News|Articles|August 18, 2026

NMA Associates MRD-Negative CR With PFS in Earlier-Line R/R Myeloma

Fact checked by: Chris Ryan

Listen

0:00 / 0:00

Key Takeaways

Trial-level WLS modeling across 12 RCTs met prespecified surrogacy criteria, supporting MRD-negative CR as a valid surrogate endpoint for PFS in earlier-line relapsed/refractory disease.
OS surrogacy was not established, with lower explained variance and a slope CrI crossing zero, consistent with dilution from subsequent therapies and longer follow-up requirements.
Eligibility focused on predominantly second- to fourth-line populations, excluding triple-class–exposed cohorts and class-specific refractoriness, improving applicability to “earlier-line” relapsed/refractory settings.
A bivariate NMA connected fragmented MRD subnetworks by incorporating PFS/time-to-progression-only studies, enabling estimation of MRD and PFS effects for regimens lacking direct MRD reporting.
Model choice mattered: DH bivariate and univariate PFS estimates aligned, while the Bujkiewicz approach produced more conservative, precise estimates, raising concern for potential over-shrinkage.

MRD-negative complete response (CR) represented a valid surrogate end point for progression-free survival (PFS) among patients with earlier-line relapsed/refractory multiple myeloma, according to a trial-level surrogacy analysis and network meta-analysis (NMA) presented at the 2026 EHA Congress.¹

In a weighted least-squares (WLS) model fit to 12 randomized controlled trials (RCTs) that reported both MRD-negative CR and PFS outcomes, MRD-negative CR and PFS were strongly associated (R = −0.85; 95% CI, −0.96 to −0.49; R² = 0.73; 95% CI, 0.24-0.93). The slope of the association excluded zero (−0.37; 95% CI, −0.53 to −0.21) and the intercept included zero (−0.14; 95% CI, −0.39 to 0.11), meeting prespecified criteria for a valid surrogate end point.

“There is a strong association between MRD-negative CR and PFS within earlier lines of therapy in [relapsed/refractory multiple myeloma], which supports MRD-negative CR as a potential surrogate end point for this population of interest,” lead study author Ola Landgren, MD, PhD, of the University of Miami in Florida, and coauthors wrote in a poster presentation of the data.

MRD-negative CR is a sensitive marker of response depth in multiple myeloma and is increasingly reported in trials; earlier in 2026, the FDA issued draft guidance on the use of MRD and R as primary end points to support accelerated approvals of multiple myeloma treatments.²

How was the surrogacy analysis conducted?

The investigators performed a systematic literature review to identify RCTs reporting MRD-negative CR at a sensitivity threshold of 10^-5 along with PFS or overall survival (OS) in earlier-line relapsed/refractory multiple myeloma, defined as trials in which more than 80% of patients were in the second- to fourth-line setting, were not triple-class exposed, and were not required to be refractory to a specific class or regimen.¹

Trial-level surrogacy was assessed via a WLS model based on the log odds ratio for MRD-negative CR and the log hazard ratio for PFS or OS. A surrogate was considered valid when the slope’s 95% credible interval (CrI) excluded zero, the intercept’s 95% CrI included zero, and the conditional variance approached zero.

NMA MRD Analysis in Myeloma: Key Takeaways

Across 12 RCTs, MRD-negative CR was strongly associated with PFS (R² = 0.73) and met criteria for a valid surrogate end point in earlier-line relapsed/refractory multiple myeloma.

The association between MRD-negative CR and OS was weaker and more uncertain (R² = 0.29 across 10 RCTs).

A bivariate NMA leveraged the MRD–PFS association to estimate MRD-negative CR treatment effects across a network of 27 studies, including regimens that lacked direct MRD data.

What did the analysis show for the MRD/OS relationship?

Among 10 RCTs reporting both MRD-negative CR and OS data, the association was weaker and more uncertain than for PFS, with an R² of 0.29 (95% CI, 0.08-0.93) in the WLS model. The intercept’s 95% CrI included zero (−0.04; 95% CI, −0.36 to 0.28), but the slope’s 95% CrI also included zero (−0.15; 95% CI, −0.35 to 0.04), and the slope was higher in the BRMA model than in the WLS and DH models.

Investigators attributed the weaker signal in part to confounding factors such as subsequent therapy, while noting that the observed association was nonetheless stronger than the R² of 0.12 reported in the earlier FDA analysis.

How did the bivariate network meta-analysis perform?

For the second objective, 27 studies, including 26 RCTs and 1 matched-pairs analysis, formed the evidence base for bivariate NMA models. Twelve studies reported both MRD-negative CR and PFS but formed fragmented subnetworks that could not be fully connected on MRD-negative CR alone, and an additional 15 studies that reported PFS or time to progression provided the bridging connections needed to complete the network.

Using dexamethasone as the reference treatment, the bivariate models leveraged the MRD–PFS association to estimate treatment effects for both end points across the fully connected network, allowing regimens without direct MRD data to be compared. Univariate PFS models and the bivariate DH model were consistent, whereas estimates from the bivariate Bujkiewicz model were more precise and conservative, which the authors noted may reflect over-shrinkage that underestimates clinically meaningful differences. The investigators concluded that a bivariate NMA can leverage the MRD-negative CR–PFS relationship to support indirect treatment comparisons that inform accelerated approvals and guideline development.

References

Landgren O, Campbell H, Towle K, et al. The relationship between minimum residual disease-negative complete response rate and progression-free survival treatment effects in relapsed refractory multiple myeloma: a novel perspective. Presented at: 2026 European Hematology Association Congress; June 11-14, 2026; Stockholm, Sweden. Abstract PF808.
US Department of Health and Human Services; US Food and Drug Administration; Oncology Center of Excellence; Center for Drug Evaluation and Research; Center for Biologics Evaluation and Research; Center for Devices and Radiological Health. Minimal residual disease and complete response in multiple myeloma: use as endpoints to support accelerated approval. FDA. January 2026. Accessed February 9, 2026. https://www.fda.gov/media/190647/download