Evidence for Premature Aging of (Clonal but also) Non-Clonal Hematopoietic Stem Cells in Patients with Acute Myeloid Leukemia (AML)

Introduction:

AML is an aggressive clonal myeloid neoplasia. Incidence increases with age and age is also the most dominant prognostic indicator for adult patients with AML. Although numerous genetic events have recently been identified in good and intermediate risk AML, understanding of the pathophysiology behind poor risk AML, which is most predominant in elderly patients and characterized by complex karyotypic abnormalities remains limited. Changes in the epigenetic signature and telomere attrition are two well recognized hallmarks of aging and genetic instability. Here, we investigate the hypothesis that poor risk AML arises from a prematurely aged, genetically instable hematopoietic stem cell population.

Methods:

We analyzed 45 newly diagnosed AML patients who were followed from diagnosis, through achievement of complete cytogenetic remission (CCR) and/or until refractory disease or relapse. Overall, 95 samples (n=88 bone marrow (BM), n=7 peripheral blood, (PB)) were available. Analysis were performed at first diagnosis (Dx, n=20), after two cycles of induction chemotherapy (post-IC, n=42), after three additional cycles of consolidation chemotherapy (post-CC, n=28) and after one year follow up (n=5). Mean age of the cohort was 50.4 years (range 21-75y). A recently developed “epigenetic aging signature (EAS)” capable to predicted to the chronological age of healthy individuals with a mean absolute deviation (MAD) of 5 years (Weidner et al., Genome Biol, 2014) was used to express the epigenetic age of the AML samples. Epigenetic changes in DNA methylation were analyzed by pyrosequencing and monochrome multiplex quantitative-PCR was used for telomere length (TL) analysis. TL was age-adapted based on a control cohort of 87 healthy donors as described previously.

Results:

A cross-sectional analysis of TL reveals substantial telomere shortening in AML patients at the time of first diagnosis compared to healthy controls (mean±SE: -0.65 ± 0.2 T/S ratio, p=0.001). When studied sequentially, the TL deficit was less pronounced in remission after IC (-0.33 ± 0.1 T/S ratio, n=34) however, still significantly shortened compared to age-adapted healthy subjects (p= 0.001). This tendency continued in patients in remission after CC (-0.18 ± 0.1 T/S ratio, n=24, p=0.03) and one year (-0.25 ± 0.1 T/S ratio, n=5, p=0.28). In comparison, in patients with active leukemia, i.e. persistent AML post-IC (-0.61 ± 0.2 T/S ratio, n=8, p=0.006) or relapsed AML post-CC (-1.0 ± 0.2 T/S ratio, n=4, p=0.04), TL was comparable to untreated, diagnostic samples.

By using EAS for age prediction, leukemic cells from AML patients with active disease at first diagnosis (+29.6 ± 8.6 years, p=0.001), in persistence (+28.9 ± 11.2 years, p=0.01) and in relapse (+86.8 ± 29.3 years, p=0.01) were found to be significantly and prematurely aged compared to their respective chronological age. In contrast, premature aging based on EAS was substantially less pronounced in CCR post-IC (+4.5 ± 1.8 additional years, p=0.01) and post-CC (+1.4 ± 2.3 additional years, p=0.52).

Conclusion: We demonstrate that hematopoietic cells from AML patients at diagnosis are characterized by premature aging indicated both by accelerated telomere shortening and EAS. Not unexpectedly, this premature aging signature of leukemic cells persists in refractory and/or relapsed patients. However interestingly, this phenomenon (although less pronounced) is not completely reverted once patients reach cytogenetic remission either post induction chemotherapy, post consolidation or one year after diagnosis, i.e. when hematopoiesis has shifted from clonal leukemic to non-clonal cells. These data support the conclusion that either normal cells in patients with AML undergo accelerated aging during conventional treatment or alternatively, AML arises from prematurely aged (yet non-clonal cells) hematopoietic stem cells.