Discussed in conjunction with question II.

FeD in patients with ESRD has been shown to have a negative prognostic impact and its correction with intravenous (i.v.) Fe improves the clinical situation and reduces morbidity, regardless of the presence of anemia.8 In addition there is growing evidence that FeD per se has negative prognostic implications in patients across the spectrum of CKD,9–21 as well as on health-related quality of life (HRQoL)22 or the general symptoms23 of patients with CKD. Since the definitions of FeD and study populations are not homogeneous, it is difficult to quantify the risk attributable to absolute or functional FeD, but they agree on an increased risk of total mortality,9–12,14,16,17,19,21 cardiovascular (CV) hospitalization,13 CV mortality,9,10,17,21 heart failure (HF)15,20 or CV events11,19 in the presence of FeD (absolute or functional) in most studies. However, these studies only demonstrate an association, because given their observational nature, causality cannot be inferred. There is also no clear evidence of the benefit of FeD correction on CV events, as the trials have been designed to assess its erythropoietic effects in the presence of anemia. The 2012 Kidney Disease Improving Global Outcomes (KDIGO) anemia guidelines6 recommend screening and monitoring of Iron parameters only for patients with anemia. In fact, the latest KDIGO controversies concluded that understanding the clinical impact of FeD and its correction, regardless of the existence of anemia, is a high priority area of research for future studies in patients with CKD.24

The diagnosis of FeD has classically been based on IST, as a marker of circulating Fe (and available to the bone marrow) and serum ferritin, considered an indicator of Fe deposits in the body. However, these indicators are considered unreliable for estimating Fe stores (ferritin is an acute phase reactant that can increase in the presence of inflammation) or for predicting response to ferrotherapy in patients with CKD. Both reticulocyte Hb (CHr) and the percentage of hypochromic red blood cells are functional parameters that can help in the study of ferrokinetics in CKD, since they are more reliable parameters for the evaluation of Fe homeostasis except in the presence of thalassemia. Currently, most hematology laboratories are capable of measuring these indicators. Additionally, they have the advantage of low cost, low variability and are not influenced by inflammation or infections. However, apart from the diagnostic evidence, there is little experience about the implementation of these markers in clinical practice. It is recommended to use the percentage of hypochromic red blood cells, but only if it is possible to process the blood sample within 6 h of collection. Probably, the combination of all these ferrokinetic parameters will allow a more accurate algorithm for the diagnosis of Fe deficiency and the erythropoietic response to ferrotherapy in renal patients. However, according to the latest controversies on the management of anemia from the KDIGO Guide,24 the widespread clinical use of both parameters is limited by the absence of universal clinical decision limits, as well as by the requirement for their determination in fresh blood samples in the case of the percentage of hypochromic red blood cells. These areas have been identified as high priority for future research, so at the present it is difficult to recommend their widespread use and to replace ferritin and IST.

In the PIVOTAL study,25 incident hemodialysis patients treated with ESA were randomized to a proactive group that received 400 mg/month of i.v. Fe sucrose until increasing ferritin levels >700 ng/ml or a TSI >40% vs the reactive group that received low-dose Fe sucrose if ferritin was <200 ng/ml or TSI 20%. After a follow-up of 2.1 years, 29.3% of the proactive group had a composite CV event (death, myocardial infarction, stroke or hospitalization for HF) versus a 32.3% of the reactive group (HR 0.85, 95% CI 0.73–1.0). An analysis of the events individually revealed a reduction in the risk of mortality, hospitalization due to HF or fatal and non-fatal myocardial infarction, with similar rates of stroke. The rates of hospitalization and/or infection were similar in the two groups. The authors concluded that, in hemodialysis patients, a high-dose i.v. Fe regimen administered proactively was superior to a low-dose regimen administered reactively and required lower doses of ESA (19% reduction), so the observed benefit could be, at least in part, due to the lower doses of ESA required.25

The FIND-CKD study26 includes 626 anemic patients with CKD no on dialysis (ND) (eGFR <60 ml/min/1.73 m2) and FeD (IST <20%, ferritin <100 ng/ml) patients were randomized into two different i.v. Fe strategies: high (400–600 ng/ml) or low (100–200 ng/ml) ferritin or FE oral. The superiority of the high ferritin strategy in preventing other anemia treatments was demonstrated,26 with no differences in safety after 1-year follow-up (serious adverse events [AEs], cardiac events, mortality, infections or progression of CKD) between the different groups.25,27

A pilot study28 analyzed the effect of a fixed dose of ferric citrate versus standard treatment in patients with advanced CKD. There were 203 patients randomized with eGFR ≤20 ml/min/1.73 m2, phosphorus ≥3.0 mg/dl, Hb >8.0 g/dl, IST < 55%. The group receiving ferric citrate significantly increased Hb, IST and serum ferritin, and significantly reduced serum phosphate and intact FGF23 (p < 0.001 for all). Compared to themusual treatment, ferric citrate treatment resulted in significantly fewer annual hospital admissions, less time spent in hospital, and a lower incidence of the combined endpoint of death, need for dialysis or transplantation (p = 0.002).28

A prospective, double-blind, randomized study in non-anemic patients with stage 3b-5 CKD and FeD investigated whether 1000 mg i.v. Fe (ferric derisomaltose –FDI–) could improve exercise capacity. After adjusting for baseline characteristics, there was no significant difference in the 6-min walk test (6MWT) between the two groups at 1 month (p = 0.736) or 3 months (p = 0.741). There were nonsignificant increases in 6MWT from baseline to 1 and 3 months in the FDI arm. Hb remained stable and there were statistically significant increases in ferritin and IST at 1 and 3 months (p < 0.001). There was a modest numerical improvement in QoL parameters. There were no adverse events attributable to FDI.29

Thus, the results point to a benefit of FeD correction in CKD. However, its effect is difficult to separate from Hb improvement or ESA dose reduction, so specifically designed randomized studies are needed to confirm these issues.

FeD may cause anemia and hyporesponsiveness to ESAs and should be corrected to ensure optimization of erythropoiesis in anemic patients with CKD.

In patients with anemia associated with CKD, ferrotherapy is aimed at ensuring adequate Fe reserves for erythropoiesis, avoiding or delaying the need for ESA and, in patients receiving ESA treatment, preventing the development of FeD and reducing ESA requirements. As mentioned above, anemia in CKD is associated with an increased risk of morbidity, mortality, and progression of CKD, so early correction is desirable.8

The recommendations of the clinical guidelines coincide on the administration of Fe to CKD patients with anemia. It is not appropriate to extrapolate the Hb thresholds applied for the initiation of ESA prescription (Hb <10 g/dl), which are lower, based on the results of clinical trials with these agents.30,31

Iron, unlike ESAs, is a factor necessary for efficient erythropoiesis, but it is not a growth factor, so there is no risk of exceeding the Hb level above the target limit29,32 (except in patients with polycythemia/polyglobulia).

It is estimated that FeD is present in approximately 46% of non-anemic patients with stable HF. The presence of FeD, independently of the presence of anemia, is associated with increased symptomatology, decreased exercise capacity, QoL, and increased mortality and hospitalizations due to HF progression.33,34

In this regard, several randomized clinical trials such as FAIR-HF,35 CONFIRM-HF,36 EFFECT-HF37 and AFFIRM-AHF38 in patients with HFrEFr and FeD (but not specifically CKD) have demonstrated the safety of Fe carboxymaltose supplementation (FCM), as well as an improvement in symptomatology and in some CV outcome variables.

In a small trial analyzing patients with CKD, HF, anemia and FeD, i.v. Fe administration was associated with improvement in myocardial functional parameters and cardiac dimensions.39 In an observational pilot study in patients with HFrEFr, FeD and renal dysfunction (mean GFR 40 ml/min),) myocardial repletion of Fe was associated with improvement in left ventricular ejection fraction and its end-systolic volume, independently of the change in Hb levels,40 which has been confirmed in another recent larger study in patients with less renal function impairment.41 In CKD patients, a strategy of high-dose FDI versus repeated low-dose Fe sucrose was associated with a more rapid increase in Hb and ferric parameters, as well as fewer CV events, including HF.42 A meta-analysis on the effect of i.v. ferrotherapy in patients with HFrEFr and FeD43 demonstrated an improvement in functional class, symptomatology, exercise capacity and QoL. It also demonstrated a reduction in the combined outcome of all-cause death and CV hospitalization, in the risk of the combined outcome of CV death and hospitalization for worsening HF, and a reduction in the risk of HF hospitalization. Another more recent meta-analysis of FCM studies in patients with HFrEF and FeD showed similar clinical results, although the beneficial effect was smaller in patients with TSAT > 20%.44

The IRONOUT HF study45 studied 225 patients with symptomatic HF and FeD who were randomized to receive 150 mg of oral Fe polysaccharide or placebo. Patients who received oral Fe failed to improve Fe stores or their functional capacity.

For all these reasons, the HF guidelines of the European Society of Cardiology (ESC) recommend considering treatment with i.v. FCM in patients with HFrEF who are symptomatic and present FeD. However, the same guidelines point out that the studies on which the recommendation is based did not have sufficient statistical power to evaluate mortality or CV events, or to analyze the independent effects of Hb levels.46

In hemodialysis patients, the PIVOTAL study has demonstrated a reduction in HF hospitalizations (this study excluded patients with New York Heart Association functional class IV HF) with proactive i.v. Fe administration versus reactive administration only in case of iron deficiency.25

All of which supports the administration of i.v. FCM in patients with HFrEFr and CKD, extrapolating the results obtained in patients with HF.

Qunibi et al.47 compared the efficacy and safety of high-dose i.v. FCM vs. oral Fe in 255 patients with ND-CKD. After 8 weeks, Hb increased more with FCM than oral Fe (1.31 g/dl vs. 0.83 g/dl, p < 0.01) and Hb increased ≥1 g/dl in a higher percentage of patients receiving FCM than oral ferrotherapy (60.4% vs. 34.7% of patients), regardless of whether they were receiving ESA. Similar results were observed in the PROGRESS study in 351 CKD-ND patients with i.v. Fe isomaltoside and similar period of follow-up.48

In the FIND-CKD study,49 the FCM i.v.-high ferritin group had higher Hb and ferritin levels than the other two groups, with no differences in terms of AEs or serious AEs between the different groups, although there was a higher rate of dropouts in the oral Fe group, especially due to gastrointestinal AEs.49 The meta-analysis by Shepshelovich et al.50 included 24 trials, 13 with 2369 patients with ND-CKD and 11 including 818 patients with CKD-5D. Patients treated with i.v. Fe were more likely to achieve an increase in Hb >1 g/dl and ferritin levels were significantly higher in all i.v. Fe groups vs. the oral Fe group. Regarding safety, the analysis demonstrated comparable rates of AEs and serious AEs between i.v. and oral Fe. However, as in many studies the limitation was that the follow-up periods were generally limited to three months, so there are insufficient data on the safety of pre-dialysis i.v. Fe administration in the medium to long term.

Only two randomized clinical trials in CKD patients have addressed the impact of i.v. Fe on outcomes other than anemia correction. The REVOKE study51 was a single-center, randomized, clinical trial that compared the effect of i.v. Fe sucrose versus oral ferrotherapy in 137 patients with CKD-ND on renal function as a primary event at 104 weeks. The study was stopped prematurely because of an increased risk of AEs in the group treated with i.v. Fe. However, some methodological limitations should be considered; most AEs occurred long after the intervention was completed, all AEs had been included (even if they were repeated in the same patient) instead of period of time to the first AE. Furthermore, the number of patients with severe AEs was similar in both groups and the difference was significant only after adjusting. The other clinical trial is the aforementioned PIVOTAL25 which demonstrated that the high-dose proactive i.v. Fe regimen was superior to the low-dose reactive regimen in reducing the rate of combined CV event and all-cause death in hemodialysis patients with anemia treated with ESAs. In this study, the proactive group achieved Hb targets earlier, ESA dose was reduced and was not associated to an increased risk of vascular access thrombosis, infections or hospitalization.

In the PIVOTAL study,25 those patients randomized to receive high and proactive doses of i.v. Fe sacharose had a lower incidence of MACE than patients assigned to low reactive doses of i.v. Fe, with a similar safety profile, demonstrating the superiority of the proactive strategy in reducing CV events.

In the FIND-CKD study in patients with ND-CKD, already mentioned, there were no differences in safety between groups. Furthermore, a post hoc analysis showed that only 21.6% of patients treated with oral Fe achieved an increase in Hb ≥1 g/dl in 4 weeks and among those who did not respond in that time, less than 30% achieved it at the end of the study, at week 52.49 This finding would suggest that an early ineffective response to treatment with oral Fe could be a reason to consider an early switch to i.v. ferrotherapy.

In the FERWON-NEPHRO study, patients were randomized to receive either a single dose of 1 g Fe isomaltoside i.v. or up to 5 doses of 200 mg Fe suscrose i.v. over a two-week period. The high single dose of Fe isomaltóside i.v. induced non inferior hematologic response in 8 weeks with low and similar rates of hypersensitivity reactions with a significantly lower incidence of the composite CV adverse event (MACE, unstable angina, HF, atrial fibrillation, hypertension or hypotension).52

The REPAIR IDA study53 randomized to 2 doses of FCM 750 mg i.v. over 1 week vs Fe sucrose 200 mg i.v. administered in up to 5 doses during 14 day period. No differences were found in the composite safety event (CV death, infarction and stroke), but there were more episodes of transient hypertension in the FCM i.v. group. In terms of efficacy, more patients in the FCM group achieved an increase in Hb ≥1 g/dl.

With these data, it can be recommended that when choosing i.v. Fe therapy in non dialysis patients, the choice should be high dose and low frequency. As recommended by the National Institute of Health and Care Excellence (NICE) guidelines,31 high dose and low frequency is a scheme in which at least 500 mg of Fe is administered in each infusion, with a maximum of two infusions, a pattern that has been shown to be safe and effective.

According to the KDIGO 2012 guidelines, ESAs should be indicated in ND-CKD patients with Hb levels <10 g/dl and in CKD-5D patients with Hb of 9.0–10.0 g/dl. The maintenance target of Hb is 10–11.5 g/dl in the population of CKD patients.54 It is also stated that Hb levels >13 g/dl should not be intentionally achieved with ESA in adult patients.

Additionally, the European Renal Best Practice (ERBP) guidelines indicate that ESAs should be indicated when the Hb concentration is <10 g/dl in all patients with CKD, proposing individualization of both initiation (earlier in low-risk active individuals) and Hb targets, although without exceeding Hb of 12 g/dl in any case.

Finally, the recent NICE guidelines31 recommend individualizing the targets also taking into account the patients’ opinion, symptoms and comorbidities. They recommend an overall Hb target of 10–12 g/dl and do not escalate ESA doses indefinitely if pre-specified therapeutic targets are not achieved.

None of the three guidelines reviewed provide results of studies with specific objectives by patient profile.

Discussed in conjunction with question XI.

In patients with ESA resistance, should treatment with high doses of ESA be maintained, even if the Hb target is not achieved?

There is no universally accepted definition of ESA hyporesponsiveness or resistance. In practice, it means not achieving a target Hb concentration despite receiving a higher than usual dose of a given ESA, or requiring very high doses to maintain a target Hb level. The main guidelines define ESA hyporesponsiveness differently: NKF-KDOQI55 and NICE56 as erythropoietin doses higher than 300 or 450 U/kg/week if given s.c. or i.v. respectively, NICE also includes a darbepoietin dose >1.5 μg/kg/week (once a known cause of resistance such as chronic bleeding or maintained comorbidity has been ruled out). The KDIGO guidelines6 define it as no increase in Hb after treatment with ESA for 1 month at appropriate doses according to weight.

Despite the absence of a precise definition of hyporesponsiveness to ESA, it has been used the ratio between the epoetin dose in IU/kg/week (or the equivalent for other ESA) and the Hb concentration in g/dl, known as the erythropoietic resistance index (ERI). Resistance is defined as an ERI greater than 12,720 IU per week of epoetin/kg body weight/g/dl of Hb.57

The prevalence of resistance to ESA treatment in patients with CKD and anemia will vary depending on the definition and the population studied. For example, recent prevalence of ESA resistance estimates a range from 12.5% (when both Hb level and ESA dose were included in the definition57), to 30.3% (when it is only considered the change from baseline in Hb with baseline Hb <11 g/dl58). Regardless of the definition, hyporesponsiveness to ESA is associated with increased all-cause mortality59–63 probably as a consequence of underlying comorbidity. Recently, a similar association has been demonstrated in HF patients treated with ESAs.64 However, a recent analysis of the DOPPS study demonstrates that ESA resistance in most cases is transient and improves with correction of its causes.65

NICE guidelines recommend considering discontinuation of ESA therapy in cases of resistance in patients requiring frequent transfusions and high doses of ESA, provided that all reversible causes of ESA resistance have been ruled out and treated.31

Increasing the dose of ESA does not improve prognosis in patients with high ERI. In fact, several studies show that the use of high-dose ESA can produce adverse clinical outcomes. In the CHOIR study,66 patients with ND-CKD were randomized to achieve a target Hb of 11.3 or 13.5 g/dl with epoetinα given once weekly. The incidence of the primary outcome (combined death, HF, stroke, and acute myocardial infarction) was higher in patients assigned to the higher Hb target. Further analysis showed that patients receiving high doses of epoetinα (≥20,000 U/week) had more adverse events.67 Failure to achieve target Hb and protocolized use of high-dose epoetin were associated with an increased risk of the combined event in unadjusted analyses. However, in adjusted models, only high-dose epoetinα was associated with a significantly increased risk of reaching the combined event, suggesting that the use of high-dose ESAs in the context of hyporesponsiveness may be more harmful than beneficial. In the TREAT study in patients with diabetes mellitus, treatment with darbepoetin-α in patients assigned to the high Hb correction group (Hb of 13.0 g/dl) doubled the risk of stroke as compared to placebo,68 and in those with baseline hyporesponsiveness was associated with an increased cardiovascular risk.60 Similar results were observed in the RED-HF study in HF patients using the TREAT inclusion criteria.69 These data suggest that the risks of administering very high doses of ESAs outweigh the benefits in patients with ESA hyporesponsiveness,70 and support the recommendations of the KDIGO guidelines of not to increase ESA doses indiscriminately in case of hyporesponsiveness.6

Post-transplant anemia (PTA) affects 30%–45% — of renal transplant recipients and it is associated with increased morbidity71–73 but only a minority of them receive ESA treatment.73 The reasons could be the lack of clear evidence on the risk-benefit of the therapy, as well as the cost of the treatments. At present there are no specific recommendations for the treatment of PTA.6

The neo-PDGF study randomized 104 patients The use of high-dose epoetinβ (4 doses of 30,000 IU, approximately 430 U/kg and dosed over 2 weeks) during the first 2 weeks post-transplant significantly increased Hb levels 1 month post-transplant compared to the no administration (mean Hb 11.1 vs. 10.5 g/dl; p = 0.038) but with no significant difference in the incidence of initial graft dysfunction or glomerular filtration rate.74 Similarly, intraoperative administration of high-dose epoetinα (40,000 IU single dose, approximately 570 U/kg) to 72 renal transplant recipients did not improve the incidence of baseline graft dysfunction, renal function at 1 month, or Hb at 1 month post-transplant.75 The limited sample size of these studies precludes a proper assessment of the potential risks of such high doses of ESA.

In a retrospective cross-sectional cohort study of 1794 renal transplant recipients, anemia (Hb <12.5 g/dl) was significantly associated with increased mortality. In patients without ESAs, spontaneous increase in Hb was associated with decreased all-cause mortality. In ESA-treated patients, improvement in anemia (Hb up to 12.5 g/dl) was also associated with decreased mortality, although achieving Hb >14 g/dl significantly increased the risk of death.76 In the CAPRIT trial,77 125 renal transplant recipients were randomized to epoetinβ with Hb targets of 13–15 g/dl (full correction) or 10.5–11.5 g/dl (partial correction). Compared — with the partial correction group, the complete correction group had a smaller decrease in estimated creatinine clearance (5.9 vs. 2.4 ml/min/1.73 m2), a lower rate of terminal CKD (21% vs. 4.8%), and longer graft survival (80% vs. 95%). There was also a significant improvement in QoL. With respect to CV safety there were no cardiac events (HF, arrhythmia or myocardial infarction) in the complete correction group as compared to the partial correction group (4 patients, 8%). These data are consistent with a study in Japan of 127 kidney transplant recipients treated with ESA for three years.78 More recently, another trial has been published in 55 kidney transplant patients with PTA at three months.79 Patients were randomized to receive epoetinβ with a target of 11.5–13.5 g/dl or no treatment for anemia. After two years of follow-up there were no differences in the progression of renal function between the two groups, with a similar rate of decline in glomerular filtration rate; there were also no differences in proteinuria or blood pressure control. A more marked improvement in QoL was indeed observed in the ESA-treated group. The small number of participants in these trials does not allow to obtain adequate information on the CV safety in trasplanted patients with higher Hb target levels.

The results of these studies, as well as those from the large observational study by Heinze et al.76 suggest that the optimal target Hb level in APT is higher than the suggested target in CKD and should probably be up to 12–13 g/dl. A larger clinical trial, designed like the CAPRIT study, with longer follow-up could help to define the target Hb level. The ALERT study demonstrates that anemia is a risk factor of renal graft loss but not for CV morbidity and mortality; Hb levels are inversely associated to graft loss (HR 0.86 [0.80–0.92] per 1 g/dl Hb p < 0.001).80

Thus, although the data on ESA treatment of anemia in post-transplant kidney patients suggest that it is benefitial, the advantage of maintaining ESAs in the immediate post-transplant period requires specific studies to substantiate its benefit.

There is pathophysiological support for the effect of hypoxia inducible factor (HIF) prolyl hydroxylase inhibitors (HPI), also called HIF stabilizers, on ferrokinetics. HIF stabilizers increase HIF expression, allowing its dimerization and translocation to the cell nucleus. This triggers gene activation that promotes an increase in the production of EPO and expression of its receptor, and other actions on ferric metabolism such as increased production of iron transporters in the cell membrane (DMT1, DcytB), increased levels of transferrin and its receptor, and blockade of hepcidin,30,81,82 favouring the absorption and mobilization of Fe and its transport to the marrow for erythropoiesis.

A recent meta-analysis included a total of 30 studies with 13,146 patients treated with roxadustat, daprodustat, vadadadustat, molidustat, desidustat or enarodustat. HIF stabilizers reduced hepcidin, ferritin and serum iron levels, while total iron binding capacity and transferrin levels increased versus those in the placebo or ESA group.83 However, the results with the different molecules (such as daprodustat or vadadustat) are not uniform which could be due to differences between the molecules or differences in protocol design.

Preclinical studies demonstrate that increased HIF expression reduces hepcidin levels, an acute phase reactant that is elevated in CKD and has been implicated in functional Fe deficits.84 A recent meta-analysis demonstrates that treatment with HIF stabilizers can reduce hepcidin levels, both versus placebo and ESA.83 In the studies compared with placebo the reduction of hepcidin is maintained in all subgroups. For both reasons we can speak of a class effect.

In the studies of roxadustat in ND-CKD, patients were stratified by C-reactive protein (CRP) quintiles and it was shown that they maintained the results of efficacy and dosing in all subgroups. Similarly, while patients with high baseline CRP treated with roxadustat maintained the same dose, those receiving ESAs required dose increases during follow-up.85

In the vadadustat studies, there was no correlation between baseline CRP levels and Hb increase.86

It should be clear that the phase 3 studies were not designed to measure the erythropoietic response in patients resistant to ESAs and that their post-hoc analysis,83 only highlights a lower impairment of the erythropoietic response by intercurrent inflammatory events in patients treated with HIF stabilizers.

There are no comparative CV safety studies between the different HIF stabilizers. Table 3 summarizes results on CV events of the different clinical trials of HIF stabilizers. After the results of the different clinical trials and analysis by the regulatory agencies84,87–90 it can be concluded that the different molecules present a non-inferiority profile with respect to MACE in DD patients, but their results in ND are more controversial. This fact would justify further studies in order to guarantee the safety of these molecules in all patients with CKD.