Hilary term is always busy for most academics, and it is not helpful to feel under the weather while I have signed up for a long-awaited DwD by Ben Schuster-Böckler, a leading computational biologist and Official Fellow of the Cellular Life theme at Reuben. I walked into the talk, titled “How close are we to detecting cancer from a drop of blood?”, thinking all I know about this topic is the infamous Theranos scandal, and I may get dizzy if it gets too technical. As a fellow “Cellular Lifer”, however, I had great hope that Ben would charm the audience with a fascinating but down-to-earth seminar, and that was exactly what he did.
Ben kicked things off with a quick rundown on cancer - what it is, its stages, and why catching it early can be a game-changer for better treatment and outcomes. He painted a picture of the current state of cancer diagnosis and screening in the UK, pointing out that, despite some solid national screening programmes (for breast, cervical, and bowel cancers), they contributed to <10% of all cancer diagnoses, compared to over 50% via GP referrals, although the screening programmes do detect more cases in earlier stages. The issue is further complicated by the fine balance between the costs and benefits of the screening programmes, and under- and over-diagnosis, the latter of which could result in unnecessary mental or physical harm for the patients to go through medical procedures for cancers that wouldn’t have caused any harm at that stage.
Then, we dove into the meat of the talk: the buzz around using “simple” blood tests to catch cancer early. It may be a simple blood draw for the patients or clinicians, but the underlying science involves complex computational biology that tries to distinguish the DNA methylation signature of cancer cells from normal cells that die (as they naturally would) and enter the blood circulation. One of the biggest challenges is that, at any given time, there would be an uncountable amount of all sorts of dead cells and their fragments with widely varying DNA methylation signatures in your blood – so the dream to detect multiple types of cancer from a drop of blood is really finding a needle in a haystack.
Beyond the Theranos scandal, there has been enormous, legitimate R&D effort put into the trillion-dollar challenge. At the forefront of the competition is the Galleri test developed by Grail, but even this poster child of the field, which, in part, motivated a USD 8 billion acquisition, fell short of expectations with suboptimal performance. While further studies are ongoing, the test seems to lack the sensitivity (i.e. the ability to correctly detect the true cancer cases), and its performance varied by populations. Most importantly, the test appears to perform worse in detecting early stages of cancer, which is the prime aim of blood-based cancer screening. It seemed like we are still a long way from the Holy Grail.
Ben ended the talk with three intriguing questions on whether we would take the Galleri test yearly, should the NHS roll it out to all adults, and how much better or cheaper does the test need to get before we all jump on board? Surely not everyone in the audience knows the technical details behind the tests (I don’t), but the talk was crystal clear and the questions were relatable. I thoroughly enjoyed the natural and lively discussion at my table with students from engineering, medical science, and mathematics backgrounds. This was followed by a buzzing Q&A session covering a diverse range of aspects of the topic, from sequencing technology to ethical issues, to the demographics of cancer and to ways to improve these tests. For a good 1.5 hours, I have totally forgotten my deadlines and troubles and enjoyed a great evening of intellectual but casual chat. Now I know what to do whenever I need a fruitful break – attend a DwD at Reuben.
