Overview
Measurable residual disease (MRD) testing detects small amounts of cancer remaining after treatment and is increasingly used to assess treatment response in multiple myeloma, a cancer of plasma cells. However, less is known about how clinicians interpret MRD results and use them to guide treatment decisions.
Using electronic health record data from the Flatiron Health Research Database, researchers analyzed 5,686 MRD tests from 2,442 patients. A large language model (LLM)-enabled approach was used to examine clinical notes and characterize physicians' interpretations, treatment rationale, and subsequent management decisions.
The proportion of tests using molecular methods increased from 14% in 2015–2019 to 52% in 2025 or later. MRD was identified as a primary driver of treatment decisions in approximately 25% of evaluable assessments, but more commonly served as supporting information rather than the deciding factor.
Within 90 days of MRD testing, treatment escalation, including consolidation, occurred following 38% of assessments, while treatment discontinuation occurred following 6%. Escalation was more common after positive MRD results, whereas discontinuation was more common after negative results. However, these treatment changes were not necessarily driven by MRD findings.
Why this matters
MRD results do not automatically translate into treatment changes. This study highlights how clinicians incorporate MRD alongside other clinical factors when making decisions about multiple myeloma treatment. Understanding this distinction may help inform future research on when and how MRD testing can best guide patient care.