CK and PK contributed equally to this work. Angiopoietin-2 (Ang-2), a member of the Angiopoietin/Tie ligand-receptor system, represents one of the specific inducers required for angiogenesis. Recent data indicate that bone marrow neoangiogenesis plays an important pathogenic and prognostic role in malignancies of the hematopoietic system. We have recently shown that circulating Ang-2 is a predictor for relapse in the setting of allogeneic hematopoietic stem cell transplantation (HSCT) in a cohort of high-risk (HR) myeloid malignancies (Kümpers, Koenecke et al.: BLOOD May 2008). The aim of this study was to validate our results in a larger cohort of patients. In contrast to our previous work we included different conditioning regimens as well as patients with Acute Myeloid Leukemia (AML) irrespective of their risk stratification according to cytogenetic (SWOG, ECOG criteria) and/ or mutational status (FLT3, NPM1). Circulating Ang-2 (in-house ELISA) was measured in serum of AML patients treated at our institution between 1991 and 2008 (n=191) immediately before standard (n=83) or reduced intensity conditioning (n= 104) for HSCT. Stem cell sources were BM (n=23), PBSC (n=163) or both (n=5) from MRD (n= 91), MMRD (n=7), MUD (n=67) and MMUD (n=26). The median age of the patients was 49 years (range 18–71). HSCT for AML (n=149) or secondary AML after MDS (n=42) was performed after primary induction failure (n=30) or for relapse (first relapse n=19, subsequent relapse n=5). 119 patients were transplanted in 1st CR; four in 2nd CR. 11 patients had no chemotherapy before HSCT. In this cohort, 114 of 159 evaluable patients suffered from HR-AML. Circulating pre-HSCT Ang-2 was elevated in patients (median (range): 1.7 (0.05–48.8) ng/ml) compared to healthy controls (0.92 (0.27–2.63) ng/ml; p<0.0001). Using univariate Cox regression analyses, patients’ age, conditioning regimen, remission status, cytogenetic/ mutational status, and Ang-2 were identified as prognostic variables. Multivariate Cox regression confirmed the independent prognostic impact of Ang-2 (hazard ratio 2.2, 95% CI 1.25–3.94, p = 0.007) and HR cytogenetics and/ or mutations (hazard ratio 1.5, 95% CI 1.09–2.11, p = 0.012) for Time to Relapse after HSCT. In this study we could show that Ang-2 is a valid and readily available biomarker to pre-estimate the risk of relapse in a large comprehensive HSCT cohort. Routine pre-HSCT measurement of Ang-2 may open new perspectives for riskadapted treatment of AML patients.

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author