Jonas Hannestad, MD, PhD, is the CMO and Head of R&D at Tranquis Therapeutics.
Can you first talk about your role as the CMO of Tranquis Therapeutics?
My role here at Tranquis has three main parts. One of them is sort of unusual for a CMO: I oversee discovery and research, which is typically more of what a CSO does. I lead the research group, which is responsible for synthesizing and screening new molecules, testing these compounds in pharmacological assays, conducting studies in animal models, and developing translational biomarkers. Understanding the mechanism of action and testing our hypothesis in animal models of disease helps us decide what potential diseases these mechanisms may be useful for, other than the initial ones we targeted. The two other roles are more traditional CMO roles. One is overseeing and leading the clinical development group and one is investor-facing.
How is the way you design trials impacted by your role in research?
Having these three roles can be challenging at times, but overseeing our research activities gives me a deep understanding of the science, which is helpful for various reasons. Before I started doing clinical research, my background was in cell and molecular biology. It is very helpful to have that understanding of the research because not only am I part of the decision about which compounds we take into the clinic, but also with regard to the development of biomarkers. Developing those biomarkers has to start early, ideally before you can get into phase one, so that they are ready for a phase two trial. A lot of that work is done either by the research group or in collaboration with the research group, because they understand how the molecules work on a molecular level, and what downstream effects you can measure and how that translates to humans.
The other other reason it is helpful to be involved so deeply in the research is that when I present to investors, I talk not only about the clinical data and our clinical plans, but also explain why we think this specific mechanism may be helpful for a disease like ALS.
What is your advice around designing a clinical trial when it comes to endpoints?
When I think about clinical trials in patients, I put them into two categories. (Phase one trials are separate as they are usually done in healthy volunteers to evaluate safety and pharmacokinetics.). One is the early, proof-of-concept phase. For example, a phase two trial where you are trying to understand what your drug does to human biology in the patient population. The second category is the phase three or pivotal program with a very specific eye towards regulatory approval. In the early phase, proof-of-concept trial, you are trying to prove that a molecule you designed to modulate a certain mechanism can actually modulate that mechanism and have an effect in humans at the dose you have chosen. That is the box you need to check. You should not focus on the disease itself but rather on the mechanism, to show that when you give x milligrams in this patient population, something related to the mechanism, like target engagement, occurs to show your drug is working on that molecular pathway.
Can you explain more about pivotal trials?
Once you have established that in the early phase, then you can move to pivotal trials. If your initial target hypothesis is correct, then modulating this target or mechanism would have a benefit on the disease process. For later trials, which can sometimes be phase two but certainly by phase three, it is all about the approvable endpoint. Even if you are working in a disease with already approved therapies and agreed-upon endpoints, you have to speak to regulators early and work with them around designing a trial with a certain sample size, duration and so forth, with the statistical power to detect the approvable endpoint. If it is a disease where nobody has received approval before, you may have the more complicated task of figuring out what the endpoint should be based on what is important for patients and what regulators would want. They are not always a perfect match and you need to meet both of those criteria.
How would you advise someone to best collaborate with statisticians and data management around trial design?
In my experience, when the clinical team has a good understanding of the disease, patient population and clinical trial design, they will sketch out a trial design with inclusion and exclusion criteria, duration, arms, randomization, endpoints, et cetera. Then they will go to the statistician and ask them to help figure out how many patients are needed to statistically power the trial. I have learned, by making the mistake myself, that it is much better if you involve the statistician early because they can provide a lot of helpful input. There are certain things with regards to statistics and data management that you don't think about when you design a trial. It is also important to involve the data management person early on. It is very tempting to collect as much data as you can on your case report forms. But people don’t always realize that you should only capture those data if they will be useful for the eventual analysis. Having tons of data is not necessarily helpful. That early involvement with clinical data management to design CRFs in the clinical database, and then also from the statistician to think about what those primary and secondary endpoints are, is extremely important.
"When you design a trial on paper, it’s tempting to put everything in there. But you have to think about the actual patients with a disease who need to come to the clinical trial site and do all these procedures and assessments."
Do you have any examples in your own career around this?
Earlier in my career, my inclination was always to get more data because it may be useful and it can’t hurt to get it. It’s like when you go through your garage and it’s hard to throw things out because you think you may need them someday even though you probably won’t. That is how many people think about capturing clinical trial data. But over the last decade, working closely with statisticians and data managers, I realized that you need to think carefully about every single aspect of the trial. If you can’t explain why a certain item on the CRF is helpful, you should consider leaving it out (excluding certain safety assessments or items required from a regulatory perspective).
The same is true, more broadly, when it comes to trial design. Is the procedure necessary? Is it burdensome to patients? When you design a trial on paper, it’s tempting to put everything in there. But you have to think about the actual patients with a disease who need to come to the clinical trial site and do all these procedures and assessments. That can be burdensome. Do you actually need everything you’re putting into the protocol?
How can CMOs best communicate with statisticians and avoid misunderstandings?
Some 15 years ago, I was working with a statistician on analyzing clinical data that I had generated in an academic setting. I learned that unless you have a very strong statistics background – and most CMOs don't – you do kind of speak a different language. Even now, 15+ years later, I still sometimes have difficulty understanding what a statistician can do and the different types of analysis they use. And sometimes it’s the other way around and I am trying to explain what symptoms a scale measures and what correlates with a reduction or improvement in symptoms and how to translate that clinical meaningfulness into a statistical test. That two-way communication can be difficult but you need to talk to them often and go over it with them over and over until you both understand each other’s point of view.
I have found that statisticians who have worked in clinical trials are easier to talk to because they are able to break it down while new statisticians or people who have only worked in the field for a few years may really know their statistics well but have a hard time translating that to people who don’t have that background.
How do you translate those elements of clinical trial design to your CFO, Head of IR or investors?
I find that part actually easier in some ways then the statisticians and data capture and so forth. Explaining to an investor who does not have a deep background in the disease is fairly easy because diseases affect human beings and you can explain how it affects their lives. You can measure symptoms. A lay audience is able to relate to clinical signs, diseases and endpoints. Biomarkers are a bit more difficult because you need to explain the molecular mechanism and how that relates to the compound.
"If you and your statistician have a good estimate of what is needed in terms of size and duration, you can go to the finance people or CEO or investors and explain why you need this much money to do this trial."
How do you balance speed, cost and strength of a clinical trial?
Often people have unrealistic expectations about what trial you can get for a certain amount of money and time. Everybody wants a trial to be less expensive and happen quicker, but without sacrificing robust readouts. Perfect is the enemy of good. A perfect trial may be 5,000 patients for five years but that is not realistic in most cases. You need to figure out the sweet spot where the trial is large and long enough that you are likely to detect the signal if there is a true effect of the drug but not so large and long that it becomes a waste of money, resources and patients’ time. Patients are offering up their time, often without benefits, for you to run your trial. If you and your statistician have a good estimate of what is needed in terms of size and duration, you can go to the finance people or CEO or investors and explain why you need this much money to do this trial.
How do you know when to toe the line in terms of fitting into the budget and when to say you need more money?
It depends on what kind of relationship you have with those decision-makers. It is never black and white. A smaller trial could work but it is not as likely to provide a clear answer. You have to explain that and say that you recommend putting more money into this to do it better and express your concern. But then you will move ahead based on the decision made even if you don’t completely agree – assuming there aren’t concerns around patient risk, of course.
Do you have an easier time communicating this because your CEO has an MD background?
At Tranquis, yes. At other jobs, when my CEO had little clinical experience it was more difficult to explain. And with the board of directors, it’s easier to explain why a clinical trial should be a certain way to the members with scientific backgrounds as opposed to those with commercial or finance backgrounds, especially when there's that pull between budget available and what the trial should look like.
"It’s like when you go through your garage and it’s hard to throw things out because you think you may need them someday even though you probably won’t. That is how many people think about capturing clinical trial data."
What surprised you most about the CMO role?
When I interviewed the CEO before taking this position at Tranquis, he proposed that I do the clinical development part, the investor facing part and also the research part, which is unusual for a CMO. I was excited because I like basic research, and that’s how I started my career. But I don't think I realized then how time consuming and challenging it would be to stay on top of all three. If I had known, I think I still would have said yes, but I might have structured things a little differently so that I could balance them better.
