Ethical Issues and Risk/Benefit Assessment of Iron Chelation Therapy: Advances with Deferiprone/deferoxamine Combinations and Concerns about the Safety, Efficacy and Costs of Deferasirox [Kontoghiorghes GJ, Hemoglobin 2008; 32(1–2):1–15.]

Dear Sir,

A paper recently published in this journal described, in the author's opinion, the ethical issues and risks/benefits of current iron chelation therapies ([1]). The paper purports to be proceedings from the 16th International Conference on Chelation (ICOC) from October 2006, at which the author, George J. Kontoghiorghes, was Chairman of the ICOC Committee and President of the Conference. However, it is unclear whether this article accurately represents the proceedings from this conference, not least because it includes information and data released subsequent to that meeting. It is also interesting to note that Kontoghiorghes is a patent co-author for, and has championed the use of, deferiprone (L1) for more than three decades. The aim of the present article is to correct some of the inaccuracies and misinterpretations of data presented by Kontoghiorghes in his paper, lest they cause undue anxiety to patients receiving treatment with, or to clinicians prescribing, the oral iron chelator deferasirox (Exjade®).

Deferasirox Tolerability Profile

It is now possible to make an evidence-based assessment of the tolerability profile of deferasirox due to the availability of published data from large, well-controlled prospective clinical trials. Kontoghiorghes asserts that data are available from only a “small number of iron-loaded patients,” but the deferasirox clinical trial program is actually the largest conducted for any iron chelator. Five studies have been published on more than 700 patients with β-thalassemia major (studies 105–108) ([2–5]), approximately 200 patients with sickle cell disease (study 109) ([6]) as well as almost 100 patients with other transfusion-dependent anemias such as myelodysplastic syndromes (study 108) ([2]). Many of these patients have received deferasirox in the core 1-year studies or the subsequent ongoing extension studies.

In clinical trials, deferasirox has been generally well tolerated with a clinically manageable safety profile over a median treatment period of 3.5 years to date ([7]). The overall discontinuation rate from study 107 was 5.7% at 1 year in patients with β-thalassemia ([5]), which is considerably lower than that reported with L1 in a similar patient population (15% at 1 year) ([8]). Kontoghiorghes lists the “major toxic side effects” of deferasirox (in Table 2), but fails to put these adverse events (AEs) in any context. For example, the most common AEs associated with deferasirox treatment were gastrointestinal problems such as nausea, vomiting and diarrhea. However, these were generally transient, of mild-to-moderate severity and could be expected given the oral route of administration. Skin rash, which occurred in 7% of patients, was also generally transient and of mild-to-moderate severity, usually occurring within the first 2 weeks and resolving with continued treatment at the same dose.

The author also lists “hepatic toxicity” as a major side effect. However, in clinical trials 6% of patients had alanine aminotransferase (ALT) values >5 × the upper normal limit (ULN) at two consecutive post-baseline visits; levels typically decreased as overall body iron burden was reduced at doses of 20–30 mg/kg/day ([9]). Only 2% of patients in the 1-year trials had elevations in ALT that were considered to be related to treatment. Similarly, “renal failure” is cited as a major side effect. Although mild increases in serum creatinine were seen in about one-third of patients receiving deferasirox in the 1-year trials, with 25% of these patients requiring dose adjustment, there have been no cases of progressive renal impairment attributed to deferasirox in the core clinical trials for which data are now available at 4 years. Importantly, in sickle cell disease patients, many of whom were naive to iron chelation therapy, increases in serum creatinine above the ULN were as infrequent in deferasirox-treated patients (2.3%) as in deferoxamine (DFO)-treated patients (3.2%) ([6]). Kontoghiorghes also cites “auditory and ocular toxicity,” however, such problems were observed in just 1% of patients during the 1-year studies and did not differ from DFO in their reported incidences.

Post-Marketing Experience with Deferasirox

Kontoghiorghes draws many of his conclusions regarding the tolerability of deferasirox from post-marketing surveillance information. However, this is a relatively new system introduced in the USA for all newly licensed products and requires that any patient receiving a prescription for a new drug must report events that occur between the issuing of prescriptions. Such events are reported on the drug labelling package irrespective of whether or not a causal link is established. More than 30,000 patients have now received deferasirox treatment outside the clinical trial setting, many of whom are elderly or have other underlying medical conditions in addition to transfusional iron overload. Thus, while post-marketing reports serve as a potential safety net for issues that may not have been observed due to the design of pre-registration studies, they should not serve as a basis for a balanced analysis of the frequency of drug-related tolerability issues.

The post-marketing issues highlighted by Kontoghiorghes include renal failure, cytopenias and hepatic failure. However, it should be highlighted that most of the fatalities resulting from renal failure occurred in patients with multiple co-morbidities and who were in advanced stages of their hematological disorders ([10]). In addition, most of the patients with cytopenias had pre-existing hematologic disorders that are frequently associated with bone marrow failure and any relationship to deferasirox treatment is uncertain ([10]). Finally, most reports of hepatic failure involved patients with significant co-morbidities such as liver cirrhosis and multi-organ failure; fatal outcomes were reported in some of these patients.

Cost of Chelation Therapy

In his review, Kontoghiorghes states that the price of deferasirox is “many times higher than that of DFO,” suggesting that this will limit its use around the world. However, a recent cost-effectiveness analysis demonstrated that the higher drug cost of deferasirox may be offset by savings either due to the cost of DFO administration (e.g., pumps, needles, etc.) or to the reduced need to treat iron overload-related complications that develop when patients are not compliant with DFO ([11]). It was estimated that treatment with deferasirox results in an additional 4.5 quality-adjusted life years (QALYs) at an additional expected lifetime cost of $US126,018 per patient, yielding a cost-effectiveness ratio of $US28,255 per QALY gained. This value is below the range generally considered acceptable in the USA ([12],[13]).

The author also promotes the use of L1 in developing countries “due to the affordable lower cost chelation therapy.” He claims that it has a “much lower toxicity rate” than deferasirox, but fails to provide scientific evidence to support this statement. Indeed, it is well known that L1 is associated with a significant risk of potentially fatal agranulocytosis that requires weekly blood count monitoring, something that is impractical or impossible in many developing countries. Deferiprone therapy has also been associated with arthropathy, which is particularly debilitating in the developing world. It can affect up to 50% of patients in such countries and can be severely erosive, requiring discontinuation in up to 20% of patients receiving L1 ([14]). The author also strongly supports the benefits of combined chelation therapy with DFO and L1, stating that it has in many cases led to “more effective and less toxic chelation treatments than previous monotherapy.” However, he does not discuss the cost implications of using two chelators as opposed to one, particularly regarding the impact this will have in developing countries.

Correcting Inaccuracies and Misinterpretations

The article contains a number of factual inaccuracies and misinterpretations of data to support Kontoghiorghes' unsubstantiated assertions. For example, he states that deferasirox depletes copper and zinc levels, but although there is weak copper and zinc binding, as with all iron chelators, no increased excretion was observed in marmosets ([15]) and no zinc or copper depletion noted in 1-year prospective trials in patients ([2],[5]). Kontoghiorghes also states that deferasirox increases aluminium absorption from the gut leading to toxicity, misquoting a paper on DFO to support this statement. He incorrectly asserts that deferasirox may increase iron absorption from the gut, despite the fact that careful preclinical studies in animal models have shown that this does not occur ([15]).

The author states that there is “no evidence” deferasirox can either achieve a negative iron balance or remove excess cardiac iron, referencing one of his own review articles to support this statement. On the contrary, a recent publication based on study 107 demonstrated that 47%, 55% and 75% of patients in high (>0.5 mg/kg/day), intermediate (0.3–0.5 mg/kg/day) and low (<0.3 mg/kg/day) transfusional iron intake categories had a reduction in liver iron at deferasirox 20 mg/kg/day, while 96% of the patients in the low iron intake category achieved a reduction at 30 mg/kg/day ([16]). Similarly, deferasirox has been shown to effectively permeate cardiac cells in vitro ([17]) to remove cardiac iron in overloaded gerbils ([18]), and, in preliminary clinical studies, to significantly improve cardiac iron levels (by T2*) ([19]) and left ventricular ejection fraction ([20]). Although Kontoghiorghes states that the deferasirox doses used in the gerbil study (100 mg/kg/day) are much higher than those prescribed in patients, they were based on a previous dose finding study ([21]) and represent 67% of human values when normalized for body surface area ([18]).

Kontoghiorghes also incorrectly suggests that the discontinuation rate due to AEs was 11% in study 109 ([6]). However, this value represents the overall discontinuation rate, with the actual rate due to AEs being 5.3% (compared with 3.2% in DFO-treated patients). In the same paragraph, referencing the deferasirox prescribing information, the author notes that thrombocytopenia, neutropenia and agranulocytosis were amongst the most common AEs leading to discontinuation, which is inaccurate and not supported by any published data.

CONCLUSIONS

As deferasirox was only first registered 1–2 years ago, there are inevitably still a number of unanswered questions about this once-daily oral chelation therapy. For example, it is not yet established how deferasirox should be optimally used at low levels of iron burden (although preliminary data suggests that it is generally well tolerated at serum ferritin levels <1000 ng/mL ([22]) or in patients with heart disease. Further work on the long-term tolerability of deferasirox is clearly required, particularly with respect to its renal effects; to date, deferasirox has been generally well tolerated over a median treatment period of 3.5 years, with no progressive increases in serum creatinine. Several studies are underway that will be important for clarifying these outstanding questions. The observations from well-controlled, prospective trials show that deferasirox is an effective and relatively well tolerated iron chelator. Due to its once-daily oral administration and acceptable tolerability profile, deferasirox is preferred by patients to DFO ([23]), the hitherto first-line therapy in Europe, America and the Middle East.

Declaration of Interest

The authors have participated in, received funding, and served on advisory boards for clinical trials on deferasirox sponsored by Novartis. The authors thank Mudskipper for technical assistance with the manuscript and Novartis for providing the required data. The authors alone are responsible for the content and writing of this article.

ACKNOWLEDGMENT

We acknowledge that this letter has not been published or submitted simultaneously for publication elsewhere.