Transfusion of human volunteers with older, stored red blood cells produces extravascular hemolysis and circulating non-transferrin-bound iron

Abstract

Transfusions of RBCs stored for longer durations are associated with adverse effects in hospitalized patients. We prospectively studied 14 healthy human volunteers who donated standard leuko-reduced, double RBC units. One unit was autologously transfused "fresh" (3-7 days of storage), and the other "older" unit was transfused after 40 to 42 days of storage. Of the routine laboratory parameters measured at defined times surrounding transfusion, significant differences between fresh and older transfusions were only observed in iron parameters and markers of extravascular hemolysis. Compared with fresh RBCs, mean serum total bilirubin increased by 0.55 mg/dL at 4 hours after transfusion of older RBCs (P = .0003), without significant changes in haptoglobin or lactate dehydrogenase. In addition, only after the older transfusion, transferrin saturation increased progressively over 4 hours to a mean of 64%, and non-transferrin-bound iron appeared, reaching a mean of 3.2μM. The increased concentrations of non-transferrin-bound iron correlated with enhanced proliferation in vitro of a pathogenic strain of Escherichia coli (r = 0.94, P = .002). Therefore, circulating non-transferrin-bound iron derived from rapid clearance of transfused, older stored RBCs may enhance transfusion-related complications, such as infection.

Trial registration: ClinicalTrials.gov NCT01319552.

Figure 1

Study design and hemoglobin levels. (A) On day 1, each human volunteer donated 2 autologous RBC units by apheresis. One unit was transfused into the same participant after 3 to 7 days of storage (ie, “fresh”); the other unit was transfused after 40 to 42 days of storage (ie, “older”). One whole blood phlebotomy was performed 3 to 7 days before the older RBC transfusion to prevent post-transfusion erythrocytosis and to control for effects of recent blood loss on laboratory parameters. Blood samples were collected 90 minutes before transfusion and at 0, 1, 2, 4, 24, and 72 hours after transfusion. (B) Data are mean ± SEM for hemoglobin levels from before transfusion to 72 hours after transfusion of either fresh or older RBCs. The P value is as specified in the figure comparing the paired area under the curve of the mean hemoglobin levels for the N = 14 volunteers from 0 to 24 hours after the fresh and older RBC transfusions. (C) The individual hemoglobin levels for each subject up to 24 hours after transfusion. Vertical arrows indicate pretransfusion time points; and horizontal dashed lines, reference range values for men (blue) and women (pink).

Figure 2

Potassium levels do not change and calcium levels decrease after transfusions of older RBCs. The mean ± SEM for serum levels of (A) potassium, (B) total calcium, and (C) corrected calcium calculated as [(0.8 × (4.0 − subject's albumin)) + serum calcium]. The vertical arrow indicates the pretransfusion time point; and dotted lines, the reference ranges. The P values are as specified in the figure comparing the paired area under the curve of the mean of the outcome parameter for the N = 14 volunteers from 0 to 24 hours after the fresh and older RBC transfusions.

Figure 3

Transfusions of older RBCs result in laboratory values consistent with extravascular hemolysis in healthy volunteers. Data are mean ± SEM for serum levels of (A) total bilirubin and (B) conjugated bilirubin from before transfusion to 72 hours after transfusion of both fresh and older RBCs. (C) The individual serum total bilirubin levels for all 14 volunteers from before transfusion to 72 hours after transfusion of both fresh and older RBCs. (D) Data are mean ± SEM for lactate dehydrogenase (LDH) and haptoglobin, from before transfusion to 72 hours after transfusion of both fresh and older RBCs. Although iatrogenic hemolysis was induced during a difficult blood draw for 2 volunteers at 1 hour after the older RBC transfusion, these samples were still included in the analysis; nonetheless, the analysis was not significantly altered by their exclusion. Vertical arrows in all panels indicate the pretransfusion time point; and dotted lines, the reference ranges (and in gray for LDH). The P values are as specified in the figure comparing the paired area under the curve of the mean of the outcome parameter for the N = 14 volunteers from 0 to 24 hours after the fresh and older transfusions.

Figure 4

Iron parameters and circulating non–transferrin-bound iron levels increase after transfusions of older RBCs in healthy volunteers. (A) The mean ± SEM and (B) individual levels of serum iron; (C) mean ± SEM and (D) individual levels of transferrin saturation; (E) increase in ferritin compared with baseline levels; and (F) increase in plasma non–transferrin-bound iron compared with baseline levels from before transfusion to 72 hours after transfusion of fresh and older RBCs. Vertical arrows indicate pretransfusion time points; and dotted lines, the reference range (the reference range for change in ferritin and non–transferrin-bound iron is 0 by definition). The P values are as specified in the figure comparing the paired area under the curve of the mean of the outcome parameter for the N = 14 volunteers from 0 to 24 hours after the fresh and older RBC transfusions.

Figure 5

Serum levels of inflammatory markers do not increase after transfusions of older RBCs compared with fresh RBCs in healthy volunteers. (A) The mean ± SEM for serum IL-6 levels, (B) CRP levels, and (C) individual levels of CRP from before transfusion to 72 hours after transfusion of fresh and older RBCs. Vertical arrows indicate pretransfusion time points; and dotted lines, the reference range. The P values are as specified in the figure comparing the paired area under the curve of the mean of the outcome parameter for the N = 14 volunteers (for IL-6) and N = 12 (for CRP; the first 2 volunteers were not tested because of inadequate sample volume) from 0 to 24 hours after the fresh and older RBC transfusions.

Figure 6

Sera obtained after transfusions of older RBCs enhance proliferation of a bacterial pathogen in vitro. (A) Bacterial growth of E coli in serum samples obtained after fresh or older RBC transfusions was determined by serial optical density measurements at 600 nm for up to 5 hours after inoculation. Each point in the graph represents the mean ± SEM of the area under the curve (AUC) of the resultant bacterial growth curve (N = 14 paired values). (B) A Pearson correlation was used to determine the relationship between the mean difference in bacterial growth between fresh and older RBC transfusions at each time point and the corresponding differences in plasma non–transferrin-bound iron levels. The P values are as specified in the figure.