FDA’s ODAC Recognizes MRD as an Accepted End Point for Accelerated Approval in Multiple Myeloma

The FDA’s Oncologic Drugs Advisory Committee voted 12 to 0 that the totality of available data supports the use of minimal residual disease (MRD) as an end point for accelerated approval of new treatments for patients with multiple myeloma.¹

“The FDA showed that MRD does fall short of true surrogacy, but that's a high bar, and that wasn't the question today. Our clinical experts and the FDA both agree that MRD does meet the criteria for accelerated approval, and that's why I voted yes. That said, we need to be cautious that once FDA guidance...gets out that MRD is acceptable for accelerated approval, it will change the incentive structure for preclinical modeling, clinical development, and early clinical trials, so that requires the FDA to be vigilant," Ravi A. Madan, MD, senior clinician at the National Cancer Institute, said.

"We talked a little bit about how that may lead to throwing the baby out with the bathwater, as financial incentives may pressure industry to hit the MRD mark or decide not to continue. On the flip side, it could raise other concerns that hitting MRD may not translate into long-term clinical efficacy. Therefore, the FDA needs to pay close attention as it always does, to safety, progression-free survival [PFS], and other relevant points like survival,” Madan added.

During the meeting, the FDA, representatives from The University of Miami’s Sylvester Comprehensive Cancer Center, and the i2TEAMM presented 3 independent data sets from a meta-analysis based on patient-level data from several clinical trials.²

The University of Miami first presented individual-level results showing that there is a high association between 12-month MRD negativity and PFS in a newly diagnosed multiple myeloma population (n = 4907; copula global OR, 4.72; 95% CI, 3.53-5.90). Associations were also seen in the transplant-ineligible (n = 3221) and transplant-eligible (n = 1686) populations (copula global OR, 6.15; 95% CI, 4.27-8.03; and copula global OR, 2.45; 95% CI, 1.40-3.51, respectively).

The association between 12-month MRD negativity and overall survival (OS) was also evaluated. The results showed that MRD negativity at 12 months is prognostic of improved long-term OS in all newly diagnosed patients (copula global OR, 4.02; 95% CI, 2.57-5.46), transplant-ineligible patients (copula global OR, 4.08; 95% CI, 2.44-5.72), and transplant-eligible patients (copula global OR, 3.78; 95% CI, 0.78-6.78).

The university also performed trial-level associations between MRD and PFS, as well as MRD and OS. A total of 9, 2-arm comparisons were available for analysis, but only 8 comparisons with 4907 patients met the data criteria for the trial-level surrogacy analysis. The eighth study, which was not included in the primary analysis because more than 20% of patients had a missing value for the primary end point definition based on MRD, brought the total sample size to 5130.

The total follow-up for PFS was 28.6 months (IQR, 19-56.4). The R²copula, R²weighted least squares (WLS; inverse variance), and R²WLS (sample size) estimates were 0.82 (95% CI, 0.62-1.03), 0.68 (95% CI, 0.45-0.91), and 0.74 (95% CI, 0.54-0.94), respectively.

The i2TEAMM also presented global odds ratio estimates regarding the measurement of individual-level associations between 9- and 12-month MRD negativity and PFS. The global odds ratio is meant to reflect the odds of a patient being alive and progression-free beyond a time point vs those without MRD negativity, adjusting for treatments. At an MRD testing sensitivity level of 10^–5, all newly diagnosed (n = 3061), transplant-ineligible (n = 2235), and relapsed/refractory (n = 1378) patients had greater odds of being alive and progression-free beyond 9 months, with respective odds ratios of 8.27 (95% CI, 6.53-10.01), 9.80 (95% CI, 5.14-14.46), and 8.24 (95% CI, 4.41-12.07).

The same was true for all newly diagnosed (n = 3009), transplant-ineligible (n = 2281), and relapsed/refractory (n = 863) patients beyond 12 months, with respective odds ratios of 9.15 (95% CI, 7.27-11.03), 11.95 (95% CI, 7.32-16.58), and 16.24 (95% CI, 5.77-26.71).

The FDA acknowledged that prior analyses have used the same methodology used by both applicants to evaluate the level of surrogacy of earlier clinical end points with long-term clinical end points. However, the FDA cited 6 limitations that may bias interpretation of the results, including the fact that:

1. The majority of treatment types included in the meta-analyses were small molecules and monoclonal antibodies and more modern therapies like CAR T-cell therapy were not included;
2. Most trials did not evaluate 9- or 12-month MRD rate as a secondary end point;
3. There was no standardization of assays and assay sensitivities across trials;
4. There were variations in trial design and follow-up times, resulting in differences in evaluation times, follow-up duration, and represented regions;
5. The patient populations were more unlike than alike, few baseline characteristics were routinely collected across trials, and the small number of trials in any particular setting limited the ability to properly evaluate subgroups;
6. The findings from the selected trials may not be entirely representative of the association between MRD-negative complete response (CR) and PFS/OS if the trials not included in this analysis are qualitatively different.

The i2TEAMM concluded that although their analysis confirmed the robust prognostic value of MRD negativity beyond CR for PFS in patients with multiple myeloma, supporting the current data, it did not support the replacement of PFS with an MRD-based end point in phase 3 trials. They did however state that the data support that treatment effects on MRD at a sensitivity of 10^–5 or better are ‘reasonably likely’ to predict treatment effects on PFS.

“Based on the results observed in this meta-analysis of multiple large, randomized studies, i2TEAMM believes there is sufficient evidence to support use of MRD as an end point for accelerated approval, with PFS maintained as a long-term end point for confirmation of clinical benefit,” the i2TEAMM wrote in the briefing document.

The University of Miami echoed the notion that the primary and subsequent analyses support MRD as an end point "reasonably likely" to predict clinical benefit.

References

1. April 12, 2024, Meeting of the Oncologic Drugs Advisory Committee (ODAC). FDA. https://www.youtube.com/watch?v=pooME9gMaL0
2. Oncologic Drugs Advisory Committee (ODAC) Meeting. Combined FDA and Applicants ODAC Briefing Document. Accessed April 12, 2024. https://www.fda.gov/media/177652/download