MRD Testing Outside the Context of a Clinical Trial

Undetectable MRD (≤1 × 10-4 CLL cells in peripheral blood or bone marrow aspirates) can be confirmed by flow cytometry or next-generation sequencing. The attainment of undetectable MRD represents an even more stringent complete remission that was prognostic for improved PFS and OS in multiple studies.[22,44-46]The use of time-limited therapy with venetoclax, or the investigational combination of venetoclax and a BTK inhibitor, has produced complete responses, and even undetectable MRD, in most patients.[22,44,47]

Testing for undetectable MRD has become a standard parameter for defining responses in all modern clinical trials for CLL. Undetectable MRD has prognostic value but its status as a predictive marker is uncertain. The potential value of MRD testing in routine clinical practice depends on whether it can be used for clinical decision making such as stopping, changing, or continuing treatment. High-level evidence for this intervention would require a prospective randomized clinical trial in which MRD was used as a predictive biomarker for a group attaining an OS advantage compared with a control group disregarding MRD status. Such evidence has not been attained. Similar to outcomes in follicular lymphoma and other indolent lymphoid neoplasms, improved PFS in patients with CLL does not directly predict OS.

A phase II trial used a combination of venetoclax and ibrutinib in patients with undetectable MRD after 1 year of therapy. Patients were randomly assigned to receive either ibrutinib maintenance therapy or treatment cessation.[47] All MRD-positive patients continued to receive ibrutinib. With a median follow-up of 34.4 months, the 1-year PFS rates were 98% for patients with undetectable MRD who ceased therapy, 96% for patients with undetectable MRD who received continued ibrutinib, and 97% for MRD-positive patients who received continued ibrutinib. All patients did well at 12 months. Knowledge of MRD status to guide discontinuation of therapy did not affect the outcome.[47]

Another phase II trial included 70 previously untreated patients who received 1 year of venetoclax plus obinutuzumab. Patients were randomly assigned to receive another year of venetoclax irrespective of MRD status, or another year of venetoclax only if they were MRD-positive. With a median follow-up of 35.2 months, the MRD rates were identical for both randomized groups, suggesting no improved efficacy with this approach, although increased adverse events were reported with an extra year of venetoclax.[48]

Before MRD is used outside the context of a clinical trial, conclusive evidence is required to establish that MRD is a predictive biomarker that can guide clinical decisions.