A matter of size: Telomere length predicts survival before allo-SCT in patients with MDS

Myelodysplastic syndromes (MDS) can be cured with allogeneic stem cell transplantation (allo-SCT) but transplant-related complications reduce long-term survival. The current genetic classification of MDS identifies patients with increased risk of relapse, whereas the risk of non-relapse mortality (NRM) is based on clinical assessment of comorbidities and end-organ dysfunction features.¹ Although several studies have suggested that MDS is linked to shorter telomeres, the potential association of telomere length with risk of relapse or NRM is yet unclear.¹

Therefore, a new study published by Myllymäki et al¹ in Blood assessed the impact of telomere length on clinical outcomes of patients with MDS prior to allo-SCT, and its potential association to genetic and transplant variables.

Study design

Whole blood DNA samples were extracted from a cohort of 1,514 MDS patients prior allo-SCT. Relative leukocyte telomere length was measured using quantitative real-time PCR (qPCR). Telomere measurements were successfully obtained from 1,508 patient samples. Recursive partitioning modeling with quantiles from 5–45% (multiples of 5) was used to subgroup patients into the following categories:

• 1st quartile: shortest telomere length
• 2nd & 3rd quartile: intermediate telomer length
• 4th quartile: longest telomere length

Genetic profiling was performed from whole blood samples by analyzing 129 known MDS-related genes.

Results

As MDS in younger patients has distinct disease characteristics, only patients ≥ 40 years were considered in the main analysis. In this cohort (n = 1,267), shorter telomeres were significantly associated with reduced:

• hemoglobin (p < 0.001)
• platelets (p < 0.001), and
• absolute neutrophil count (p < 0.001)

Bone marrow blast counts (p = 0.27) and Revised International Prognostic Scoring System (IPSS-R) cytogenetic group distribution (p = 0.62) were not different among patients with different telomere lengths.

There were no significant differences in telomere length in patients treated with DNA methyltransferase inhibitors, chemotherapy, or a combination of the two versus those without such pretreatment. Similarly, there was no difference in telomere length for patients having received chemotherapy or radiation for MDS compared with naïve MDS patients.

 Analysis of MDS somatic mutations showed the following significant associations with telomere length:

• Longer telomers with mutations in SRSF2 (p < 0.001), DNMT3A (p < 0.001), or STAG2 (p = 0.04)
• Shorter telomeres with mutations in PPM1D (p = 0.01), WT1 (p = 0.03), and ATM (p = 0.04)

The authors proceeded to evaluate the clinical outcomes of patients with different telomere length. In a multivariate analysis, intermediate and short telomeres were independent predictors of worse overall survival (OS) when compared with long telomeres (short versus long telomers, HR = 1.49; 95% CI, 1.22–1.83; p < 0.001; intermediate versus long telomers, HR = 1.34; 95% CI, 1.12–1.60; p = 0.001). Known prognostic factors for survival such as, age, Karnofsky index < 90, very high IPSS-R, TP53 mutations, JAK2 V617F, RAS/tyrosine kinase pathway mutations, and cord blood grafts remained significant in this model.

Shorter and intermediate telomeres, allo-SCT prior to 2008, and JAK2 V617F mutations were associated with a higher NRM rate but not with disease relapse. Predictive of disease relapse were TP53 mutations and very high IPSS-R risk scores.

Regardless of the conditioning regimen, patients with longer telomeres had significantly superior OS. However, in patients with shorter telomeres, the risk of NRM was linked to myeloablative conditioning (MAC; p < 0.001) but also to fludarabine/melphalan-based reduced-intensity conditioning (RIC) (p = 0.002). The NRM effects were not seen in the intermediate or long telomere groups.

With regards to comorbidities, although the Karnofsky score distribution was not different between patients with different telomere length, the hematopoietic cell transplantation-specific comorbidity index (HCT-CI) score was lower in patients with longer telomeres when compared with those with shorter telomeres (p = 0.005). This was linked to higher rates of pulmonary or hepatic dysfunction and infection prior allo-SCT, in patients with shorter telomeres. In patients without comorbidities, who received MAC, shorter telomeres were significantly associated with decreased OS (p = 0.0046) due to a higher rate of early NRM in patients with shorter telomeres (p = 0.02).

No significant increase in severe acute graft-versus-host disease (aGvHD) incidence among telomer groups was observed. However, patients with shorter telomeres had a significantly reduced OS after developing severe aGvHD due to the higher rate of NRM.

Conclusion

The results of this study indicate that telomere length is independently associated with OS and NRM in MDS patients prior to allo-SCT. More specifically, short telomeres were linked to inferior OS and higher NRM, irrespective of age and disease characteristics. Decreased outcomes were more pronounced in patients receiving MAC or melphalan-based RIC conditioning, suggesting that such regimens should be used cautiously in this patient subpopulation. Furthermore, short telomeres were associated with higher rates of end-organ diseases, indicating that regimens for this patient subset should focus on minimizing transplantation-related toxicity. Further data from prospective clinical trials are needed to validate the impact of telomere length on OS and NRM in MDS.