From Renal Siderosis Due to Perpetual Hemosiderinuria to Possible Liver Overload Due to Extravascular Hemolysis: Changes in Iron Metabolism in Paroxysmal Nocturnal Hemoglobinuria (PNH) Patients On Eculizumab.

Abstract 4031

Poster Board III-967

Iron metabolism in PNH patients is dominated by perpetual iron loss consequent to the chronic complement-mediated intravascular hemolysis; thus, they are prone to develop iron deficiency rather than iron overload, even in presence of large transfusional requirement. Eculizumab (Ecu) has proven effective for the treatment of intravascular hemolysis in PNH patients, resulting in reduction and even abolishment of transfusion requirement and improvement of signs and symptoms of intravascular hemolysis; however, Hb gain is heterogeneous among patients, in most cases due to residual C3-mediated extravascular hemolysis hampering Hb normalization. The goal of our study was to identify possible modifications in iron compartmentalization associated with Ecu treatment and possible clinical consequences. We evaluated iron metabolism in 5 untreated PNH patients and 23 who were receiving Ecu (of whom 4 have been also studied before treatment), combining biochemical parameters with a semiquantitative T2* MRI technology. MRI was performed using four gradient-echo sequences and one spin-echo sequence; signal intensity (SI) was measured on images obtained with each sequence by means of three regions of interest placed in the renal cortex, liver, spleen and at the level of the para-spinal muscle, resulting in a semiquantitative SI value (Grandon et al., Radiology 1994). Within the total patient cohort (regardless they were or were not on Ecu), there was a significant correlation between liver SI and serum ferritin (P<0.001), while kidney SI correlated with the presence of hemosiderinuria (HS, P<0.001). All untreated PNH patients showed similar MRI findings, with significant renal cortex siderosis and normal SI in liver and spleen. This was consistent with overt intravascular hemolysis, as confirmed by biochemical routine testing, and consequent perpetual hemosiderinuria; as expected, all these patients had abundant HS. In contrast, the 23 PNH patients on Ecu showed a distinct and heterogeneous pattern. All patients showed a normal renal SI, with the exception of 2 cases who have recently started Ecu and 2 experiencing Ecu breakthrough; these 4 patients had normal hepatic and splenic SI. All of them (but none of those with normal renal SI) had persistent HS, while only the latter 2 had increased LDH; we conclude that these 4 patients have had residual intravascular hemolysis, and that HS was more sensitive than LDH to identify recent history of intravascular hemolysis. In contrast, the majority of patients showed increased hepatic SI: we found 6 cases with moderate and 5 cases with severe iron overload; in some patients, high hepatic SI was associated with increased SI in the spleen. The 4 patients evaluated before and during treatment showed pre-treatment renal siderosis which progressively disappeared after months of Ecu therapy; in 2 of them, who had a longer exposition to Ecu, moderate hepatic iron overload was demonstrated. Hepatic SI significantly correlated with serum ferritin (P<0.05), but not with transferrin saturation nor with LDH. Iron overload was predictable as a result of persistent transfusional need only in two patients with partial response to Ecu; however, within the whole cohort, patients with suboptimal hematological response (i.e., those with persistent Hb<11) were more likely to develop severe hepatic iron overload (P=0.02). Thus, we hypothesized that iron overload in these patients may be pathophysiilogically linked to persistent extravascular hemolysis; we found a direct correlation between liver SI and both % of C3+ PNH RBCs (P=0.02) and absolute reticulocyte count (P=0.02), which were considered markers of extravascular hemolysis (Risitano et al, Blood 2009). In conclusion, we show by T2* RMI that untreated PNH patients have significant renal siderosis, which tends to disappear during Ecu treatment as a result of the blockade of intravascular hemolysis. However, such blockade of urinary iron loss may render PNH patients susceptible to liver iron overload resulting from transfusions, as well as from residual extravascular hemolysis. While is still not clear the proportion of patients developing clinically significant iron overload requiring specific treatment, we provide evidence that iron metabolism substantially changes during eculizumab treatment, and C3-mediated extravascular hemolysis may play a major role in this process.