Skip to Main Content

When Sam Altman, the CEO of artificial intelligence firm OpenAI and a well-known tech investor, saw the pandemic coming, he started investing in biotech companies that could combat it, and fast. Eventually, he had funded roughly 20 different efforts, a mix of investments and philanthropy.

“I think people always try to help in any situation the way they know how,” Altman said. “The one thing I know how to do is fund companies.”


But as he talked to these companies, the tech veteran, who ran the legendary accelerator YCombinator until 2019, was in for a shock. One executive told him a trial would cost about “two.” Two million dollars, Altman asked? That didn’t sound bad. No, the exec said. He meant $200 million.

The fruit of Altman’s frustration: a new effort to speed up Covid-19 clinical trials that is being unveiled Tuesday. As part of that effort, TrialSpark, a startup that is aimed at lowering the cost — and increasing the speed — of clinical trials, will work with companies and researchers seeking to accelerate Covid-19 research.

TrialSpark’s model is based on finding patients digitally, often through social media, and using telemedicine, remote data collection, and even at-home testing and specimen collection to test drugs at home. 


“We’re pushing ourselves … to think about things for Covid differently, more urgently,” said Mark Fishman, a Harvard Professor who was the founding president of the Novartis Institutes for Biomedical Research, who has come on board to help with the project. “You know, this is not your standard trial. It has to be done quickly, but it has to be done well. And you’ve seen already the problems of trials being done or reported in a slipshod fashion.” 

The effort, known as Project Covalence, got its start on March 21, when Altman fired off a tweet asking for help setting up clinical trials. This is the most expensive and time-consuming part of the process of drug development. 

Among the people he connected with was Benjamine Liu, the CEO and co-founder of TrialSpark, which has raised $87 million from investors including Sequoia Capital, Thrive Capital, and John Doerr. Soon, it enticed Michael Mina of the Harvard T.H. Chan School of Public Health to use the platform.

The project’s first study, run by Mina, will be an effort to understand how Covid-19 has moved through Massachusetts using diagnostic tests that check for both the presence of the SARS-CoV-2 virus, which causes Covid-19, and the presence of antibodies to the virus, which indicate that patients have been exposed to the virus and recovered.

Mina said the research will differ from other so-called “serological studies” because it will use sampling methods analogous to political polling techniques to get a representative group of people.

“We all understand that you can’t go to one small community and extrapolate how that population is going to vote, because it can be very biased,” Mina said. The study, he said, “will give us a measure of the effectiveness of all the strategies various policymakers are putting in place.”

Liu, the TrialSpark founder, said the company uses Facebook and Instagram to recruit patients among other methods. For Mina’s trial, large databases akin to this that might be used for political polling are being used. 

“We want to be deliberate as the information is collected to make sure we continue to expand the cohort in a way that it’s demographically quite representative,” Liu said.

Part of the advantage, Liu said, is that the platform can remove the need to “reinvent the wheel” every time a trial is started. Also, TrialSpark, he said, can conduct studies outside a hospital, which is important during outbreaks when hospitals are overburdened and patients do not want to go to the hospital because it could increase their risk of becoming infected.

For grant-funded and not-for-profit efforts, like Mina’s and a study sponsored by the biotech resTORbio that will be included in Project Covalence, Liu says that TrialSpark will run studies at cost. For for-profit efforts, TrialSpark will charge its normal rate.

But it won’t always mean that things speed up, Fishman said. “Some of what we’re trying to do is actually put the brakes on things and tell people, look, I’m sorry, that’s going to take you two more months.” That’s another reality of drug development: haste makes waste.

  • The key is finding the appropriate bio markers that track disease progression. That allows you to rule out medicines that do not work cheaply. For COVID-19 we know know some of the important bio markers are IFN-1, IL-6, D-Dimer, and Ferritin. If we can screen for effective bio markers systemically then that should allow a speed up in research.

  • I am no scientist or Dr. but this book is all about sharing data to speed up a Doctor’s cure of a rare disease. Time saving ideas to speed any vaccine.
    Doctor With Rare Disease And No Answers Decides To … – NPR
    David Fajgenbaum was diagnosed with Castleman disease as a medical student. In Chasing My Cure, he recounts crowd-sourcing his own treatment with a global network of doctors, scientists and patients.

    ‘The Beautiful Cure’

  • Good ideas. One stakeholder not mentioned — folks who are being treated now and are running low on options. A close family member is battling (so far, holding her own) cancer. Development of two drugs that could be very, very useful in her battles down the road was accelerated this year — both combined stage 1 and stage 2 trials. If they work, they’ll be available 6 to 9 months earlier than originally expected. I’ve done a lot of political and marketing polling and see real promise in aggressively using polling techniques to recruit test subjects.

    All that said, I occasionally get hired to help review and if possible partially salvage medical research data (NOT in drug trials but sometimes in behavioral research and medical device work). It is darn hard to write, test, and oversee protocols that might be running in three or five locations or institutions despite the quality of the institutions and the staffs doing the work. Herding cats comes to mind.

    Tough to take shortcuts that don’t bite you at the end. Getting all this to work in a big way requires more than just making recruitment more efficient. But it is a great start.

  • @Richard O, @Buzz, @GordP, it is correct that technology alone cannot solve the search for a vaccine. Nothing and no one in the article set such an expectation.

    Where your bias is showing – an attitude of anti-innovation, that some one from Silicon Valley cannot help change and improve medical process and development that has existed for decades. It’s ironic that I defend SV; I’ve worked there over 20 years and know its biases and issues well. But SV “hunts” innovation like no other industry. Consider how a partnership, each bringing their best talents to the task, might achieve the main goal of a globally-delivered vaccine faster vs. no partnership and a lot of knuckle-dragging nay-saying. Medicine and health care are brimming with technology, but sometimes hampered by slow adoption by doctors and administrators, too old to try new ways. Example: doctor visits and exams via the internet, adoption rapidly ramping during the shutdown by necessity. This capability has been available at least 20 years, and could have been helping many more people sooner, had attitudes within the medical community toward innovation, partnership and new technology been more open to and accepting of new ideas.

    • Telemedicine was not reimbursed by the federal government or insurers until the pandemic. That is the reason why it was not adopted sooner- not the reasons you claim.

    • Buzz, actually, many telemedicine procedures were reimbursed before COVID-19. Medicare and 47 states reimbursed for some things by summer of 2019. The pandemic vastly loosened the rules, but the technology was there and reimbursements were probably doubling annually. Some physical therapy is now reimbursed in some cases, for instance. I can’t see us ever going back. Does anyone really want to warehouse old folks in nursing homes if they could be cared for in a residential setting?

      Broadband networks are getting better and more reliable. Technology for exams is getting better daily. Even years ago, it was clear that many ophthalmology exams could be done as well or better remotely than in-person. Hospitals, medical personnel, government, and insurers have been given a huge push by the pandemic as well. The issue to me is that backsliding (in search of dollars) could be the norm unless we dig in our heels.

  • The effort of companies like TrialSpark can only go as far as finding patients digitally, and assist with data management or devise smart tracking & analysis systems. It would however still not be representative of the general population, as not everyone is hooked up to the digital network. These tech companies certainly can not be involved in any medical exercise of actual drug administration & testing, etc. – for the total lack of medical expertise. Drug trials require far more than a desk, desk-top, lap-top and the internet. Sorry Silicon Valley, you don’t know it all.

  • Sigh…Another well meaning but naive tech investor that thinks great success in tech will seemlessly transfer to another area that they have little experience. It’s okay Michael Jordan wasn’t that great at baseball either.

    No one would ever let one of these guys perform heart surgery even if they were claiming they could “hack” “revolutionize” “Disrupt” (pick your favorite Silicon Valley buzz word) the field of cardiac surgery, so why don’t they think the same Logic applies to drug development?

    While serosurveys in which there is little to no risk can be performed at home, interventional studies are a completely different beast. There is no way that an IRB or regulator would ever allow trial subjects themselves to self administer novel experimental compounds at home without available Medical monitoring.

    • My thoughts exactly, Buzz. The attempt at using technology for medical drug testing is indeed “cute” but definitely not a “game changer” (another barf). You are quite right that new drug trials are not suitable for medical remote control ideas.

Comments are closed.