Kelvin P. Lee, MD, discusses the nuances of an effective molecular tumor board, explains the practical application of precision oncology genomics, and highlights the pathways that have the potential to change the treatment landscape for precision oncology.
Precision medicine involves more than molecularly sequencing a tumor and matching it with a targeted therapy, according to Kelvin P. Lee, MD, who argued that bulk sequencing may become an antiquated approach when single-cell RNA sequencing technology becomes available because it can provide a clearer picture of the complexity and heterogeneity of tumors.
“Once we move away from bulk sequencing, which essentially says, ‘This whole tumor is all the same’ and get down to a much finer specificity that [enables us to identify] tumor cells that have this mutation and look like this vs tumor cells that have that mutation and look like that, we will get more and more sophisticated in how we treat patients. The more heterogeneous the tumor––and they become more heterogeneous as they are later in treatment––the harder they are to treat, and the less helpful our current precision genomics technology is in that setting,” Lee said in an interview with OncLive® following an Institutional Perspectives in Cancer webinar on Precision Medicine.
In the interview, Lee, director of the Indiana University (IU) Simon Comprehensive Cancer Center, the H.H. Gregg Professor of Oncology, professor, Department of Medicine, Division of Hematology/Oncology, associate dean for cancer research, IU School of Medicine, discussed the nuances of an effective molecular tumor board, explained the practical application of precision oncology genomics, and highlighted the pathways that have the potential to change the treatment landscape for precision oncology.
Lee: The take-home message is that every tumor should be sequenced. If you don’t do it now, you may not be able to do it later. The information can be useful in designing a treatment plan, but it also helps us understand that cancer. The more information we have on lung cancers from patients, the better we can design new therapies and move that field forward. If a patient has a tumor specimen that you can send to precision genomics, send it because maybe we don’t have an answer today, but maybe in 2 weeks we will have an answer, and that becomes important for that patient.
At IU there is a precision medicine team, and our patients are referred to them. That precision medicine team sets up the testing and the [molecular] sequencing and then does tumor boards, where they go through the results, they talk about what abnormalities and what mutations were found and then they go through what [mutations] are potentially actionable and how we want to utilize [treatment for] those [alterations]. The nice thing about tumor boards is that all the experts are in the room, and now that we have extended [ours] to be virtual, not only can we do precision medicine downtown at University Hospital, but now we have clinics that are in our suburban and metropolitan regions. We’re moving those facilities out farther, so we can do precision medicine, hopefully, out in the community. Hopefully, with the expansion, primary care physicians and community oncologists can also utilize this [opportunity to] get the referrals and then sit in virtual tumor boards and get answers for their patients.
The key piece is not only doing the [molecular] sequencing and understanding what the mutations are, but also what you’re doing with that information. For many of the mutations, the therapeutic that might be applicable to that patient or that mutation are in clinical trials. An effective precision medicine program has to not only have the expertise as to what these mutations are, what these established targets or established therapeutics are that we might be able to use, but also has to have an ongoing real-time knowledge of all the clinical trials that are available and in the literature.
Not only are there new compounds that a patient might be eligible for in a clinical trial, but there are also drugs that have been used for other things that are being repurposed for targeting mutations that we had not previously expected those drugs to be able to do. The precision medicine team must be aware of the literature, not only published literature but abstracts and journal or meeting work that says, “Maybe you can use this malaria drug to target this mutation,” which may be something that had not been previously understood. That is, overall, a key piece of what makes precision medicine so effective and so important, but so difficult to do. It’s not just sequencing stuff, and then saying, “There it is” and then figuring out how to manage that [patient].
Next-generation sequencing is now allowing us to get at whole genomes, instead of just testing for BRCAmutations or [other single-gene mutations]. We are identifying other mutations, and now instead of having to re-sequence somebody’s tumor, we have all that data that’s there. If something pops up that later becomes this polymorphism or this mutation that is important in this cancer, we have that data; we can go back and look at those aspects. What it allows us to do is really what I would consider the next generation of precision medicine: to understand how mutations work together.
For immunotherapy, [tumor] mutational burden has been a key driver of whether checkpoint inhibitors are important. Now as we understand more of what a person’s cancer has, in terms of mutations, our informatics, our artificial intelligence, and our machine learning technology is poised to take that data and say, “if you have these 2 mutations: Are they compensating for each other, and do you have to target both?” Now, if a patient has 1 mutation, they get this 1 drug, but biology is much more complicated. There are mutations that may work in concert with each other that may develop other additional vulnerabilities that we didn’t anticipate because one mutation is causing the cell to do something and another mutation is stressing the cells, so maybe there is a target that’s not either one of those 2 mutations but that is in the pathways that those mutations are driving that can be gone after. The exciting thing for me is understanding that. I’m an immunologist, so understanding the complexity of cells, because immune systems see lots of things simultaneously, is really what we are looking forward to.
It’s a very active process; patients are identified at the beginning of the week, and then the team asks: What are the mutations? What are the [alterations] that are actionable? Then they begin to sort through what treatment options are available and what the adverse effects [AEs] are. PharmDs have to ask: What are the AEs, particularly for experimental agents? What are the interactions with other drugs? How do we get these things and what literature supports the use of this agent? Then, are there odd things that we have to understand? Are there subsets of patients who have particularly bad responses that had been reported in the literature? All that gets pulled together.
Our tumor board is later in the week where all that information is presented: the patient case is presented, the genomic description or the description of the genetic changes are reviewed, and then the treatment options are also reviewed if they’re available, and the strength of each [drug] in terms of the data that says that this drug would be particularly good in this patient, or this drug would not be something that we’d be looking for in this patient. A lot of what makes tumor boards effective is that research. It’s not just, I have a piece of paper says I can give this drug and then you are done. It really is a lot of thoughtful research that goes into understanding the options that a patient might have.
The case study showed what the right process is to analyze the data you get to reach a meaningful action plan. As the technology goes forward, and as our ability to detect things gets more sophisticated, as we start to move toward single-cell sequencing, for example, RNA sequencing, when we start to look at the epigenome, we will have substantially more data than we have now. Then we’ll start looking at what the patient’s immune system looks like when we start [molecularly] sequencing that. The amount of data that will be collected and the kinds of data that will be collected will grow exponentially.
The key piece of molecular tumor boards and the key piece of precision medicine is: How do you analyze that data? The analytical pipeline is going to be the same, the structure is going to be the same. How do you act? How do you take that data in? How do you analyze it, and then how do you use that analysis to come to specific treatment recommendations for that patient? That framework, that pipeline is going to be the same regardless of what the data coming in is. The key thing that was important in that whole process of going through these case studies is to recognize what the steps were that were taken from the very beginning, from the actual case itself where the patient comes in with their history. How were those data put together and analyzed? How was that used to make decisions for the treatment plan for that patient? It’s the structure of the analytical process that was the most important aspect of going through those case studies.
The data suggests that with early genomic testing, when you apply it to precision medicine to identify therapeutics, the anti-cancer effect, or the ability to impact a person’s cancer is greater on early diagnosis than late diagnosis. There probably are a variety of biologies that are implicated by that. When a tumor has been exposed to lots of things, it probably not only has its initial mutations, but probably has a lot of adaptations that have happened because of chemotherapy that has been given that we don’t necessarily pick up; it may not be a genomic abnormality, but it may be overexpression of a particular gene that confers resistance. As those tumors become more resistant to therapy, the initial driver mutations may become less important as things go forward.
It speaks to biology also, because in the beginning, probably, in tumors that have not been treated or not been heavily pretreated at the time of diagnosis, they’re probably less genetically complicated, so maybe they have just one mutation. Maybe all of them have that one mutation, you treat them with a drug, and they all die. As tumors go along, they become much more heterogeneous, so instead of one population of cancer cells, now you have 75 different populations or tribes, for example, that are living, and some of them are sensitive, some of them are different. Some of them have different mutations, and precision genomics is moving towards being able to understand complex tumors, such that some of the cells have one mutation and some of them have a completely different set of mutations. We don’t pick that up right now, simply because we don’t have the single-cell RNA sequencing technology yet, although that’s coming.
We are beginning to see precision medicine in the context of immunotherapy. In that sense, the change in framework is it’s not the cancer alone that’s important because not only do you have to sequence the cancer, and understand it’s genetic makeup, you also have to sequence the immune system—the normal part of the patient that is essentially the effector part. Instead of giving chemotherapy, where you’re saying, “I have a drug, I know everything about that drug, and all I need to do is figure out what’s going on in the cancer,” if I can find that this drug will hit this piece of the cancer, then let’s put those two together.
For immunotherapy, you have the cancer, which is doing stuff and it’s dynamic, and it’s activating the immune system and suppressing the immune system. We have to understand that, and some of the suppression that it does is not because it is suppressing the immune system, it’s making the normal tissue around it suppress the immune system.
People say that cancer is a non-healing wound, so essentially, for wound healing, you don’t want your immune system to fire up and start destroying all the tissue around a healing wound. Otherwise, you’ll never heal. The normal body has perfectly good mechanisms to shut off your immune response and cancers take advantage of that, but that phenomenon is not in the cancer. It’s in the surrounding tissue. You have to look in the surrounding tissue to see what’s going on there.
Then, you have to look at the immune system because the immune system is the thing that’s going to kill the cancer cell, and people have different immune systems. It’s very clear that there are lots of genetic variabilities in that. Maybe that genotype within somebody’s immune system is not good at getting activated by this immunotherapy. Maybe we should try something different. It’s another level of complexity that precision medicine has a tremendous role in, but it becomes that much more complicated because now you’re not just looking at the cancer, we’re now looking at the cancer, the cancer’s effect on its surrounding microenvironment, and the immune system’s ability to target that cancer and perhaps live in that environment that surrounds the tumor. It’s an additional level of analysis and complexity. That’s coming though. With the expansion of immunotherapy that will be a much bigger piece of what we do in terms of therapy. Those kinds of analyses and guidance by precision medicine will be a key component of how we deploy immunotherapy in patients with cancer.