Introduction: Acquired aplastic anemia (AA) is typically characterized by pancytopenia and bone marrow (BM) failure mostly due to an autoimmune attack against the hematopoietic stem cell compartment. Thus, AA patients frequently respond to immunosuppressive therapy (IST). Hypoplastic myelodysplastic syndrome (hMDS) frequently mimics clinical and morphological features of AA and proper clinical discrimination of hMDS from AA sometimes remains difficult. Interestingly, some cases of hMDS respond at least partially to IST and on the other hand, AA can clonally evolve to hMDS. Telomeres shorten with each cell division and telomere length (TL) reflects the replicative potential of somatic cells. Whereas it is proposed that TL can to some degree discriminate hereditary subtypes of bone marrow failure syndromes from classical acquired forms, the role of TL for disease pathogenesis in hMDS remains unclear. In this study, we therefore aimed to investigate accelerated TL shortening as a possible (bio-)marker to distinguish hMDS from AA.

Patients and Methods: TL of BM biopsies at diagnosis of 12 patients with hMDS and 15 patients with AA treated in the University Hospital Düsseldorf were analyzed. Mean age was 45.2 years in AA patients and 65.2 years in patients with hMDS. Confocal Q-FISH protocol was used for TL measurement as published previously (Blood, 2012). TL analysis was performed in single-blinded fashion. 28 patients (range 18-80 years) with newly diagnosed M. Hodgkin without BM affection were used as controls for linear regression and calculation of age-adapted TL difference. For the analysis of the data, we made use of a recently developed mathematical model of TL distribution on the stem cell level allowing us to extrapolate mean TL shortening per year (TS/y) based on the individual TL distributions of captured BM biopsies.

Results: Using the controls to adjust for age, we found that age-adapted TL was significantly shortened both in patients with AA (median: -2.96 kb, range -4.21 to 0.26, p=0.001) and patients with hMDS (median: -2.26, range -3.85 to -0.64, p=0.005). In direct comparison, telomere shortening was more accelerated in patients with AA as compared to hMDS (p=0.048). Next, we analyzed the TL shortening per year (TS/y) based on the individual telomere distribution. Calculating the extrapolated TL shortening per year (TS/y), we found significant increased TS/y in AA patients (mean±SD: 235,8 bp/y±202,9, p=0.001) and hMDS patients (120,5±41,7 bp/y, p=0.0001) compared to controls (37,5±18,9 bp/y). Interestingly, the extrapolated rate of TS/y remained stable at different ages in hMDS patients as observed in healthy controls. In contrast, TS/y in AA patients showed a strong age-dependence with a maximum of TS/y in patients younger than 30 years (R²=0.42, p=0.008). Finally, we set to test whether TS/y can be used to identify AA or hMDS patients. Using 150 bp TS/y as a cut-off (4-fold the mean of controls), we found significantly more AA patients (10/15, 66.7%) had accelerated TL shortening compared to hMDS (1/12, 8.3% p=0.005).

Conclusion: We provide first retrospective data on TL in patients with hMDS using confocal Q-FISH. Age-adapted TL is significantly shorter in patients with AA compared to hMDS. Interestingly, telomere shortening per year is both significantly increased in AA as compared to hMDS patients as well as in both groups compared to controls. The rate of telomere shortening TS/y might represent a new marker in patients with bone marrow failure syndromes that allows to discriminate AA from hMDS patients pending prospective validation.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.