Trends in Ferritin Can Be Dramatically Different From Trends in Total Body Iron and Could Lead to Erroneous Decisions in Iron Chelation Management and Discourage Adherence in Chronically Transfused Patients,

Abstract 3203

Ferritin trends are used as surrogates for change in total body iron in patients with transfusional iron overload who are on chelation therapy. They are often used to infer patient adherence with prescribed therapy and for recommending changes. Population studies of ferritin show a 70% correlation with liver iron. The aim of this study was to determine whether the trends in ferritin adequately reflect the change in liver iron concentration (LIC) in individual patients. We retrospectively evaluated ferritin and LIC for 10 years in 40 patients with transfusion dependent anemia (23 with transfusion dependent thalassemia, 12 with sickle cell anemia, 2 with congenital dyserythropoetic anemia, 2 with Diamond Blackfan Anemia and one with sideroblastic anemia). Ferritin levels are evaluated every three weeks at each transfusion and liver iron concentration (LIC) by MRI at approximately annual intervals. The LIC values in mg/g dry weight (dw) are derived by MRI. The trends for both LIC and ferritin were evaluated at each period between the sequential MRIs. We used the average of all ferritins in a four month window centered on the date of the MRI for comparison to the LIC. The overall correlation between ferritin and LIC was similar to other published results (r²=0.69).

When ferritin and LIC were plotted against time for each patient, the ferritin trend clearly predicted the LIC trend during certain periods of time (Example figure 1 segment A) and did not during other periods (Figure 1 segment B).

View largeDownload slide

View largeDownload slide

Close modal

The trend in ferritin correctly predicted the trend in LIC all of the time in 55% of patients (22/40). In 45 % of the patients (18/40) the ferritin trend did not correlate with the LIC in over half of the observational periods. In 37.5 % (15/40) of patients during at least one observation period the direction of change was dramatically different. Of these, the direction of change was opposite in 12.5% (5/40). In 22.5 % (9/40) the changes were disproportionate. Six of these patients showed a period during which there was a slight decrease in ferritin but a significant decrease in LIC. In two there was a significant increase in LIC with only a minimal rise in ferritin. In one, with a significant increase in ferritin the LIC increased minimally.

While the ferritin was decreasing the LIC and ferritin trends correlated much better than when the ferritin was increasing. This implies that when ferritin levels increase it is a particularly poor tool for assessing change in iron overload.

It is clear from this analysis that over certain periods of time, even up to four years, the trends in ferritin can be opposite in direction to the change in total body iron, as derived from LIC. This could lead to inappropriate changes in therapy and incorrect assumptions by health care providers about patient adherence. It is accepted that poor compliance with chelation therapy is the greatest barrier to effective management of iron overload. If only ferritin is used to assess changes in total body iron, patients could be discouraged by their apparent poor response to therapy even though their LIC may actually be decreasing. Serial assessment of total body iron burden by direct measurement of LIC is essential for proper management of patients with transfusional iron overload.