Abstract 630: Detection of copy number variations in breast cancer using targeted sequencing without normal tissue controls

Abstract Cancer exome sequencing including recent studies of TCGA and ICGC consortia revealed the landscapes of somatic mutations and copy number variations (CNVs) in multiple cancers. CNVs represent an important genetic change that may contribute to tumorigenesis, tumor growth and metastatic spread. HER2 amplification in breast cancer represents a paradigm of a molecular target and personalized cancer therapy. As part of the breast cancer routine diagnostics, HER2 status is determined by immunohistochemistry (IHC) and subsequent in situ hybridization (ISH) when necessary. HER2 breast cancer patients are known to benefit from blockage of HER2 signaling, for example by treatment with the therapeutic antibody trastuzumab. A large cohort of frozen breast cancer tissues (184 patients) was investigated by semiconductor sequencing. A customer-designed panel of 154 amplicons was used including regions of mutations and CNVs known to be recurrent from preceding exome-wide studies. After sample-wise normalization, a null distribution of copy numbers was estimated using outlier robust statistics and the sequencing data of all patients and amplicons. Significance of gene amplifications was assessed by comparison with the null distribution leading to cutoff values of 2.77 copies for single hypothesis testing (p < 0.05) and 3.59 copies for multiple hypotheses testing (p < 0.05/154). First, we evaluated the accuracy of targeted sequencing and the new algorithm compared to the gold standard of IHC and ISH. The amplicon panel included three regions of HER2 located in exons 19, 20 and 21. Based on the target region in exon 20 and the single hypothesis cutoff, we obtained a sensitivity of 83% and a specificity of 95% in the detection of HER2 amplifications. Correlations between the copy number estimated from the coverage of exon 19, 20 and 21 where high (Pearson correlation R = 0.92, 0.96 and 0.97). Second, we did an unsupervised search for gene amplifications with the 154 amplicons of the sequencing panel using the multiple testing method. We detected gene amplifications that were abundant in at least 8 patients (>5% of the cohort) in the following 17 genes: CCND1, CDH1, CDKN1B, ERBB2, FGFR1, GRB7, MDM2, MED1, MYC, PAK1, PIK3CA, PIK3R3, RB1, STARD3, TBL1XR1, TSHZ2 and ZNF703. In conclusion, the new algorithm allowed detecting of CNVs from semiconductor sequencing without the need of normal tissue controls. Thus, using targeted sequencing, CNV detection and simple somatic mutation detection can be integrated in an easy and efficient manner. Citation Format: Jan Budczies, Volker Endris, Nikola Bangemann, Thomas Wolf, Jens-Uwe Blohmer, Abrecht Stenzinger, Manfred Dietel, Wilko Weichert, Carsten Denkert. Detection of copy number variations in breast cancer using targeted sequencing without normal tissue controls. [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 630. doi:10.1158/1538-7445.AM2015-630