Does an increased probability of graft-vs-host disease improve the survival of patients with adult T-cell leukemia-lymphoma? A simulation analysis using a Markov model

Adult T-cell leukemia-lymphoma (ATL) is a peripheral T-cell lymphoma caused by human T-cell lymphotropic virus type I. Previous retrospective studies suggested the presence of graft-vs-ATL (GV-ATL) effects associated with acute graft-vs-host disease (GVHD), which suggests that intentional induction of acute GVHD to facilitate the efficacy of GV-ATL effects might reduce the risk of relapse and contribute to improved survival after allogeneic hematopoietic stem cell transplantation. However, it is impractical to conduct a prospective study to assess the benefit of intentional induction of acute GVHD. In a simulation analysis using a Markov model, we assessed the impact on overall survival (OS) of changing the probability of overall acute GVHD and severe acute GVHD. When the probability of grade III-IV acute GVHD changed from 0% to 100%, the expected 2-year OS rate changed from 57.0% to 21.8% (48.2% of baseline at 2 years). When the probability of all grades of acute GVHD changed from 0% to 100%, the expected 2-year OS rate changed from 48.2% to 49.5%. In conclusion, increasing the probability of overall acute GVHD was beneficial only when the proportion of grade III-IV acute GVHD was lower than that in the original data. Based on the simulation results, preventive management of grade III-IV acute GVHD is mandatory to achieve the clinical benefit associated with grade I-II GVHD.