Dialysis Unplugged: Ferumoxytol – “Take A Chance on Me?”

Remember the ABBA song: "Take a Chance on Me"?

In 2008, Tejas Patel, a clinical research fellow that worked in my research group at the Brigham, and I published a paper in the American Journal of Kidney Disease on the safety of ferumoxytol in treating iron deficiency anemia in CKD patients.

The paper reported the phase 3 experience from a double-blind placebo-controlled, crossover, multicenter study of a single 510-mg dose of ferumoxytol versus saline as placebo. 750 patients with CKD stages 1 to 5 and 5D. The intervention comprised of an IV injection of either 17 mL of ferumoxytol or saline placebo over 17 seconds on day 0 and the alternate agent on day 7. Of 750 randomly assigned patients with CKD, 60% were not on dialysis therapy.

The main results of the study were the following:

• Of 420 adverse events reported; 242 in 152 patients (21.3%) with ferumoxytol and 178 in 119 patients (16.7%) with placebo.
• The incidence of related adverse events was 5.2% with ferumoxytol and 4.5% with placebo.
• The most common related adverse events after each treatment included symptoms related to the injection/infusion site, dizziness, pruritus, headache, fatigue, and nausea.



(Click to enlarge)

• Serious adverse events occurred in 21 patients (2.9%) after ferumoxytol and 13 patients (1.8%) after placebo.

We concluded: “Ferumoxytol is well tolerated and has a safety profile similar to placebo in anemic patients with CKD stages 1 to 5 and 5D.”

Since then several papers, review articles, and conference posters have reported that ferumoxytol is well tolerated in CKD patients, and following FDA approval for ferumoxytol, my only reason for not using ferumoxytol was price – it was priced well above the other approved intravenous iron medications on the market.

A few weeks back Tejas Patel, now a nephrology attending in Boston, sent me the abstract of a safety analysis on ferumoxytol; I tracked the paper down and am sharing some of these results, because you may not have seen it. I certainly sat up in my seat after reading the paper.

The paper published February 15, 2012 by George R. Bailie in American Journal of Health-System Pharmacy is unfortunately unavailable open access (the abstract is here). Ailie is a Professor of Nephrology Pharmacy at Albany College. I thought that it was a terrific paper.

Bailie used the Freedom of Information Act to collect all of the adverse effect (AE) reports to the US FDA that cited iron sucrose, ferric gluconate, high- and low-molecular-weight iron dextran products, or ferumoxytol from October 2009 through June 2010. He used two different ways to calculate AEs (by a per-unit-sold basis and in terms of 100-mg dose equivalents (DEq) of iron). Bailie also obtained commercial ales data from a market research vendor.

The key results were:

• A total of 197 reported AEs were identified (a cumulative rate of 14.1 AEs per million units sold).
• The rates of all AE classifications combined ranged from 5.25 to 746 per million units sold for iron sucrose and ferumoxytol, respectively; using the other method of calculation, the rates ranged from 5.24 per million DEq (iron sucrose) to 147 per million DEq (ferumoxytol).
• Relative to iron sucrose and sodium ferric gluconate, ferumoxytol was associated with significantly elevated risks of death (odds ratio [OR], 475 and 156, respectively; p < 0.0001), serious nonfatal AEs (OR, 263 and 121, respectively; p < 0.0001), and all evaluated AE classifications combined (OR, 142 and 109, respectively; p < 0.05).

Strengths and Weaknesses

Frankly, I liked the study for several reasons. 

• Bailie was upfront about who funded the study (American Regent). 
• He recognized that since ferumoxytol is newly introduced, the much higher reports of AE may reflect the weber effect. [“Weber Effect”: The peak reporting for events in a drug on market occurs within the first 2 years of approval during the initial 5 year marketing period. In 1984, Dr J. C. P. Weber plotted the mean number of ADR reports for seven non-steroidal anti-inflammatory drugs (NSAIDs) over the first 5 years of marketing and showed that they peaked at 2 years and then declined rapidly. Notably, Hartnell has replicated the Weber effect]. 
• Bailie also acknowledges that there may be imprecision in how AEs are reported and also in the classification of AEs.

The bottom-line is that the staggeringly high rates of AEs with ferumoxytol compared to the other iron preparations might be due to the Weber effect, so we need to be careful in not over-reacting. However, the magnitude of the difference is concerning. Since the main reason for using ferumoxytol is convenience-of-dosing and not cost in most dialysis patients and probably in the majority of non-dialysis CKD patients I don't see a rationale for using the drug. So, in the case of ferumoxytol the answer to “Take a Chance on Me?” is "No, Not Yet", because "First Do No Harm."