Insights into Cancer Drug Development and Trials

0:00:01 - Announcer

You are listening to the National University Podcast.

0:00:10 - Kimberly King

Hello, I'm Kimberly King. Welcome to the National University Podcast, where we offer a holistic approach to student support, well-being and success- the whole human education. We put passion into practice by offering accessible, achievable higher education to lifelong learners. Today we are discussing cancer drug development and clinical trials. According to the National Library of Medicine, as a result of the unprecedented challenges imposed by the COVID-19 pandemic on enrollment to cancer clinical trials, there's been an urgency to identify and incorporate new solutions to mitigate these difficulties. So what they're saying is that the concept of decentralized or hybrid clinical trials has rapidly gained currency and, given that it aims to reduce patient burden, increase patient enrollment and retention and preserve the quality of life, while also increasing the efficiency of trial logistics, therefore, the clinical trial environment is moving toward remote collection and assessment of data, transitioning from the classic site-centric model to one that is more patient-centric.

Some interesting information is coming up on today's show. On today's episode, we're discussing cancer drug development and clinical trials, and joining us is National University's associate faculty instructor, Brooks Ensign. Brooks has taught at National University for 13 years and has 20 years of experience in the pharmaceutical industry. Most of his courses are biostatistics courses and he previously developed and taught clinical research courses for a clinical research degree program at National, currently supporting a clinical trial for a drug for leukemia and sponsored by a cancer drug development company, and we welcome him to today's podcast. How are you?

0:02:04 - Brooks Ensign

Thank you, Kim, delighted to be here. Thank you.

0:02:08 - Kimberly King

Excellent. Why don't you fill our audience in a little bit on your mission and your work before we get to today's show topic?

0:02:15 - Brooks Ensign

Sure, so I'm honored to be associated with National University. I actually took my very first college course from National University many years ago and did serve in the military and the Navy and am very supportive of National University's mission in providing adult education, especially to veterans and military students. My background includes teaching for National University and in the School of Health Professions, teaching biostatistics, also healthcare finance and healthcare administration courses, health economics. But yes, previously taught and helped develop courses for National in a degree program, a former degree program in clinical research, FDA regulations, regulatory affairs, et cetera. So my mission is joint. It's both teaching but also bringing in my knowledge from my business career in the pharmaceutical and medical device industries.

0:03:27 - Kimberly King

Great, fascinating. So today we are talking about cancer drug discovery and clinical trials. And so, Brooks, how did you get involved with clinical trials?

0:03:38 - Brooks Ensign

Yes. So after graduating from business school, I joined the pharmaceutical industry, not per se in a clinical trial role, I was in a business and finance role. It was actually a very interesting combination of strategic finance, mergers and acquisitions, corporate development, business development and market research. So I was very lucky to be involved with colleagues throughout the company. But in that experience, and later with other biotech and pharmaceutical companies, I learned that and, as was explained to me, really the product certainly the development stage companies are selling is not per se the pill, but it's actually the clinical data and often the companies are working in clinical trials for a future product and so the success or failure of the drug and potentially the success or failure of the company is all about the design and then execution and results of the clinical trial. So, having seen some ups and downs, disappointments and successes in clinical trials from the business side, I got more involved in understanding the nuances and complexity and then, especially in teaching biostatistics and then developing courses for National in clinical research, I learned more about it.

0:05:01 - Kimberly King

Well, that's interesting. And boy, what a time to be involved in clinical trials. What brought you back to National University?

0:05:09 - Brooks Ensign

Yes. So, as I said, I first took a college course actually when I was still in high school on a scholarship with National University in the early 80s and then, after serving as a junior officer in the Navy in the Western Pacific, I went to graduate business school, earned my MBA and joined the pharmaceutical industry and I had started to teach for other colleges. But I was really excited to learn about an opportunity at National, actually teaching mostly, but not exclusively, in the biostatistics course, undergraduate biostatistics course. One of the and one of the most rewarding aspects of that among many is that that is a prerequisite course for nursing and other healthcare administration programs. So I was able to help people succeed in that course and then move on to important things like their nursing degree or other healthcare degrees and then help them really launch their healthcare careers. So, in particular, some of these students are combat medics or corpsmen from the Navy and the Army, and the opportunity to help military students and veterans definitely resonated for me.

0:06:26 - Kimberly King

Oh, I can imagine. Let's take a go back a step and talk about what are clinical trials and why are they so important.

0:06:35 - Brooks Ensign

So clinical trials, in contrast to, say, case studies or observations or what are sometimes called natural experiments- clinical studies are designed upfront to test something prospectively, test a hypothesis. Typically you have an experimental group and a control group and that way you can have more confidence in the conclusion because it was tested upfront. So drug development has relied for decades on clinical trials and the calculations of statistical significance and clinical meaningful data to support the approval of drugs based on safety and effectiveness.

0:07:29 - Kimberly King

Got it. And there are ethical issues in clinical research, and how are those addressed?

0:07:36 - Brooks Ensign

Yes, there are important ethical issues in clinical research. So we're talking about experiments with humans, so the safety issues are very important. There's informed consent, where typically there's a long legal disclosure to make sure that the participants they're called subjects in clinical trials and clinical trials are aware of the risks. It's not considered ethical to test something that could be hazardous when there's a safer standard of care. So if there's already an appropriate and reasonable treatment for something, it creates a higher bar to introduce a new treatment if there's risks with that. Some of the interesting ethical issues that I've seen and sometimes the decision is made one way versus the other way.

So in a clinical trial with a control group in the general audience we sometimes call that the sugar pill or the placebo, but better thought of as the standard of care without the experimental drug, in order to have what we call double blind. In general, you want the patient, the subject and the investigator to not know whether the subject is getting the experimental treatment or the placebo. That's easy enough when it's a pill, but in some cases it might- For example, I dealt in the field of brain surgery and so introducing the idea of the placebo, the standard of care, would require that you have a control group, receive what's called a sham surgery, so a fake surgery such that the patient thinks he or she received the actual surgical treatment.

I heard a, and that may surprise people, might make people uncomfortable, but I heard a very passionate defense of that approach from a physician who explained that really that's necessary because that's how science has advanced, and he actually explained in an article that if it weren't for such scientific advances, we would still be using leeches and obsolete treatments from centuries ago that we now know do not work. However, sometimes that's not considered ethical. For example, I participated, or actually one of my children participated in a clinical trial where the treatment was delivered with an IV and so with children to receive a sham IV infusion was considered not ethical. So it's a judgment call that clinicians and other observers and experts make based on the risk, benefit and the ethical standards.

0:10:52 - Kimberly King

And I can imagine. Yeah, there's a lot of endless paperwork as you are entering into clinical trial. My dad was part of one. He had pancreatic cancer and it's also kind of a way to give back to the medical-

0:11:07 - Brooks Ensign

I’m sorry to hear that. Yes, exactly.

0:11:10 - Kimberly King

But so why is it critical to select an appropriate primary endpoint?

0:11:16 - Brooks Ensign

So in sports, we know that in order to score a touchdown you need to get the ball past into the end zone. But in a clinical study, success or failure needs to be defined based on exactly what are we measuring and what difference in that measurement will be considered important. And in a clinical study that probably a number of clinical variables are being measured, but selecting the one that matters the most that's called the primary endpoint, and then other things that are measured are called the secondary endpoints. But when I've been involved or I've observed the data analysis after a study and sometimes companies will say, well, it was successful with the secondary endpoint, but not the primary endpoint. Well, you have to define upfront what success is, and so it's critical to define the primary endpoint, and it's also should be something that is relevant ultimately to the doctors and the patients. That will therefore change the treatment pattern.

0:12:34 - Kimberly King

So you did mention the secondary endpoint- So what's the difference? I guess is my question about between the primary and secondary.

0:12:37 - Brooks Ensign

Sure, so if you think about like heart rate and blood pressure or something like that, so one of those might be the primary endpoint, one might be a secondary endpoint. It's more complex than that, but those are two biological measurements and there probably are going to be several biological measurements. The primary endpoint is the is the standard that the FDA is looking at for success or failure.

0:13:11 - Kimberly King

Okay, so what's the difference between statistically significant and clinically meaningful, and how are both of these terms measured?

0:13:21 - Brooks Ensign

Yes. So, and this is something that I focus on in teaching statistics, because the standard that we teach in statistics is statistical significance. Statistical significance means probably not random. So, based on the numbers, is it probably a random occurrence or is it probably not a random occurrence? And typically in our course, but also the FDA and generally in science- not always- but generally, there's a standard of odds of random occurrence less than 5% such that if you say it's statistically significant, it could still have happened randomly, but the odds of that occurring are less than 5% and therefore we feel comfortable concluding that it didn't happen randomly.

However, that doesn't necessarily mean that it's clinically meaningful. So, in order to walk into a doctor's office or I'm speaking as an industry representative or in order to make a scientific announcement, really the entry level threshold is you have statistical significance. So statistical significance, however, does not mean clinically meaningful. Clinically meaningful is will it change the doctor's protocol or therapy? So, for example, if a change in blood pressure of one point, very small, is considered statistically significant, that could be statistically significant if you have a lot of data. As I was told by a senior faculty member when I started teaching this material, tell the students that anything could be statistically significant if there's enough data. So if you have thousands and thousands of patients or data points, a very small difference could be statistically significant but may not necessarily be clinically meaningful.

Another theme I want to address in this talk is because I focused on in the title on cancer and so far I'm really talking about clinical trials in general difference between clinical trials in general and cancer clinical trials. Actually, where I'm working right now- cancer - there's urgency because patients may have a very serious prognosis, if not a risk of death. And so with cancer there's the investigators allow more flexibility and experimentation. Phase one in a normal non-cancer clinical trial is normally considered to be healthy volunteers, so they're not even people with the disease, whereas, as an oncologist friend of mine said recently, in a phase one trial in cancer the patients are really the sickest of the sick, so that's a big difference. And then there's something called the crossover study. So and I teach this in my stats class Often in cancer trials patients and their family members may really want to participate in the study because there's a chance that this experimental treatment may actually have a benefit.

But in a statistics design generally you want to have the experimental group and the control group. So that creates an ethical issue in that you're trying to help as many people as possible. So there are statistics tools such as repeated measure, ANOVA, but anyway, where you start out with group A getting the drug, group B gets the control or the placebo, say for 30 days, and then for 30 days there's a washout period to remove the effect of the drug, and then group B gets the drug and group A gets the placebo or standard of care.

And that way you can get both your ethical results and your statistics results using the statistics tool. And then another difference is in a cancer trial, a phase two trial could be pivotal for registration and approval due to the urgency of getting the drug out to the market, whereas in other clinical trials a larger phase three trial might be required.

0:18:09 - Kimberly King

Okay, oh well, thank you for explaining that. So you just kind of talked a little bit about the different types of trials, meaning that the A trial and the B trial. Is there another type of trial, or is that what you meant by the A trial?

0:18:26 - Brooks Ensign

and the B trial and the placebo, yes. So in the medical device field it's probably easier to think because they use simpler terms. They start out with a smaller trial. They call that the pilot trial. So you're testing something, maybe for the first time. There isn't a prior study, we don't have a lot of information about how the device works, etc. Small numbers, maybe one or two, probably one investigation site, and then if there's encouraging data, then they'll move to a larger, what's called pivotal. Pivotal means that it is statistically powered. It's large enough to have statistical significance for an FDA submission.

In the drug industry it's yes, it's more complicated. So the phase one trial is first in human. So the drug has not been necessarily used in humans before. So it's highly experimental and typically you're looking for safety, to make sure that there isn't toxicity. And yes, certainly in cancer we're also looking for signs of efficacy, signs of benefit. But it's a smaller study and what you're doing is basically determining whether it's worth investing and it's worth the risk of moving into a larger study to get more powerful.

By powerful I mean impressive statistics that will- This word power actually in the statistics world means that you can avoid, you have a higher chance of avoiding the conclusion that the drug didn't work when actually the drug did work. So that then. So phase two is going to be larger, often focused on and with more centers and focused on dose ranging to find the right dose, and then ideally, in phase two you have positive data and you've have selected your dose, then the gold standard that the FDA is looking for is two independent, well-controlled phase three pivotal studies. So these are the large studies that typically have hundreds or thousands of patients, depending on the indication and the specifics, in order to get FDA approval. And then phase four is after FDA approval. Often they're either required by FDA requirements for some sort of post approval, post market surveillance, or often they're supported by marketing objectives to expand the use of the drug. But yes, the stages and phases of clinical trials are an important distinction.

0:21:24 - Kimberly King

Okay, so that is important and I guess and you kind of talked a little bit about this, the considerations and selecting the study population that you said something about safety. Are there other considerations?

0:21:36 - Brooks Ensign

Yes.

So, having been both on the patient side with my son who participated in a trial, but then also more often I've been on the industry side, the investigator or the company really can't select the healthiest patient for the study.

Reason being the healthiest patient, who has the mildest condition, doesn't necessarily have the urgency and probably is adequately treated by the standard of care the existing drugs, or at least doctors are going to try that first. On the other hand, you may have to start with the sickest patients, because that's where the urgency is, the patients who have failed other therapies. However, you don't want the worst of the worst, because, or the sickest of the sickest, because in that case the patients have such a poor prognosis that it may not be a good chance to show success with your drug. So somewhere in the middle is ideal. But, for example, I can speak a little bit about my current company. We are dealing with third line relapse, refractory advanced leukemia, so very sick patients, and then the hope would be to move into earlier stage where the broader market is, with data that supports that, and so that's a typical approach in studies, certainly in cancer but other types is show benefit, say in the more serious disease, and then try to expand the market by moving into larger patient populations.

0:23:29 - Kimberly King

Okay, and that makes sense. So the issues to address in the basic study design what are those issues?

0:23:37 - Brooks Ensign

Yes, so number of patients, and that relates to power. So I tell my students that, again, power has a statistical definition about avoiding a false conclusion. But typically the power of the study is the size of the study, and so my day job in the industry is usually in finance. So power then relates to cost, because if you add more patients and more sites to the study, it becomes more and more expensive. Something that's already expensive is more expensive. Length of treatment how long will the patients, the subjects, receive the treatment? How often will they be treated? How often will they be tested and measured? And then the amount of follow-up. So some studies can look at the acute treatment and short-term follow-up. Some studies require a long-term follow-up, say of years rather than months, which then makes the study more difficult to fund, and a longer time frame. Randomization is important. How is it actually going to be double-blind to both the investigator and the patient, etc. How many subgroups, how many, how many investigation sites? So yes, these are a number of issues in study design.

0:25:05 - Kimberly King

Wow. Well, you have certainly been sharing some great interesting information. We have to take a quick break, so just stay with us. More in just a moment We'll be right back. And now back to our interview with National University's Brooks Ensign, and we are discussing cancer drug discovery and clinical trials. And, brooks, thank you for sharing the knowledge that you have, both you know really with your son's situation, and then also for teaching this and then really letting us know about clinical trials. And I guess my next question is how do we manage the randomization process?

0:25:44 - Brooks Ensign

Yes. So randomization is important because you want group A and group B, the, the experimental and the control group or, if there's several groups, to be comparable, so comparable age, comparable disease state, comparable mix of genders and ethnicity. Otherwise there's going to be retrospectively possibly and this is reported in the study to show all these various traits, including medical traits, to show that the groups were comparable. Otherwise a critic would say well, of course group A did better because they were healthier or younger or thinner or whatever, and so ideally the randomization process produces groups that are comparable. So there's different and so there's a stratification where you have different assignments. But ideally, and unfortunately no randomization or assignment process is going to be perfect, so there is a chance that group A might be younger than group B. There are statistical techniques that can control for that, as the term is used, but ideally that type of difference is avoided in the assignment process.

0:27:16 - Kimberly King

Okay, so can you briefly describe the blinding process, and how does this apply to cancer research?

0:27:23 - Brooks Ensign

Yes. So in general in studies we want both the doctor and the patient to be unaware of whether the patient is receiving the experimental drug versus the placebo. The reason is that it could impact the results. If the observer, the investigator, thinks that it's the experimental drug, might think that the results are better, and even with the patient, there's something called the placebo effect and the patient may have, say, psychological conclusions based on thinking that he or she is getting the experimental versus control drug and that may impact the variables measured. So ideally nobody knows. This may not be possible as again I talked about the ethical aspects if it's an invasive treatment and also in cancer. My experience is that in general all the patients in the beginning are actually patients with the disease, because there's an urgency of seeing how it works in that particular patient type. And so then the comparison is to what we call historical controls or literature-based controls, how investigators think that patients would do without the drug.

0:28:56 - Kimberly King

Can you talk a little bit about the sample size and how that's determined?

0:29:02 - Brooks Ensign

Yes, so there are various factors that go into that. It's critical because, again, I approach this often as a finance person- Large sample size requires more money and all this is a lot of money, but it may be an extreme amount of money and time. So the determinants of sample size really probably the number one determinant is how different do you think the experimental treatment will be versus control? If you think that it's going to be a noticeable difference with less and the other factor, another key factor is variability. So if there's less variability and there's a profound difference between the treatments, then that can probably be measured with a small sample size. On the other hand, if you're talking about lots of variability- in biology we generally think of clinical data and human data as having variability- and then a more modest what we call effect size then, that's going to require a larger sample size in order to be powered for the appropriate conclusion.

0:30:22 - Kimberly King

And is that what is considered a baseline assessment?

0:30:27 - Brooks Ensign

Well, baseline assessment also is something that needs to be done up front. So baseline assessment is understand at the beginning. What were the patient variables? So at the very end of my intro biostats class there's an example in our materials of a back pain study. But really it could be anything, but this happens to be a back pain study Group A, group B and Group C and I asked the students who's going to do best in this study, and it seems like a silly question. No one's going to know. But we do know - the people who have less back pain at the end of the study are probably going to be the people who had less back pain going into the study. So if they're healthier going into the study, they're likely going to be healthier coming out of the study. So that's the measurement at baseline in order to understand the difference, are the patients comparable, etc. Really understanding what are the patients like as we go into the study.

0:31:32 - Kimberly King

Okay, that makes sense. Interesting. How does the recruitment of study participants differ in cancer and cancer trials from other trials?

0:31:44 - Brooks Ensign

Yes, so would again talk about how, in general clinical studies, at first the phase one trials in healthy volunteers and in cancer it's patients who urgently probably want the treatment. But what I've seen on the industry side is, in order to help the drug treat the appropriate patient, you're often looking for a very specific patient profile to make sure that the odds of success for the drug to ultimately help patients is supported by the treatment of the right type of patient. So it's a difficult judgment call.

0:32:34 - Kimberly King

Okay, so every case is different, I would imagine.

0:32:39 - Brooks Ensign

Yes.

0:32:40 - Kimberly King

And what about- how survival analysis used in clinical trials?

0:32:45 - Brooks Ensign

Yes. So ultimately, as family members or as patients, the goal is survival- living longer- and so that's called overall survival, or another type of survival is progression-free survival. It may be measured in years, in which case that would be a great benefit. However, sometimes in cancer it's measured in months. One issue with survival and I think I'll talk about this a little bit more is, survival is a skewed data, skewed variable, and so there's the need to focus on the median rather than the mean for statistical considerations, because the mean is skewed by the small number of outliers who have high survival.

However, another issue is that measuring survival may take a long time, and the industry, the investigators, are trying to determine success faster than watching for a slow variable like overall survival. So that leads to what we call surrogate or proxy endpoints, something that can be measured more quickly, say a cellular or a biological assay or some kind of blood test or something, a marker. That, for example, what we call a complete response. That may, we hope, correlate with longer survival, but it's something that we can measure upfront. However, sometimes I've seen companies get in trouble by aggressively marketing their surrogate marker, when actually the patient's goal and the FDA's goal is ultimately cure or overall survival.

0:34:36 - Kimberly King

So would that be something with having issues in the data analysis? Is that where this is leading to?

0:34:43 - Brooks Ensign

If you've seen some of the statistics, so, yes, in the data analysis, for example and I was given an example by a friend who's an oncologist the survival. So, just like with income and wealth, those are positively skewed data, points, variables, excuse me. So most people with income are somewhere near the middle, but Bill Gates and Jeff Bezos, et cetera, are way out in the very high income category. Right, not many of those people, but they have a lot of income and wealth. So we call that skewed data. So when you talk about income or housing values, income generally, we talk about the median. It's better to use the median versus the mean, because the mean, the average, is skewed, it's pulled away by the outliers on the high end. So survival data is the same way in cancer.

Most patients probably have some kind of small result, but maybe a few patients have a dramatic increase in survival. Okay so, but what's more representative for the family and the patient, the new patient? You know you can hope to have one of those extremely good results, but it's probably more likely to have a typical result. So I teach this with reference to a drug where the difference in mean survivals was quite dramatic, but the difference in median survivals was modest. But what I understand from oncologists is. This is often understood and really the FDA should focus on the median rather than the mean due to the statistical issues with the mean. But the company was able to get the FDA to look at the difference in mean survival, which was much more dramatic, and therefore to get the drug approved based on that.

0:36:53 - Kimberly King

Okay, okay, interesting. So regulatory issues in cancer clinical trials versus other types of trials what are those?

0:37:04 - Brooks Ensign

So I've seen that I would say you're dealing with the urgency, the need to get things to the market faster. Oh, and also there's but there is more yeah therefore more latitude both in study design and treatment decisions by the physician, because if the patients are really at the possibly at the end of life, everybody wants to do whatever we can to help these patients. So there's more latitude. Oh, and something I should say is that a top officer at the FDA in the oncology area has encouraged this greater latitude and earlier decision making, and actually it's suggested that this has led to earlier approval, so it's been beneficial. Something that I've dealt with recently is something called the FDA's Project Optimus. Project Optimus is an FDA initiative to encourage companies, or require companies, to study more doses early on in order to select the best dose of a targeted therapy. So, for example, in the study that I've been dealing with recently, we've talked about the Project Optimus requirements to identify the dose with more careful consideration of various doses.

0:38:25 - Kimberly King

Wow, okay. I mean again, you're seeing this in real time, so that Project Optimus sounds really interesting. What problems have you seen in clinical trials personally?

0:38:34 - Brooks Ensign

Yes, I encourage students to look at a website, improvingmedicalstatistics.com, and there's other sources, but and this has examples for students everything from high school up to graduate school but it basically chronicles, unfortunately, there's mistakes, bad designs, incorrect interpretations, incorrect conclusions, incorrect comparisons, and this website chronicles these and helps… In teaching biostatistics, I try to encourage students to be encourage their critical thinking and not just listen to a conclusion but think about, what was the study design? Is this an appropriate study design, appropriate conclusion? Et cetera. And I think this website is interesting to read because there's a lot of examples of things that should have been done differently.

0:39:43 - Kimberly King

And Brooks, is this what your company is working on right now? What are you doing?

0:39:49 - Brooks Ensign

Yes, so my company. I don't wanna say too much about it for sensitivity, but my company is a public company, Aptos Biosciences, and our clinical work is in severe leukemia, acute myeloid leukemia, and our experimental drug is being used with in third line relapsed refractory patients. So they've already, unfortunately, failed other therapies. So this is really the sickest of the sick, the most urgent prognosis for leukemia patients. I actually have a friend who I had a friend, excuse me I lost three years ago to aggressive leukemia. It was about a week between her first doctor's visit and her passing, and so when I work on this study or support this study, it means a lot to think that we could do something to help such patients.

So I did have a little quotation from our company it's a phase one two open label study for relapsed refractory leukemia and in a recent announcement we are announcing dose expansion in combination with another drug, Venetoclax, and we're encouraged by our early data and there's more information on our website. But we're hoping to expand the use and expand our clinical trials in combination with other drugs and be able to treat more patients and show successful data, ultimately get the drug approved.

0:41:28 - Kimberly King

Well, I wish you the best of luck with your company. You are doing God's work there and I'm looking for that cure. So I do wish you all the best and I thank you for sharing all of your knowledge, and if you want more information, you can visit National University's website at nu.edu. And thank you so very much for your time today, Brooks.

0:41:48 - Brooks Ensign

Thank you. Thank you for the opportunity.

0:41:53 - Kimberly King

You've been listening to the National University podcast. For updates on future or past guests, visit us at nu.edu. You can also follow us on social media. Thanks for listening.