Serum Ferritin and Cardiac/Liver Magnetic Resonance Imaging In Evaluating Iron Overload for Patients with Bone Marrow Failure Conditions Undergoing Non-Myeloablative HSCT

Abstract 1331

Several groups have identified iron overload, in terms of raised pre-transplant serum ferritin levels, as an independent adverse prognostic factor for patients undergoing myeloablative HSCT. While serum ferritin has been used as a common marker in clinical studies to evaluate the impact of iron overload following allogeneic transplantation, there are limitations to its use with it being an acute phase reactant, as well as its lack of specificity for predicting end-organ toxicities. Patients undergoing HSCT for bone marrow failure (BMF) syndromes usually have a significant red cell transfusion history, and although the majority of these patients receive non-myeloablative HSCT regimens, it is unclear as to the impact of iron overload in these patients on subsequent transplant outcomes.

In order to address these questions, we performed a prospective study evaluating the pre-transplant serum ferritin together with concurrent T2* cardiac magnetic resonance imaging (MRI) and R2 liver MRI in 18 patients with BMF syndromes undergoing allogeneic HSCT. The diagnosis of the patients included MDS (RCMD/hypoplastic MDS) =10, acquired aplastic anaemia =7, fanconi anaemia =1. The median age of the patients at transplantation was 42 years, and all patients received a T-cell depleted non-myeloablative HSCT. All patients were transfusion dependent pre-HSCT, with a median number of red cell transfusions of 45 (range: 8–115). Pre-HSCT ferritin was performed within 2 weeks of HSCT, and the results were correlated with albumin and C-reactive protein to reduce the impact of ferritin as an acute phase reactant. T2* and R2 MRI were similarly performed within 2 weeks of HSCT.

The median pre-HSCT ferritin was significantly raised at 2119 ug/l(range: 559–12235). In contrast, the T2* cardiac MRI was normal for all but one patient who had evidence of mild cardiac iron overload. All patients had a corresponding cardiac echocardiogram performed with an ejection fraction within normal limits. For the liver T2* MRI, 7 patients had evidence of none or mild hepatic iron overload, while 11 patients had moderate to severe iron overload. There was no correlation between pre-HSCT transfusion burden and serum ferritin levels. Furthermore, there was no correlation between either the transfusion burden or serum ferritin, and the T2*MRI readings.

In terms of HSCT outcome, the median time to neutrophil engraftment was 14 days. 2 patients had primary graft failure and only 1 patient died within 100 days due to an intra-cerebral haemorrhage. No patients had any clinical features of hepatic veno-occlusive disease (VOD), and 5 patients had evidence of grade I-II acute grade versus-host-disease. Data were also collected on the incidence of bacterial, fungal and viral infections post-HSCT for the cohort. There was however no significant association between transfusion burden, serum ferritin or T2* imaging and any of the HSCT outcomes (engraftment/day 100 TRM, GvHD, VOD or infections).

In the context of heavily transfused BMF patients receiving allogeneic HSCT, serum ferritin does not correlate with end-organ deposition of iron. Despite the high transfusion burden in our cohort of patients, cardiac deposition of iron appears minimal while hepatic iron deposition is significant in a large proportion of patients. Reassuringly, a raised iron overload by either of the above mentioned parameters had no effect on HSCT outcomes. Our findings highlight the limitations of using serum ferritin as a marker of iron overload pre-HSCT. The role of active pre-HSCT chelation of BMF patients receiving non-myeloablative HSCT regimens remains unclear, and further studies are warranted.