Tackling a ‘Genius Virus’

Pharmaceutical Executive, Pharmaceutical Executive-08-01-2020, Volume 40, Issue 8

Patrick Soon-Shiong, the billionaire biotech entrepreneur and owner of NantWorks, a network of health startups, discusses his companies’ COVID-19 vaccine candidate and its “fighting chance” to outsmart the virus’s hidden tricks through merging supercomputing and modernized DNA delivery.

Dr. Patrick Soon-Shiong is a physician, surgeon, researcher, philanthropist, and owner of the LA Times. From 1997–2010, he served as founder, chairman, and CEO of American Pharmaceutical Partners (APP) and Abraxis BioScience, where he developed the drug Abraxane, which received FDA approval for the treatment of metastatic breast cancer, lung cancer, and advanced pancreatic cancer. In 2008, Soon-Shiong sold APP to Fresenius for approximately $4.6 billion. Two years later, he sold Abraxis to Celgene for about $3.8 billion.

Soon-Shiong currently serves as chairman and CEO of NantWorks, an “ecosystem of companies” aiming to create transformative global health information and a next-generation pharmaceutical development network.

Under the NantWorks umbrella, Soon-Shiong is also chairman and CEO of NantKwest, an immunotherapy company focusing on natural killer (NK) cells, and ImmunityBio, a company that aims to deploy a broad portfolio of biological molecules as an integrated cancer vaccine platform targeting multiple tumor types without the use of high-dose chemotherapy.

Since the outbreak of COVID-19, Soon-Shiong has been applying his various resources toward understanding and combatting the virus. In May, ImmunityBio’s COVID-19 vaccine candidate was one of 14 companies selected for Operation Warp Speed, a national program to accelerate COVID vaccine development. The candidate is the first human adenovirus (Ad5) vaccine designed to deliver both spike (S) and nucleocapsid (N) DNA for potential long-lasting humoral and cell-mediated immunity.

Pharm Exec caught up with Soon-Shiong to talk about how his companies are progressing in developing a COVID-19 treatment and his goal to produce billions of doses to tackle the virus worldwide.

PE: Can you talk about how you became involved in the fight against COVID-19?

Patrick Soon-Shiong: My companies, Abraxis BioScience and American Pharmaceutical Partners, were working back in 2010 on the hypothesis that we could drive memory T cells for both infectious diseases and for cancer. We had developed Abraxane, which is now owned by Bristol Myers Squibb. Unfortunately, much as I tried to explain the motivation behind my developing Abraxane was to avoid high-dose chemotherapy by a paradigm change of using this protein nanoparticle as an immuno-modulating agent rather than a cytotoxic drug, the concept was not understood. By the early 2000s, I came to realize that high-dose chemotherapy, still the current standard of care, was counterintuitively the wrong therapy and would do little more than induce metastasis and drive resistance toward an incurable outcome. What my colleagues did not realize is that Abraxane was a protein nanoparticle designed to penetrate the cancer microenvironment and activate the macrophages as well as induce what are known as DAMPS (damage-associated molecular patterns or activators of the innate immune system) to expose the tumor to our immune system.

That stimulated in me as far back as 2008 that what we’re doing with regard to chemotherapy was all wrong, that high-dose chemotherapy was wiping out our immune system. Only by activating the immune system could we ever have a chance to find a cure.

But I couldn’t convince anybody. I basically threw up my arms and said, fine, I’m selling APP to Fresenius, I’m selling Abraxane to Celgene, and we’ll go stealth for a decade and prove this hypothesis.

For the past decade, NantWorks has spent close to $3 billion building five to eight manufacturing plants with full R&D facilities on properties spanning about 40 acres in Colorado, California, and Chicago. We needed to collect within NantWorks all the tools in the armamentarium that could prove this hypothesis. That meant we needed to develop a vaccine vector that could drive the code of whatever cargo or transcript is necessary into the human being, whether it is a tumor-associated antigen, a new epitope, or a COVID sequence, into the dendritic cell and allow that dendritic cell to pump out this code as long as possible, so that we can activate the T and B cell memory.

For this vaccine vector, we selected and modified an adenovirus vector, which we call the second-generation adenovirus and which, by the way, has eluded every company, including Merck, Johnson & Johnson, and AstraZeneca, who with Oxford University in the UK is still developing a first-generation adenovirus.

First-generation adenoviruses produce in us adenoviral fibers and in so doing completely limit the capability of the dendritic cell to pump out its code, because these fibers alert the immune system in a healthy person that there is something foreign inside the dendritic cell, and therefore kills that dendritic cell.

The breakthrough was to develop an adenovirus that had all the propensities of the first-generation adenovirus, that is, being able to get into the dendritic cell and drive the code, but to remove the capability of that adenovirus to make adenoviral fibers, (or proteins) so that it remains stealth. All of a sudden you have an adenovirus that can actually pump out its code for as long as six months, which is remarkable. To us that was the breakthrough.

The next step was to figure out what happens as you age. As you age, your T cells die, your NK cells die or diminish. The molecule that up-regulates growth of these cells is IL-15, so we brought in a mutant of IL-15, which is called N-803, and ran that through clinical trials, specifically in bladder cancer. And that has received a breakthrough therapy status from FDA; it met the primary endpoint in bladder cancer. When you add to that a NK cell, which we could grow in an unlimited supply, and modify that NK cell to target the tumor, you now have a triple whammy, what I call the triangle offense (a nod to Soon-Shiong’s love of the LA Lakers and the famous offensive strategy run by former head coach Phil Jackson). You have a dendritic cell to activate the B cell. You have the memory T cell, and now you have a targeted NK cell.

Using that, we were able to get a complete remission, now lasting over six months, in metastatic pancreatic cancer; a complete remission, now lasting over a year, in triple-negative breast cancer, that had failed all previous therapy; and complete remission now lasting four years in Merkel cell carcinoma, that had failed all therapy. So, we have evidence that the hypothesis was correct.

When COVID-19 came along, it appeared that everything we were doing in cancer and infectious diseases was ready made, thank God, for COVID. Our manufacturing capacity was ready because we were growing NK cells. So we jumped into action. But one advantage we had over other companies is we made a scary discovery about COVID when we were doing molecular dynamic modeling of the receptor-binding domain (RBD), which is the tip of the spear of the virus’s spike protein, to the ACE2 receptor, which is the receptor in the lung or the body.

We discovered that this genius virus has figured out a way to hide its RBD, or the business end of the virus. In partnership with Microsoft, we now have a hundred terabytes of data, using a couple of thousand GPUs, and by analyzing this interaction, we discovered that the spike protein when given by itself, unfortunately, hides this RBD. Using an unmodified spike in combination with a first-generation adenovirus vector is a huge limitation to this approach. The second-generation adenovirus, however, is stealth to the dendritic cells; it’s hidden from T-cell attack. You have a much longer time for the appropriate cargo to dispense.

PE: How does your vaccine differ from other adenovirus approaches?

Soon-Shiong: Ours is a second-generation adenovirus (Ad) that includes the nucleocapsid gene as well as the spike gene, allowing for a broader immune response. Other Ad approaches use spike only and this could allow the coronavirus to mutate and escape the immune response.

By including nucleocapsid as well, we significantly reduce the ability of the virus to escape. The others are also using first-generation adenovirus, but boosting with this same vaccine may not be possible and won’t work as well if in people previously exposed to the adeno “common cold virus.” Ours avoids the preexisting immunity issue altogether, allowing multiple boosting/dosing if necessary [as in our cancer trials].

PE: What’s your timeframe for moving forward with this?

Soon-Shiong: By January 2020, we had undertaken the sequence. By March, we had confirmed that the antibodies recognized our protein construct. By April, we created finished dosage form of both constructs; we submitted our IND (investigational new drug) and went immediately into scale-up. We have a magnificent facility already activated that has a capacity now of 100 million doses by the end of the year. And so, in Q3 we will do first-in-human studies, and then we’re off to the races with everybody else, and we will hopefully be on the path to the billion doses.

PE: Can you talk a little bit about the supercomputing power you are deploying to understand the COVID virus?

Soon-Shiong: Around 2012, NantWorks took over National LambdaRail, which ran the Large Hadron Collider and was behind the 12,000-mile fiber-optic network that links physicians with data from big science projects such as The Human Cancer Atlas. That was the basis of us building our supercomputing network. Now, in collaboration with Microsoft, which has been a fantastic partner, we have one of the largest graphics processing unit (GPU) clouds.

We’re leveraging Microsoft’s Azure platform to perform a highly detailed computational analysis of the spike protein structure of the SARS-CoV-2. A digital blueprint of the spike protein obtained via a process called cryo-electron microscopy was published in February by researchers at the University of Texas and the NIH. Our teams have applied a technique called molecular dynamics to the blueprint. Molecular dynamics analyzes the physical movements of the virus components at the atomic level over an extended period of time and runs a series of computationally intensive simulations that result in a detailed model of the most likely solution structure of the spike protein.

Microsoft and ImmunityBio’s engineers and scientists deployed a cluster containing over 1,250 NVIDIA V100 Tensor Core high-performance GPUs specifically designed for machine learning and other computationally intensive applications. Similarly, ImmunityBio has deployed its 320 GPU cluster, which has been optimized for molecular modeling of proteins, antibodies, antivirals, and targeted small molecule drugs. This has given us an amazing amount of information about where to attack this virus. We’re looking into neutralizing antibodies and finding the best sites on the virus to attack. I anticipate we’ll make some exciting announcements in this area soon.

PE: That immense amount of computing power notwithstanding, you noted that SARS-CoV-2 is a “genius virus.” Can you elaborate?

Soon-Shiong: It is a genius virus. What’s happened is that just one mutation on the RBD has increased the infectivity of this thing by 20 times. That’s why it’s so effective.

What is encouraging, however, is that we have discovered that patients who have recovered 11 years on from the first SARS-Cov outbreak have memory T cells to the nucleocapsid (N) phosphoprotein.

This validated our decision to add N into our construct so that we can generate memory T cells, because while spike and RBD will mutate, N can’t mutate.

The science of this virus is fascinating. Integrating this computational science with modernized DNA systems and modernized genomic devices, such as the second-generation adenovirus that can actually induce this code in the dendritic cell, gives us a real fighting chance to create memory T cells against SARS-Cov-2.

PE: From an operational point of view, how are you managing to balance your resources in terms of fighting COVID-19 and continuing with your work in cancer?

Soon-Shiong: I hope that our government recognizes that a second-generation human adenovirus, which uniquely includes N in addition to S, warrants support because there’s no way that we as a small privately held company could do this alone.

My next challenge then is not fighting the virus but receiving the resources to scale to a billion doses. It is making sure that we stand as tall as the large pharma companies that have been funded by the government to date. The Gates Foundation has analyzed our adenovirus approach and we are in negotiations for their support, for which we are very grateful. But this will not go far enough to manage the potential billion-dollar cost to see this all the way through and manufacture billions of doses for the world.

PE: We just wanted to touch on your role as owner of the LA Times. That puts you in a unique position to communicate the progress of your scientific work on COVID-19. How much do you feel that is your responsibility to do?

Soon-Shiong: I have no control, and rightly so, of the editorial content of the newspaper. To avoid any conflicts, it’s best that this kind of news is broken by a newspaper like The New York Times or The Wall Street Journal. I’m very aware that I have to be careful as the owner of the LA Times.

It’s important that the editorial staff follow the science rather than the specific research in which I’m involved.

Julian Upton is Pharm Exec’s European and Online Editor. He can be reached at jupton@mjhlifesciences.com.