Comparison of Outcomes of Elderly Patients with Acute Myeloid Leukemia (AML) in CR1 Receiving Reduced Intensity Allogeneic Stem Cell Transplantation (RICT) from Matched Sibling (MSD) or Unrelated Donors (URD)

BACKGROUND. Comparative data on the results of RICT from MSD or URD in elderly patients (pts) are scarce. We describe outcomes of AML CR1 pts > 50 years (yrs) allografted from MSDs or URDs as part of a multinational, prospective, controlled study (NCT00342316), which aims to compare RICT vs chemotherapy. The study started in 2004 and completed enrollment (339 pts) in 2016. In this report, we compare the impact of MSD or URD and therefore we focus only on the time period when both MSD and URD transplants were performed within the study (2010-2016).

PATIENTS AND PROCEDURES. Included were 136 pts (53 females, 83 males). Median age was 64 (50-70) and 62 (52-70) yrs in URD and MSD groups, respectively. Pts were high (54%) or intermediate (IR) (46%) risk, donors were URD (n=81), and MSD (n=55). Prespecified high risk criteria were: poor risk cytogenetics, secondary or therapy-related AML, blasts >15% after induction #1, or FLT3-ITD+/NPM1-. Eight URDs were <8/8 HLA matched. Reduced intensity conditioning was busulfan/fludarabine based in 92% of transplants (Tx). Anti thymocyte globulin (ATG) pre Tx was used in 93% of URD and 25% of MSD pts. Pts received PBSC (94%) or BM (6%). Distribution of high/intermediate risk pts were: URD 58%/42% and MSD 48%/52% (p=0.29.)

STATISTICS: Overall survival (OS) and leukemia free survival (LFS) after Tx were estimated by means of the Kaplan-Meier method, whereas cumulative incidence functions were used for non-relapse mortality (NRM) and relapse incidence. The Cox regression model was used to analyze incidence rates.

RESULTS. Median time from CR to Tx in URD and MSD groups was 4.0 (range 1.1-8.7), and 3.1 (0.8-10) mo, respectively (p<0.001). URD pts received more chemo pre Tx: ≥3 courses in 85% vs 60% (MSD group) p=0.001. Graft rejection occurred in 1 pt.

Median follow-up from Tx for living pts was 23 (4-78) and 25 (3-82) mo for URD and MSD pts, respectively. During follow-up, 36 pts relapsed, 24 URD (30%) and 12 MSD (22%) and 47 pts died, 23 URD (28%) and 24 MSD (44%). Causes of deaths in the URD and MSD groups were AML (n=18,11), GvHD (1,5), infections (2,5) and other (2,3).

Estimated 2-year OS for all pts was 64% (95% CI 54-72%). For the subgroups of high and IR risk pts OS at 2 yrs was 49% (36-61%) and 82% (69-90%), respectively. For MSD and URD groups OS were 53% (38-67%) and 72% (60-81%); hazard ratio (HR) 1.66 (0.94-2.95; p=0.08). If adjusting for confounders [risk category (high vs . IR), country (Canada/other vs . Sweden), gender, age] the HR for OS was 1.57 (0.80-3.08; p=0.19). LFS was similar in URD and MSD groups: 53% (38-66%) vs 63% (50-73%); adjusted HR 1.21 (0.65-2.26; p=0.54). The 2-yrs cumulative incidence (CI) of relapse (NRM competing) was 23% (13-35%) in the MSD group and 32% (21-43%) in the URD group; adjusted HR 0.71 (0.32-1.60; p=0.41). The 2-yrs CI of NRM (relapse competing) was 24% (13-37%) in the MSD group and 5% (2-11%) in the URD group, adjusted HR 3.45 (1.03-11.6; p=0.045).

Grade III-IV acute GvHD and chronic extensive GvHD were less common in the URD group, 5% vs 21% (p=0.009) and 10% vs 45% (p <0.001), respectively. At 12 mo after CR1, 18% of URD pts and 58% of RD pts were on corticosteroids (p<0.001).

CONCLUSION. No significant differences in OS, LFS or relapse rate were found between MSD and URD transplants. However, RICTs from URDs were associated with less acute and chronic GvHD, and less NRM. These findings may be explained by lower donor age, more chemo, longer time from CR to Tx in MUDs and probable contribution of ATG. Our data suggest that URD RICT may be used safely in older patients with AML with satisfactory OS.