We performed a multicenter retrospective study of the significance of the addition of cytarabine (CA) or thiotepa (TT) to the myeloablative standard regimen consisting of total body irradiation (TBI) and cyclophosphamide (CY).
Among the patients who underwent allogeneic hematopoietic stem cell transplantation between 2,000 and 2,010 in the four institutions related to Yokohama City University Hematology Group, a total of 365 patients who used the TBI/CY-based regimen were included to the cohort. Conditioning regimens were distributed to three groups: TBI/CY group (TC, n=82), TBI/CY/CA group (TCC, n=90), and TBI/CY/TT group (TCT, n=193). A standard TBI and CY regimen consisted of 12 Gy of TBI with four fractions and 60 mg/kg of CY for two days. Two days of CA (2 g/m2/day) or TT (200 mg/m2/day) were added to the TBI/CY regimen. A minimum dose reduction was permitted according to an older age or poorer co-morbidity. Overall survival (OS) and disease-free survival (DFS) were estimated by the Kaplan-Meier method. Gray's test was employed for comparison of cumulative incidence curves for relapse and non-relapse mortality (NRM). NRM and relapse were competing events. Adjusted multivariate Cox regression models were employed for each disease stage.
The median age was 42 years old (16–62). Diagnoses were acute myeloid leukemia (AML, n=219), acute lymphoid leukemia (ALL, n=105), and myelodysplastic syndrome (MDS, n=41). Complete remission or refractory cytopenia with multilineage dysplasia was defined as the standard stage (n=211), while the others were advanced (n=154). The stem cell sources were related bone marrow (RBM, n=79), related peripheral blood (RPB, n=51), unrelated bone marrow (UBM, n=161), and unrelated cord blood (UCB, n=74). As for patient characteristics, there was a deviation in the transplant year (P<0.001), disease stage (P=0.001), and stem cell source (P=0.022). Among the 195 patients with a previous transplant year (2000–2005), 187 patients (95.9%) received the TCT regimen, while only 6 patients with the TCT regimen were included in the recent transplant (2006–2010). Among 82 patients of the TC group, 62 patients (75.6%) were defined to be in the standard stage.
The cumulative incidence of engraftment was 89.5% with a median of 17 days (8–48). Cumulative incidences of acute graft-versus-host disease (GVHD) (grade 2 to 4, at day 50) and chronic GVHD (at 1 year) were 47.0 and 42.3%, respectively, with no significant difference between the three conditioning regimens. The median follow up period among survivors was 61.4 months (12.4–144). Outcomes at 5 years in each group (TC/TCC/TCT) were as follows: OS rates were 51.2, 34.4, and 40.3 (P=0.107), respectively; DFS rate were 50.8, 31.5, and 38.5 (P=0.177), respectively; NRM rates were 23.8, 35.5, and 40.5 (P=0.045), respectively; relapse incidences (RI) were 21.4, 29.6, and 20.8 (P=0.193), respectively. In standard stage, hazard risks (HR) for each regimen (TC is referent) were as follows; OS (TCC, HR=1.19, 95%CI: 0.64–2.01, P=0.70; TCT, HR=1.08, 95%CI: 0.66–1.78, P=0.75), NRM (TCC, HR=1.32, 95%CI: 0.61–2.85, P=0.48; TCT, HR=1.55, 95%CI: 0.82–2.94, P=0.17), and RI (TCC, HR=0.78, 95%CI: 0.26–2.33, P=0.65; TCT, HR=0.69, 95%CI: 0.28–1.72, P=0.42). In advanced stage, HR for each regimen (TC is referent) were as follows; OS (TCC, HR=0.94, 95%CI: 0.48–1.81, P=0.85; TCT, HR=1.16, 95%CI: 0.62–2.17, P=0.64), NRM (TCC, HR=1.39, 95%CI: 0.51–3.77, P=0.52; TCT, HR=2.68, 95%CI: 1.06–6.78, P=0.04), and RI (TCC, HR=1.05, 95%CI: 0.44–2.50, P=0.92; TCT, HR=0.69, 95%CI: 0.30–1.57, P=0.37). In AML, the TC group showed superior OS and NRM compared with the TCT group (54.0 vs. 36.9%, P=0.034; 23.1 vs. 42.4%, P=0.023, respectively). In MDS, the TCT group showed a lower RI compared with the TCC group (0 vs. 34.5%, respectively, P=0.012).
The benefit of the addition of CA or TT to the TC regimen was not demonstrated for each risk group. Especially in AML, TCT lead to inferior survival compared with TC. However, a bias in patient characteristics has to be considered, and a prospective study is needed in the future.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.