Chasing The Grail: Bill Gates, Jeff Bezos, Illumina, and Google Ventures Are Betting This Company Will Find A Cure For Cancer

A conversation with Jeff Huber, CEO of Grail

The Grail team outside their SF HQ on “Grail Day” — the first day of the company’s life earlier this spring.

Jeff Huber lost his wife Laura to cancer last fall, a loss made even more devastating by the knowledge he had gained through a mid-career shift into life sciences at Google, and board work with the gene sequencing pioneer Illumina. Just as his wife’s cancer was metastasizing beyond the reach of science, Huber was working with the Illumina board to spin out a company that promised to detect and ultimately provide the tools to beat cancer before it could spread throughout a person’s body.

After his wife’s death, Huber became CEO of the newly spun-out company. Dubbed Grail, it is backed by more than $100 million from Bill Gates, Jeff Bezos, Google, and others. Grail’s unofficial debut came via a moving commencement speech, “Find A Better Way,” that Huber gave at his alma mater in May.

Huber is best-known in the technology world as the senior engineering leader behind a 12-year hit parade of massive data businesses at Google, including Google Ads, Apps, and Maps. At Grail, he’s moved from mapping the world to mapping genomes, using Illumina’s cutting edge HiSeq sequencers — which, working in a pod of ten, can sequence 18,000 genomes in a year, a throughput of roughly one every few hours. Huber’s plan is to spin up scores and possibly hundreds of sequencing machines, each of which spit out a gigabit of data per second during operation. His long-term goal is nothing short of breathtaking — the creation of a living map of human biology, a platform spanning data and biology capable of delivering personalized cures not only for cancer, but quite possibly many other seemingly insoluble diseases.

We spoke to Huber at Grail’s San Francisco headquarters last week.

Your commencement speech at Illinois was powerful.

Thank you. The overall experience was amazing, but there were certainly parts of it that were hard to deliver. When the opportunity came up, at first I said no, since it felt too soon and too raw. But as I thought about it more, it felt like an important message to share.

I find your move to Grail interesting given your large-scale computational background. How fluent were you in science and biology?

I’m becoming fluent, but I started that direction three years ago when I moved into Google X. I ramped up the big data side of Google Life Sciences, now called Verily.

Jeff Huber (photo Grace Huber)

How did you move from Google X to Grail?

The decision to go focus on life sciences at Google came because I was at my 10‑year anniversary. That seemed like a good opportunity to pause and reflect.

If I were contemplating another decade at Google, I didn’t want to just stumble into it. I wanted to be thoughtful. If I’m going to bite off another decade, I want to make sure I’m going into it with a lot of energy and passion.

It was really the early days of Ads, the early days of Apps, the early days of Maps that I found most exciting, and where I think I had the biggest impact for Google, from building the teams, and helping set the strategy and direction.

Also, learning new things is incredibly motivating for me. I wanted to continue challenging myself to learn, as opposed to just getting into a rut and turning the crank on the next big system, where I’d be feeling like I was doing the same thing over again.

As I thought about things I wanted to learn, it felt like the intersection of computer science and life science had the biggest potential to have an impact on the world. Biology is increasingly being digitized by technologies like genome sequencing.

If you can take things that were previously analog, incredibly complex systems like biology, and if you can digitize them and then bring the power of computation and machine learning to them, you can accelerate understanding and accelerate science dramatically.

You could have taken a different tack when you announced Grail — and not set the bar at curing cancer. But you chose to make the stated goal of the company a major moonshot. Are you that certain you can do it?

We certainly hope so. Timing matters, and the intersection of genome sequencing and the computation that is possible now, with new technologies like machine learning, it feels like we are at the right time to make this happen.

There’s a hell of a lot of work to be done around the fundamental understanding of the underlying biology. We’re going to have to do one of the largest scale clinical trials ever done if we are going to make the case for early detection of cancer. But it feels like now it’s finally possible. It wouldn’t have been possible five years ago.

How important is it to your investors that you follow the well-worn path of monetization that large healthcare companies take — create therapies and drugs, patent them, monetize them?

Grail’s purpose is to save lives. Our mission is to detect cancer early when it can be cured. We need to make the gears of capitalism work for us. We need to build an attractive business, a strong, viable business with good revenues and good margins. But it’s in service of the purpose, and in service of the mission.

The way I’ve described it to our team here, and to investors, is that Grail’s purpose is to save lives. Our mission is to detect cancer early when it can be cured. That’s the way we achieve the purpose. You could say “Oh, that’s so noble, it should be a philanthropy.” But I just haven’t seen those models work.

We need to make the gears of capitalism work for us. We need to build an attractive business, a strong, viable business with good revenues and good margins. But it’s in service of the purpose, and in service of the mission.

Do you know how you’ll make money at this? Or are you more like at the early stage of Google Maps — you didn’t know exactly how Maps was going to end up making money, you just knew that turning the physical world into data was going to be extremely useful and you’d figure out how to make money with it later?

The key insight in the early days of Maps was to think about it as a platform instead of a product.

That’s very NewCo. How will you be a platform? If you’re streaming a gigabit per sequencer per second, with possibly hundreds of these machines all pumping out data — that’s rarefied air in terms of the amount of data that you’re creating. Are you going to make that data available to authenticated platform partners?

We’re still figuring out the path, but given the underlying asset that we’re creating, this is going to be a capital‑intensive business. We’re going to spend on the order of a billion dollars doing the large-scale clinical trials and building the underlying database of the genome as reflected in circulation in the blood.

The reason that the blood is interesting is it’s an integrator across many things going on in the body, and specifically what it means to have cancer, which is our focus.

Ultimately, what we want to get to is a blood test that’s integrated in the medical system. Today you go in for your annual physical exam, they already do a blood draw, and they tell you things like glucose and cholesterol levels and basic blood chemistry.

There should be one more thing you can do, which is the Grail test, which gives you the chance to detect early whether you have cancer.

I imagine this approach might identify many other kinds of diseases as well. But do people really want to know if they have them? There’s already an ethical debate about these issues …

We should give people choice to know or not. You’ll likely want to know, if there’s something you can do about it.

Historically, many of these diseases haven’t had good therapies, but increasingly they do now. Look at cancer for example. If it is detected at stage 1 or stage 2 — with today’s treatments, it has 80 to 90 percent positive outcomes, where lives are saved.

But cancer caught later — at stage 3 or stage 4 — has exactly the opposite outcomes — 80-to-90 percent negative outcomes. So … do you want to detect cancer early?

Definitely. So will knowing the molecular makeup of a cancerous mutation — as you will once you sequence them — drive personalized new therapies?

Yes. There’s still a huge, huge way to go, but in the last five years, the new category of treatments called immunotherapies have made great strides. Contrast that with chemotherapies , the historic treatment model.

Chemo is a chemical treatment. It’s poison. The chemotherapy model is to “kill you as much as possible — without actually killing you.” It’s a very blunt instrument that is essentially killing fast growing cells preferentially over stable cells, which is why your hair falls out during chemo. And your stomach lining is fast growing — which is why patients often have terrible nausea and stomach problems during chemo.

Chemo is relatively untargeted. The new generation of immunotherapies — which are at a very early stage — basically enable your immune system to do its job. You’ve got 37 trillion cells in your body. They’re dividing all the time. You have errors all the time. Your immune system is able to detect those errors and eliminate them.

Cancer starts with literally one cell that has the right combination of mutations that let it evade the immune system — the immune system either can’t see it, or the cell is able to co-opt the immune system to bring it resources to grow faster. Immunotherapies help the immune system target the cancer. They will become more and more personalized over time.

So when I go to the doctor for my annual physical, I’ll find out if cancer has developed in my body that year?

In an ideal future state, you do a Grail test, and then let’s say you get the unfortunate news back that you have cancer.

The good news is that we’ve discovered it early, and we see exactly the molecular signature of what is driving your cancer. So here is a synthesized immunotherapy. It could be a shot, a vaccination. It could be a prescription. You get your synthesized immunotherapy that is targeting exactly what you have — it starts targeting cancer at an early stage before it’s become incredibly complex — when there’s a clear target to shoot at.

It isn’t chemotherapy. Your hair doesn’t fall out. You don’t puke your guts out.

You take your prescription or vaccine, and you might feel like you have a flu for a couple of weeks while your immune system wakes up and does its job. If we can, in the future, get a cancer diagnosis to being about as eventful as having the flu, that would be a good outcome.

You take your prescription or vaccine, and you might feel like you have a flu for a couple of weeks while your immune system wakes up and does its job. If we can, in the future, get a cancer diagnosis to being about as eventful as having the flu, that would be a good outcome.

Indeed it would be. Will Grail be in the business of that synthesized immunotherapy?

I don’t expect so. There’s a big industry of people working on therapeutics. And there’s a huge amount of work to do. Remember, I’m describing a future possible state we can get to a deeper understanding of cancer and next-generation treatments. This is not something that’s possible today or next year.

You could be in the business of providing the information to a platform-based ecosystem that builds those therapies.

We’re at a super early stage, but that’s the direction that makes the most sense to me right now.

Is curing cancer really an information processing problem?

That’s the biggest challenge that we’re looking at. How do you accumulate enough data, enough quality of data with the correlation to clinical records, and phenotypic information to be able to make significant correlations?

Grail Values

Let me get back to the question I asked earlier. You could have announced that you were doing a really important life sciences company spun out of Illumina. You didn’t have to reveal that the goal of the company was to beat cancer, and you didn’t have to name it Grail. Why did you set the true north of the company so high?

I joined the board of Illumina about two years ago. After I joined, I had a front row seat — I saw the technology coming along. The discussion that we had at the board level was, “The potential for impact here is huge. This is a moral and ethical imperative that it happen.”

I had seen the technology previously, and was a strong proponent of it. But at the time, my wife was going through late-stage cancer treatment. I was her primary caregiver. There’s no way I could possibly be involved. Unfortunately, Laura then passed away. But then there was an opportunity to evaluate “OK, should I approach this? Should my role be more fundamental?”

Was it called the Grail at that point?

There were a set of candidate names. When Grail came up, it instantly resonated.

I know this is sensitive, but clearly your wife’s loss must have played into your motivations to lead Grail. During the time she was in treatment, you knew of Grail’s potential — but that’s all it was, potential. That must have been deeply frustrating to you.

If Grail had existed five years ago, even three years ago, I am wholly, wholly confident that her journey, her outcome could have been … completely different. She would be here today. When she was diagnosed, she vowed, and I vowed along with her, that we were going to do absolutely everything possible to fight. She has probably one of the most studied cancer cases in the history of the world at this point in terms of the diagnostic testing that we did.

With all of the state-of-the-art testing available today, it yielded insights, and it affected her path of treatment, but it wasn’t enough. The reason it wasn’t enough is because it was caught so late stage. It had spread too far, and it was moving too fast.

Cancer is an evolutionary disease. It’s a disease of mutation. It starts with a single mutation, a single cell, but if you catch it after it’s gone exponential …

As an information problem, it’s way ahead of your ability to create information solutions to it.

Yeah. I felt like we were always chasing it.

Snapchat recently raised $1.3 billion. As much as I think Snapchat’s cool, they’re not exactly curing cancer. Funding a company like Grail, you need investment at a Snapchat level. Where is the next billion coming from? Do you worry about that?

I definitely worry about it. We need 10‑year or 20‑year investors. Back to one of your questions earlier — of why we declared that cancer is the thing that we’re going to tackle. It provides much stronger sense of purpose and mission. We’re not wandering the wilderness. We have a very clear, and specific focus around what we’re doing, and it’s one that is a huge importance to the world.

It must be great for attracting the right people to the business.

It attracts the right kind of team. It’s a very mission-driven organization. Everyone knows exactly what we’re doing, and the reason that people get up in the morning excited is because they see and feel the impact of it. They’ve had an experience with cancer themselves, or they’ve had a loved one who they’ve lost to cancer.

It’s incredibly motivating, and it helps us recruit the best people in the world.

Then from an investor perspective — people are looking for great investments, but if you can add, “Here’s the positive impact you’re having in the world while you’re making money,” that’s the next dimension.

###

If you want to share this story, please hit “Recommend” below. It really helps us spread the word. Also, this story is sent first to readers of NewCo’s just-debuted weekly newsletter. Want to get it first? Subscribe free here.