The FDA Oncologic Drugs Advisory Committee March 7 voted to accept the metric of “minimal residual disease,” or MRD, as a basis for approval of a drug for the treatment of acute lymphoblastic leukemia.
In an 8:4 vote, the committee recommended broadening the indication for the Amgen drug Blincyto (blinatumomab) to include the treatment of MRD-positive B-cell precursor ALL.
ALL is a small indication. There are about 2,500 new cases among adults in the US, and about half of these patients are found to be MRD-positive after initial chemotherapy. Since not all of these patients are offered treatment for MRD, the indication likely applies to several hundred new patients per year.
However, MRD is a factor in all forms of leukemia, multiple myeloma, some lymphomas, and, potentially, via measurement of tumor-associated cell-free DNA, in solid tumors.
The agency didn’t specify to the public—nor, presumably, to the committee members—whether it’s considering a full approval or an accelerated approval for the indication.
ODAC’s recommendation signals a significant evolution in the agency’s approval criteria since historically—from the 1950s on—approvals of therapies for the treatment of ALL have been based on measures demonstrating achievement and maintanence morphological complete remission.
While drugs were being approved based on these criteria, hematologist/oncologists who treat ALL have been using MRD to determine the patients’ risk of recurrence and thus guide treatment decisions. If FDA follows the committee’s advice, which it almost certainly will, the approval of this supplemental Biologics License Application will mark the first-ever approval of a treatment targeting MRD.
Blincyto, which is sponsored by Amgen, is a bispecific CD19-directed CD3 T cell engager.
Though having MRD after initial therapy for ALL is bad, it’s not known whether eliminating it would improve the patients’ outcomes.
Worse, it’s not clear whether questions related to the role of MRD could ever be answered conclusively, and, more than anything, ODAC discussion pointed to the challenge of making regulatory decisions in the midst of a permanent revolution driven by technology. As science and technology changes, scientific questions are being rendered moot before answers are able to emerge.
“Yes, MRD is bad, and MRD does define a high-risk population, but we do not know the appropriate cutoff,” said Andy Chen, leader of the Lymphoma Program at the Oregon Health & Science University Knight Cancer Institute, who served as a temporary voting member of the committee.
“Even though we know that MRD is bad and it defines a poor risk group, we don’t really have randomized data yet to say that eradication of MRD improves outcomes, and that leads to the second part of this question—and that’s a more difficult question for us—is whether or not this presentation here is strong enough in the absence of that,” said Chen, who voted against approval.
FDA seemed disinclined to put all the cards on the table. The words “reasonably likely to predict,” which signal that the agency is considering an accelerated approval, were never uttered. This could be only because the agency wasn’t interested in hearing the ensuing discussion about surrogacy.
However, some interesting dance steps could be observed.
Richard Pazdur, director of the FDA Oncology Center of Excellence and acting director of the Office of Hematology and Oncology Products, prompted the sponsor to describe the confirmatory trials that would tease out the role of blinatumomab treatment and of transplantation in the MRD-positive population.
Sponsors rarely volunteer such information so as not to prompt ODAC members to say, “Well, the trials are ongoing, so why not wait? Come back in a few years.” Since the indication is rare—there are just a few hundred such cases a year—the wait would take the sponsor well into the 2020s.
The randomized trials addressing the question of transplantation are ECOG1910 and COGAALL1331. The agency usually requests that confirmatory trials be in progress if an application is to qualify for an accelerated approval.
In another hint that the agency may be thinking of an accelerated approval, the committee was asked whether “blinatumomab provides a potential benefit that outweighs the risks from the treatment.” The word “potential” could be interpreted as code for “reasonably likely to predict.”
Blincyto was granted an accelerated approval in December 2014 for the treatment of Philadelphia chromosome-negative relapsed or refractory BCP ALL on the basis of the CR rate, duration of CR, and proportion of patients with an MRD-negative CR or CR with partial hematological recovery (CRh) within two cycles of treatment in a single-arm trial.
In July 2017, the agent received a regular approval for treatment of relapsed or refractory BCP ALL in adults and children.
ODAC members struggled with Amgen’s data in part because about 77 percent of the MRD-positive patients went on to receive transplants after treatment with Blincyto. The contribution of transplants ultimately confounded the result of the Blincyto trial.
“There is no drug that’s ever been approved in ALL that specified whether a transplant should subsequently be performed or not,” said Aaron Logan, assistant professor of clinical medicine in the Division of Hematology/Oncology at the UCSF Helen Diller Comprehensive Cancer Center. Logan presented the clinical perspective as part of the Amgen presentation.
“The way I view this is this is a decision that needs to be individualized, and that individualized decision makes the assessment of the role of the transplantation as is trying to be performed today very complicated, because transplantation is not two buckets,” Logan said to the advisory committee. “It’s not no-transplant or transplant. It’s no-transplant or which of these 12 different kinds of transplants did the patient undergo? Was it myeloablative? Was it reduced-intensity? What kind of donor did you use? How did you manage the immune suppression after the transplant?
“And were your decisions about any of these things informed by your knowledge of the patient’s MRD? In the current era, all of those things are influenced for every patient on an individual basis by knowledge of their MRD.
“Our field has already accepted 10-4 as a threshold for specifying patients as being high-risk for relapse. If they remain MRD-positive at the time of the transplant, we are less likely to do a reduced-intensity transplant. If they remain MRD-positive at 10-4 or higher after transplant, we are more likely to rapidly taper their immune suppression. These are things that cannot be captured by the type of analyses presented today, and therefore I don’t think that this should be the focus of the discussion.
“The discussion should be, ‘Did this drug take a high-risk population—MRD-positive patients—and enable them to go on to potentially curative therapy?’ I think it has been demonstrated that a very high percentage of patients have been able to go on to a potentially curative allogeneic transplantation.”
Amgen’s application appears to have been inspired in part by a paper by Don Berry et al. that appeared in JAMA Oncology last year. The paper presented a meta-analysis of 39 publications based on data on 13,637 patients.
“The value of having achieved MRD negativity is substantial in both pediatric and adult patients with ALL,” the paper concludes. “These results are consistent across therapies, methods of and times of MRD assessment, cutoff levels, and disease subtypes. Minimal residual disease status warrants consideration as an early measure of disease response for evaluating new therapies, improving the efficiency of clinical trials, accelerating drug development, and for regulatory approval.
“A caveat is that an accelerated approval of a particular new drug using an intermediate endpoint, such as MRD, would require confirmation using traditional efficacy endpoints.”
The FDA analysis of this paper points out that the authors had pooled “nonrandomized responder analyses, i.e., irrespective of treatment received, patients with MRD-negative disease have longer [event-free survival] and [overall survival].”
“Additionally, because the analysis included studies with different cut-offs to determine MRD-negativity, the results do not address what level of MRD identifies the high-risk group or what level of MRD identifies the group with good long-term prognosis,” the agency’s report states. “Further, because the description of the patient population did not include whether patients were in Complete Remission 1 or later, whether the patients had true CR or marrow remission with incomplete hematologic recovery, or how [hematopoietic stem cell transplantation] use was addressed, it is not clear to what population specifically these results would apply.”
Amgen’s data was based on the BLAST trial, which—despite being the largest prospective trial in patients with MRD-positive ALL—was a single-arm phase II study. As such, the trial couldn’t measure progression in a convincing manner. Instead, the sponsor presented a propensity score analysis comparing outcomes with a historical control cohort.
The BLAST study [MT103-203] was a single-arm trial of up to 4 cycles of blinatumomab for treatment of patients with BCP ALL in CR or CR with partial platelet recovery and MRD > 0.1%. The primary efficacy endpoint of MT103-203 was complete MRD response (defined as absence of detectable MRD using an assay with a sensitivity < 0.01%) after 1 cycle of blinatumomab.
There were 116 patients treated with blinatumomab. From this group, FDA identified 87 patients in CR with hematologic recovery and baseline MRD > 0.1%, including 61 patients in CR1, 25 in CR2 and 1 in CR3. A complete MRD response was achieved by 69 patients (79%; 95% CI: 70%, 88%).
The estimated median hematological RFS was 22.3 months (25.6 months for patients in CR1 and 11.0 months for patients in CR2 or CR3).
A propensity score analysis for the patients in first remission (with or without hematopoietic recovery) in Study MT103-203 and in Study 20120148, a larger retrospective cohort study sponsored by Amgen, demonstrated that the RFS for the patients treated with blinatumomab was significantly greater than in the historical controls (p<0.0001 by log-rank; median 35.2 months vs 8.3 months, respectively).
How ODAC voted—and why
FDA asked the Oncologic Drugs Advisory committee to discuss a possible threshold for MRD cutoff. The Amgen study included patients with MRD > 0.1%. Committee members said they didn’t feel comfortable setting the cutoff threshold.
The voting question read: “Do the results of MT103-203 demonstrate that for patients with ALL in CR who have MRD > 0.1%, treatment with blinatumomab provides a potential benefit that outweighs the risks from the treatment?”
Acting industry representative
vice president of oncology development
As a non-voting member, I would like to congratulate and thank FDA and the sponsor for bringing forward and advancing the consideration of minimal residual disease and how we really begin to integrate this into assessing patients, considering how we use it to make decisions around therapy, and ultimately how it will potentially become an outcome measure for clinical care.
I think, clearly, the field is moving from the days of laying on of hands, radiographic evaluation, microscopic evaluation, and now molecular determinations of disease and disease response, and that’s going to become incredibly, and increasingly, important over the next decade.
Chief, Myeloid Malignancies Section
National Heart, Lung and Blood Institute
I voted Yes. I share the desire to have randomized study and better quality evidence about the confounding impacts of transplant, but I believe MRD-positive patients need treatment now, and want to have options for them while we’re working out the confounding influence of transplant.
Leader, Lymphoma Program
Knight Cancer Institute
Oregon Health & Science University
I voted No. I do believe that MRD is an important marker, and it should be used in studies going forward.
I thought that the results from this phase II study was too confounded by transplant to say, for certain, that there’s a significant clinical benefit, and I thought the numbers for patients who did not get transplant were too small to make any conclusion.
Assistant professor of medicine
Division of Hematologic Malignancies and Cellular Therapy
Duke Cancer Institute
I voted Yes.
I do think that there was significant data that was presented that showed that use of this drug in this setting is able to convert patients from an MRD-positive status to an MRD-negative status.
I think that there was data presented that was suggestive that having an MRD-negative status is beneficial, regardless of whether or not you’re going to transplant after receiving blinatumomab.
I do know that the reason I voted yes, however, was because the question was worded as a “potential benefit.”
I do not think, as Dr. Chen mentioned, that there is significant evidence suggesting that this is, for sure, definitively the way that we should go, in terms of treatment. I also think that it’s important to look at the data from the randomized trials that were upcoming, that were discussed, because I don’t think that, for example, if this was a question of whether or not it should be approved for this indication, I probably would have voted no in that setting, but I do think there’s enough data to suggest a potential benefit.
The other thing that I would like to comment is I would like to see more data about the potential adverse effects in patients who receive blinatumomab and then go on to transplant, because I feel that was not adequately presented, and as one person earlier in the conversation noted, a lot of the historical data was from 2000, where transplant in 2000 is very different from transplant in 2009, which is very different from transplant now.
And so, I think a more granular look at that detail and data is needed.
I voted Yes. I think that, as Dr. Sung said, patients probably benefit from being MRD-negative going to transplant. I don’t think that patients who are MRD-positive after the chemotherapy and then get Blincyto and become MRD-negative are quite the same, but it still looks like there’s some benefit over not being MRD-positive at the time of transplant.
So that’s why I voted yes. And I would like to add I’d like to see more randomized trials; I agree with that.
senior research administrator and CLIA operations director
Clinical Research Division Fred Hutchinson Cancer Research Center
I also voted Yes. I thought that the survival benefit in MRD-positive patients was there, although we can’t exactly quantify it, and we don’t know what the exact MRD cutoff is.
We don’t know how much MRD you can live with and not relapse. I felt that the data was sufficient, in the 10-1 to 10-4 population, to provide this additional treatment option to patients and their providers.
Professor of biostatistics and bioinformatics
Duke University Medical Center
I voted No, contrary to my previous peers.
And the reason I voted no was mostly because I wasn’t totally convinced that you can interpret the data, because the outcomes are being confounded due to HSCT, which limits, obviously, the interpretations of the results, and even though the study met its primary endpoint, I believe that additional follow-up are needed for the 203 study and additional analysis may help to adjust for confounding.
Professor of medicine
MD Anderson Cancer Center
Department of Thoracic Head & Neck Medical Oncology
I also voted No, and it was because mostly a question of interpretation of the intent of the question here.
My interpretation of the intent was that we were asked to vote on this, and the question was indicative of our intent to improve the drug in this indication. Therefore, that tainted my vote, although I think there is potential benefit, and I think there is plenty of data, and a clear benefit for this patient to have some therapy in the setting of MRD, I do not feel that we have the exact definition of the population that benefits.
For example, it was adequately phrased CR1 versus later CR, and also the confounding factor of transplant, as everybody else mentioned, was not clarified by the analysis. It was not feasible to clarify it, I think.
Bruce J. Roth
professor of medicine,
Division of Oncology
Washington University School of Medicine St. Louis
I ended up voting Yes. I actually wanted to vote Yes and No, and then I wanted to abstain. I voted Yes, because I think it met the primary endpoint, I think fairly impressive, almost 80% of the people MRD converted, and also voted yes kind of in the back of my mind for the patients who do not have transplant as an option, as another option to reduce MRD, and hopefully have something else available down the road.
The No part of my brain said that I am not convinced of clinical benefit from what was presented, and I think it was an impossible task to take this heterogeneous group of historical controls and try get anything out, and that’s, in fact, I believe what happened. I’d be very interested to see the results for the upcoming randomized trials to confirm that that MRD conversion actually does end up resulting in improved clinical benefit.
Professor of medicine
The University of Chicago
Section of Hematology/Oncology
Department of Medicine
I voted Yes. I think, perhaps the most simplistic way, after hearing the data and reviewing this, that the drug is currently approved for treating refractory ALL, and the way I see it, MRD-positive is a form of refractory ALL.
It’s a different mechanism of measurement, as we’ve heard, and it will probably get, as has been discussed, even more sensitive over time, but I think that that is an indicator of persistent and refractory disease, and I would be swayed by the predominance of clinical evidence that even with using it as a bridge transplant, that it was still valuable, since patients who get transplanted and are MRD-positive going in, have a less good outcome than those who go in MRD-negative.
Associate professor of medicine and oncology
Mayo Clinic Rochester
I voted Yes, really on the three pillars.
One, is that MRD used to be clearly predicting or identifying patients at the risk of relapse. We can argue about a cutoff, but multiple publications and data show that MRD is important in ALL.
Secondly, 203 study demonstrated that blinatumomab can actually convert patients from being MRD-positive to negative, and does it in a significant proportion of the patients.
The problem, thirdly, was the clinical benefit. And here, just like others, I struggled a lot with the conversion from MRD-positive to negative truly corresponds to a clinical benefit. But I think, overall, looking at the evidence, there is a reasonable probability that indeed it does help.
Finally, biologically, I think about being MRD-positive like almost being tied to the railroad tracks, and you see this train coming, and you see the lights far, far away, and you can think about this, “I’m going to wait until the train comes closer and use my ammunition then, or maybe I try to do something earlier to stop this train,” and biologically I cannot help thinking that early intervention could be of help here.
Professor of statistics and biostatistics
Dana-Farber Cancer Institute
Harvard T.H. Chan School of Public Health
Apparently, Dr. Roth and I agree on everything except the vote.
I voted No, and I voted No, primarily because, for me, there is still uncertain benefit in the patients who are eligible for transplant after their CR.
I don’t think of the subgroup, necessarily, as the ones who got transplant, but the ones who, you know, after that CR, could get transplant. I think that what’s difficult to sort out here, is that different analyses show a different level of benefit for blinatumomab before the transplant.
So, for me, it doesn’t quite reach the level of labeling evidence.
I think that in most trials, most of us are willing to approve an indication when a subset that you’re particularly worried about is small, but this is a large subset, and it leaves sort of open the question of whether they should be treated or go right to transplant.
There was a claim that, one hopes, that the deeper the response that inducing less residual disease prior to transplant will lead to a longer and better outcome for transplant, but I think that remains to be shown.
Bosworth Chair for Cancer Biology
Director, Center for Cancer and Immunology
Research professor of pediatrics and microbiology, immunology and tropical medicine
Children’s National Health System
The George Washington University School of Medicine and Health Sciences
I voted Yes. Again, like Dr. Sung, the keyword for me, in this question, was “potential benefit” that outweighs the risks from the treatment.
I think we all agree that MRD-positive patients need treatment. The sponsor gave a very good risk-benefit ratio, I think the study met its primary endpoint, and, for me, the data they showed in response to my question about the patients who did not go on to BMTs, and those 22% responding patients who did not go on to BMT, they clearly had an excellent RFS, especially compared to the absolutely dismal prognosis or outcome for the patients who did not respond and did not go onto transplant.
Obviously, we will await the data of the COG and ECOG randomized trials, and my one caveat would be, as we move, if we do move forward in this MRD setting, that we do need to look at the incidence of CD19-negative disease and the non-responders understanding that there are other CD19-directed therapies these patients might not be eligible for after this therapy.