Secondary Logo

Share this article on:

A Multicenter, Randomized Clinical Trial of IV Iron Supplementation for Anemia of Traumatic Critical Illness*

Pieracci, Fredric M. MD, MPH; Stovall, Robert T. MD; Jaouen, Brant BA; Rodil, Maria BA; Cappa, Anthony MD; Burlew, Clay Cothren MD; Holena, Daniel N. MD; Maier, Ronald MD; Berry, Stepheny MD; Jurkovich, Jerry MD; Moore, Ernest E. MD

doi: 10.1097/CCM.0000000000000408
Clinical Investigations

Objective: To evaluate the efficacy of IV iron supplementation of anemic, critically ill trauma patients.

Design: Multicenter, randomized, single-blind, placebo-controlled trial.

Setting: Four trauma ICUs.

Patients: Anemic (hemoglobin < 12 g/dL) trauma patients enrolled within 72 hours of ICU admission and with an expected ICU length of stay of more than or equal to 5 days.

Interventions: Randomization to iron sucrose 100 mg IV or placebo thrice weekly for up to 2 weeks.

Measurements and Main Results: A total of 150 patients were enrolled. Baseline iron markers were consistent with functional iron deficiency: 134 patients (89.3%) were hypoferremic, 51 (34.0%) were hyperferritinemic, and 64 (42.7%) demonstrated iron-deficient erythropoiesis as evidenced by an elevated erythrocyte zinc protoporphyrin concentration. The median baseline transferrin saturation was 8% (range, 2–58%). In the subgroup of patients who received all six doses of study drug (n = 57), the serum ferritin concentration increased significantly for the iron as compared with placebo group on both day 7 (808.0 ng/mL vs 457.0 ng/mL, respectively, p < 0.01) and day 14 (1,046.0 ng/mL vs 551.5 ng/mL, respectively, p < 0.01). There was no significant difference between groups in transferrin saturation, erythrocyte zinc protoporphyrin concentration, hemoglobin concentration, or packed RBC transfusion requirement. There was no significant difference between groups in the risk of infection, length of stay, or mortality.

Conclusions: Iron supplementation increased the serum ferritin concentration significantly, but it had no discernible effect on transferrin saturation, iron-deficient erythropoiesis, hemoglobin concentration, or packed RBC transfusion requirement. Based on these data, routine IV iron supplementation of anemic, critically ill trauma patients cannot be recommended (NCT 01180894).

1Department of Surgery, Denver Health Medical Center/University of Colorado School of Medicine, Denver, CO.

2Department of Surgery, University of Pennsylvania School of Medicine, Philadelphia, PA.

3Department of Surgery, Harborview Medical Center/University of Washington School of Medicine, Seattle, WA.

4Department of Surgery, University of Kansas Medical Center, Kansas City, KS.

* See also p. 2140.

Supported, in part, by grant NTI-ICU-008-01 from the National Trauma Institute and by Prime Award 08302001 from the Department of the Army.

Dr. Pieracci received support for article research from the National Trauma Institute (NTI)/Department of Defense. His institution received grant support from NTI (funding source). Dr. Stovall, Mr. Jaouen, Ms. Rodil, Dr. Cappa, Dr. Burlew, Dr. Holena, Dr. Maier, Dr. Berry, Dr. Jurkovich, and Dr. Moore received support for article research from the NTI. The U.S. Army Medical Research Acquisition Activity, 820 Chandler Street, Fort Detrick, MD 21702–5014, is the awarding and administering acquisition office. The opinions or assertions contained herein are the private views of the authors and are not to be construed as official or as reflecting the views of the Department of the Army of the Department of Defense.

For information regarding this article, E-mail: fredric.pieracci@dhha.org

Anemia is nearly universal among critically ill patients and associated with poor outcomes (1). The etiology of ICU anemia is multifactorial, including acute and chronic blood loss, hemodilution, and inflammation. Inflammation induces cytokine-mediated alterations in erythrocyte longevity (2), erythropoietin synthesis and sensitivity (3), and iron metabolism (4). Specifically, iron is diverted from bone marrow sites of erythropoiesis into ferritin within the reticuloendothelial system (RES). This alteration in iron trafficking appears to be mediated at least in part by up-regulation of the iron-regulatory protein hepcidin, which inhibits ferroportin-mediated release of iron from cells (5).

Our group previously reported that the vast majority of critically ill surgical patients become both hypoferremic and hyperferritinemic within 48 hours of ICU admission (6). Furthermore, over one half of these patients demonstrate evidence of iron-deficient erythropoiesis (IDE), as indicated by an increase in erythrocyte zinc protoporphyrin (eZPP) concentration (7). We subsequently confirmed these findings among a separate subgroup of critically ill trauma patients (8).

Iron-prescribing habits are inconsistent among intensivists (9, 10), and few investigations have addressed the utility of iron supplementation of critically ill patients (11, 12). It is unclear if iron supplementation is sufficient to overcome the cytokine-mediated, functional iron debt observed in critically ill patients and thereby deliver iron to the bone marrow in quantities sufficient to impact erythropoiesis meaningfully. The relationship between iron supplementation and risk of infection also remains unclear (13).

In a single-center randomized controlled trial (RCT), we observed a trend toward decreased packed RBC (pRBCs) transfusion requirement for patients who received enteral iron supplementation (ferrous sulfate 325 mg thrice daily) as compared with placebo (14). However, many critically ill patients are unable to tolerate enteral medications, and the inflammatory response associated with critical illness impairs enteral absorption of iron (15). The hypothesis of the current multicenter RCT was that IV iron supplementation 1) improves functional iron deficiency, 2) improves IDE, 3) decreases the pRBCs transfusion requirement, and 4) does not increase the risk of infection.

Back to Top | Article Outline

MATERIALS AND METHODS

This was a multicenter, randomized, single-blind, placebo-controlled trial involving four state-verified, American College of Surgeons–certified, level I trauma centers: Denver Health Medical Center, Denver, CO (lead study center); Harborview Medical Center, Seattle, WA; University of Pennsylvania Hospital, Philadelphia, PA; and University of Kansas Medical Center, Kansas City, KS. Enrollment occurred from May 2011 through September 2013. The trial was funded by an investigator-initiated grant from the National Trauma Institute (NTI-ICU-008-01). Institutional review board (IRB) approval was obtained separately by the U.S. Army Medical Research and Materiel Command, Office of Research Protections, Human Research Protection Office (proposal log number 08302001), the lead study center, and each satellite center. The trial was conducted in accordance with the recommendations of the Consolidated Standards of Reporting Trials Statement (16) and registered with the U.S. National Institutes of Health (ClinicalTrials.gov Identifier NCT 01180894).

Eligible patients included those admitted to the ICU with a primary diagnosis of trauma. The inclusion criteria were 1) anemia (latest hemoglobin concentration < 12 g/dL); 2) age 18 years old or older); 3) less than or equal to 72 hours from ICU admission; and 4) expected ICU length of stay (LOS) more than or equal to 5 days. The hemoglobin concentration was chosen based on the lead study center’s laboratory definition of anemia. The interval between admission and enrollment was selected to minimize the influence of interventions (e.g., pRBCs transfusion) prior to study enrollment and maximize exposure to the study drug. The expected ICU LOS was selected to increase the injury severity of the sample, prolong the follow-up period, and allow time for any treatment effect to accrue. This time period is consistent with previously validated definitions of prolonged critical illness (17–19). Assessment of the expected ICU LOS was determined by the trauma attending intensivist upon ICU admission.

The exclusion criteria were 1) active hemorrhage requiring pRBCs transfusion; 2) iron overload (serum ferritin concentration ≥ 1,000 ng/mL, serum iron concentration > 160 μg/dL, or serum transferrin saturation ≥ 50%) or any condition associated with iron overload (e.g., hemochromatosis and aceruloplasminemia); 3) active infection, defined using Centers for Disease Control and Prevention criteria (20); 4) chronic inflammatory conditions (e.g., systemic lupus erythematosus, rheumatoid arthritis, and ankylosing spondylitis), 5) preexisting hematologic disorders (e.g., thalassemia, sickle cell disease, hemophilia, von Willibrand’s disease, or myeloproliferative disease); 6) macrocytic anemia (admission mean corpuscular volume [MCV] ≥ 100 fL); 7) current or recent (within 30 d) use of immunosuppressive agents; 8) use of any recombinant human erythropoietin formulation within the previous 30 days; 9) pregnancy or lactation; 10) legal arrest or incarceration; 11) prohibition of pRBCs transfusion; 12) stay of more than or equal to 48-hour duration in the ICU of a transferring hospital; 13) history of intolerance or hypersensitivity to iron; and 14) moribund state in which death was imminent.

Patients were screened and enrolled Monday–Friday. Following informed consent from the patient or legally authorized representative, subjects were randomized to receive either iron sucrose (Venofer; Luitpold Pharmaceuticals, Norristown, PA) 100 mg IV or placebo, thrice weekly (Monday, Wednesday, and Friday if enrolled on Monday, Wednesday, or Friday; Tuesday, Thursday, and Saturday if enrolled on Tuesday or Thursday) for up to six doses or until ICU discharge, whichever occurred first. Due to the deep red color of iron sucrose, the unavailability of an IV placebo of similar color, and the logistical difficultly in shrouding the bags and tubing, the placebo consisted of an equal volume (100 mL) of normal saline. Thus, study group assignment was unblinded to both subjects and healthcare providers who administered the study drug and blinded to the research team abstracting and analyzing the data. Randomization was accomplished by the investigational pharmacy at each satellite site using a computer-generated block pattern. The randomization was unblinded to the investigators only after completion of data accrual.

Selection of iron formulation was based on safety and efficacy information from the outpatient setting (21–23), as well as the limited data from critically ill patients (11, 24). Dosing of iron sucrose ranged widely in studies of critically ill patients, from 140 (11) to 300 mg (24) weekly. Approximately 25 mg/d of iron is necessary to support erythropoiesis during health (25), but this number may increase markedly during critical illness. In the study by van Iperen et al (11), IDE persisted despite IV iron supplementation with 20-mg iron sucrose daily for 14 days among critically ill patients treated with recombinant human erythropoietin, suggesting that the dose of iron may have been too low. Conversely, increasing the weekly dose of iron sucrose above 300 mg raised concerns for both iron overload and infection. Based on these considerations, a weekly dose of 300 mg was selected. An Investigational New Drug Application (IND) was obtained from the Food and Drug Administration (FDA), Department of Health and Human Services for use of iron sucrose for the non-FDA-approved indication of ICU anemia (FDA IND 109877).

Sample size calculations were performed using SAS version 9.1 (SAS, Carey, NC; proc power), using α = 0.05 and β = 0.20. Functional iron deficiency was defined as a low serum transferrin saturation in the setting of either normal or elevated serum ferritin concentration. Prior analysis of a sample of critically ill surgical patients revealed a baseline transferrin saturation of 10% with a SD of 8.5 (8). An increase in the transferrin saturation to 20% among the iron group was considered meaningful clinically. This returned a total sample size of 26. The mean eZPP concentration (SD) among the placebo group in our RCT of enteral iron supplementation was 78.3 micromol:mol heme (12.9) (14). An approximate 10-point decrease would be required to normalize the eZPP (normal range, 0–69 micromol:mol heme), thus correcting IDE. This returned a total sample size of 56 subjects. The prevalence of RBC transfusion among critically ill patients with an ICU length of stay more than or equal to 7 days is approximately 70% (1). A 10% reduction in the risk of transfusion from 70% (p 1) to 60% (p 2) was considered meaningful clinically. This returned a total sample size of 172 subjects. The prevalence of any infection among critically ill patients with an ICU length of stay more than or equal to 5 days is approximately 50% (14). A 10% increase in the risk of infection from 50% (p 1) to 60% (p 2) was considered meaningful clinically. This returned a total sample size of 204 subjects. Based on the aforementioned calculations, we targeted a total sample size of 200 subjects. We also planned a subgroup analysis of iron markers at baseline, day 7, and day 14 among subjects with an ICU LOS of more than or equal to 14 days (who would have received all six possible study drug doses), estimated to involve 50 subjects total (25%).

Baseline demographics included enrollment site, age (yr), gender, mechanism of injury (motor vehicle crash, pedestrian struck, fall, assault, penetrating, and other), comorbidity score, admission Acute Physiology and Chronic Health Evaluation (APACHE) II score, admission Injury Severity Score (ISS), time from ICU admission to enrollment (d), baseline estimated blood loss (EBL), and the occurrence and amount of previous pRBCs transfusions. EBL was calculated by summing blood losses from indwelling drains, invasive procedures, and tubes collected from phlebotomy. The comorbidity score was adapted from Charlson et al (26).

After enrollment, subjects were followed for 42 days or until hospital discharge, whichever occurred first. Outcome variables were 1) number of total doses of study drug received; 2) hemoglobin concentration (measured at baseline and daily thereafter); 3) hematologic and iron markers (measured at baseline and weekly thereafter); 4) pRBCs transfusions; 5) infections and infection type; 6) antibiotic exposure; and 7) mortality.

Additional hematologic markers included the MCV (Siemens Advia 2120, 80–100 fL; Siemens Healthcare Diagnostics, Inc, Tarrytown, NY) and the percentage reticulocytes (Siemens Advia 2120, 0.5–1.5%). Iron markers included the serum iron concentration (Siemens Vista Dimension 1500, 40–150 ug/dL), serum ferritin concentration (Siemens Vista Dimension 1500, 28–365 ng/mL), serum transferrin saturation (Siemens Vista Dimension 1500, 18–55%), and eZPP concentration (ARUP Laboratories, Salt Lake City, UT, 0–69 mmol:mol Heme). The serum transferrin saturation was used as an estimate of the supply of iron readily available for deposition in the bone marrow for erythropoiesis (27). During normal erythropoiesis, iron is chelated to protoporphyrin IX to form heme. When inadequate iron is delivered to the bone marrow, zinc is substituted for iron, forming zinc protoporphyrin. An elevated eZPP is diagnostic of IDE and reflects the bone marrow iron supply regardless of total body iron (7, 28). Study drug was discontinued after three doses if there was evidence of iron overload on the first weekly (study day 7) labs: Serum ferritin concentration more than or equal to 1,000 ng/mL, serum iron concentration more than 160 μg/dL, or serum transferrin saturation more than or equal to 50%. Laboratory markers were drawn at 4 AM and study drug was administered at 9 AM.

The decision to transfuse pRBCs was made by the attending intensivist. ICU protocols at each institution specified the hemoglobin transfusion threshold as 7.0 g/dL in the absence of shock or acute coronary syndrome (29). Nosocomial infections were defined according to the U.S. Centers for Disease Control and Prevention guidelines (20), with the exception of ventilator-associated pneumonia (VAP), which was defined as clinical suspicion for pneumonia along with a lower respiratory tract culture with more than or equal to 105 colony forming units/mL (30). Specific infections analyzed included VAP, bacteremia, and urinary tract infection (UTI). Antibiotic days were defined as the product of the days of antibiotics and the number of independent antibiotics received.

No interim analysis was planned. Unanticipated problems and related unexpected adverse events were reported to both the lead study center IRB and the U.S. Army Medical Research and Material Command, Office of Research Protections, Human Research Protection Office within 5 days. The safety of the study was also monitored by an independent, external Data Safety Monitoring Board, which consisted of a medical intensivist, a surgical intensivist, and a biostatistician. The committee met biannually and submitted written reports to the IRBs, FDA, and funding agency.

All statistical analyses were performed using SAS Version 9.1 (SAS). Data are expressed as median (range) or number (%). Differences in the medians of continuous variables were assessed using the Wilcoxon rank test. Proportions of categorical variables were compared using the chi-square test. When expected cell counts were less than 5, Fisher exact test was used. An a priori analysis of iron markers at baseline, day 7, and day 14 was planned, thus no post hoc adjustment for multiple comparisons was applied. Linear regression (sum of squares) was used to control for the effect of potential confounders on the continuous outcomes of iron markers. Normal distribution of iron markers was confirmed using the Kolmogorov–Smirnov test. Overall model significance was calculated using the F statistic, with the significance of individual variables calculated using the t statistic. Logistic regression was used to control for the effect of potential confounders on binary outcomes. The overall contribution of the fitted model to predicting variability in the outcome of interest was assessed using the likelihood ratio chi-square test. The independent contribution of individual variables was assessed using the Wald chi-square test and expressed as the odds ratio (OR) with 95% CIs. Model fit was assessed using the Hosmer-Lemeshow goodness-of-fit chi-square statistic, with p value of more than 0.05 indicating acceptable model calibration (31). The α error level was defined at 0.05, with statistical significance set at p value of less than 0.05.

Back to Top | Article Outline

RESULTS

The derivation of the final sample size is shown in Figure 1. Of the 1,298 patients screened, 150 (11.6%) were randomized. The most common reason for study exclusion was expected ICU LOS less than 5 days (n = 907, 69.9%). Of the 210 patients who were approached, 151 agreed to participation (71.9%). Of the patients who were enrolled, one subject was immediately withdrawn from the study prior to randomization after it was realized that he was already enrolled in another interventional study. Demographics of patients who agreed to versus declined study participation were similar with respect to enrollment site (p = 0.27), gender (p = 0.96), mechanism of injury (p = 0.17), APACHE II score (p = 0.70), and ISS (p = 0.87). Enrolled patients were significantly younger than patients who declined enrollment (39 yr vs 48 yr, respectively, p = 0.03). Of the 150 subjects enrolled, 57 (38.0%) received all six possible consecutive doses and thus constituted the a priori subgroup.

Figure 1

Figure 1

Baseline characteristics of both the iron and placebo groups are shown in Table 1. There was no difference between groups in enrollment site (p = 0.63), age (p = 0.71), mechanism of injury (p = 0.73), comorbidity score (p = 0.15), ISS (p = 0.28), time from ICU admission to study enrollment (p = 0.12), number of doses of study drug (p = 0.94), and number of missed doses (there were no missed doses in either group). In the iron group as compared with the placebo group, there was a significantly increased proportion of male patients (77.3% vs 61.3%, p = 0.03), a significantly greater baseline EBL (196 mL vs 57 mL, respectively, p = 0.02), and a significantly greater EBL/study day (75.9 mL vs 53.6 mL, respectively, p = 0.04). Furthermore, there was a trend toward an increased admission APACHE II score in the iron as compared with placebo group (23.1 vs 20.9, respectively, p = 0.10).

TABLE 1

TABLE 1

Both iron and hematologic markers for the sample at baseline suggested functional iron deficiency. The median baseline serum iron concentration was 18 ug/dL (range, 5–137), and 134 subjects (89.3%) were hypoferremic. The median baseline ferritin concentration was 247.0 (range, 18.0–967.0), and 51 subjects (34.0%) were hyperferritinemic. Only two subjects (1.3%) were hypoferritinemic at baseline (serum ferritin concentration, < 28 ug/mL), suggesting underlying total body iron deficiency as an etiology for IDE. Study drug was withheld in three subjects (3%) during the study because of a serum ferritin concentration more than 1,000 ug/dL on study day 7, two (2.7%) in the iron group, and one (1.3%) in the placebo group.

The median baseline transferrin saturation was 8% (range, 2–58%), and 133 subjects (88.7%) had a low transferrin saturation. The median baseline eZPP was 68 umol:mol heme (range, 29–187), and 64 subjects (42.7%) had IDE at baseline as evidenced by an elevated eZPP. Furthermore, baseline IDE was associated with severity of injury; the median APACHE II score was higher for patients with baseline IDE as compared with subjects without baseline IDE (24 vs 20, respectively, p = 0.01). The median MCV at baseline was 91.0 (range, 89.0–96.0) and the median reticulocyte count was 1.3 (range, 0.4–5.9).

Iron markers at baseline, study day 7, and study day 14 among the subgroup of subjects who received all six doses of study drug (n = 57) are shown in Figure 2. Both groups remained hypoferremic at all time points, and there was no significant difference in serum iron concentration between the iron and placebo groups at any time point.

Figure 2

Figure 2

Both the iron and placebo groups began in the normal range for ferritin concentration (282.8 ng/mL vs 316.2 ng/mL, respectively, p = 0.99), increased into the hyperferritinemic range at day 7, and increased further at day 14. The iron group, as compared with the placebo group, had a significantly increased ferritin concentration at both day 7 (808.0 ng/mL vs 457.0 ng/mL, respectively, p < 0.01) and day 14 (1,046.0 ng/mL vs 551.5 ng/mL, respectively, p < 0.01).

The transferrin saturation remained below the normal range for both groups at all time points. At day 7, the iron group, as compared with the placebo group, had a significantly higher transferrin saturation, although it remained below the normal range (15% vs 11%, respectively, p = 0.02). There was no significant difference in transferrin saturation between the iron and placebo groups at day 14 (16% vs 13%, respectively, p = 0.23).

The eZPP concentration began at the high end of normal range for both the iron and placebo groups (67.0 umol:mol heme vs 68.0 umol:mol heme, respectively, p = 0.73). It then continued to rise at both day 7 and day 14. At no time point was a significant difference observed in eZPP concentration for the iron group as compared with the placebo group.

Because pRBCs transfusions contain iron, we repeated analyses of the relationship between iron supplementation and iron markers after controlling for any prior transfusion. The only independent association observed was between iron supplementation and serum ferritin concentration on day 7 (degrees of freedom [df] = 2, model F = 5.23, p = 0.01, t for iron supplementation = 2.62, p = 0.01) and day 14 (df = 2, model F = 4.45, p < 0.01, t for iron supplementation = 2.79, p < 0.01). Controlling for prior pRBCs transfusion did not unmask any association between iron supplementation and serum iron concentration, transferrin saturation, or eZPP concentration.

The hemoglobin concentration at baseline and over the subsequent 14 days among the subgroup of patients who received all six doses of study drug (n = 57) is shown in Figure 3. At baseline, the hemoglobin concentration was significantly higher for the placebo group as compared with the iron group (9.9 g/dL vs 8.8 g/dL, respectively, p = 0.03). At no other time point was a significant difference in hemoglobin concentration observed between groups.

Figure 3

Figure 3

Iron supplementation did not decrease the transfusion requirement. During the study period, 55 subjects in the iron group (73.3%) and 47 subjects in the placebo group (62.7%) received at least one pRBCs transfusion (p = 0.16). After controlling for baseline EBL, there remained no difference in risk of transfusion between groups (df = 2, model chi-square = 6.2, p = 0.04, OR for iron supplementation = 1.56, 95% CI, 0.77, 3.17, p = 0.21). As compared with the placebo group, the iron group had a significantly lower percentage of transfusion-free days per subject (84.0% vs 90.5%, respectively, p = 0.02).

We next examined both the percentage of subjects transfused on each study day and the EBL on each study day. The only study day for which there was a significant difference in percentage of patients transfused between the iron and placebo group was study day 1 (29.3% vs 13.3%, respectively, p = 0.02). The only significant differences in EBL between the iron and placebo groups were prior to study entry (196 mL vs 57 mL, respectively, p = 0.02) and on study day 1 (87 mL vs 24 mL, respectively, p = 0.04).

Approximately one half of subjects had received at least one pRBCs transfusion prior to study entry (n = 43 [57.3%] iron vs 37 [49.3%] placebo, p = 0.34). In order to address the possibility of confounding by baseline transfusion, we performed a subgroup analysis of the 70 subjects who had not received a pRBCs transfusion prior to study entry. There was no difference in baseline EBL between the iron and placebo groups (62 mL vs 41 mL, respectively, p = 0.36). Furthermore, there was no difference in the proportion of subjects transfused during the study period between the iron and placebo groups (65.6% vs 52.3%, respectively, p = 0.27).

Iron supplementation did not increase the risk of infection (Fig. 4). At least one infection occurred in 44 subjects in the iron group (58.7%) and 52 subjects in the placebo group (69.3%, p = 0.17). The median number of infections per subject was 2 for both groups (p = 0.18). Specifically, there was no difference in the risk of pneumonia (p = 0.55), bacteremia (p = 0.95), UTI (p = 0.66), or infections from other sources (p = 0.83). Both the antibiotic days (p = 0.64) and antibiotic days per study day (p = 0.55) were equivalent between groups. Finally, there remained no association between study group and infection risk in subgroup analysis of subjects who received all six doses of study drug (92.6% for the iron group and 90.0% for the placebo group, p = 0.73).

Figure 4

Figure 4

Additional outcomes are summarized in Table 2. By univariate analysis, there was a trend toward an increased mortality in the iron group as compared with the placebo group (9.3% vs 2.7%, p = 0.09). This trend persisted in subgroup analysis of subjects who received six doses of study drug (11.1% vs 3.3%, p = 0.25). However, after controlling for gender, baseline APACHE II score, and baseline EBL, there was no independent association between study group and mortality (df = 4, model chi-square = 8.8, p = 0.04, OR for iron supplementation = 1.12, 95% CI, 0.92, 1.33, p = 0.67), suggesting that baseline differences between groups confounded this relationship. Neither ICU LOS (p = 0.53) nor hospital LOS (p = 0.50) differed between study groups.

TABLE 2

TABLE 2

In order to access for confounding by enrollment site, we repeated analyses for iron markers, hemoglobin concentration, pRBCs transfusion, and infection among the subgroup of subjects enrolled from the lead study center (n = 119). All outcomes remained unchanged (data not shown).

Back to Top | Article Outline

DISCUSSION

In this multicenter RCT of anemic, critically ill trauma patients, iron sucrose 100 mg IV thrice weekly for up to 2 weeks did not impact functional iron deficiency, IDE, anemia, or pRBCs transfusion requirement. Iron supplementation at this level did not increase the risk of infection.

Most ICU patients become anemic soon after admission and remain so for several months thereafter (32). Although the hemoglobin transfusion threshold has been lowered in most ICUs (33), pRBCs transfusion remains a common occurrence, as evidenced by the 68% prevalence observed in our sample. The complications associated with pRBCs transfusions (34–39) underscore the importance of developing alternative strategies that enhance endogenous erythropoiesis.

Iron supplementation represents one such strategy that is based on the observation that ICU patients rapidly develop a functional iron deficiency, characterized by hypoferremia, low transferrin saturation, hyperferritinemia, and IDE. These phenomena are believed to be secondary to up-regulation of the iron-regulatory protein hepcidin, which in turn down-regulates the basolateral membrane protein ferroportin, trapping iron within both duodenal enterocytes and macrophages within the RES (40–42). Without readily available iron substrate, erythrocytes cannot be produced and anemia persists. Although the teleological rationale for this phenomenon remains unknown, sequestration of iron from invading microorganisms has been posited (13). Iron supplementation of critically ill patients may have additional, nonhematologic benefits, including improved cardiac function in patients with congestive heart failure (43).

Currently, iron-prescribing habits among intensivists are sporadic (9), and evidence-based recommendations are needed (44). Despite the aforementioned potential benefits of iron supplementation, several concerns exist, including pharmacy expense, oxidative damage, and infection (13). The encouraging results observed in our RCT of enteral iron supplementation (14) led us to conduct the current investigation, which represents to our knowledge the first multicenter RCT of IV iron supplementation of critically ill patients. An intermediate dose of iron was selected in an effort to balance the benefit on bone marrow iron delivery against the risks of iron overload.

Iron markers drawn at baseline confirmed functional as opposed to total body iron deficiency. This functional iron deficiency persisted despite iron supplementation. Iron supplementation increased the transferrin saturation only minimally on day 7 of supplementation, while significantly increasing the ferritin concentration on both days 7 and 14 of supplementation. Furthermore, IDE persisted for both groups as evidenced by a similar and elevated eZPP concentration. In sum, these findings suggest that iron supplementation at the current level was sequestered as ferritin within the RES as opposed to delivered to the bone marrow in quantities sufficient to impact erythropoiesis.

It follows that iron supplementation neither increased the hemoglobin concentration nor reduced the pRBCs transfusion requirement. Although there was a decrease in transfusion-free days for the iron group as compared with placebo group, this discrepancy was explained by differences observed both at baseline and on study day 1. At baseline, the iron group had lost more blood and started at a lower hemoglobin concentration, despite randomization. Furthermore, on study day 1, the iron group lost significantly more blood and incurred significantly more pRBCs transfusions. It would be unlikely that differences at these time points would be explained by study group assignment, as only one dose of study drug would have been given at this point. That the iron group was likely more severely ill at baseline as compared with the placebo group is highlighted by a trend toward both a higher baseline APACHE score and unadjusted mortality in the former group.

The relative iron deficiency observed in our sample despite markedly elevated ferritin concentration underscores the limitations of this marker for diagnosing iron overload in critically ill patients. By contrast, the transferrin saturation likely represents a more accurate marker of readily available iron for deposition in the bone marrow (or consumption by invading microorganisms). In the outpatient setting, transferrin saturation more accurately predicts response to recombinant erythropoietin than ferritin (27), and misinterpretation of hyperferritinemia to indicate adequate iron availability may have impacted negatively trials of recombinant erythropoietin among critically ill patients (45–47). In the current trial, there was no relationship between ferritin concentration and infection, and iron supplementation did not increase the risk of infection, despite substantially increasing the ferritin concentration.

At present, routine iron supplementation of anemic, critically ill patients cannot be recommended. It is possible that larger doses of iron may overcome both the serum and bone marrow iron debt, although the safety of such a strategy remains unstudied. In this trial, we chose a single-dosing scheme, irrespective of transferrin saturation. An alternative strategy employed commonly in dialysis-dependent outpatients involves redosing of IV iron until the transferrin saturation reaches and is maintained at 30–50%, a range that appears optimal for bone marrow iron delivery (48). However, in the setting of the hepcidin-mediated altered iron trafficking characteristic of critical illness, additional iron may only accumulate as ferritin in storage, increasing further the risks of complications without meaningfully increasing bone marrow iron deposition. Consequently, the ultimate treatment strategy for functional iron deficiency during critical illness may involve a combination of goal-directed iron supplementation, targeted to both transferrin saturation and eZPP, and hepcidin antagonism (49). The efficacy of this strategy should be evaluated in terms of both improving anemia and reducing the pRBCs transfusion requirement, the latter of which continues to decrease in the age of restrictive transfusion (29). Finally, the majority of subjects in this study were discharged from the ICU prior to receiving the full 2-week treatment course, suggesting that most patients will not remain critically ill long enough to reap any potential benefits of iron supplementation. However, even in the subgroup of patients with an ICU LOS more than or equal to 14 days, no discernable benefit of iron supplementation was observed.

Our trial was limited by baseline differences in groups despite randomization, practice variability between centers, and generalizability outside of the critically ill trauma patient. Furthermore, only single blinding could be achieved reliably due to the color of iron sucrose. Hepcidin concentration was not measured due to prohibitive cost; thus, any correlation between hepcidin concentration, degree of functional iron deficiency, and response to iron supplementation could not be addressed. Enrollment was 75% of the target sample size of 200. Finally, the proportion of subjects enrolled from each study center was not equal, although the same number of subjects per group was enrolled from each center.

Back to Top | Article Outline

CONCLUSIONS

This RCT documented severe functional iron deficiency in anemic, critically ill trauma patients. However, iron supplementation of such patients with iron sucrose 100 mg IV thrice weekly for up to 2 weeks did not impact functional iron deficiency, IDE, anemia, or pRBCs transfusion. Iron supplementation at this dose increased the serum transferrin saturation only marginally, although not into the normal range, and increased the serum ferritin concentration significantly. Iron supplementation did not increase the risk of infection. Based on these data, routine IV iron supplementation of anemic, critically ill trauma patients cannot be recommended.

Back to Top | Article Outline

REFERENCES

1. Vincent JL, Baron JF, Reinhart K, et al.ABC (Anemia and Blood Transfusion in Critical Care) Investigators. Anemia and blood transfusion in critically ill patients. JAMA. 2002;288:1499–1507
2. DeBellis RJ. Anemia in critical care patients: Incidence, etiology, impact, management, and use of treatment guidelines and protocols. Am J Health Syst Pharm. 2007;64:S14–21; quiz S28
3. Napolitano LM. Epoetin alfa in the critically ill: What dose? Which route? Crit Care Med. 2009;37:1501–1503
4. Pieracci FM, Barie PS. Diagnosis and management of iron-related anemias in critical illness. Crit Care Med. 2006;34:1898–1905
5. Ganz T. Hepcidin, a key regulator of iron metabolism and mediator of anemia of inflammation. Blood. 2003;102:783–788
6. Pieracci FM, Eachempati SR, Henderson P, et al. Prevalence of hypoferremia and iron-deficient erythropoiesis in anemic critically ill patients, and correlation with severity of illness. Crit Care Med. 2006;34:A132
7. Hastka J, Lasserre JJ, Schwarzbeck A, et al. Central role of zinc protoporphyrin in staging iron deficiency. Clin Chem. 1994;40:768–773
8. Cappa AR, Jaouen BM, Moore EE, et al. Prevalence of hypoferremia and iron-deficienct erythropoiesis in anemic critically ill trauma patients early after admission, and correlation with severity of illness. J Surg Res. 2012;179:A199
9. Lasocki S, Garnier M, Ley L, et al. [How is iron prescribed in French ICUs: A practice survey]. Ann Fr Anesth Reanim. 2008;27:909–914
10. Lapointe M. Iron supplementation in the intensive care unit: When, how much, and by what route? Crit Care. 2004;8(Suppl 2):S37–S41
11. van Iperen CE, Gaillard CA, Kraaijenhagen RJ, et al. Response of erythropoiesis and iron metabolism to recombinant human erythropoietin in intensive care unit patients. Crit Care Med. 2000;28:2773–2778
12. Thomas J, Martinez A. Blood conservation in the critically ill. Am J Health Syst Pharm. 2007;64:S11–S18
13. Pieracci FM, Barie PS. Iron and the risk of infection. Surg Infect (Larchmt). 2005;6(Suppl 1):S41–S46
14. Pieracci FM, Henderson P, Rodney JR, et al. Randomized, double-blind, placebo-controlled trial of effects of enteral iron supplementation on anemia and risk of infection during surgical critical illness. Surg Infect (Larchmt). 2009;10:9–19
15. Nemeth E, Rivera S, Gabayan V, et al. IL-6 mediates hypoferremia of inflammation by inducing the synthesis of the iron regulatory hormone hepcidin. J Clin Invest. 2004;113:1271–1276
16. Moher D, Schulz KF, Altman DCONSORT Group (Consolidated Standards of Reporting Trials). . The CONSORT statement: Revised recommendations for improving the quality of reports of parallel-group randomized trials. JAMA. 2001;285:1987–1991
17. Lipsett PA, Swoboda SM, Dickerson J, et al. Survival and functional outcome after prolonged intensive care unit stay. Ann Surg. 2000;231:262–268
18. Nasraway SA Jr, Hudson-Jinks TM, Kelleher RM. Multidisciplinary care of the obese patient with chronic critical illness after surgery. Crit Care Clin. 2002;18:643–657
19. Takala J, Ruokonen E, Webster NR, et al. Increased mortality associated with growth hormone treatment in critically ill adults. N Engl J Med. 1999;341:785–792
20. Garner JS, Jarvis WR, Emori TG, et al. CDC definitions for nosocomial infections, 1988. Am J Infect Control. 1988;16:128–140
21. Fishbane S, Kowalski EA. The comparative safety of intravenous iron dextran, iron saccharate, and sodium ferric gluconate. Semin Dial. 2000;13:381–384
22. Silverstein SB, Rodgers GM. Parenteral iron therapy options. Am J Hematol. 2004;76:74–78
23. Van Wyck DB, Cavallo G, Spinowitz BS, et al. Safety and efficacy of iron sucrose in patients sensitive to iron dextran: North American clinical trial. Am J Kidney Dis. 2000;36:88–97
24. Georgopoulos D, Matamis D, Routsi C, et al.Critical Care Clinical Trials Greek Group. Recombinant human erythropoietin therapy in critically ill patients: A dose-response study [ISRCTN48523317]. Crit Care. 2005;9:R508–R515
25. Bleackley MR, Wong AY, Hudson DM, et al. Blood iron homeostasis: Newly discovered proteins and iron imbalance. Transfus Med Rev. 2009;23:103–123
26. Charlson M, Szatrowski TP, Peterson J, et al. Validation of a combined comorbidity index. J Clin Epidemiol. 1994;47:1245–1251
27. Gaweda AE, Bhat P, Maglinte GA, et al. TSAT is a better predictor than ferritin of hemoglobin response to Epoetin alfa in US dialysis patients. Hemodial Int. 2014;18:38–46
28. Metzgeroth G, Adelberger V, Dorn-Beineke A, et al. Soluble transferrin receptor and zinc protoporphyrin–competitors or efficient partners? Eur J Haematol. 2005;75:309–317
29. Hébert PC, Wells G, Blajchman MA, et al. A multicenter, randomized, controlled clinical trial of transfusion requirements in critical care. Transfusion Requirements in Critical Care Investigators, Canadian Critical Care Trials Group. N Engl J Med. 1999;340:409–417
30. Croce MA, Fabian TC, Mueller EW, et al. The appropriate diagnostic threshold for ventilator-associated pneumonia using quantitative cultures. J Trauma. 2004;56:931–934; discussion 934–936
31. Kleinbaum DG, Kupper LL, Muller KE:. Applied Regression Analysis and Other Multivariable Methods. The Duxbury series in statistics and decision sciences. 1988Second Edition Boston, MA, PWS-Kent Pub,:pp 718, xviii
32. Bateman AP, McArdle F, Walsh TS. Time course of anemia during six months follow up following intensive care discharge and factors associated with impaired recovery of erythropoiesis. Crit Care Med. 2009;37:1906–1912
33. Napolitano LM, Kurek S, Luchette FA, et al.American College of Critical Care Medicine of the Society of Critical Care Medicine; Eastern Association for the Surgery of Trauma Practice Management Workgroup. Clinical practice guideline: Red blood cell transfusion in adult trauma and critical care. Crit Care Med. 2009;37:3124–3157
34. Hill GE, Frawley WH, Griffith KE, et al. Allogeneic blood transfusion increases the risk of postoperative bacterial infection: A meta-analysis. J Trauma. 2003;54:908–914
35. Malone DL, Dunne J, Tracy JK, et al. Blood transfusion, independent of shock severity, is associated with worse outcome in trauma. J Trauma. 2003;54:898–905; discussion 905–907
36. Bochicchio GV, Napolitano L, Joshi M, et al. Outcome analysis of blood product transfusion in trauma patients: A prospective, risk-adjusted study. World J Surg. 2008;32:2185–2189
37. Bochicchio GV, Napolitano L, Joshi M, et al. Blood product transfusion and ventilator-associated pneumonia in trauma patients. Surg Infect (Larchmt). 2008;9:415–422
38. Offner PJ, Moore EE, Biffl WL, et al. Increased rate of infection associated with transfusion of old blood after severe injury. Arch Surg. 2002;137:711–716; discussion 716–717
39. Silliman CC, Moore EE, Johnson JL, et al. Transfusion of the injured patient: Proceed with caution. Shock. 2004;21:291–299
40. Nicolas G, Chauvet C, Viatte L, et al. The gene encoding the iron regulatory peptide hepcidin is regulated by anemia, hypoxia, and inflammation. J Clin Invest. 2002;110:1037–1044
41. Weinstein DA, Roy CN, Fleming MD, et al. Inappropriate expression of hepcidin is associated with iron refractory anemia: Implications for the anemia of chronic disease. Blood. 2002;100:3776–3781
42. Nemeth E, Valore EV, Territo M, et al. Hepcidin, a putative mediator of anemia of inflammation, is a type II acute-phase protein. Blood. 2003;101:2461–2463
43. Silverberg DS, Iaina A, Schwartz D, et al. Intravenous iron in heart failure: Beyond targeting anemia. Curr Heart Fail Rep. 2011;8:14–21
44. Piagnerelli M, Vincent JL. Role of iron in anaemic critically ill patients: It’s time to investigate! Crit Care. 2004;8:306–307
45. Corwin HL, Gettinger A, Fabian TC, et al.EPO Critical Care Trials Group. Efficacy and safety of epoetin alfa in critically ill patients. N Engl J Med. 2007;357:965–976
46. Corwin HL, Gettinger A, Pearl RG, et al.EPO Critical Care Trials Group. Efficacy of recombinant human erythropoietin in critically ill patients: A randomized controlled trial. JAMA. 2002;288:2827–2835
47. Corwin HL, Gettinger A, Rodriguez RM, et al. Efficacy of recombinant human erythropoietin in the critically ill patient: A randomized, double-blind, placebo-controlled trial. Crit Care Med. 1999;27:2346–2350
48. Minutolo R, Locatelli F, Gallieni M, et al. Anaemia management in non-dialysis chronic kidney disease (CKD) patients: A multicentre prospective study in renal clinics. Nephrol Dial Transplant. 2013;28:3035–3045
49. Guo W, Bachman E, Li M, et al. Testosterone administration inhibits hepcidin transcription and is associated with increased iron incorporation into red blood cells. Aging Cell. 2013;12:280–291
Keywords:

anemia; iron-deficient erythropoiesis; iron supplementation; randomized clinical trial; red blood cell transfusion

Copyright © by 2014 by the Society of Critical Care Medicine and Wolters Kluwer Health, Inc. All Rights Reserved.