The role of HLA-matched unrelated transplantation in adult patients with Ph chromosome-negative ALL in first remission. A decision analysis

Abstract

The efficacy of unrelated transplantation for patients with ALL who lack an HLA-matched sibling remains unclear. We performed a decision analysis to determine the efficacy of myeloablative transplantation from a genetically HLA-A, -B, -DRB1 allele-matched unrelated donor for patients with Ph chromosome-negative ALL aged 21-54 years. The transition probabilities were estimated from the Japan Adult Leukemia Study Group studies (ALL93; n ¼ 80, ALL97; n ¼ 82), and the Japan Marrow Donor Program database (transplantation in first CR (CR1): n ¼ 177). The primary outcome measure was the 10-year survival probability with or without quality of life (QOL) adjustment. Subgroup analyses were performed according to risk stratification based on the WBC count and cytogenetics, and according to age stratification. In all patients, unrelated transplantation in CR1 was shown to be superior in analyses both with and without QOL adjustment (40.8 vs 28.4% and 43.9 vs 29.0%, respectively). A similar tendency was observed in all subgroups. The decision model was sensitive to the probability of leukemia-free survival following chemotherapy and the probability of survival after transplantation in standard-risk and higher-aged patients. Unrelated transplantation in CR1 improves the long-term survival probability in patients who lack an HLA-matched sibling. However, recent improvements in treatment strategies may change this result.

Unrelated bone marrow transplantation or immediate umbilical cord blood transplantation for patients with acute myeloid leukemia in first complete remission

“…For the QOL‐adjusted analysis, we adjusted the 5‐year survival rate by including the presence or absence of persistent GVHD rather than using QOL‐adjusted life years. Under these conditions, we decided to use a simple decision analysis model instead of a Markov model, because the latter has limited advantages in this situation . In addition, a large number of patients are required for a Markov model to determine appropriate TPs that change with time, and the number of our patients, particularly of those who received UCBT and 7/8 UBMT, might not have been large enough to construct a valid Markov model.…”

“…Given that prospective comparisons of UCBT vs. UBMT are unlikely to be performed because of the highly complex nature of such a trial, search for an alternative approach was warranted. We therefore employed a decision analysis, which is a statistical technique used to aid medical decision‐making when conditions of uncertainty exist . With this methodology, we attempted to evaluate which transplantation procedure should be prioritized for patients with high‐ or intermediate‐risk AML in CR1 who lack a matched related donor.…”