Anaemia in cardiac surgery – a retrospective review of a centre's experience with a pre‐operative intravenous iron clinic

Pre‐operative anaemia is associated with higher rates of transfusion and worse outcomes, including prolonged hospital stay, morbidity and mortality. Iron deficiency is associated with significantly lower haemoglobin levels throughout the peri‐operative period and more frequent blood transfusion. Correction of iron stores before surgery forms part of the first pillar of patient blood management. We established a pre‐operative anaemia clinic to aid identification and treatment of patients with iron deficiency anaemia scheduled for elective cardiac surgery. We present a retrospective observational review of our experience from January 2017 to December 2019. One‐hundred and ninety patients received treatment with intravenous iron, a median of 21 days before cardiac surgery. Of these, 179 had a formal laboratory haemoglobin level measured before surgery, demonstrating a median rise in haemoglobin of 8.0 g.l‐1. Patients treated with i.v. iron demonstrated a significantly higher incidence of transfusion (60%) compared with the non‐anaemic cohort (22%) during the same time period, p < 0.001. Significantly higher rates of new requirement for renal replacement therapy (6.7% vs. 0.6%, p < 0.001) and of stroke (3.7% vs. 1.2%, p = 0.010) were also seen in this group compared with those without anaemia, although there was no significant difference in in‐hospital mortality (1.6% vs. 0.8%, p = 0.230). In patients where the presenting haemoglobin was less than 130 g.l‐1, but there was no intervention or treatment, there was no difference in rates of transfusion or of complications compared with the anaemic group treated with iron. In patients with proven iron deficiency anaemia, supplementation with intravenous iron showed only a modest effect on haemoglobin and this group still had a significantly higher transfusion requirement than the non‐anaemic cohort. Supplementation with intravenous iron did not improve outcomes compared with patients with anaemia who did not receive intravenous iron and did not reduce peri‐operative risk to non‐anaemic levels. Questions remain regarding identification of patients who will receive most benefit, the use of concomitant treatment with other agents, and the optimum time frames for treatment in order to produce benefit in the real‐world setting.


Introduction
The worldwide prevalence of anaemia is estimated to be 25% [1]. Pre-operative anaemia is associated with higher rates of transfusion and prolonged hospital stay, morbidity and mortality [2][3][4][5][6][7]. This is particularly evident in cardiac surgery where blood loss and transfusion requirements are higher. Even relatively modest rates of peri-operative transfusion have been associated with increasing morbidity and mortality following cardiac surgery [8,9]. The presence of iron deficiency, with or without anaemia, is associated with significantly lower haemoglobin (Hb) levels throughout the peri-operative period, and this group of patients receive transfusions more frequently and receive a greater total number of red cell units than non-iron-deficient patients [10,11]. Patient blood management comprises a series of patient-centred and evidence-based preventative and reactive measures intended to reduce bleeding and blood transfusion. The international consensus guidelines on the management of peri-operative anaemia and iron deficiency recommend treatment of pre-operative iron deficiency anaemia with intravenous (i.v.) iron if the surgery is planned within 6 weeks [12]. This is also supported by the National Institute for Health and Care Excellence (NICE) quality statement of 2015 [13] and the 2020/21 Commissioning for Quality and Innovation (CQUIN) framework, which incentivises the treatment of iron deficiency anaemia before surgery. Introduction of a multimodal patient blood management programme has been associated with improved outcomes in terms of reduced rates of transfusion, morbidity and mortality [14] and it has been suggested that treatment of pre-operative anaemia may improve outcome and reduce cost [15,16]. However, such recommendations present significant organisational challenges in an often busy setting before surgery. Additionally, a recent detailed metaanalysis of patient blood management interventions found that, while patient blood management measures reduce bleeding and requirement for transfusion, they did not significantly reduce morbidity or mortality and there was a lack of evidence of true cost effectiveness [17].
In 2010, we performed an audit of the prevalence of anaemia in our centre, a high-volume stand-alone cardiothoracic unit. Similar to national data reported [2] we found a prevalence of anaemia of 30%. As a result of this, we set up a pre-operative anaemia clinic for the diagnosis and treatment of iron deficiency anaemia. Our clinic started to see patients in late 2016. We decided to perform a retrospective review of outcomes of patients attending our service, summarising our experience from January 2017 to December 2019.

Methods
We performed a retrospective review of data from all patients undergoing elective cardiac surgery between 1 January 2017 and 31 December 2019, a proportion of whom attended for pre-operative i.v. iron treatment. All elective cardiac and aortic surgical patients at our centre with proven iron deficiency and anaemia are invited to attend for iron, according to the approach described below and in online Supporting Information Figure S1. This service evaluation was registered with our Trust research and development department; ethical approval was waived in view of its observational nature. All patients scheduled for elective cardiac surgery attended for baseline screening before surgery, including a full blood count. Where Hb was < 130 g.l -1 , an automatic order for the addition of further tests was made, including ferritin and transferrin saturation (TSAT), B12/folate and C-reactive protein (CRP). Patients with Hb < 130 g.l -1 , whether male or female, and with either a ferritin level < 30 ng.ml -1 , or ferritin < 100 ng.ml -1 and TSAT < 20%, were identified. A review of their medical records was performed by the lead nurse for the anaemia service to assess for the presence of contra-indications to iron treatment. In the absence of any contra-indication, the patient was invited to attend pre-operatively to receive a single dose of i.v. iron isomaltoside 1000 (20 mg.kg -1 ) (Monofer, Pharmacosmos A/S, Holbaek, Denmark). Our hospital covers a large geographical area of Merseyside and North Wales but, despite this, all patients were offered the opportunity to attend.
Our local policy now stipulates that i.v. iron should be administered 4-6 weeks in advance of surgery if possible, but during the study period iron was occasionally offered up to 2 weeks before admission for surgery. Patients were admitted to our day surgery unit and reviewed by the lead nurse for the i.v. iron service. After confirmation of the absence of any contra-indication, iron isomaltoside 1000 was administered, followed by a period of monitoring. The first six patients treated by the service received a dose of iron calculated using the Ganzoni equation; total iron deficit (mg) = patient weight (kg) 9 (target Hbactual Hb in g.dl -1 ) 9 2.4 + iron stores, with target Hb of 13 g.dl -1 and iron stores of 500 mg used for all patients. This was later changed so that all patients received 20 mg.kg -1 . Following discharge from the day surgery unit, the patient attended for surgery as planned, and their Hb was checked before surgery.

Discussion
The main findings of our study were that, in patients with proven iron deficiency anaemia, supplementation with i.v. iron showed only a modest effect on Hb; supplementation with i.v. iron was not associated with improved outcomes compared with anaemic patients who did not receive i.v.
iron, and did not reduce peri-operative risk to non-anaemic levels.
The median rise of Hb in those patients given iron supplementation was only 8 g.l -1 and this was despite administration of a full replacement dose of iron a median of 3 weeks before surgery. It is possible that a larger rise in Hb would have been seen if the patients had been given more time to generate a response; however, this represents the reality of work within the constraints of the UK NHS. Our experience mirrors that of a large UK study addressing the feasibility of a larger UK-wide i.v. iron initiative [18], where iron was administered a median of 33 days before surgery, with a mean Hb rise of 8.4 g.l -1 . When considering the response of patients according to their presenting Hb we found that, while there was an increase in the median Hb before surgery in all groups, this was most pronounced in the cohort with the most severe anaemia, and less pronounced in those with mild anaemia (Hb of 120-129 g.l -1 ). As expected, a greater degree of anaemia was associated with a lower median ferritin and TSAT. As the degree of anaemia reduced, both these markers improved. Our practice has been to treat all patients, regardless of sex, where Hb is < 130 g.l -1 , but our data suggest that the optimal benefit, in terms of Hb increase, may favour treatment where Hb is < 120 g.l -1 , the practice in many other centres.
Six patients received a dose of iron less than 1000 mg following dosing using the Ganzoni formula (two patients received 600 mg, three received 700 mg and one received 900 mg). This group was treated a median of 22 days before surgery and demonstrated a median rise in Hb of 10 g.l -1 . They were included in the overall analysis.
Anaemic patients treated with i.v. iron demonstrated significantly higher rates of transfusion, stroke and requirement for haemofiltration when compared with the non-anaemic cohort of those with a presenting Hb > 130 g.l -1 . The risks, which are known to be greater in Table 1 Characteristics of patients who were anaemic and received intravenous (i.v.) iron, those who were anaemic but did not receive i.v. iron and patients who were not anaemic. Values are number (proportion) or median (IQR [range]). It is worth noting that only 39% of patients managed to attain a pre-operative Hb > 130 g.l -1 . There are many  factors which may contribute to this. Our current practice does not include administration of any erythropoiesisstimulating agent. There is concern regarding the risk of an increase in thromboembolic events in patients treated with an erythropoiesis-stimulating agent. It is unclear whether the addition of an erythropoiesis-stimulating agent would improve effectiveness of i.v. iron in the medium term before surgery, while also demonstrating an acceptable riskbenefit.
The role of hepcidin and the hepcidin-ferroportin axis in iron metabolism and the development of iron deficiency is increasingly understood [19]. Hepcidin is a hormone secreted by the liver to control the transfer of iron into the circulation via degradation of the ferroportin transporter. In physiological conditions, it is released when iron is plentiful, with the intention of preventing iron absorption and mobilisation, and secretion should conversely be reduced in the presence of iron deficiency. Despite this, in the presence of chronic inflammation, hepcidin release is frequently increased. Iron is an essential element to some pathogens, and reduction in its availability within the circulation may be protective, although ultimately leading to iron-restricted erythropoiesis and anaemia. As a consequence, many patients will receive limited benefit from oral iron supplementation, as dietary absorption is prevented by elevated hepcidin levels. It may also be true that not all patients will have a good response to i.v. iron therapy. A recent study of patients admitted to ICU showed that hepcidin levels were predictive of a response to i.v. iron [20].
There are limitations in our service that may contribute to the lack of an increase in Hb. The time period for identification and treatment of patients with iron deficiency before surgery is frequently limited. Our patients underwent treatment a median of 21 days before surgery, which may limit the time for treatment to be maximally effective.
However, a recent study by Klein   peak in ferritin levels [21], and 5-14 days to an increase in Hb [22]. surgery when patients attended pre-operative assessment clinics. Avoiding delays in diagnosis by encouraging early screening of patients for anaemia, or the development of same-day treatment clinics, is something we are pursuing and that may benefit this cohort. Community-based services could also increase the opportunity for intervention, as could regional delivery of services, allowing centralised treatment of patients from several hospital Trusts. With regional 'buy-in', the opportunities to assess and identify iron deficiency early, before referral, may allow optimisation of the timing of treatment.
We found that administration of i.v. iron to patients with iron deficiency anaemia before cardiac surgery was The role of iron is more than just in the formation of haemoglobin; iron has significant involvement in mitochondrial function; oxygen transport; DNA metabolism; host defence; and apoptosis [19]. Rossler et al. [27] recently demonstrated that iron deficiency, even in the absence of anaemia, is associated with higher mortality in patients undergoing cardiac surgery. With iron deficiency alone, mortality increased from 2% to 5%.
Where anaemia was also present, mortality increased from 4% to 14% and there was an increase in serious adverse events, major cardiac and cerebrovascular events,  [28]. In the cardiac surgical population, the measurement of this is more difficult, because the surgery performed is likely to lead to a significant improvement in quality of life, and identifying improvement solely related to iron replacement is therefore more challenging.
The delivery of a successful patient blood management programme requires education of clinicians, and our experience represents a 'real-world' setting where clinical autonomy regarding prescribing exists outside of a prescriptive study framework, particularly in terms of triggers to transfusion. The impact of intervention on one pillar of blood management is unlikely to produce a benefit without intervention on the other two. Where programmes collectively address all three pillars, there is evidence of a reduction in transfusion of red cells, and a lower complication and mortality rate [14]. We endeavour to educate staff according to recent consensus guidance, including the Frankfurt Patient Blood Management consensus document [29] and 2017 EACTA/ESCTS guidance [30], but ultimately decisions are made by the individual clinician.
Our experience raises several questions that future research may answer. To confer benefit in terms of reduction in requirement for transfusion and morbidity, do we need to achieve Hb > 130 g.l -1 before surgery or should we seek out a more individualised target Hb? Is anaemia alone an appropriate identifier of the 'at-risk' patient or are we looking at the wrong trigger for treatment? Should we be considering ferritin levels and/or TSAT as a trigger to treatment, rather than Hb alone? Development of anaemia occurs at a relatively advanced stage of iron deficiency, and it is not our practice to screen all patients for the presence of iron deficiency. As many patients demonstrate a fall in Hb during the peri-operative period, the presence of undiagnosed iron deficiency may hinder their ability to return to pre-operative levels as swiftly, potentially influencing their overall recovery. Should we screen and treat all patients with iron deficiency, regardless of their preoperative Hb?
The potential benefits of iron replacement in the cardiac and wider surgical population have been described as significant in terms of the impact on patient outcomes, duration of hospital stay and overall costs of patient care.
There is increasing experience that treatment is more complex than solely replacing iron, and simply setting up a service for administration may not be sufficient. Identifying patients who will benefit most, those who require concomitant treatment with other agents, and the optimum time frames in research populations, will hopefully help to create protocols that will transfer this potential benefit into the real-world setting.