Background and Objectives: Sickle cell anemia (SCA) is characterized by chronic haemolysis and frequent painful vaso-occlusive episodes, eventually leading to multiple organ damage, including the kidneys. Chronic haemolysis is frequent in SCA patients receiving regular transfusions. Not all renal disease in patients with SCA is due to sickle cell nephropathy (SCN), but the incidence of renal failure increases as patient survival improves. Since Magnetic Resonance Image can provide a unique evaluation of the abdomen in a single exam, the aim of this study was to assess renal iron deposits in patients with SCA, correlating these values with transfusion burden, liver and heart T2* and blood LDH levels.

Material and methods: Twenty patients with SCA older than 20 years of age and transfusion burden were studied in a 1.5T scanner (Avanto, Siemens Medical Systems, Erlangen, Germany) using T2* technique and compared to age- and sex-matched normal controls. For liver T2* analysis, regions of interest (ROI) were manually drawn in the liver, using Viewing software (Leonardo - Siemens) avoiding all major visible vessels. For heart T2* analysis, a full-thickness ROI was manually defined in the interventricular septum (routinely chosen to avoid T2* artifacts from the cardiac veins, liver, and lungs). For renal T2* analysis, ROI were manually drawn in the cortex, using Viewing software (Leonardo - Siemens) avoiding the medullary region. All parameters are presented as mean ± standard deviation. Spearman's rank test was used to assess the correlation between renal T2* and blood LDH levels. The Ethics Committee for Research approved this trial.

Results: Mean renal T2* was significantly lower (p < 0.001) in the SCA patients [16.4 ± 11.8 ms] compared with healthy controls (47.9 ± 11.1 ms). However there was no correlation between renal T2* and heart (p = 0.500) and liver T2* (p= 0.578). Median liver and myocardial T2* in this population was 3.96 ± 3.58 ms and 36.87 ± 8.06 ms. Cortico-medullary signal differences in renal iron loading were observed. Low signal intensity in the cortex relative to the medulla was observed in thirteen patients (65%). SCA patients who showed no cortico-medullary difference in signal intensity had a mean T2* of 30.9 ± 6.5 ms, which was significantly lower (p < 0.001) than that of the group of patients showing a hypointense cortical signal (8.6 ± 2.8 ms). We investigated the relationship between cortico-medullary differences and haemolysis using LDH values as a marker of haemolysis. Renal T2* showed a significant correlation with LDH values. Mean LDH values were significantly different between patients with and without cortico-medullary differences (p = 0.004), being 470.6 ± 123.1 IU/L for those with no cortico-medullary differences, and 1313.0 ± 788.6 IU/L for those with hypointense cortex. The seven highest LDH values in this cohort of 20 patients corresponded to seven patients with cortico-medullary signal intensity differences. Conclusion: Our study in SCA shows that iron deposition in the kidney is not related to systemic iron overload, but to haemolysis. There appears to be a clear localization of iron deposition within the kidney, with the deposits being found in the cortex and not in the medulla, as found by T1 and T2-weighted MRI. Our results suggest that haemolysis is a major determinant of renal iron load in SCA patiests.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.