Analysis
AI & Data
CLL measurable residual disease: why treatment context matters
In a single-centre retrospective cohort, the link between undetectable MRD and progression-free survival differed by treatment setting.
Technical Review of a peer-reviewed original retrospective study, 10.3389/fonc.2026.1918537. CFCM editorial synthesis, not independent scientific review or clinical validation.
Standfirst: A 2026 real-world CLL cohort shows why the prognostic meaning of an undetectable blood MRD result depends on assay sensitivity, time of sampling and the treatment strategy.
What does eight-colour ERIC flow actually measure?
Mihalyova and colleagues analysed 177 unique patients over 198 treatment episodes (749 peripheral blood samples and 64 bone marrow aspirates). Their ERIC-style tube included CD81, CD43, CD79b, CD22, CD5, CD19, CD20 and CD3. Approximately a million acquired events per analysis yielded more than 200,000 evaluable leukocytes. The paper states an LoD calculation of 20 divided by evaluable leukocytes, multiplied by 100. Acquired events and evaluable leukocytes are not interchangeable denominators.
A signal specific to clinical context
Undetectable MRD occurred in 111 of 198 episodes. In newly diagnosed CLL across mixed regimens, median progression-free survival was 58.0 months with undetectable MRD and 34.7 months with detectable MRD (HR for detectable versus undetectable 2.28, 95% CI 1.20–4.32). In relapsed or refractory CLL the corresponding association was less precise (HR 1.86, 95% CI 0.81–4.27). Figure 1 compares MRD strata within each setting. The two panels cannot establish a treatment effect for venetoclax or a difference between settings. Continuous BTK inhibitor therapy may maintain good disease control despite persisting measurable CLL cells, unlike fixed-duration venetoclax-based regimens where deep MRD response is a central treatment endpoint.
Where interpretation must stop
The retrospective single-centre design, heterogeneous regimens, sampling schedules and small subgroups preclude a new treatment-discontinuation algorithm. The reported median blood LoD of 0.0032% cannot be generalized to every instrument. Always report specimen, timepoint, evaluable-leukocyte denominator and actual LoD before comparing groups.

Daily Learning Pill
MRD negativity is an assay- and therapy-context observation, not a universal biological absence.
Read the primary study. Explore CFCM Resources and Troubleshooting.