NEW YORK – All 12 experts on an US Food and Drug Administration advisory panel on Friday agreed in a unanimous vote that available evidence supports the use of minimal residual disease (MRD) as an accelerated approval endpoint in multiple myeloma clinical trials.
At a meeting of the FDA's Oncologic Drugs Advisory committee, patients, doctors, and advocates for multiple myeloma research all expressed hope that the proposal to use MRD as an earlier clinical trial endpoint could bolster drug development in the disease setting and even bring about curative treatments. Market observers also saw the advisory committee's vote as a positive for diagnostic companies like Adaptive Biotechnologies and other MRD test providers as pharmaceuticals companies increasingly use MRD as an endpoint in drug trials.
In multiple myeloma, a type of antibody-producing white blood cell proliferates in the bone marrow and crowds out other types of blood cells. The overabundance of these plasma cells leads to anemia, excessive bleeding, and poor immune function. Myeloma also forms osteolytic lesions, or weak spots in the bone, which cause significant pain, fractures, and increased calcium levels in the blood, leading to impaired kidney function or kidney failure.
Prior to the 2000s, the standard treatment for multiple myeloma was chemotherapy followed by an autologous stem cell transplant, and the disease had a very grim prognosis. In the late 90s, the five-year survival rate for newly diagnosed multiple myeloma patients was around 35 percent and had changed little since the 1960s.
The landscape shifted, however, as new breakthrough treatments began to emerge, beginning with the repurposing of thalidomide, which was approved for multiple myeloma in 1997. Since then, treatment options have proliferated to include numerous drugs, biologics, and combination regimens, and more recently, CAR T-cell therapies. These new therapies have significantly extended patient survival, which is now about 10 years for newly diagnosed patients. However, the disease remains incurable, with a five-year survival rate of less than 60 percent.
Ironically, it is the success of newer treatments in extending patients' survival that now presents a challenge for researchers and drug developers. The FDA has considered a multiple myeloma drug's ability to improve patients' progression-free and overall survival to approve it under its regular pathway, and the agency has granted accelerated approvals to treatments based on patients' overall response rates and duration of response on them. However, overall response rates in drug trials can be very high — often approaching 100 percent in newly diagnosed and relapsed and refractory settings — because most patients do initially respond to standard-of-care therapies.
"We are now seeing response rates with single agents in a very relapsed patient population that are as high as those of the combination regimens evaluated [previously] in earlier-line settings," Bindu Kanapuru, the FDA's associate director of therapeutic review said at the meeting. "Developing new drugs or therapies in multiple myeloma has become challenging with the availability of highly effective regimens, and demonstrating statistically significant differences in overall response rates may require unfeasibly large clinical trials."
Despite more active treatments entering the market in the last three decades, multiple myeloma remains incurable. Patients frequently relapse, and their remission periods become shorter with every line of therapy.
During the open public hearing portion of the ODAC meeting, Linda Huguelet, a 14-year multiple myeloma survivor and co-leader of the Chattanooga Multiple Myeloma Networking Group, compared her treatment journey to being a frog hopping from lily pad to lily pad in a pond, where each floating leaf is a new treatment. When she was first diagnosed in 2010, Huguelet received Bristol Myers Squibb's thalidomide derivative Revlimid (lenalidomide), Janssen's Velcade (bortezomib), dexamethasone, and bisphosphonate to stabilize bone damage to her spine from the disease. Since then, she has cycled through many treatments.
"Having treatment options is always key with this disease because relapse is almost inevitable for every patient," Huguelet said. "The goal is to maximize each treatment and to buy time for more treatments, or lily pads, to become available."
In April 2023, Huguelet was "overjoyed" to land a slot for Janssen's CAR T-cell therapy Carvykti (ciltacabtagene autoleucel) and is now looking forward to her one-year MRD evaluation. "I'm optimistic, but also realistic that at this point in my journey I have used up many of the lily pads in the pond," Huguelet said. "Having additional treatment options is very personal to me and to all myeloma patients."
However, drugmakers working on better, potentially curative treatments say that the clinical trials testing these therapies' impact on progression-free and overall survival require an average of ten years to read out. MRD — a measure of the reduction of cancer cells in the body following treatment — has emerged as a potential intermediate or surrogate marker to gauge the efficacy of multiple myeloma drugs. Clinical trial investigators now routinely collect MRD data on patients and already use it as an exploratory or secondary endpoint in drug trials. The FDA's task to its advisory committee was to weigh the available evidence on MRD and determine if it can be a reliable and early predictor of survival endpoints and support faster regulatory decisions of multiple myeloma drugs.
Making of a surrogate endpoint
For regular FDA approval, a new drug must have substantial evidence of effectiveness based on a direct measure of clinical benefit. In cancer, the gold standard is typically overall survival. For accelerated approval, the FDA will accept evidence based on a surrogate endpoint that is reasonably likely to predict clinical benefit. According to Nicole Gormley, FDA associate director of oncology endpoint development, meta-analysis data including patient-level and trial-level data from multiple trials are needed to validate an intermediate or surrogate endpoint for accelerated approval.
Gormley cautioned at the meeting that the use of a surrogate endpoint that reads out earlier than a survival endpoint carries some risk and cited examples where surrogate biomarkers ultimately didn't reflect long-term clinical outcomes for patients. In the Cardiac Arrhythmia Suppression Trial, conducted in the late 1980s and 1990s, investigators sought to test whether suppression of post-myocardial infarction pre-ventricular contractions correlated with long-term clinical outcomes and overall survival. Arrythmias had been accepted as a prognostic biomarker based on multivariate analyses. However, in the trial, the use of antiarrhythmic agents increased arrhythmic death and cardiac arrest 3.6-fold despite all patients tolerating the drug and demonstrating premature ventricular-contraction suppression. If suppression of premature ventricular contractions post-myocardial infarction had been relied upon as a surrogate endpoint, "disastrous consequences could have occurred," Gormley said.
To avoid such consequences, the FDA has established procedures for the use of surrogate biomarkers as early endpoints for accelerated approval, including requiring that sponsors conduct confirmatory trials to prove that a therapy approved on an accelerated clip has clinical benefit. Under the Consolidated Appropriations Act of 2023, the agency now has authority to require that a confirmatory trial is underway before granting accelerated approval, and there is a formal, expedited withdrawal procedure for drugs approved through the accelerated approval pathway if the confirmatory trial fails to verify the expected clinical benefits.
In multiple myeloma, overall response rates based on biomarkers such as M protein in the urine and bone marrow biopsy have supported accelerated approval of new therapies. MRD gauged using next-generation sequencing or multiparametric flow cytometry promises a far more sensitive snapshot of drug response than conventional assessments, as seen in two meta-analyses the FDA asked ODAC members to consider.
Ola Landgren, a hematologist at the University of Miami Health System, presented a meta-analysis from multiple randomized-controlled Phase II and Phase III confirmatory trials in patients with newly diagnosed multiple myeloma in which MRD negativity was a primary, secondary, or exploratory endpoint. All studies used the FDA-cleared Adaptive Biotechnologies clonoSEQ 2.0 diagnostic test or multiparametric flow cytometry for MRD analysis, both of which have a sensitivity of 1 in 100,000 cells.
Their analysis found a strong individual-level association between achieving MRD negativity a year after initiating treatment and both progression-free and overall survival. The University of Miami group concluded that the data supported MRD as a prognostic marker for determining multiple myeloma patients' long-term outcomes, and that as an endpoint, MRD is likely to predict whether patients will benefit from treatment.
Another meta-analysis by the International Independent Team for Endpoint Approval of Myeloma MRD (i2TEAMM) included 20 randomized multiple myeloma trials for which researchers were able to obtain patient-level data, including clinical and pathological baseline values, tumor response data, MRD data, progression and survival data, and other relevant information. The trials all had MRD measurements based on validated multiparametric flow cytometry or next-generation sequencing tests with a sensitivity of 1 in 10,000 cells or better.
The i2TEAMM concluded that patients who achieved MRD negativity after a complete response had a much better long-term prognosis than those who did not, supporting MRD negativity as a surrogate endpoint for predicting outcomes. The FDA's Kanapuru said that MRD data used to support accelerated approval in multiple myeloma should be robust and come from a validated assay.
"The objective is to evaluate the strength of the association between the treatment effect on the surrogate [endpoint] and the treatment effect on the true endpoints," Gormley added. "In other words, if a treatment improves MRD-negative [complete response] over the control arm, will a similar improvement be observed in [progression-free survival] and [overall survival]?"
Gormley explained that if a strong trial-level association is achieved for the biomarker, the endpoint would be validated as a surrogate and could be used to support approval of a drug under the regular pathway. However, few early oncology endpoints meet that standard. For accelerated approval, weak trial-level associations are acceptable if the association is strong at the patient level.
The FDA generally agreed with the University of Florida and i2TEAMM's conclusions. The agency's reviewers found strong patient-level associations across all populations for MRD-negative complete responses at nine and 12 months after beginning treatment and weak-to-moderate trial-level associations for progression-free survival in most disease subpopulations.
The ODAC members, after reviewing this data, also unanimously agreed with the FDA, and all voted 'yes' to the question: "does the evidence support the use of MRD as an accelerated approval endpoint in [multiple myeloma] clinical trials?"
Rachel Ershler, a clinical reviewer for the FDA, noted that if MRD is to be accepted as a surrogate endpoint for accelerated approval, there are two options for clinical trial designs. First, sponsors could use a traditional two-trial approach, starting with a single-arm trial in the late-line setting followed by a randomized confirmatory trial in an earlier-line setting to convert to regular approval.
Alternately, drugmakers could fold everything into a single trial, in which data from MRD testing, supported by duration of response data from a randomized trial in an earlier-line setting, could be used for initial accelerated approval, and then longer follow-up data on progression-free survival and overall survival from the same trial could be used to confirm clinical benefit. In either case, accelerated approval could be withdrawn if the benefit is not confirmed via progression-free and overall survival endpoints.
Shifting incentives, future questions
While the advisory committee was generally in agreement on the strength of the data supporting MRD as an endpoint for accelerated approval, they raised several concerns while discussing the evidence. For example, some advisors pointed out that it could shift the incentives for drugmakers, leading to some level of "gamesmanship" in optimizing development around the MRD endpoint while being less concerned about long-term outcomes. On the other hand, committee members noted that a shorter development timeline associated with MRD as a surrogate endpoint could motivate companies to do more clinical trials, leading to more potential new therapies for patients.
The advisory committee's views about MRD generally bode well for diagnostic companies in this space, Mark Massaro and Vivian Bais, analysts from the investment bank BTIG, wrote in a note to investors on Friday.
"We believe [Adaptive's] FDA-cleared clonoSEQ assay is well positioned to benefit from today's vote and be the 'assay of choice' to handle much of the expected increased demand of using MRD in more clinical trials in multiple myeloma drug development," they wrote, adding that Adaptive expects to now book more partnerships with drugmakers wanting to track MRD in drug trials for melanoma and other diseases. Natera, which sells the patient-specific multiplex PCR NGS-based Signatera MRD test and is the market leader in the solid tumor space, may also "find a way to benefit from today's vote," having shown some data in hematologic cancers, the analysts added.
During the meeting, the committee discussed the optimal sensitivity tests measuring MRD as clinical trial endpoints should have, which in the meta-analyses had been one cell in 100,000. Ershler noted that while patients might prefer a more sensitive test to guide their individual treatment choices, that is the lowest sensitivity that can be reliably obtained from patient samples with current technology in a clinical trial setting. "You want a sensitivity level that allows you to discriminate between the two treatments, and [at one in 1,000,000], by way of example, you may have just one or two patients," Ershler said. "Whereas if you use [one in 100,000], you might have 15 to 20 patients, and that might allow for more discriminating power."
Lastly, reviewers noted that the two meta-analyses providing evidence on MRD lack data from CAR T-cell therapies. Neil Vasan, a professor of hematology and oncology at Columbia University Medical Center, pointed out that the meta-analyses include trials of small molecules and antibodies, and wondered if the conclusions drawn from them can be extrapolated to CAR T-cell therapies or therapies with mechanisms that have not been tested yet in clinical trials for multiple myeloma. CAR T-cell therapies can cause toxicities over a longer timeline, Vasam noted, adding that "in a nine-month or 12-month assessment for MRD that may or may not be reflected."
In response, Ershler acknowledged that there are gaps in the current data, including on CAR T-cell therapies and in disease settings such as smoldering myeloma and precursor conditions. She speculated that a repeat of the meta-analyses might not be necessary to shed light on those situations and that regulators might be able to extrapolate relevant insights from them. "Other information can help supplement and allow us to, within the regulatory agency, have confidence that [MRD] can be used in these other settings," Ershler said.
During the portion of the meeting for public comments, Vinay Prasad, a hematologist-oncologist and an epidemiology and biostatistics professor at the University of California, San Francisco, dissented with the committee on the utility of MRD as an endpoint for accelerated approval. Rather than addressing the FDA's central question about the adequacy of evidence for MRD as a surrogate endpoint, he argued that because multiple myeloma patients survive an average of 10 years following diagnoses and are typically in their late sixties, there is no unmet need for new therapies and novel agents should not be eligible for accelerated approval at all. "The biggest problem is that unsafe drugs will come to the US market," he said, arguing that approvals of new multiple myeloma drugs should be based on progression-free survival. "This population needs to be shielded from risk precisely because the outcomes are good."
While acknowledging that within the FDA's accelerated approval framework there is always a risk that a drug approved quickly based on surrogate endpoints can later prove ineffective or unsafe, committee members responded to Prasad's concerns by pointing out that many effective multiple myeloma drugs had already been approved via the accelerated pathway based on overall response rates and that the FDA would continue to consider the totality of evidence, including safety and tolerability, for any approval.
"We know a lot about the risks of action," said Christopher Hourigan, director of the Virginia Tech Fralin Biomedical Research Institute Cancer Research Center, adding that the committee and FDA must also weigh the "risks of inaction" on the MRD endpoint.
"Imagine the future of drug development where we're using a non-high sensitivity measure of anti-myeloma response. We can't push any trials for new entities forward for drug development because we don’t have the appropriate tools to measure the efficacy of those therapies," he said. "There's harm to inaction."
Despite Prasad's comments, during the vote, committee members remained optimistic that the use of MRD as an early endpoint in multiple myeloma drug trials will herald treatments allowing patients to live longer, better, and with fewer relapses, or even cure their disease.
"I want to live 50 or 60 or 70 years," said Michael Riotta, the patient representative on the advisory committee. "If [MRD] can bring a drug to market faster, as an educated patient, I'll take that risk. If there's a drug out there that's on a clinical trial and MRD negativity is its endpoint, and it's going to give me maybe 18 months or 24 months, I'm going to jump at it."