Figure 3.
Change in 5-year OS probabilities with allo-HCT prior to AML transformation, compared with no allo-HCT, derived from the multistate models with landmarks by 6 months from diagnosis to 24 months. The multistate model was used to predict 5-year OS for reference patients defined by patient sex, age at diagnosis, and CPSS risk group at diagnosis in (A) 40-year-old, (B) 50-year-old, (C) 60-year-old, and (D) 70-year-old patients, respectively, for landmark times 6, 12, 18, and 24 months since diagnosis (x-axis). Predictions are conditional on surviving without transformation to AML until the landmark time. At each landmark time, the difference between the 5-year OS predicted for patients who have undergone allo-HCT without prior transformation to AML at any time between diagnosis and the landmark time and that of patients who have not transformed to AML and have not been transplanted (yet) is displayed in the y-axis. Positive values indicate advantage of undergoing allo-HCT, and negative values suggest that not-transplanted patients have better prognosis.