Dataset Information

FFPE breast tumour blocks provide reliable sources of both germline and malignant DNA for investigation of genetic determinants of individual tumour responses to treatment.

ABSTRACT:

Background

Bio-banked formalin-fixed paraffin-embedded (FFPE) tissues provide an excellent opportunity for translational genomic research. Historically matched blood has not always been collected as a source of germline DNA. This project aimed to establish if normal FFPE breast tissue could be used as an alternative to blood.

Methods

Exome sequencing was carried out on matched tumour tissue, normal breast tissue and blood on five patients in the START trial. Retrieved samples had been archived at different centres for at least 13 years. Following tissue macro-dissection and DNA extraction, targeted exome capture was performed using SureSelect Human All Exome v5 reagents (Agilent). Illumina paired-end libraries were prepared from the captured target regions and sequenced on a HiSeq2500 (Illumina) acquiring 2?×?75 bp reads. Somatic variants were called using the MuTect software analysis tool and copy number abnormalities (CNA) were identified using CNVkit. Targeted sequencing and droplet digital PCR were used to validate somatic variants and CNA, respectively.

Results

Overlap of somatic variants and CNA called on tumour versus blood and tumour versus normal breast tissue was good. Agreement in somatic variant calling ranged from 76.9 to 93.6%. Variants with an allele frequency lower than 10% were more difficult to validate irrespective of the type of germline DNA used. Pearson's correlation coefficients for paired comparisons of CNA using blood or normal tissue as reference ranged from 0.70 to 0.94.

Conclusions

There is good correlation between the somatic mutations and CNA called using archived blood or normal breast tissue as germline reference material.

SUBMITTER: Wilkins A

PROVIDER: S-EPMC6022520 | biostudies-literature | 2018 Aug

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

FFPE breast tumour blocks provide reliable sources of both germline and malignant DNA for investigation of genetic determinants of individual tumour responses to treatment.

Wilkins Anna A Chauhan Ritika R Rust Alistair A Pearson Alex A Daley Frances F Manodoro Floriana F Fenwick Kerry K Bliss Judith J Yarnold John J Somaiah Navita N

Breast cancer research and treatment 20180426 3

<h4>Background</h4>Bio-banked formalin-fixed paraffin-embedded (FFPE) tissues provide an excellent opportunity for translational genomic research. Historically matched blood has not always been collected as a source of germline DNA. This project aimed to establish if normal FFPE breast tissue could be used as an alternative to blood.<h4>Methods</h4>Exome sequencing was carried out on matched tumour tissue, normal breast tissue and blood on five patients in the START trial. Retrieved samples had ...[more]

PMID: 29700676

Similar Datasets

Project description:The reliability of differential gene expression analysis on formalin-fixed, paraffin-embedded (FFPE) expression profiles generated using Affymetrix arrays is questionable, due to the high range of percent-present values reported in studies which profiled FFPE samples using this technology. Moreover, the validity of gene-modules derived from external datasets in FFPE microarray expression profiles is unknown. By generating matched gene expression profiles using RNAs derived from fresh-frozen (FF) and FFPE preserved breast tumors with Affymetrix arrays and FF/FFPE RNA specific amplification-and-labeling kits, the reliability of differential expression analysis and the validity of gene modules derived from external datasets were investigated. Specifically, the reliability of differential expression analysis was investigated by developing de-novo ER/HER2 pathway gene-modules from the matched datasets and validating them on external FF/FFPE gene expression datasets using ROC analysis. Spearman's rank correlation coefficient of module scores between matched FFPE/frozen datasets was used to measure the reliability of gene-modules derived from external datasets in FFPE expression profiles. Independent of the array/amplification-kit/sample preservation method used, de-novo ER/HER2 gene-modules derived from all matched datasets showed similar prediction performance in the independent validation (AUC range in FFPE dataset; ER: 0.93-0.95, HER2: 0.85-0.91), except for the de-novo ER/HER2 gene-module derived from the FFPE dataset using the 3'IVT kit (AUC range in FFPE dataset; ER: 0.79-0.81, HER2: 0.78). Among the external gene modules considered, roughly ~50% gene modules showed high concordance between expression profiles derived from matching FF and FFPE RNA. The remaining discordant gene modules between FF and FFPE expression profiles showed high concordance within matching FF datasets and within matching FFPE datasets independently, implying that microarrays still require improved amplification-and-sample-preparation protocols for deriving 100% concordant expression profiles from matching FF and FFPE RNA.

Dataset Information

FFPE breast tumour blocks provide reliable sources of both germline and malignant DNA for investigation of genetic determinants of individual tumour responses to treatment.

Background

Methods

Results

Conclusions

Publications

FFPE breast tumour blocks provide reliable sources of both germline and malignant DNA for investigation of genetic determinants of individual tumour responses to treatment.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets