In patients with iron overload (e.g., hemochromatosis, thalassemia, DBA), elevated pancreatic iron levels and
fat infiltration are frequently observed. In transfusion dependent thalassemia (TDT) patients, diabetes mellitus
(DM) develops after the age of 20 y with a prevalence of 6% to 50% depending on age, ethnicity and chelation
treatment regimen.
Objective: To quantitatively explore the relationship between pancreatic iron or fat content and parameters of
glucose metabolism in patients undergoing regular iron monitoring of liver, heart, and pancreas by
biomagnetometry and MRI.
Materials and Methods: Pancreatic iron (R2*) and fat content (FC) were measured by standard gradient echo
sequences and analyzed by chemical-shift relaxometry (= echo time dependent signal magnitude fit). Timely
correlated glucose and insulin kinetics could be studied by oral glucose tolerance tests (oGTT) in 19 patient
measurements. The following parameters were determined: fasting glucose level (G-0), HOMA index, and the
model derived oral glucose insulin sensitivity (OGIS) index (Mari et al, 2001).
Results: Mean pancreatic R2* rate and FC were obtained as 240±195s -1 (in vivo iron concentrations of about
1000±800μg/g) and 28±18%, respectively. The highest R2* rate (660s -1 ) was found in a patient with IGT/DM.
Pancreatic R2* significantly correlated with G-0 or OGIS (p<0.01) but FC did not correlate. Bivariate prediction
of OGIS was found to be significant (r 2 =0.61) for both R2* (p=0.002) and FC (p=0.02).
Conclusion: Pancreatic iron and fat content does not only indicate end-stage impaired glucose tolerance but
may also predict the risk of developing diabetes in patients with iron overload early on.