No abstract

Purpose: With the advent of highly efficacious time-limited combination treatments of targeted agents in chronic lymphocytic leukemia (CLL), minimal residual disease (MRD) assessment has gained importance as a measure for therapeutic success and as a surrogate for progression-free survival. The currently most widely used method is multicolor flow cytometry, which detects circulating CLL cells in the peripheral blood. However, it seems to be less sensitive for the detection of MRD in the lymph node compartment. Patients and Methods: To evaluate whether a cell-free approach can overcome this limitation, we performed serial assessments of circulating tumor DNA (ctDNA) in patients with CLL treated with obinutuzumab, acalabrutinib, and venetoclax in the phase II CLL2-BAAG trial. Patient-specific variability, diversity, joining (VDJ) rearrangements as well as somatic driver mutations were tracked before, during and after treatment by digital droplet PCR in blood plasma. Furthermore, these were systematically compared to matched flow cytometry data. Results: In the 381 sample pairs, ctDNA and flow cytometry yielded highly concordant results. However, clone-specific ctDNA was detected in 44 of 152 samples (29%) that were assessed as undetectable MRD (uMRD) by flow cytometry (defined as less than one CLL cell in 10,000 normal leukocytes). 29 ctDNA-negative samples showed detectable MRD >10–4 by flow cytometry. Also, somatic driver mutations were detected with a similar sensitivity compared with patient-specific VDJ rearrangements in plasma. In patients with predominantly nodal residual disease, ctDNA compared favorably with 4-color flow cytometry and seemed to more accurately reflect the entire disease burden across compartments. Conclusions: On the basis of these findings, ctDNA-based MRD assessment appears to be a promising method to complement cell-based MRD approaches like flow cytometry that focus on circulating CLL cells in the peripheral blood.