Rationale for minimal residual disease (MRD) assessment and role in treatment decision-making for patients with multiple myeloma (MM)

DR LOVE: Welcome to Oncology Today, current role of minimal residual disease assessment in the management of Multiple Myeloma and Chronic Lymphocytic Leukemia, this is medical oncologist Dr Neil Love. For this program, I first met with Dr Shaji Kumar from the Mayo Clinic in Rochester, Minnesota to talk about the use of MRD in multiple myeloma and to begin I asked him to discuss the rationale for MRD assessment.

DR KUMAR: So, minimal residual disease assessment is something that has been used extensively in hematological malignancies, most importantly in the setting of acute leukemia. It has been used for changing therapy as well as in CML.

Now in myeloma, MRD testing has been employed for quite a while, but it has really become mainstream only in the recent years since the updated IMWG response criteria was published back in 2016. And broadly, we have used DNA sequencing based approach or a morphology by flow cytometry and both these technologies have evolved considerably over the past few years as well.

So I think where we are right now with myeloma is that we are confident that we can detect the presence of 1 in 10⁵ or 10⁶, basically 1 in a million myeloma cells, with good degree of certainty in the setting of myeloma.

Now, this has all been based on bone marrow aspiration or bone marrow studies and obviously, that has a disadvantage that it could be patching involvement and you could have a false negative test even though there’s residual cells. There is ongoing studies looking at evaluating the peripheral blood for minimal residual disease or evidence of residual disease, both using the flow cytometry approaches as well as using circulating tumor DNA.

So far, it seems like circulating tumor DNA probably gives only about 1 log lower level of detection than what can be achieved in the bone marrow. So there are other approaches looking at the monoclonal protein by mass spectrometry which can get to about 10- to 100-fold more sensitive than what has traditionally been achievable with the immunofixation.

DR LOVE: So, before we go into a lot of the specifics, just to take a step back in terms of the concept of attempting to minimize the burden of disease, which has really been a concept out there.

And coming from the perspective of the general medical oncologist where in solid tumors often you use single agents in a non-curative situation, people have always kind of questioned, is it really going to play out in the long run, particularly as it relates to survival. Any thoughts about that?

DR KUMAR: So I think we have to think about, from a general practitioner, general medical oncologist perspective, I think there are two important questions — I mean the one most important question is what is the value of the MRD test result for the patient? And two, what is the value for the practitioner who is treating the patient in terms of making treatment decisions?

So, from a prognostic standpoint, which is what the data tells the patient, there is wealth of evidence now from multiple Phase III trials and also from meta-analysis, including one recently published in Blood Advances, which shows there’s no doubt the people who get to be MRD-negative had a better progression-free survival, they have a better overall survival in newly diagnosed disease, relapse disease, and also in the transplant setting and the non-transplant setting for the newly diagnosed patients. So, the prognostic value is very clear.

The second question that the practitioner will wonder about with the data in front of them is what do we do with it? Are we going to change anything? And I think that's where we don’t really have enough data to say that if you are MRD-negative, can we decrease the treatment or stop the treatment? Or, if you’re MRD-positive, do we need to treat you more?

DR LOVE: Can you talk a little bit about the conventional clinical approaches we’ve had to measuring minimal residual disease?

DR KUMAR: So, in the past, we have looked at either the protein that the myeloma cells secrete, or the myeloma cells themselves in the bone marrow as ways of measuring the disease burden. And the way to detect the monoclonal protein was in the majority of the patients using serum protein electrophoresis with immunofixation or in the urine using protein electrophoresis or in patients who only make light chain using the serum free light chain assay. And, also, in addition to this, we would also look at the number of plasma cells in the bone marrow.

Now, combining all these different parameters, we used to group these patients as having achieved a partial response, very good partial response, complete response or a stringent complete response, again depicting an increasing depth of response. Now in patients with a partial response, it is basically a 50% reduction in the tumor burden. In the VGPR, again a 90% reduction in the tumor burden measured by the serum or urine protein studies. And when you no longer can detect the monoclonal protein in the serum or the urine with immunofixation, and your bone marrow plasma cells in the bone marrow is less than 5%, you have a complete response. And if you are able to demonstrate that the residual plasma cells are completely not clonal, and you have a serum free light chain ratio that's normal, then the term becomes a stringent complete response.

Now, when we revised the response criteria, particularly because we knew that the patients who are even in that stringent complete response, they still can begin to progress and the median progression-free survival in many of these studies on the order of 4 to 5 years. So that tells us that there’s still significant amount of tumor cells left behind and we need to detect those tumor cells. And that is where the need for the minimal residual disease testing came on. And again, with the experience from before showing that MRD can be detected using flow or next-generation sequencing, and that it does predict survival outcomes, it was incorporated as another layer of response in the setting of myeloma. And that's where we said, if somebody has to be in a complete response and their bone marrow should have less than 1 in 100,000 myeloma cells, detected either by sequencing or by flow cytometry, to say that a particular patient has achieved MRD-negative status.

And one thing, which, again, many of these initial criteria did not include, was the role of imaging. And we know that in myeloma we can have extramedullary disease which may not be detected by the conventional methods. So, the new response criteria also incorporated a MRD + imaging MRD negativity or negativity status, which essentially means negative PET, negative MRD.

Defining and measuring MRD; assays for MRD analysis in MM

DR LOVE: Can you talk a little bit about how MRD is actually measured and defined?

DR KUMAR: Absolutely. So, the way we measure MRD, it's built upon the original foundation of the conventional responses. So we certainly have to measure the M-spike with the electrophoresis, and that has to be negative for somebody to be MRD-negative. And then the second piece is, you look at the bone marrow plasma cells and there are two that have been taken. One is using the flow cytometry and this what we call the next generation flow and this is something that has been again developed by the European groups and uses a multiple antigens in a test-tube setting. And this can detect myeloma cells with the sensitivity of 1 in 100,000 to 1 in a million. And the next approach is next-generation sequencing. And this one basically looks at the immunoglobulin’s which, again, we know is unique for the tumor clone and by determining the identification of that clone in the newly diagnosed setting or the baseline, you can then follow the sample over time if that clone has disappeared with the sensitivity of, again, up to 1 in a million.

Now, the serum protein electrophoresis also has gone to a next phase because now we can use mass spectrometry to identify these very, very small amounts of monoclonal protein in the serum, probably about 10- to 100-fold more sensitive than the serum protein electrophoresis. Now, how does this correlate with the flow cytometry-based approaches? We still don’t have a lot of data to say if they are replaceable. But the data that we have so far is that they are probably have to be used in conjunction with each other.

So, they’re hoping in the future that the minimal residual disease definition will not just include the serum protein electrophoresis or immunofixation-based assessment of the monoclonal protein, but rather use a mass spec-based approach to detect monoclonal protein in addition to either flow or next-generation sequencing-based approach to the bone marrow plasmacytosis. And then finally, the imaging studies, particularly the PET scan-based imaging that would be a valuable adjunct to these 2 measurements.

Incorporation of MRD status into the revised International Myeloma Working Group response criteria

DR LOVE: Can you talk about how this has now been incorporated into the IMWG response criteria?

DR KUMAR: So, when they redefined the response criteria back in 2016, it had 3 main focus. Well, the first was to define what does it mean to be bone marrow MRD-negative? And that we defined as less than 1 in 100,000 nucleated cells. Now, how can we get there? We can get there by using a next generation flow cytometry or a next-generation sequencing approach, both of them will give this information.

Now, there are studies that have compared these two and there’s always going to be some picked by one versus the other.

So they seem to be quite comparable for what we need for defining the MRD-negative, based on this definition.

There are ongoing studies that are looking at 1 in a million to see if that translates to better outcome. And it seems like the deeper you go, the better the outcome’s going to be. But right now, the criteria uses a 1 in 100,000.

The second important aspect of the criteria was the inclusion of the imaging studies with the PET CT. And the third one is a very important concept in myeloma, which is the sustained MRD-negativity. Now, before we started doing MRD testing routinely, there’s a concept of this sustained complete response where the patients not only achieve a complete response, but stay in the complete response for year or two and those patients actually do the best.

So the same concept now has also been shown, and there were abstracts at ASH and some recent publications showing that not only getting to MRD-negative, but maintaining that MRD negativity, at least for a year, seems to predict for much better outcomes compared to if you lose it rapidly.

So, one other important aspect of the response criteria was that we know that the stringent complete response we just developed, was based on again the normal free light chain ratio and no clonal plasma cells in the marrow by less than sensitive techniques and what we’re using right now. So, in a way, by going to the MRD negativity, we have lost the value of the stringent complete response and we anticipate that this particular response segment or level, will probably disappear over time and we will be focusing primarily looking at a lot of the immunofixation or mass spec and the MRD test in the bone marrow.

DR LOVE: So, just to clarify. Right now, does it seem that flow and sequencing are producing similar outcomes?

DR KUMAR: Yes. Both of them have — they’re equally capable of defining the 1 in 100,000. Whether both can reliably detect 1 in a million, again I think the sequencing may be slightly more sensitive than the flow, especially in the setting where we don’t have a significant number of cells collected in the bone marrow aspirate. But if you are able to do flow on about 2 to 5 million cells, then we are approaching parity in terms of the sensitivity.

Perspective on the use of MRD testing in routine clinical practice

DR LOVE: Before we go on and talk more about the rationale here, of course we’re thinking about the oncologist in community-based practice, just kind of curious at this point do you see a role for MRD outside a clinical trial?

DR KUMAR: Now that's an important question and I think that we all struggle with every day. I think the important question is, to do something outside of the clinical trial in the routine practice, you should be able to act upon the result. And so, that's where we don’t have prospective trials showing that if you act on the result your outcome is going to be different.

Now, having said that, we also have now increasing amount of evidence showing that patients with high risk abnormalities or high-risk myeloma, those patients actually have a much better outcome if they’re able to get to MRD-negative, or I should say that the high-risk patients getting MRD negativity have outcomes that are comparable to standard risk patients.

Now we have extrapolated the data to say that given that most of our current approaches seems to not benefitting the mass standard risk patients, we should try and get them to be MRD-negative.

So in my own practice, I think for patients with high-risk myeloma I routinely do MRD testing at the end of a planned therapy or a planned phase of therapy to see whether I should do something different to try and get them to be MRD-negative. But for the standard risk patients, I think there is a possibility that we could potentially harm some patients by escalating therapy based on MRD results because we all have patients with standard-risk disease who may have residual M-protein who still go 4, 5, 6 years without disease progression.

DR LOVE: So, it certainly does fit into the longstanding paradigm of treatment of myeloma. I know a lot of docs in practice are ordering it. What are the options available in terms of specific assays? And is there any difference in terms of reimbursement, or accuracy, or any other issues?

DR KUMAR: So right now, commercially, people can order a flow-based assay. The majority of the flow-based assays are usually done by the larger institutions, stem cells for in-house samples. For example, in our own lab, we know that outside samples come in for MRD testing by flow. There’s also the next-generation sequencing test that's available commercially that also can be ordered in the clinical practice.

Now, as I said before, I think for the way we define MRD negativity right now, they’re fairly quite comparable based on all the data we have. So I think both can be ordered. Now one of the disadvantages with the next-generation sequencing is we do need a baseline sample in order to identify the clones so that we can track them. Whereas, flow cytometry-based approaches, we can, even if there is no baseline sample, we are still able to try and identify a clone.

Impact of MRD status on outcomes in the IFM 2009 trial evaluating early versus late autologous stem cell transplant (ASCT) for patients with newly diagnosed MM

DR LOVE: So, let’s talk about what we know in terms of MRD and its value in clinical trials, looking at treatment efficacy. Of course one issue that we’ve been hearing about, to me — I hear a lot of people in practice bringing up this scenario — transplant versus not. So let’s talk about what we know there.

DR KUMAR: Definitely. So I think what we know for sure, again as the IFM trial had shown, that patients getting the stem cell transplant have a higher rate of MRD negativity. And this clearly translates to a better progression-free survival. And there was updated data from the IFM from 2009 at ASH and showed that bortezomib/lenalidomide/dexamethasone induction therapy, single transplant, one year of lenalidomide maintenance, we are looking at almost a 4.5- to 5-year median PFS for these patients.

Now, one thing that is interesting, and the question that always comes with these, what if you had the induction therapy and you’re already MRD negative by stem cell transplant. And the answer from the IFM trial would be that even in patients who get to be MRD-negative, transplant still adds benefit. And even in patients who get to be MRD-negative at the end of the stem cell transplant, if you ask the question, do you need maintenance or don’t need maintenance, again, the MRD-negative patients still seem to have benefited from some additional therapy.

And I think part of it is related to the fact that — it's not only a question of getting to be MRD-negative, but also staying there, and that's probably where the maintenance therapy allows patients to continue with that response.

DR LOVE: When the data was first presented from the IFM French trial, it kind of looked like if the patients were MRD-negative they didn’t benefit for a transplant. With this new dataset, were there new findings there that suggested it was more effective?

DR KUMAR: No. I think both MRD-negative and MRD-positive patients seems to have benefitted from the continued therapy. And I think the important thing is also to look at this whole thing as a single package of treatment. I think it's the 3 phases, all of it contributes to deepening response across time. And finally translates to that overall improved PFS that we see with this modality.

DR KUMAR: Even in the FORTE trial, where patients who got the KRd versus KRd plus transplant, they both seem to have almost similar, at least in the initial ones, almost comparable depth of response by MRD, but the patients who got the transplant actually had more durable responses. Again suggesting that it does add to contribute to the durability or response.

Effects of MRD assessment on therapeutic decision-making in MM

DR LOVE: What about MRD as a way to assess the difference between conventional treatments, for example, Rd versus DRd?

DR KUMAR: So I think this is the place where I think the MRD assessment is going to have the maximum impact in myeloma that is in context of clinical trials before it becomes fully accepted in routine practice. So, two things. One, I think it gives us a very good way of comparing between different regimens. And now with almost all clinical trials routinely incorporating MRD assessment, we are starting to see that all these trials, addition of newer drugs, for example, like the MAIA trial where the daratumumab was added to lenalidomide/dexamethasone. Or even the CASSIOPEIA trial where daratumumab was added to bortezomib/lenalidomide/dexamethasone. All these trials have shown that the MRD negativity is a very good way of assessing the efficacy of their treatment approach.

So, there’s increasingly a push and a desire to have MRD as a surrogate endpoint for long-term outcomes in myeloma. So, what is happening is, right now with the newly diagnosed myeloma initial therapies, we are talking about PFS that's exceeding 4 or 5 years and portably the overall survival will likely exceed 10 years. And it will be impossible to do clinical trials with this kind of endpoints.

So, we are hoping that there would be approval by the regulatory authorities to use the MRD negativity as a surrogate endpoint so that we can have early read outs, which will allow us to do these clinical trials much more effectively, and also bring the newer therapies, as we saw at ASH — a whole bunch of them are out there right now — we want to try and bring them up to the earlier settings.

DR LOVE: What’s been the response of regulatory authorities to this concept?

DR KUMAR: They have been quite accepting of the whole concept. I think what we really need is a very formal analysis demonstrating the consistency of the finding. Also, what they want to see is a quantitative relationship between these and the actual outcomes.

So, there are a couple of ongoing efforts that are being driven by different groups where we have combined the data from multiple Phase III trials across the globe and they’re hoping that analysis of that combined dataset will move us forward in terms of, again, submitting to the regulatory authorities a formal request for approval.

DR LOVE: Can you see any other path to this being accepted, other than its relationship to survival? I would imagine theoretically that these patients would have better quality of life with delayed relapse. Are we going to need to show correlation of survival?

DR KUMAR: I believe we will have to show the correlation with survival, primarily because those have been the traditional endpoints that the regulatory authorities have been looking at.

Now you also bring up a very important point in terms of the quality of life. Now, the problem in myeloma right now is that most patients continue on therapy until disease progression. And so, whether they are MRD-negative or -positive, they are still on therapy. What we really need is data showing that if you’re MRD-negative and you can stop therapy, then we can certainly make the argument that you have lower cost, better quality of life, and can we use that as a way to lead to approval of drugs and treatments.

DR LOVE: Interesting.

What do we know right now about how MRD can be used to change treatment?

DR KUMAR: So, there are a lot of clinical trials ongoing right now trying to ask that question. For example, there’s a trial going on through SWOG that is looking at daratumumab/len maintenance post-transplant versus len maintenance. And in that trial, after 2 years of maintenance, patients who are MRD-negative are going to be randomized to discontinuing treatment versus continuing on treatment and asking question, is MRD negativity a good way to stop therapy?

There have been other smaller trials like the MASTER trial that they have looked at in the concept — or in the context of newly diagnosed myeloma therapy to see if we can discontinue treatment if you become — sustained MRD negativity.

And there is also trials, like what the ECOG is trying to do with the OPTIMUM trial where patients, after 1 year of len maintenance, if they are still MRD-positive, they are going to get randomized to continuing the same maintenance versus adding a second drug to the lenalidomide maintenance to see if we escalate therapy at that point, can we decrease the likelihood of failure.

So, those are the prospective trials that are ongoing. But, until then, I think based on what we know, most of us feel quite comfortable using this in the setting of high-risk disease. So if you have somebody with high-risk myeloma, they had a stem cell transplant, and they come back day 100, they are still MRD-positive, would we do something different for that patient? And I think that's one area where I think, especially with some of the recent data showing the role of tandem transplant, we might alter a therapy to say okay, maybe we want to consider a tandem transplant in this high-risk patient who’s still MRD-positive after the first transplant.

Again, very limited data, mostly in the context of trials where we have done subgroup analysis. But I think the data is convincing in the sense when you look at patients with adverse cytogenetics, who tend to become MRD-negative, their outcomes get to be pretty close to the standard risk, especially in the more recent studies that have been done.

Consideration of the benefits and risks of treatment as factors in determining the value of MRD assessment

DR LOVE: You had a slide in one of your talks comparing the biology versus therapeutic issues in MRD. Can you comment on that?

DR KUMAR: I think one of the mistakes that we make is we just think about MRD in absolute terms. But I think it's very important for us to consider that in the context of the disease biology, and as we have seen, obviously, the patients with high-risk disease seems to have the maximum benefit from getting to be MRD-negative.

Now, whether that could be a reflection on the fact that the lower the clone size, less likely they’re going to get newer mutations and that may be one of the reasons why those patients derive more benefit.

Now, at the same time, we also know that more intense therapies can lead to deeper responses, but they also can have more treatment toxicity, which might, in turn, limit the duration of treatment that you can give to these patients.

So it's a complex interplay between the disease biology on one hand and the treatment toxicity and the efficacy of the treatment on the other hand. And it's that balance that basically — or those factors which determine how much value the MRD results are going to have in a given patient.

DR LOVE: In CLL, we have a situation, to me, that is a little bit hard for me to understand, but a very interesting, where you have two very effective therapies, venetoclax-based therapy and Bruton’s tyrosine kinase, BTK therapy, both extremely effective, almost a tossup clinically. And yet, BTK usually doesn’t lead to MRD and venetoclax does. Do you have any similar situations in myeloma where some drugs tend to lead to MRD more than others, or some treatments lead to more MRD?

DR KUMAR: No, I don’t think we have a correlative situation in myeloma. But in myeloma, it's a question of the combinations. So I think one of the highest response rates or the MRD rates that we have seen so far is when you use a 4-drug induction therapy followed by an autologous stem cell transplant. Now it might be different when we talk about drugs which have very specific targets, for example, venetoclax.

You can actually get MRD-negative disease even the use of just venetoclax and dexamethasone without intense therapies.

So, I think we’re going to find both paradigms in the setting of myeloma.

So, the setting of immunotherapy, especially the CAR T and the bispecifics, we are seeing really deep responses, again without use of combinations, and including MRD negativity. So, basically, we have 3 situations here. One, highly intense therapy, therapies like immunotherapy or a therapy that is targeted towards a disease biology like the venetoclax and translocation 11;14.

Optimal time for evaluating MRD status for patients with MM; MRD status as a surrogate endpoint in clinical trials

DR LOVE: So, let’s talk a little bit about practical issues in using MRD, beginning with when you should order it.

DR KUMAR: So, currently the true MRD negativity is defined as bone marrow MRD test that is negative and a serum/urine immunofixation that is negative. So the current recommendation is to order the bone marrow test once the immunofixation in the serum and the urine becomes negative. Because that way, you avoid having to repeat the bone marrow. It has become a little bit of a challenge because we know some of the immunotherapy approaches like CAR T can give a patient MRD negativity within a month of treatment while they’re still have monoclonal protein. And it is partly because the monoclonal protein takes some time to get out of the body because of some of those endothelial recirculation that happens.

So the disadvantages of delaying is that you won’t call somebody bone marrow MRD-negative until you actually have the bone marrow. So that's a little bit of a conundrum at this point. But at the same time, from a patient perspective, if you want to do one bone marrow and get the result, that is most likely, then you want to do it in when the serum and the urine immunofixation becomes negative.

So, outside of clinical trials, I think our recommendation is wait until the immunofixation is negative before you do it.

DR LOVE: What about repeat MRD assessment?

DR KUMAR: It's the same issues as with the original one. Obviously, we want to be convinced that you’re going to act upon the test result before you order that. And the same principal certainly applies here as well.

The advantage of repeating an MRD testing is it gives us a good sense of the sustainability of the MRD in a given patient. We know that if a patient who was MRD-negative a year ago and a year later, now the patient is still MRD-negative, we know that the outcomes are going to be very good.

Now, this is the approach that is also being taken in some of the clinical trials where the discontinuation of therapy is not based on a single MRD negativity, but rather a sustained MRD negativity over a couple of different occasions. But I think in the non-clinical trial setting, again, I do not recommend repeating MRD testing on a regular basis to see if it is starting to become positive because we don’t even know whether we should start treating somebody when they become positive versus what we do currently.

DR LOVE: How does imaging response fit in here? Do you have patients who don’t appear to be completely responding on imaging who are MRD-negative? Is imaging really an effective way to measure response?

DR KUMAR: It is. And I think as we, especially in the relapse setting where we have a lot of extramedullary disease, PET CT can be quite useful. Now, both MRI and PET CT have some unique advantages for each of them. However, what the French group had shown in the context of the IFM 2009 is that the MRI changes tends to take a lot more longer compared to the PET CT and the PET CT appears to be a much more dynamic marker in terms of assessing response in the short term.

DR LOVE: Anything you want to say about where you see things heading with MRD assessment, and particularly how it's going to be looked at in clinical trials?

DR KUMAR: I think the most immediate thing that's likely to happen is that MRD is going to become a surrogate marker or surrogate endpoint in the context of clinical trials, I would anticipate that happening in the next 2 to 3 years.

I think we have already adopted the routine clinical use of MRD testing, especially in the high-risk patients. Now, we will definitely get data in the near — hopefully, in the next few years, in terms of its practical value in altering therapy, and once that happens, I think the community will embrace the MRD testing in whole.

Then the next thing is expanding on the methodologies. Now the problem with the current MRD testing is that you have to do a bone marrow and that is obviously a painful procedure for the patients. So I anticipate that in the next 3 to 5 years we will be using or developing blood-based testing, whether it be tumor DNA, mass spec-based assessment of extremely low levels of the protein in terms of determining whether the patient is MRD-negative. And that will also allow us to do serial monitoring of these patients.

DR LOVE: I want to clarify one thing. Did you say that venetoclax in myeloma leads to MRD negativity more commonly than other therapy?

DR KUMAR: Well, in the 11;14 patients — so I cannot say it's compared to other therapies, but for a single drug to produce MRD negativity, that is something we see with venetoclax in the setting of 11;14 translocated patients.

DR LOVE: As long as we’re bringing up single drugs, what about daratumumab and anti-CD38 in general?

DR KUMAR: So, as a single agent, the MRD negativity, yes, we do see some but again, most of it have been studied in the setting of relapse disease where the majority of the patients are less likely to get to be MRD-negative.

Case: A man in his early 60s with t(4;14) MM attains MRD-negative status after treatment with daratumumab/bortezomib/dexamethasone/lenalidomide followed by an ASCT

DR LOVE: So, let’s go through some of your cases. I know these are from your practice, beginning with this 61-year-old man.

DR KUMAR: So, this gentleman basically presented with some rib pain. Was diagnosed with myeloma, predominantly light chains, kappa light chain myeloma and had almost 90% plasma cells in the bone marrow with a translocation 4;14. Multiple lytic lesions on imaging studies. Elevated LDH. Elevated beta-2 microglobulin.

So, this patient initially received treatment — again, presented with some renal insufficiency as well. Received 1 cycle of — 2 cycles of daratumumab/bortezomib and dexamethasone and then after cycle 2, we added lenalidomide. So, essentially a quadruplet, 4 cycles. Achieved a complete response. Then underwent stem cell collection and we actually collected enough cells for this patient to have up to 3 transplants, thinking that we might go in for a tandem autotransplant, based on some of the more recent trials showing the benefit.

So, underwent the single stem cell transplant. Came back day 70ish, 200. Was MRD-negative at that time. And so, we had a discussion as to whether we should be doing a tandem or do we go directly to maintenance? And based on the fact that he was already MRD-negative, the patient preferred to, and I agreed, to proceed directly to maintenance. And the maintenance that we wanted to do for this patient was a PI plus IMiD. So we talked about using bortezomib/lenalidomide, but he wanted to continue to travel, so he decided to go in for ixazomib instead of bortezomib so that he can continue to maintain his lifestyle and that's where he is right now.

DR LOVE: So, how long has he been on maintenance? And have you repeated the MRD or do you plan to repeat it?

DR KUMAR: I plan to repeat that. Again, depending upon whether we are planning on discontinuing therapy at some point. Right now, he’s only about 7 months out from transplant maybe at a year out, depending upon how the clinical setting, we might bring that up again. But if at some point we are talking about discontinuing one or more of these, then I think the MRD testing at that point might be of value, again showing that this has been sustained over time.

DR LOVE: Let me ask you another question because we always think of these ideal scenarios, but what about the patient who’s not tolerating therapy well? I’m not talking about high risk/standard risk situation. For example, maintenance therapy, they’re having a lot of trouble with lenalidomide. The patient really wants to stop therapy. Do you think MRD could be useful in that situation?

DR KUMAR: It's an important question. And again, as we talked about before, the lack of prospective data definitely does limit what we can do.

Case: A man in his mid-60s with MM receives bortezomib/lenalidomide/dexamethasone and an ASCT followed by maintenance lenalidomide

DR KUMAR: We can talk about one other patient of mine who originally had presented with smoldering myeloma and then about 6 years later he actually developed symptomatic disease, new lytic bone lesions. Was started on bortezomib/len/dex, 6 cycles induction therapy, VGPR. Then went on to have a single autologous stem cell transplant. Was basically immunofixation, only positive. Flow cytometry. Had an extremely tiny 0.0004% monotypic lambda plasma cells.

The patient was started on len maintenance. And again, this was a standard-risk myeloma, and subsequently became immunofixation-negative as well.

Now a year later, now he has this terrible diarrhea, which is unpredictable. Comes for a few days goes away. Has had extensive GI workup. No clear evidence of any other pathology. This is slightly related to the lenalidomide. So we basically talked about discontinuing therapy. We repeated some of the testing. He’s MRD-negative in the bone marrow. PET CT-negative. Immunofixation-negative. So we decided, based on this particular setting, especially that his MRD negativity is almost sustained now, we decided to discontinue the len and he felt much better after discontinuing the lenalidomide.

DR LOVE: Just out of curiosity, do you consider in patients who, not so much related to MRD, but just patients who are having difficulty on len maintenance, do you consider switching to a proteasome inhibitor like ixazomib alone without an IMiD or another IMiD?

DR KUMAR: I think that would certainly be a consideration. I think there’s been a little data with pomalidomide as a maintenance therapy, so it's not something that we have used as much. But the ixazomib as an oral proteasome inhibitor or bortezomib given every other week, carfilzomib given weekly or every other week, have all been tried in the setting of maintenance. Certainly something worth considering. I think, given this particular patients’ scenario, given the standard-risk disease, we didn’t feel that compelled that we wanted to try something different. Had it been a high risk disease, we could argue he may already have been on the PI to start with. But in someone, still had residual disease on the testing, then I certainly would have considered switching to another drug.

Case: A woman in her early 60s with relapsed/refractory MM receives daratumumab/lenalidomide/dexamethasone

DR LOVE: So, let’s finish out with your 60-year-old woman. What happened with her?

DR KUMAR: So, this is another patient who essentially presented with bone pains, hypercalcemia. Had a IgA lambda M-spike and 40% bone marrow plasma cells.

So this was a standard risk myeloma with trisomies, basically hyperdiploid clone. Had CyBorD, bortezomib/cyclophosphamide/dex induction therapy, single autotransplant, len maintenance. So 3 years later she presented with increasing IgA. At that time, she basically was started on dara/len/dex. Now this patient, just to note, had discontinued the maintenance after 1 year. So she was not refractory to len.

Now, after 1 cycle she was already in a VGPR. And after 12 cycles, we repeated the MRD testing, or repeated the bone marrow, and the MRD testing by flow showed 0.004% plasma cells. No other evidence of disease. And the main question for her was, do we just continue with all these drugs or can be actually kind of dialed down on the therapy to go down to more of, either 1 drug or a lower dose of these drugs. And I think here again, I think there is very little data, particular direction is better than the others, but I think the showing — the fact that she has very little disease in the bone marrow is very reassuring. And I think I feel very comfortable switching to more of a maintenance mode here, again with a much lower dose of the len, with the hope that hopefully we can discontinue one of these.

DR LOVE: So, what’s the follow-up on the patient? Have you already decided what you’re doing?

DR KUMAR: Yes. So this patient is now continuing on the monthly dara and is actually on 10 mg of len for 3 out of 4 weeks. And my hope is that if in another year we may be able to just discontinue 1 of these, either the monoclonal antibody or the IMiD, again depending upon the patient preference.

DR LOVE: So, I mean here, whether you’re extending survival, and of course who knows whether, in the long run what’s going to happen, but I guess you could say you’ve improved her quality of life. I imagine she’s maybe happier now.

DR KUMAR: Absolutely. Yes. I think that's the key thing. A lot of the trials really — it's close enough, integrating that quality of life, the depth of response and the survival into one simple formula, which is hard.

DR LOVE: Any final thoughts you have?

DR KUMAR: I think the MRD testing is really a big step in myeloma. And I think increasing tools and increasing data from different trials is going to help us use this as a — I think it's going to be a very valuable tool in the management of myeloma patients in the years to come.

Role of MRD assessment in the management of chronic lymphocytic leukemia (CLL)

DR LOVE: I next met with Dr Philip Thompson from the University of Texas MD Anderson Cancer Center to talk about MRD evaluation in CLL, and to begin I asked him to discuss the background behind the use of MRD assessment.

DR THOMPSON: So essentially the whole concept of minimal residual disease is the ability to detect a malignant clone that is present below the threshold for normal morphologic analysis sensitivity. So using acute myeloid leukemia as an example, if you look at a post-induction chemotherapy marrow, complete remission is defined as having less than 5% blasts with recovery of normal hematopoiesis. And you still have a very large leukemic burden if you have 4% or 1% blasts. And if you still have disease but undetectable with standard techniques and you don’t give further therapy, then the patient is going to relapse.

So in acute leukemias that has enormous therapeutic significance in that the disease grows very rapidly and you want to try and eradicate that and move to some kind of consolidation therapy such as a transplant in order to try and cure the patient. In CLL, in days gone by, we didn’t have highly effective therapy, so the likelihood of achieving very deep remissions was low. But also there was a little bit of therapeutic nihilism, which people didn’t try to achieve deep remissions because they felt they could just kind of manage this disease as a chronic disease with a little bit of chlorambucil, and when the disease grew again a little bit more chlorambucil.

Now, with modern therapeutics, chemoimmunotherapy like FCR, venetoclax, you can achieve deep remissions and potentially prolonged treatment-free intervals and perhaps even cure in a subpopulation of patients. And that has made the ability to accurately detect small amounts of disease important prognostically. So the most established use of minimal residual disease analysis in CLL is kind of post time-limited therapy, like chemoimmunotherapy with FCR, you do a blood test or a bone marrow. The most commonly-used techniques with flow cytometry will detect around 1 CLL cell in 10,000 normal cells. And if you have undetectable MRD at that threshold this is a favorable prognostic feature.

Where I think the field will go, and we’re probably not quite there yet, is that we will see adjustment in the treatment strategy for an individual patient based on MRD results. At the moment it’s best established as a prognostic tool post treatment, but increasingly I think we’ll use it serially to be able to track individual patient’s disease burden over time, get a sense of in vivo sensitivity to the therapeutic regimen the patient’s on and potentially adjust the duration of therapy and/or add or remove therapies based on the results. I hope that’s where we will get to.

DR LOVE: Again, at maybe a macro or naïve level, it kind of reminds me of tumor markers in solid tumors, PSA, CA-125. Why is that that kind of approach doesn’t work with the hematologic cancers?

DR THOMPSON: Well, I mean we have some kind of crude biomarkers in some hematologic malignancies. For example, beta2 microglobulin is an example. So in CLL, in multiple myeloma, that has been shown to be of prognostic significance, the level of the beta2 microglobulin. Also, we’ve shown in patients receiving ibrutinib therapy that the degree of reduction in beta2 microglobulin during therapy correlates with progression-free survival. So it’s just that it’s quite a crude biomarker. It’s dependent on renal excretion and things like that.

The good thing about MRD measurements is you’re directly measuring the tumor cells. And so that provides a very precise measurement, quantification, of disease burden. You don’t have to infer anything from the results. So this is a quantitative test. And in multiple studies after chemoimmunotherapy, after venetoclax-based treatment, it’s been shown that quantity of MRD matters. It’s not just detectable versus undetectable. Essentially, the less you have the longer your remission is going to be, which is not particularly surprising.

It’s not the only thing that matters. So traditionally the way that we assessed disease debulking if you will, after treatment for CLL, was with physical examination, routine lab work, and CT scans, particularly looking for reduction in all lymph node sizes to 1.5 cm or less and normalization in the size of the spleen and liver, and then absence of a morphologic infiltrate in the bone marrow. And so patients who had a complete remission, that fulfilled all of those criteria, would have longer progression-free survival than patients who did not achieve that.

But interestingly, when you look at data sets from, for example the CLL8 study of FC versus FCR and more recently the MURANO study of venetoclax/rituximab versus bendamustine and rituximab, you can see there are a large number of patients who are what we call partial remission but with undetectable MRD. So they’ll have a lymph node that’s bigger than 1.5 cm or they’ll have mild splenomegaly. And when you look at the group of patients that has complete remission with undetectable MRD compared to the group of patients with partial remission and undetectable MRD the outcomes are very similar in terms of progression-free survival. So it certainly seems that the MRD test, which is very, very objective, is a better prognostic determinant than CT scan.

But it’s not the only thing that matters. I kind of use the analogy with a patient when I talk to them about MRD of using this as a tool to predict how long they’ll stay in remission. We need to know, number 1 how much disease is leftover in the most precise way that we can. But we also need to know how fast it’s going to grow. I tell them like if you can imagine a train speeding towards the edge of a cliff the amount of MRD is how far away the train is from the edge of the cliff, and then the rate of growth is how fast the train is going.

So when you look at — there’s some really nice data from the German CLL Study Group CLL14 study, as well as the MURANO study, both of these with time-limited venetoclax plus CD20 antibody therapy, showing that you can kind of track the clonal growth rates after treatment. And in CLL14 you could clearly see that patients with unfavorable disease biology, like unmutated IGHV, 11q deletion, 17p deletion, these patients, their clone grows quicker. And we saw that after FCR as well, and in the CLL8 study. So disease biology still matters even when you know the MRD level.

Impact of MRD status on clinical outcomes and prognosis in CLL

DR LOVE: I remember recently somebody talking about high and low MRD and actually cutoffs for it.

But can you talk about that idea?

DR THOMPSON: Yeah. So this is the concept I was alluding to earlier –

DR LOVE: Right.

DR THOMPSON: — which is this is a quantitative test. It’s not a yes/no answer. It gives you a precise quantity of the residual disease in whichever compartment you’re assaying, whether that’s blood or bone marrow. So the definitions of high and low MRD have really been developed basically to make it I guess easy to have a cutoff. But they don’t have any particular biologic significance. But when you look at CLL8 data, which was chemoimmunotherapy based, and venetoclax/rituximab data from the MURANO study, both of them show that if you have more than 1% MRD, which is high MRD, then your progression-free survival is much shorter than if you have between 0.01%, which is the detection threshold of the tests that were being used, to 1%, also called low MRD, those patients have a better outcome than those with high MRD. But of course if you have undetectable, less than 0.01%, then you have the best outcomes.

DR LOVE: Do you think there’s a logic in this clinically? I know it’s very early, but it’s a really tough decision. You see that MRD and you stop, and people get uncomfortable about it. Is it worth looking at high or low?

DR THOMPSON: I mean this is a really important question. So essentially there was some really granular data generated in both CLL14, which was venetoclax/obinutuzumab, and MURANO, which was venetoclax/rituximab. So they did an MRD test every 3 months during therapy. So you could see different patterns within individual patients as to what happens to their MRD level over time.

There were some people where the first time you looked it was undetectable, and it remained undetectable. But then you have other people where it went down more slowly and then continued to go down. You had some where it went down a little more slowly and then plateaued. And then you have another group of patients where it went down and then maybe even became negative, but then reemerged during venetoclax therapy.

So those are all different groups of patients biologically. So when I think about this, I think that we can infer several things from those serial MRD data. First of all I would say if you get early undetectable MRD, so say after 3 months undetectable minimal residual disease, you’ve got a very sensitive clone and you’re likely to have a very prolonged remission after time-limited therapy.

And certainly there’s data from the CLARITY study, which was ibrutinib and venetoclax, which showed that when they measured MRD after 2 months of venetoclax patients who had a greater than 3-log reduction in their MRD levels were much more likely to achieve undetectable MRD. So not surprisingly the degree of debulking early on correlates with your likelihood of achieving undetectable MRD, and then probably with the likelihood of long-term remission. So I think if you get early undetectable MRD it makes sense to have a time-limited approach, and perhaps you don’t even need to treat for a full 1 year or 2 years. Perhaps you could stop earlier.

Then there’s the group of patients where — so on CLL14 they showed that at the end of combination therapy with obinutuzumab they showed that there were a group of patients, so about 74% of patients had undetectable MRD in blood at 10^-4sensitivity. Of the MRD-positive patients about half of them, during the next 6 months of venetoclax monotherapy, had continued gradual reduction in the MRD during that time, indicating that they still were sensitive to venetoclax, and their disease burden was going down during continued therapy.

Now to me, that is a group of patients that could potentially benefit from a longer duration of therapy because their disease burden is going down. Perhaps we do them a disservice by stopping at that time, when they might get a deeper remission if we continued the drug. And that would be particularly the case if they’re not having any toxicity from the drug.

I guess if they’re having rising levels of MRD during venetoclax treatment, that indicates the emergence of a resistant clone of cells. Now what to do about that is a little bit of a fraught question because really in the field we don’t intervene on the basis of just asymptomatic minimal residual disease levels. So you can certainly make an argument that well okay, this isn’t actually going to change our management, and for many patients that might be true. But I think there’s a group of patients perhaps with high-risk disease biology, where you may be considering some cellular therapy at some time, like an allogeneic stem cell transplant, where knowing that the disease is present and starting to develop resistance to venetoclax might be important in informing your kind of subsequent treatment decisions.

And then I guess the fourth group of patients is the one that has the kind of plateauing level of minimal residual disease during treatment. And it’s unclear whether that group of patients is going to benefit from long-term maintenance therapy with venetoclax to kind of control the disease from growing or whether you should just stop treatment at the end of a fixed period of time, as was done in the studies. But ultimately if we want to use this to make treatment decisions, we probably are going to have to design some kind of interventional studies that are MRD-directed therapy to be able to show that it makes a difference to patients outcomes.

There is 1 study which is a good example of this, where some data were presented at ASH, and that’s the CAPTIVATE trial. So this was a front-line study of ibrutinib plus venetoclax. So patients were divided into 2 cohorts, one of which was just time-limited therapy for 3 cycles of ibrutinib monotherapy, and then 12 cycles of combination. The other cohort had MRD-directed therapy, so after they completed their 12 months of combination therapy they had an MRD assessment. If they were MRD undetectable at the completion of therapy they were then randomized to either receive ibrutinib monotherapy as maintenance or no therapy and just observation.

And so that will look and see, in the group of patients where you achieve undetectable MRD, does maintenance therapy still prolong the progression-free survival further. So far there have been very few relapses in either arm, so we don’t know the answer to that yet, although all of the relapses have occurred in the no maintenance arm.

And then the second group of patients were the MRD-positive patients, and they were randomized to receive ibrutinib monotherapy as maintenance or ibrutinib plus venetoclax ongoing. And the data that Dr Wierda presented at ASH last year now, showed that you were more likely to deepen your response during that second year of therapy, i.e. become MRD undetectable if you had ongoing ibrutinib and venetoclax combination compared to ibrutinib monotherapy.

So this is just 1 example of a kind of MRD-directed treatment approach. Another example is from the CLARITY study that I mentioned earlier, where they used early undetectable MRD results at 6 months as a trigger to reduce the duration of therapy. So if you had undetectable MRD at 6 months they would stop after 12 months rather than going on for a full 2 years. So there are different ways of doing it, using it as an early-treatment cessation trigger, using it to extend therapy in people who are MRD positive, but ultimately all of these things are — to probably gain wider acceptance we’re going to have to show some kind of time to event endpoint benefit from these approaches.

Rationale for the difference in MRD rates between Bruton tyrosine kinase-based and venetoclax-based regimens

DR LOVE: So I’m just kind of flashing on prostate cancer and how great the biomarker PSA is, and yet they don’t go into any kind of thinking like here, the complexity of the way you guys are thinking about it. It’s so different from the way the GU people do. But anyhow, maybe you can explain something that just kind of seems a little strange to me, which is that BTK inhibitors and venetoclax/anti-CD20, seem clinically like they have some similar benefit. Certainly the benefits of BTK is incredible, and yet the MRD rate with BTK is not that high, and it’s pretty significant with venetoclax.

DR THOMPSON: Yeah.

DR LOVE: So that’s question number 1, why is that. And then question number 2 is it looks to me like the MRD rates when you combine venetoclax and BTK don’t jump up that much. I mean some, but not like from say 60 to 95 or something.

DR THOMPSON: Sure.

DR LOVE: It seems like it goes up — and why hasn’t it gone up? It makes sense if you’re adding BTK. But again, I don’t even know why BTK works so well and yet it doesn’t cause MRD. Cytostatic or something?

DR THOMPSON: I mean it’s a great question. So the first thing I would say is that the way that these drugs have been developed are kind of representative of slightly different treatment paradigms. So if you take a BTK inhibitor like ibrutinib or acalabrutinib for CLL, essentially you’re giving indefinite maintenance therapy. You’re treating the patient in the same way as you would treat a patient with chronic myeloid leukemia, you take the drug indefinitely to suppress clonal growth.

The fascinating thing about these drugs is that, as you rightly pointed out, most patients still have detectable disease. Sometimes it’s just minimal residual disease, but sometimes it’s prolonged lymphocytosis above the normal range, which doesn’t seem to have any negative prognostic implications when it occurs. There was a really nice analysis a few years ago from Ohio State University that showed that genomically patients who have very prolonged lymphocytosis, they often have low-risk genetics, and they don’t have a higher incidence of resistance mutations present. So they seem to be able to suppress clonal growth, but of course if you stop the drug, and you have a significant amount of disease, like say a patient develops atrial fibrillation and you have to stop ibrutinib, they still have 30% involvement in the bone marrow, then of course their disease is going to come back.

In contrast, if you are using venetoclax and you achieve undetectable minimal residual disease after therapy, and then you stop treatment as part of the therapeutic design, you can expect a prolonged treatment-free remission, similarly to what you would get after say FCR chemoimmunotherapy. So different treatment paradigms.

As for why do patients not eradicate the disease with BTK inhibitors when they do with venetoclax, particularly in the bone marrow, that’s a challenging question that I don’t think we’ve really fully come to grips with.

To the next point, venetoclax plus CD20 monoclonal antibodies, in MURANO the MRD undetectable rate was 60% in the peripheral blood in front line; with obinutuzumab/venetoclax 74% in the peripheral blood at 10^-4sensitivity. In the ibrutinib/venetoclax studies, like CAPTIVATE, which is a multicenter study, MD Anderson study led by my colleague Dr Nitin Jain, we saw MRD undetectable rates in the bone marrow with that combination in the high 60s in the MD Anderson study on intent to treat analysis. And in CAPTIVATE I think 72%. So I guess similar rates of undetectable MRD to what you get with venetoclax/obinutuzumab, but you’re comparing blood to bone marrow, and certainly bone marrow analysis is more sensitive for detecting residual disease than blood. So that’s not a kind of absolute direct comparison.

Then the other thing I would say is the complete remission rates, so the likelihood of having all of your lymph nodes less than 1.5 cm, seem to be higher with BTK plus BCL2 inhibition, in the 70s, compared to venetoclax plus antibody, where it’s much lower in the relapse setting. On the MURANO study there were much more partial remissions with undetectable MRD than there were complete remissions with undetectable MRD. And so that seems to be one point of difference, that there’s a greater nodal reduction with BTK inhibitor plus venetoclax than there is with venetoclax plus antibody. Now how important that is, as I was alluding to earlier, certainly the MRD seems to be a more important prognostic marker than lymph node size on CT.

Effect of current CLL treatment regimens on MRD levels

DR LOVE: Just out of curiosity, what’s the MRD rate of monotherapy with obinutuzumab, and also rituximab?

DR THOMPSON: Rituximab, it’s in the low single digits. It’s really not a potent anti-CLL therapy. So there’s no study of obinutuzumab monotherapy, but the best data come from the German CLL Study Group CLL11 trial, where they compared chlorambucil monotherapy to chlorambucil/rituximab and chlorambucil/obinutuzumab. And in that study the MRD undetectable rate in the chlorambucil/obinutuzumab arm was 20% in bone marrow.

DR LOVE: Interesting.

DR THOMPSON: And that’s also been shown in the iLLUMINATE study, which compared chlorambucil/obinutuzumab to ibrutinib/obinutuzumab, and it was in the ballpark, again, there in the low 20s. It was a little higher with ibrutinib/obinutuzumab, up to the mid 30s, I think 35%. So certainly you do get undetectable MRD with obinutuzumab. I mean chlorambucil’s not a particularly effective therapy, as we’ know, so I think most of the heavy lifting is being done by the obinutuzumab there. So you’d expect 20% to 25% just with obinutuzumab.

DR LOVE: And within that, both for obinutuzumab and the other agents, is there a difference in MRD in patients who have a more white count-type disease as opposed to nodal disease?

DR THOMPSON: Yeah, so I’ve never seen data that confirms that. It seems to be that disease biology is probably important, but more so with particular types of therapy. So for example, on the ibrutinib/venetoclax study we didn’t find any kind of biomarker to predict the likelihood of undetectable MRD, so even people with 17p deletion were having similar rates of undetectable MRD to those without. However, when they stopped therapy people with unmutated IGHV and 11q or 17p deletion will grow faster and relapse sooner.

In contrast, with FCR with chemoimmunotherapy you have less likelihood of undetectable MRD if you have high-risk genomics, particularly unmutated IGHV and 17p deletion.

DR LOVE: What kind of MRD rates do you see with FCR?

DR THOMPSON: Yeah, so it’s significantly less than what we’re seeing with venetoclax-based therapy, so a lot of the older trials didn’t report true intent to treat analyses. They often only did MRD testing in patients where they were trying to confirm complete remissions. But I think in the German CLL Study Group, in CLL8, it was in the mid 60s in the blood and the mid 40s in the marrow. So definitely lower than what you get with venetoclax-based therapy in the front line.

DR LOVE: The reason I was asking about obinutuzumab is people talk about the fact that it has a greater effect on the white count than it does on the nodes.

DR THOMPSON: Sure.

DR LOVE: So I was wondering if that was kind of reflected in the MRD.

DR THOMPSON: Yeah, no. I don’t think so. But definitely what you say is true, it’s a much less effective drug to reduce lymph node size than either a BTK inhibitor or venetoclax.

Assays for analyzing MRD in patients with CLL

DR LOVE: All right, so let’s jump into the part of this that always is a mystery to me. Maybe you can explain it. I was thinking you could explain it to me like I’m a fellow, but maybe explain it to me like I’m a patient, which is how you actually do — what kind of assays you do, how the assays are done, and where things are heading in terms of measurement. Also how you choose, for a doc in practice, how you choose which assay.

DR THOMPSON: Yeah. So basically there are 3 main ways that MRD analysis is done. The one that’s by far in widest use is flow cytometry, and essentially the technology is pretty straightforward. So essentially CLL cells have different proteins on the surface than a normal B lymphocyte has. And the concept of flow cytometry is identifying the unique pattern of surface protein expression that a CLL cell has that allows you to distinguish it from a normal cell. And the way you do that is use antibodies that target particular proteins that will stick to the surface of the cell, and they have a fluorochrome attached to it, which is something that when you excite it with a laser it emits light of a particular wavelength that can then be detected by your machine.

And then what you do is you aspirate the sample through a narrow hole in the machine, the laser excites it, and then fluorochromes that are bound to the proteins on the cell will be detected. So you can determine what the pattern of those proteins on the cells that you see are. So essentially the limitations to sensitivity of flow are the number of cells that are in the sample, and then how unique this what we call immunophenotype is or the protein expression of the cell is.

So the most widely used assay was developed in 2007 by Andy Rawstron as part of the European Research on CLL consortium. And that can detect around 1 CLL cell in 10,000 normal cells. So that’s pretty sensitive. But they’ve actually optimized that test, so now the assay that Andy runs in all of the UK studies is actually sensitive to 1 in 100,000 normal cells, so a whole order of magnitude more sensitive than the previous test. So they’re reporting MRD down to 10-5, or 1 in 100,000 on their trials now.

Another way of doing it is what’s called allele-specific oligonucleotide PCR. That’s often done in clinical trials, but it really isn’t done much in routine practice. The reason being is it’s really not suitable for doing in your average lab. You have to sequence the individual patient’s CLL cells to determine exactly what their immunoglobulin heavy chain rearrangement is in the cells that makes the cells unique and different from other cells. Then you have to design a PCR assay to detect that clone. So that’s not widely used.

Another assay which is now available was approved earlier, in I think around August of 2020 by the FDA, is a next-generation sequencing-based test. So the way this works is every person has a huge repertoire of different B lymphocytes, which allows us to recognize any different pathogen that might try and cause us harm. And this diversity comes from rearrangement of immunoglobulin heavy chain variable diversity in joining regions into huge numbers of different possible combinations. So we actually have something like 10¹¹ different possible IGH rearrangements.

And so when you sequence the IGH gene you can see that there will be a huge spike of identical ones, which is the patient’s disease. And this works in CLL. It works in myeloma. It works in acute lymphoblastic leukemia. And so you can then — you don’t have to with next-generation sequencing make a patient-specific test. This is like a universal test. But you need to have a pretreatment sample, which you don’t with flow. So to do NGS-based MRD testing you have to have a pretreatment sample so you know what specific rearrangement you’re tracking, and then you can measure it post treatment, either in blood or in bone marrow.

And this is a very sensitive test, so this can detect down to about 1 in a million cells, so 1 CLL cell in a million normal cells, so really only sensitivity is limited by the amount of DNA that you’ve got in a sample. So I guess it is a limitation that you have to get a pretreatment sample, but you can even scrape a bone marrow slide to obtain the pretreatment sample. So most patients will have something in the pathology lab that you can use for a pretreatment sample if you want to do NGS-based MRD testing.

Case: A man in his mid-60s with CLL and a p53 mutation receives ibrutinib and venetoclax on a clinical trial and attains undetectable MRD

DR LOVE: So let’s talk about some of these patients that you brought in today, beginning with this 65-year-old man.

DR THOMPSON: Yeah. So this is a patient that had CLL diagnosed back in 2009. He received standard chemoimmunotherapy with fludarabine/cyclophosphamide/rituximab. He achieved an unconfirmed complete remission, MRD testing wasn’t done.

He then relapsed in 2016. At that time, he actually had a more extensive chromosomal genomic analysis. He had a TP53 mutation that was identified on next-generation sequencing. So he was started on salvage treatment with ibrutinib.

And then in 2018 he enrolled on a clinical trial that we had open at our center with venetoclax as consolidation therapy for high-risk patients on ibrutinib. So the aim of this study was to achieve undetectable MRD and then have a time-limited therapy. And patients could be treated for up to 2 years, but they had the option of stopping venetoclax and ibrutinib early if they had a rapid response. So another example of a kind of MRD-directed approach.

So at 12 months his CT scan showed that he still has persistent lymph node enlargement, so 1.7 cm lymph node on CT, although it had reduced significantly since starting the venetoclax. Before his treatment with venetoclax began he had about 3% minimal residual disease in the bone marrow, which is kind of within what you would expect from a patient on ibrutinib. But after 6 months of combination therapy he had undetectable MRD in his bone marrow. So this is a rapid responder, very sensitive to venetoclax, no detectable MRD after 6 months. And then at 12 months, again, no MRD.

Now by the trial design, because he was not in complete remission on CT, he was supposed to continue on for 24 months of therapy. So if we only had the CT and not the MRD we would have said okay, he’s a partial response, we continue both drugs. Now he had a couple of equivocal symptoms that suggested possible ibrutinib-induced arrhythmias. They weren’t serious. He did have a couple of episodes of presyncope. He had a Holter monitor that showed frequent ventricular ectopic beats.

So given that we knew that he was a rapid MRD responder with earlier treatment with undetectable MRD we made the decision to stop his ibrutinib. We thought it wasn’t worth taking the risk of continuing the drug and potentially exposing him to the risk of serious cardiac arrhythmias.

So this is just an example of where knowing that this patient had a rapid response using MRD testing, which we wouldn’t have been able to know any other way, gave us confidence to change the management. It altered the risk/benefit ratio of his treatment, and it helped us make a decision about what to do with his ibrutinib.

DR LOVE: So a couple questions about this case, and I don’t know whether you know this from other similar cases, but when you have a patient who’s MRD negative and they have like a node or something, and you biopsy the node, does it look like CLL or is it like dead cells or what’s going on?

DR THOMPSON: That’s a really good question. I have done it once, and we don’t routinely do this, just because it’s invasive and it doesn’t necessarily change what you do. So I had 1 patient on another ibrutinib/venetoclax trial who had this waxing and waning enlargement of an axillary node, and it was a decent size. It was about 3 cm on her CT scan. And we kind of ummed and ahead about what to do this lymph node because everything else was going well. She had developed undetectable MRD. And so we ended up taking the node out, and it was all just kind of scarring and hemorrhagic change, actually having kind of bleeding into the lymph node, presumably because of the ibrutinib.

DR LOVE: Wow.

DR THOMPSON: I was worried that it may represent another disease like breast cancer, which was the main reason to take it out. Because everything else was shrinking. But yeah, so that’s an anecdote. One thing that I was interested in exploring was in other diseases like diffuse large B-cell lymphoma that don’t have a large circulating component and there’s a lot of nodal disease, it’s been shown that circulating tumor DNA can be detected in the blood, and it’s a really useful MRD marker. In fact, it’s probably more useful than PET scanning in diffuse large B-cell lymphoma. And this has been shown in numerous solid tumors, as well, to be a really useful biomarker of treatment response and likelihood of metastasis.

So we were interested in whether in CLL using circulating tumor DNA analysis might identify a group of patients that had disease in lymph nodes but didn’t have it in the blood or bone marrow. So we used a next-generation sequencing MRD assay that I was referring to earlier looking at plasma samples, bone marrow samples, and blood samples in patients after FCR treatment. And we didn’t actually end up finding anybody that had circulating tumor DNA positive who didn’t have detectable disease in the blood or marrow. So I guess at least in that particular setting, in CLL in the front-line treatment, we weren’t able to show that circulating tumor DNA analysis was more useful than traditional MRD analysis in the blood or marrow.

Directly biopsying nodes that are persistently enlarged, or actually surgically removing them, I think would be the best way of answering that question. Doing ctDNA analysis, of course, is noninvasive, but it can’t directly tell you what’s in that lymph node. But yeah, there’s not been a good analysis that I’m aware of to answer that question.

Considerations in the selection of an MRD assay and the value of MRD assessment in clinical practice

DR LOVE: So oncologists in practice who hear about a case like this go “wow, that makes a lot of sense.” And yet I hear a lot of investigators going this is not really ready for primetime. So why is that? And if somebody wants to use an MRD assay in a, for example, situation like this, same thing with myeloma when I was talking to Dr Fonseca, everybody’s thinking what about MRD in patients having problems on maintenance therapy, lenalidomide, whatever. It makes sense.

So is that a rational thought? And if somebody wants to do it, is there a specific assay? What assay do you recommend?

DR THOMPSON: Yeah, so I’m going to answer the second part first. So both flow cytometry and next-generation sequencing are now FDA approved, so you can order either of them. You can order them from the blood, and you can order them from the bone marrow. The challenge, I think, with MRD analysis is that I think people want to have a very like cut and dried decision. So at this particular timepoint in therapy, if the MRD is X, you should do Y, because that will lead to better results. And unfortunately we don’t have those data.

I’ve had discussions with other investigators, where we’ve talked about the importance of MRD. So essentially people have said exactly that. Look, it’s great. It’s really interesting. I can see that maybe in individual cases it’ll be useful, but I’ve never seen a clinical trial that says intervening on the basis of MRD results will make a difference to how long a patient survives. I mean we’ve seen 1 study, the German CLL Study Group did a study with lenalidomide where patients got chemoimmunotherapy. If they had high-level MRD after chemoimmunotherapy they received lenalidomide or nothing. And of course lenalidomide extended the progression-free survival, but it didn’t make any difference in survival.

So ultimately I think what you want to show is that intervening on the basis of an MRD result will improve not just progression-free survival but survival. And it’s going to be really hard to design a study to do exactly that, particularly because, I guess, of the diversity of treatments that we’re giving, all of the different time points we can analyze. And how do you integrate disease biology with MRD results? I think it’s one of those situations where I think probably there’s a lot of art to it. You have to kind of look at it in an individual patient and try and ascertain that individual patient’s risks and kind of think about it and integrate it with their disease biology.

But I guess the other thing that is different about CLL to a disease like acute lymphoblastic leukemia is in acute leukemias by far your best chance of doing well is to cure the patient first, first-line therapy. And in that situation if you’ve got MRD that’s bad. You need to do something about it, get them MRD negative, get them to transplant. Otherwise they’re going to relapse rapidly and die.

In CLL we know you can be MRD positive and just continue ibrutinib or acalabrutinib for 5 or 10 years. So there’s much less kind of urgency, I guess, in terms of acting on some of the results. Which is why I think about the importance of it in particular patient situations, where it may actually change management, particularly those who have high-risk disease biology, and you might actually be needing to think about transplantation. Or alternatively those with low-risk biology who have very rapid responses, and you may be able to stop earlier and avoid treatment-related toxicity.

Perspective on the use of MRD analysis for patients receiving venetoclax with obinutuzumab

DR LOVE: Getting back to this issue of is MRD ever rational outside a clinical trial setting? In general practice? How about the patient who gets obinutuzumab/venetoclax as up-front therapy, getting an MRD assay at a year? Yes or no?

DR THOMPSON: Well I mean I would do it because I like to know. I mean I do it every 3 months because like I was saying to you, I can track that patient over time. I can know how quickly — because it’s not just what it is at 1 year, it’s how quickly does it get there if it’s negative at 3 months.

DR LOVE: I’m just talking about outside a trial, like a doc in practice. They’re using venetoclax/obinutuzumab all the time. So should they get it every 3 months? I mean that’s what you’re doing.

DR THOMPSON: I mean I like to have the information because I think it’s prognostically significant. Do I always act on the results? For example in someone who’s still MRD — I mean what the crux of that question is if a patient is MRD positive at 12 months are they going to benefit from a longer duration of therapy. That would be the main decision that you could make on the basis of an MRD test, a single MRD test at 12 months.

And a lot of people will just say well, there’s no data that tells me that that makes any difference to this patient’s outcome, so I’m not going to do the test. And that’s a completely reasonable way to think about it. But that’s why I kind of was drawing some attention to the data that were presented at ASH which shows that there are different groups of patients. So if you had say MRD of 0.1% at 6 months after completion of obinutuzumab, and then at 12 months, after another 6 months of venetoclax monotherapy your MRD was 0.01%, so it’s continuing to go down. Now that’s a scenario where it makes rational sense that this patient may benefit from a longer duration of therapy because it might continue to go down further.

But if you only get that snapshot at 12 months, and it’s 0.01%, you don’t know if it was negative after 6 months and it’s on the way up, or if it was positive at 6 months and on the way down, or if it’s plateauing. So I think it’s actually harder to make the decision on the basis of a single timepoint in CLL with a therapy like venetoclax in some ways.

But I think it may be problematic for people in general practice if they don’t know the data from some of these studies really well to kind of counsel patients on what the results mean. And they may be uncomfortable with doing that and having to say to a patient well I can still detect the disease there, but we’re going to stop. That’s kind of a challenging conversation to have.

DR LOVE: Well we do observation on people who have CLL if they are not symptomatic and they have the disease.

DR THOMPSON: Yup. For sure.

DR LOVE: And it’s one thing too when you talk about big populations, and then you get down to individual patients. Like your patient develops these cardiac problems. And maybe it’s not going to change globally the algorithm but in individual cases I could see where maybe it’d be useful to have the information.

This concludes our program. Special thanks to Drs Kumar and Thompson, and thank you for listening. This is Dr Neil Love for Oncology Today.