Donor Lymphocyte Infusion Vs. Second Allogeneic HCT In Relapsed AML After Allografting: A Landmark Analysis From The Acute Leukemia Working Party Of EBMT

Best management of AML relapsing after allogeneic HCT is controversial. While enrollment in clinical trials is ideal, such option would generally exclude these patients from participation. Accordingly, offering donor lymphocyte infusion(s) (DLI) or a second allogeneic (2^nd-allo) HCT are commonly pursued.

Patients (pts), disease, and treatment characteristics related to 1^st-first allograft are shown (Table 1). We compared outcomes of allografted pts receiving DLI (n=321, male=57%) or 2^nd-allo (n=297, male=54%) within 100 days from relapse/progression after 1^st-allo HCT between 01/2000-12/2011. Pts were not in remission at time of DLI or 2^nd-allo and the majority of DLI-treated received 1 dose [1 dose=220, 2 doses=68, ≥3 doses=33). Forty-six pts received a 2^nd-allo after DLI. PBSC was more commonly used for the 2^nd-allo (PBSC=277, BM=14, CB=4, unk=2). Pts receiving DLI were older [50 (18-71) vs. 43 (18-72) yrs, p<0.0001] and DLI was offered sooner [26 (1-100) vs. 44 (1-99) days, p<0.0001].

Median F/U (all pts) from time of post allo-HCT relapse was 17 (0.3-125) months. Two-year overall survival (OS) from intervention (DLI or 2^nd-allo) was poor regardless of modality offered (DLI=17±2% vs. 2^nd-allo=13±2%, p=0.14). However, 2-year OS from relapse (intervention as time-dependent variable after relapse) favored DLI [HR=1.35 (95%CI: 1.13, 1.61, P=0.001]. Multivariate analyses (Cox proportional hazard model) to evaluate OS from time-of-relapse or from time-of-intervention (DLI or 2^nd-allo) were performed. OS (from relapse) was worse when a 2^nd-allo was offered as the preferred choice [HR=1.28 (95%CI:1.05, 1.56), p=0.01]. Also, the following variables at time of 1^st-allo HCT were associated with worse OS (from relapse): age>46 yrs [HR=1.27 (95%CI:1.04, 1.55), p=0.02], use of MRD [HR=1.34(95%CI:1.11, 1.62), p=0.002], relapse <6 months from allografting [HR=0.48 (95%CI:0.40, 0.58), p<0.0001]. PBSC as cell source at 1^st allograft was associated with a strong trend for worse OS [HR=1.35 (95%CI:1.00, 1.81), p=0.05]. OS (from time of DLI or 2^nd allo-HCT) was better when median time from relapse to intervention of > 33 days [HR=0.79 (95%CI:0.65, 0.96), p=0.01], worse for age>46 yrs [HR=1.25 (95%CI:1.03, 1.53), p=0.03], URD [HR=1.40 (95%CI:1.15, 1.69), p=0.0006], relapse <6 months from allografting [HR=0.50 (95%CI:0.41, 0.60), p<0.0001], presence of grade II-IV acute GVHD before relapse [HR=1.34 (95%CI:1.04, 1.73), p=0.03], and PBSC as cell source [HR=1.39 (95%CI:1.03, 1.86), p=0.03]

Notwithstanding poor outcomes whether DLI or a 2^nd-allo is offered, especially when relapse occurs within ≤ 6 months, a preliminary multivariate analyses shows that DLI use, younger age, MRD, lack of prior GVHD, and relapse > 6 months from allografting were associated with better OS. Further analysis on factors affecting physician’s choice of modality and prior cytoreductive therapy(ies) is warranted to shed light on the complex decision making process for post-allograft relapse management.

No relevant conflicts of interest to declare.