Incidence of adverse events in iron‐deficient pregnant women and surgical patients undergoing intravenous iron treatment with ferric isomaltose or ferric carboxymaltose: A systematic review

Iron deficiency anemia (IDA) remains the most common form of anemia worldwide, with pregnant women particularly at risk due to increased iron demand.1-4 Despite increased awareness and a slight decline since the early 2000s, approximately 37% of pregnant women worldwide still suffer from anemia, which is associated with an increased risk of peripartum hemorrhage and contributes to around 19% of maternal deaths.5, 6 This has prompted global health efforts, including the Millennium Development Goals, to improve maternal outcomes.7 Surgical patients with iron deficiency with or without anemia (ID(A)) are also vulnerable, especially in the perioperative setting. Intravenous iron (IVI) is an established element of Patient Blood Management, offering advantages over oral iron, which is often poorly tolerated due to gastrointestinal adverse events (AEs).8-17 Newer third-generation IVI formulations like ferric carboxymaltose (FCM) and ferric (der-)isomaltose (FDM) allow for high-dose administration with potentially lower rates of AEs.18-22 Still, specific AEs such as hypophosphatemia (HP), hypersensitivity reactions (HSRs), and anaphylaxis remain concerns.23, 24 HP, frequently subclinical, is linked to altered iFGF23 metabolism and can cause symptoms overlapping with ID, including fatigue or osteomalacia.25-33 Distinguishing HSRs from anaphylaxis is challenging due to overlapping symptoms and inconsistent terminology.34-36 The umbrella term hypersensitivity is generally used to encompass allergic reactions of the immediate type (IgE-mediated, anaphylaxis) as well as the delayed type (e.g., exanthema). Anaphylaxis is an acute, systemic reaction triggered by the rapid release of mediators from mast cells and basophils, leading to potentially life-threatening symptoms such as airway obstruction, circulatory collapse, and cutaneous manifestations. It can have an allergic cause (IgE-mediated, immediate type hypersensitivity) or be non-allergic in origin. Complement activation-related pseudoallergy (CARPA) has been proposed as a possible mechanism of non-allergic anaphylaxis, though its relevance in pregnancy—where complement activity is elevated—remains unclear.35-38 Existing reviews did not require lab-confirmed ID and did not report on HP, HSRs, or anaphylaxis specifically.21, 39 As IVI use is increasing, assessing these AEs in patients with verified ID is essential to avoid unnecessary risks. Standard markers like ferritin and transferrin saturation (TSAT) are critical for diagnosis.40 This systematic review aims to evaluate the incidence and severity of HP, HSRs, and anaphylactic reactions following IVI treatment with FCM or FDM in pregnant and surgical patients with laboratory-confirmed ID. In addition, the study investigates the reporting quality and classification of these AEs across eligible trials. We prospectively registered the study protocol in PROSPERO with identifier CRD42023428997.41 We conducted a systematic search of MEDLINE (via Ovid), CENTRAL (via Cochrane Library), ClinicalTrials.gov, the WHO International Clinical Trials Registry Platform (ICTRP), and from the website of regulatory bodies (e.g., European Medicine Agency (EMA)) Clinical Trials Register (EUCTR) up to May 23, 2023. Detailed search strategies are provided in Table S1. Search results were imported into EndNote, and duplicates were removed before transferring the records into Covidence, where additional duplicate entries were identified and removed. Two independent reviewers screened all references at the title and abstract level against predefined exclusion criteria. Disagreements were resolved through consultation with a third reviewer or group discussion. Full-text articles of potentially eligible studies were then retrieved and assessed independently by two reviewers. As recommended by the Cochrane Handbook for Systematic Reviews of Interventions, we sought additional information for included studies, such as clinical study reports or data from EUCTR, to ensure comprehensive analysis.42 We included studies published after 2000 that reported the incidence of AEs associated with IVI treatment for ID(A) in pregnant women (aged ≥18 years) or non-pregnant surgical patients (aged ≥18 years). Eligible studies included randomized controlled trials (RCTs) and cohort studies that compared FCM or FDM with oral iron preparations, placebo, standard care (e.g., red blood cell transfusion), or each other in head-to-head comparisons. Inclusion required laboratory-confirmed ID based on hematological parameters, such as ferritin levels or TSAT. Studies published in English or German were eligible for inclusion. Exclusion criteria encompassed studies involving health conditions associated with secondary ID(A), such as sickle cell disease, malaria, or patients requiring hemodialysis. Data extraction was performed using a predefined, piloted Excel sheet. Each extraction was independently verified by a second reviewer, with discrepancies resolved through discussion or consultation with a third reviewer, as in the screening process. Extracted data are presented in Table S2. Initial plans to extract AEs data based solely on explicit mentions in publications proved inadequate for addressing our primary research question. An allergist was consulted to define the terms “anaphylaxis” and “HSR” based on their possible clinical symptoms. The resulting standardized symptom list is provided in Table S4. Using this framework, we extracted the number of symptom events reported in each study. Where possible, we supplemented this with additional information obtained outside the primary publications. Additional methodological details regarding AE assessment were also extracted to identify differences in outcome measurement and reporting. Specific criteria included the use of standardized taxonomies (e.g., MedDRA [Medical Dictionary for Regulatory Activities]) and potential limitations such as thresholding or selective reporting (e.g., focusing only on the intervention group). We also evaluated whether the methods sections of included studies adequately described protocols for assessing unwanted events, including harms and AEs. This approach aimed to highlight discrepancies in AE reporting, providing a clearer picture of the challenges in interpreting outcomes related to HP, HSRs, and anaphylaxis in the included studies. The quality of included studies was evaluated using the Critical Appraisal Skills Programme (CASP) checklists for RCTs and cohort studies.47 Two independent reviewers conducted the assessments, with disagreements resolved through discussion or, when necessary, consultation with a third reviewer. Our search identified 6731 records across the specified databases and registers. After the automated removal of 734 duplicates using EndNote and Covidence, an additional four duplicates were manually removed, resulting in 5993 records for title and abstract screening. Two independent reviewers assessed these records for eligibility, excluding 5625 records that did not meet the criteria. This left 368 reports, from which 363 full-text articles were retrieved and assessed for eligibility. Reasons for further exclusion in process are detailed in the PRISMA flow diagram (Figure 1). Among the included studies, three out of four studies in pregnant women used FCM,48-50 while Hansen et al. used FDM.51 For surgical patients, six studies used FCM,52-57 and one study used FDM.58 The multi-center study by Breymann et al. was conducted across six countries (South Korea, Russia, Australia, Singapore, Sweden, and Switzerland), while four studies were conducted in Europe (Denmark, Netherlands, Italy, and Spain), two in India, three in Asian countries (Hong Kong, Korea, and Singapore), and one in Australia. ID was diagnosed using various criteria across studies. Four studies used serum ferritin alone as a predefined threshold, while five studies combined ferritin with TSAT. One study used TSAT as the sole diagnostic criterion.55 Another study did not specify a hematological parameter for ID but excluded anemia not related to ID(A), allowing for its inclusion in this systematic review.49 The characteristics of the included studies are summarized in Table 1, with details of excluded ongoing studies provided in Table S3. Breymann et al. (2017)48 NCT01131624 Pregnant women with IDA aged 18 years or older within their 2nd or 3rd trimester Singleton only FCM 31.4 (±5.77) FS 30.9 (±5.39) FCM, single or multiple administration, n = 123 Cumulative dosage: 1000 mg for [Hb] >9 g/dL; 1500 mg for [Hb] ≤ 9 g/dL Number of administrations depending on pre-pregnancy body weight Chawla et al. (2022)49 CTRI/2017/06/008884 FCM 26.43 (±3.6) FS 26.26 (±3.46) Hansen et al. (2023)51 NCT03188445 Pregnant women with ID aged 18 years or older within their 2nd trimester Singleton only FDM 30.7 (±4.5) FF-AC 31.3 (±4.6) Khalafallah et al. (2018)50 ACTRN 12613000853741 FCM 28.7 (±5.99) FS 28.5 (±6.3) Thin et al. (2021)52 NCT03295851 [Hb]male <13 g/dL [Hb]female <12 g/dL, serum ferritin levels <100 μg/L or 100–300 μg/L and TSAT ≤20% FCM 59.2 (±12.4) FF 55.2 (±23.3) Froessler et al. (2016)11 ACTRN12611000387921 [Hb]male <13 g/dL [Hb]female <12 g/dL Serum ferritin levels <300 μg/L TSAT <25% FCM 64 (±15) SoC 68 (±15) Kim et al. (2023)54 NCT04898569 [Hb]male <13 g/dL [Hb]female < 12 g/dL Serum ferritin levels <300 μg/L TSAT ≤25% Placebo n = 43 Talboom et al. (2023)55 NCT02243735 [Hb]male <13 g/dL [Hb]female <12 g/dL TSAT ≤20% FCM 72 (IQR 63–79) FF 70 (IQR 61–81) FCM, single or multiple administration, n = 96 Cumulative dosage: Between 2000 mg (for [Hb] ≤10 g/dL and body weight over 70 kg) and 1000 mg (for [Hb] >10 g/dL and body weight 35–70 kg) Fung et al. (2022)58 NCT03565354 [Hb] <13 g/dL Serum ferritin <30 μg/L or 30–100 μg/L TSAT ≤20% FDM 68.4 (±6.8) SoC 69.8 (±12.6) NA Serum ferritin levels <100 μg/L (and displaying normal CRP values) FCM 67.6 (±9.5) SoC 66.0 (±10.9) [Hb]male <13 g/dL [Hb]female <12 g/dL Serum ferritin levels <30 ng/mL, TSAT ≤ 20% FCM 72.9 (±11.1) SoC 70.8 (±10.3) FCM, n = 111 Median total dose 1000 mg (1275 ± 430.1) None of the included studies explicitly reported HP as an AE, nor did they provide phosphate level measurements. Breymann et al.48 recorded 10 cases of reduced phosphate levels among 123 FCM-treated pregnant women, but these were deemed clinically irrelevant. Attempts to infer HP occurrences based on reported symptoms associated with HP proved impractical. HP is asymptomatic in most cases, and when symptoms do occur, they are diverse, often nonspecific, and may overlap with symptoms of ID(A).26-31 Consequently, no aggregation of symptoms potentially indicative of HP was conducted. At first glance no anaphylactic reactions seemed to be reported among the 775 patients treated with FCM or the 120 patients treated with FDM across the included studies. But the reason for this was that three anaphylactic reactions were just not clearly stated as “anaphylaxis” by the study authors. Therefore, we decided that the following events should be labeled as anaphylactic reaction: n = 1 bronchospasm48 as “probable anaphylactic reaction”; n = 2 “bronchospasm, pruritus, and flushing”51 as “possible anaphylactic reaction.” Table 2 provides an overview of all AEs reported in the included studies. HSRs were reported in two studies. Breymann et al.48 documented one case of bronchospasm classified as an HSR related to FCM administration. Hansen et al.51 reported two HSR cases with symptoms including bronchospasm, pruritus, and flushing following FDM administration. Table 3 provides a detailed summary of the reported HSRs, listing the symptoms that led authors to classify the events as HSRs. Notably, Breymann et al.48 described a single symptom (bronchospasm) for their reported HSR, while Hansen et al.51 identified three symptoms (bronchospasm, flushing, and pruritus). The third column of Table 3 lists all reported symptoms potentially associated with HSRs, as derived from our methodology. A comprehensive list of symptoms linked to HSRs can be found in Table S4. The remaining nine studies did not report any HSRs. However, three studies reported general symptoms without attributing them to HSRs.49, 50, 53 Table 3 highlights where additional sources outside the primary publications were utilized. Figure S1 illustrates the discrepancies among reported symptoms, those potentially associated with HSRs, and those with no association to HSRs. Among the included studies, seven48, 51, 52, 54, 56-58 classified HSRs as AEs, while three50, 52, 56 studies designated anaphylactic reactions as AEs. HP was considered an AE in four studies. One study52 categorized HSRs, HP, and anaphylactic reactions collectively as AEs, whereas three studies did not label any of these as AEs. None of the included studies explicitly mentioned the occurrence or incidence of HSRs, HP, or anaphylactic reactions as predefined outcomes in their methods sections. Most harm-related outcomes and AEs described in the methods sections were broadly focused on surgical complications, the need for red blood cell transfusions, or general side effects and treatment tolerability. Two studies51, 54 applied limitations to AE reporting, either by setting a threshold for reporting or by including only severe events. One study51 reported only AEs occurring at a frequency exceeding 5%, a detail not mentioned in the publication but identified in an accompanying EudraCT protocol. For one56 cohort study, no information was available regarding potential limitations in AE reporting. Of the 11 included studies, only three used an established taxonomy for reporting AEs: MedDRA (used by Breymann et al.48 and Hansen et al.51) and CTCAE (used by Talboom et al.55). Two studies52, 57 restricted AE reporting to the intervention group only. All other studies48-51, 53-56, 58 documented and reported AEs, if they occurred, consistently for both intervention and control groups, ensuring comparability. Table 4 summarizes key aspects of AE reporting across studies, with a focus on general AE documentation practices. Limitations, such as selective reporting or inconsistent outcome definitions, are highlighted. HP: yes HSRs: yes Anaphylactic reaction: nob/possible yes HP: yes HSRs: yes Anaphylactic reaction: noc/propable yes HP: no HSRs: no Anaphylactic reaction: no HP: yes HSRs: no Anaphylactic reaction: yes HP: yes HSRs: yes Anaphylactic reaction: yes HP: no HSRs: no Anaphylactic reaction: no HP: no HSRs: yes Anaphylactic reaction: no HP: no HSRs: no Anaphylactic reaction: no HP: no HSRs: yes Anaphylactic reaction: no Number of transfusions Surgical complications Postoperative mortality HP: no HSRs: yes Anaphylactic reaction: yes HP: no HSRs: yes Anaphylactic reaction: no The quality assessment conducted using the Critical Appraisal Skills Programme (CASP) identified methodological shortcomings in both cohort studies included in the review. Of the two, Calleja et al.57 demonstrated greater statistical precision compared to D'Amato et al56 For the RCTs, despite some methodological limitations, most studies were of high quality. A common weakness across the included RCTs was inadequate blinding, with inconsistencies in the blinding of participants, investigators, and outcome assessors. Four studies (Breymann et al., Hansen et al., Khalafallah et al., and Talboom et al.) were notable for their strong methodological rigor. These studies were distinguished by a clearly defined research question, appropriate randomization techniques, and comprehensive outcome reporting.48, 50, 51, 55 In contrast, the study by Chawla et al.49 exhibited more significant limitations. Key issues included problems with randomization processes and incomplete blinding, which resulted in a lower quality rating compared to the other RCTs. Tables S5 (for RCTs) and S6 (for cohort studies) provide a detailed critical appraisal of the included studies, summarizing their strengths and weaknesses based on CASP criteria. To the best of our knowledge, this is the first systematic review to focus on the detection of AEs such as HP, HSRs, and anaphylactic reactions following the administration of third-generation IVI formulations, FDM, and FCM, where laboratory confirmation of ID is required prior to administration. From the 11 included studies, no conclusions could be drawn regarding the occurrence of HP following the administration of FCM and FDM. This absence of data is attributed to the subclinical nature of HP, coupled with the lack of reported phosphate levels in most trials. Only one study noted 10 cases of low phosphate levels in the FCM treatment group, which were deemed clinically irrelevant.48 A total of three potential anaphylactic reactions were reported among 895 patients treated with FCM or FDM. One possible case of anaphylaxis occurred in 775 patients receiving FCM, and two likely cases occurred in 120 patients receiving FDM. Since these three events were not clearly stated as anaphylactic reactions by the study authors, it might be tough for non-allergologists to identify those events as anaphylactic reactions afterwards. Therefore, a standardized approach with a clear definition of and classification of adverse events is necessary for authors reporting study designs as well as researchers trying to interpret incident cases regarding adverse events. Recent studies with larger datasets, such as those by Pandey et al. and Kennedy et al., do not contradict our findings. The primary reason for the higher number of included studies in their reviews is that they did not require laboratory confirmation of ID, unlike our review.21, 39 Pandey et al. concluded that IVI formulations were to oral iron regarding the of AEs, but did not the AEs into specific them only as or they did not the AEs of HP, HSRs, or anaphylactic review also included studies using IVI formulations other FCM and FDM, which the inclusion of additional studies. Kennedy et al. compared the incidence of HSRs following the administration of FDM and of studies a lower incidence of HSRs with FDM compared to FCM, but they concluded that additional head-to-head studies were necessary to these also noted the lack of data on the frequency and severity of HP following IVI These studies highlight the need for research that not only the of IVI but also systematic and reporting of AEs. the these AEs, such as provide for research and could intervention studies in data and reporting of AEs which are by For HP, the absence of a specific at which HP further some symptoms of HP overlap with those of ID, which can it to the two, especially since HP often Despite the challenges associated with and consistently reporting adverse events the administration of iron (IVI) to be generally more only a severe events were a clinical for hypersensitivity reactions remains due to their and often overlapping the reported events were and only three anaphylactic reactions were and the incidence of clinically HSRs was The of our Systematic is the search of published studies, clinical study reports from the website of regulatory bodies (e.g., European Medicine Agency and entries of clinical This approach a comprehensive assessment of the incidence of AEs. only studies with laboratory-confirmed ID were This that only studies patients with anemia of iron deficiency were included and appropriate intervention regarding the cause of The assessment of reported AEs was due to reporting of AEs in studies. This approach a on the as well as the reporting (e.g., to a study group or a of events studies. a standardized for the of these is A of this systematic in its is the low number of included studies due to the of laboratory confirmation of ID, which led to the exclusion of studies at full-text screening. The to focus only on FCM and FDM also may have to the lack of the of included The review also highlights the of among as to for additional such as clinical study reports, were For more in research regarding the incidence of AEs, it is to standardized and reporting for AEs. information among researchers also be for the quality and of in this these it can be concluded that the associated with IVI are likely by the particularly for the treatment of iron deficiency anemia in both pregnant and surgical research should standardized and AE reporting to improve across studies and ensure more data quality of symptom and data quality and and study search and consultation and data quality and and study search and and study search and and study study data quality and were to the and the of the The authors for and of and by This research no specific from any in the or from and from and and no of The data that the of this study are available from the Table S1. Search and Table S2. Extracted Table S3. characteristics of ongoing studies. Table S4. of hypersensitivity Table assessment for RCTs. Table assessment for cohort studies. S1. Search WHO International Clinical Trials Registry Platform and Clinical Trials Register The is not for the or of any information by the authors. should be to the for the