Leukemic Stem Cell Quantification Is Of Prognostic Value In Newly Diagnosed Patients In Chronic Phase Chronic Myeloid Leukemia (CML-CP) Receiving Nilotinib Therapy: Results From The ENEST1st Stem Cell Substudy

Leukemic stem cells (LSCs) are considered to be very important therapeutic targets in CML, as it has been shown both in vitro and in vivo that tyrosine kinase inhibitor (TKI) therapy is not able to eradicate these entirely. We previously reported that LSC burden at diagnosis has significant prognostic impact on therapy outcome in imatinib and dasatinib treated CML patients. The ENEST1st study (NCT01061177) is focused on examining the role of first-line nilotinib therapy in patients in CML-CP. This translational, multinational, multicenter ENEST1st substudy addresses the predictive value of the stem cell profile in these patients.

Bone marrow (BM) and/or peripheral blood (PB) samples were collected from newly diagnosed CML-CP patients before and after 1 and 3 months of nilotinib treatment. The LSC burden was analyzed by FISH and flow cytometry (FC). With the FISH method, CD34⁺ cells were pre-selected with paramagnetic beads after which they were fractionated into CD38⁺ (upper 80%, “progenitor”) and CD38^- (lowest 5%, “stem cell”) pools using flow sorting. The proportion of Ph+ cells was assayed by counting 1,000 cells with interphase FISH using specific BCR-ABL1 probes. In the FC method, the stem cell compartment was defined as the lowest 1% of CD34⁺CD38^- cells. LSCs were distinguished from normal hematopoietic stem cells (nHSCs) by either higher light scatter properties, and/or aberrant expression of CD7, CD11b and CD56, and/or higher CD45 and CD90 expression. Patient samples were assigned to as “residual nHSCs present” or “no residual nHSC present” at diagnosis and as “residual LSCs present” or “no residual LSCs present” at 1 and 3 months. Altogether, 48 patients from 6 European countries were investigated.

By FISH analysis at diagnosis, the proportion of BCR-ABL+ cells in the stem cell compartment varied markedly in individual patients (1%-100%) and was lower (85%) than in progenitor (96%) or whole BM fractions (96%, p<0.05). This LSC burden correlated significantly with high WBC count (r=0.44, p=0.008), hemoglobin (r=-0.51, p=0.002), spleen size (r=0.51, p=0.002) and PB (r=0.57, p=0.0002) and BM (r=0.55, p=0.0005) blast percentage. Weak correlation was also found with Sokal score (r=0.37, p=0.03). By FC method, similar to FISH, the range of LSCs was 0%-100% with a median of 73%. Forty percent of patients had detectable residual nHSC at diagnosis. These patients had significantly lower Sokal score (p=0.001), smaller spleen size (p=0.006), lower platelet count (p=0.04) and lower PB blast percentage (p=0.001). With both methods no prognostic relationship existed between LSC burden and eosinophil/basophil counts. Interestingly, by FISH method, the LSC burden at diagnosis correlated also with BCR-ABL transcript levels according to the international scale at 3 (r=0.39, p=0.03), 9 (r=0.41, p=0.01) and 18 months (r=0.42, p=0.02). Borderline significant associations were also found at 6 (p=0.06) and 12 (p=0.07) months. All patients lacking major molecular remission at 12 or 18 months, had >80% of BCR-ABL+ cells in the stem cell fraction at diagnosis. When LSCs were analyzed by the FC method, patients with residual nHSCs at diagnosis had significantly lower BCR-ABL levels at 3 (0.27 vs 1.74, p=0.025) and 6 (0.13 vs 1.21, p=0.03) months, but this was not reflected in lower BCR-ABL levels at 12 or 18 months. During nilotinib therapy, the proportion of BCR-ABL⁺ cells by FISH decreased rapidly both the in progenitor and stem cell compartments. At 3 months the median LSC percentage was 0.28%, and it did not differ significantly from the values detected either in the progenitor fraction (0.34%) or whole BM (0.29%). Similarly, only a few patients had detectable LSCs left during follow-up time-points with the FC method.

LSC burden at diagnosis reflects the biology of the disease in newly diagnosed CML-CP patients. It also carries a prognostic value in first-line treated nilotinib patients, and a substantial number of patients not meeting the optimal response criteria at the 12-month time-point have high LSC burden at diagnosis. Furthermore, the presence of residual nHSCs at diagnosis is predictive for molecular response at 3 and 6 months. Nilotinib therapy markedly decreases LSC pool during the first 3 months of treatment. Noteworthy, the reduction is as potent in the LSC compartment as it is in more mature progenitor cell or whole BM fractions.

Richter:Novartis: Consultancy, Research Funding, Speakers Bureau; Bristol Myers Squibb: Consultancy, Speakers Bureau. Fioretos:Novartis: unrestricted research grant Other. Giles:Novartis: Consultancy, Research Funding. Hochhaus:Novartis: Consultancy, Honoraria, Research Funding, Travel Other; BMS: Consultancy, Honoraria, Research Funding; Pfizer: Consultancy, Honoraria; Ariad: Consultancy, Honoraria. Ossenkoppele:Novartis: Consultancy, Research Funding, Speakers Bureau; Bristol Myers Squibb: Consultancy, Speakers Bureau. Porkka:Novartis: Consultancy, Research Funding, Speakers Bureau; BMS: Consultancy, Research Funding, Speakers Bureau. Wolf:Novartis: Honoraria, Research Funding; Pfizer: Honoraria; Bristol-Meyers Squibb: Honoraria. Janssen:Novartis: Consultancy, Research Funding. Mustjoki:Novartis: Consultancy, Speakers Bureau; BMS: Consultancy, Speakers Bureau.