Optimal Serum Ferritin Levels for Iron Deficiency Anemia during Oral Iron Therapy (OIT) in Japanese Hemodialysis Patients with Minor Inflammation and Benefit of Intravenous Iron Therapy for OIT-Nonresponders

Abstract

Background: We determined optimal serum ferritin for oral iron therapy (OIT) in hemodialysis (HD) patients with iron deficiency anemia (IDA)/minor inflammation, and benefit of intravenous iron therapy (IIT) for OIT-nonresponders. Methods: Inclusion criteria were IDA (Hb <120 g/L, serum ferritin <227.4 pmol/L). Exclusion criteria were inflammation (C-reactive protein (CRP) ≥ 5 mg/L), bleeding, or cancer. IIT was withheld >3 months before the study. ΔHb ≥ 20 g/L above baseline or maintaining target Hb (tHB; 120-130 g/L) was considered responsive. Fifty-one patients received OIT (ferrous fumarate, 50 mg/day) for 3 months; this continued in OIT-responders but was switched to IIT (saccharated ferric oxide, 40 mg/week) in OIT-nonresponders for 4 months. All received continuous erythropoietin receptor activator (CERA). Hb, ferritin, hepcidin-25, and CERA dose were measured. Results: Demographics before OIT were similar between OIT-responders and OIT-nonresponders except low Hb and high triglycerides in OIT-nonresponders. Thirty-nine were OIT-responders with reduced CERA dose. Hb rose with a peak at 5 months. Ferritin and hepcidin-25 continuously increased. Hb positively correlated with ferritin in OIT-responders (r = 0.913, p = 0.03) till 5 months after OIT. The correlation equation estimated optimal ferritin of 30-40 ng/mL using tHb (120-130 g/L). Seven OIT-nonresponders were IIT-responders. Conclusions: Optimal serum ferritin for OIT is 67.4-89.9 pmol/L in HD patients with IDA/minor inflammation. IIT may be a second line of treatment for OIT-nonreponders.

Keywords: ferritin; hemodialysis; hepcidin-25; inflammation; iron deficiency anemia; oral iron therapy.

Figure 1

Protocol of iron therapy. Fifty-one consecutive hemodialysis (HD) patients with iron deficiency anemia (IDA) and minor inflammation were first treated with oral ferrous fumarate (50 mg/day). At 3 months after oral iron therapy (OIT), the patients were classified into two groups; OIT-responders and OIT-nonresponders. OIT was continued in 39 OIT-responders for another 4 months. OIT was switched to intravenous iron therapy (IIT; saccharated ferric oxide: 40 mg × 13 times for another 4 months) in the remaining 12 OIT-nonresponders. All patients simultaneously received a continuous erythropoietin receptor activator (CERA) during the study period.

Figure 2

Change of Hb levels in OIT-responders and OIT-nonresponders. The Hb levels were increased at 3 and 6 months after OIT in OIT-responders. In OIT-nonresponders, the Hb levels remained unchanged at 3 months after OIT but increased with IIT at the end of the study as a whole IIT-group. Data are expressed as mean ± standard error of the mean (SEM). Comparison of two nonparametric data groups was analyzed by the Mann-Whitney U test. * p < 0.01, vs. data at 0 month, ** p < 0.01, vs. data at 3 months, *** p < 0.05, vs. data at 3 months, # p < 0.05, vs. OIT-nonresponders.

Figure 3

Change of Hb levels in IIT group before and after IIT. Of the 12 OIT-nonresponders, the levels of Hb were significantly increased in seven (58.3%) of the 12 patients at the end of the study, whereas the Hb levels remained unchanged in the five IIT-nonresponders. Data are expressed as mean ± SEM. Comparison of two nonparametric data groups was analyzed by the Mann–Whitney U test. * p < 0.01, vs. IIT-nonresponders.

Figure 4

Sequential change in the levels of Hb, serum ferritin and hepcidin-25 in OIT-responders during OIT. The levels of Hb rose linearly with a peak at 5 months after OIT and then slightly decreased at the end of the study. Serum levels of ferritin were decreased from baseline at 1 month, and then rose continuously from 2 to 7 months after OIT. Serum levels of hepcidin-25 were similar to baseline at 6 months but significantly increased at the end of the study. Serum levels of hepcidin-25 were positively correlated with those of ferritin in the OIT-responders (r = 0.869, p = 0.0002). Data are expressed as mean ± SEM. Comparison of 2 means was determined by the Man–Whitney U test. # p < 0.01, vs. data at 0 month.

Figure 5

Correlation of Hb and serum ferritin between 1 and 5 months after OIT in OIT-responders. Ccorrelation between the levels of Hb and serum ferritin was analyzed in 39 OIT-responders using mean values for Hb and serum ferritin at every month obtained from OIT-responders as a whole. The levels of Hb were positively correlated with serum levels of ferritin till 5 months after OIT in 39 OIT-responders. The correlation equation calculated by Pearson’s correlation coefficient test was y = 0.0945x + 9.23, where y = Hb: x = serum ferritin.

Figure 6

Change in serum levels of ferritin at 0 and 6 months after initiation of the study in OIT-responders, IIT-responders and IIT-nonresponders. Serum levels of ferritin were significantly higher in the IIT-nonresponders than in the OIT-responders and the IIT-responders, whereas there was no difference in serum ferritin levels between the OIT-responders and the IIT-responders. Comparison of three nonparametric data groups was analyzed by the Tukey–Kramer test. * p < 0.05, vs. IIT-responders, # p < 0.01, vs. OIT-responders.

Figure 7

Correlation between serum levels of hepcidin-25 at the start of OIT and ΔHb at 3 and 6 months after OIT. Serum levels of hepcidin-25 at the start of OIT were negatively correlated with ΔHb at 3 months (r = −0.282, p < 0.05) (A) and at 6 months (r = −0.392, p < 0.01) (B) in 51 HD patients. The correlation equation was calculated by Pearson’s correlation coefficient test. White circle: OIT-responders, black circle: OIT-nonreponders.

Figure 7

Correlation between serum levels of hepcidin-25 at the start of OIT and ΔHb at 3 and 6 months after OIT. Serum levels of hepcidin-25 at the start of OIT were negatively correlated with ΔHb at 3 months (r = −0.282, p < 0.05) (A) and at 6 months (r = −0.392, p < 0.01) (B) in 51 HD patients. The correlation equation was calculated by Pearson’s correlation coefficient test. White circle: OIT-responders, black circle: OIT-nonreponders.

Figure 8

Serum levels of hepcidin-25 at the start of OIT between 21 OIT-responders with ΔHb ≥ 20 g/L and 18 OIT-responders achieving target Hb but ΔHb < 20 g/L at 3 months after OIT. Serum levels of hepcidin-25 at the start of OIT were significantly lower in the 21 OIT-responders with ΔHb ≥ 20 g/L than those in the 18 OIT responders who achieved target Hb (120–130 g/L) but ΔHb < 20 g/L. Comparison of two nonparametric data groups was analyzed by the Mann–Whitney U test. White circle: OIT-responders with ΔHb ≥ 20 g/L, black circle: the 18 OIT-responders with ΔHb < 20 g/L. * p < 0.05 vs. OIT-responders with ΔHb < 20 g/L.