Shifting Clinical Trial Design Towards Improving Patient Enrollment - Michael Morris

November 17, 2023

Alicia Morgans engages with Michael Morris to discuss the transformation of clinical trial design to improve patient access and applicability of results. Dr. Morris highlights the complexity of current clinical trials, which often involve extensive documentation and stringent eligibility criteria, limiting participation to a narrow patient population and placing a heavy burden on both participants and research centers. They explore the need for simplification and decentralization, allowing trials to be more inclusive and reflective of real-world settings. The conversation also touches on the innovative steps taken during the COVID-19 pandemic to maintain trial continuity, such as telehealth and local care, which could inform future practices. Dr. Morris emphasizes the importance of collaboration among stakeholders to streamline trials, focusing on essential data collection and leveraging real-world care settings as control arms. 


Michael Morris, MD, Prostate Cancer Section Head, GU Oncology, Memorial Sloan Kettering Cancer Center, New York, NY

Alicia Morgans, MD, MPH, Genitourinary Medical Oncologist, Medical Director of Survivorship Program at Dana-Farber Cancer Institute, Boston, MA

Read the Full Video Transcript

Alicia Morgans: Hi, I'm so excited to be here with Professor Michael Morris of Memorial Sloan Kettering to talk about how we can improve access and design of clinical trials to ensure that more patients have the ability to participate. Thank you so much for being here with me today.

Michael Morris: Thanks for having me.

Alicia Morgans: Wonderful. So I would love to hear from your perspective how we really change the paradigm of clinical trial design to ensure that we are testing the people who really need to be tested, and then ensure that we can generalize that data to the populations that we care for.

Michael Morris: Well, if you think about the average clinical trial that we either design or participate in, you're talking about a document that's generally hundreds of pages. And contained within those hundreds of pages are generally many pages that describe eligibility, inclusion, and exclusion criteria to create a relatively tightly defined and internally controlled patient population that's going to be on the trial. And then the interventions are explicitly outlined to the greatest detail of what drugs will be applied and the doses and the dose reductions and what to do in the event of toxicity and highly descriptive and prescriptive imaging and assessment schedule of lab draws, tox assessments, et cetera.

And in that whole endeavor of those many hundreds of pages that you need to read, and the patient needs to read a 30 to 40 page consent form. And we know that most patients do not read that whole document. And we know also that those documents have to be written at an appropriate level of medical understanding, that in the end you get a clinical trial that involves an enormous amount of work for the investigators. They need to have a dedicated nursing staff, they need to have dedicated research staff to do copious data entry. And then there's usually an auditing or monitoring process, and then more data entry and more queries. And those queries take time from your research staff and from the investigators. And from a patient perspective, multiple visits to a participating physician or cancer center or practice, multiple visits in which they have to either lose time at work, lose time taking care of their dependents, or just lose their own time at a point in their disease where they have little time to spare.

Participating in that trial frequently for long days getting PKs, long days getting EKGs, long days getting biopsies, missed work for imaging opportunities. And so you need to not only have participants who have all of those resources to spare, but also the study centers, which also are well-resourced. So you're getting these very refined populations, a lot of unnecessary data collection, and you're excluding the participation of large segments of society from a patient and from a provider perspective who could be participating in clinical trials but simply can't do so as a result of that. So there's a big initiative on the part of NCI, FDA, other stakeholders like ASCO, EORTC, patient advocacy groups. I think there's general consensus that this is an unsustainable system and especially in the post COVID era, we now have staffing shortages, nursing shortages, and even the whole pattern of getting your oncologic care.

And this whole centralized system of accessing trials is a pattern that is really yesterday's treatment pattern. So there's a lot of now effort to collaborate in order to simplify clinical trials from the design on up and also how those clinical trials are conducted so that there could be perhaps more local care, more distributed care in order to decentralize the resources and allow patients to access clinical trials in their communities more feasibly and easily.

Part of those solutions, I do make a contribution to the NCI effort to do that by co-directing a new unit called the Clinical Trials Innovation Unit. But that's only one piece of a broader vision on the part of the NCI. To simplify trials, make data collection focused on the essentials, have more of a real world assessment on both the treatment or the intervention side and on the control arm side in order to make trials more feasible, reduce the resources that are needed to access them and reduce those hurdles, and make them ultimately have a result that you can get faster so that you can get the data to, or the new therapy or new intervention or new biomarker, whatever's being tested to the patient and their clinicians that much faster.

Alicia Morgans: So this is a fantastic effort, and I think I've heard talk about this over the last number of years because we absolutely want to do this, and we needed to do this. In the setting of COVID and the pandemic, we actually didn't always have the opportunity to have our patients in person. What are some of the nuts and bolts steps that groups are taking now to try to make this a reality?

Michael Morris: Your raising of the COVID era is so informative, and if we think back, we've already taken many of those steps, it's just a question of sliding back to old bad habits. We did allow patients to get a lot of their care locally on clinical trials. We did allow their visits to be conducted by telehealth. We did allow them to access clinical trials, nursing assistants, and evaluation by telehealth. We did allow IRBs to aggregate, especially a lot of the just routine AE information and then reported in bulk. We did allow passes on some really more of the trivial protocol deviations in order to keep patients safe. We reduced the amount of times that they needed to come into the centers. We allowed more flexibility in terms of where they could get their imaging done. So those are all things that we've actually in a very practical way have tested.

And the reality is that most clinical trials were just fine in terms of the data collected and the adequacy of whatever was being studied and answering the question. The question now is why can't we just continue to do that and why are we losing that imaginativeness, that innovation, that resilience, that ability to think of new ways to bring trials to patients rather than bringing patients to trials. And I think that when you think about what are the operational things that we have yet to explore, some of those have to do with how we collect data and how we enter data. It's a very cumbersome process, but there's some really exciting and innovative new ways of directly getting into the EHR or by pushing data out of the EHR into a study database framework. How we design trials still in large part excludes the role of telehealth.

And largely that's increasingly difficult to employ because state by state licensure issues really do interfere with the ability to have clinical trials incorporate telehealth because of those state dependent licensure regulations. But there are many things that we can do in terms of simplifying designs, removing a lot of the data collection that ultimately is not needed, minimizing the biospecimen collection to the essential correlatives as opposed to a lot of studies in which those biomarkers are sitting on shelves unanalyzed or poorly analyzed in the future. So there's much work to be done, but I think that we've actually done the three-year study just as a pragmatic matter of living through the pandemic and continuing clinical trials in their operations during the pandemic.

We just have to actually make that standard and routine and not like, "Oh yes, that was in the golden age of clinical trial pragmatism, which happened to be in the horror of the worst pandemic that we've seen." We should be able to extract those golden nuggets that we had and put them in a post pandemic era, which is a challenge, honestly. It's a challenge to get people to think differently.

Alicia Morgans: It is, and also to reassure that the data will still be as robust and that the findings can be interpreted in the ways that we need to interpret our clinical trials. But I'm glad that they have you on the job to help with that process. Now, one other piece that I think is really interesting that I've heard murmurings about and I think is starting to come into clinical trial design is the pragmatic control arm that may be based on real world data or maybe based on historic controls. Can you speak to that and the potential use of these control arms in the regulatory approval process, which is actually part of what I understand the goal might be for some of this work?

Michael Morris: Sure. When we think about control arms, we're really generally speaking, talking about therapy that the patient could access every day under the routine care that they are receiving through their clinicians. Now, that raises a number of possibilities. First of all, the control arm could just be routine care and not have a very prescriptive set of circumstances, dose reductions, treatment selections, et cetera. It could be simply get your care as you would always get your care. And indeed, if you did that for the study evaluations for the control arm, you'd need to do that for the investigational arm as well. You can't oversample in one arm and undersample in another. And that's a legitimate strategy is at least for the evaluations, apply that across the board in the clinical trial. And indeed in studies in which you're comparing approved drug A in a different disease setting with approved drug B in an existing approved setting, the entire trial could be pragmatic in many respects for both arms, but certainly for the control arm of many studies you could have that simply be their routine care.

So that's a pragmatic control arm and it really is a user-friendly way of doing things. And it also gives you very real world data as opposed to what some have discussed in relation to some randomized trials accusing the trial of having a weak control arm or an insufficient control arm or an inactive control arm. This at least you could say the control arm just represents the real world reality of a patient in that space getting treated. The second is a more synthetic control arm in which you're using historical data to predict what these patients would look like in models. And I think that that requires discussion with regulatory agencies to see what the acceptability would be. But certainly in terms of pragmatic control arms, agencies have already weighed in on that. The Pragmatica-Lung trial, which is an NCI study, looking at second line checkpoint inhibition in refractory non-small cell lung cancer, was designed in many respects as an innovative collaboration with the FDA. So purely pragmatic trial.

The whole protocol document is around 40 to 50 pages, and the patients are receiving approved therapy on both arms, and the agency is comfortable putting their endorsement on that kind of an approach where the patients in as much as possible are receiving a standard within their own framework, their own doctors. The safety assessments are pretty stripped down. There are no dictated study evaluations in terms of imaging or labs. It's just standard of care. That certainly has been fine by the regulatory agency. And I think it is an example of how a good dialogue with the regulator upfront can lead to these innovative designs that ultimately I think everybody is interested in seeing, including a regulator, because the FDA doesn't want to make trials uninformative except for very narrowly defined populations, and they would like to see barriers reduced in terms of accessing clinical trials. And they also have an interest in internationalizing the whole endeavor.

Alicia Morgans: That all makes sense, and it's really important, I think in the example that you just pointed out, this is the lung cancer setting, that both arms are really being run in that pragmatic way. I could imagine that if only the control arm is either a real world dataset, let's say, or historical control, but perhaps more so with a real world dataset, and then you have an intervention arm that's enrolling and getting something that they couldn't get in standard of care, there might be an effect where if there was a control arm in that interventional setting, the control arm would even do better than the real world control arm. Because there is this effect that as we enroll patients in clinical trials, they do better either way, whichever arm.

Michael Morris: Right, because they have more visits, they have more support frequently with the whole study staff that is committed towards maintaining them. And at least in a pragmatic trial, they are getting their routine care as they always would. And so you can get a sense of whether a regimen is tolerable, a regimen is efficacious in that context. And that leads also to the whole discussion of dose optimization, which is frequently when you transition from the heavily supportive environment of a clinical trial into the real world, you find that the dose isn't optimized because those patients have toxicity management issues that in the clinical trial setting were manageable. But in the real world setting leads to illness. And that is a whole separate topic, but getting to a real world dose that is better defined before you get the drug approved. And that too is a big priority, dose optimization.

Alicia Morgans: Absolutely. It's so interesting to think about how we're making this happen and doing so in a way that really continues to include those rigorous statistical assessments and the thought around all of these things that can make it more challenging. And I commend you and the team for pushing us in the direction because I don't think that it's something that we've actively moved towards, though we've talked about this, I think, for a long time the steps have been minimal.

Michael Morris: Well, if you think of how in the age of chemotherapy dose was defined, that model just may not be appropriate anymore. You would dose escalate to your maximum tolerated dose so that you are now dosing it as you moved out based on a handful of patients, the barely tolerable dose. And that's just not what we necessarily need, especially in today's biologic era or targeted therapy era. And when you're talking about how many hundreds of thousands of patients are taking a dose of a drug because of the way that we do phase one and phase two studies traditionally that are just barely tolerable when we could be dosing at the dose that's most optimal as opposed to barely tolerable, and the biologically effective dose in today's day and age may be quite different from the maximum tolerated dose. This clearly is something that should be at the forefront of everyone's mind. And our phase one studies need to be a lot more complete as opposed to just reaching an MTD and a DLT as to what the real optimal dose is.

Alicia Morgans: I could not agree more. So as we move away from these maximally tolerated doses and these dose limiting toxicities into a way of really controlling the cancer without having as many side effects as are maximally tolerated, I think we are hopefully going in a direction that is closer to not just efficacy in a trial, but really effectiveness in our patient population because that's where we practice every day. So I commend you for this work, and I would love to hear final thoughts on where you go from here.

Michael Morris: I think that we need to really pay attention to this rare opportunity. You have alignment of patients, regulators, and government. And I think, you notice I left out to some degree sponsors because sponsors want, I think, to reduce barriers to trials, but then they are also risk averse. And so we need to take advantage of this alignment and really push the field forward now when we have the opportunity to do so. But now the effort is in our hands as the folks who design trials, the sponsor's hands, who ultimately are going to pay for and assume the risk of studies. And we are the two groups that need to focus on getting this done in design, implementation, and sponsorship because I think that there is an alignment of the other stakeholders to have these types of trials and to have this goal achieved.

Alicia Morgans: Wonderful. Well, we are taking steps to make that happen, and you are a big part of that process and those efforts. So I thank you for your time and for sharing your expertise today.

Michael Morris: Thank you, Alicia. It's a pleasure being here.