Clinical trials demystified, part 2: Dr. Kurzrock (2021)

Julie Clauer 0:00
Hi everyone. I’m Julie Clauer, Manju George, Adrian Terek, and I have been putting this session together because it’s part of our desire to demystify clinical trials. And one of the things that really needs to be demystified, particularly for people like me who are not scientists, is you look at a trial, you see results, and you’re like, What the heck does this mean? And so that’s what this session is about. It’s really around understanding trial results and when they’re posted or presented. And it’s really more general. It’s not specific to a specific trial. It’s more generally the kinds of things that are reported. And we’re using examples to bring that up. So Dr. Kurzrock is an amazing researcher, but she’s also very good at explaining things, as you’ve seen in previous talks she’s given us, and so she is going to be doing most of the talk, kind of talking us through the different parts of trial results. But first what we wanted to do is go over just some fundamentals, because all of us are coming in with different levels of understanding. So Adrian is going to take us through just some of the basics of terminology and things, so that when Dr.Kurzrock joins us, she will be able to go into the results more specifically. In terms of questions, we’ll be doing questions off of the chat. So if you have questions, please put them in the chat, and at the end, I think Manju will be the navigator of those and be able to ask those questions of Dr. Kurzrock. So I think that’s it, but I’m very excited, because anything that helps me understand this better is excellent for me, so hopefully everyone else feels the same way. Okay. Adrian, take it away.

Adrian Terek 1:46
Thanks for the wonderful introduction, and thanks everyone for joining us here today. So as Julie mentioned, I’ll just be doing a very brief overview of clinical trials themselves, some of the terminology, and then Dr. Kurzrock will dig into the more interesting stuff. So, what are clinical trials? Clinical trials are essentially the process in which new treatments are tested and they go through a certain set of phases. So in Phase I, what they’re really trying to do is they’re learning about the correct dose, and they’re really looking heavily at safety. They don’t want to expose people to treatments that cause them undue harm when we don’t know if that treatment actually has a benefit yet, and so they’ll often really focus on that in a Phase I trial. But they do still track efficacy, even in those trials. It’s just that they’re not quite statistically powered enough to give us a good reading, as opposed to just an idea of what we might see in the future. From there, they extend it to a larger Phase II trial, and that’s where they really design the trial to test the efficacy, and if, from there, the trial’s efficacy is still looking good, then they’ll run a much larger Phase III trial, where they generally are comparing it against an already approved, established treatment to try and see if it’s either better than that treatment or safer than that treatment. And so, you know, all treatments will go through that process of checking for safety and getting dose found, then starting to look at efficacy, and then finally, a large trial to prove that it’s better than what’s already out there.

Adrian Terek
And when it comes to finding trials, there’s a lot of good resources that you can use. Clinicaltrials.gov, of course, is the government mandated trial website that contains every trial in America. There are some different research groups that help track specific types of trials. So, for instance, the Cancer Research Organization does track immunotherapy specific trials. Your local cancer center, you can always use your local doctor as a great resource. So just ask them, what kind of trials are being run in this center here and do you think any of them would be a good fit for me? And then, of course, patient organizations. I’m a big fan. Colontown, of course, is a striking organization where there’s so many ways to learn about really promising research and trials. And a couple of things to note is that when you’re looking at which clinical trials to participate in, one of the primary things you should be considering is what’s known so far. So if you’re looking at trials, you generally want to look at ones that have already got some data behind them. It’s usually going to be a better roll of the dice if you find a trial where they’ve already done some of this research, and you have a rough idea what to expect in terms of safety and in terms of efficacy.

Adrian Terek
So clinical trials, when they’re testing these things, they have actually structured criteria. So they’re not just saying this drug works and this one doesn’t, but they’ve really got a set of criteria that’s been worked on over the years to be able to look at what’s going on with the treatment in a trial. And the primary one they use is for looking at the size reductions that happen in cancer treatments is RECIST 1.1. It’s gone through a lot of changes over the years. That’s why we’re at 1.1. What they’re essentially looking at is how a treatment affects the amount of tumor that someone has in their body. And so the way it works is you’ll get into the trial, and they’ll do what’s called a baseline scan. They will look at the cancer in the patient’s body, and they’ll select a couple of targeted locations. So usually about two per organ, and they will measure those specific tumors, give the patients the treatment, and then a few months later, they will take another scan and measure those same tumors just to see what’s happened to them. And so they then break it down into four different criteria. So the first is what’s called a response, and the subset of that is two different types of responses. One is a partial response, meaning that the targeted tumors shrunk by at least 30% but didn’t quite go away. And then a complete response is when all of the tumors on the scan have completely disappeared. And so whenever you hear the word response, just think the cancer shrank by at least 30% and if they talk about a complete response, it means that, according to the scans, the cancer is gone.

Adrian Terek
The next set of criteria is stable disease. So this is their way of judging that the treatment may have affected the cancer, made it to grow a lot slower, or maybe shrunk it a little bit, but not quite enough to be classified. And so they view stable disease as when the targeted tumors that they’re measuring didn’t quite shrink by 30% or even grew a little bit in between the start of treatment and that follow up scan, but didn’t quite grow past the 20% mark. So the assumption there is that if, over that period of time, the tumor only grew 20% then maybe the treatment really slowed down the growth of that tumor, which can still have a benefit on how people feel and how long they survive. And then finally, progressive disease is when the tumors have changed by greater than 20% so they’ve grown more than 20% from the baseline scan, and usually that’s an indication that the treatment is not really effective. It’s not doing what it was hoped it would do. And so they basically look at these baseline scans, compare it to the new one, and classify what happened to each patient who was given this treatment. And then they report the data and aggregate in clinical trial reports in various presentations and conferences and in journals.

Adrian Terek
And then just some definitions of these measurements. It’s usually reported as a rate. And so what they mean by that is the proportion of patients who are given a treatment whose tumors had a response or were shrunk by greater than 30% and so the proportion is the key. So what they do is they look at, okay, if we give 100 people this treatment, how many of them have their cancer shrink? And if 40 out of 100 had their cancer shrink, then they would say it has an overall response rate of 40% and they can use that data and kind of extrapolate what the likely outcomes are for patients in the future. Next is disease control rate. So this is a combination of all the patients who responded, plus the patients who had stable disease. So basically it’s adding together the patients whose cancer shrunk with the patients whose cancer didn’t quite shrink enough to be a response, but also didn’t grow enough to suggest the treatment wasn’t doing anything at all. And when you add those together, the idea is that it gives you an estimate on how many people get at least some sort of benefit from this treatment. Duration of response is another thing you’ll see quite often, DOR, and what they’re looking at there is the second metric when cancer shrinks that people care about is, of course, how long does it stay shrunk for? If I have a patient who is given a treatment and their cancer shrinks 80% but it starts growing back immediately, and it’s back to its original size in two months, it may not be having as much of an effect, a positive effect, for that patient. Whereas you have some of these amazing treatments that can shrink someone’s cancer for months and even years, and so at that point, it really starts to get very exciting. And so they also track how long the average response is.

Adrian Terek
And then finally, we have the survival metrics. So these are the two ones that doctors really pay attention to the most, most of the time. So progression free survival is essentially how long after a patient starts a treatment until they either pass away or their cancer starts to get worse. And getting worse, again that is the definition of progressive disease. So the cancer grew by at least 20%. So what they look at here is, if you give a 100 patients Treatment A and on average, they don’t count as their disease progressing or getting larger for eight months, then it would have a progression free survival average of eight months, meaning that half the patients did worse than that, and half the patients did better. And so progression free survival is important because it gives doctors an estimation on how long a treatment can work for. And then finally we have overall survival, which is simply how long did the patient survive. And so if you give someone a treatment and they stopped the treatment after six months, and then they pass away two years later, then the overall survival for that treatment was actually two and a half years, not just the six months. The whole goal of a treatment is to keep people alive for longer and just because a patient started to progress on a treatment doesn’t mean that they’re going to pass away a week later. They can still be reaping the benefits of that treatment for long periods of time afterwards. And so these are sort of the five main metrics that doctors look at when they’re judging how well a treatment works. And the two they’re primarily going to look at is progression free survival and overall survival, because that lets them know, on average, what they can expect to happen when a patient is put on that treatment. They can look at a PFS of nine months and know that, on average, if a patient gets given that treatment, their cancer will stop getting worse for around nine months. Overall survival, again, if it’s three years, they know that, on average, if they give a patient this treatment, they’ll survive for that three years. Now, one key consideration is, I’m saying on average quite often, but it’s not the average that most people think of. It’s a different kind of average called a median. And what a median is is essentially the middle number. So if I look at 100 patients who are given a treatment, and half of them pass away in under six months, and half of them survive longer than six months, then the median is that middle number. And so I always try and view it very simply as half did worse, half did better. So whenever you see in a clinical trial where they say that median overall survival was 13 months, don’t take that to mean that if you go on that treatment, you’ll only survive 13 months. What it means is actually that half the patients survived less than that. But of course, very importantly, it means that half the patients survived longer, and sometimes they survive significantly longer. So it’s really important to not look at these numbers and say that’s me, because it’s essentially a range that you should consider.

Julie Clauer 12:50
Adrian, how does it work with overall survival if you’re on a different treatment after the trial?

Adrian Terek 12:55
So technically, that’s taken into account. All that happens is when you join a trial, they will continue to give you the treatment until your disease progresses, in which case they assume that the treatment is no longer beneficial, and they’ll just follow your timeline until eventually you can end up passing away from the disease, and in between you stopping that trial and passing away, you might do three more treatments but the assumption is that it’ll be an even distribution across the patient cohort. So what I mean by that is, if 100 people joined a trial, and all of those patients leave the trial, because at some point the treatment stops working for them, and all of them then go on to a secondary treatment, they’re all going to go on to a variety of different treatments, and it’ll all kind of even out, meaning that the assumption that they see at the very end from overall survival, because all the other factors will even out, the difference must be because of that treatment that they got initially.

Manju George 14:06
Dr. Kruzrock has just joined.

Dr. Kruzrock 14:09
Hi, how are you?

Manju George 14:11
We are good. Hi, so nice to have you. Dr. Kruzrock, do you want to give a very brief introduction about yourself, and then we can start with this slide that is shown?

Dr. Kruzrock 14:20
Sure. So my name is Dr. Razelle Kurzrock. I’m a medical oncologist with expertise in clinical trials and in precision medicine. I’ve worked as a department chair for an early phase clinical trials department at MD Anderson, probably one of the largest departments of its type in the world. And then I’ve worked at University of California, San Diego, heading up the precision medicine program. And currently I’m Chief Medical Officer for the Worldwide Innovative Network for Personalized Cancer Therapy. And so it’s a pleasure to speak to you.

Manju George 15:07
Okay, thank you. So then we can start with this slide.

Dr. Kruzrock 15:10
So clinical trial data, I think you can see it. What we try to put into our clinical trials is a trial schema, which is just a plan, the patient characteristics, which is, what kind of patients, how old, what kind of disease they might have, what kind of genomic findings they might have, adverse events, any side effects, and a table of results. There will be a discussion of results too, but tables often give you that information in a shorter form. The figure of the results. Figures are often very helpful, graphics, in understanding and then future directions, conclusions, what we learned and where we want to go in the future. This is the ANCHOR trial, which is a Phase II study in first line BRAF mutant colorectal cancer. And I think first of all, why was this study made possible in the first place? Or why was there interest in this study? One is because a subset of colorectal patients, it’s a minority, but a significant minority, have BRAF mutations. Patients with this mutation do not do well on classic chemotherapy, but we now have drugs that can attack the BRAF mutation. These drugs can be very effective in other cancers, and it’s been a little bit more challenging to get them to be effective in colorectal cancer. And one of the reasons we think it’s been more challenging in colorectal cancer is because there may be more than one thing going on, and so this is why this study was set up this way. And what you see is that the patients are actually getting three drugs. They’re getting a drug called encorafenib, which can directly affect the BRAF mutation, another drug called binimetinib, which can indirectly affect the BRAF mutation, and a third drug, cetuximab, which is an EGFR inhibitor, which has been approved for a long time for colorectal cancer. And the reason that’s added in is because it’s believed that in colorectal cancer, when you attack the BRAF, the tumor tries to recover, and one of the ways it tries to recover is by over-activating the EGFR pathway, and that’s why the cetuximab was added in, to interfere with the feedback loop that may be still driving the cancer. 95 patients have been enrolled. The objective was to look at response rate, progression free survival, overall survival, safety, PK, and so forth. And I think we can go on to the next slide. And by the way, I’m happy to answer your questions if there are questions from anybody, or I don’t know if you want to wait till the end.

Manju George 18:29
Yeah. So basically, with these slides, what we want is for you to tell us, when you see a trial schema like this, what are the points that as patients, what we should think about, for each of the sections, like of the different parts of the clinical trial. Okay, next slide.

Dr. Kruzrock 18:46
Yeah. So, if we go back to the last slide, I think the thing that you should be thinking about is, why did they choose these drugs? And I tried to outline that.

Manju George 18:58
Okay.

Dr. Kruzrock 18:59
Encorafenib is the BRAF inhibitor, but the other two work on the pathways, but indirectly, and that’s why they chose those drugs. This is the BREAKWATER study, and this also provides a schema. You can see the schemas are a little bit complicated, but this is another way to look at some of these questions in the BRAF mutated patients. So they are also looking at encorafenib, that BRAF inhibitor, but they’re giving it just with cetuximab, which is the EGFR inhibitor that will impede the feedback loop that comes when you try to cut off the BRAF. And then they’re looking at it with and without chemotherapy. And so there are different arms. They’re looking at FOLFIRI, which is a classic chemotherapy, FOLFOX, another classic chemotherapy, then randomizing to different arms. And so I think the main question would be, again, why are they using these drugs? It is simply they’re trying to figure out which of the regimens may be more powerful. So we know that the chemotherapy works in colorectal cancer, but doesn’t work for everybody. And then the questions are, if you add these targeted drugs to patients that have BRAF mutations that normally would not do well with any of these chemotherapy regimens, how would they do with the chemotherapy and the targeted drugs together.

Manju George 20:45
And then I was wondering about the safety lead in, when we see that, what are the things that we should keep in mind?

Dr. Kruzrock 20:50
Well, I think for the safety lead in, the reason for that is because they’re combining a lot of drugs together, so they’re giving classic chemotherapy, but now they’re adding in two targeted drugs, the encorafenib and the cetuximab, and that’s true for FOLFIRI and that’s true for FOLFOX. So they want to know if they can do that safely, and then, will they have to reduce doses, or is there any danger signals for doing that? So they want to put together a very comprehensive regimen for colorectal cancer, but especially when you combine it with chemotherapy, you want to make sure that that regimen is not too toxic for the patients to tolerate. And there’s a lot of drugs here put together, so that’s why they have the safety lead in.

Manju George 21:42
This from a different trial, but a table just showing how the patient characteristics are depicted.

Dr. Kruzrock 21:46
So this table, they give you the basic characteristics of the patients, and so they give you the age, and you want to get a feel if this study is an unusual study in some way, or does it reflect colorectal patients? So the median age here is 62 and I think that really does reflect colorectal patients. And then there’s a broad range anywhere from 28 to 83 so they included quite old people. And then, unfortunately, we know that younger people, for reasons that we don’t know yet, are getting colorectal cancer. So the youngest patient is 28. This is something we really didn’t see much, going back, not that we never saw it, but it was really unusual to see people in their 20’s and 30’s with colorectal cancer and now we see it more and more frequently, so I think the study is reflecting that. Then we’re looking at the number of women. I would say that I’m a little worried about this study because there’s not very many women in it. There’s 31 patients and four women. You know, one thing I’d be asking is, why is that? Why was the bias so strong towards men?

Dr. Kruzrock
ECOG, performance status, PS is a performance status. And so what performance status does is it is a standardized scale that tells you very quickly how well the patients are doing. And the lower the number, the better the patients are doing. So in other words, if the number is zero, the patient is perfect. They don’t know anything is wrong with them. And if the number is five, they’re dead and everything in between is gradations, and these are standardized. Zero and one is definitely considered good performance status. So a lot of these patients were one. I’d actually like to know how many were zero as well, but I’m going to assume maybe that the others were all zero, and then the next thing they look at is PD-L1, and they’re looking at a combined score where you look at the immune cells and the tumor cells, and the PD-L1 is a measure of prediction for immunotherapy, and so anything above 1% is considered positive. And so you see that 22 out of the 31 patients had some prediction for immunotherapy, and then they looked at the number of prior lines of therapy and how many patients had various prior lines. So what you should know, in general, is the more the prior lines of therapy, the harder it is to get responses. So in a trial that has lots of prior lines of therapy for most patients, you have to take that in consideration when you’re looking at a response rate. The response rate may be compromised by how difficult the patients are with lots of prior therapy. And so you might think that if the trial has a decent response rate, if you brought it in even earlier in the disease, it might be a really substantial response rate.

Dr. Kruzrock
Then disposition means are the patients continuing on treatment or not? And it gives you the numbers for each, continuing on treatment, not continuing on treatment, and then gives you some of the reasons that they may have discontinued. They may have discontinued because the tumor was progressing, the tumor was growing. That’s true in 16 patients. They may have discontinued because they have toxicity. AE is adverse events, which is toxicity. They may have discontinued because the physician decided that they should stop the therapy. I don’t really know what that means. Usually the physician decides it’s because they’re having side effects, or their disease is progressing, or the patient really just doesn’t want to continue for some personal or other reasons. So, the physician decision is a little hard for me to gauge. And then at the bottom, they tell you how much time has gone by until they cut off the data, and that’s seven months, with the range of 1.9 to 11.9 months. So that tells me it’s not that long of a period of time until they cut off the data. So there may be a lot of missing data points. It’s still pretty early in this study.

Manju George 26:20
One point that I want to say for the people who are listening, is that we have taken these images or data from different trials. So we are not talking about one specific trial. These are just sample representations of different parts of trial results. And Dr. Kruzrock is helping us understand what is present in all of these, just to make sure that people are not confused that we are talking about one particular trial.

Dr. Kruzrock 26:45
Yeah. I think that’s a very important point, because the last table that you just showed me, I don’t know what trial it was, but I’m gonna guess it had immunotherapy in it, because they had an immunotherapy marker. But of course, I could be wrong. So yes, these are definitely different trials. This is another trial where they gave encorafenib, which is that BRAF inhibitor, together with binimetinib, which is another inhibitor of the pathway, and we’re back again to Cetuximab. So this theme comes up again and again where they’re trying to interfere with all the pathways, and they gave this to 95 patients. And what this table tells you is the characteristics of those patients. So if we just look at the top row as an example, Stage IV at study entry, then they give you the number as N95, so it is 95 patients, so it’s all of them are Stage IV, and then in brackets is 100 so that’s 100% and so this means that everybody in the trial had Stage IV, which is metastatic disease. And then, for the sake of time, I’m not going to go through every one of these rows, but let’s just skip down to the bottom row, and again, reading it the same way, they’re trying to say has the patient had prior systemic therapy, which systemic therapy is chemotherapy or targeted therapy, or something other than surgery or radiation, and they’re saying that 18 of the 95 patients [18.9%] have had prior systemic therapy, and then they give what kinds of systemic therapy. So 17 had adjuvant, that’s 17.9%, 3 had neoadjuvant, that’s 3.2%, and 2 had therapy for locally advanced disease.

Dr. Kruzrock
Let me just say that I’m not clear when I look at this, which is maybe explained in the text, but I’ll tell you something I’m not clear about. So one of the possibilities is the row with prior systemic therapy is different than the row with adjuvant, neoadjuvant, and locally advanced. So maybe they’re trying to talk about prior systemic therapy in the metastatic setting, and 18 patients had prior systemic therapy. And then, in addition, they’re telling you that 17 had adjuvant therapy, three had neoadjuvant and locally advanced. I’m not sure what they mean by that, but two patients had locally advanced disease. Or it could be that the last row is a subset of the row before. So the adjuvant, neoadjuvant, locally advanced, may be a subset of the prior systemic therapy. So when you look at tables like this, I would hope that they had footnotes that helped explain things. But there’s an example of a row where, just by looking at the table, I’m not sure what they’re saying.

Dr. Kruzrock
So now they’re telling you about adverse events, and they’ve drawn this out as a bar. They’re telling you that this molecule, DSA-201, has robust activity in colorectal cancer, and now they’re telling you about the side effects, the adverse events, and they’ve presented this graphically, and I think graphically, people can see it easier. So the x-axis, the axis at the bottom, you see the percent of patients that had a specific type of side effect, and then on the y-axis, on the left hand side, you see the type of side effect. So let’s just start with nausea. And then there’s one more way they divide it, Grade I or II, which is really not usually too bad. Of course, nobody likes to be nauseated, but from a clinical standpoint, it’s not potentially life threatening or anything like that. And then the darker blue, the deeper blue is grade greater than three, and these are problematic side effects. Usually, if the grade is three or more, you actually can’t continue the drug. You have to lower the dose or discontinue the drug. So what you can see is that over 55% of patients had a low grade nausea. So this is a very common thing with this drug. And then maybe another 7% of patients had more severe nausea. And then, if we go through all the side effects, the only other side effects which are related to each other, if you look at the third bar, it’s neutrophil count decrease. Those are the infection fighting cells. And if you look at the last bar, it is white blood count decrease. The neutrophils are a subset of the white count, and they all fight infection. And here you can see that half or over half of the patients that had that side effect actually had it and it was severe. So that suggests being prone to infection is a problem with these drugs, and we have ways to ameliorate that now. We can give drugs like Neulasta or a G-CSF, which promotes the growth of the white blood counts, but we have to understand that this drug has that as a problem.

Dr. Kruzrock
Okay, so here again, we’re looking at safety. Adverse events is how many patients had side effects. Not all of them are really terribly bad. But let’s just start with a row, the top row, treatment related AEs, adverse events. So why is it important for them to say it’s treatment related? Well, because sometimes the way it’s presented is all adverse events. In other words, cancer patients can get sick even without having a treatment, just because they have cancer, and then they can get sick in the middle of a clinical trial when they’re taking a treatment and the way in which they’re sick may have nothing to do with the drug you’re giving them. They just got sick because the cancer caused a problem. And so the treatment related AEs is important here, because these are adverse events that the doctor thought were related to the treatment. The way we rate those is possibly, probably or definitively related, and this is just our best clinical judgment. But even if the doctor rates it as possibly related, it will get included. The only thing that will not get included is if the doctor says not related. So now you see the AEs. Out of 31 patients, 28 had treatment related AEs, and that’s 90%. So, at first, that sounds really high but then we need to look below it and see what is the grade of these AEs. If they’re like grade I or II, maybe that’s not a big deal that 90% of patients had something. You know, aspirin causes problems too. But then I look at this and I see grade III, 12 patients or 39%. Remember I said grade III and IV is already very serious AEs, or often serious, and then there’s one grade V AE, which is death, and it was felt that the death was related to the treatment, not to the cancer, or at least possibly related to the treatment. So, not knowing this drug, I think it has a fair number of side effects. Now, on the other hand, not that many patients discontinued the drug only to stop the drug. So that means that either the doctors were able to figure out a way to lessen the side effects, of course, we have many medicines. If you’re nauseous, we can give you medicine. We have many different medicines that can help side effects. And maybe the other way that the doctor lessened the side effects is simply reducing the dose. They go through all the AEs. I’m guessing that this is the pembrolizumab/lenvatinib study. So they go through what are the common adverse events? Hypothyroidism, that the thyroid doesn’t function. Hyperthyroidism, where it’s overactive and so forth. Hepatotoxicity, which means side effects where the liver was damaged, it may not be permanent, usually isn’t permanent, but where we saw with blood tests that the liver was damaged.

Dr. Kruzrock
So this gives you the outcome, or pieces of the outcome that are important. And then I want to give you a little bit of insight here of what to look for. So ORR is objective response rate, and those are the strongest responses, complete remission, and partial remission. And 22 patients out of 32 had an objective response rate, and then they give you the 95% confidence interval, which is a statistical calculation, because even with a large number of patients, but especially with a small number of patients, you really cannot be sure that 22 out of 32 is right. So they’re giving you 22% of patients had a response. So they could have made that a little clearer by saying percent of patients, but now I realize that’s 22% of patients that had a response, and statistically, that can mean anything from 9% to 40%. Then the second row, I think this is a really important row to be aware of what it means, and they have a little “a” so they may have explained it, but I don’t have the explanation in front of me. So now they say disease control rate. And they’ve added a third parameter. In addition to complete remission, which means the disease went away completely and a partial remission where there was over 30% regression, they’ve added in stable disease. And I personally think that stable disease is very important for patients, because if you can prevent the tumor from growing, that’s a really good thing. But there is a caveat, and I’m going to come back to it. So 47% of patients now with this expanded parameter that includes stable disease had what we call disease control rate. But I have a problem here which that superscript “a” over the stable disease may answer, but let me tell you what you need to look for. Some trials use stable disease at two months, and I don’t think that should be used, but a lot of trials use it to give the impression that it’s a better trial. And the problem I have with stable disease at two months is, by the criteria, you can have a 19% increase in your tumor at your first restaging at two months, and it’s called stable disease. So I don’t think most patients or physicians really think that a 10% or 19% increase in tumor at two months has any meaning whatsoever. So some people use stable disease at four months, which I think is a little better. I personally never use anything but stable disease at six months. I think for patients to stay stable without tumor growth for six months has meaning. I think to stay stable for two or four months is just like it’s not good enough as far as I’m concerned. Duration of response just tells you how long are the responses lasting. NR means not reached, that the median has not been reached, and then you have the range is 2.1+ months to 10.4+ months. So let me explain the plus sign. The plus sign means those responses are ongoing. And that’s actually really important, because if you have a response that’s 2.1 months, that’s not very meaningful, but if it’s 2.1+ months, it means that it’s ongoing when they cut off the data, so you don’t know if the patient is going to fail next week—fail to respond to this therapy—or if they’re going to keep on responding for the next 10 years. So when you have a plus sign, it is very different than when you don’t have a plus sign. A 2.1 month duration of response to me is almost meaningless but a 2.1+ duration of response, well, I don’t know what that means. It may be really good, maybe not, but we don’t know. And 10.4+, this is getting to be a longer duration. It makes me feel better that this is ongoing and that, you know, the patient is approaching a year and still doing well? So plus signs are very important. Progression free survival, the median is 2.3 months, and then they give the 95% confidence intervals. So the average patient here, by the way, median and average are not the same thing, but I’m going to use the words interchangeably. They’re definitely mathematically quite different. Just for discussion, I’m going to say that the average patient had progression of disease after 2.3 months, and you’re going to say to me, well, that’s pretty lousy. I’m going to say I agree with you, but this is missing data because there may be a subset of patients that did very well. And so the real question is, is there a subset of patients that did very well even though most patients clearly didn’t. So with median, at least 50% of the patients have progressed at 2.3 months. Now I’m going to tell you that I suspect that there isn’t a very large subset of patients that did well, and that’s because the 95% confidence interval is 2 to 5.2 months. So there may be some outliers, but there’s not a lot of outliers.

Dr. Kruzrock
And then OS, overall survival, same numbers. Median is 7.5 months, and the 95% confidence intervals are 3.9 months to not reached. Remember, they may not have had much follow up on the study, and so I don’t know what not reached means, but whenever they ended the study 50% of the patients had not died yet. So of course, it depends. Did they end the study after five years? Well, that would be pretty good if 50% didn’t die yet. Did they end the study after 10 months? That would be not so great if most patients died. So it’s not 50% because it’s a 95% confidence interval. This tells you when the majority of patients have died. So they’re looking at confirmed objective response rate, I don’t know what they mean with this study, but it usually means that they had at least one month later another imaging. Now, the other thing it could mean in a particular study is that they had some central review. So I’m not sure what they meant in this study. But the bottom line is that 47.8% of patients had a confirmed objective response. So remember, an objective response is either a complete response where the disease went away altogether, or a partial response where there’s at least 30% regression. So in the field of oncology and especially colorectal, a 48% response rate is considered quite decent. Of course, we would like to see a much higher response rate but 48% is respectable. And then at the lower part, second row, you have a disease control rate of 88%. So the objective response rate is 48% and the disease control rate is 88%. But remember what I told you before? 88% sounds terrific, right? 88% of patients had their disease controlled. But that only depends on what they meant by stable disease. And if they meant stable disease at two months and they included that, and that includes 19% increase in the tumor at two months, that 88% doesn’t mean a thing to me. On the other hand, if they had more stringent criteria and they included complete remission, partial remission and stable disease of at least six months or longer, I would think that that 88% would be quite impressive. So that’s some of the little nuances I think you should be looking for.

Dr. Kruzrock
And these are the Kaplan Meier Survival Curves. These are progression free survival, how long was the patient alive from the start of therapy until the disease got worse, and then the overall survival, how long was the patient alive period. And what you can see at six months on the left side, the progression free survival is 22% so really not very impressive. And the overall at six months is 46% so again, not very impressive. Now I want to tell you, you might ask, why do we use a Kaplan Meier Curve, and what are those little tick marks that you see in the curve? And maybe I can just draw something that will help you see what I’m talking about, this little tick mark here and the tick mark there. See those little marks there? Yeah. So what does that mean? Because that’s actually pretty important, and it’s actually the reason that we use Kaplan Meier Curves. So one of the difficulties, one of the real challenges in analyzing clinical data, is something I alluded to before. If the patient dies at three months, that’s really bad. But if the study data cut off was at three months, and the patient is still alive at three months, that’s a really different story, because you don’t know if that patient’s going to die at four months or if that patient is going to live another 10 years. So what these tick marks are are their patients that have ongoing response and then in the case of the right hand curve, the tick marks are patients that are still alive at the time of data cut off. In doing the math, the statistics of doing progression free survival and overall survival, as you might imagine, is very different if the patient is still alive at three months and that was the data cut off than if the patient died at three months and that was an event. And you can’t statistically analyze them that way. So for years, that was a huge challenge to doing statistical data for patients in clinical trials or under treatment. Kaplan Meier was the solution. They figured out a statistical way to incorporate what we call censored data. That’s data that is incomplete because the response is ongoing or because the patient is still alive. I personally know the statistics, the math behind it quite well. I think Kaplan Meier is definitely not perfect, but it is the best that we have and without Kaplan Meier, everything that we analyze would be wrong. So this is very important to incorporate that censored data.

Dr. Kruzrock
Yeah. So this a very important figure. It’s called a waterfall plot. And waterfall plots tell you which patients are responding and which are not. Everything below the horizontal line, all of this stuff, shows how much the tumor regressed, how much of the tumor shrank. And then you see this other line here, I’m sorry, I’m not the best artist here, but everything below that line is a partial remission. And then when you go down to here, you have complete remissions, because it’s 100%. So what you have, everything above the line, which you see on the left hand side, means that the patient grew, and this is just a percentage of how much the tumors grew or how much the tumors shrunk. This drug has nice activity. What you can see is it is just full of people that shrunk their tumors, you know, all the way till here, everybody on down that. That’s a lot of partial remissions, and some of them are complete remissions as well. So this drug definitely has robust activity in colorectal cancer. And this is just the same thing again, except what they’ve done is done some nice colors for you to help you figure out how good the responses are. But it’s just the same graph. And then this graph is telling you how many days the patient responded. And then if you have an arrow at the end, it means that the response is ongoing. So for instance, right here is an arrow. This patient is still responding. In contrast, this patient here is not still responding, but this just tells you, at the time of data cut off, how long were these responses ongoing. And the other thing I think you need to notice is that the PRs are in green, the partial remissions are in green. And not surprisingly, there are no complete remissions in this study, two of the partial remissions have the longest response as well.

Dr. Kruzrock
So this is again showing in a different graphic, the percent change from baseline in the measurements. And if the line goes up, that means the tumor is growing. And if the line goes down below that zero, that means the tumor is shrinking. So most of these patients had shrinkage of tumor. And then you can see how long it was ongoing. And then the other thing they show you is the patients that had had prior HER-2 treatment and the ones that had not had prior treatment to see, can you see a difference? I don’t think there’s enough patients in each group to see a difference, but they’ve color coded that orange versus blue so that you can see possibly a difference. So this is a very mathematically, it’s not an easy graph by any means. And so let me try to explain what this graphic means. This is a forest plot. Okay, let me just show you this. This is the middle line, and if any of the lines that you see do not cross that middle line, they are likely to be statistically significant, at least in a univariate analysis. I think what they’ve gone and done here is a multivariate analysis where you put all the data together. Everything to the left of the line means the patients did better, and everything to the right of the line here, sorry for my drawing, means that the patients did worse. So in other words, they’re looking at all these different things. The performance status, is the tumor in the right colon or the left colon, and they’re putting it all together mathematically to see, is the experimental group better, or is the control group better? And again, if it’s to the left of the line, that vertical line, then the patients did better on the experimental group.

Manju George 53:08
And then when you see the trial results, what are the things that patients can look for?

Dr. Kruzrock 53:14
So I think the important things are, to me, How safe was the drug? How well tolerated was the drug? What was the response rate with the drug? How long did the responses last? Duration of response or progression free survival? Are there FDA approvals in other cancers? What is the biology of the responses? So I think those are some interesting questions. So these are the take home points that I just sent you, and I see there’s a little dot at the bottom, but this is what, yeah, it could use a little cleanup, but let me go through it with the grammar cleanup not done. So I addressed this point in the beginning. How long is stable disease? Two months is not meaningful. Six months is meaningful. A lot of trials incorporate stable disease. It makes the trial look really good. But if I don’t know how long the stable disease is, if it’s two months, I would take all of that with a grain of salt. Number two, is a trial biomarker based? We already know that, especially with targeted drugs, trials that use a targeted drug without a biomarker, those trials don’t do very well, and unfortunately, in colorectal cancer, I see a lot of trials with targeted drugs, without a biomarker. We recently published a paper in Nature Cancer that you might want to look at. It’s called Missing The Target. It’s a play on the word missing about why we should be doing biomarker based trials. Number three, a lot of people think about phase I trials as safety trials. This is maybe an old fashioned way of thinking about phase I trials. It was true 20 years ago. They are still safety trials, but an essential component of phase I today is understanding the efficacy of the drug, and how the disease biology plays into that efficacy. And the FDA has now approved two drugs, ceritinib and pembrolizumab after phase I. So this old way of thinking that phase I is not about activity, that was good 20 years ago. That is no longer the case. Number four, randomized trials are a gold standard because they reduce confounders and can discriminate between regimens that appear similar. So that is very important. Now I’m going to tell you something that is my feeling, and some people might agree, but a lot of people might not agree. So this is the next statement that we should be looking for regimens with substantial outcome improvement, and if the improvement is substantial enough, a randomized trial may not be ethical anymore. So we should be aiming high. If we don’t know the answer, we need a randomized trial. There is no other way to sort out the answer when you don’t know the answer, but when the drugs are tremendously effective, I don’t think a randomized trial is ethical, and I’ll give you a noncancer example is insulin for diabetes. Nobody would think that we should have ever randomized people with diabetes to insulin versus nothing, because insulin is dramatically effective and with nothing, everybody dies. So what I’m leaving you with is that if we can’t tell the difference between regimens, there’s no way better than a randomized trial, but we should be aiming high for substantial improvement, and if we reach our goals, we may not need a randomized trial, or it may no longer be ethical.

Manju George 57:29
Thank you very much. I think we have gone five minutes past the time. I’m not sure about your schedule, do you have to leave?

Dr. Kruzrock 57:39
I have five minutes. If there’s any questions, I’m happy to answer them, or whatever you’d like.

Speaker 57:47
I have a question.

Manju George 57:48
Yeah, go ahead.

Speaker 57:51
So, hi. Thank you Dr. Kruzrock. So very quickly, I see some clinical trials that say first in human.

Dr. Kruzrock 58:00
Right?

Speaker 58:01
Is there a difference between that to phase I, or is there like a pre-phase I? How does that work?

Dr. Kruzrock 58:10
No, a first in human is a phase I. And usually what it refers to is a totally new drug that has never been used in human beings before. It’s been tested in animals, but just never been used in human beings before. So it is a type of phase I, but it’s first in human. Other phase ones which you could also consider first in human, but they’re usually not called that way, is, let’s say, for the first time ever somebody is going to give pembrolizumab with trametinib or with binimetinib. That’s also a phase I because it’s the first time that combination was ever given. But we usually don’t say it’s a first in human. When we say first in human, we usually mean a brand new drug that is not FDA approved.

Speaker 59:04
One more very quickly. When you said about the stable disease being two months versus the six months, can that be related to the progression free survival or that’s totally different?

Dr. Kruzrock 59:51
It can be and mathematically it will be. They’re calculated in a different way. But obviously, if you have stable disease that is stable for a lot longer, you’re going to have a longer progression free survival. So they’re not the same parameter, but mathematically they will be related. And my main thing with the stable disease of two months, or at least two months, is that I just don’t think that’s very meaningful for patients. Stable disease for six months, you may say, Well, I’d like to be stable more than six months, and I agree with that, but at least it now becomes something that I think might be meaningful. It’s a minimum of six months to be called stable disease. So it might be stable disease for a year or two or whatever, but a minimum of six months, to me, is the parameter that I like to see. But a lot of studies incorporate two months, and some of them incorporate four months.

Manju George 1:00:00
So I’ll just ask one question from the chat. So what is considered different lines of therapy when you are considering a clinical trial?

Dr. Kruzrock 1:00:25
Yeah, so that’s actually a really good question. Most of the time, it is how many lines of therapy you’ve had in metastatic disease. So let’s say you had FOLFOX, that’s one line of therapy. Then let’s say you had FOLFIRI, that’s a second line of therapy. Then let’s say you had immunotherapy, that’s a third line of therapy. So that’s how it is most of the time. Sometimes, though, people will do it differently. If you had adjuvant therapy, like you had your colon cancer removed, and you were free of disease, and they gave you some chemotherapy to prevent the disease recurrence. Some people will call that a line of therapy. Okay, most of the time it’s just the number of lines in metastatic disease, but it’s a good idea to look at the methods to see if they defined it.

Manju George 1:01:23
So I think we can stop there, right? I mean, we’d probably like to have you again, but this was very helpful, and what we’ll do is that we’ll collect some questions, and then we can email you, depending on how many questions we get. So thank you so much for your time.

Dr. Kruzrock 1:01:38
My pleasure. Great questions, and I hope it’s been helpful. And take care. It’s amazing. I met you through Twitter, and it’s just an amazing thing.

Manju George 1:01:51
Thank you.

Dr. Kruzrock 1:01:52
Okay. Take care.

Manju George 1:01:53
Take care. Bye. Take care. Thank you all for attending.

Julie Clauer 1:01:56
If anybody wants to stay on and ask questions here too, we can and to collect them for her, in addition.

Manju George 1:02:01
Yeah, and then we have Adrian, right? So he’ll be able to answer many questions.

Julie Clauer 1:02:08
And one thing that I just would add, which it’s not really on topic, but it’s kind of on topic, is, I think there’s also looking at all the results in the lens of your objectives too, because the results that are published are based on what the study’s objectives are, and the studies trying to achieve, which is for mass amounts of people, right? It’s for a group of people. But when you look at it and say, Okay, what is it that I’m looking for in a trial? You can find some good stuff in there. You know, that kind of gets down to it. Like the forest plot, for instance, it showed where the mets were, and it was significantly better for people with less than two places or whatever it was. I don’t remember the specifics, but you can look at that and go, Oh, I fit into there. So that means this might be more interesting for me than it is when you look at the whole study overall. So I think that there is an element of that, or like, for instance, I looked at the spider plot to say, how quickly will I find out if this trial is working? Because, for me, that was kind of something I really cared about in terms of being able to move on to the next thing, because I was lining up other trials. So there’s some things there that are reported in these results, but the researchers don’t really care about them necessarily as much as an individual patient might. So you kind of have to know what you’re looking for, then you can kind of dig in and find stuff in there that’s valuable.

Adrian Terek 1:03:30
Definitely, definitely. I can, if you guys don’t mind, I can answer a couple of the questions that didn’t get addressed, just from the chat.

Manju George 1:03:39
Go ahead.

Adrian Terek 1:03:40
So, Lindsay, you asked about using a trial coordinator at your center or to go out on your own to look for a trial. I think both. I actually lean heavily towards doing it on your own, simply because trial coordinators are often going to be heavily biased. They’re going to be looking specifically and most knowledgeable about, of course, the trials that are happening at their own center. And it may end up being that the best trial for you might be at a center just one town over, and they may not pay too much attention to that. The other thing is that trial coordinators very much focus on whether the patient qualifies for the trial, not whether it’s a good trial for them, right? Those are technically different questions. There’s a lot of trials that patients qualify for and, generally speaking, the newer a treatment, the less stringent the criteria is. They’re just looking to get some base data, and they’ll take any sort of anybody who was willing to do it. So if you go through a trial coordinator, it’s a great way to learn about what’s going on at your own center. But when it comes to deciding which clinical trial to participate in, I think the questions are a lot larger and places like Colontown actually do a great job in helping people get those answers, because it doesn’t just say you might qualify, but it says, here’s reasons why this trial is actually a really good trial for you to consider. So I hope that’s a useful answer there.

Adrian Terek
The other one, a great question from Robin Lane about how there’s a lot of research going on, but it seems to be going into smaller and smaller groups. And so why would scientists work that way? And it’s a great question, and it actually kind of goes to show how the funnel of research works. So what they do is like Dr. Kruzrock said, you have these biomarker based trials, but that didn’t always exist. In fact, we weren’t even able to really measure biomarkers in patients, you know, 30 years ago. And so they would take this treatment and test it in a large group of patients that had one simple correlation between them. They all had, let’s say, colorectal cancer. And they would say, okay, 20% of the patients had their cancer disappear, and in 80% of patients it did nothing. And so then the next question becomes, okay, what’s unique about these 20 patients who had their cancer disappear? Something’s different about those patients where it worked than the ones that it didn’t work for. And so they start to dig into that, and they essentially identify a subgroup, and they say, Oh, it turns out this group of patients who did really well, happen to have this thing like called KRAS G12C, for instance. And they’ll then say, Okay, now we can do a new trial of this drug in patients with just that mutation and see what happens. And then they might see that it’s much more effective in that specific subpopulation. We brought up MSI versus microsatellite stable, which is the great paradigm to consider. Immunotherapies were initially given like in most other cancers, to patients who just had the kind of cancer. They gave pembrolizumab and nivolumab and the other PD-1 inhibitors to just large swaths of CRC patients, and what they saw is similar to what they saw in the other cancer types, which is a small subset within that group of CRC patients did exceptionally well. And they started to ask, why is that? And that’s when they discovered, okay, one of the things we’re noticing that’s common amongst this group that’s doing really well is they have a huge amount of mutations in their cancer. And they then started to apply just logic, we know that the immune system recognizes non-self cells based on sort of the amount of antigens that are non-self that it’s releasing. So the more mutated a cell is, the more likely it is your immune system will recognize it in the first place. And so they now had sort of preclinical and medical rationale and a biomarker that explained why that subgroup did better. And so the overall answer to your question is not that they actually go out and look for these tiny, small groups of patients. Everyone wants a drug that works for all patients that have CRC. But as knowledge progresses and we realize how unique each person’s cancer actually is, they’ll take a treatment and try it in a large group, see that it only works in some patients, and try and figure out who that smaller group is, and then they’ll try it in that smaller group and hopefully figure out what the details are in that group.

Manju George 1:08:03
Great answer Adrian. I was trying to find that journal article that she talked about Missing Targets in Cancer. But I don’t have through my university library. I don’t have access, so maybe we can ask her to give us a PDF. I think that’d be kind of interesting to read. Adrian, there is one more question. Can you choose your arm when there are different arms? Do you want to answer that also?

Adrian Terek 1:08:26
That’s a very interesting question. If it’s randomized, you can’t. So randomization is done through computer software so nobody has any sort of say in how that works. Theoretically, when it comes to trials that are not randomized, they’re just testing three different arms with different combinations, then I have spoken to patients who have worked with their doctor to say, okay, this trial has six different arms and for various reasons, maybe one of the three out of the six arms includes a treatment that has a side effect that my particular patient might not be a good candidate for, but they still want to participate in this trial. So can we at least put them in one of these two arms? And I’ve definitely heard of that actually happening where they do that, because when you have multiple arms, it is up to the investigator’s discretion which arm you end up in, right? It’s not a roll of the dice. That’s only with randomized trials where you don’t have a choice.

Julie Clauer 1:09:25
And then also, can you talk about crossover Adrian?

Adrian Terek 1:09:29
Crossover. Yeah, so crossover is when they want to some degree incentivize people to do randomized trials because as people get more knowledgeable, they recognize that often in a randomized trial, you very much want to be in the group with the newer treatment. And so what they do is if you end up in the trial arm that’s not with the experimental treatment and it’s with the treatment that’s either a placebo or what they know is not very effective, so what they’ll do is that they’ll have you participate in the trial in that arm. You’ll get that treatment until it stops working, and then immediately after it stops working, they’ll allow you to essentially join the other arm, so you can crossover to the other arm and get the experimental treatment. Now that I think is quite useful for patients. It does make the data a lot trickier, though. So I think that part of crossover is often a relic of what Dr. Kruzrock was mentioning about how randomized trials are not quite as important as they once were, and I see less and less of them because, very often drugs are getting close to the point of approval, far before they even need to do a full phase III randomized trial. The data is so overwhelming that they can’t recruit patients. Patients just won’t join the trial when they can find, hopefully, another way to get that drug because they already believe so much that it’s going to be helpful. Yeah. So crossover compounds the data hugely because the number one thing that doctors look at is overall survival. And so if they are trying to compare, let’s say a chemotherapy arm where they have a pretty good understanding that, you know, the average patient is going to survive, let’s say one year, I’m just picking a random number out of a hat, and they’re comparing that to this pembrolizumab arm, if they allow the patients the moment they progress, so let’s say they go on the chemotherapy arm, they go on the treatment, and after four months, they progress and immediately switch over to the pembrolizumab arm, they may actually be gleaning all of the benefits from a survival perspective, from the pembrolizumab. And if that’s true, then when you try and look at the two different arms, the survival curves, they might look very similar, and so you might actually reach the wrong conclusion that pembrolizumab doesn’t increase survival compared to chemotherapy, but that’s because most of the patients in the chemotherapy arm moved over and got pembrolizumab, and if you would just let them not get pembrolizumab, then they may have passed away much sooner. And so it really, really confounds the data. Like I said, it’s something that I don’t see very much, very often anymore, simply because it makes it so difficult for anyone to figure out what the actual truth is.

Julie Claur 1:12:12
Another thing I just want to note, and you might be able to shed some good light on this, is when you look at this data, I know, when I first started looking at it, I’m like, five months survival, what the heck? I want like ten year survival. But those numbers are not like you’re gonna die in five months. So I think, you know, it goes to the median point, and it also goes to, like, these trials are looking at one thing, one part of treatment, and usually it’s for very far down the line. You saw like two or three lines before this. You’re at the point where it’s like later in line. But I think there’s an important element to it. Trials aren’t like one and done, maybe if you get complete response, it’s one and done. But you have to think of your treatment as like a long journey, and as Todd Scott in Colontown explained once, if you get four months, four months, four months, and you add those together, you end up with a really long duration. So, those numbers that used to scare me, and you say, oh, 22% is a good response. And it’s like, what do you mean that’s a good response? But when you think of chemo, chemo actually is around 30% depending on where it is. So standard of care is around 30% and we know by manipulating and doing multiple treatments and doing local treatments and everything, more than 30% of people have a good kind of response on chemotherapy and standard of care treatment. So just something to note, because I think sometimes the numbers get scary, or look scary, but they’re actually not as scary as they appear.

Adrian Terek 1:13:56
There’s a really beautiful essay that was written. I can’t remember who wrote it, but it was called The Median is Not the Message. It’s from a gentleman who had aggressive type of cancer that normally you wouldn’t survive very long for, and I do believe he survived for 10 or 15 years, so way better than you would normally expect. But beautiful, beautiful essay describing why median can be somewhat misleading for patients, especially when it comes to viewing things from a hope perspective, right? I always relate median to say half did worse, half did better, and there’s no reason to think you won’t be in the half that did better, you know? So what you have to realize is that medians is inherently left skewed. So what I mean is, if someone says the median survival was 10 months, like in your example, half the patients would have passed away between zero and 10 months. So there’s only a total of a 10 month window to exist within that one half. The other half survived longer than 10 months. And that window can be massive. That can be from 10 months to 12 years, right? And so you’re going to see a large difference in how long that can play out for patients. And then, like you mentioned, clinical trials also give you a lot of possibility to do future treatments or alter your treatment, right? Some patients will do a clinical trial, and they’ll have a certain amount of shrinkage, which suddenly allows them to become eligible for surgery, and then that surgery ends up making a massive difference to their survival. And even recently, there’s a lot of interesting data coming out about patients who go on certain vaccine trials. And like most of the vaccine trials, they don’t really seem to provide a lot of benefit out the gate. There’s almost nobody who has their cancer shrink due to the vaccines. But they’re starting to notice this weird effect where people are having much, much better responses to the follow up chemotherapy. So they go on this vaccine trial. They are there for three months. They get their baseline scan, get the vaccine, get their three month scan. The doctor says your cancer is still growing. Vaccine didn’t do anything. And the doctor says, Okay, well, you’re not going to just stop treating you. Let’s put you back on one of the known, you know, chemotherapy regimens. And you get put on this regimen, and the doctor thinks it’s not going to have a huge effect, it’ll just be helpful. And suddenly, and suddenly they notice, like, you know, massive amounts of decreases in tumor size. And they’re like, this is not what we regularly see. And so there’s this idea that, even treatments that you’ve had before can have a pretty strong effect on the future treatments that are available. Manju also talks a lot about rechallenge, so the idea that your tumor is this constantly evolving, I don’t want to say organism, because it’s not technically separate, but you can go on a chemotherapy that’s very effective the first time and at some point it stops working. You can then go do a clinical trial, and let’s say it keeps you stable for nine months, or what have you, and then it stops working, and you have to make your next decision. A lot of times, patients can go back on that original chemotherapy that worked for them and have another good response to it, because their tumor’s biology has changed or reverted back, or some of the resistance has gone away. And so you can recycle treatments as well. So there’s a lot of hope out there. I don’t want patients to look at an overall survival curve, look at a point on that curve and say, that’s me, because that’s not actually quite how it plays out. It’s often very different. And, like I said, there’s no reason to think that you won’t be on the other side of the median, and not the earlier side.

Manju George 1:17:18
I just want to add that we have seen effective chemo being better the second go around after people have gotten on immunotherapy, and especially for MSS patients, when the immunotherapy hasn’t worked and they come back on chemo, then suddenly the chemo is working so much better. So we’ve seen instances of that. And then I think there’s one more question where Laure is asking that there are two arms in a trial, can the patient be on the two arms at the same time when the two arms have two different drugs?

Adrian Terek 1:17:49
No. So the whole point of each arm is to stratify patients into a specific group that they can then analyze those results for. And so the kind of thing you’re describing there, where one arm is for an AKT1 mutation and another one’s for BRAF V600e, what they’re essentially saying is we believe that there’s some medical reason that patients like yourself will potentially benefit from this drug if you have an AKT1 mutation, or if you have a BRAF V600e mutation, and we’re not sure how much benefit each one of these mutations confers. So they’ll break people up into groups and give the drugs to each of those groups. And at the end, they’ll say, okay, for the patients who had the BRAF V600e mutation, this is how well the drug worked, and for the patients who had the AKT1 mutation, the drug worked twice as well. And so they might then say, okay, moving forward, we’re going to really focus on the patients with the AKT1 mutation, because that’s where the biggest benefit was seen. 

Manju George 1:18:45
Yeah, I remember from her post in the BRAF group. So I think in this case, the patient has both the AKT and the BRAF mutation. So that’s why she was interested in whether she could get both the drugs. But I think what Adrian was saying is when they are randomized and there are two arms, or even if they are not randomized and there are two distinct arms, in general, patients won’t be able to get drugs from both arms at the same time.

Adrian Terek 1:19:12
Yeah, exactly. And it’s interesting because this kind of brings up, like Dr. Kruzrock mentioned the idea of biomarker driven trial. Her and I might disagree a little bit on this, because I know reading a lot of Dr. Kruzrock’s work, that she also knows that combinations are the sort of way forward. Pure biomarker mutation driven trials are actually very often a failure because there’s so many different pathways that are being activated in cancer, and just stopping one of them often isn’t enough, but there’s been a ton of data from precision medicine protocols over the last few years that show that combinations can be extremely effective. There was just a recent one in GBM, which is a very aggressive type of brain tumor, and they basically took a basket of like seven different drugs, and they did molecular testing for all the patients, and then for each mutation they had where a drug was assigned to that basket, that patient would get that drug. And so if a patient shows up and has four of those mutations, they would literally get all four of those drugs, and the results were quite exciting. It seems like the future of precision medicine is, of course, being precise. It means the goal is to be able to analyze someone’s cancer or even their immune system or both, and say, Okay, for the best effect, we can attack your cancer in four different ways with these four different drugs. But right now, we’re still at that point where they’re mostly just trying to see how effective one drug is for one mutation and often that’s not the most effective way, because, again, there’s there’s just so many different pathways involved that just saying this person has mutation x and this drug blocks mutation x, therefore benefit is usually not the case, unfortunately. The very first thing I said was the key for picking trials is look for trials that already have some data. Preclinical evidence exists for every trial. So when I pull from clinicaltrials.gov and I see, you know, 18,000 trials testing treatments and combinations in cancer, every single one of those, to be approved by the ethics review board, has shown preclinical evidence. There is medical rationale for why every single one of those trials should be done. And yet, as we also know, most trials don’t provide a huge amount of benefits. So the key to me is always look for trials that have some early evidence that they’re doing something for people, and then you can kind of say, okay, based on that, which one of these five trials that I’ve learned about, that have all shown some benefit, which one suits me the best?

Manju George 1:21:54
I just wanted to say this. You know, from other conversations with Dr. Kruzrock, what she means, what I have understood, is that any time any trial has any results, you have those people who respond to the treatment versus the people who don’t respond to the treatment, her idea of the precision medicine part is to understand why the people who responded responded, so that then you can identify biomarkers that may be associated with the response and then use that to screen the additional people to see whether, in the large number of people that are being recruited, how many have those biomarkers which will make it more likely that they would respond to the treatment that is being tested. 

Adrian Terek 1:22:35
Oh, then yeah, I misunderstood, and I could agree completely then, because the whole goal of precision medicine is to identify why the people who benefited did benefit. And so you can predictively identify those people in the future, right? And so, I can imagine a world, wishful thinking maybe, but I can imagine a world where people get their tumor sampled, and maybe an immune system sample and run it through a machine, and the machine spits out an entire profile of that person and says, based on that profile of these biomarkers, these are the four treatments you should sort of combine to get the absolute best outcome. We don’t want to be flying blind forever. And I think huge strides have been made already in the last 10 years when it comes to biomarker driven research and trials.

Manju George 1:23:23
I think maybe this is a good time to wind it up. I hope that this was useful to people, and it’ll be posted in Colontown as well as Colontown University. So if you, as you watch it again, and you have questions, you’re free to post them, and then at some point, we can take them to Dr. Kruzrock. So Adrian, thank you so much for your expertise and for your time. And Julie, thank you for organizing it.

Adrian Terek 1:23:48
It’s a pleasure. Thank you. 

Julie Clauer 1:23:49
Yeah.

Manju George 1:23:50
So take care.