- Professor Dame Sue Hill, Chief Scientific Officer, NHS England
- Dr Rachel Butler, Consultant Clinical Scientist, University Hospital of Wales
- Professor Ian Cree, Consultant Pathologist, University Hospitals Coventry and Warwickshire NHS Trust
- Dr Clare Verrill, Consultant in Cellular Pathology, Oxford University Hospitals NHS Foundation Trust
We will now look at the particular requirements for histopathologists when processing solid tumour samples for DNA extraction for whole genome sequencing. It is vitally important to follow all current NHS England-approved protocols, as the work done at the histopathology stage has implications for the quality of DNA that can be extracted from the tumour sample, and therefore on the success of whole genome sequencing. These processes are being optimised as new information emerges, so you should always refer to the most up-to-date protocols.
Prof Dame Sue Hill: For cancer patients, the skill of the histopathology team in selecting and preparing tissue is vital. Selecting tissue with a high neoplastic content should ensure a high yield of DNA with sufficient quality to lead to successful whole genome sequencing. This will enable a good quality comparison between the tumour and the patient’s normal genome from the blood DNA sample, so allowing any genetic changes associated with the cancer to be identified.
For the first stage of processing, the tumour sample must be collected from surgery and taken to the histopathology laboratory within two hours of excision. Alternatively, if this is not possible, the sample could be vacuum packed, kept on ice or refrigerated to extend the time between excision and laboratory handling. This is particularly relevant for tumour samples that are taken by the local delivery partners, which then need to be transported to the GMC’s histopathology department. For the 100,000 Genomes Project, fresh frozen samples should be collected where biologically possible, as DNA extracted from fresh frozen tumour consistently gives better whole genome sequencing results compared to FFPE samples. However, it is recognised that some specimens will not be suitable for fresh frozen sampling. For example, if the tumour is small, then sampling at this stage will compromise the routine diagnostic reporting.
Dr Clare Verrill: To ensure adequate fixation of samples for DNA sequencing, there are several things you need to think about. So firstly you need to ensure you’ve got robust arrangements within your institution to get samples from theatre to the pathology laboratory as this enables rapid action by the pathologist. So at this stage, they’ll take their fresh frozen and their optimised formalin-fixed paraffin-embedded samples, and the FFPE samples will go into the genomic block. If appropriate, the pathologist at this stage will also slice, or open the specimen, and this is particularly important to get the formalin right to the centre of the specimen, as otherwise it only penetrates about 1 mm an hour. You need to think about things such as the size of the pot you’re putting the specimen into and you need to ensure you have an adequate volume of neutral buffered formalin, and convention suggests about 10 times the volume of formalin to the volume of the specimen. I mean, the key message is, if you don’t fix the specimen adequately, the DNA will degrade due to autolysis. You really only have one chance to get it right.
Specimen cut-up should take place in a clean environment. For each tumour sample use a clean cut-up board, either disposable cardboard or plastic, use sterile disposable blades and clean forceps. The specimen should be inked, measured and weighed and sliced according to the Royal College of Pathologist’s guidelines for the specific tumour type being processed.
On slicing, the location of the tumour should be determined and the histopathologist should make a decision whether the selection of a fresh frozen sample is possible without compromising histological diagnosis and staging. Fresh frozen tissue can be selected using a punch biopsy or scalpel. Where possible, more than one sample should be selected. The sample should be taken from an area of minimal necrosis and inflammation and high neoplastic content, as a minimum of 40% tumour DNA is required for the whole genome sequencing.
The site of the selection of the block should be noted. Different blocks can be indicated on a photograph or by using different coloured inks. After this, the selected sample must be rapidly snap frozen in liquid nitrogen or isopentane and stored in a labelled CryoTube at -80° Celsius. Where possible at the time of cut-up, a second sample adjacent to the fresh frozen sample should be harvested, either by punch or in a slice, and placed immediately into 10% buffered formalin at pH 7. The selected tissue should be fixed for a maximum of 24 hours and this will form the genomic block for future whole genome sequencing. The genomic block should be labelled with the unique laboratory accession number. It is recommended that both samples are processed and assessed for cellularity and neoplastic content so that the best sample can be sent for DNA extraction.
The 100,000 Genomes Project cancer pipeline focuses on tumours from a number of different sites on the body including lung, ovarian, breast, colorectal and prostate. This list will change over the course of the project so always check the latest specifications. When handling these tumour samples, histopathologists will need to make adjustments for the characteristics of each tumour type.
Prof Ian Cree: Lung cancers are usually removed as relatively small tumours. It’s particularly important that the resection margins are respected and, if the tumour is incised, to ensure adequate fixation. That should be done through a margin that doesn’t matter. The colorectal cancer requires the bowel to be opened to allow the fixative to penetrate the lumen of the bowel and get to the cancer. The cancer if it’s very large may also need to be incised, but again it’s very important to respect resection margins so that the optimal information can be given to the surgeon and the patient on how close those margins are to the actual cancer.
Formalin-fixed tissue should be processed and embedded avoiding prolonged periods in formalin. Histopathological diagnosis of FFPE samples is made from H&E stained slides cut from paraffin blocks, typically 4 micrometres in thickness. H&E staining is used to identify and estimate the percentage of neoplastic cells. For DNA extraction, the selected marked areas of tumour should contain at least 40% viable tumour and areas with minimal necrosis and inflammation should be selected. Areas of high neoplastic content should be marked by a fully trained and competent practitioner in histological assessment. The estimated tumour content should be recorded by the histopathologist and reported in the cancer data set. Samples from these high neoplastic areas can be obtained for DNA extraction by macrodissection of mounted sections or punch cores.
Dr Clare Verrill: When you’re considering tumour samples to go forward for DNA extraction, obviously the genome sequence that will come out at the end depends on having a high tumour cell content entering the process at the beginning. So we need to think about marking and selecting an area with a high percentage of viable tumour cells to the total number of cells within the marked area. The minimum you would want is 40% tumour cell content and you want to avoid areas of necrosis because they really affect the DNA quality. And also you want to avoid areas of high inflammatory infiltrate, the reason being that inflammatory cells contain DNA as well and you’ll be sequencing large numbers of inflammatory cells rather than tumour cells. And it will really push your ratio of tumour cell content down.
Dr Rachel Butler: The pathologist has to ensure that there is sufficient tumour present. So in the samples that are going to be passed on for DNA extraction, those are the two things they have to be thinking about – is there tumour there and is there enough tumour tissue there for DNA extraction? Because if there isn’t, then there really isn’t any point in passing them on for DNA extraction because we’re not going to get enough DNA out to go on for DNA sequencing.
A digitally image of the slide from which the tumour DNA has been extracted should be captured, ideally showing the selected marked area. If 1 mm punches from the formalin-fixed block are being submitted for DNA extraction, the marked area where these have been taken from should be captured as part of the image. Slides should be digitally scanned in high resolution, ideally at 40 times magnification. If a digital scanner is not available, then JPG or TIFF images at sufficient resolution for histopathological review should be obtained.
Dr Clare Verrill: The impact of the 100,000 Genomes Project is massive for pathology. It’s the single biggest revolution that I will see in pathology in my career. So at the moment, as a pathologist, I can look at a group of tumours down the microscope and they all have the same diagnosis, but I know there are differences, and at the moment I can’t look beyond that. And that’s what we’ll be able to do with 100,000 Genomes Project. We’ll be able to unpick tumours, we’ll be able to better classify them, and for patients that means we’ll be able to potentially identify therapeutic targets. They may be individual therapeutic targets for that patient and their tumour, and also we’ll be able to provide better prognostic information for those patients.