Ben Liu, CEO of Trial Spark

TrialSpark is a health technology startup that is reimagining the way clinical trials are performed. By partnering with independent and community physicians, TrialSpark transforms their practices into high-quality FDA compliant trial sites. TrialSpark recruits diverse patient pools to these sites and powers their trials using a roaming cohort of certified research coordinators, optimized by software, data and technology. The Pulse sat down with co-founder and CEO Ben Liu to learn more about the company’s origins, latest accomplishments, and vision for the future of clinical trials.

The Pulse: Can you share a bit about your background and how it influenced your journey to becoming an entrepreneur and starting TrialSpark?

BL: I trained as a computational biologist. Like many others, I was really excited about the confluence of big data applications in health care mixed with imaging, clinical data, and AI/ML to potentially transform drug discovery. In my graduate work, my lab discovered a few candidate drugs. When we talked to pharma companies about exploring a clinical trial, we were exposed to how expensive and time consuming that process was. It was eye-opening to learn how they review hundreds of drug candidates a year, and that a new clinical trial could cost hundreds of millions of dollars across all phases.

My core realization was this: as technology gets better and better, our ability to discover candidate drugs is getting more efficient, but the cost and time of a clinical trial is only getting more expensive – and the trial bottleneck is only going to get worse. That was the main motivation behind starting TrialSpark and our core mission to bring new treatments to patients faster and more efficiently.

The Pulse: Startups in this space are trying to accelerate drug approval through many different approaches: from clinical trial operations to AI-powered pharma R&D and real world evidence. What motivated you to take TrialSpark in its current approach, specifically with clinical trials?

BL: There are so many companies tackling the “AI in drug discovery” part – that was my background as well. But no one was really tackling full-stack how to run clinical trials cheaper and faster. I realized that if we weren’t going to do it, no one else was going to do it with the same urgency. If you can run trials cheaper and faster, you enable more people in the ecosystem to take more shots on goal. Time is the #1 driver – if you can save a month from a $1.2B/year drug with a 20-year patent life, that’s worth hundreds of millions of dollars.

The Pulse: How can clinical trials be improved?

BL: There are a couple of issues with the way the industry is currently structured. Pharma companies hire CROs to run trials, but trials are really run at the site level. In any trial, each site operates differently – one may collect data using a paper case report form, and another may use an electronic system to collect on the site level. They all have different operating procedures, and so that CRO layer really exists to manage variability – in sending a human for every few sites, collating data, and sharing it back to the sponsor. That means there’s always latency to where the sites see the data, to where that CRO layer collates the data, to when it’s shared back to the sponsor. There are also a few misaligned financial incentives: the CRO business model is a cost-plus model, so the more variability there is at the site level, the more billable hours they can charge to the sponsor.

There is also a site shortage – we need two orders of magnitude more physicians involved in clinical trials to fill up all the trials on clinicaltrials.gov. For example, if you’re running a dermatology study today, every pharma and biotech company are using the same sites. They’re all competing for the same patients, and after a few years all the eligible patients will have already been picked off.

The Pulse: How does TrialSpark’s platform address these issues?

BL: We take a full stack approach, realizing that we need to own the process end-to-end in order to really align incentives. Our model is this: we find where the patients are, the doctors that see those patients, and create an FDA-compliant trial site embedded in that doctor’s practice, which will allow us to reach 98-99% of new patients. When we create these trial sites, they all can operate the same way through our own end-to-end tech platform that manages logistics, data collection, and analysis – this helps to eliminate site level variability. We also employ full time research coordinators that go around to our trial sites to do visits at the site level. It’s like a half-AirBnB model where we enable any doctor to create a trial site in their practice.

The Pulse: Today, TrialSpark can complete trials 6 months faster than traditional CROs. Can you give us an idea of how you achieve this?

BL: One big area we’re more effective on is study start-up. Traditionally, when you’re running a Phase 2 trials with 50 trial sites, it can take 3-8 months to start a single site because each will have its own contracting processes, negotiated budget, and IRB. For us, it’s just one rate card, one budget, and one IRB across all of our network. That process itself can save 3-5 months just from start-up. What really matters is how quickly can you get all of your sites up, because until then you’re not running your trial at full capacity. To add on that, you can only complete your trial as soon as your last patient is enrolled. We can save a ton of time this way over a traditional CRO.

The Pulse: How are current stakeholders (physicians, sponsors, patients) responding to TrialSpark’s solution

BL: We’ve been really surprised by how excited physicians are in partnering with us. We actually found that the doctors that are most successful care less about the financial aspects – rather, they’re really excited about being able to continue to do research and be a principal investigator while being in private practice. They’re excited to give their patients more access to new treatments typically found only at major centers.

From a sponsor perspective, I find that everyone is rooting for a faster, more efficient way to do this. They resonate with the idea of democratizing access to clinical trials and reaching more patient populations. There is appropriate risk aversion in our industry for anything that’s new because of the stakes involved. But we’ve been really encouraged by the number of sponsors who want to join us in reimagining the how we can reach more patients and to run their trials more efficiently.

For patients, we’ve gotten a lot of great feedback and testimonials from those participating in our trials. There is a deeply personal experience of a patient’s journey throughout a clinical trial that we’re able to improve on through meaningful interactions with caregivers. We’re excited to explore broader initiatives such as observational registry studies to expand value for patients.

The Pulse: The theme for this year’s Wharton Health Care Conference is “New Frontiers in Health Care”. What are some upcoming trends in the clinical trial space that you find exciting? How do you see TrialSpark participating and leading in those areas?

BL: In the clinical trials space, we’re excited about the growing trend towards using adaptive or basket type trials, which can enable the testing of multiple hypotheses per unit trial. They can have multiple treatment arms in a certain type of indication or have patients moving from one arm to another based on preset criteria, which can significantly increase the efficiency of a trial. Traditionally, it’s hard to run an adaptive trial if trial sites can’t read out data in real time, so we’re excited about how our platform can play a big role here.

We’re also excited about explosion of innovation in drug discovery. As more drug candidates get developed over time, we believe TrialSpark’s platform can help find a home for these assets, rapidly scale trials, and get them to patients.

Interviewed by Vincent Yeh, December 2019