Associated research

Plain language summary of: The ATOM-Seq sequence capture panel can accurately predict microsatellite instability status in formalin-fixed tumour samples, alongside routine gene mutation testing. Kanishta Srihar, Arief Gusnanto, Susan D. Richman, Nicholas P. West, Leanne Galvin, Daniel Bottomley, Gemma Hemmings, Amy Glover, Subaashini Natarajan, Rebecca Miller, Sameira Arif, Hannah Rossington, Thomas L. Dunwell, Simon C. Dailey, Gracielle Fontarum, Agnes George, Winnie Wu, Phil Quirke & Henry M. Wood 

Background:

Microsatellite Instability (MSI) is when cells have a high number of mutations. It happens in a number of different cancers and is linked to different outcomes and responses to chemotherapy and immunotherapy when it is compared to microsatellite stable cancers.

It is recommended that MSI testing is done as a part of standard care for patients with certain types of cancers including bowel cancer (where MSI is seen in approximately 12-17% of tumours) to help guide the type of treatment offered to the patient.

What did we do?

This study is part of a larger study called the GeneFirst ATOM-Seq capture panel which is a new approach to testing for MSI. This study aimed to evaluate the ATOM-Seq technology and how effective it is in establishing the MSI status.

What did we find?

The GeneFirst ATOM-Seq technology is ideally suited to test for MSI, and performs as good as, or better than other methods. It can easily be incorporated into a research or diagnostic setting.

A plain language summary of: Neo-adjuvant FOLFOX with and without panitumumab for patients with KRAS-wt locally advanced colon cancer: results following an extended biomarker panel on the FOxTROT Trial embedded phase II population. J.F. Seligmann, D. Morton, F. Elliott, K. Handley, R. Gray, M. Seymour, B. Glimelius, L. Magill, C.J.M. Williams, P. Quirke, D. Bottomley, H.M. Wood,K. Murakami, A.D. Beggs,N.P. West, FOxTROT Collaborative Group

Background: FOxTROT is a clinical trial that is looking at whether treating bowel cancer patients with chemotherapy before and after surgery can help stop the bowel cancer from coming back. A smaller subsection of the trial looked at a targeted cancer treatment called panitumumab as well as chemotherapy and was for people who didn’t have a change in a gene called RAS. The first stage of the trial ran between 2008 and 2016. The results showed that treating patients with chemotherapy before surgery stopped 1 in 5 cancers from coming back and there was still a benefit 5 years after treatment.

Before being treated for bowel cancer, biomarker testing is done. This is a test done in a laboratory that examines a sample of blood, tissue or other bodily fluid to detect specific molecules and amongst other uses, can be used to help plan a patient’s treatment.

Biomarker testing directly influences what treatment is used for patients with certain types of tumours called microsatellite instability-high, RAS and BRAF wild-type and targeted treatments are available for patients with BRAF-mutant, KRAS mutant and HER2 amplifies tumours.

This study aimed to test whether adding the targeting therapy drug panitumumab to chemotherapy before surgery in RAS/BRAF wild-type patients and patients with specific biomarkers (biomarker hyperselection) reduced their risk of their bowel cancer returning.

What did we find? In total 269 patients were enrolled into the trial. In RAS/BRAF wild-type tumours, treating patients with panitumumab and chemotherapy before surgery showed a trend towards  a reduction in cancers returning. The effect of using panitumumab was more obvious in biomarker hyperselected patients.

Plain language summary of: Associations between AI-Assisted Tumor Amphiregulin and Epiregulin IHC and Outcomes from Anti-EGFR Therapy in the Routine Management of Metastatic Colorectal Cancer. Christopher J.M. Williams; Faye Elliott; Nancy Sapanara; Faranak Aghaei; Liping Zhang; Andrea Muranyi; Dongyao Yan; Isaac Bai; Zuo Zhao; Michael Shires; Henry M. Wood; Susan D. Richman; Gemma Hemmings; Michael Hale; Daniel Bottomley; Leanne Galvin; Caroline Cartlidge; Sarah Dance; Chris M. Bacon; Laura Mansfield; Kathe Young-Zvandasara; Ajay Sudan; Katy Lambert; Irena Bibby; Sarah E. Coupland; Amir Montazeri; Natalie Kipling; Kathryn Hughes; Simon S. Cross; Alice Dewdney; Leanne Pheasey ; Cathryn Leng ; Tatenda Gochera; D. Chas Mangham; Mark Saunders; Martin Pritchard; Helen Stott; Abhik Mukherjee; Mohammad Ilyas; Rafael Silverman; Georgina Hyland; Declan Sculthorpe; Kirsty Thornton; Imogen Gould; Ann O'Callaghan; Nicholas Brown; Samantha Turnbull; Lisa Shaw; Matthew T. Seymour; Nicholas P. West; Jenny F. Seligmann; Shalini Singh; Kandavel Shanmugam; Philip Quirke 

Background:

Advanced bowel cancer, also referred to as stage 4 bowel cancer, metastatic bowel cancer or secondary bowel cancer, is cancer that has spread from the bowel to other organs in the body.

Advanced bowel cancers cannot be cured however doctors can sometimes offer a combination of therapies (such as chemotherapy) to increase a patient’s survival time. Not all therapies are appropriate for all patients so cancer doctors must decide which treatment they think will be most effective for individual patients.

One way of deciding upon the treatment that is most suitable for a patient is to look at levels of particular proteins within a tumour. When tumours are removed by surgeons, pieces of them are fixed in formalin (a type of solution made up of water and gas) and then put in paraffin blocks. This allows researchers to analyse the tumour samples using a technique called immunohistochemistry to measure the levels of proteins in a tumour sample.

Studies over the past few years have suggested that tumours with high levels of two proteins called amphiregulin (AREG) and epiregulin (EREG) respond well to chemotherapy agents called panitumumab and cetuximab.

What did we do?

Researchers and doctors from eight cancer centres across the UK recruited patients who consented to having their tumour samples analysed for levels of AREG and EREG proteins. The researchers worked, in collaboration with Roche Diagnostics, to develop a computer-based method of quantifying AREG and EREG, known as artificial intelligence (AI).

What did we find?

The results of the study showed that patients, who had high levels of the AREG/EREG proteins, and who were treated with panitumumab or cetuximab, lived longer before their cancer progressed, and also survived longer overall, than patients who had low levels of AREG/EREG proteins.

What will we do next?

The researchers involved in this study are now planning to run a clinical trial, with the aim of providing sufficient data to warrant the introduction of measuring AREG and EREG levels in routine clinical practice.

Patients with high levels of AREG/EREG will subsequently be offered panitumumab and cetuximab, either on their own, or combined with other chemotherapy agents.

A plain language summary of: Generalized biomarker prediction from cancer pathology slides with self-supervised deep learning: A retrospective multi-centric study. Jan Moritz Niehues, Philip Quirke, Nicholas P. West, Heike I. Grabsch, Marko van Treeck, Yoni Schirris, Gregory P. Veldhuizen, Gordon G.A. Hutchins, Susan D. Richman, Sebastian Foersch, Titus J. Brinker, Junya Fukuoka, Andrey Bychkov, Wataru Uegami, Daniel Truhn, Hermann Brenner, Alexander Brobeil, Michael Hoffmeister, Jakob Nikolas Kather

Background:

Deep Learning is a type of Artificial Intelligence in which computers are taught to process large volumes of data, often in a superior way to humans. Deep Learning models are able to recognise complicated patterns in pictures, text, sounds and other data to provide insights and predictions.

In pathology, Deep Learning can be used to predict common DNA changes in cancers from routinely collected histopathology tissue slides. In bowel cancer, these include Microsatellite Instability (MSI) and KRAS/NRAS/BRAF mutation status . These insights enable doctors to diagnose hereditary cancer syndromes and make decisions on a patient’s treatment.

This paper examines whether Deep Learning can be used to accurately predict these DNA changes in patients with bowel cancer.

What did we do?

Using tissue slides from patients with bowel cancer, we compared six different Deep Learning models to predict key biomarkers required in the treatment pathway including MSI and mutations in BRAF, KRAS, NRAS and PIK3CA.

What did we find?

Our study found that Deep Learning can be used to accurately predict MSI and BRAF mutations. Current standard of care testing is expensive and time consuming. A move to Deep Learning for these biomarkers would save both time and money. However, even with powerful Deep Learning models, using current technology it is not possible to predict the mutation status of KRAS, NRAS and PIK3CA accurately enough to change testing pathways for these markers.

Plain language summary of: Evaluation of CD3 and CD8 T-Cell Immunohistochemistry for Prognostication and Prediction of Benefit from Adjuvant Chemotherapy in Early-Stage Colorectal Cancer Within the Quasar Trial. Christopher J.M. Williams, Richard Gray, Robert K. Hills, DPhil, Michael Shires, Liping Zhang, Zuo Zhao, PhD, Tracie Gardner, Nancy Sapanara, Xiao-Meng Xu, Isaac Bai, Dongyao Yan, Andrea Muranyi, Sarah Dance, Faranak Aghaei, Gemma Hemmings, Michael Hale, Uday Kurkure, Christoph Guetter, Susan D. Richman, Gordon Hutchins, Jenny F. Seligmann, Nicholas P. West, Shalini Singh, Kandavel Shanmugam and Philip Quirke

Background:

Around 80% of patients with stage 2 bowel cancer and 50% of patients with stage 3 bowel cancer are cured by bowel cancer surgery. Giving patients chemotherapy after their surgery, known as adjuvant chemotherapy, improves the chance of stage three bowel cancer patients surviving for five years free from cancer. However, it is not known how much of a benefit it offers to patients with stage 2 bowel cancer.

Indicators are used by doctors to decide whether a patient with stage 2 bowel cancer will benefit from adjuvant chemotherapy, however, they are only moderately accurate. Therefore, newer more accurate indicators of whether chemotherapy will be beneficial to this group of patients is needed.

The immune system plays a large role in controlling the spread of cancer. Certain immune cells, when present in bowel cancer tumours, are known to improve the outcomes of patients, particularly CD3 and CD8.

The aim of the study is to determine whether CD3 and CD8 cells in bowel cancer tumours can predict how beneficial chemotherapy will be to patients.

What did we do?

We analysed the tissue taken from 868  bowel cancer tumours for CD3 and CD8 immune cells and we measured the density of the CD3 and CD8 cells.

What did we find?

We found that examining the CD3 and CD8 immune cells in bowel cancer tumours can provide a strong indication of the risk of a patient’s cancer returning, known as cancer recurrence.

Plain language summary of: Automated curation of large-scale cancer histopathology image datasets using deep learning. Lars Hilgers, Narmin Ghaffari Laleh, Nicholas P West, Alice Westwood, Katherine J Hewitt, Philip Quirke, Heike I Grabsch, Zunamys I Carrero, Emylou Matthaei, Chiara M L Loeffler, Titus J Brinker, Tanwei Yuan, Hermann Brenner, Alexander Brobeil, Michael Hoffmeister, Jakob Nikolas Kather

Background:

Histopathology is the diagnosis and study of diseases and tissues. It involves examining tissue and cells under a microscope. Digital pathology includes the creation of  digital slides. The images on glass slides are captured with a scanning device to provide a high-resolution image. The image can then be viewed on a computer screen.

Over the past years, there has been an increase in the number of digital histopathology image data that are available. There has also been a rapid growth in the use of Artificial Intelligence (AI) to analyse digital histopathology in cancer. AI has a lot of uses in pathology to help with the diagnosis and prognosis of cancer, however there are some limitations to its use. For example, in bowel cancer surgery, over 25 tissue slides can be generated from one specimen taken during surgery. However, AI models are generally trained to assume one tissue slide represents the entire patient case.

Manually sorting and labeling images is a very time-consuming process. It can hinder the use of AI from big studies on tissue samples. In this study, we addressed this issue by developing a deep learning (DL)-based method for the automatic curation of large pathology datasets using several slides per patient.

What did we find?

This study found that using a low-resolution thumbnail image is sufficient to classify the type of slide in digital pathology accurately. This can help researchers to make the large resource of existing pathology archives accessible to modern AI models with only minimal manual work.

Plain language summary of: Preoperative Chemotherapy for Operable Colon Cancer: Mature Results of an International Randomized Controlled Trial. Dion Morton, Matthew Seymour, Laura Magill,Kelly Handley, James Glasbey, Bengt Glimelius, Andy Palmer, Jenny Seligmann, Søren Laurberg, Keigo Murakami, Nick West, Philip Quirke, and Richard Gray, on behalf of the FOxTROT Collaborative Group 

Background:

The standard treatment given to bowel cancer patients, who are considered to have moderate to high-risk cancer, is surgery, followed by chemotherapy. This is referred to as adjuvant chemotherapy, meaning chemotherapy that is given after the primary treatment (which is surgery). Although these patients receive adjuvant chemotherapy, 20-30% of patients’ cancers come back (also known as cancer recurrence) and those cancers are usually incurable.

Giving chemotherapy before surgery, known as neoadjuvant or preoperative chemotherapy, has improved outcomes for patients in other gastrointestinal cancers, and it also has the potential to also improve outcomes for bowel cancer patients.

This study aims to provide an examination of whether neoadjuvant chemotherapy has advantages over postoperative chemotherapy for patients with bowel cancer.

What did we do?

Patients who were part of the trial were randomly allocated to two groups. Group 1 underwent six weeks of neoadjuvant chemotherapy plus 18 weeks of postoperative chemotherapy. Group 2 underwent 24 weeks of postoperative chemotherapy.

What did we find?

Six weeks of chemotherapy before surgery can be given safely to patients without increasing the risk of complications following surgery. It also provided significant tumour regression (meaning the tumour decreases in size) which is a strong predictor for a lower risk of the cancer returning.

Conclusion:

Six weeks of neoadjuvant chemotherapy should be considered as a treatment option for patients with stage 3 and 4 bowel cancer.

Plain language summary of: The genomic landscape of 2,023 colorectal cancers. Alex J. Cornish, Andreas J. Gruber, Ben Kinnersley, Daniel Chubb, Anna Frangou, Giulio Caravagna, Boris Noyvert, Eszter Lakatos, Henry M. Wood, Steve Thorn, Richard Culliford, Claudia Arnedo-Pac, Jacob Househam, William Cross, Amit Sud, Philip Law, Maire Ni Leathlobhair, Aliah Hawari, Connor Woolley, Kitty Sherwood, Nathalie Feeley, Güler Gül, Juan Fernandez-Tajes, Luis Zapata, Ludmil B. Alexandrov, Nirupa Murugaesu, Alona Sosinsky, Jonathan Mitchell, Nuria Lopez-Bigas, Philip Quirke, David N. Church, Ian P. M. Tomlinson, Andrea Sottoriva, Trevor A. Graham, David C. Wedge & Richard S. Houlston 

Background:

Bowel cancer is the third most common cancer worldwide and is a common cause of cancer related death. Despite this, a detailed description of its genomic landscape does not exist. The genomic landscape refers to the pattern of gene mutations (changes) across the genome (the complete set of genetic information) in a specific cancer type.

What did we do?

We performed whole-genome sequencing of 2,023 bowel cancer samples from participants in the 100,000 Genomes Whole-gene sequencing is a method used in a laboratory and carried out by pathologists to determine the entire genetic makeup of a specific organism or cell type.

The 100,000 Genomes Project is a British project to sequence and study the role our genes play in health and disease. Launched in 2012, the project aimed to bring the benefits of personalised medicine to the NHS.

What did we find?

The analyses provided a highly detailed genomic analysis of bowel cancer. The findings included:

• Over 250 genes that are linked to cancer progression (cancer that becomes worse or spreads in the body), which have not previously been linked to bowel cancer or other cancers.
• Bowel cancer can be categorised into four subgroups based on genomic features, each associated with different likely outcomes.
• Rare molecular subgroups are identified.
• The study found that different parts of the colorectum (the large intestine and rectum) have distinct patterns of genetic changes. This suggests that there may be different causes for cancer in these areas. The bacteria Escherichia coli (also known as E-Coli) and a specific type of genetic change (called the SBS93 signature) are linked to these patterns. This indicates that lifestyle factors, such as what we eat and whether we smoke, might increase the risk of developing bowel cancer.
• Many bowel cancers show changes that help them evade the immune system, especially in certain types called hypermutant tumors.

Finding specific mutations that can be targeted for treatment—like those in less common groups such as BRCA1 and IDH1—shows how whole-genome sequencing can improve personalized treatment for patients.

Plain language summary of: Combined endoscopic and laparoscopic surgery (CELS) for early colon cancer in high-risk patients. Morten F. S. Hartwig, Mustafa Bulut, Jens Ravn-Eriksen, Lasse B. Hansen, Rasmus D. Bojesen, Mads Falk Klein, Henrik L. Jakobsen, Morten Rasmussen, Bo Rud, Jens-Ole Eriksen, Susanne Eiholm, Anne-Marie K. Fiehn, Phil Quirke & Ismail Gögenur 

 Background:

Bowel cancer is the third most common cancer. It causes almost 10% of all cancer-related deaths across the world. Bowel cancers are cured by having an operation to remove the part of the bowel that is affected by cancer: this is known as a surgical resection.

Having a surgical resection can cause complications and an increased risk of death in people who are frail or who are elderly.

This paper describes a new method of operation called Combined Endoscopic and Laparoscopic Surgery (CELS). This method allows surgeons to perform an operation endoscopically (through inserting a tube with a small camera on the end into your body through a natural opening such as the anus) and laparoscopically (also know as key hole surgery, where the surgery is performed through small incisions in the body).

At present, the use of CELS is aimed at patients who are at an increased risk of death from standard surgical operations.

The aim of this paper was to examine the feasibility of performing Combined Endoscopic and Laparoscopic surgery on patients with early stage bowel cancer, who are considered high risk of poor outcomes from surgery.

Results:

This study found that the use of CELS was well tolerated and effective.

Conclusion:

CELS could be considered by surgeons as an additional surgery method to cure people of bowel cancer.

A plain language summary of  Tumour-infiltrating Lymphocytes, Tumour cell Density and Response to Neoadjuvant Short-Course Radiotherapy in Rectal Cancer: A translational sub-study from the MRC CR07 clinical trial. Jonathan P. Callaghan, Ross Jarrett, Alice C. Westwood, Jon Laye, Philip Quirke, Derek R. Magee,  Daniel Bottomley, David Sebag-Montefiore, Lindsay Thompson,  Angela Meade, Heike I. Grabsch and Nicholas P. West

Background:

Rectal cancer is often treated with radiation before surgery to lower the chance of the cancer coming back. But not all patients respond to radiation in the same way, and doctors don’t yet have a reliable way to predict who will benefit most.

This study looked at two possible measures: the number of tumour cells (TCD) and the number of immune cells called lymphocytes (TILs) in tissue samples taken before treatment.

What did we do?

Researchers looked at tissue samples from people with rectal cancer to see if the number of tumour cells and immune cells could help predict how well a patient would respond to radiation treatment.

The tumour cells were counted manually and Artificial Intelligence was used to measure immune cells (specifically lymphocytes) in the tissue.

What did we find?

We found that radiation reduced both tumour and immune cells. Importantly, patients’ chances of survival were linked to how many of these cells were present before treatment.

Conclusion:

The study suggests that measuring tumour and immune cell levels could help doctors better predict outcomes and find the most suitable treatment plans. The number of cancer cells found in the tissue sample may be useful both for predicting how well radiation will work and for estimating long-term survival chances. The number of immune cells found in the tissue samples seems to be more useful for determining a patient’s likely outcome than for predicting how a patient responds to treatment.