30P - Targeted mRNA sequencing of small formalin-fixed and paraffin-embedded breast cancer samples for the quantification of immune and cancer-related genes

Date 04 May 2017
Event IMPAKT 2017
Session Welcome reception and Poster Walk
Topics Breast Cancer
Translational Research
Basic Principles in the Management and Treatment (of cancer)
Presenter Bruno Sinn
Authors B.V. Sinn1, S. Loibl2, R. Kronenwett3, J. Furlanetto2, K. Krappmann3, T. Karn4, B. Kerns5, K. Weber2, M. Schmidt6, C. Denkert7
  • 1Department Of Pathology, Charité – Universitätsmedizin Berlin, Berlin/DE
  • 2-, German Breast Group Forschungs GmbH, Neu-Isenburg/DE
  • 3-, Sividion Diagnostics GmbH, Cologne/DE
  • 4Department Of Gynecology And Obstetrics, Johann-Wolfgang Goethe University, 60590 - Frankfurt/DE
  • 5-, HTG Molecular Diagnostics, Inc., Tucson/US
  • 6Department Of Gynecology, Johannes-Gutenberg Universität Mainz, Mainz/DE
  • 7Department Of Pathology, Charité – Universitätsmedizin Berlin, 1007 - Berlin/DE




To evaluate the use of targeted RNA sequencing (RNA-seq) for application to formalin-fixed and paraffin-embedded (FFPE) breast cancer core biopsies from clinical trials.

Patients and Methods

50 FFPE samples of pre-treatment core biopsies from patients of the neoadjuvant GeparQuattro trial were used. For 13 genes, matching quantitative PCR data was available (ESR1, ERBB2, CCL5, CD80, CD8A, CTLA4, CXCL9, CXCL13, FOXP3, IDO1, ANXA1, ARID1A, and PDGFRB). One tissue section (5 µm) for each sample was processed with the HTG EdgeSeq library preparation system using the mRNA Oncology biomarker panel (2,560 genes) and sequenced on an Illumina NextSeq instrument. PCR and RNA-seq data were compared for the 13 genes available on both platforms and a summarized 8-gene immune-score was calculated. We performed additional analyses (unsupervised clustering, differential gene expression analysis) to assess the overall biological signal.


Library preparation was successful in 48 of 50 samples. Of those, 45 were successfully sequenced and passed quality control. Pearson's correlation coefficients for the matched genes ranged from 0.959 to 0.243 with coefficients above 0.6 in 8 of 13 genes. The correlation of the summarized immune score was 0.889. Levels of ESR1 and ERBB2 were highly concordant with the PCR-based classification (p < 0.001) and an exploratory analysis of global differential gene expression according to ESR1-status yielded known targets of ESR1. However, unsupervised clustering of the RNA-seq data resulted in incomplete separation according to estrogen receptor status.


The HTG EdgeSeq system is applicable to small FFPE tissue biopsies with a high rate of success in mRNA sequencing. Overall, there was a good concordance between the sequencing data and quantitative PCR data. The strength of the biological signal, especially its correlation to clinical endpoints, warrants further investigation. The method is a promising approach to targeted mRNA sequencing using small FFPE tissue samples from clinical trial cohorts.

Clinical trial identification