Letter to the editor: in response to: Richard F Pollock & Patrick Biggar. Indirect methods of comparison of the safety of ferric derisomaltose, iron sucrose and ferric carboxymaltose in the treatment of iron deficiency anemia

Dear Editor,

We read with interest the analysis of Pollock and Biggar [1] concerning the comparative safety of ferric derisomaltose/iron isomaltoside 1000 (FDI), ferric carboxymaltose (FCM), and iron sucrose (IS). However, we believe the work has some notable deficits. In particular, insufficient detail is given concerning the applied methodology, without which the report lacks the transparency necessary for the reader to fully comprehend the strength, and ultimately the validity, of the results.

Firstly, no protocol was presented for the study, and in particular, the literature search process is not described in detail. For example, what search terms were used, how many studies were identified in total, and what were the inclusion and exclusion criteria applied to reduce the total number to 21? Furthermore, since two of the included studies of FDI/FCM [2,3] are unpublished and thus not yet available in the public domain (PubMed/Google Scholar), they have yet to undergo peer review and editorial assessment. This represents a deviation from the specified data sources and needs clarification. While the authors write that they searched specifically for data using SMQ (standardized MedDRA queries) coding to describe serious drug-related adverse events and severe hypersensitivity reactions (HSRs), the data extraction methods are not specified. Even the reason for excluding Adkinson’s FCM-ferumoxytol trial [4] on the basis of it lacking subclassification of adverse events (AEs) is not clear, since the report contains clear tabular data on severe hypersensitivity reactions and serious cardiovascular events, differentiating these from the other AE data. The authors state that they used a variety of statistical techniques, expecting that ‘reporting of HSRs using SMQs would only be available in a subset of studies,’ so why were other FCM/IS studies not included? Moreover, can the authors even be certain that no suitable studies using SMQ coding were overlooked, especially in view of the fact that, whereas FDI and FCM are newer products, IS has been on the market in Europe since 1950. Our own search in the public domain (restricted to PubMed) revealed several randomized controlled trials of FCM versus placebo or oral ferrous sulfate, respectively, with a total of well over 1,000 patients [5–8], in which AEs were meticulously classified, graded and reported according to MedDRA terms. In all these studies, a total of one hypersensitivity reaction was reported to be related to FCM treatment. Why, or according to what criteria, these trials were excluded from the analysis is entirely unclear.

Secondly, according to the authors, one aim of their study was to use ‘more robust and less biased data sources.’ Interestingly, however, they presented no bias analysis for the included studies. As specified by the Cochrane Handbook of Systematic Reviews of Interventions [9], ‘Problems with the design and execution of individual studies of healthcare interventions raise questions about the validity of their findings; empirical evidence provides support for this concern’ and ‘It is important to assess risk of bias in all studies in a review irrespective of the anticipated reliability in either the results or the validity of the included studies. For instance, the results may be consistent among the studies, but all the studies may be flawed.’ In this instance, the large majority of the included studies were funded by the same commercial source. This is an example of possible bias that needs to be openly described and assessed: 16 of the 19 publications presented in Table 1 [1] acknowledge funding from the same commercial company, and in the remaining three, although there is no clear indication of the funding source, employees of the same company are among the authors. While the authors briefly note that they attained much of their data from the manufacturer (who also funded the present analysis), a bias assessment would present transparent information to allow readers to make their own evaluation, especially in view of the fact that two of the studies are unpublished material. In the absence of such an assessment, the validity of the present analysis must be questioned. For the purpose of transparency, it would be prudent for the authors to declare that the analysis was based largely on the manufacturer’s data set, and performed on behalf of the manufacturer.

Thirdly, three extension trials [10–12] were included in the data analysis. The authors state ‘in these instances, the HSR data were obtained from the extension period only to avoid double counting.’ What exactly is meant by this is unclear, as the total population of the data analysis for FID (n = 3,922) is the same as the total population of all the studies listed in Table 1 [1] , among which the three extension trials are listed. Since the total number of subjects included in the extension trials was 170, whereas over 1000 patients participated in the parent trials, the question of exactly which patients make up the denominator in the analysis, and whether HSRs from the parent studies were included, is of considerable importance. Therefore – and especially since at the time of the analysis, some of the data concerned were unpublished (Jan 2020) – the readers are left guessing as to exactly how the authors arrived at their conclusions and whether these are indeed significant. Furthermore, the inclusion of extension trials may in itself have potential implications in terms of bias, since the parent trials effectively ‘remove’ most individuals who suffer severe/serious drug-related AEs during induction.

While the authors mention that the diversity of the available data in terms of study designs and the lack of head-to-head studies pose limitations, we suggest that the considerable heterogeneity of the studies in terms of dosage, duration, population (underlying disease, age, presence/absence of anemia, etc.) should have been explicitly addressed and presented in the form of a heterogeneity analysis and given more prominence as an important limitation of the study.

In conclusion, although we are well aware of the difficulties posed by heterogenous studies and the current lack of head-to-head data on intravenous iron safety, in view of the lack of transparency regarding the applied search methodology, as well as possible undisclosed biases in data selection and issues concerning the homogeneity of the studies analyzed, we strongly suggest that the results of the analysis should at best be interpreted with caution.

Declaration of interest

A Aksan has received research funding, consulting and congress fees from Immundiagnostik AG and Vifor.

J Stein has received consulting fees from Pharmacosmos, Vifor, Abbvie, Bristol Myers Squibb, Dr Schar, Falk, Ferring, Fresenius Kabi, Immundiagnostik, Janssen Medice, MSD, Pfizer, Shire, Takeda, Thermofisher; is a Board member for Pharmacosmos, Vifor, Abbvie, Bristol Myers Squibb, Dr Schar, Ferring, Fresenius Kabi, Immundiagnostik, Janssen, MSD, NPS, Takeda, Shield; and receives Lecturing fees from Vifor.

A Dignass has received consulting fees from Abbott, Amgen, Fresenius Kabi, Hexal, Gilead, Galapagos, Pharmacosmos, Vifor, MSD, Ferring, Tillotts, Janssen, Pfizer, Roche/Genentech, Takeda, Falk; receives Lecturing fees from Falk, Ferring, Amgen, MSD, Abbott, Janssen, Pfizer, Pharmacosmos, Vifor, Kompetenznetz CED, Takeda, Tillotts; is involved in manuscript preparations for Falk, Tillotts, Janssen, Pfizer; involved in educational presentations for Pharmacosmos, Falk and Ferring.

The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

Reviewer disclosures

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Acknowledgments

The authors wish to acknowledge J Collins (ICCC Rhein-Main, Germany) for writing and editorial assistance.

Correction Statement

This article has been republished with minor changes. These changes do not impact the academic content of the article.

Additional information

Funding

This paper was not funded.