Dr. Tom Novak is the Vice President of Strategic Partnerships at Cellular Dynamics International (CDI), one of our corporate neighbors here at the Buck Institute. He received his PhD in the Division of Biology at the California Institute of Technology and did his postdoc in Immunobiology at the Yale University School of Medicine. Dr. Novak has extensive industry experience as a pharmaceutical executive. Before joining CDI, Dr. Novak was the Associate Director of CNS Molecular Biology at Wyeth-Ayerst in Princeton, the Senior Director and Head of Discovery Technologies at Roche in Palo Alto, and the Senior Vice President of R&D at Fate Therapeutics. He also serves as co-chair of Tissue Engineering and Biomaterials Section at Alliance for Regenerative Medicine.
Dr. Novak’s spoke recently at the Buck Institute about CDI’s role in large-scale induced pluripotent stem cell (iPSC) production based on the California Institute for Regenerative Medicine (CIRM) Human iPSC Initiative. The initiative is an ambitious project whose ultimate goal is to develop a BioBank of well-characterized iPSC lines. CIRM awarded CDI $16 million to create three iPSC lines for each of 3000 healthy and diseased donors (9000 cell lines in total). iPSC lines will be derived from donated tissue samples obtained from healthy people and patients suffering from Alzheimer’s disease, cardiovascular diseases, neurodevelopmental disabilities, autism spectrum disorders, liver diseases, eye diseases and respiratory diseases. In addition, CIRM also awarded the Coriell Institute close to $10 million to set up a repository to bank and distribute the iPSC lines, of which CDI was the primary subcontractor. Dr. Novak described the workflow of the Human iPSC Initiative from “tissue collectors (sample collecting)” to “derivers (making iPSCs)” and to “repository (storing iPSCs)”. He also discussed the challenges in each area and the potential for collaborations with scientists at the Buck.
Q: The methods used to reprogram iPSCs require the use of genes that also play roles in cancer development. What is the FDA’s attitude toward iPSCs as a therapeutic?
TN: The FDA often finds itself between a rock and a hard place when it comes to evaluating new therapeutic modalities like stem-cell-derived cellular therapies. They are accused of unconscionable delays by patients desperate for hope and excoriated (and accused of being in Big Pharma’s pocket) when they approve therapies that are then quickly withdrawn from the market due to unanticipated side effects. When it comes to stem cells, I think the FDA has been less innovative and aggressive than Japan, Korea and China. Part of the FDA’s caution is due to the nature of the genes used in reprogramming, but it’s mostly due to the newness of the approach and the small number of patients that have been treated. In Asia, it is a lot easier to move from preclinical to clinical for iPSCs. In fact, the first iPSC clinical trial for macular degeneration began in Japan last year. We are probably five years from that happening in the US.
I think the challenge in getting regulatory approval is that neither they nor the company who developed the technology know what an adequate safety package is. And that highlights a key point about drug development—cost in industry is not as important as uncertainty. If you tell me that your clinical trial is going to cost $100 million, but if we do A, B and C we will get approval, that’s fine, I can spend the money. But if you tell me that there is no certainty for approval, I am not going to invest it, even it only costs me 1/5 of that price. So there are lots of things that need to be learned.
As more and more cells get put into people, we will be able to see what the safety risk is. The FDA is certainly willing to talk to industry and people in the field; for example, as Brian Kennedy mentioned, I am a member of the Alliance for Regenerative Medicine lobbying group. We spend a lot of time talking to the FDA, and they have been very receptive to the issues we’ve raised. They want this technology to move forward, but there are a lot of unknowns at this point. Regardless of how one generates iPSCs, the cells proliferate, but once they are done differentiating they stop dividing. So a key question is, how can you define the residual undifferentiated population in a therapeutic pool? Because you can never say there are zero undifferentiated cells. We will get more and more comfortable with that number as more differentiated cells get to put into people and no tumors arise. And that will come in time.
Q: What technologies does CDI have to conduct high-throughput iPSC reprogramming?
TN: There is a heavy reliance on automation. But there are some trade-offs. For example, we told tissue collectors that we want a total of 48 mL of blood: 40 mL for reprogramming and 8 mL for infectious disease testing. A lot of the tissue collectors would say, “I can reprogram from 1 mL of blood!” We can too. But when you are doing 3000 samples, it’s difficult to go back to the donors and collect the samples again since the donors have a relationship with the tissue collectors, not with CDI. So our argument is, we need a certain amount of the starting material so we can do a couple of repeats in case some samples fail to survive. As I said in my talk, if you are doing this manually, you can pick six clones from one donor, look at them every day and decide: “this one needs to be fed” or “this one is over growing”. But we can’t handle them manually; we have to do it on a robot, which means “you all are going to be split on Wednesday whether you like it or not.” You will lose some cells, and that’s why we have to get enough starting material. The automation gives you the ability to run through a lot more samples in the same amount of time. But you probably take a hit in the efficiency, and your failure rate will go up. Our current manual failure rate is about 10%. But we have no idea what the failure rate is going to be when everything is fully automated. We certainly hope it’s no higher than our manual procedure.
Q: Will automation be the future for large scale iPSC reprogramming in the industry?
TN: I think the future is going to be large-scale differentiation. And the question is what your starting material is going to be. Is it going to be autologous or allogeneic? Autologous means taking your cells and putting them back into your own body. They are genetically matched with you, and your body is not going to reject them. Allogeneic means making iPSCs from my cells and then injecting them into you. So your body is going to reject my cells if you and I are not closely related. Even if you and I are closely related, we are not identical, so you will likely be on immunosuppressant drugs for the rest of your life.
We have made some GMP (Good Manufacturing Practices) lines that are called HLA (Human Leukocyte Antigen) super donors. The histocompatibility loci that govern rejection are very polymorphic because we are all different. And we are almost always heterozygous for a particular HLA gene–that is we usually express two different alleles. But we found individuals that are homozygous for all three key HLA alleles that govern rejection, so in this case you only have to match three genes instead of six genes. These so-called “super donor” iPSC lines can be used to reduce immunogenicity. So, if those lines worked, you may need 200 lines for the US populations, 90-100 lines for Japan, and those will be the lines that everything will be made from. Therefore, you would not have to use autologous lines, which are a challenge because it means that every batch is a product. That is not a system that Pharma companies like. Pharma likes to make a big batch, test a small percentage of it and ship them out so everyone can use them. But if it’s autologous, every patient needs his or her own batch and product, which gets expensive and time consuming. The goal is to make as many doses as possible from a differentiation run. At least, that’s how Pharma is thinking now because it matches how they produce small molecule drugs and biologicals. Some people think the future is going to be the use of autologous iPSCs. I think the future is going to be large scale of differentiation and purification of allogeneic iPSCs, and we know we can do that well.
Q: Could you discuss any collaborations between CDI and scientists at the Buck Institute?
TN: We don’t have any yet. And the reason is our lab here is only used to do the CIRM project. We are not allowed to use it for anything else even if there are no samples coming in for two months. What we have been starting to do is talking to Buck scientists who study diseases for which we make cells that might be therapeutically useful. For example, we make dopaminergic neurons from Parkinson’s. So I have been talking to Xianmin Zeng and Julie Andersen since they have Parkinson’s model systems up and running. Can we perhaps collaborate to test our cells in their model systems? We make differentiated cells very reproducibly, but we don’t know if these cells have the right specifications to provide a therapeutic benefit. This it is not our focus right now. Our focus is producing and selling cells to Pharma companies for use in drug safety screening. So we are starting to have those conversations about evaluating our cells in efficacy models.
And it has been challenge of course: our cells are expensive and academic grants are tight. There are a lot of experiments that your labs would like to do, but we need to get paid and labs don’t always have the money. So we are trying to figure out a collaboration model that is going to benefit both of us, where we get data that we won’t obtain otherwise, and you get a paper and a grant. As for our long-term plans in Novato, we don’t know what happens when the CIRM grant ends (late 2017). When it ends, it is possible that the Buck says, “Sorry CDI, we have faculty with lots of grant money and people and we want the space back”, and we leave. But what might happen instead is that we would want to remain at the Buck and do CDI, rather than CIRM, work. We hope the Buck will be happy to have us remain here because it’s a source of income for the Institute. I don’t think the plan will be to move everyone back to Madison because most of the people wouldn’t go, and the robot took months to set up. So the hope is we stay for the long term. Right now we have two and half more years on this grant. By the end of 2017 we need to have all 9000 iPSC lines in the bank, and that’s all I’m focusing on at this time.
Q: What brought you to CDI?
TN: That is a long story. It is a very typical “Pharma” story, by which I mean pharmaceutical companies change all the time. In 2009, I was at Roche Palo Alto and was asked to lead a global team to evaluate using stem cells as a research tool for safety and screening. We actually did the first deal with CDI in 2009. So I knew the CDI folks since 2009 because the safety group in Palo Alto wanted to use their cardiomyocytes. In 2010, Roche shut down the Palo Alto site, so I joined a stem cell company in San Diego called Fate Therapeutics as their head of R&D. I got laid off there a year later because it was a start-up and they were downsizing. When that happened, CDI called me immediately and said, “We’d like you to join CDI.” They wanted me to join them after Roche Palo Alto closed but they wanted my wife (also a scientist) and me to move to Madison, and we didn’t want to leave California. This time they said, “we want you to stay in California as our point person and help us get money from CIRM.” I took the offer because I didn’t have to move. I love Madison (as a midwest cities go), but I don’t want to leave California if I don’t have to. CDI has known me since I was in Roche. I guess I made a good impression on them!
Q: You have been at a large company like Roche and small start-up like CDI. Could you give some pros and cons for each of those?
TN: The advantage of a large company is they usually have a lot of revenue. The trade-off used to be that if you work for a large company, you take job security and you give up the chance to be a millionaire because there are not a lot of stock options being handed out. In a start-up, you took a lot of risk, but you also had a possibility to get rich because you had options that could be worth a fortune. That has changed over the last 20 years. Big companies are not safe anymore, and small companies aren’t getting the valuations they did in the past. So there is no guarantee for lifetime employment.
One challenge working in big companies is that it takes long time to get a decision made. In small companies, you get a meeting with three people, 10 minutes to make a decision, and that’s it, you’re done. The down side of small companies is their funding is from venture capital, and VCs want their money back at some point. So you want to get something that is attractive to the investors. What is the mechanism to get their money back? IPO (initial public offering, i.e. “going public”), acquisition by a Pharma, or a large deal with a Pharma that brings in a lot of cash. So small companies can be risky because they are small and they don’t have the revenue that a large company has. Another thing is, large companies move slowly, and sometimes it is hard to know what your contribution is. So you have to decide.
Personally I don’t think I would work for a VC funded company again. CDI is different because it’s funded by high-net-worth individuals that worked together at a previous company (NimbleGen). In a VC funded company, every two months the board comes to have a meeting with you. So every two months you have to explain to them what you do and why you are doing it. The constant changes can be very frustrating. But in small companies, you can make decisions quickly. You don’t have people sitting around with their feet on the desk. There is something to be said about uncertainty focusing the mind. I’m actually thankful that I have worked in big and small companies as both offer useful lessons in what to do and what not to do.
Q: What skills do you think our postdoc should get in order to succeed in industry?
TN: That’s an excellent question. It’s unfortunately true that it’s getting more and more difficult for postdocs to find jobs commensurate with their skills and training. Regardless of the job market, the one thing that will make you stand out to a hiring manager is a solid record of achievement (papers, fellowships, grants) and demonstrated excellence in some area of science. (Also, if you don’t get strong recommendations from your graduate and post-doctoral advisors, you’ll be in trouble!) That may be more important than your actual field of study. Obviously an oncology group, for example, will prefer to hire people with expertise in that area; however, they’d also consider candidates with a strong background in cell biology, signal transduction, regulation of gene expression, apoptosis, etc. People in industry move around all the time, and I expect that to continue. So being flexible is crucial.
It’s also true that drugs are discovered and developed by teams. Being able to work productively in a large group is a necessity. As for specific skills, it’s hard to say because science changes so fast. However, I think a facility with statistics, especially the ability to manipulate and extract actionable information from large datasets will become even more important in the future. Once you have a job, you’ll be amazed at how little you know (if you’re an American, like me). Pharmacology, physiology, and anatomy become as important as cell and molecular biology. Once compounds advance to the clinic, other disciplines predominate: regulatory affairs, clinical pharmacology, drug safety, clinical operations, and a whole lot more. Very few people get advanced training in these disciplines (they are usually learned on the job), so there are endless possibilities to switch focuses in industry. Learning some pharmacology and human physiology never hurt, and having experience handling animals (rodents, primarily) is also a plus. I may be biased but I think having human stem cell experience will also be a sought-after trait for the next 5-10 years.