Telomere shortening correlates with leukemic stem cell burden at diagnosis of chronic myeloid leukemia.

Telomere length (TL) in peripheral blood (PB) cells of patients with chronic myeloid leukemia (CML) has been shown to correlate with disease stage, prognostic scores, response to therapy, and disease progression. However, due to considerable genetic interindividual variability, TL varies substantially between individuals, limiting its use as a robust prognostic marker in individual patients. Here, we compared TL of BCR-ABL-, nonleukemic CD34+CD38- hematopoietic stem cells (HSC) in the bone marrow of CML patients at diagnosis to their individual BCR-ABL+ leukemic stem cell (LSC) counterparts. We observed significantly accelerated telomere shortening in LSC compared with nonleukemic HSC. Interestingly, the degree of LSC telomere shortening was found to correlate significantly with the leukemic clone size. To validate the diagnostic value of nonleukemic cells as internal controls and to rule out effects of tyrosine kinase inhibitor (TKI) treatment on these nontarget cells, we prospectively assessed TL in 134 PB samples collected in deep molecular remission after TKI treatment within the EURO-SKI study (NCT01596114). Here, no significant telomere shortening was observed in granulocytes compared with an age-adjusted control cohort. In conclusion, this study provides proof of principle for accelerated telomere shortening in LSC as opposed to HSC in CML patients at diagnosis. The fact that the degree of telomere shortening correlates with leukemic clone's size supports the use of TL in leukemic cells as a prognostic parameter pending prospective validation. TL in nonleukemic myeloid cells seems unaffected even by long-term TKI treatment arguing against a reduction of telomere-mediated replicative reserve in normal hematopoiesis under TKI treatment.