Comparative rates of adverse events with different formulations of intravenous iron

Oral iron replacement is the standard therapy in iron-deficiency anemia (IDA) [1]. However, 59% of patients have gastrointestinal toxicity [2]. With impaired iron uptake from the gastrointestinal tract (in anemia of chronic disease (ACD) [3] or after bariatric surgery), suboptimal responsiveness to exogenous erythropoietin (in chronic renal failure) [4], in patients with cancer receiving chemotherapy [5], or when oral iron is poorly tolerated, IV iron therapy is the preferred mode of repletion [2, 5]. Although effective in increasing hemoglobin [6-10], the relative safety of the available IV iron preparations is not well documented. We examined the comparative safety of IV iron formulations used at hospitals associated with our institution. Among 619 unique patients who received IV iron over a 2-year period, we found 32 adverse events (AEs), ranging from urticaria to chest pain. There were no serious AEs or anaphylactic-type reactions. In a multivariate model, there was no difference in AE rates between low-molecular-weight iron dextran (LMWD) and ferric gluconate; however, iron sucrose had significantly higher odds ratio of AEs (OR = 5.7; 95% CI = 1.6–21.3). Our data suggest that AE rates with IV iron are acceptable. More widespread use of LMWD, in particular, which can be given safely as a total dose infusion (TDI), should be considered. Five formulations of IV iron are approved by the United States Food and Drug Administration: LMWD (INFeD in the United States; approved at 100 mg dose), high-molecular-weight iron dextran (HMWD or Dexferrum; approved at 100 mg dose), ferric gluconate (Ferrlecit; approved at 125 mg dose, given over eight sessions for 1 g), iron sucrose (Venofer; approved at 100–400 mg dose, given in three or more sessions for 1 g), and ferumoxytol (Feraheme; approved at 510 mg dose, given in two sessions for 1 g). LMWD, although approved in the United States at 100 mg dose, can be given safely as a TDI of 1 g or more in one session. During the 2-year study period, 619 unique patients received at least one treatment cycle of IV iron. A total of 3,174 infusions of IV iron were given. The mean age of patients was 50.7 years; 81.1% were women and 70.6% were White. Four hundred seventy-eight patients (77.2%) were seen at a tertiary center and 470 (75.9%) were treated for IDA. Three hundred ninety-three patients (63.5%) received ferric gluconate, 121 patients (19.5%) received LMWD, 96 patients (15.5%) received iron sucrose, and nine patients (1.5%) received HMWD. Ferumoxytol was not used at our institution hospitals during the years of this study. A total of 32 AEs were reported in association with IV iron. AEs were seen in 14 patients who received ferric gluconate, three patients who received LMWD, 11 patients who received iron sucrose and four of those treated with HMWD (see Table I). All AEs were mild or moderate and included urticaria/rash, facial flushing, lip numbness, lightheadedness, nausea, swelling/erythema at the injection site, rigors, fevers, and chest pain (see Table II). No severe AEs or anaphylactic-type reactions were noted. All were easily managed and none were complicated by long-term sequelae. Patients who experienced AEs were more likely to be younger (P = 0.003), had ACD rather than IDA (P = 0.04), and were treated in a tertiary setting (P = 0.003). Finally, in a multivariate model with AEs as outcome, there was no difference in rates of AEs between LMWD and ferric gluconate; however, iron sucrose had significantly higher odds ratio of AEs (OR = 5.7; 95% CI = 1.6–21.3) than LMWD or ferric gluconate (see Table III). In our comparison of IV iron formulations in a nondialysis population, there were no serious AEs associated with IV iron use. LMWD was similar to ferric gluconate and better than iron sucrose with respect to AE rates. HMWD, although rarely used, was associated with a high rate of AEs (44.4%), consistent with prior observations [11]. A majority of our patients (75.9%) were being treated for IDA, and the remainder patients (24.1%) were treated for functional iron deficiency associated with ACD (in most cases cancer). Six of seven randomized controlled studies show an increased efficacy of IV iron over oral iron in anemic patients with cancer receiving erythropoiesis-stimulating agents [5, 12-17]. High doses of IV iron overcome the hepcidin-mediated block to iron absorption and iron utilization in patients with chronic disease [18]. Studies including more than 30 million doses of IV iron have demonstrated that HMWD has substantially higher frequencies of serious AEs than other iron formulations, including anaphylactic-type reactions [19]. When HMWD and LMWD are considered together, “iron dextran” is associated with increased risks of serious AEs [20], whereas an evaluation of LMWD alone found serious AEs to be uncommon [21]. Recently, Auerbach et al. demonstrated the safety of a single dose of LMWD for the treatment of IDA. Five hundred seventy infusions of 1 g each were administered over an hour, with a total of 41 AEs in 22 patients. There were no anaphylactic-type reactions and no serious AEs [22]. Our study had limitations. Although based on a large cohort, it was retrospective and may have been susceptible to information bias. Moreover, too few patients received HMWD to be included in the multivariate analysis. Our data in a nondialysis population are consistent with previous reports of safety of LMWD in patients with chronic renal failure and/or on dialysis [23, 24]. Indeed, the risk of mild to moderate AEs seen in ours and previous studies [19] is not higher than the risk of commonly prescribed medications such as proton-pump inhibitors [25]. LMWD has the advantage of being delivered safely as a TDI of iron in one session that has favorable implications for resource utilization, cost, and time benefits for both patients and hospitals [24, 26]. These findings suggest that more widespread use of LMWD (INFeD) should be considered. A prospective clinical trial of the different IV iron formulations that is powered to capture both AEs and efficacy seems warranted. We performed a retrospective review of all adult patients (age ≥ 18 years) who received IV iron at the Dana-Faber/Brigham and Women's Cancer Center from April 1, 2008 to March 31, 2010. Participating hospitals included the Dana-Farber Cancer Institute (Boston, MA), a tertiary oncology hospital, Brigham and Women's Hospital (Boston, MA), a tertiary general hospital, and four affiliated community hospitals in the greater Boston metropolitan area. We used pharmacy records to identify all patients who received IV iron at least once during the study period. Patients on dialysis were excluded. We extracted age, gender, race (White vs. non-White), ethnicity (Hispanic vs. non-Hispanic), indication for iron therapy (ACD or iron-deficiency anemia), iron formulation received and type of treatment center (tertiary or community) from medical records, and the occurrence and management of AEs from nursing records. All data were de-identified. Our study was deemed exempt from review by the institutional review board of the Partners Healthcare System. Each administration of HMWD and LMWD was preceded by a test dose of 25 mg (0.5 ml) via slow IV push over 30 sec or diluted in normal saline over 5 min, followed by observation for 1 hr before administration of the full dose. Typically, a full dose of 1–2 g of LMWD or HMWD was completed over 2.5 to 3 hr. Ferric gluconate was administered by IV infusion over 1 hr at a dose of 125 mg per session. Iron sucrose administration varied by dose. Doses up to 200 mg were given IV push over 2–5 min, 300 mg doses were given over 1.5 hr, and 400 mg doses were given over 2.5 hr. AEs were defined as any undesirable sign, symptom, medical condition, or experience occurring during or shortly after IV iron administration. In calculating rates of AEs, each patient was counted once in the denominator and had only one opportunity to be counted (categorically) in the numerator. Each IV iron infusion was a chance for an AE. We fit a multivariate model with AE as outcome and iron formulation as predictor. On the basis of a priori knowledge, we controlled for patients' age, gender, race, ethnicity, and diagnosis. Treatment with HMWD was excluded from the multivariate model (this preparation was very infrequently used; see Table I). Significance was set at the level of 0.05. Analyses were conducted with the SAS 9.2 statistical package (SAS Institute, Cary, NC). Maureen M. Okam had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. The authors are thankful to Professors Nancy Berliner and Ronald P. McCaffrey for critical review of the manuscript. Maureen M. Okam* , Elyse Mandell* , Nathanael Hevelone , Rachel Wentz*, Ainsley Ross*, Gregory A. Abel§ ¶, * Division of Hematology, Brigham and Women's Hospital, Boston, Massachusetts, Department of Medical Oncology, Dana-Farber Cancer Institute, Boston, Massachusetts, Department of Surgery, Brigham and Women's Hospital, Boston, Massachusetts, § Division of Population Sciences, Department of Medical Oncology, Dana-Farber Cancer Institute, Boston, Massachusetts, ¶ Center for Leukemia, Department of Medical Oncology, Dana-Farber Cancer Institute, Boston, Massachusetts.