FATE and Role of Peripheral Blood CD26+ Leukemia STEM CELLS at Diagnosis in Chronic Myeloid Leukemia Patients: FINAL Results of Prospective Flowers Study

Background: In a cross-sectional study we previously demonstrated that in peripheral blood (PB) of chronic myeloid leukemia (CML) patients leukemia stem cells (LSCs) CD26+ are detectable by flow-cytometry at diagnosis, during TKI therapy and during treatment free remission (TFR). No prospective data are available regarding the behavior of PB CD26+LSCs from diagnosis and the correlation, if any, between the bulk of this staminal compartment at diagnosis with the attainment of molecular response.

Methods: We here present final results of a prospective Italian multicenter study including newly diagnosed chronic phase (CP) CML patients centrally monitored by flow-cytometry for PB CD26+LSCs quantification from diagnosis up to 24 months of TKI treatment.

Results: 242 consecutive CP-CML patients were enrolled (132 imatinib, 72 nilotinib and 38 dasatinib). The bulk of CD26+LSCs at diagnosis varied between patients with a median value of 7,1454 cells/µl (range 0,0126-698,746 cells/µl; IQR 2,18-33,26 cells/µl). During TKI treatment, we observed a consistent and rapid reduction of them achieving median values of 0,0132 cells/µl (IQR 0-0,034 cells/µl), 0,011 cells/µl (IQR 0-0,031 cells/µl) and 0,0071 cells/µl (IQR 0-0,0259 cells/µl) at 3, 12 and 24 months, respectively. No statistically significant differences in terms of CD26+LSCs log-reduction were noted according to the type of TKI treatment at any time points evaluated. However, a significant correlation between a low amount of CD26+LSCs at diagnosis and an optimal molecular response at 3, 12 and 24 months (BCR::ABL1<10% and BCR::ABL1<0.1%, respectively) was documented. Indeed, CML patients with optimal molecular response at 3 months had a median CD26+LSCs of 6,21 cells/µl (IQR 1,79-31,50 cells/µl) while suboptimal responders patients showed a median of 19,87 cells/µl (IQR 5,37-39,81 cells/µl) (p=0.03); moreover, patients with BCR::ABL1<0.1% at 12 and 24 months, revealed median CD26+LSCs at diagnosis of 5,50 cells/µl (IQR 1,81-22,64 cells/µl) and 6,05 cells/µl (IQR 1,79-29,90 cells/µl) respectively, compared to suboptimal responders showing 16,87 cells/µl (IQR 2,82 -71,77) and 20,52 cells/µl (IQR 4,24-106,91) (p=0.004, p=0.009). Additionally, evaluating the cohort of CML patients who switched TKI treatment after failure with respect to patients who did not change TKI, we observed that the former had a significantly higher median CD26+LSCs at diagnosis (14,59 cells/µl; IQR: 3,76-46,00 cells/µl) compared to the no-switch group (median of 5,82 cells/µl; IQR: 2,35-26,70 cells/µl) (p=0.034). Three ranges of CD26+LSCs correlating to molecular response were identified: <3,21 cells/µl (1° tertile); between 3,21-19,21 cells/µl (2° tertile); >19,21 cells/µl (3° tertile). In particular, considering the molecular response at 3 months the incidence of CML patients with BCR::ABL1<10% was 93.5% in the first CD26+LSCs tertile, while 78.8% in the third tertile (p=0.027). At 12 months the incidence of optimal response in the first tertile was 78.5% and 62.8% in the third one (p=0.015). At 24 months the two incidences were 90.8% and 77.9%, respectively (p=0.079).

Conclusions: This prospective study demonstrated a rapid rate of reduction of CD26+LSCs during TKI treatment, however confirming their long-lasting persistence even if at very low levels. For the first time, a correlation between the amount of CD26+LSCs at diagnosis and the response to TKI treatment, was documented. Given these results, the bulk of CD26+ LSCs at diagnosis could represent an easily and rapidly measurable, new prognostic tool for predicting TKI response.