S885 IN SEARCH OF THE RESIDUAL LEUKAEMIC CLONE IN CHRONIC MYELOID LEUKAEMIA PATIENTS IN TREATMENT‐FREE REMISSION

Abstract

Background:Approximately half of the patients with chronic myeloid leukaemia (CML) who stop tyrosine kinase inhibitors (TKIs) after a stable deep molecular response (DMR) remain in treatment free remission (TFR), while the remaining patients need to re‐start TKI therapy. In an updated report of the TWISTER study we demonstrated the persistence of BCR‐ABL1‐positive cells with falling levels of residual disease, detected by highly‐sensitive patient‐specific genomic DNA PCR. The characteristics of the leukemic cells that remain after successful TKI treatment are not known. Understanding which cells persist despite TKI therapy could inform strategies aimed at recruiting more patients to TFR.Aims:To determine the lineage of residual BCR‐ABL1‐positive cells from patients in TFR.Methods:Twenty CML patients in TFR with undetectable BCR‐ABL1 transcript (uMR4.5) for 1–11 years (median 3.0 years) provided peripheral blood for fluorescence‐activated cell sorting into granulocytes (CD16+CD66+), monocytes (CD14+), B cells (CD3‐CD19+), T cells (CD19‐CD3+), and NK cells (CD3‐CD19‐CD56+). B‐cells were then separated into mature‐naïve (CD27‐IgD+) and memory cells (CD27+IgD‐ switched memory; CD27+IgD+ non‐switched memory; CD27‐IgD‐ double‐negative). DNA nested Q‐PCR was performed on sorted cells using 20 replicates of 500 ng of DNA (total 10 ug) reaching a sensitivity of MR6.2 (one in ∼2 million cells) in the various cell fractions.Results:BCR‐ABL1 DNA was detected in total leukocytes from 14/20 TFR patients with a median value of MR5.6 (range MR5.1‐MR6.2). Even when BCR‐ABL1 DNA was not detected in total leukocytes it was sometimes detected in sorted fractions (4/6), likely reflecting concentration of measurable residual disease (MRD) in the relevant population. BCR‐ABL1 DNA was detected in B cells (n = 18 patients), T cells (n = 11), NK cells (n = 5) and monocytes (n = 4). Notably, BCR‐ABL1 was not detected in T cells if the B cell fraction was negative and BCR‐ABL1 values were consistently higher in B cells than in T cells (median MR4.9 vs MR5.7; P = 0.01). Only one patient in TFR for 8 years had undetectable BCR‐ABL1 DNA in all of the fractions analysed. BCR‐ABL1 DNA was not detected in granulocytes in any of the 20 patients. To verify whether BCR‐ABL1 positive lymphocytes identified during TFR are arising pre‐TKI treatment, we analysed 11 patients at the time of diagnosis. A variable fraction of B (median 1.9%, range 0.1–12%) and T lymphocytes (1.7%, range 0.009–36%) were leukaemic (BCR‐ABL1 DNA). BCR‐ABL1 mRNA was expressed in the sorted cell fractions at levels similar to the DNA values. Subsequent sorting of B lymphocytes from TFR patients into naïve and memory subsets showed that both fractions contained BCR‐ABL1 DNA. The absolute number of naïve B‐cells did not change during years in TFR, but the absolute number of BCR‐ABL1‐positive naïve B‐cells significantly decreased with longer duration of TFR.Summary/Conclusion:Lymphocytes are part of the BCR‐ABL1 clone at diagnosis. Their relative contribution to MRD measured in the blood increases over time in TKI‐responsive patients as BCR‐ABL1+ granulocytes rapidly decline. Diminishing numbers of leukemic naïve B cells in TFR support the hypothesis that leukemic lymphocytes are long‐lived. In contrast, no patient in stable TFR had BCR‐ABL1‐positive granulocytes. The lineage‐specific assessment of MRD needs to be further explored as a means to improve the prediction of TFR.