The right combination for early detection

Early detection is without a doubt one of the most immediate ways to improve cancer survival.

For most cancers, there is currently no consistent early detection process available and, even where screening programmes do exist, sampling is often infrequent and still requires extensive follow-up to verify cases. To meet the potential of early detection we must invest in new technologies that are suitable for high-frequency sampling, or even ongoing health surveillance.

While a system that constantly monitors our bodies for cancer may sound farfetched, the development of smart phones means ongoing health monitoring has rapidly become a part of everyday life. Through over a million health apps, people are recording heart rates, sleep patterns, diet, digestive health, blood glucose, sexual activity and much more. While the quality of some of these technologies may be questionable, the developments in this area serve to demonstrate both a popular desire for ongoing health monitoring, and a willingness to participate in the collection of such data. So why not for the early detection of cancer?

The need for collaboration

In an upcoming session at this year’s conference – ‘How to detect: emerging technologies for understanding signals over time’ – I will be joined by Stephen Friend of Sage Bionetworks, Shan Wang from Stanford University and Daniel Heller of the Memorial Sloan Kettering Cancer Center to discuss the efforts being made to ensure early detection through the use of tools that monitor health over time.

The ultimate vision of such approaches is to build a more complete picture of patient biology that establishes a robust, individualised baseline for health and provides high sensitivity detection of developing diseases.

Early detection screening today is a compromise between risk and resource, and testing is typically only offered to at-risk groups every few years.

To be useful, these technologies need to meet several key requirements. They need to be affordable, accessible, and can’t overburden our healthcare systems. They should depend on methods that are acceptable to most patients, ideally being non-invasive. And, they should be multi-factorial – drawing from across as wide a range of available health data and medical expertise as possible. Achieving success in this area will also depend on the direct involvement of the public, reflecting a growing desire for people to take an active and informed role in monitoring and managing their own healthcare.

Developing technologies for the early detection of cancer is a highly collaborative process. In addition to medical and research expertise in cancer, there is a real need for specialists in biomarker detection, data collection and analysis, and often engineers to develop testing devices. Exploring the applications for field-asymmetric ion mobility spectroscopy (FAIMS) devices, for example, was a key aspect of my own initial efforts in cancer detection.

This need for collaboration, has been a key driver for Dr Stephen Friend, who, following years of work in cancer research, co-founded Sage Bionetworks – a non-profit which aims to provide tools that foster collaborative, large-scale biomedical research.

Three key factors

Reliable early detection of cancer relies on three key factors; timing, sensitivity, and specificity. By definition, checking patients for cancer more often reduces the average length of time between a cancer starting and the first opportunity for it to be detected. In turn, this increases the likelihood of finding cancer early. Of course, healthcare resources are limited, and cancer tests are often lengthy, expensive, require specialist skills, and can even pose their own risks to a patient’s health. As such, early detection screening today is a compromise between risk and resource, and testing is typically only offered to at-risk groups every few years.

Limited resource is also one of the reasons why specificity is key. Early detection tests that aren’t specific may find a lot of cancers – but that could well include those indolent malignancies that won’t cause harm. As such they’ll also result in a lot of healthy people being referred for further testing, or treatments that they don’t need. This represents a significant additional healthcare burden as well as unnecessary stress for patients and their families.

As we have come to understand cancer more, the difficulty of effectively identifying early-stage cancers has become increasingly apparent. While an advanced metastatic cancer has many genetic and metabolic differences compared to healthy cells, the distinction between an early-stage cancer and a typical cell, particularly in an ageing patient, is much more subtle. A patient may have many cells containing benign genetic aberrations and we need to be able to determine which have the potential to evolve into life threatening cancers.

As well as providing more opportunities to detect cancer, increasing the frequency of testing has synergistic benefits for sensitivity and specificity. Multiple time points can be used to define trends over time, which could indicate changes that are too subtle to be identified as unusual when using a sparser dataset. Similarly, the accumulation of many data points during good heath can help to establish a clearer concept of how an individual’s biology varies over time, providing a more robust background against which to compare new data and detect cancer indicators.

While an advanced metastatic cancer has many genetic and metabolic differences compared to healthy cells, the distinction between an early-stage cancer and a typical cell is much more subtle.

While devices and digital solutions have clear roles to play in the future of early detection, there is also progress to be made through innovative biotechnologies. Probes, for example, that – unlike radiation and X-rays – are safe and suitable for frequent use could be developed for cancer detection. These could be used frequently and would dramatically increase sensitivity. By comparison to the current approach, which looks for fractional changes to existing biomarkers, this could provide a more conclusive ‘all-or-nothing’ result. In the context of non-invasive biomarkers this can include exogenous volatile organic compound (EVOC) probes, or more complex nanoparticles such as those under investigation by Dr Daniel Heller of the Memorial Sloan Kettering Cancer Center.

Showtime…

There are various challenges to overcome, and significant progress on early detection is likely to require continued international collaboration across diverse fields of expertise.

This is why I’m excited to be joined by Stephen, Shan and Daniel for the session I am chairing at the conference. Each will bring a wealth of knowledge and expertise in developing technology and tools for early detection.

Many approaches are being explored for tests and probes that could be developed for public screening programmes. The aim is that at-home cancer testing that integrates with health data collected through smartphones could soon be available to facilitate early detection across the full range of cancer types.

Author:
Billy Boyle is one of the original co-founders of Owlstone Inc, spun out of the University of Cambridge in 2004, where he was an engineering graduate.

In 2016 Billy helped to establish Owlstone Medical Ltd. to explore the medical applications of field-asymmetric ion mobility spectroscopy (FAIMS) technology and the development of non-invasive breath biomarkers. This effort was partly inspired by his wife who was diagnosed and later died of colon cancer as a result of a late diagnosis.