The prognosis of ovarian cancer (OC) remains poor. Thus, the present study aims to identify independent prognostic factor in OC patients. OC gene expression studies GSE26712 and TCGA-OV were included in this study. Prognosis-associated differentially expressed genes (DEGs) between normal ovarian tissue and OC were identified. LASSO Cox proportional hazards regression model was conducted and a prognostic signature was constructed based on these DEGs. The predictive ability of the signature was analyzed in the training set and test set. The prognosis performance of the signature was compared with CA-125 and HE4. Gene set enrichment analysis (GSEA) was conducted to identify relevant mechanism. 332 DEGs were identified, out of which 64 DEGs were significantly correlated with the overall survival (OS) of OC patients, and 5 DEGs (IGF2, PEG3, DCN, LYPD1 and RARRES1) were applied to build a 5-gene signature. Patients in the 5-gene signature low-risk group had significantly better OS compared to those in the 5-gene high-risk group (p=0.0004) in the training set. Similar results were found in the test set, and the signature was also an independent prognostic factor. The prognosis performance of the 5-gene signature was significantly better than that of CA-125 and HE4. GSEA suggested that OC samples in the 5-gene high-risk group were significantly enriched in WNT/β-catenin signaling and epithelialmesenchymal transition. We developed and validated a 5-gene signature that might be used as an independent prognostic factor in patients with OS.

Background: Triple negative breast cancer (TNBC) is a heterogeneous disease with several molecular subtypes: basal-like1 (BL-1), basal-like 2 (BL-2), mesenchymal (M), and luminal androgen receptor (LAR). Molecular evolution of TNBC through chemotherapy selection pressure is well recognized but poorly understood. In addition, approximately 20% of TNBCs respond to PD-1 or PD-L1 inhibitors. It has been observed that heavily pre-treated patients may not respond well to immunotherapy. This study was carried out to perform immune profiling of paired primary and recurrent TNBC. Here we report the result of the first 10 paired tissue pilot analysis. Methods: Twenty specimens were identified through an IRB-approved protocol via the City of Hope Biospecimen Repository (2002-2015). Two brain and one bone metastasis specimens were not included due to technical difficulty. Formalin-fixed paraffin embedded (FFPE) sample blocks were cut into 5-mm thick slides and labeled with the following antibodies: CD4, CD8, CD3, FOXP3, CD20, CD33, Pan-CK, and PD-1 using the multiplex IHC opal method. Image acquisition and cell counting were carried out using PerkinElmer Vectra automated quantitative pathology imaging system and inForm software analysis (PerkinElmer, Waltham, MA). mRNA expression profiling was performed using Affymetrix Human Genechip 2.0. Raw data were normalized and processed using Expression Console. Using Vanderbilt TNBC sub-classification tool, we have sub-classified the 20 primary and recurrent TNBC specimens. Tumor mutation burden (TMB) was generated through FoundationOne® platform. Result: A total of 17 samples were analyzed (M, 5; LAR, 3; BL-1, 4; BL-2, 5). M-subtype had a significantly lower tumor-infiltrating CD3+ T cells (p=0.005), CD8+ T cells (p=0.024), CD4+ T cells (p=0.065) and CD4+FOXP3+ Treg cells (p=0.054), irrespective of the site of metastasis. CD20+ B cells were particularly enriched in BL-1 subtype (p=0.0013, 23.5% of 17 samples). Of 17 samples, 8 had TMB. Seven had low TMB (<10 mut/Mb) and one had intermediate TMB (11 mut/Mb, LAR subtype). The tumor with intermediate TMB had the highest quantity of tumor-infiltrating CD3+ T cells, CD8+ T cells, CD8+PD1+ T cells, and CD4+FOXP3+PD1+ Treg cells compared to the 7 tumors with low TMB. Compared with recurrent tumors, primary tumors had a significantly higher percentage of tumor-infiltrating T cells (TIL). To validate multiplexed IHC results, these samples were evaluated by a licensed pathologist at City of Hope using the International TILs Working Group 2014 guidelines, and there was a good correlation between percent of TILs and CD3+ T cells by IHC approach. Conclusion: To our knowledge, this is the first study linking tumor immune cell profiles with the TNBC 4 subtypes. Distinctive immune cell patterns were observed among 4 TNBC subtypes. M subtype had significantly lower TILs, which may indicate poor response to checkpoint inhibitors. Further analysis of a total of 50 paired TNBCs is currently underway. Contact information: Yuan Yuan MD PhD, Email: yuyuan@coh.org Citation Format: He T-F, Yost S, Schmolze D, Wang R, Rosario A, Tu T, Chu P, Lee P, Yuan Y. Immune profiling of paired primary and recurrent triple negative 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 P3-05-02.