This was a retrospective observational study of 190 patients with stage 5 chronic kidney disease requiring renal replacement therapy (haemodialysis or haemodiafiltration) from two UK renal networks (University Hospitals of Leicester NHS Trust and Hull University Teaching Hospitals NHS Trust) from September 2015 to April 2017 plus a year of follow up. The study was part of service improvement in the implementation and assessment of a change in therapy and therefore ethical approval was not required. However, all the principles of the declaration of Helsinki and the ICH and good governance were followed. All subject data were de-identified.
The study examined the efficacy and safety of IV iron sucrose (IS) compared with IV iron isomaltoside 5% (IIM). In both networks, patients were first receiving IS and were then switched to IIM. The primary objective of Hb non-inferiority was assessed by comparing values between the 6-months before change in medication (i.e. patients receiving IS), versus the 12-months after (i.e. patients then receiving IIM). Data were collected over one continuous time frame with no break in therapy. Safety objectives included assessing the incidence of moderate-to-severe hypersensitivity reactions, serious cardiovascular events and other adverse reactions. Data were collected monthly by the dialysis nurses for assessing and implementing the care. Where the exact date of change in therapy was not clear, the time frame was calculated using the middle date as the day between the last dose of IS and the first dose of IIM.
The study was conducted in the local haemodialysis units contained within the umbrella of both NHS Trusts. Fourteen units in total were involved (2 main and 12 satellites) who all follow the same standard protocol for the implementation of care. Patients with a C-reactive protein (CRP) > 50 or with confounding factors for erythropoiesis were excluded from the data set (including, myeloma or other underlying malignancy, recent blood transfusion, recent hospital admission, active infection, active bleeding). Patients who were participating in any other anaemia-based research study were also excluded.
IV irons as well as ESAs are routinely given to patients in order to maintain adequate Hb levels within a target range of 100–120 g/L as per the UK national guidance guidance6,22. In this study, both the IV iron (IIM or IS) and ESA (epoetin alpha or darbepoetin alpha) therapies were used continuously and contiguously and were given on the first or last dialysis session of the week, often in tandem. Initial dosing of the IV irons were based on serum ferritin (< 200 mcg/L) and for maintenance dosing this was a level of less than 300 mcg/L, and if applicable a transferrin saturation (TSAT) of less than 20% irrespective of an inflammatory state. The routine blood samples collected monthly were assessed to decide on the adjustment or temporary suspension of the IV iron and or/ESA dosing based on local policy. For the IV iron, the upper limit for temporary suspension was set at 500 mcg/L, and again each month this was reassessed when it dropped below 300 mcg/L at which point, if required, it could be restarted. The ESA dose was based on the Hb level and was independently adjusted irrespective of IV iron dosing as per departmental protocol. The usual dose of IS or IIM ranged from 0 to 100 mg weekly, in very few cases it was higher than 200 mg weekly, as deemed by the caring physician. Local policy for the use of IV iron in the dialysis setting is based on maintenance dosing with the aim of a serum ferritin level of 300 to 500 mcg/L as noted above.
Blood samples for each month were collected pre-dialysis as per standard departmental protocol. Laboratory tests included biochemical profile: sodium, potassium, bicarbonate, chloride, urea, creatinine and albumin; and bone profile: corrected calcium, phosphate, alkaline phosphatase and PTH; the haematology panel included Hb, serum ferritin, TSAT, baseline B12 and folate and reticulocyte haemoglobin concentration (CHr). Finally, CRP was obtained at baseline and at routine dialysis follow-up visits. The dialysis efficacy was calculated using the urea reduction ratio (URR).
Adverse events were collected from two sources: the formal electronic (‘yellow card’) reporting system within each trust accessible through and; a confirmation call to all dialysis units to corroborate this information and discover potential reactions that were not reported previously. Adverse events were categorised as: hypersensitivity reactions, cardiovascular reactions and others. Hypersensitivity reactions were defined as mild if the patient had symptoms that required no interventions, moderate if the patient required antihistamine or steroids and/or oxygen, severe if the patient required volume replacement, adrenaline or hospital admission.
Statistical analysis
The continuous data and descriptive data were expressed as mean SD or median with interquartile values depending on normality. Categorical data were expressed as number or percentage. For haemoglobin, ferritin and weekly dose of iron and EPO, baseline values of time zero (T0) are expressed as a mean of values over the preceding 6 months (T=-6 to T=-1 inclusive). Between-period comparisons were performed using one-way, repeated measures ANOVA for normally distributed data and the Friedman’s test for non-normally distributed data. Adjustment for multiple comparisons was made using Dunnett’s or Dunn’s tests for normally and non-normally distributed data respectively, using T0 as the control column. Sensitivity analyses were performed using the same methods described above but utilising the last observation carried forward (LOCF) technique to account for missing data. All statistical analyses were performed using GraphPad Prism software v7.0e. A p value of less than 0.05 was considered statistically significant.