Spitting to fight cancer is a cunning gambit

Hello there,

We’re still a long way from a world without cancer, but there are a bunch of startups working hard to take us closer to that dream.

Today’s featured startup has learned lessons from teams that have come before it, as it works towards an affordable home-testing kit for prostate cancer, using just a quick spit of saliva.

Scroll down to read all about GambitBio.

Meanwhile, the great news that the UK is back in the EU’s Horizon programme will no doubt be celebrated by many of the future startups we’ll feature in this newsletter.

– Martin

👀 You saw them first on PreSeed Now

There have been some notable developments this week from companies we’ve previously featured:

• A.I. governance startup Enzai, which we covered way back in May last year, has raised a €4 million round

• Chatloop, one of the startups out there at the moment rethinking what a web browser can be, has raised £2.1 million and been featured in The Times and on Sky News. We featured them nine months ago.

• Spacetech startup SpaceDOTS, which we featured in March, has also been profiled by The Times this week.

PreSeed Now gives you an eye on interesting startups, way before you see them elsewhere. And our Startup Tracker updates our paying subscribers on them all, whenever they have big news.

GambitBio wants to change how we think about cancer tests

Cancer detection and treatment has improved enormously over the years, but there’s still much scope for catching tumours earlier. 

One case in point: saliva is emerging as a potentially powerful front in the fight to find and treat more cancers at an earlier stage. It’s a field of research that goes back to the 1950s, but in recent years research and entrepreneurship around it has grown significantly.

One startup that says it’s found a particularly promising route to a future where people can check themselves for signs of cancer is London-based GambitBio.

The startup is developing a home self-test kit for early cancer detection. Following in the wake of home fertility tests, DNA tests, food sensitivity tests, and Covid tests, they want to “see a world where people aren't scared of cancer anymore,” as co-founder and CEO Tiffany Ma puts it.

The plan is to sell tests that involve users spitting into the test device, which displays lines similar to a Covid test, but designed to denote signs of specific types of cancer. The first product they intend to bring to market will be focused on prostate cancer detection. 

The plan is for the test to have a number of lines that indicate the likelihood of a cancer diagnosis, based on the test. Once they have data from initial users, the team will introduce a smartphone app to accompany the test.

“Say you have 10 lines on one strip and several lines come up. We don't want to just say ‘there's a 70% chance’ because each marker on the line has a slightly different weighting,” says Ma.

“So we want to make use of the smartphone app where users can snap a photo if there seems to be any signal of a positive result. Then our app can give a very preliminary interpretation of the results.”

Eventually, the startup wants to sell its tests through shops like Boots in the UK and Target in the US, as well as via a direct-to-consumer subscription. Customers would be able to buy tests specific to their own health profile, each targeted at a different cancer.

The accuracy challenge

‘You have Covid’ and ‘you have cancer’ are–for most people in most situations–diagnoses with very different implications for their future health. 

How is GambitBio sure this test is going to be accurate enough that it’s not terrifying healthy people or giving cancer sufferers a false sense of security?

“We control how sensitive the test is, but there's always a trade-off,” says Ma. “If you make a test more sensitive, it is also more likely to increase false positives. 

“So we've been doing interviews with different potential end users, and people would rather prefer to have a false positive than a false negative. They would rather think they might have cancer and then go to the GP and check that they don't.”

And GambitBio will use a number of different biomarkers, rather than just one, to improve the accuracy of the result.

“We created a huge systematic literature review of a panel of potential very specific biomarkers for prostate cancer. And the interesting thing is that there are certain types of biomarkers that can be detected way before the cancer is actually advanced in its stage,” says co-founder and chief science officer Susan Kilgas.

“That offers us the possibility to detect it very early on and give reassurance to the customers that when they detect it early on, they can get proper treatment. We aim to always create this panel of markers that indicate early stage disease. 

“No single marker has the ability to predict cancer, per se. So that's why we need to look at specific combinations of different types of biomarkers.”

And the GambitBio team is aware that the smartphone app’s accuracy needs to be good too. With this in mind, they’re planning tests to account for variables like the quality and position of the camera.

Why look at proteins?

Ma explains that cancer detection often involves looking at the levels of a gene or a protein, which can be an unreliable indicator. 

“We're combining looking at the levels of a protein with changes to that protein. Because in the disease world when a protein is made, and it's out there in our body, it can get modified, depending on whether or not your cells in tissue are diseased. So we're looking at the levels of the protein and what is stuck on these proteins.

“It’s not a new thing. This is the bread and butter of cell biology. It's just that it's never been used in the context of diagnosing cancer yet, because there are a lot of logistical hurdles and regulatory hurdles towards validating these biomarkers, so people haven’t been incentivised to do so.”

Ma adds that while DNA and RNA sequencing has been cheap and easy to do for some time, the ability to look at these protein changes has improved greatly in recent years and they provide a much better insight into the likelihood of cancer than DNA and RNA alone.

Co-founder and CTO Shilin Chen explains that proteins are more reliable to test in saliva: 

“DNA is oftentimes very unstable when they're in the extracellular environment, for example, your saliva, especially with all the proteases, all the enzymes that might break them down. Proteins are relatively more stable in that case and last for longer, and they're way easier biomarkers to be handled versus DNA.”

Background

The GambitBio story has its beginnings in California. Ma and Chen met as biology and bioengineering students at UC Berkeley. Being so close to San Francisco and Silicon Valley, they got bitten by the startup bug, but Ma wanted a PhD before she dived into entrepreneurship.

So Ma came to Oxford to work on a PhD. It was in Oxford where she met Kilgas and the startup’s other co-founder, Cristiano Peron.

Inspired by how readily people tested themselves for Covid during the worst days of the pandemic, they pitched the idea behind GambitBio at a university competition and were surprised to discover that it hadn’t been done in this way before. They wondered why.

“We asked around, and the answer was that people don't expect to diagnose cancer from a simple lateral flow test, even though it can be done. It's just been underestimated,” says Ma. 

“Also, when teams tried to do something like this, there was something wrong with the sample size when they tried to do protein studies - they were too small.”

Over time, GambitBio evolved from a side project into the startup Ma had envisaged starting back at Berkeley.

Next steps

The GambitBio team is currently developing its second prototype test and lining up a partnership with a lateral flow test manufacturer with a view to be ready for mass production.

But products like this take time to get right and cleared for commercial launch. Ma estimates the saliva-based prostate cancer at home self-test kit could come to market in 2026 or 2027.

Subscribe to Premium to read the rest.