Factors governing the erythropoietic response to intravenous iron infusion in patients with chronic kidney disease; a retrospective cohort study

doi:10.21203/rs.3.rs-2584173/v1

Download PDF

Research Article

Factors governing the erythropoietic response to intravenous iron infusion in patients with chronic kidney disease; a retrospective cohort study

https://doi.org/10.21203/rs.3.rs-2584173/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background

This study aimed to evaluate the erythropoietic response to parenteral iron in iron-deficient anaemic patients with normal and impaired kidney function and determine the factors that affect response.

Method

Recipients of parenteral iron from the CKD and pre-operative clinics of a tertiary kidney centre were evaluated. No patients were receiving erythropoiesis stimulating agents. The key outcome was the change in haemoglobin(∆Hb) defined as the difference between the pre-infusion and post-infusion haemoglobin(Hb) measured within 1- 3 months of iron infusion. Propensity-score matching, and multivariate linear regression adjusted for baseline demographics, comorbidities and iron indices were used to assess response to parenteral iron and the impact of kidney impairment versus the nature of iron deficiency (absolute iron deficiency [AID] vs functional iron deficiency [FID]) on response.

Results

732 subjects (834 infusions) were included (mean age 66±17years, 56% females and 87% White). Median haemoglobin rise was significantly lower with lower kidney function (CKD0-2, 13[IQR, 2-23]g/l; CKD3-5; 7[IQR,1-15]g/L, P<0.001). When groups with different degrees of renal impairment were propensity-matched according to whether iron deficiency was due to AID or FID, the severity of renal impairment was no longer a factor affecting haemoglobin response (unmatched [14.2 vs 8.3]g/l, matched[14.1vs13.6] g/L). However, comparing AID and FID after propensity matching for the degree of renal impairment showed a significantly lower response in those with FID (unmatched AID vs FID 16.0 vs 6.6 g/L, matched 14.2 vs 6.6g/L).

The multivariate analysis showed a positive association between ∆Hb and body mass index, total iron dose and glomerular filtration rate, whereas a negative association was observed between haemoglobin response and Black ethnicity, Asian ethnicity, diabetes, baseline Hb, mean corpuscular volume, and Functional iron deficiency

Conclusion

The nature of iron deficiency rather than the severity of CKD has a stronger impact on haemoglobin response with an attenuated response seen in functional iron deficiency irrespective of the degree of renal impairment

Iron deficiency anaemia (IDA) can be the result of reduced iron intake, reduced intestinal absorption or chronic blood loss resulting in absolute iron deficiency(AID). In addition to this, a state of functional iron deficiency (FID) occurs when there is inadequate iron available for erythropoiesis in the presence of adequate iron stores [1] This state is commonly associated with inflammation[2, 3].

In Chronic kidney disease (CKD), apart from being an independent risk factor for death, IDA is associated with reduced quality of life, cardiovascular disease (CVD), hospitalizations, and cognitive impairment[4–6]. Pre-operative patients also suffer from IDA which has been associated with higher peri- and postoperative morbidity and mortality[7, 8].

Iron supplementation in IDA reduces the need for blood transfusion, length of hospital stays and the costs to healthcare providers[9]. In CKD, it reduces the requirement for erythropoietin stimulating agents(ESA) [10]. In the management of IDA, it is important to differentiate between AID and FID as pathophysiology and response to therapy differ between these two conditions [11, 12]. In the normal population, AID is defined as serum ferritin < 30µg/l [13] and in CKD_3 − 5 as ferritin < 100 µg/l with either transferrin saturation (TSAT) < 20%; percentage of hypochromic red blood cells > 6% or reticulocyte Hb content (CHr) < 29pg[14–16]. FID is defined by normal or raised ferritin concentrations but low transferrin saturation.

Current management guidelines recommend parenteral iron for anemic patients with CKD₅ receiving hemodialysis or non-dialysis CKD_3 − 5 patients with either severe iron deficiency, inadequate response or poor tolerance to oral iron supplementation[17, 18]. In the perioperative setting, parenteral iron is recommended for patients intolerant of or non-adherent to oral iron, or when the preoperative period is too short for oral iron to effectively correct the pre-operative anaemia[9, 19].

Currently, none of the available tests has adequate discriminatory power to identify patients who will have a good response to parenteral iron[20].

The pre-operative anesthetic department in our institution had observed that surgical patients with lower estimated glomerular filtration rate(eGFR) tended to have a poorer response to parenteral iron than those with normal eGFR. Determining the validity of and explanation for this observation underpinned the rationale for the current study. This retrospective observational study was therefore conducted to evaluate the response to parenteral iron in patients with differing degrees of kidney function and to identify reasons for differential response.

Study population

The analysis included subjects who received parenteral iron ferric derisomaltose (Monofer®; Pharmacosmos A/S, Holbaek, Denmark) between 2017 and 2019 for the treatment of anaemia in the CKD and pre-operative clinics of a tertiary nephrology centre in Northwest England. Subjects were excluded if they were receiving erythropoietin stimulating agent (ESA) or had no recorded post-infusion haemoglobin (Hb) within 1–3 months of the iron infusion.

The study procedure was compliant with the principles of the declaration of Helsinki and approved by the local institutional review board, of the Northern Care Alliance NHS Group (Ref: S21HIP50). As this was a retrospective observational study using routinely collected and fully anonymized data, the need for individual patient consent was not considered necessary.

Data collection

Anonymized data concerning demographic characteristics, comorbidities, Hb, iron parameters and red-cell indices were recorded. Subjects were categorized according to the nature of their iron deficiency into those with functional iron deficiency (FID) and those with absolute iron deficiency (AID) and stratified based on the severity of CKD (KDIGO CKD-stage).

According to the World Health Organization anaemia was defined as Hb < 130 g/L in Men and Hb < 120 g/L in women[21].

The dosing protocol for pre-operative and CKD parenteral iron therapy is shown in Table 1

Table 1

The pre-operative and CKD parenteral iron infusion protocol
Pre-Operative Regime			CKD Regime
Pre-infusion Hb	Parenteral iron dose by weight^α		Pre-infusion Hb	Parenteral iron dose by weight ^β
Pre-infusion Hb	< 50kg	> 50kg	Pre-infusion Hb	< 50kg	50kg-69.9kg	70kg-100kg	> 100kg
Hb > 80g/L	500mg	500mg	Hb > 100g/L	20mg/kg	1000mg	1500mg	2000mg
Hb < 80g/L	20mg/kg	1000mg	Hb < 100g/L	20mg/kg	1500mg ^µ	2000mg ^µ	2000mg
^α,Doses up to 500mg are infused over 15minutes while doses exceeding 500mg are infused over 30 minutes;^βDoses up to 1000mg are infused over 15minutes while doses exceeding 1000mg are infused over 30 minutes;^{, µ}administered in 2 divided doses, maximum individual dose of 20 mg/kg divided into an initial dose of 1000mg followed by the remaining dose of 500mg-1000mg given 1 week later

Outcomes and predictor variables

The primary outcome was the change in haemoglobin (∆Hb) at 1–3 months post parenteral iron, defined as the difference between the pre-infusion Hb and the post-infusion Hb measured at this time. This was first compared among the various stages of CKD, then between subjects with FID vs AID. A propensity-score matching (PSM) method was used to balance out the impact of the degree of kidney impairment and other baseline covariates that may affect haemoglobin response enabling the analysis of the effect of iron deficiency type on response.

Two different definitions of FID and AID were used in this study. This was because the FID definition differs between CKD and non-CKD populations due to the effects of CKD and chronic inflammation on iron metabolism [22]. Since our cohort consisted of both CKD and non-CKD patients, we included both definitions in our analysis.

According to the international consensus statement on the perioperative management of anaemia and iron deficiency, AID was defined as ferritin < 30µg/l whereas FID was defined as ferritin ≥ 30µg/l and TSAT < 20%. In CKD_3 − 5 population, AID was defined as ferritin ≤ 100µg/l and TSAT < 20%. whereas FID was defined as ferritin > 100µg/l with TSAT < 20%[22].

The total dose of iron infused (TDI) for each infusion episode was also recorded. In patients with less severe anaemia, their TDI was administered in a single infusion. In contrast, those with more severe anaemia requiring higher TDI received their infusion in two divided doses one week apart (Table 1).

The comorbidities recorded included diabetes and CVD defined as a composite of ischemic heart disease, myocardial infarction, heart failure, cerebrovascular disease and peripheral vascular disease. These factors were chosen because of their influence on iron metabolism[23–25]. Kidney function was assessed using the modification of diet in renal disease (MDRD) eGFR formula. Subjects were categorized into normal eGFR (eGFR ≥ 60 ml/min/1.73m²), CKD₃ (eGFR 30-59ml/min/1.73m²), CKD₄ (eGFR 15-29ml/min/1.73m²) and ESKD (eGFR < 15ml/min/1.73m² including those on renal replacement). Iron infusion-associated reactions were categorized into mild, moderate, and severe reactions:

Mild reactions; defined as localized infusion site reactions that did not result in discontinuation of infusions
Moderate reactions; defined as systemic symptoms leading to discontinuation of infusions but not to hospitalization
A severe reaction; defined as systemic symptoms requiring hospitalization for ≥ 24 hours.

Statistical analysis

Descriptive statistics of categorical data were summarized as frequencies and percentages. Continuous variables were summarized as mean ± standard deviation (SD), or median and interquartile range (IQR). Inter-group comparisons were made using a one-way ANOVA or the Kruskal-Wallis's rank test. The Chi-square or Fisher’s exact test was used to compare categorical variables.

Propensity score matching analysis

To evaluate the impact of AID and FID and the effect of eGFR on response to parenteral iron, we conducted separate PSM analyses to provide clearer insights into the effects of kidney function, FID and AID on the response to parenteral iron. In the first PSM analyses, separate propensity scores (PS) were generated for matching subjects with eGFR ≥ 60 vs eGFR < 60; eGFR ≥ 60 vs eGFR < 30; eGFR = 30–59 vs eGFR < 30 (ml/min/1.73m²). PS were calculated using the following variables: age, gender, diabetes, baseline Hb, baseline ferritin, baseline TSAT and presence or absence of FID. In the second PSM analysis, separate PS models were used to assess the effect of the two forms of IDA (FID vs AID). To ensure subjects met the definitions of FID, those with both normal ferritin and TSAT (> 20%) were excluded from the inferential analysis. The PS were estimated using the following variables: age, gender, diabetes, baseline Hb, and baseline eGFR. To assess the effect of kidney impairment on ∆Hb, the type of iron deficiency was included in the covariates used to generate PS. Conversely, to evaluate the impact of the nature of iron deficiency on ∆Hb, the degree of kidney impairment (baseline eGFR) was one of the covariates used to generate PS. By doing this we hoped to minimise the confounding effect of kidney impairment while evaluating the effect of FID on response and vice versa.

We conducted radius matching on the generated PS with a calliper of 0.05 (using 0.25 standard deviations of the logit of PS)[26, 27]. All PSM analyses were conducted with the psmatch2 Stata user-written function for conducting PSM analysis. The mean bias, Rubin’s B and Rubin’s R were used to ensure covariate balance after matching [28]. A Rubin’s B < 25% and Rubin’s R between 0.5-2 are considered to indicate a sufficiently balanced matching [28]. In all the PSM analyses, Rubin's B and Rubin's R suggested high-quality matches (Table 2).

Table 2

*Propensity score matching analysis of the impact eGFR and type of IDA on* ∆Hb *following parenteral iron.*
Treated vs Control (n)	eGFR ≥ 60 (n = 178) vs eGFR < 60(n = 544)				eGFR ≥ 60 (n = 151) vs eGFR < 30(n = 387)				eGFR 30–59(n = 153) vs eGFR < 30 (n = 379)				AID^α(n = 230) vs FID^π (n = 489)				AID^β(n = 482) vs FID^γ (n = 285)
Matched Covariates	Age, sex, diabetes, CVD, baseline Hb, ferritin, TSAT, FID/AID				Age, sex, diabetes, CVD, baseline Hb, ferritin, TSAT, FID/AID				Age, sex, diabetes, CVD, baseline Hb, ferritin, TSAT, FID/AID				Age, sex, diabetes, CVD, baseline Hb, eGFR				Age, sex, diabetes, CVD, baseline Hb, eGFR
Before & After PSM	U		M		U		M		U		M		U		M		U		M
Mean Bias %	51.3		4.0		57.2		4.9		21.7		1.9		31.8		6.2		27.3		4.4
Rubin’s B	147.6		15.7		170		17.5		66.3		6.8		89.7		16.2		66.8		10.8
Rubin’s R	0.89		0.63		0.98		0.98		1.12		1.14		0.59		1.21		0.61		1.04
P-value	< 0.001		0.947		< 0.001		2.33		< 0.001		1.00		< 0.001		0.270		< 0.001		0.894
∆Hb (g/L) (T:C)	T	C	T	C	T	C	T	C	T	C	T	C	T	C	T	C	T	C	T	C
∆Hb (g/L) (T:C)	14.2	8.3	14.1	13.6	14.1	7.7	12.5	12.2	9.8	7.6	9.4	9.6	16.0	6.6	14.2	6.6	11.7	6.6	10.3	6.6
Radius matching within a caliper of 0.25 x SD of the propensity scores (0.05). Rubin’s B; absolute standardized difference of the means of the linear index of the propensity score in the treated and (matched) control group; Rubin's R, the ratio of treated to (matched) control variances of the propensity score index; A Rubin’s B < 25% and a Rubin’s R between 0.5 and 2 after PSM indicates a sufficiently balanced matching; ∆Hb, change in Hb; AID, Absolute iron deficiency; α, Absolute iron deficiency defined as ferritin < 30µg/l; β, absolute iron deficiency defined as ferritin ≤ 100µg/l & TSAT < 20%; C, control; CVD, cardiovascular disease; eGFR, estimated glomerular filtration rate(ml/min/1.73m2); FID, functional iron deficiency; π, functional iron deficiency defined as ferritin ≥ 30µg/l & TSAT < 20%; γ, functional iron deficiency defined as ferritin > 100µg/l & TSAT < 20%; M, matched; T, treated; TSAT, transferrin saturation(%); U, unmatched.

Multivariate regression analysis

The wider predictors of response to parenteral iron were assessed using multivariate linear regression. Predictor covariates were chosen based on previous knowledge of their effect on anaemia, iron metabolism and erythropoiesis. They included age, gender, ethnicity, body mass index (BMI), diabetes, CVD, TDI, baseline eGFR, baseline mean corpuscular volume (MCV) and the nature of iron deficiency (FID vs AID). To avoid the potential confounding of HB response by inadequate parenteral iron dose, we calculated the total iron deficit for each subject using the Ganzoni formula [29]. We then evaluated the ∆Hb in patients who received a parenteral iron dose that was equivalent to their calculated iron deficit (within 200mg of the calculated deficit). Statistical analysis was performed with Stata software version-14 licensed to the University of Manchester.

A total of 1341 Iron infusion episodes in 952 subjects were evaluated. After the exclusion criteria were applied, a total of 834 Iron infusions in 732 Subjects were included in the final analysis. Mean age was 66 ± 17years (range 18–98), 56% were female, 87% were white, 41% were diabetic, and 36% had CVD (Fig. 1). 570 (78%) patients were from the CKD iron clinics and 162 (22%) came from the pre-operative iron clinics. The mean ± SD total dose per infusion was 18.2 ± 3.65mg/kg body weight. 88% received their total dose as a single infusion while 12% required two infusions one week apart. The baseline characteristics are presented in Table 3. The majority of patients had CKD₄(n = 343; 47%), followed by CKD_0 − 2 (n = 180; 25%). Very few patients (n = 6) were receiving renal replacement therapy. Baseline Hb (median, 107g/l; IQR, 98–115) was similar across the CKD categories, but baseline ferritin (median, 67µg/l; IQR, 24–179), was higher with worse CKD-stage.

Table 3

Baseline characteristics of subjects stratified by CKD stage
Characteristics	TOTAL	Normal eGFR (eGFR ≥ 60)	CKD3a (eGFR 45–59)	CKD3b (eGFR 30–45)	CKD4 (eGFR 15–29)	CKD5 (eGFR < 15)	RRT
Proportion	100%	23%	8%	14%	48%	7%	0.70%
Number of infusions	834	191	64	118	398	57	6
Number of Subjects	732	180	55	101	343	47	6
Age, (yr) (mean ± SD)	66 ± 17	60 ± 18.0	63 ± 17	69 ± 16	70 ± 16	67 ± 13	53 ± 9
Female, N(%)	465(56)	109(57)	38(59)	65(55)	230(58)	21(36)	2(33)
BMI, Kg/M2, (mean ± SD)	29 ± 6.6	28 ± 6.6	27 ± 5.3	29 ± 6.1	30 ± 6.9	28 ± 6.7	29 ± 6.7
Caucasian N(%)	722(87)	152(80)	58(91)	112(95)	347(87)	47(82))	6(100)
Diabetes N(%)	345(57)	46(24)	26(41)	63(53)	179(45)	30(52)	1(17)
CVD (Stroke IHD, HF, PVD) N(%)	299(36)	28(15)	29(45)	58(49)	159(40)	23(40)	2(33)
∆Hb (g/L), Median(IQR)	8(1–17)	13(2–24)	8(4–15)	7.5(1–17)	8(1–14)	4(-3-10)	6(1–13)
Baseline Ferritin (µg/l), Median(IQR)	67(24–179)	18(8–50)	59(26–154)	64(25–172)	96(36–202)	189(102–428)	195(58–336)
Change in ferritin (µg/l), Median(IQR)	359(170–475)	204(109–358)	234(110–374)	315(208–500)	341(191–498)	378(221–534)	477(269–829)
Baseline TSAT, Median(IQR)	12(8–16)	7(5–10)	11(7.5–4.5)	12(9–14)	14(11–17)	16(12–19)	17.5(15–19)
Change in TSAT, Median(IQR)	10(5–17)	12(5–18)	13(8–18)	10(6–17)	10(5–16)	8(5–15)	13(7–19)
MCV (FL), Median(IQR)	88.7(83.9–92.8)	82.2(76.7–87.6)	89(84.8–92.7)	89(84.6–94.4)	90.3(86.2–93.6)	90.5(86.8–94.4)	91(89.7–94.6)
MCH (pg), Median(IQR)	28.8(27.0-30.4)	26.4(23.6–28.2)	28.7(27.0-30.1)	28.8(27.3–30.7)	29.3(27.9–30.6)	29.7(28.4–30.6)	30.6(28.6–31.7)

Values are represented as, N (%), mean ± SD or median (IQR) unless otherwise stated; ∆Hb, Change in haemoglobin; AID, absolute iron deficiency defined as defined as Ferritin < 30; CKD, chronic kidney disease; CVD, cardiovascular disease defined as a composite of stroke, ischemic heart disease, heart failure and peripheral vascular disease; FID, functional iron deficiency defined as Ferritin ≥ 30 and TSAT < 20% BMI, Body mass index; Hb, haemoglobin; HF, heart failure; IHD, ischemic heart disease; MCV, mean cell volume (Ref-range = 84–105 FL); MCH, mean corpuscular haemoglobin (Ref-range 27–32 pg); PVD, peripheral vascular disease; TSAT, transferrin saturation; yr, year.

At baseline, using the classical definitions of AID and FID, a total of 252 (33%) infusions were administered to patients with AID (ferritin < 30µg/l) whereas 496 (33%) infusions were given to patients with FID (ferritin ≥ 30µg/l with TSAT < 20%) (Table 4). When stratified using the CKD definitions of FID and AID, those with AID (ferritin ≤ 100µg/l with TSAT < 20%) received 444(60%) infusions whereas those with FID (ferritin > 100µg/l with TSAT < 20%) received 290(40%) infusions (We excluded patients with ferritin > 100 but with TSAT > 20 from this analysis) (see additional file 1).

Response to parenteral iron

The median(IQR) ∆Hb in the cohort was 8g/l (1–17). There was a negative correlation between ∆Hb and baseline ferritin, haemoglobin, TSAT, and MCV (p < 0.001) (Figs. 2a, 2c-2e) whereas the ∆Hb was positively correlated to baseline eGFR(p < 0.001) (Fig. 2b) as had been noted by the anesthetic team. Patients with the classical definition of FID (ferritin ≥ 30µg/l and TSAT < 20%) had a significantly lower response compared to those with AID (∆Hb = 6g/L [IQR 0–13] vs 16g/L [IQR 6–23])p < 0.001 (Table 4.) A similar outcome was observed with the CKD definitions of FID and AID (∆Hb = 7g/L [1–13] vs 11g/l [2–20])p < 0.001(Additional file 1).

Table 4

Baseline haemoglobin and iron indices stratified by CKD stage and nature of iron deficiency (FID defined as Ferritin ≥ 30 and TSAT < 20%)
Variables	TOTAL		Normal –CKD₂ (eGFR ≥ 60)		CKD_3a (eGFR 45–59)		CKD_3b (eGFR 30–45)		CKD₄ (eGFR 15–29)		CKD₅ (eGFR < 15)
	FID	AID	FID	AID	FID	AID	FID	AID	FID	AID	FID	AID
Age, years	68 ± 16	64 ± 18	64 ± 18	57 ± 18	60 ± 17	70 ± 17	70 ± 15	67 ± 17	70 ± 15	71 ± 18	67 ± 14	81 ± 2
BMI, Kg/M²	29 ± 7	29 ± 7	27 ± 6	29 ± 7	28 ± 6	26 ± 4	30 ± 7	28 ± 6	30 ± 7	29 ± 7	28 ± 7	22
Baseline Hb(g/L)	109 ± 14	107 ± 15	112 ± 13	108 ± 15	117 ± 16	106 ± 22	107 ± 15	105 ± 14	107 ± 14	108 ± 15	105 ± 12	111 ± 9
Baseline Ferritin µg/l	122(59–267)	14(8–20)	81(46–261)	10(6–18)	114(59–180)	16(8–25)	135(68–246)	15(11–24)	123(60–255)	18(12–24)	206(90–410)	24(22–27)
Baseline TSAT (%)	13(10–15)	7(5–11)	9(6–12)	6(4–8)	11(8–14)	5(4–10)	12(11–14)	9(6–11)	14(11–16)	10(7–14)	15(12–17)	11(9–13)
Baseline MCV	90.2(85–94)	84(78–89)	85.3(80–91)	80(75–86)	89(86–93)	86(80–87)	91(85–96)	86(84–91)	91(86–94)	88(84–91)	91(87–96)	83(82–84)
Baseline MCH	29(28–31)	27(25–29)	27(24–30)	26(23–28)	29(28–30)	27(24–28)	29(28–31)	28(26–30)	29(28–31)	28(27–30)	30(28–31)	28
Iron dose (mg/kg)	18 ± 4	18 ± 3	18 ± 3	18 ± 2	19 ± 4	18 ± 3	18 ± 4	18 ± 4	18 ± 3	18 ± 3	18 ± 4	19 ± 3
∆Hb	6(0–13)	16(6–23)	6(-1-21)	17(5–27)	7(3–14)	12(5–20)	5(-2-14)	16(9–20)	7(1–13)	15(6–22)	5(-4-9)	-1.5(-5- 2)
Infusions (subjects)	496(476)	252(250)	65(63)	121(120)	39(37)	18(18)	69(65)	37(36)	274(264)	74(74)	44(42)	2(2)
Values are represented as, N (%), mean ± SD or median (IQR) unless otherwise stated; ∆Hb, Change in haemoglobin; AID, absolute iron deficiency defined as defined as Ferritin < 30; CKD, chronic kidney disease; CVD, cardiovascular disease; FID, functional iron deficiency defined as Ferritin ≥ 30 and TSAT < 20% BMI, Body mass index; Hb, haemoglobin; HF, heart failure; IHD, ischemic heart disease; MCV, mean cell volume (Ref-range = 84–105 FL); MCH, mean corpuscular haemoglobin (Ref-range 27–32 pg); TSAT, transferrin saturation

The PSM analysis was conducted in two phases (Fig. 3). First, we compared responses among patients with different stages of renal impairment while adjusting for the presence or absence of FID and other baseline parameters using PSM. Then we compared the response between patients with FID vs AID while the degree of renal impairment was adjusted by creating a propensity match for eGFR and other baseline variables between the groups. In the first PSM analysis, 178 subjects with eGFR ≥ 60 ml/min/1.73m² were matched to 544 subjects with eGFR < 60 ml/min/1.73m². The groups were matched by age, gender, diabetes, baseline ferritin, baseline haemoglobin, baseline TSAT and FID.

Before PSM, the mean ∆Hb was significantly higher in those with eGFR ≥ 60 ml/min/1.73m² compared to those with eGFR < 60 ml/min/1.73m² (14.2g/L vs 8.3g/L). But after PSM, the difference in ∆Hb between the two groups was no longer apparent (14.0g/L vs 13.6g/L). A similar result was observed when subjects with eGFR ≥ 60ml/min(n = 150) were compared to those with eGFR < 30ml/min(n = 387) (∆Hb, before-PSM = 14.1g/L vs 7.7g/L after-PSM = 12.5g/L vs 12.2g/L). The result was no different in subjects with eGFR 30-59mil/min(n = 153) vs eGFR < 30ml/min (n = 379) (∆Hb before-PSM = 9.8g/L vs 7.6g/L after-PSM = 9.4g/l vs 9.6g/L).

The second PSM analysis compared the ∆Hb in subjects with FID to those with AID while they were matched by age, gender, diabetes, baseline Hb and eGFR.

PSM analyses were conducted with two different definitions of FID, the traditional definition and the CKD-related definition. With the traditional definition (Ferritin > 30µg/l), the average rise in haemoglobin before PSM was almost 10g/l higher in those with AID(n = 238) compared to subjects with FID(n = 491) (16.3g/L vs 6.6g/L). This difference persisted after matching (14.2g/L vs 6.6g/L). A similar pattern was noted with the CKD definition of FID (ferritin > 100µg/l and TSAT < 20%) (∆Hb before-PSM = 11.7g/L vs 6.6g/L; after-PSM = 10.3g/L vs 6.6g/L).

To eliminate the impact of suboptimal iron dosing on response, we estimated the proportion of subjects who received adequate total iron dose by calculating the iron deficit for each patient using the ganzoni formula. 86%(716) of the TDI were judged to have been optimally dosed while 14%(118) were suboptimal. We then evaluated the impact of eGFR categories and nature of iron deficiency on ∆Hb in subjects judged to have received optimal iron dose only (Fig. 4). The result was similar to the main analysis above and showed that irrespective of the degree of renal impairment, FID was associated with a significantly lower response even with optimal iron repletion.

Multivariate analysis

To determine the predictors of response to parenteral iron we conducted a multivariate regression analysis, as explained in the methods section (Fig. 5). The factors associated with a positive ∆Hb were BMI, coefficient of ∆Hb (∆HBβ) = 1.51(95%CI; 0.82–2.21)p < 0.001; total iron dose ∆Hbβ = 2.36 (95%CI;1.11–3.62)p < 0.001 per 5mg/Kg dose increase; baseline eGFR, ∆Hbβ = 0.45 (95%CI: 0.07–0.83)p = 0.02 per 10ml/min/1.73m² increase whereas Black ethnicity, ∆Hbβ = -7.21(95%CI; -12.43 - -1.99)p = 0.007; Asian ethnicity ∆HBβ= -3.01 (95%CI; -5.97–0.05)p = 0.04 (White ethnicity as reference); diabetes, ∆HBβ= -2.86 (95%CI; -4.68 - -1.05)p = 0.002; baseline Hb, ∆Hbβ = -2.96 (95%CI; -3.6- -2.3) p < 0.001; baseline MCV, ∆Hbβ= -0.45g/l (95%CI; -0.67–0.23)p < 0.001; and FID ∆Hbβ= -6.15g/l (95%CI; -8.21 - -4.09)p < 0.001(compared to AID) were associated with a negative ∆Hb.

Incidence of adverse reaction to parenteral iron.

Infusion-associated reactions occurred in 7 of the 834 infusion episodes (0.8%) (Table 5). Of these, one (0.1%) was a serious allergy (hemodynamic instability requiring hospitalisation but recovered). One was mild localised pruritus while the other five were ‘Fishbane’ reactions [30, 31], manifesting as flushing and mild chest or shoulder pain and which promptly resolved on cessation of infusion. One patient known to have angina had an episode of chest pain. This was judged unrelated to the infusion.

Table 5

Safety and tolerability of parenteral ferric derisomaltose
Infusion associated reactions	Severity	Infusion discontinued
Hypotension, Nausea, and erythema	severe	Yes
Itching	Mild	No
Flushing, back pain	Moderate	Yes
Nausea	Moderate	Yes
Chest tightness	Moderate	Yes
Dizziness and tingling	Moderate	Yes
Shortness of breath, Flushing, and paresthesia	Moderate	yes
Chest pain in a known angina	Moderate	Yes
Serious allergy 0.1% (1/834) labile/Fish bane 0.7%(6/834); Mild, local reaction infusion not discontinued; moderate, systemic symptoms but patient not hospitalized; Severe, systemic symptoms patient hospitalized.

Although anaemia in CKD patients has been attributed to the effects of erythropoietin deficiency and uremia, both of which are dependent on the severity of CKD, the impact of FID alone on Hb response has not been studied. In this study, we used PSM analyses to isolate the effect of FID from that of kidney impairment on Hb response. First, we compared responses among patients with different CKD stages while adjusting for FID. Then we compared the responses between patients with FID vs AID while adjusting for the degree of renal impairment. Furthermore, combining data from pre-operative and CKD cohorts allowed us to assess Hb responses across a wider range of patients representing all levels of kidney function.

Our study showed a mean ∆Hb of 8g/L after a mean iron infusion dose of 18mg/kg. This was significantly higher than that reported by Sivakumar et al (3.5-4.5g/l) [32] but lower than the 14-16g/l reported by the authors of the randomized controlled trial of ferumoxytol vs. ferriccarboxymaltose for the treatment of IDA(FIRM) trial [33]. The differences are due to the nature of the population studied. Whereas our study cohort included both CKD and non-CKD patients, Sivakumar et al studied only CKD and peritoneal dialysis patients, while the FIRM study recruited subjects with IDA of any aetiology, the majority of which had gastrointestinal or abnormal uterine bleeding while excluding subjects with FID or dialysis-dependent CKD[33].

In our study, subjects with normal or well-preserved eGFR (CKD_0 − 2), showed a better response than those with CKD_3 − 5 (∆Hb = 14.01g/L vs 8.3g/L) which was in keeping with the observations of the pre-surgical anaemia clinic team that formed the rationale for this study. However, this difference was no longer apparent after PSM corrected for the effect of FID. Crucially, this pattern held true across the various levels of CKD (Figs. 3a-3c). On the contrary, when subjects were grouped based on FID vs AID and matched on their level of kidney function, there was a significantly better response in those with AID compared to those with FID after matching (Figs. 3d-3e). Taken together, these findings suggest that response to parenteral iron is far more dependent on the type of IDA (FID vs AID) than on the severity of kidney impairment. The strength of this finding was further emphasized after excluding subjects judged to have received a sub-optimal iron dose based on the Ganzoni equation (Fig. 4). It was clear that subjects with severe kidney impairment responded as well as their counterparts with mild or no kidney impairment if they had a similar type of iron deficiency, be it AID or FID.

FID, due to hepcidin upregulation in chronic inflammatory states (including CKD) is a key factor reducing iron availability for erythropoiesis and inducing erythropoietin resistance [30]. This is probably the reason for the poorer responses in subjects with FID than those with AID irrespective of their kidney function.

The pre-operative intravenous iron to treat anaemia before major abdominal surgery (PREVENTT) trial found that parenteral iron administered 10–42 days before elective major abdominal surgery was not superior to placebo in preventing death or reducing transfusions in anemic patients[34]. However, it reported an improved secondary outcome of reduced re-admission in the iron group [34]. The study had limitations, not least that the type of anaemia and the nature of iron deficiency was not defined. Our study suggests that consideration of the nature of a patient’s iron deficiency will identify those who are likely to achieve the desired haemoglobin response after iron infusion.

Regarding the wider risk predictors of Hb response to parenteral iron, Black and Asian ethnicity, lower BMI, diabetes, lower total iron dose, lower eGFR, higher baseline MCV and FID were independently associated with poor response (Fig. 5).

The relationship between BMI and Hb response may be because subjects with higher BMI received a higher TDI. We believe this is related to parenteral iron dose being calculated based on actual body weight (including adipose mass) rather than lean body weight coupled with the fact that parenteral iron preparations are highly polar, thus their volume of distribution is independent of adipose mass. Consequently, subjects with higher BMI would have received higher doses than those with lower BMI [35]. This is supported by the fact that the rise in ferritin post-infusion increased as BMI increased from < 25kg/m² through 30kg/m² to > 30kg/m²; p = 0.0007.

This study also highlighted the safety of modern parenteral iron preparations. As shown in recent studies, available preparations have been associated with a much lower risk of serious adverse events(SAE) compared to more historic iv iron preparations[36, 37]. Sivakumar et al reported an adverse reaction rate of 0.54% with ferric derisomaltose infusion[32]. A large meta-analysis of 103 studies, including 10,390 patients treated with IV iron compared to 4044 patients treated with oral iron, 1329 treated with no iron, 3335 treated with placebo and 155 who received IM iron, reported an SAE rate of 0.3% with iv iron.

Clinical implications

Our findings carry important implications for the management of anemic patients. First, patients with preoperative anaemia and FID are less likely to respond adequately to parenteral iron. These patients should be investigated and treated for reversible causes of FID. Conversely, patients with AID can be anticipated to have a better ∆Hb after parenteral iron, irrespective of their kidney function. Therefore, it is key that the causes of anaemia and the nature of iron deficiency are determined. This should be incorporated into the pre and peri-operative pathways allowing time for investigation, diagnosis and treatment. When considering CKD patients, the degree of kidney impairment has less impact on response than the nature of iron deficiency. However, it should be remembered that although parenteral iron may have a reduced effect on ∆Hb in patients with FID its effect on other parameters such as cardiac function, and exercise tolerance is not addressed in this paper and will require further exploration with prospective studies.

Strengths and limitations of the study

The strength of this study stems from the relatively large number of patients evaluated and the use of PSM to enable the study of the effect of FID on ∆Hb independent of kidney function and vice-versa. However, some limitations are worth considering. First, the retrospective observational nature of the study implied that the risk of sampling bias could not be avoided. However, this was minimized with the PSM methodology. Second, data were not collected on intra-operative blood loss in the pre-operative patients. These losses would impact treatment rise in haemoglobin. Thirdly, it is possible that several pre-operative anaemia patients were only cured of the underlying causes of their anaemia following surgery, thus, a better Hb response was not possible due to continued blood loss pre-surgery.

AID and FID are better predictors of response to parenteral iron than kidney function. The total parenteral iron dose is positively correlated to haemoglobin response with the best responses seen in AID patients with lower ferritin and haemoglobin. The presence of FID rather than the level of kidney function per se attenuates the response to parenteral iron. Parenteral ferric derisomaltose is safe and associated with minimal risk of serious adverse reactions, and therefore adequate doses should be given to anemic patients to increase their chance of optimal response.

AID: Absolute iron deficiency

BMI: Body mass index

CKD: Chronic kidney disease

CVD: Cardiovascular disease

eGFR: Estimated glomerular filtration rate

ESA: Erythropoietin stimulating agent

ESKD: End stage kidney disease

FID: Functional iron deficiency

Hb: Haemoglobin

IDA: Iron deficiency anaemia

IQR: Interquartile range

MCV: Mean corpuscular volume

MDRD: Modification of diet in renal disease

PS: Propensity score

PSM: Propensity score matching

SAE: Serious adverse events

TDI: Total dose infusion

TSAT: Transferrin saturation

∆Hb: Change in haemoglobin

Availability of data and materials

The datasets used and/or analyzed during the study are available from the corresponding author on reasonable request.

Ethics approval and consent to participate.

The study was reviewed and approved by the ethics committee of the Northern Care Alliance research and innovation institutional review board (Ref: S21HIP50) and carried out in accordance with the Declaration of Helsinki. As this was a retrospective observational study using routinely collected and fully anonymized data, the need for individual patient consent was not considered necessary and was therefore waived by the Northern Care Alliance research and innovation institutional review board.

Consent for publication

Not applicable.

Acknowledgements

We are grateful to Pharmacosmos for providing an unrestricted grant that helped facilitate the data collection necessary for the analysis. Pharmacosmos was not involved in the design of the study, collection, analysis, and interpretation of data and in writing the manuscript.

The data underlying this article will be shared on reasonable request to the corresponding author

Competing interest

C. A. Chukwudeclares no competing interest.

H. Gilbody declares no competing interest

O. Wickens declares no competing interest

C. Carroll declares no competing interest

S. Bhandari declares no competing interest

P.A Kalra reports fees for lecturing and advisory board attendance for Astra Zeneca, Napp, Mundi Pharma, Boehringer Ingelheim, Novonordisk, Pfizer, Pharmacosmos, Vifor; Consultancy work with Bayer, Astellas, Otsuka and Unicyte; Institutional research funding from Vifor, GSK, Bergen Bio and Evotec

Funding

This research was part funded by an unrestricted grant from Pharmacosmos.

Authors' contributions

CAC: collected and analysed data, wrote, reviewed and edited the manuscript

HG and OW collected parts of the iron infusion data

PAK and CC Conceptualised, supervised the study and edited the manuscript

SB Reviewed and edited the final manuscript

Gaweda AE. Markers of iron status in chronic kidney disease. Hemodialysis International. 2017 Jul;21:S21–7.
Mercadel L, Metzger M, Haymann JP, Thervet E, Boffa JJ, Flamant M, et al. The relation of hepcidin to iron disorders, inflammation and hemoglobin in chronic kidney disease. PLoS ONE. 2014 Jun;9(6).
Ganz T. Hepcidin, a key regulator of iron metabolism and mediator of anemia of inflammation. Blood. 2003 Aug;102(3):783–8.
Clevenger B, Richards T. Pre-operative anaemia. Anaesthesia. 2015 Jan;70. DOI: 10.1111/ANAE.12918
Nakhoul G, Simon JF. Anemia of chronic kidney disease: Treat it, but not too aggressively. Cleveland Clinic Journal of Medicine. 2016 Aug;83(8):613–24.
Vlagopoulos PT, Tighiouart H, Weiner DE, Griffith J, Pettitt D, Salem DN, et al. Anemia as a risk factor for cardiovascular disease and all-cause mortality in diabetes: The impact of chronic kidney disease. Journal of the American Society of Nephrology. 2005;16(11):3403–10.
Musallam KM, Tamim HM, Richards T, Spahn DR, Rosendaal FR, Habbal BSN A, et al. Preoperative anaemia and postoperative outcomes in non-cardiac surgery: a retrospective cohort study. The Lancet. 2011;378:1396–407.
Fowler AJ, Ahmad T, Phull MK, Allard S, Gillies MA, Pearse RM. Meta-analysis of the association between preoperative anaemia and mortality after surgery. British Journal of Surgery. 2015 Oct;102(11):1314–24.
National institute of health and care excellence(NICE). Blood transfusion. 2015.Available from: www.nice.org.uk/guidance/ng24
Besarab A, Coyne DW. Iron supplementation to treat anemia in patients with chronic kidney disease. NATURE REVIEWS | NEPHROLOGY. 2010;6:699–710.
Auerbach M, Ballard H. Clinical Use of Intravenous Iron: Administration, Efficacy, and Safety. Hematology. 2010 Dec;2010(1):338–47.
Batchelor EK, Kapitsinou P, Pergola PE, Kovesdy CP, Jalal DI. Iron Deficiency in Chronic Kidney Disease: Updates on Pathophysiology, Diagnosis, and Treatment. J Am Soc Nephrol. 2020;31(3):456.
Muñoz M, Acheson AG, Auerbach M, Besser M, Habler O, Kehlet H, et al. International consensus statement on the peri-operative management of anaemia and iron deficiency. Anaesthesia. 2017 Feb;72(2):233–47.
Gafter-Gvili A, Schechter A, Rozen-Zvi B. Iron Deficiency Anemia in Chronic Kidney Disease. Acta Haematologica. 2019 May;142(1):44–50.
Mikhail A, Brown C, Williams JA, Mathrani V, Shrivastava R, Evans J, et al. Renal association clinical practice guideline on Anaemia of Chronic Kidney Disease DOI: 10.1186/s12882-017-0688-1
National institute of health and care excellence(NICE). Chronic kidney disease: assessment and management [Internet]. NICE guidance. 2021 Aug [cited 2022 Jun 13]. Available from: https://www.nice.org.uk/guidance/ng203
National institute of health and care excellence(NICE). Chronic kidney disease: assessment and management guidelines [Internet]. NICE guideline. 2021 Aug [cited 2022 Mar 18]. Available from: www.nice.org.uk/guidance/ng203
Kidney disease improving global outcomes (KDIGO). Clinical Practice Guideline for Anemia in Chronic Kidney Disease. Kidney international Supplements. 2012 [cited 2022 Mar 18]. ;2(4):279–334.
Muñoz M, Acheson AG, Auerbach M, Besser M, Habler O, Kehlet H, et al. International consensus statement on the peri-operative management of anaemia and iron deficiency. Anaesthesia. 2017 Feb;72(2):233–47.
Goddard AF, James MW, Mcintyre AS, Scott BB. Guidelines for the management of iron deficiency anaemia. Gut. 2011;60:1309–16.
World Health organisation. Haemoglobin concentrations for the diagnosis of anaemia and assessment of severity. [Internet]
Gafter-Gvili A, Schechter A, Rozen-Zvi B. Iron Deficiency Anemia in Chronic Kidney Disease. Acta Haematologica. 2019 May;142(1):44–50.
Deray G, Heurtier A, Grimaldi A, Launay Vacher V, Isnard Bagnis C. Anemia and Diabetes. American Journal of Nephrology. 2004;24(5):522–6.
Suárez-Ortegón MF, McLachlan S, Price AH, Fernández-Balsells M, Franch-Nadal J, Mata-Cases M, et al. Decreased iron stores are associated with cardiovascular disease in patients with type 2 diabetes both cross-sectionally and longitudinally. Atherosclerosis. 2018 May;272:193–9.
Bi Y, Ajoolabady A, Demillard LJ, Yu W, Hilaire ML, Zhang Y, et al. Dysregulation of iron metabolism in cardiovascular diseases: From iron deficiency to iron overload. Biochemical Pharmacology. 2021 Aug;190. DOI: 10.1016/j.bcp.2021.114661
Lunt M. Selecting an Appropriate Caliper Can Be Essential for Achieving Good Balance With Propensity Score Matching. American Journal of Epidemiology. 2014 Jan;179(2):226–35.
Rosenbaum PR, Rubin DB. Constructing a Control Group Using Multivariate Matched Sampling Methods That Incorporate the Propensity Score. 2012 DOI: 10.1080/00031305.1985.10479383
Rubin DB. Using Propensity Scores to Help Design Observational Studies: Application to the Tobacco Litigation. Health Services & Outcomes Research Methodology. 2001;2:169–88.
Ganzoni A. Intravenous iron-dextran: therapeutic and experimental possibilities. Schweiz Med Wochenschr. 1970 [cited 2022 Mar 7]. ;100(7):301-3.
Auerbach M, Ballard H, Glaspy J. Clinical update: intravenous iron for anaemia. Leukemia . 2007;21:627–32.
Rampton D, Folkersen J, Fishbane S, Hedenus M, Howaldt S, Locatelli F, et al. Hypersensitivity reactions to intravenous iron: guidance for risk minimization and management. Haematologica. 2014 Nov;99(11):1671.
Sivakumar C, Jubb VM, Lamplugh A, Bhandari S. Safety of Intravenous Iron-Cosmofer and Monofer Therapy in Peritoneal Dialysis and Non-Dialysis-Dependent Chronic Kidney Disease Patients. Peritoneal Dialysis International. 2019;39(2):192–5.
Adkinson NF, Strauss WE, Macdougall IC, Bernard KE, Auerbach | Michael, Kaper RF, et al. Comparative safety of intravenous ferumoxytol versus ferric carboxymaltose in iron deficiency anemia: A randomized trial. Am j Hematol. 2018;93:683–90.
Richards T, Baikady RR, Clevenger B, Butcher A, Abeysiri S, Chau M, et al. Preoperative intravenous iron to treat anaemia before major abdominal surgery (PREVENTT): a randomised, double-blind, controlled trial. The Lancet. 2020 Oct;396(10259):1353–61.
Danielson BG. Structure, chemistry, and pharmacokinetics of intravenous iron agents. Journal of the American Society of Nephrology. 2004 Dec;15(SUPPL. 2). DOI: 10.1097/01.ASN.0000143814.49713.C5
Avni T, Bieber A, Grossman A, Green H, Leibovici L, Gafter-Gvili A. The Safety of Intravenous Iron Preparations: Systematic Review and Meta-analysis. Mayo Clin Proc. 2015;90(1):12–23.
DeLoughery TG. Safety of Oral and Intravenous Iron. Acta Haematologica. 2019 May;142(1):8–12.

Competing interest reported. C. A. Chukwu declares no competing interest. H. Gilbody declares no competing interest O. Wickens declares no competing interest C. Carroll declares no competing interest. S. Bhandari declares no competing interest P.A Kalra reports fees for lecturing and advisory board attendance for Astra Zeneca, Napp, Mundi Pharma, Boehringer Ingelheim, Novonordisk, Pfizer, Pharmacosmos, Vifor; Consultancy work with Bayer, Astellas, Otsuka and Unicyte; Institutional research funding from Vifor, GSK, Bergen Bio and Evotec

Additionalfile1.csv

Download PDF

Version 1

posted

You are reading this latest preprint version

Factors governing the erythropoietic response to intravenous iron infusion in patients with chronic kidney disease; a retrospective cohort study

Status:

Version 1

Abstract

Figures

Introduction

Materials And Methods

Study population

Data collection

Outcomes and predictor variables

Statistical analysis

Propensity score matching analysis

Multivariate regression analysis

Results

Response to parenteral iron

Multivariate analysis

Discussion

Clinical implications

Strengths and limitations of the study

Conclusion

Abbreviations

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1