Abstract
Background:
Two weekly infusions of ferric carboxymaltose (FCM) are commonly prescribed for treatment of iron-deficiency anemia. However, administration of FCM increases intact levels of fibroblast growth factor 23 (FGF23), which causes hypophosphatemia due to renal phosphate wasting, calcitriol deficiency and secondary hyperparathyroidism. The adverse effects of FCM on mineral metabolism and bone health emerged from case reports and secondary analyses of trials. Data on these safety signals with FCM in clinical practice are limited because markers of mineral and bone metabolism are not routinely checked.
Methods:
To obtain real-world experience with effects of FCM on mineral and bone metabolism, we conducted a prospective observational study of 16 women who were managed at a single-center hematology clinic for iron-deficiency anemia. From October 2016 to February 2018, all participants received two weekly infusions of FCM at a hematology infusion clinic. We hypothesized that FCM would decrease phosphate, increase intact FGF23 (iFGF23), and decrease c-terminal FGF23 (cFGF23). Secondary outcomes were changes in hemoglobin, iron indices, urine fractional excretion of phosphate (FePi), parathyroid hormone (PTH), calcitriol, calcium, osteocalcin, and bone-specific alkaline phosphatase (BAP). FCM was administered at weeks zero and one, and we measured laboratory values at weeks zero, one, two, and five of the study. We used linear mixed models to analyze the significance of the changes in laboratory values over time.
Results:
After two FCM infusions, nearly all (14 of 16) participants developed hypophosphatemia. iFGF23 increased, cFGF23 decreased, and phosphate decreased significantly from week zero to week two (iFGF23 increased by +134.0% [40.6, 305.8], p <0.001; cFGF23 decreased by −516.3% [−1332.7, −142.7], p = 0.002; phosphate decreased by −49.8 ± 15.4%, p <0.001). There was also a significant increase in FePi, PTH, and BAP and a significant decrease in calcitriol and calcium from week zero to week two. There was no significant change in osteocalcin during this time period. iFGF23, but not PTH, was independently associated with decreased phosphate. iFGF23 was also significantly associated with decrease in calcitriol from week zero to week two. Elevation in BAP suggests disordered bone mineralization in response to FCM therapy.
Conclusion:
In this prospective observational study of women with iron deficiency anemia, two FCM infusions significantly altered markers of bone mineralization and mineral metabolism. The results suggest that FCM should be used cautiously in the treatment of iron-deficiency anemia.
Keywords: Hypophosphatemia, Iron-deficiency anemia, Ferric carboxymaltose, Fibroblast growth factor 23, Bone-specific alkaline phosphatase
1. Introduction
Iron-deficiency anemia (IDA) is a common medical condition affecting over one billion people worldwide.1,2 IDA leads to significant morbidity, including fatigue, shortness of breath and weakness. Intravenous (IV) iron preparations are an effective method to correct iron deficiency (ID) without the gastrointestinal side effects and daily pill burden associated with oral iron therapy.3 Traditional IV iron preparations require multiple infusions or require a test dose to evaluate for allergic reactions.4 By stabilizing the iron within carboxymaltose carbohydrate polymer, ferric carboxymaltose (FCM) allows for a large single dose of iron, which is then deposited in the reticuloendothelial system.5 Consequently, treatment with FCM allows for administration of large doses of elemental iron in a single infusion over a short time period. Given its ease of use, FCM is a frequently used IV iron preparation for treatment of ID.
Emerging data point to a link between ID, IDA, iron repletion and phosphate metabolism. Fibroblast growth factor 23 (FGF23) is a bone-derived hormone that lowers serum phosphate levels by promoting urinary phosphate excretion and through inhibition of calcitriol production. Both FGF23 transcription and post-translational cleavage of the newly synthesized FGF23 protein determine levels of circulating bioactive FGF23.6,7 Animal and human studies demonstrate that FGF23 transcription is elevated in ID and IDA, paralleled by an equivalent increase in FGF23 cleavage. As a result, serologic measurements in ID and IDA demonstrate normal levels of bioactive intact hormone, as measured by intact FGF23 (iFGF23) assays, and increased breakdown of c-terminal fragments, detected by c-terminal FGF23 (cFGF23) assays that measure both the intact hormone and c-terminal fragments.6,8 The absence of hypophosphatemia in healthy individuals with ID and IDA supports that post-translational FGF23 cleavage matches increased FGF23 transcription in these settings.
With most IV iron preparations, correction of ID and IDA decreases FGF23 transcription and normalizes levels of FGF23. However, prior studies demonstrate that treatment with FCM results in elevation in iFGF23 levels, phosphaturia, calcitriol deficiency and hyperparathyroidism with the consequent development of hypophosphatemia.7,9 FCM may affect phosphate handling by shifting the balance between FGF23 production and FGF23 cleavage. By reducing the stimulus for FGF23 transcription, treatment of ID and IDA with FCM likely decreases the production of FGF23, but it may concurrently decrease iFGF23 cleavage and inactivation. By reducing inactivation more than production of iFGF23, FCM may paradoxically increase bioactive iFGF23. Elevated iFGF23 promotes urinary phosphate excretion, decreases calcitriol, and increases parathyroid hormone (PTH), which result in hypophosphatemia after FCM treatment.7,9
Prior studies examining the effects of FCM on FGF23 and other markers of mineral metabolism were case reports or secondary analyses of randomized controlled trials. Here, we present a real-world, observational prospective study examining the effect of two weekly FCM infusions on markers of mineral metabolism over five weeks in women with IDA. We hypothesized that iron repletion with two weekly doses of FCM infusions would result in an increase in iFGF23, decrease in cFGF23, and reduction in phosphate levels. Prior studies suggest that repeated FCM doses may also affect bone health.10–14 Osteocalcin and bone-specific alkaline phosphatase (BAP) are bone formation markers, which are produced by osteoblasts and are elevated with abnormal bone mineralization.15 We also hypothesized that BAP and osteocalcin would increase in response to FCM infusions.
2. Methods
2.1. Source population
From the hematology clinic, we recruited participants who had IDA, were at least 18 years of age, and were scheduled to receive two 750 mg weekly infusions of FCM at the Rube Walker Blood Center at Northwestern Medical Hospital as part of their treatment regimen for IDA. Recruitment occurred between October 2016 and February 2018. ID diagnosis was determined by the participants’ hematologist, based on serum iron, ferritin and transferrin saturation values prior to study enrollment. Although the definition of ID may have varied by provider, all participants in this study had laboratory values at visit 0 consistent with accepted definitions of iron deficiency (low ferritin <30 ng/l and low transferrin saturation <20%).16 IDA was diagnosed based on accepted definitions of anemia, defined as hemoglobin <12 g/dl, in the presence of ID.17 The need for FCM treatment was determined by the prescribing hematologist. Exclusion criteria included inability to consent, pregnancy, or a treatment plan that included only one infusion of FCM. The Northwestern University Institutional Review Board approved the study, and all participants provided written informed consent.
2.2. Study design
Participating individuals were scheduled for four study visits over five weeks (Figure 1). At week zero, study coordinators completed a medical history, collected vital signs, and took blood and urine samples. Following sample collection, participants received the first FCM infusion. FCM was given per center protocol, with 750mg of FCM in 250 ml of normal saline. At the week one study visit, weight and blood pressure were measured and blood and urine samples were obtained. After sample collection, participants received the second FCM infusion. All participants attended two follow-up visits for collection of vitals and laboratory measurements at weeks two and five.
Figure 1. Schematic of study design.
Sixteen participants were included in the study. All participants received two infusions of ferric carboxymaltose at weeks zero and one after blood and urine sample collection. The participants were followed for a total of five weeks, with routine study visits at weeks zero, one, two and five. Participants who had a low phosphate level (<2.5 mg/dl) at week five were followed in additional safety visits every four weeks until their phosphate reached a level of 2.5 mg/dl or higher. Eight participants needed extended phosphate monitoring. By week 17 of the study, all phosphate levels had normalized.
Due to the frequency and severity of hypophosphatemia in this study, defined as a phosphate level <2.5 mg/dl, prescribing and monitoring patterns of FCM changed during the study period.18 Periodic monitoring of phosphate was implemented by the providing physician, however this was not standardized. Patients were not prescribed two weekly infusions of FCM if they were considered high risk for FCM-induced hypophosphatemia due to an underlying condition or pre-infusion laboratory evaluation. If two infusions were prescribed, the time between the two doses was often greater than one week. Recruitment was limited to sixteen participants.
In response to hypophosphatemia that developed in the study, the initial study design was modified to include a safety protocol for the monitoring and management of hypophosphatemia. Individuals who had a low phosphate level <2.5 mg/dl at week five were followed in additional monthly visits until their phosphate reached a level of 2.5 mg/dl or higher. Those with a phosphate level between 1.0 mg/dl and 1.5 mg/dl were prescribed phosphate supplements if symptomatic, and those with a phosphate level ≤1.0 mg/dl were prescribed phosphate supplements regardless of symptoms. We prescribed potassium, sodium phosphates (Phos-NaK) 280-160-250 mg at an initial dose of one packet three times daily. The dose was increased to two packets three times daily if tolerated. At each visit, trained study coordinators evaluated hypophosphatemia-related symptoms, which included muscle cramps, parasthesias, and respiratory complaints.
At each of the study visits, blood and urine samples were collected for analysis. Samples were not included in the data analysis if they were drawn after an iron infusion. The Northwestern Memorial Hospital laboratory performed same-day testing of hemoglobin, iron indices, PTH, and creatinine using standard, automated techniques. Random urine collections were analyzed by the Northwestern Memorial Hospital laboratory for urinary creatinine and phosphate. Fractional excretion of phosphate (FePi) was calculated from measurements of random urinary creatinine and phosphate, using the equation: urinary phosphate × serum creatinine/phosphate × urinary creatinine.
Additional serum and protease-free EDTA plasma were collected and stored in −86°C Ultra-Low Freezers for biomarker analyses. The Immunoassay and Biomarker Laboratory at the Diabetes Research Institute, University of Miami, measured iFGF23, cFGF23, osteocalcin and BAP. Samples were shipped in batches on dry ice and first-thaw samples were used for analysis of these measurements. iFGF23 and cFGF23 were measured in plasma by ELISA using kits from Immutopics, Inc. (San Clemente, CA) following the manufacturer’s instructions. Intra-assay coefficients of variability (CV), based on measurements of duplicate samples, were 6.5% for iFGF23 and 5.3% for cFGF23. The inter-assay CVs were measured by analyzing control samples on each ELISA plate, and were 3.3% for iFGF23 and 4.2% for cFGF23. Osteocalcin and BAP were measured from first-thaw stored samples from week zero, week one and week two using reagents from Quidel Corporation (San Diego, CA). Additional samples were shipped on dry ice to the Department of Laboratory Medicine, University of Washington for calcitriol (1,25-dihydroxy vitamin D2) testing, using immunoaffinity extraction and liquid chromatography-tandem mass spectrometry (LC-MS/MS).19
2.3. Exposure and outcomes
The study exposure was two FCM infusions. The primary outcomes were changes in phosphate, cFGF23 and iFGF23 after FCM infusions. Secondary outcome measurements included changes in hemoglobin, iron indices, urine FePi, PTH, calcitriol, calcium, osteocalcin and BAP. Safety endpoints were also analyzed, including incidence and severity of hypophosphatemia, associated symptoms and need for phosphate supplementation.
2.4. Statistical analysis
Baseline characteristics are presented as a mean ± standard deviation (SD) for normally distributed continuous variables, while non-normally distributed variables are presented as a median with the interquartile range (IQR). Categorical variables are presented as proportions. Repeated laboratory measures are presented for each study week in table and graphical form for weeks zero to five. Mean or median change of laboratory values over-time were calculated as the mean or median difference in measurements from week zero to weeks one, two and five for each individual. To test the significance of the change in laboratory values over time, we used linear mixed models for repeated measures with an unstructured covariance structure. Non-normally distributed variables were natural log-transformed for these analyses. Time was considered a fixed effect. The models were unadjusted. To assess the relative contributions of iFGF23, PTH and calcitriol in the development of FCM-induced hypophosphatemia, we used multiple linear regression models. The dependent variables were the percent decrease in phosphate and calcitriol and were calculated as percent change from week zero to the nadir of each individuals’ values over the five week study period. The independent variables were percent changes in log iFGF23, log PTH, and calcitriol from week zero to week two. We summarized safety endpoints by number and proportion of participants who developed hypophosphatemia, experienced symptoms or required phosphate supplementation.
3. Results
3.1. Baseline characteristics
All sixteen individuals recruited into the study were women. Of the sixteen women, eight were black, seven were white, and one was Hispanic (Table 1). The mean hemoglobin ± SD at week zero was 9.7 ± 1.1 g/dl. Iron indices for all individuals were consistent with IDA, with low serum iron, transferrin saturation, and ferritin. Two women had an eGFR <60 ml/min/1.73m2. The mean or median values for phosphate, PTH, calcitriol and iFGF23 levels were within normal range at week zero (Table 2). However, at week zero, the median (IQR) serum cFGF23 was markedly elevated (484.2 RU/ml [253.3, 993.6]).
Table 1.
Baseline demographic and laboratory characteristics (N = 16)
| Demographics | Characteristics |
|---|---|
| Age, years | 48.8 ± 13.9 |
| Race, n (%) | |
| Non-Hispanic White | 7 (43.8) |
| Non-Hispanic Black | 8 (50.0) |
| Hispanic | 1 (6.2) |
| Other | 0 (0) |
| Vitals | |
| Systolic blood pressure, mmHg | 122.1 ± 17.7 |
| Diastolic blood pressure, mmHg | 77.7 ± 11.3 |
| Heart rate, bpm | 73.6 ± 11.6 |
| Body mass index, kg/m2 | 30.0 ± 8.8 |
| Iron deficiency parameters | |
| Hemoglobin, g/dl | 9.7 ± 1.1 |
| Iron, μg/dl | 35.2 ± 13.1 |
| *Transferrin saturation, % | 8.7 ± 3.7 |
| Ferritin, ng/ml | 11.2 ± 7.3 |
| Current oral iron supplement, % | 7 (43.8) |
| Etiology of iron-deficiency anemia | |
| Abnormal uterine bleeding | 7 (43.8) |
| Gastrointestinal blood loss | 3 (18.8) |
| Malabsorption | 2 (12.5) |
| Unknown | 4 (25.0) |
| Other | |
| CRP, mg/dl | 0.10 (0.10, 0.10) |
| eGFR, ml/min/1.73m2 | 91.6 ± 26.7 |
N = 15 for transferrin saturation due to missing values.
Continuous variables are presented as mean ± standard deviation for normally distributed data or as median (interquartile range) for skewed data. Categorical variables are presented as proportions.
Abbreviations: bpm, beats per minute; CRP, C-reactive protein; eGFR, estimated glomerular filtration rate.
Table 2.
Effects of FCM therapy on hematologic parameters and markers of mineral metabolism
| Variable | Week 0 | Week 1 | Week 2 | Week 5 |
|---|---|---|---|---|
| Hematologic Variables | ||||
| Hemoglobin, g/dl | n = 16 | n = 14 | n = 16 | n = 16 |
| 9.7 ± 1.1 | 10.5 ± 1.0 | 11.5 ± 1.2 | 12.1 ± 1.0 | |
| Iron, μg/dl | n = 16 | n = 15 | n = 16 | n = 16 |
| 35.2 ± 13.1 | 82.5 ± 22.2 | 101.1 ± 20.3 | 77.2 ± 20.5 | |
| Transferrin saturation, % | n = 15 | n = 15 | n = 16 | n = 16 |
| 8.7 ± 3.7 | 21.7 ± 6.4 | 31.4 ± 8.3 | 24.8 ± 6.1 | |
| Ferritin, ng/ml | n = 16 | n = 15 | n = 16 | n = 16 |
| 11.2 ± 7.3 | 396.6 ± 192.5 | 526.3 ± 343.0 | 207.1 ± 120.8 | |
| Markers of mineral metabolism | ||||
| Phosphate, mg/dl | n = 16 | n = 15 | n = 16 | n = 16 |
| 3.3 ± 0.8 | 2.0 ± 0.7 | 1.7 ± 0.7 | 2.4 ± 1.0 | |
| Fractional excretion of phosphate, % | n = 15 | n = 14 | n = 16 | n = 15 |
| 11.3 ± 6.6 | 18.4 ± 9.6 | 19.9 ± 7.8 | 23.6 ± 11.1 | |
| Serum calcium, mg/dl | n = 16 | n = 15 | n = 16 | n = 16 |
| 9.5 ± 0.3 | 9.3 ± 0.4 | 9.2 ± 0.4 | 9.4 ± 0.5 | |
| Serum parathyroid hormone, pg/ml | n = 16 | n = 15 | n = 16 | n = 16 |
| 45.0 (37.0, 59.0) | 56.0 (41.0, 66.0) | 80.0 (57.5, 104.0) | 68.5 (52.0, 101.5) | |
| Calcitriol, pg/ml | n = 15 | n = 15 | n = 15 | n = 16 |
| 51.0 ± 24.9 | 30.2 ± 23.1 | 21.3 ± 18.1 | 38.8 ± 15.5 | |
| Plasma intact FGF23, pg/ml | n = 16 | n = 15 | n = 16 | n = 16 |
| 69.5 (55.2, 94.4) | 147.8 (102.8, 206.0) | 193.0 (123.6, 238.2) | 97.3 (75.1, 108.8) | |
| Plasma C-terminal FGF23, RU/ml | n = 16 | n = 15 | n = 16 | n = 16 |
| 484.2 (253.3, 993.6) | 154.3 (110.9, 259.6) | 165.0 (131.3, 263.3) | 123.2 (91.0, 170.7) | |
| Osteocalcin, ng/ml | n = 16 | n = 15 | n = 16 | |
| 7.1 ± 4.5 | 6.8 ± 3.4 | 7.1 ± 3.8 | NA | |
| BAP, mcg/l | n = 16 | n = 15 | n = 16 | |
| 17.9 ± 7.5 | 19.0 ± 7.9 | 20.4 ± 8.6 | NA | |
Continuous variables are presented as mean ± standard deviation for normally distributed data or as median (interquartile range) for skewed data.
Missing values are due to samples insufficient for analysis and one protocol violation during week one (labs were drawn after the iron infusion).
Abbreviations: FGF23, fibroblast growth factor 23; BAP, bone-specific alkaline phosphatase; NA, not available.
3.2. Effects of FCM on hemoglobin and iron indices
All participants completed the two FCM infusions and four planned study visits. Each participant received 1500 mg of elemental iron over two FCM infusions. The individual patient values for each laboratory measurement over the study period are presented in Supplemental Figure 1. Mean ± SD hemoglobin levels peaked at week five, with an increase of +2.4 ± 1.3 g/dl from week zero (Figure 2). Iron indices also increased, peaking at week two. From weeks zero to two, mean ± SD increase in iron levels was +65.9 ± 18.3 μg/dl, increase in transferrin saturation was +23.3 ± 7.1%, and increase in serum ferritin was +515.1 ± 339.9 ng/ml.
Figure 2. Graphs of hemoglobin and iron values over time.
Changes in laboratory values over time. Values are presented as mean ± standard error. Intravenous iron was administered at week zero and week one after blood and urine were collected.
3.3. Effects of FCM on iFGF23 and cFGF23
Median iFGF23 rose with each FCM infusion, peaking at week two (Table 2, Figure 3). Values significantly increased from week zero through all subsequent study visits. The median (IQR) change in iFGF23 from week zero to one was +79.0 pg/ml (29.3, 114.4), p < 0.001, from week zero to two was +114.9 pg/ml (39.9, 156.6), p < 0.001, and week zero to five was +16.4 pg/ml (−0.4, 45.8), p = 0.014 (Table 3).
Figure 3. Graphs of intact fibroblast growth factor 23, c-terminal FGF23, phosphate, and fractional excretion of phosphate over time.
Changes in laboratory values over time. Continuous variables are presented as mean ± standard error for normally distributed data and as median and interquartile range for skewed data. Intravenous iron was administered at week zero and week one after blood and urine were collected.
Table 3.
Changes in bone and mineral values after FCM infusions
| Variable | Time frame | Absolute change from week 0 | Percent change from week 0 | p-value |
|---|---|---|---|---|
| iFGF23, pg/ml | Week 0 → 1 | +79.0 (29.3, 114.4) | +77.8 (16.6, 201.0) | <0.001 |
| Week 0 → 2 | +114.9 (39.9, 156.6) | +134.0 (40.6, 305.8) | <0.001 | |
| Week 0 → 5 | +16.4 (−0.4, 45.8) | +22.3 (−0.7, 87.0) | 0.014 | |
| cFGF23, RU/ml | Week 0 → 1 | −247.8 (−638.7, −91.6) | −306.0 (−1167.6, −138.8) | 0.001 |
| Week 0 → 2 | −310.1 (−750.8, −116.5) | −516.3 (−1332.7, −142.7) | 0.002 | |
| Week 0 → 5 | −324.6 (−876.3, −123.2) | −575.6 (−1350.5, −203.7) | <0.001 | |
| Phosphate, mg/dl | Week 0 → 1 | −1.3 ± 0.7 | −38.4 ± 17.2 | <0.001 |
| Week 0 → 2 | −1.6 ± 0.6 | −49.8 ± 15.4 | <0.001 | |
| Week 0 → 5 | −0.9 ± 0.8 | −28.7 ± 22.2 | <0.001 | |
| FePi, % | Week 0 → 1 | +7.8 ± 7.6 | +104.3 ± 118.8 | 0.001 |
| Week 0 → 2 | +9.6 ± 6.7 | +126.1 ± 102.0 | <0.001 | |
| Week 0 → 5 | +13.2 ± 11.3 | +164.8 ± 148.8 | <0.001 | |
| PTH, pg/ml | Week 0 → 1 | +5.0 (−5.0, 13.0) | +9.8 (−10.2, 44.4) | 0.104 |
| Week 0 → 2 | +23.5 (13.0, 38.0) | +51.8 (20.4, 90.1) | <0.001 | |
| Week 0 → 5 | +25.5 (1.0, 48.0) | +56.4 (4.4, 88.0) | 0.009 | |
| Calcitriol, pg/ml | Week 0 → 1 | −18.3 ± 33.6 | −21.2 ± 88.4 | 0.028 |
| Week 0 → 2 | −32.0 ± 29.1 | −54.1 ± 47.6 | <0.001 | |
| Week 0 → 5 | −16.6 ± 25.5 | −17.1 ± 42.8 | 0.051 | |
| Calcium, mg/dl | Week 0 → 1 | −0.2 ± 0.3 | −2.5 ± 3.4 | 0.010 |
| Week 0 → 2 | −0.3 ± 0.4 | −3.1 ± 4.3 | 0.012 | |
| Week 0 → 5 | −0.2 ± 0.4 | −1.6 ± 4.0 | 0.136 | |
| Osteocalcin, ng/ml | Week 0 → 1 | −0.4 ± 2.2 | +10.7 ± 42.1 | 0.548 |
| Week 0 → 2 | +0.0 ± 4.0 | +29.6 ± 116.1 | 0.971 | |
| BAP, mcg/l | Week 0 → 1 | +1.0 ± 1.3 | +6.9 ± 10.8 | 0.009 |
| Week 0 → 2 | +2.5 ± 2.3 | +14.4 ± 14.5 | <0.001 |
Results are reported as mean ± standard deviation for normally distributed data or as median (interquartile range) for skewed data.
P-values are from linear mixed-model repeated-measures analyses with time as a fixed effect and test significant change over time.
Abbreviations: iFGF23, intact fibroblast growth factor 23; cFGF23, c-terminal FGF23; FePi, fractional excretion of phosphate; PTH, parathyroid hormone; BAP, bone-specific alkaline phosphatase.
Median cFGF23 was elevated at week zero, but more than halved after the first FCM infusion and reached its nadir at week five (Table 2, Figure 3). From week zero, median (IQR) cFGF23 decreased significantly by −247.8 RU/ml (−638.7, −91.6) at week one (p = 0.001) and remained significantly low at weeks two and five (−310.1 RU/ml [−750.8, −116.5], p = 0.002; −324.6 RU/ml [−876.3, −123.2], p < 0.001, respectively) (Table 3).
3.4. Effects of FCM on phosphate and FePi
The mean ± SD phosphate at week zero was 3.3 ± 0.8 mg/dl. After the FCM infusions, phosphate levels decreased with a nadir at week two, when the mean ± SD value was 1.7 ± 0.7 mg/dl (Table 2, Figure 3). Phosphate decreased significantly from week zero to each subsequent visit, by −1.3 ± 0.7 mg/dl at week one (p < 0.001), by −1.6 ± 0.6 mg/dl at week two (p < 0.001), and by −0.9 ± 0.8 mg/dl at week five (p < 0.001) (Table 3).
The mean ± SD FePi at week zero was 11.3 ± 6.6%. After the FCM infusions, FePi increased and remained elevated through week five, when the mean ± SD FePi was 23.6 ± 11.1% (Table 2, Figure 3). The mean ± SD FePi was significantly elevated from week zero to weeks one, two and five (+7.8 ± 7.6%, p = 0.001; +9.6 ± 6.7%, p <0.001; +13.2 ± 11.3%, p <0.001 respectively) (Table 3).
3.5. Effects of FCM on PTH, calcitriol and calcium
PTH rose after the FCM infusions, peaking at week two during the follow up period (Table 2, Figure 4). The rise in median (IQR) PTH was not significant from week zero to one (+5.0 pg/ml [−5.0, 13.0], p = 0.104), but was significant after the second iron infusion, from week zero to two (+23.5 pg/ml [13.0, 38.0], p <0.001) and from week zero to five (+25.5 pg/ml [1.0, 48.0], p = 0.009) (Table 3).
Figure 4. Graphs of parathyroid hormone, calcitriol and calcium over time.
Changes in laboratory values over time. Continuous variables are presented as mean ± standard error for normally distributed data and as median and interquartile range for skewed data. Intravenous iron was administered at week zero and week one after blood and urine were collected.
Calcitriol declined after the infusions, with the nadir at week two (Table 2, Figure 4). The decline in calcitriol was significant from week zero to one and week zero to two in the study (−18.3 ± 33.6 pg/ml, p =0.028; −32.0 ± 29.1 pg/ml, p <.001 respectively) (Table 3). At week five, calcitriol levels had increased and were not significantly lower than baseline values (−16.6 ± 25.5 pg/ml, p = 0.051).
Calcium also declined after the infusions with the lowest values at week two (Table 2, Figure 4). The decline in calcium was significant from week zero to one and week zero to two (−0.2 ± 0.3 mg/dl, p = 0.010; −0.3 ± 0.4 mg/dl, p = 0.012 respectively) (Table 3). However, by week five mean ± SD calcium levels were no longer significantly different compared to baseline values (−0.2 ± 0.4 mg/dl, p = 0.136).
3.6. Effects of FCM on osteocalcin and BAP
Osteocalcin and BAP were measured in all participants at weeks zero, one and two. Osteocalcin did not change significantly during the study (from week zero to one, −0.4 ± 2.2 ng/ml, p = 0.548; from week zero to two, 0.0 ± 4.0 ng/ml, p = 0.971) (Table 3, Figure 2). However, mean BAP significantly increased from week zero to week one (+1.0 ± 1.3 mcg/l, p = 0.009) and from week zero to week two (+2.5 ± 2.3 mcg/l, p <0.001).
3.7. Regression analyses modeling decline in phosphate and calcitriol
The regression analyses for the development of hypophosphatemia and calcitriol deficiency are presented in Table 4. In the univariate analysis associating percent decrease in phosphate with change in log iFGF23, log PTH, and calcitriol from week zero to week two, both iFGF23 and PTH were associated with phosphate change. Elevations in both hormones were associated with a decrease in phosphate levels. Calcitriol was not associated with phosphate in the univariate analysis. In the multivariate model including change in log iFGF23, log PTH, and calcitriol from weeks zero to two, only iFGF23 remained significant. In the model evaluating percent decrease in calcitriol with change in log iFGF23 and log PTH from weeks zero to two, only iFGF23 was significant in the univariate and multivariate analysis. Elevated iFGF23 levels were associated with a decrease in calcitriol. PTH was not significantly associated with the change in calcitriol levels.
Table 4.
Linear regression analyses modeling factors associated with changes in phosphate and calcitriol levels
| Unadjusted model | Adjusted model | |||||||
|---|---|---|---|---|---|---|---|---|
| Estimate | SE | t value | p-value | Estimate | SE | t value | p-value | |
| Dependent variable: Percent decrease in phosphate from week 0 to nadir* | ||||||||
| Week 0 → 2 | ||||||||
| Log iFGF23 % change | 0.43 | 0.15 | 2.95 | 0.011 | 0.45 | 0.14 | 3.23 | 0.010 |
| Log PTH % change | 0.57 | 0.25 | 2.28 | 0.039 | 0.39 | 0.24 | 1.65 | 0.133 |
| Calcitriol % change | −0.11 | 0.06 | −1.79 | 0.101 | −0.03 | 0.06 | −0.44 | 0.669 |
| Dependent variable: Percent decrease in calcitriol levels from week 0 to nadir** | ||||||||
| Week 0 → 2 | ||||||||
| Log iFGF23 % change | 0.97 | 0.27 | 3.52 | 0.004 | 0.96 | 0.28 | 3.38 | 0.006 |
| Log PTH % change | −0.29 | 0.60 | −0.49 | 0.632 | −0.21 | 0.44 | −0.49 | 0.636 |
Adjusted models adjusted for log iFGF23, PTH, and calcitriol.
Adjusted models adjusted for iFGF23 and PTH.
Abbreviations: iFGF23, intact fibroblast growth factor 23; PTH, parathyroid hormone.
3.8. Safety data
87.5% (14 of 16) of the participants developed hypophosphatemia during the study (Table 5). In eight participants, hypophosphatemia resolved by week five, four weeks after the second FCM infusion. However, the hypophosphatemia persisted longer in eight participants, requiring extended monthly safety visits (Figure 1). Four participants experienced severe hypophosphatemia with phosphate levels ≤1.0 mg/dl. Per protocol, these participants were prescribed oral phosphate supplementation. No participants had symptoms attributable to the hypophosphatemia. Phosphate normalized by four months in all participants.
Table 5.
Safety data
| Patient characteristic during follow up: | N (%) |
|---|---|
| Phosphate ≤ 2.5 mg/dl | 14 (87.5) |
| Lowest phosphate 1.0 – 2.5 mg/dl | 10 (62.5) |
| Lowest phosphate ≤ 1.0 mg/dl | 4 (25.0) |
| Required phosphate supplementation | 4 (25.0) |
| Symptomatic hypophosphatemia | 0 |
4. Discussion
In this observational physiologic study of women with IDA, we examined changes in phosphate, FGF23, and bone formation markers after two weekly FCM infusions. Treatment of ID and IDA with FCM rapidly decreased cFGF23 levels, suggesting a decrease in FGF23 transcription. Despite reduced cFGF23 levels, FCM treatment resulted in a paradoxical increase in iFGF23 levels, suggesting less proteolytic degradation of iFGF23. Consequently, there was significant renal phosphate wasting with resultant hypophosphatemia, which occurred in nearly all patients in this study (87.5%).
Prior studies also show a high incidence of hypophosphatemia with FCM infusions. In three previous studies comparing FCM to other IV iron formulations, over 50% of participants who received FCM developed phosphate levels < 2.0 mg/dl.7,9,20 While no one in our study had symptoms attributed to hypophosphatemia, several case reports have documented symptomatic FCM-induced hypophosphatemia, with patients experiencing weakness, fatigue, parasthesias and vomiting.21–23 Certain conditions may worsen or prolong FCM-associated hypophosphatemia, such as pre-existing vitamin D deficiency, hyperparathyroidism, and malnutrition.21,22 In contrast, kidney disease may attenuate FCM-induced phosphaturia, as the kidney’s capacity to filter and excrete phosphate decreases as eGFR declines.9
Our study corroborates two previous physiological studies examining the effects of FCM on mineral and bone parameters in iron-deficient individuals. The first study was a randomized controlled trial comparing one dose of FCM to iron dextran in 39 iron-deficient women.7 Participants who received FCM similarly developed significantly elevated iFGF23, hypophosphatemia, and phosphaturia. Ten out of 17 participants developed serum phosphate levels <2.0 mg/dl. By day 35, all but six participants had normal phosphate levels, and by day 80, all phosphate levels had normalized. In this study, participants’ PTH levels remained unchanged. The second study was a randomized controlled trial of 1,997 iron-deficient individuals comparing two doses of FCM to ferumoxytol.9 Out of 997 participants who received FCM, 50.8% developed serum phosphate levels <2.0 mg/dl, and 10% developed severe hypophosphatemia with levels <1.3 mg/dl. At the end of the five week study, 29.1% of participants remained hypophosphatemic with serum phosphate <2.0 mg/dl. Compared to the participants in the prior study who received one infusion of FCM, individuals in this study who received two infusions developed more severe hypophosphatemia, increased elevations in iFGF23, and significant hyperparathyroidism, suggesting worse laboratory derangements with multiple infusions.7 In both of these studies, FCM infusions were associated with lower calcitriol and calcium levels. Our study demonstrates similar results with patients experiencing hypophosphatemia, iFGF23 and PTH elevation, and decreased calcitriol and calcium levels after two FCM infusions. In our study, participants developed more profound hypophosphatemia, with their nadir phosphate level lower than in the previous two studies.
We found that iFGF23 was the primary factor for FCM-induced hypophosphatemia. In the multiple regression models, only iFGF23, not PTH, was independently associated with hypophosphatemia from weeks zero to two. Furthermore, hypophosphatemia developed at week one when iFGF23 had increased significantly but before PTH had risen. iFGF23 also appears to be the main mediator of calcitriol suppression. Calcitriol declined early at week one with the corresponding rise in iFGF23. Calcitriol remained low at week two despite ongoing hypophosphatemia and elevated PTH, which should stimulate calcitriol production. The regression models also support this conclusion, as only iFGF23 was independently associated with decreased calcitriol. iFGF23 can have opposing effects on serum PTH levels. iFGF23 directly inhibits PTH secretion from the parathyroid gland.24 However, the release of the inhibitory effect of calcitriol on PTH production and reduction in serum calcium secondary to reduced calcitriol levels may lead to PTH elevation.25 Our results demonstrated that PTH rose in week 2, despite elevated FGF23 levels, suggesting that the iFGF23-induced reduction in calcitriol was the dominant effect. We conclude that the early rise of iFGF23 caused phosphaturia and suppressed calcitriol production, which in turn, resulted in hypocalcemia and a delayed rise in PTH at week two. Although it has been proposed that FCM infusions may decrease iFGF23 cleavage, the exact biologic mechanisms behind this hypothesis remain unknown.
While FCM has significant short-term effects on renal phosphate handing, it may also have long-term effects on bone mineralization. There are several case reports of elevated alkaline phosphatase, decreased bone density, biopsy-proven osteomalacia, and insufficiency fractures in patients who were given repeated doses of FCM for chronic IDA.10–14,26 FCM-induced hyperparathyroidism may increase bone turnover and low levels of phosphate may decrease mineralization. It is unclear if these bone changes are permanent. In three case reports, individuals with FCM-induced osteomalacia had improvement of their bone imaging after phosphate supplementation and cessation of FCM, suggesting a reversible component of the bone pathology.11–13 Burosumab, a human recombinant monoclonal antibody that blocks FGF23 binding to the FGFR2/Klotho receptor used for the treatment of X-linked hypophosphatemia, has also been shown in a case report to normalize phosphate levels and improve bone mineralization in FCM-induced hypophosphatemia and osteomalacia.26
In our study, we evaluated markers of bone formation, osteocalcin and BAP, in response to FCM infusions. Both osteocalcin and BAP are produced by osteoblasts and are released with bone formation.15,27 Osteocalcin did not significantly change with the FCM infusions. However, BAP significantly increased after each infusion. Elevated BAP has been associated with decreased bone mineral density and is elevated in rickets and ostomalacia.15,28 Our study provides evidence of subclinical bone changes after just two weekly FCM infusions.
Many uncertainties exist with regard to management of FCM-induced hypophosphatemia in clinical practice. Challenges include monitoring phosphate levels, identifying symptoms of hypophosphatemia, treating low phosphate levels, identifying individuals at risk for poor outcomes with FCM, and ensuring bone health with repeated infusions. Many patients develop hypophosphatemia that persists for several months, suggesting that phosphate monitoring, if performed, would need to be done for extended periods of time. The symptoms of hypophosphatemia are largely non-specific and overlap with ID, and thus may not be recognized. Given the severity of renal phosphate wasting, phosphate repletion with oral supplements may be ineffective and may necessitate IV phosphate repletion.10,21,23 Repeated FCM infusions seem to increase the risk of bone disease, however, many patients need more than one dose of IV iron to replete their iron stores and require long-term IV iron supplementation for chronic blood loss. Clinics may be unaware of these challenges and unprepared for extended phosphate monitoring and management. We found that after initiation of this study, prescribing physicians at our institution were more cautious in prescribing two infusions of FCM. We also noted that if FCM was used, then baseline and follow-up phosphate measurements were added to usual assessments in some instances but not in a standardized approach. The clinical management of FCM-induced hypophosphatemia requires further investigation.
Despite having a large amount of laboratory data including bone formation markers over time and no loss to follow-up, we acknowledge several limitations. The current study is a small, observational single center study with all participants receiving FCM therapy. This study consists of a relatively limited population of middle-aged women. While we measured bone formation markers, we did not measure bone resorption markers. In addition, we were unable to study the effects of FCM on patients with CKD. Further study is required in this patient population, as CKD may modulate the effects of FCM on markers of mineral metabolism.9 Despite these limitations, we were still able to demonstrate significant changes in markers of bone mineralization and mineral metabolism after two FCM infusions in a clinical setting.
5. Conclusions
Although described in the literature, hypophosphatemia is not a clinically well-known side effect of FCM therapy, and many physicians may be unaware of the complications associated with FCM use. In addition, there are little data on how to monitor and manage these complications, making FCM use challenging. The clinical consequences of transient hypophosphatemia may be minor for most patients, however, symptomatic hypophosphatemia and bone pathology may increase with certain comorbidities and long-term administration of FCM.10–14 Further clinical and preclinical research is needed to understand the biologic mechanisms that result in iFGF23 elevation after FCM infusions and the clinical effects on our patients.
Supplementary Material
Figure 5. Graphs of osteocalcin and bone-specific alkaline phosphatase over time.
Changes in laboratory values over time. Values are presented as mean ± standard error. Intravenous iron was administered at week zero and week one after blood and urine were collected.
Acknowledgements
This study was supported by grants P30DK114857, R01DK102438 (TI), T32DK108738 (RF), and a National Kidney Foundation of Illinois Young Investigator Grant (RM). Research reported in this publication was also supported, in part, by the National Institutes of Health’s National Center for Advancing Translational Sciences, Grant Number KL2TR001424 and by the National Institutes of Health’s National Center for Advancing Translational Sciences, Grant Number UL1TR001422. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. The views expressed in this article are those of the authors and do not necessarily reflect the position or policy of the Department of Veterans Affairs or the United States government.
Abbreviations:
- FCM
ferric carboxymaltose
- FGF23
fibroblast growth factor 23
- iFGF23
intact fibroblast growth factor 23
- cFGF23
c-terminal FGF23
- FePi
fractional excretion of phosphate
- PTH
parathyroid hormone
- BAP
bone-specific alkaline phosphatase
- IDA
iron-deficiency anemia
- ID
iron deficiency
- IV
intravenous
- CV
coefficients of variability
- LC-MS/MS
liquid chromatography-tandem mass spectrometry
- SD
standard deviation
- IQR
interquartile range
Footnotes
Disclosures
RM has interest in Abbot Laboratories, AbbVie, Inc. and Teva Pharmaceuticals Industries Ltd, and has received honoraria from Akebia/Otsuka. TI has received consultant honoraria from Kyowa Kirin and LifeSci Capital. MW has received research support, honoraria or consultant fees from Akebia, Amag, Amgen, Ardelyx, DiaSorin, Keryx, and Shire. VD receives research funding from Keryx Biopharmaceuticals and has received research funding from Vifor Pharma and consulting honoraria from Keryx Biopharmaceuticals, Vifor Pharma, Luitpold, and Amgen. BS has had consulting meetings with Celgene, Apex Oncology, and Kartos. The remaining authors declare that they have no relevant financial interests.
References
- 1.Zimmermann MB, Hurrell RF. Nutritional iron deficiency. Lancet (London, England). 2007;370(9586):511–520. [DOI] [PubMed] [Google Scholar]
- 2.McLean E, Cogswell M, Egli I, Wojdyla D, de Benoist B. Worldwide prevalence of anaemia, WHO Vitamin and Mineral Nutrition Information System, 1993–2005. Public health nutrition. 2009;12(4):444–454. [DOI] [PubMed] [Google Scholar]
- 3.Auerbach M, Macdougall I. The available intravenous iron formulations: History, efficacy, and toxicology. Hemodialysis international International Symposium on Home Hemodialysis. 2017;21 Suppl 1:S83–s92. [DOI] [PubMed] [Google Scholar]
- 4.Van Wyck DB, Mangione A, Morrison J, Hadley PE, Jehle JA, Goodnough LT. Large-dose intravenous ferric carboxymaltose injection for iron deficiency anemia in heavy uterine bleeding: a randomized, controlled trial. Transfusion. 2009;49(12):2719–2728. [DOI] [PubMed] [Google Scholar]
- 5.Keating GM. Ferric carboxymaltose: a review of its use in iron deficiency. Drugs. 2015;75(1):101–127. [DOI] [PubMed] [Google Scholar]
- 6.Wolf M, White KE. Coupling fibroblast growth factor 23 production and cleavage: iron deficiency, rickets, and kidney disease. Current opinion in nephrology and hypertension. 2014;23(4):411–419. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Wolf M, Koch TA, Bregman DB. Effects of iron deficiency anemia and its treatment on fibroblast growth factor 23 and phosphate homeostasis in women. Journal of bone and mineral research : the official journal of the American Society for Bone and Mineral Research. 2013;28(8):1793–1803. [DOI] [PubMed] [Google Scholar]
- 8.Yamazaki Y, Okazaki R, Shibata M, et al. Increased circulatory level of biologically active full-length FGF-23 in patients with hypophosphatemic rickets/osteomalacia. The Journal of clinical endocrinology and metabolism. 2002;87(11):4957–4960. [DOI] [PubMed] [Google Scholar]
- 9.Wolf M, Chertow GM, Macdougall IC, Kaper R, Krop J, Strauss W. Randomized trial of intravenous iron-induced hypophosphatemia. JCI insight. 2018;3(23). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Schaefer B, Glodny B, Zoller H. Blood and Bone Loser. Gastroenterology. 2017;152(6):e5–e6. [DOI] [PubMed] [Google Scholar]
- 11.Bartko J, Roschger P, Zandieh S, Brehm A, Zwerina J, Klaushofer K. Hypophosphatemia, Severe Bone Pain, Gait Disturbance, and Fatigue Fractures After Iron Substitution in Inflammatory Bowel Disease: A Case Report. Journal of bone and mineral research : the official journal of the American Society for Bone and Mineral Research. 2018;33(3):534–539. [DOI] [PubMed] [Google Scholar]
- 12.Klein K, Asaad S, Econs M, Rubin JE. Severe FGF23-based hypophosphataemic osteomalacia due to ferric carboxymaltose administration. BMJ case reports. 2018;2018. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Moore KL, Kildahl-Andersen O, Kildahl-Andersen R, Tjonnfjord GE. Uncommon adverse effect of a common medication. Tidsskrift for den Norske laegeforening : tidsskrift for praktisk medicin, ny raekke. 2013;133(2):165. [DOI] [PubMed] [Google Scholar]
- 14.Urbina T, Belkhir R, Rossi G, et al. Iron Supplementation-Induced Phosphaturic Osteomalacia: FGF23 is the Culprit. Journal of bone and mineral research : the official journal of the American Society for Bone and Mineral Research. 2018;33(3):540–542. [DOI] [PubMed] [Google Scholar]
- 15.Maffioli P, Derosa G. Overview of Biochemical Markers of Bone Metabolism In: Patel VB, Preedy VR, eds. Biomarkers in Bone Disease. Netherlands: Springer, Dordrecht; 2017:3–21. [Google Scholar]
- 16.Bermejo F, García-López S. A guide to diagnosis of iron deficiency and iron deficiency anemia in digestive diseases. World J Gastroenterol. 2009;15(37):4638–4643. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.World Health Organization, Centers for Disease Control and Prevention. Assessing the iron status of populations: Second edition, including literature reviews. 2007.
- 18.Food and Drug Administration. Guidance for Industry: toxicity grading scale for healthy adult and adolescent volunteers enrolled in preventive vaccine clinical. 2007. [DOI] [PubMed]
- 19.Laha TJ, Strathmann FG, Wang Z, de Boer IH, Thummel KE, Hoofnagle AN. Characterizing antibody cross-reactivity for immunoaffinity purification of analytes prior to multiplexed liquid chromatography-tandem mass spectrometry. Clinical chemistry. 2012;58(12):1711–1716. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Detlie TE, Lindstrom JC, Jahnsen ME, et al. Incidence of hypophosphatemia in patients with inflammatory bowel disease treated with ferric carboxymaltose or iron isomaltoside. Alimentary pharmacology & therapeutics. 2019;50(4):397–406. [DOI] [PubMed] [Google Scholar]
- 21.Blazevic A, Hunze J, Boots JM. Severe hypophosphataemia after intravenous iron administration. The Netherlands journal of medicine. 2014;72(1):49–53. [PubMed] [Google Scholar]
- 22.Fierz YC, Kenmeni R, Gonthier A, Lier F, Pralong F, Coti Bertrand P. Severe and prolonged hypophosphatemia after intravenous iron administration in a malnourished patient. European journal of clinical nutrition. 2014;68(4):531–533. [DOI] [PubMed] [Google Scholar]
- 23.Mani LY, Nseir G, Venetz JP, Pascual M. Severe hypophosphatemia after intravenous administration of iron carboxymaltose in a stable renal transplant recipient. Transplantation. 2010;90(7):804–805. [DOI] [PubMed] [Google Scholar]
- 24.Takashi Y, Fukumoto S. FGF23 beyond Phosphotropic Hormone. Trends in endocrinology and metabolism: TEM. 2018;29(11):755–767. [DOI] [PubMed] [Google Scholar]
- 25.White KE, Econs MJ. Fibroblast Growth Factor-23 In: Rosen CJ, ed. Primer on the Metabolic Bone Diseases and Disorders of Mineral Metabolism. 7 ed. 2025 M Street, NW, Suite 800, Washington, DC 20036: Americal Society for Bone and Mineral Research; 2008:112–116. [Google Scholar]
- 26.Amarnani R, Travis S, Javaid MK. Novel use of burosumab in refractory iron-induced FGF23-mediated hypophosphataemic osteomalacia. Rheumatology (Oxford, England). 2020. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Song L. Calcium and Bone Metabolism Indices. Advances in clinical chemistry. 2017;82:1–46. [DOI] [PubMed] [Google Scholar]
- 28.Ross PD, Knowlton W. Rapid bone loss is associated with increased levels of biochemical markers. Journal of bone and mineral research : the official journal of the American Society for Bone and Mineral Research. 1998;13(2):297–302. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.