Reduced Intensity Stem Cell Transplantation for Accelerated-Phase Myelofibrosis

Background

Circulating peripheral blasts ≥1% have long been considered an unfavorable feature for patients with primary myelofibrosis. Recent findings (Masarova et al. Cancer 2020) suggested more differentiated impact of blasts on outcome. However, accelerated-phase (AP) myelofibrosis, which is currently defined by circulating blasts 10-19%, usually confers worse outcome. The outcome of allogeneic stem cell transplantation for AP myelofibrosis has not been evaluated yet.

Patients and Methods

Thirty-five out of 349 patients with primary or secondary myelofibrosis undergoing reduced intensity allogeneic stem cell transplantation were reported as AP (10-19% blasts) at time of transplantation. Outcome of these patients was compared to patients with circulating blasts: 0% (n=135), 1-4% (n=146), and 5-9% (n=33). Conditioning consisted of busulfan/fludarabine, fludarabine/melphalan, or fludarabine/TBI2Gy.

Results

Characteristics. The median blast percentage in the AP group was 14% (10-19%). More patients in the AP group appeared to have secondary myelofibrosis (40%) compared with patients with the 0% blasts (21%), the 1-4% blasts (31%), and 5-9% blasts group (30%, P=0.08). The median hemoglobin levels decreased with blast group: 9.7g/dl (0%), 9.5g/dl (1-4%), 9.4g/dl (5-9%), and 9.0g/dl (AP). In contrast, median leukocyte counts appeared to be increased: 7.2, 9.5, 16.9, and 13.6 x 10 ⁹/l, respectively. More patients in the AP group presented with constitutional symptoms (71%). There was no difference in frequency of driver mutations (P=0.57) and presence of high-risk mutation profile (defined as ASXL1, SRSF2, IDH1/2, EZH2; P=0.93). Most patients in each group received matched unrelated donor transplants (P=0.61).

Survival. The median follow-up of all patients was 6 years. Follow-up was similar across the 4 groups (P=0.30). 5-year overall survival (95% confidence interval) according to blast group was 66% (58-73%) for the 0%, 62% (53-71%) for the 1-4%, 66% (50-81%) for the 5-9%, and 68% (53-83%) for the AP group (P=0.92). Median overall survival was not reached for all groups, except for 1-4% blasts group (17.1 years). 10-year long-term follow-up showed survival rates of 64% for the 0%, 58% for the 1-4%, 66% for the 5-9%, and 68% for the AP group. In terms of relapse-free survival, 5-year outcome was comparable (P=0.95) showing 57% (48-66%) for the 0%, 52% (43-60%) for the 1-4%, 55% (37-73%) for the 5-9%, and 52% (34-69%) for the AP group. Median relapse-free survival was 7.9 years, 5.7 years, 6.5 years, and 9.2 years, respectively. Taking blasts as continuous variable in spline function analyses on survival, no significant effect was identified overall, while after 15% blasts, risk for death appeared to increase consistently.

Non-relapse mortality and relapse. In terms of non-relapse mortality, no difference was found between the groups (P=0.33). 5-year outcome was 25% (19-31%) for the 0%, 33% (25-40%) for the 1-4%, 31% (15-47%) for the 5-9%, and 17% (5-30%) for the AP group. In terms of cumulative incidence of relapse, the AP group showed 5-year outcome of 31% (15-47%) compared to 18% (12-24%) for the 0%, 16% (10-22%) for the 1-4%, and 14% (1-28%) for the 5-9% group (P=0.17). Pairwise comparison showed significant difference between AP and the other groups (P=0.03). Spline function analyses using blasts as continuous variable and 3 knots showed no significant effect on non-relapse mortality but suggested increased risk of relapse for AP myelofibrosis (P=0.04).

Multivariable analysis. Adjusting for clinical (diagnosis, leukocyte and platelet counts, age, constitutional symptoms) and molecular-genetic (driver mutation genotype, high-risk mutations), Cox model on survival (with the 0% group as reference) showed comparable risk of death across blast groups with hazard ratios of 1.06 (0.70-1.62) for the 1-4%, 1.11 (0.56-2.32) for the 5-9%, and 0.95 (0.49-1.86) for the AP group. Independent factors for worse outcome were age, platelet and leukocyte counts, and CALR/MPL-unmutated genotype.

Conclusion

Reduced intensity allogeneic stem cell transplantation for AP myelofibrosis was associated with excellent and similar survival and non-relapse mortality in comparison with other blast groups and after adjustment for other risk factors, offering long-term survival (>10 years) for more than half of patients at AP. Relapse incidence appeared to be increased for AP myelofibrosis.