Blinatumomab is a bispecific T-cell engager (BiTE) antibody construct targeting CD3 and CD19, resulting in T-cell-mediated lysis of malignant B cells. Currently the main application of blinatumomab is the treatment of relapsed or refractory B-cell precursor acute lymphoblastic leukemia (BCP-ALL). Although promising results were obtained therapy with blinatumomab fails in 30-60% of patients. CD19 downexpression is shown to be one of the possible mechanisms of tumor cells adaptation and subsequent escape from treatment. Moreover, loss of CD19 is a serious obstacle for flow cytometric minimal residual disease (MRD) assessment, necessary for control the leukemia response to therapy. The aim of the present study was to evaluate changes of CD19 expression in children with BCP-ALL treated with blinatumomab and possible monitoring of MRD using additional markers.
Patients and methods. From December 2015 to July 201739 (18 F/21M) patients with median age of 8 years (8 mo-20 y) with relapsed/refractory (R/R) BCP-ALL completed at least one course of blinatumomab. 5 patients (13%) carried various types of KMT2A-gene rearrangements and in 3 (7,5%) children ETV6-RUNX1 fusion transcript was detected. MRD monitoring was performed by 8-color flow cytometry. After CD19-directed treatment CD22 and CD24 were added to conventional antibodies' combinations for MRD detection (as suggested by Cherian S. et al, Cytometry B. 2016) to avoid negative influence of potential CD19 loss on cytometric data analysis and interpretation. MRD-negativity was defined as undetectable at the sensitivity level of 0.01%.
Results. 20 out of 39(51%) patients continue in complete continuous remission(CCR) after one or two blinatumomab courses, followed by alloHSCT.
In 8 (40%) patients in CCR, MRD was never detected since completion of first blinatumomab course, and they were excluded from further analysis. In other 12 pts were found to be MRD-positive at least once after blinatumomab administration. Among them in 5 patients leukemic blasts were still CD19-positive and in 7 patients residual leukemic cells were CD19-negative.
Among 19 out of 39 patients (49%) who were resistant to(n=7) or developed bone marrow relapse after blinatumomab (n=12), in 12 (63%) cases tumor cells at relapse were CD19-positive, whereas 5 children experienced CD19-negative relapse. Among them only CD19-positive MRD was detected in one case, only CD19-negative residual blasts - in two cases and in other two children both types of CD19 expression were found atdifferent time-points.
Two additional patients, who had CD19-negative MRD, then developed relapse as a lineage "switch" from BCP-ALL to acute myeloid leukemia (AML). In these patients myeloid switch developed in a different ways: one patient carried KMT2A-AF4 fusion gene which is more or less typical for switching AL, the second patient with germline KMT2A hadvery small AML population at the time of relapse diagnosis. Thus in this case positive selection of myeloid subclone during the blinatumomab treatment could be the main pathway of AML development.
Conclusions. In our group the remission rate after blinatumomab treatment was about 50%. In those children, who didn't respond and/or developed subsequent relapse, CD19 downexpression seems to be one of the significant pathways for tumor escape from blinatumomab - 26% of relapses were CD19-negative, and in 42% of all relapsed/resistant casesCD19-negative blasts were detected at least on MRD level. Rather frequent loss of CD19 could be a serious obstacle for flow cytometric MRD monitoring, but application of expanded antibodies panel and use of 10-12-color investigation allows residual blasts detection in nearly all cases. Other mechanisms of leukemia resistance including myeloid switch (even in KMT2A-germline patients) can cause blinatumomab failure.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.