Abstract 3208

Poster Board III-145

In murine models, telomere erosion promotes chromosomal instability via breakage-fusion-bridge cycles, contributing to the early stages of tumorigenesis. However, direct evidence that short telomeres predispose to cancer development in humans is lacking. In acquired aplastic anemia, evolution to malignant clonal disorders is a major complication after immunosuppressive therapy, affecting up to 15 percent of patients at 10 years. We investigated whether telomere length measured at diagnosis predicted clonal evolution in these patients. Telomere length was measured from DNA extracted from peripheral blood leukocytes collected at disease presentation in 183 consecutive patients enrolled in successive clinical trials for immunosuppressive regimen as first line therapy for severe aplastic anemia at the Clinical Research Center, National Institutes of Health (ClinicalTrials.gov identifier numbers, NCT00001964, NCT00260689, and NCT00061360) and 164 healthy volunteers. Leukocyte telomere length of aplastic anemia patients at diagnosis was in the normal range and was not shorter than in healthy controls (ANOVA-F test). Telomere length was corrected for age and patients were separated into two groups: patients with short telomeres (in the lowest quartile) and long telomeres (other quartiles). Telomere length was a critical and independent predictive biomarker for evolution to myelodysplastic syndrome, especially monosomy 7, and acute myeloid leukemia (AML) in patients with acquired aplastic anemia (Multivariate Cox Proportional Hazard Model, P=0.006). Patients with short telomeres had six-fold higher probability to develop clonal malignant disease than did patients with longer telomeres. Bone marrow cells of aplastic patients were cultured in vitro for short term in the presence of cytokines and high-dose granulocyte-colony stimulating factor (G-CSF) and cells of patients with short telomeres (n=5) showed increased telomere-free chromosomal ends in comparison to cells of patients with long telomeres (n=6), by fluorescence in situ hybridization (FISH; P<0.0001). Spectral karyotyping (SKY) revealed that cultured bone marrow cells of patients with short telomeres exhibited aneuploidy and translocations, including Robertsonian translocations, which were not found in cells of patients with long telomeres. Bone marrow cells at diagnosis were further evaluated for the presence of monosomy 7 cells using interphase FISH in 73 patients. Telomere length inversely correlated with the frequency of monosomy 7 cells: the shortest the telomeres, the highest the percentage of aneuploid cells at diagnosis (Pearson r=-0.5110; P=0.0009). We further employed bone marrow cells of clinically healthy individuals carrying loss-of-function telomerase mutations and with extremely short telomeres (n=5) as a model for telomere dysfunction in hematopoietic cells in the absence of human disease. In vitro culture of these cells yielded aberrant karyotypes by SKY, including translocations and aneuploidy, and end-to-end chromosomal fusions by FISH. These results indicate that telomere length at diagnosis predicts evolution to myelodysplasia and leukemia in patients with acquired aplastic anemia treated with immunosuppression. Our findings support the hypothesis that short and dysfunctional telomeres restrain stem cell proliferation and predispose for malignant transformation by selecting stem cells that are prone to chromosomal instability. This is the first prospective study to demonstrate that short telomeres in human hematopoietic cells promote chromosomal instability in vitro and predispose to malignant transformation in humans.


Cooper:NIH-Pfizer: Research Funding.

Author notes


Asterisk with author names denotes non-ASH members.