Image: Dr Maggie Cheang, Group Leader of the Integrative Genomics Analysis in Clinical Trials group at the ICR
Since the 1990s, it’s been known that mutations in specific genes are responsible for certain types of cancer. For example, in 1995, a team of researchers at The Institute of Cancer Research, London, published a paper announcing that they had successfully identified the BRCA2 gene that, when faulty, can increase the risk of developing breast cancer by more than 50 per cent.
Less than 20 years later, the first whole cancer genome was sequenced, opening up the field of cancer genomics. Further research revealed that cancer cells acquire DNA changes over time, meaning that each individual’s cancer can have a unique genetic signature. What’s more, the expression – or lack of expression – of certain genes allows researchers and clinicians to group cancers into subtypes.
Tumour cells’ genes have a significant role in determining how the disease grows, whether it develops resistance to drugs and how it responds to different treatments. For example, in 2022, a study led by Dr Maggie Cheang, Group Leader of the Integrative Genomics Analysis in Clinical Trials group at The Institute of Cancer Research (ICR), identified five new subtypes of hormone receptor-positive HER2-positive breast cancer tumours based on their genetic and molecular profiles. She and her colleagues were able to demonstrate that one of these subtypes did not respond well to a standard treatment.
In another related study, they found that pre-surgery treatment with drugs called aromatase inhibitors converted cancers to lower-risk subtypes, improving patient outlooks and informing tailored treatments after surgery.
Making treatments more targeted
Increasingly, researchers worldwide are finding new examples of cancer subtypes that require a different treatment approach. Treating everyone the same is no longer appropriate in many cases, as cancer biology is too varied among individuals.
Next-generation sequencing (NGS) has revolutionised cancer research by making it quicker and less costly to sequence multiple DNA or RNA samples simultaneously.
However, the price is still relatively high. In addition, the clinical significance of many identified abnormalities is not yet known, so the findings might just cause additional concern without being useful.
So, what is the solution? Developing a standardised, clinical-grade classification method using the minimal number of gene sets that can provide the information necessary to predict patient outcomes in specific treatment contexts.
Taking a genomic tool from concept to clinical reality
Dr Maggie Cheang has made significant contributions to the clinical application of breast cancer panel tests, including co-inventing the PAM50 test. This test quantifies the expression of 50 key genes, building on the foundational knowledge of the five intrinsic molecular subtypes of breast cancer. The concept of breast cancer intrinsic subtypes is established and recognised in both research and clinical communities. The PAM50 test represents a logical advance in understanding the interplay between a cancer’s subtype and patient outcomes.
Throughout her PhD and postdoctoral tenure, Dr Cheang played a pivotal role in both co-developing the PAM50 test and leading its clinical evaluation studies, which facilitated its transition to the commercial version, Prosigna®. Her contributions began in North America and extended to the UK, where she established a research group at the ICR. Here, alongside work on her newly developed biomarker tests, she focused on demonstrating PAM50’s additional benefits.
She said:
“Genomics classifiers give us a more refined, detailed look at individual tumours. We can determine the tumour characteristics of each patient so that we can offer a more precise medicine to them.”
When she was first approached to demonstrate the clinical value of breast cancer intrinsic subtypes in 2003, shortly after this was first reported by Professor Charles Perou and colleagues in 2000, Dr Cheang was the Data and Bioinformatics Manager at the University of British Columbia. In this role, she was responsible for linking patients’ clinical data with their biological data and histological images of more than 5,000 breast cancer samples, and then modelling these data to patient outcomes.
Collaborating with Professor Perou and Professor Torsten Nielsen, she orchestrated all the data and statistical analyses in this project, demonstrating for the first time the clinical value of subclassifying breast cancer in a large human population series into more subgroups beyond the traditional markers of ER, PR and HER2. This work was published in Clinical Cancer Research in 2004, when it became the fourth most cited paper in the history of the journal.
Dr Cheang’s interest had been piqued, alongside the rest of the research community, and she decided that she wanted her PhD project to be on developing optimal diagnostic methods to classify breast cancer subtypes in clinics. With a degree in cell biology and genetics, and a master’s in data science and bioinformatics, she was well-placed to take on this interdisciplinary work.
“I was thrilled,” Dr Cheang said. “I had come to realise that my research could contribute meaningfully to improvements in patient outcomes. I wanted to develop a clinical-grade, reproducible test that could help people. My research interests were broad, ranging from the development of more costly assays to the formulation of methodologies that empower the wider community, especially in resource-constrained regions, to conduct biomarker diagnostic testing.”
Dr Cheang pioneered the introduction of a biologically pertinent scoring system in breast cancer biopsies based on evaluating a cell proliferation marker called Ki67. This system provides clinical pathologists worldwide with a cost-effective method for differentiating between high- and low-risk variants of ER-positive luminal subtypes—the most common molecular subtype of breast cancer. The methodology was swiftly incorporated into the St. Gallen International Breast Cancer guidelines and used by the American Society of Clinical Oncology (ASCO) for risk classification and treatment determination in newly diagnosed breast cancer cases.
During her senior PhD year and early postdoctoral fellowship, Dr Cheang co-invented the PAM50 test. Using machine learning for feature selection from global gene expression profiles across multiple breast cancer tumor series, she and the multi-institute research team refined the focus to 50 genes, making it feasible to perform the in-vitro diagnostic test using standard clinical specimens. Additionally, Dr. Cheang led the biostatistical designs for numerous clinical validation studies, establishing the test's reliability and reproducibility to predict patient outcome.
The patent was licensed to NanoString Technologies, which further developed the test for its nCounter genomic platform, commercially releasing it as Prosigna®. This innovation was later acquired by Veracyte, which secured an exclusive license. The test has been integrated into several international clinical practice guidelines, including those by ASCO, the German Association of Gynecological Oncology (AGO), the St. Gallen International Breast Cancer Conference, the European Society for Medical Oncology (ESMO), the National Comprehensive Cancer Network (NCCN) and the National Institute for Health and Care Excellence (NICE).
Sparing patients from unnecessary treatment
Prosigna® has been used for more than a decade following FDA approval for its use in postmenopausal individuals with hormone receptor-positive breast cancer involving up to three lymph nodes. It offers clinicians a tool to assess the risk of disease recurrence following hormone therapy. In practical clinical applications, the Prosigna® assay has consistently identified patients at low risk of recurrence, whether they have no lymph node involvement or one to three positive lymph nodes. This capability aids clinicians in determining when patients might safely forego adjuvant chemotherapy.
Dr Cheang explained:
“If, for example, someone has a five-year recurrence risk of less than 5 per cent, they may not need or want the toxicity of chemotherapy. However, someone with a risk of more than 25 per cent will likely want chemotherapy to minimise the likelihood of the disease returning.”
This has not always been the approach. Historically, if clinicians removed a large tumour, they would be very likely to give the patient chemotherapy. However, a biological test might now reveal such a tumour to be slow-growing, despite its size, meaning that the person does not need to undergo chemotherapy and face unpleasant side effects.
Further tests on the horizon
Dr Cheang is now using her expertise to help other researchers at the ICR create new biological tests and computational algorithms in a range of cancer types. These include a predictive test for people with advanced sarcoma and a new test for people with breast cancer. The process by which she co-developed Prosigna® is highly transferable, and her 10 years of work on developing the test have taught her the mistakes and unnecessary steps to avoid.
“Now we have robust biological technology and knowledge, we can sequence tumours very quickly,” said Dr Cheang. “I am very optimistic that we will soon be able to develop even more sophisticated tests to direct targeted treatments or spare patients from cytotoxic treatments, particularly in breast cancer.”
Dr Cheang is proud of her work on PAM50/Prosigna® and delighted that it is helping people with breast cancer around the world. She is also extremely grateful to have been listed as a co-inventor despite still being a student at the start of the test’s development.
“I have to thank my supervisors for having faith in me and the other students,” she said. “They demonstrated the significance of inclusivity and proper acknowledgement of students and postdocs where applicable. Recognising students’ contributions to research can be highly motivational – a practice I actively endorse in my lab.”
Dr Cheang’s sentiment is echoed by the ICR, which owns most of its IP but has a contribution system that allows it to acknowledge the work of its students. The ICR is committed to educating and training the researchers and clinicians of the future, knowing that their work will be essential to its mission of defeating cancer.
comments powered by