Abstract

Background

The 100,000 Genomes Project is a Government funded initiative to create the infrastructure for genomic medicine in the NHS. It aims to sequence 100,000 genomes on patients with rare disease or cancer. For patients with cancer, tumour samples are required for sequencing together with germline DNA extracted from blood. Traditionally tissue samples are placed in formalin fixative to maintain morphology for histological analysis, however, this damages DNA and makes the interpretation of whole-genome sequences (WGS) unreliable. This study analysed the quality of DNA and WGS generated from tissues using an alternative fixative and assessed the feasibility of introducing this into routine practice.

Method

Biopsy-sized tissue samples were taken from surgical excision specimens from patients who had consented for use of their tissue in research. The tissue biopsies were placed in formalin (FFPE) or the alternative fixative PAXgene^®, and matched samples were snap-frozen in liquid nitrogen (FF). Tissues were processed; H&E sections assessed for morphology, and DNA was extracted for WGS on the Illumina platform.

Results

Morphology was compared between samples FF or fixed in formalin or PAXgene^®: the two fixatives gave comparable morphology which was superior to the FF tissue. DNA yields were similar for all sample conditions, but pre-sequencing QC metrics were poorer for FFPE samples. WGS showed good coverage uniformity in the PAXgene^® samples, comparable to the FF samples, whilst coverage was uneven in the FFPE samples with many unique variants introduced due to the fixation.

Conclusion

PAXgene^® fixed tumour samples give WGS of comparable quality to FF tissue without the generation of numerous unique variants which make FFPE samples unsuitable for reliable WGS. The ease of use of PAXgene^® compared to FF tissue makes it more suitable for wider-scale implementation in the clinical setting.