The Ratio of Pre-Transplant Serum Ferritin to the Units of Blood Transfused and Not Serum Hepcidin Strongly Correlates with Outcome After Allogeneic Hematopoietic Cell Transplantation In Patients with AML and MDS.

Abstract 1307

Iron homeostasis is controlled through a coordinated activity of multiple proteins through different signaling pathways with hepcidin being the central regulatory hormone. Iron overload in recipients of allogeneic hematopoietic cell transplantation (HCT) is associated with an adverse outcome. As causes of disturbed iron homeostatsis are multifactorial with transfusional iron playing a key role, we analysed prospectively the outcome after HCT with pre-transplant patterns of potential interactions between the units of red blood cells (URB) transfused, serum hepcidin (SH), and ferritin (SF). For this purpose, 3 quotients were generated (SF/URB, SF/SH, SH/URB). All parameters were measured 7 days pre-HCT with a median C-reactive protein of < 5 mg/l. From February, 2008–January, 2010, 111 consecutive patients (pts) [58 males/53 females, median age 57 (range 19–75) years] with AML in remission (n=85) and MDS (n=26) who underwent HCT at the University of Leipzig were studied. Donors were matched related (MRD) in 23 (21%) and matched unrelated (MUD) in 88 (79%) pts. Preparative regimen was conventional conditioning (CC) with 12 Gy TBI/cyclophosphamide 120 mg/kg in 41 (40%) and RIC with fludarabin 30 mg/m²/day for 3 days and2 Gy TBI in 70 (60%) pts. Median SH measured by C-ELISA was 282 (range 6–1595) ng/ml. Median SF was 1869 (range 36–17995) ng/ml with SF >1000 ng/ml present in 91 (82%) pts. Median URB was 24 (0-95) units. There was a poor linear correlation of SH with SF (r=0.58) and URB (r=0.35). After a median follow up of 16 (5-28) months, OS, EFS, TRM, and RI were 66%, 53%, 17%, and 30% respectively. In multivariate analysis, SH, SF, and URB had no impact on OS, EFS, RI or TRM. The incidences of acute (a) GvHD > grade II was 48%. It was associated with MUD (p=0.004) and the presence of an HLA-mismatch (p=0.02). Chronic (c) GvHD occurred in 46% of pts with 33/104 (32%) pts suffering from extensive cGvHD. RIC (p=0.001), MUD (p=0.06), and a higher SF (p=0.02) correlated with an increased incidence of cGvHD. SH, and URB had no influence on aGvHD or cGvHD. A higher SF (p=0.02) and CC (p=0.001) correlated with the 93 viral, bacterial and fungal infections > grade 3 (according to common toxicity criteria) in the first 100 days post-HCT. SF/SH, and SH/URB had no impact on outcome. On the other hand, SF/URB < 75% correlated strongly with an improved OS (p=0.06), and EFS (p<0.0005) as well as a lower TRM (p=0.02), and cGvHD (p=0.06) in both uni- and multivariate analysis. EFS in pts with SF/URB < 75% was 60% versus 33% for those with a higher SF/URB (p=0.01) mainly because of a significantly lower TRM of 11% for SF/URB < 75% versus 33% for SF/URB > 75% (p=0.002). In the absence of relapse, the probability of one-year survival in a patient with SF/URB < 75% is 86% compared to 50% for a patient with SF/URB > 75% (p=0.001). Median pre-transplant SF of 1844 (range 567–6842) ng/ml was significantly lower (p=0.003) and median URB of 26 (8-95) units higher (p=0.005) in pts with SF/URB < 75% compared to pts with higher SF/URB [median SF 2573 (range 463–17995) ng/ml, median URB 19 (range 2–63) units]. SH in both SF/URB groups was similar (p=0.7). SF/URB was not influenced by age, gender, diagnosis, type of conditioning and donor, number of previous therapies, CMV risk status, and female to male HCT. This simple ratio, if validated in larger studies, might provide clinicians prior to HCT with a reliable tool for assessment of TRM after HCT for AML and MDS. Unlike ferritin alone which is a sequel of intracellular iron metabolism, this ratio reflects some of the often simultaneously disturbed iron sensor pathways for hepcidin regulation such as erythropoietic activity, inflammation and iron stores in pts with haematological disorders undergoing HCT.