The Future of Cancer Diagnostics vol. 1

Cancer is the cause of 1 in 8 deaths globally. In the United States, there are 1.6 million new cancer cases in 2016 with about 600,000 deaths every year. It is also well known that early detection is instrumental in effective treatment and management of the disease. For example, lung cancers diagnosed at stage one generally has a five-year survival rate of over 50% — if the same cancer is diagnosed just two years later, the survival rate drops to under 10%. In additional to improving clinical techniques, technologies targeting presymptomatic diagnosis — where the cancer is found even before any symptoms appear — is currently the hot topic for pharmaceutical and healthcare companies alike. However, since symptoms are what drive people to the hospital, people that believe they are healthy tend to have a much lower tolerance for medical invasiveness. Thus, one of the main challenges of these technologies is that they would have to be somehow be incorporated into the everyday lives of normal, healthy people. For today’s article, I will be looking at one startup tackling this problem in cancer diagnostics, a topic that is very close to to my own PhD research area.

Freenome is a California based startup that utilizes AI to recognize certain disease markers in blood samples. By detecting certain patterns of DNA fragments in a blood sample, Freenome’s algorithms aim to predict very early stage cancer and other disease developments. The startup recently closed a $65 Million Series A and is currently backed by Google Ventures, Polarias Partners and other investors. It has ongoing partnerships with UCSF, Moores Cancer Center, and the Massachusetts General Hospital.

On an invasiveness scale, blood tests are actually about as non-invasive as you can get while still acquiring large amounts of useful information. Getting your blood taken won’t be as convenient as a cancer diagnosing skin patch (which would be incredible) but plenty of people do give regular blood samples for other reasons. If the test is inexpensive and accurate enough, I could see it being implemented as an optional portion of an annual physical checkup. However, since extracting free DNA from blood samples in a daunting task in itself, I am concerned on how much blood each test would require- which will affect the invasiveness of this test. Perhaps, Freenome could seek a partnership with the next startup that I will review in this series (hint hint). Also, previously, I had said “accurate enough” because the test does not have to be 100% conclusive; I’m sure the founders didn’t aim for Freenome to be the final confirmation of whether a person has cancer or not. It just has to be good enough to warrant further diagnostics with potentially more expensive and invasive technologies, like MRI/CT. One side note is that, due to the Theranos scandal, any lifesciences startup involving blood test just has that extra burden of proof in demonstrating their technologies to investors. However ,with strong partners, investors, and a solid technology base, Freenome is definitely one of the life science startups to watch out for in 2018.

Edit: Part II and part III of this series are now out!

Related posts:

If you could have dinner with anyone from the past, present, or future who is dead or alive who would it be? I never really knew how to answer this question up until a few years ago where I came…

Social media takes a more important place in our lives now than ever. We could not imagine a life without Facebook, Instagram, Twitter and the list goes on. Social media channels are not solely used…

Yesterday I scheduled over 300 display fonts to my Pinterest account (feel free to follow me). A crazy and time-consuming experiment to see if I can get some traction on my account. It’s not…