality, and data from several clinical trials based on identified molecular alterations can be assessed and can help address the question of whether personalized treatment based on genomics produces superior survival outcomes compared with unselected treatment. Targeted treatment-based clinical trials have shown benefit in molecular subgroups of patients. Of further interest, it appears that genomics and immunotherapy are coupled to each other, since the immune system recognizes neo-antigens produced by the mutanome. Hence, some of the most important markers predicting response to immunotherapy are genomic markers, PDL1 amplification (for PD-1/PD-L1 checkpoint inhibitors), and tumor mutational …

The maker revolution, in which 3D printers and other innovative technologies let us produce many types of prototypes and products quickly and easily, is profoundly affecting research and industry.

Purpose The Herceptin Adjuvant study is an international multicenter randomized trial that compared 1 or 2 years of trastuzumab given every 3 weeks with observation in women with human epidermal growth factor 2–positive (HER2+) breast cancer after chemotherapy. Identification of biomarkers predictive of a benefit from trastuzumab will minimize overtreatment and lower health care costs. Methods To identify possible single-gene biomarkers, an exploratory analysis of 3,669 gene probes not expected to be expressed in normal breast tissue was conducted. Disease-free survival (DFS) was used as the end point in a Cox regression model, with the interaction term between C8A mRNA and treatment as a categorical variable split on the cohort mean. Results A significant interaction between C8A mRNA and treatment was detected (P < .001), indicating a predictive response to trastuzumab treatment. For the C8A-low subgroup (mRNA expression lower than the cohort mean), no significant treatment benefit was observed (P = .73). In the C8A-high subgroup, patients receiving trastuzumab experienced a lower hazard of a DFS event by approximately 75% compared with those in the observation arm (hazard ratio [HR], 0.25; P < .001). A significant prognostic effect of C8A mRNA also was seen (P < .001) in the observation arm, where the C8A-high group hazard of a DFS event was three times the respective hazard of the C8A-low group (HR, 3.27; P < .001). C8A mRNA is highly prognostic in the Hungarian Academy of Science HER2+ gastric cancer cohort (HR, 1.72; P < .001). Conclusion C8A as a single-gene biomarker prognostic of DFS and predictive of a benefit from trastuzumab has the potential to improve the standard of care in HER2+ breast cancer if validated by additional studies. Understanding the advantage of overexpression of C8A related to the innate immune response can give insight into the mechanisms that drive cancer.

Abstract Background: RNA-Seq from total RNA in FFPE tissue can be more challenging due to limited capture of partially degraded RNA. Exome-capture based RNA-Seq may circumvent such problems and allow reproducible complete molecular characterization of low-quality RNA from small clinical samples. Methods: HER2 negative patients within the GeparQuinto trial were treated with neoadjuvant anthracycline-taxane-based chemotherapy +/- bevacizumab. Patients with bevacizumab therapy had a significantly higher pCR rate, especially within the triple negative subgroup. We performed exome-capture RNA-Seq on 5µm FFPE sections from pre-therapeutic cores of 400 HER2 negative samples from this trial. In a prospectively planned, blinded study we correlated molecular subtypes and metagenes for proliferation, stroma, MHC2, and VEGFA with clinical and histopathological data. Molecular subtypes were defined using the AIMS methods. Metagenes were calculated as mean values corresponding to previously described gene clusters after platform transfer (Rody et al. 2011 PMID 21978456, Hu et al. 2009 PMID 19291283) and then z-transformed. Results: 296 samples with RNA-Seq data were classified as either of high (n=226) or of limited quality (n=70). For 22 samples RNA yield was insufficient and 82 did not pass initial QC. 121 (41%), 63 (21%), 34 (11.5%), 46 (15.5%), and 32 (11%) samples were defined as basal-like, HER2-enriched, luminal A, luminal B, and normal-like, respectively. Subtyping was robust with regard to gene filtering, normalization, and sample quality. ER and PR status from local IHC strongly correlated with gene expression (overall correctness 84% and 80% for ER, and 85% and 74% for PR, in samples with high and limited quality, respectively) and luminal subtypes (95% ER positive). Proliferation metagene correlated with histological grade (median -0.73, -0.39, and 0.53 in G1, G2, and G3, respectively; P<0.001) and MHC2 metagene correlated strongly with TIL counts (Rho=0.53, P<0.001). Among the high quality samples response rates (49.3% pCR overall) differed significantly by subtype, with higher pCR rates in basal-like (68.9%) and HER2-enriched (45.5%) than in luminal B (35.7%), luminal A (17.9%), and normal-like (20.0%). MHC2- (OR 1.60, 95%CI 1.21-2.12, P=0.001), proliferation- (OR 2.88, 95%CI 2.00-4.16, P<0.001), and VEGFA-metagenes (OR 1.92, 95%CI 1.41-2.60, P<0.001) were significant predictors for pCR. In a multivariate logistic regression (adjusted for bevacizumab treatment and hormone receptor status) both VEGFA metagene (OR 2.59, 95%CI 1.40-4.77, P=0.002) and the interaction between the VEGFA-metagene and bevacizumab treatment arm (P=0.023) significantly predicted pCR. Conclusions: Exome-capture RNA-Seq allows robust genomic characterization of clinical samples with limited FFPE material from core biopsies, and molecular subtypes and immune metagenes are predictive for pCR. The VEGFA metagene is a specific predictor for response to neoadjuvant bevacizumab treatment. Citation Format: Karn T, Meissner T, Weber K, Sinn B, Denkert C, Budczies J, Nekljudova V, Fasching PA, Holtrich U, Schem C, Solbach C, Hartmann A, Röcken C, Untch M, Young BM, Willis S, Leyland-Jones B, von Minckwitz G, Loibl S. Blinded molecular subtyping analysis from RNA-Seq of FFPE samples in the GeparQuinto trial reveals predictive value of VEGFA metagene for bevacizumab treatment [abstract]. In: Proceedings of the 2017 San Antonio Breast Cancer Symposium; 2017 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2018;78(4 Suppl):Abstract nr P2-09-02.

Students are ready to embrace the technology-driven 21st century. If we make learning fun and challenging—and then get out of the way—they will readily learn the skills needed to succeed in the future.

Abstract Introduction Mutational processes can be characterized by unique combinations of mutation types in the form of mutational signatures and have been associated with age, known mutagenic exposures, defects in DNA maintenance, or the APOBEC family of cytidine deaminases. We asked whether mutation signatures could be extracted from DNA sequence information in a targeted 434 gene panel covering 297 breast cancer specimens. Materials and Methods Targeted whole exome sequencing (Illumina, 2x50bp) of a 434 gene panel was performed on a set of 297 primary and metastatic breast tumor samples. Tissue of origin included breast (56%), liver (15%), lymph node (10%), lung (3%) and others (16%). Alignment was done with BWA against the human reference hg19 and variant calling was performed using VarDict. Germline variants were filtered based on allele frequencies, cohort specific population frequencies, as well as using 1000 Genomes and ExAC population frequencies. For somatic signature inference, only single nucleotide variants were retained. Panel specific trinucleotide frequencies were computed and normalized towards whole genome frequencies and somatic signatures were inferred using deconstructSigs method. Results We identified a total of 26 signatures from the set of 30 known signatures in our patient samples. Due to the small panel size, there was only a limited number of mutations available per patient to infer somatic signatures. On average, we identified two somatic signatures per sample. Most common mutation signatures identified were: Signature 1 (90.8%) - result of an endogenous mutational process initiated by spontaneous deamination of 5-methylcytosine; Signature 6 (21.8%) - defective DNA mismatch repair; Signature 15 (15.6%) - defective DNA mismatch repair; Signatue 7 (9.9%) - ultraviolet light exposure; and Signature 10 (6.5%) - altered activity of POLE. An APOBEC specific signature was identified in 20 (7%) samples. APOBEC positive samples showed significantly higher tumor mutational burden (10.7 vs. 5.7 mutations/mb) as compared to APOBEC negative samples (p<=0.001). PIK3CA was found to be mutated in 80% of APOBEC positive samples, compared to 36% of APOBEC negative samples. In addition, we found higher rates of mutations in TP53 (70% vs. 50%), MLL3 (50% vs. 19%) and MLL2 (25% vs 14%) of APOBEC positive patients. Response rates of APOBEC positive patients were significantly worse than of APOBEC negative patients, with 50 percent of patients having progressive disease compared to 25 percent of APOBEC negative patients(p=0.07, borderline). Conclusions We demonstrate the feasibility of a targeted sequencing approach to extract somatic mutation signatures from breast tumor samples, and we highlight the potential of using the APOBEC signature to predict therapeutic responses. Citation Format: Meissner T, Amallraja A, Willis S, Harris R, Leyland-Jones B, Williams C. APOBEC mutation signature in breast cancer correlates with tumor mutation burden and poor responses to therapy [abstract]. In: Proceedings of the 2017 San Antonio Breast Cancer Symposium; 2017 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2018;78(4 Suppl):Abstract nr PD8-10.

The extent of tumor-infiltrating lymphocytes (TILs), along with immunomodulatory ligands, tumor-mutational burden and other biomarkers, has been demonstrated to be a marker of response to immune-checkpoint therapy in several cancers. Pathologists have therefore started to devise standardized visual approaches to quantify TILs for therapy prediction. However, despite successful standardization efforts visual TIL estimation is slow, with limited precision and lacks the ability to evaluate more complex properties such as TIL distribution patterns. Therefore, computational image analysis approaches are needed to provide standardized and efficient TIL quantification. Here, we discuss different automated TIL scoring approaches ranging from classical image segmentation, where cell boundaries are identified and the resulting objects classified according to shape properties, to machine learning-based approaches that directly classify cells without segmentation but rely on large amounts of training data. In contrast to conventional machine learning (ML) approaches that are often criticized for their "black-box" characteristics, we also discuss explainable machine learning. Such approaches render ML results interpretable and explain the computational decision-making process through high-resolution heatmaps that highlight TILs and cancer cells and therefore allow for quantification and plausibility checks in biomedical research and diagnostics.

Abstract Background: EarlyR is a prognostic gene signature score in ER+ breast cancer (BC) computed from the expression values of ESPL1, SPAG5, MKI67, PLK1 and PGR using a nonlinear mathematical formula. EarlyR has been validated in multiple cohorts profiled on Affymetrix and Illumina microarrays and by RNA-seq. This study sought to assess the prognostic features of EarlyR in a cohort of E2197. Patients and Methods: Illumina DASL assay was used to measure gene expression in FFPE tissue of primary BC from a case-cohort sampling subset of women in E2197 treated with doxorubicin plus docetaxel (AT) or doxorubicin plus cyclophosphamide (AC). ER+ patients received hormone therapy at physician's discretion. After 79.5 months median follow-up, disease-free survival was 85% in both treatment arms. Among patients centrally reviewed with sufficient RNA material for the DASL assay, 319 with ER+ status and assessed for EarlyR are included in the analytic cohort. EarlyR scores and pre-specified risk strata (≤25=low, 26-75=intermediate, >75=high) were computed, while blinded to clinical data. The analysis endpoint was disease-free survival (DFS), defined as the time from randomization to date of invasive BC recurrence or death from any cause within 8 years. Weighted Cox proportional hazards models were used to associate EarlyR score or risk strata with DFS. Variances of the estimated coefficients were adjusted to account for the case-cohort design. Results: The distribution of the EarlyR risk groups was 59% low, 11% intermediate and 30% high risk in this ER+ cohort. The continuous EarlyR score was significantly prognostic of DFS up to 8 years after randomization (p = 0.02). Patients with low EarlyR score (≤ 25) had significantly lower risk of BC recurrence within 8 years (p = 0.031, univariate HR=0.562, 95%CI: 0.334-0.948) compared to those with high EarlyR score (> 75). Analysis within the AC arm showed that patients with low EarlyR score had significantly lower risk of 8-year BC recurrence (p = 0.023, univariate HR=0.392, 95%CI: 0.175-0.878) compared to those with high EarlyR score. Within the AT arm there was no significant difference in 8-year DFS prognosis between any of the EarlyR risk groups. Conclusions: This study confirmed the prognostic significance of EarlyR using FFPE tissue in a cohort of patients treated with AC chemotherapy from E2197. Patients with high EarlyR score who were treated with AC had significantly higher risk of recurrence than low EarlyR score patients treated with AC. On the other hand, prognosis of high EarlyR score AT-treated patients was not significantly lower than the prognosis of low EarlyR score AT-treated patients. Further study in a larger cohort is needed to assess the relative benefits of AC versus AT within the EarlyR high risk group and the EarlyR low risk group. Citation Format: Badve S, Wang V, Willis S, Leyland-Jones B, Gokmen-Polar Y, Shulman L, Martino S, Sparano J, Davidson N, Goldstein L, Buechler S. Independent validation of EarlyR gene signature in E2197: A randomized clinical trial comparing doxorubicin plus docetaxel to doxorubicin plus cyclophosphamide as adjuvant chemotherapy in breast cancer [abstract]. In: Proceedings of the 2017 San Antonio Breast Cancer Symposium; 2017 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2018;78(4 Suppl):Abstract nr P1-06-08.

New editor Scooter Willis reflects on his journey to becoming an electrical engineer and how we can promote the work of students who've found their passion.

Recent advances in deep learning for image recognition have spawned numerous challenge-based learning competitions in which participants can use a low-cost GPU graphics card to accomplish goals that required expensive resources in the recent past. Students are encouraged to explore this exciting new field of research by entering these competitions.

Abstract INTRODUCTION: RAC1-GTPase which transduces signals from cell surface integrins, have been implicated in metastasis. We reported that Wnt-beta-catenin pathway (WP) that signals metastasis (BMC Cancer, 2013), is one of the salient genetic features of Triple-Negative Breast Cancer (TNBC) (PlosOne, 2013).AIM: We demonstrated that TNBC cells acquire integrin-directed metastasis-associated (ID-MA) phenotypes following an upregulation of the WP (Oncotarget, In Press). Here we examined how WP signals are transduced in the context of ID-MA phenotypes in TNBC.METHOD: We documented gain and amplification of RAC1 gene in Breast Invasive Carcinoma subtypes from cBioPortal. The outcome for RFS was studied in the Hungarian ER-ve BC cohort.Mechanistically, we studied fibronectin-directed (1) migration, (2) matrigel- invasion, (3) RAC1 activation, (4) actin dynamics (confocal microscopy) and (5) podia-parameters using pharmacological agents (sulindac sulfide), genetic tools (beta-catenin siRNA), WP modulators (Wnt-C59, XAV939), RAC1 inhibitors (NSC23766, W56) and WP stimulations (LWnt3ACM, Wnt3A recombinant) in a panel of 6-7 TNBC cell lines, RESULTS: The collective percentage of gain and amplification of RAC1 were (1) 31% of total 1105 breast invasive carcinoma samples, (2) 29% of total 594 ER+ve samples, (3) 39% of total 174 ER-ve samples, (4) 38% of total 120 HER2+ve samples and (4)35% of total 82 TNBC samples (brca/tcga/pub2015; Cell 2015).In invasive ductal BC subtypes, gain and amplification of RAC1 were (1) 32% of total 201 Luminal A samples, (2) 37% of total 122 PAM50 Luminal B samples, (3) 47% of total 51 PAM50 Her2-enriched samples and (4) 33% of total 107 PAM50 Basal-like samples. In invasive lobular cancers, gain and amplification of RAC1 were 24% of total 127 samples.Involvement of WP in different TNBC cells was tested following stimulation by LWnt3ACM and Wnt3Arecombinant protein and different inhibitors of WP by both qRT-PCR and WB for beta-catenin, active beta-catenin, cMYC, cyclin D1and WP specific several stem cell markers. The WP attenuation, which (a) decreased cellular levels of beta-catenin, as well as its nuclear active-form, (b) decreased fibronectin-induced migration & invasion, (c) altered actin dynamics and (d) decreased podia-parameters was successful in blocking fibronectin-mediated RAC1/Cdc42 activity. Both Wnt-antagonists and RAC1 inhibitors blocked fibronectin-induced RAC1 activation and inhibited fibronectin-induced ID-MA phenotypes following WP stimulation by LWnt3ACM and Wnt3Arecombinant protein. High expression of RAC1 was associated with poor outcome for RFS with HR=1.48 [CI: 1.15-1.9] p=0.0019 in the Hungarian ER-veBC cohort.CONCLUSION:In TNBC model, the activation of RAC1 signals downstream of WP mediated ID-MA phenotypes. The identification of the functional relationship between RAC1 signaling and the WP activation in the control of ID-MA mechanistically explains how the activation of WP in TNBC is associated with the high metastatic incidences and a dismal outcome. Citation Format: Dey N, Carlson JH, Jepperson T, Willis S, De P, Leyland-Jones B. Gain and amplification of RAC1 GTP-ase in BC: Explaining alterations in patients by experiments using TNBC model [abstract]. In: Proceedings of the 2016 San Antonio Breast Cancer Symposium; 2016 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2017;77(4 Suppl):Abstract nr P6-08-07.

Purpose Identification of single-gene biomarkers that are prognostic of outcome can shed new insights on the molecular mechanisms that drive breast cancer and other cancers.Methods Exploratory analysis of 20,464 single-gene messenger RNAs (mRNAs) in the Molecular Taxonomy of Breast Cancer International Consortium (METABRIC) discovery cohort indicates that low expression of FGD3 mRNA is prognostic for poor outcome. Prognostic significance of faciogenital dysplasia 3 (FGD3), SUSD3, and other single-gene proliferation markers was evaluated in breast cancer and The Cancer Genome Atlas (TCGA) cohorts.Results A meta-analysis of Cox regression of FGD3 mRNA as a continuous variable for overall survival of estrogen receptor (ER)-positive samples in METABRIC discovery, METABRIC validation, TCGA breast cancer, and Combination Chemotherapy in Treating Women With Breast Cancer (E2197) cohorts resulted in a combined hazard ratio (HR) of 0.69 (95% CI, 0.63 to 0.75), indicating better outcome with high expression. In the ER-negative samples, the combined meta-analysis HR was 0.72 (95% CI, 0.63 to 0.82), suggesting that FGD3 is prognostic regardless of ER status. The potential of FGD3 as a biomarker for freedom from recurrence was evaluated in the Breast International Group1-98(BIG1-98; Letrozole or Tamoxifen in Treating Postmenopausal Women With Breast Cancer) study (HR, 0.85; 95% CI, 0.76 to 0.93) for breast cancer-free interval. In the Hungarian Academy of Science (HAS) breast cancer cohort, splitting on the median had an HR of 0.49 (95% CI, 0.42 to 0.58) for recurrence-free survival. A comparison of the Stouffer P value in five ER-positive cohorts showed that FGD3 (P = 3.8(E-14)) outperformed MKI67 (P = 1.06(E-8)) and AURKA (P = 2.61(E-5)). A comparison of the Stouffer P value in four ER-negative cohorts showed that FGD3 (P = 3.88(E-5)) outperformed MKI67 (P = .477) and AURKA (P = .820).Conclusion FGD3 was previously shown to inhibit cell migration. FGD3 mRNA is regulated by ESR1 and is associated with favorable outcome in six distinct breast cancer cohorts and four TCGA cancer cohorts. This suggests that FGD3 is an important clinical biomarker. (C) 2017 by American Society of Clinical Oncology

Abstract Background: EarlyR is a prognostic gene signature score in ER+ breast cancer (BC) computed from the expression values of ESPL1, SPAG5, MKI67, PLK1 and PGR using a novel algorithm. EarlyR has been validated in multiple cohorts profiled on Affymetrix and Illumina microarrays. This study sought to verify prognostic features of EarlyR in a cohort of BIG 1-98. Patients and Methods: Illumina DASL assay was used to measure gene expression in FFPE tissue of primary BC from a case-cohort sampling subset of postmenopausal women in BIG 1-98 treated with adjuvant endocrine therapy (letrozole or tamoxifen). Chemotherapy treatment was at the discretion of individual physicians and patients. Among the 1218 patients centrally reviewed with sufficient RNA material for the DASL assay, 1174 with ER+ status and assessed for EarlyR are included in the analytic cohort. EarlyR scores and pre-specified risk strata (≤25=low, 26-75=intermediate, >75=high) were computed, while blinded to clinical data. The analysis endpoints included distant recurrence free interval (DRFI) defined as time from randomization to BC recurrence at a distant site within 8 years and BC free-interval (BCFI) defined as time from randomization to first invasive BC recurrence at a local, regional or distant site or invasive contralateral BC within 8 years. Weighted proportional hazards models (univariate and multivariate, stratified by treatment assignment) were used to adjust for Kaplan-Meier, hazard ratio estimates and Wald test statistics to obtain unbiased analyses and to give consistent estimates. Results: The distribution of the EarlyR risk groups was 67% low, 19% intermediate and 14% high risk in this ER+ cohort. EarlyR was prognostic for 8-year DRFI (P-trend=0.008). Patients with high EarlyR risk score (>75) had significantly increased risk of distant recurrence within 8 years (univariate HR=1.73, 95%CI: 1.14-2.64) compared to low EarlyR risk group (≤25). The estimated 8-year DRFI (95%CI) is 84%(80%-88%) for high risk vs. 91%( 89%-92%) for low risk, corresponding to an absolute DRFI risk reduction of 7% (low vs high). EarlyR is also prognostic of 8-year BCFI in ER+ (P-trend=0.002) with the estimated 8-year BCFI (95%CI) 79%(75%-84%) for high risk vs. 88%( 86%-89%) for low risk. Consistent results were observed in ER+, HER2- (P-trend=0.01 for DRFI, P-trend=0.004 for BCFI), in ER+, LN- (P-trend=0.05 for DRFI, P-trend=0.03 for BCFI) and ER+, LN+ (P-trend=0.08 for DRFI, P-trend=0.03 for BCFI) subsets. Conclusions: This study confirmed the prognostic significance of EarlyR using FFPE tissue from the BIG 1-98 trial. In analyses of all ER+ patients and subsets LN-, LN+ and HER2-, EarlyR classifies 65%-70% of patients as low risk, 11-16% as high risk, and < 20% as intermediate risk. In these subsets, the size of the low risk group is larger and the size of the intermediate risk group is smaller than those reported for commercially available signatures. EarlyR identifies a set of high-risk patients with relatively poor prognosis who may be considered for additional treatment. The clinical utility of EarlyR requires further study. Citation Format: Buechler S, Gray KP, Gökmen-Polar Y, Willis S, Thürlimann B, Kammler R, Leyland-Jones B, Badve SS, Regan MM. Independent validation of EarlyR gene signature in BIG 1-98: A randomized, double-blind, phase III trial comparing letrozole and tamoxifen as adjuvant endocrine therapy for postmenopausal women with hormone receptor-positive, early breast cancer [abstract]. In: Proceedings of the 2016 San Antonio Breast Cancer Symposium; 2016 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2017;77(4 Suppl):Abstract nr P4-12-01.

Abstract Background: Prognostic factors are capable of providing information on clinical outcomes at the time of diagnosis; they are usually indicators of growth, invasion, and metastatic potential (Gasparini G. et al. 1993; Hayes DF. et al. 1998). Tissue marker is one of the prognostic factors; to date, only a small proportion of markers are ultimately clinically useful, including Ki-67 and HER2. Faciogenital dysplasia 3 protein (FGD3), a putative regulator of cell morphology and motility, has been shown to regulate cell migration. In a study of 3,256 tumors, low expression of FGD3 mRNA indicates poor outcome (Hayakawa M. et al. 2008; Scooter W. et al. 2014). However, an immunohistochemistry (IHC) study to evaluate FGD3 protein expression has not been done. We hypothesize that the expression levels of FGD3 protein by IHC might improve the prediction of patient outcomes, and FGD3 might be a potentially prognostic marker of breast cancer. Materials and Methods: 322 cases of breast cancer Tissue Microarrays (TMA) (BR1504a, BR1505b, HBre-Duc068Bch-01, and BR20837) were purchased from US Biomax, Inc (Rockville, MD). Rabbit polyclonal antibody against FGD3 was purchased from Novus Biologicals, LLC (Littleton, CO). Tissue cores were stained for FGD3 with Dako’s EnVision + Dual Link System. Image acquisition was performed using an Olympus camera and software. FGD3 protein expression by IHC was quantitatively determined in the range of 0-4. Unpaired t-test was used for data analyses. Results: 1) Benign tumor and breast adenocarcinomas in lower stages showed strong expression of FGD3, whereas the breast adenocarcinomas in higher stages showed mild ~ weak expression. 2) Invasive breast cancer in stage IIA showed strong FGD3 expression compared to matched metastatic carcinoma which showed mild expression of FGD3. 3) FGD3 protein levels for tumors (n=135) with N0 indicating no regional lymph node metastasis were compared with tumors with lymph node metastasis (N1-3; n=98) and corresponding matched metastatic tissue (n=103). An unpaired t-test comparing N0 vs. N1-3 showed lymph node metastasis is associated with lower FGD3 protein levels (p<0.0001). Conclusion: FGD3 protein expression levels within breast tumors were different according to metastatic status. Our IHC results suggest the possibility of FGD3 to be a prognostic marker in patients with breast cancer. Citation Format: Yuliang Sun, Scooter Willis, Xiaoqian Lin, Justin Achua, Casey Williams, Brian Leyland-Jones. Is FGD3 a potentially prognostic marker for breast cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 3926. doi:10.1158/1538-7445.AM2017-3926