**2. The principles and history of citrate anticoagulation in hemodialysis**

The anticoagulant properties of citrate have been known since the late 1800s already and are based on its capacity to chelate calcium ions. Ionized calcium (iCa) is an important co-factor at several steps in the coagulation cascade and, in that role, was formerly called coagulation factor IV. Addition of citrate to whole blood leads to formation of stable calcium-citrate complexes, thereby lowering the concentration of ionized calcium. At iCa levels below 0.5 mmol/L, clotting becomes impaired; at levels below approximately 0.3 mmol/L, coagulation is virtually blocked. This principle has been applied for storage of red cells in transfusion medicine since the early 20th century and later on for blood cell apheresis and lipid apheresis. Citrate physiologically occurs in the human body. It is an intermediate metabolite in the mitochondrial Krebs cycle, and all human cells that possess mitochondria can generate and metabolize citrate, particularly those tissues that are rich in mitochondria, such as the liver.

The first mention of citrate for anticoagulation in hemodialysis dates back to 1961 [5]. Traditionally, regional citrate anticoagulation in hemodialysis involves infusion of trisodium citrate into the arterial line of the extracorporeal circuit in sufficient quantities to lower iCa levels to around 0.25 to 0.35 mmol/L in order to substantially inhibit coagulation. In the venous limb of the dialysis tubing, ideally close to the point of blood reinfusion into the patient, calcium is substituted in the form of a calcium chloride or calcium carbonate infusion. This calcium substitution primarily serves to raise the iCa concentration in the blood to safe levels before the blood re-enters the patient's circulation, but there is another aspect to it as we shall see later. Classically, a calcium-free dialysate is used in this setting so

such as long half-life, lack of an antidote, or high cost, and all of them increase the bleeding

The primary purpose of anticoagulation during hemodialysis is to prevent clotting of the blood while it is traveling through the blood tubing and dialyzer. Against this background, the cornerstones of optimal anticoagulation for hemodialysis are complete suppression of the activation of the clotting cascade, strict limitation to the extracorporeal circuit, absence of

Limitation of anticoagulation to the extracorporeal circuit, also known as regional anticoagulation, is important because it eliminates the increased bleeding risk associated with systemic anticoagulation. Originally, this was accomplished by infusing heparin into the arterial line of the blood circuit and antagonizing its anticoagulant effect by infusing its antidote protamine into the venous line. Since protamine's half-life is shorter than heparin's, the anticoagulant effect may return after the dialysis procedure, increasing the bleeding risk. Also, this mode of anticoagulation is not suitable for HIT type II patients because of the heparin administration. Regional anticoagulation by infusing the arachidonic acid derivative prostacyclin into the arterial line is based on this molecule's inhibitory effect on thrombocyte aggregation and its short half-life of only a few minutes. The downsides are its vasodilatatory properties, which can cause significant hypotension during the treatment, and its prohibitive cost. Regional citrate anticoagulation is an alternative to these two methods that also confines anticoagulation to the extracorporeal circuit but does not come with the disadvantages mentioned above. In fact, it conveys a set of additional advantages

that go above and beyond merely providing regional anticoagulation.

**2. The principles and history of citrate anticoagulation in hemodialysis** 

The anticoagulant properties of citrate have been known since the late 1800s already and are based on its capacity to chelate calcium ions. Ionized calcium (iCa) is an important co-factor at several steps in the coagulation cascade and, in that role, was formerly called coagulation factor IV. Addition of citrate to whole blood leads to formation of stable calcium-citrate complexes, thereby lowering the concentration of ionized calcium. At iCa levels below 0.5 mmol/L, clotting becomes impaired; at levels below approximately 0.3 mmol/L, coagulation is virtually blocked. This principle has been applied for storage of red cells in transfusion medicine since the early 20th century and later on for blood cell apheresis and lipid apheresis. Citrate physiologically occurs in the human body. It is an intermediate metabolite in the mitochondrial Krebs cycle, and all human cells that possess mitochondria can generate and metabolize citrate, particularly those tissues that are rich in mitochondria,

The first mention of citrate for anticoagulation in hemodialysis dates back to 1961 [5]. Traditionally, regional citrate anticoagulation in hemodialysis involves infusion of trisodium citrate into the arterial line of the extracorporeal circuit in sufficient quantities to lower iCa levels to around 0.25 to 0.35 mmol/L in order to substantially inhibit coagulation. In the venous limb of the dialysis tubing, ideally close to the point of blood reinfusion into the patient, calcium is substituted in the form of a calcium chloride or calcium carbonate infusion. This calcium substitution primarily serves to raise the iCa concentration in the blood to safe levels before the blood re-enters the patient's circulation, but there is another aspect to it as we shall see later. Classically, a calcium-free dialysate is used in this setting so

risk as they are administered systemically.

serious side-effects, and low cost.

such as the liver.

as not to compromise anticoagulation due to calcium influx from the dialysate [6]. This setup of regional citrate anticoagulation is depicted in **Figure 1**.

Fig. 1. Conventional setup of regional citrate anticoagulation in hemodialysis.

A question of central importance is how plasma citrate concentrations relate to iCa concentrations. We analyzed the data from 21 regional citrate anticoagulation treatments performed at Renal Research Institute facilities in New York, USA, in 10 patients, during which 4% trisodium citrate (136 mmol/L) was infused into the arterial line and iCa measured before the dialyzer. Blood flow rates were 350 mL/min in 4 treatments, 400 mL/min in 13 treatments, and 450 mL/min in 4 treatments. Hematocrit and iCa were measured 13 minutes into the treatment using an Abbott i-Stat point-of-care analyzer. Hematocrits ranged from 28% to 39% (average, 33.6%). Citrate infusion rates ranged from 140 to 480 mL/h, and iCa ranged from 0.27 to 0.68 mmol/L (average, 0.38 mmol/L). Plasma citrate concentrations were calculated based on citrate infusion rates and calculated plasma flow rates. **Figure 2** illustrates the relationship between pre-dialyzer blood iCa activity and plasma citrate concentration. As can be seen, a plasma citrate concentration of >3.5 mmol/L is typically required to bring iCa levels to below 0.3 mmol/L. The exact citrate concentration necessary depends mainly on the individual patient's plasma calcium and protein (primarily albumin) concentrations. Total calcium in the serum comprises a protein-bound and a free (ionized) fraction, and the equilibrium concentrations of each can be estimated based on the respective dissociation constant [7-10]. Likewise, free citrate reacts with free calcium to form calcium-citrate complexes, again with a known dissociation constant [11]. Strictly, the multi-ionic milieu of the plasma should be considered, but reducing the relationships to calcium, protein, and citrate is a fair approximation. In clinical practice, these relationships are, however, not calculated. Instead, the citrate infusion rate is generally based on empirical knowledge and in most cases only tailored to the patient's blood flow rate. As can be expected, this may occasionally lead to citrate concentrations that are either too low to provide sufficient anticoagulation, or unnecessarily high. To assess the individual situation, pre-dialyzer (some groups use post-dialyzer) iCa levels can be measured in the plasma to ascertain that they are within the desired target range of approximately 0.25 to 0.35 mmol/L. If they are not, adjustments to the citrate infusion rate can be implemented and the iCa levels reassessed. Likewise, the post-dialyzer iCa concentrations are not known in clinical practice, and the rate of calcium substitution is based on empirical knowledge. Routinely, systemic iCa levels are measured in the patient at multiple time points during the treatment, and the calcium substitution rate is adjusted to counter drops or rises in systemic iCa concentration. Each adjustment usually necessitates a reassessment of iCa levels after 15 to 30 minutes to monitor its effect.

Citrate Anticoagulation in Hemodialysis 221

obvious advantages, however, there are several additional benefits to using regional citrate anticoagulation. One of these appears to be improved biocompatibility of the dialysis procedure: comparing heparin anticoagulation with citrate anticoagulation, Böhler et al. found that citrate anticoagulation reduced complement activation, neutropenia and lactoferrin release with the use of cuprophane dialyzers, and significantly inhibited neutrophil degranulation with the use of polymethyl methacrylate membranes [12]. Likewise, Gritters and colleagues compared anticoagulation using unfractionated heparin, low molecular weight heparin and citrate in a randomized crossover trial and found that citrate anticoagulation suppressed the dialysis-associated degranulation of polymorphonuclear cells and platelets. Furthermore, pro-atherogenic oxidized low-density lipoprotein levels were reduced by a median of 26% after only one week on citrate anticoagulation [13]. In view of the heightened inflammatory state of chronic hemodialysis patients, the reduction of oxidative stress, complement and cell activation associated with citrate dialysis may be a relevant benefit with regard to reducing the high cardiovascular morbidity in these patients. Hofbauer et al. compared anticoagulation with unfractionated heparin, low molecular weight heparin and citrate during dialysis with a single-use polysulfone dialyzer and used scanning electron microscopy to quantify the degree of membrane-associated clotting [14]. The highest degree of cell adhesion and thrombus formation was observed with unfractionated heparin, and it was only slightly reduced with the use of low molecular weight heparin. With regional citrate anticoagulation, on the other hand, thrombus formation was found to be negligible, indicating a far superior anticoagulation using citrate compared to both unfractionated and fractionated heparin. Gabutti et al. employed a randomized controlled cross-over design to compare standard heparin dialysis with regional citrate anticoagulation, dosed to achieve a similar degree of coagulation activation, and study the effects on complement activation and interleukin-1 beta release. In this setting, complement activation was slightly but significantly higher in the citrate dialysis group, but at the same time, interleukin-1 beta release was markedly reduced. Citrate can, and often is, dosed higher in regional citrate anticoagulation than was done in this study, and it stands to reason that with such higher citrate concentrations, complement activation would have been lower than with standard heparin dialysis, associated perhaps with a further decrease in interleukin-1 beta secretion. In line with Hofbauer's results mentioned above, regional citrate anticoagulation appears to allow for markedly prolonged filter patency times in continuous dialysis [15-18]. Lastly, a recent study by Oudemans-van Straaten and colleagues found higher patient and kidney survival in critically ill patients on citrate versus low-molecular weight heparin [19]. On top of these

benefits, citrate is a relatively inexpensive compound compared to heparin.

The single biggest concern with regional citrate anticoagulation is the development of potentially life-threatening systemic calcium derangements. Acute changes in systemic iCa can develop quickly when calcium elimination across the dialyzer (in the form of free calcium and calcium-citrate complexes), calcium release from the metabolism of calciumcitrate complexes, and calcium substitution (from the calcium infusion and/or the dialysate, if a calcium-containing dialysate is used) are mismatched. From this concern springs the need to monitor, at least initially, systemic iCa levels fairly closely during regional citrate anticoagulation. This, along with the more complex setup, presents a significant strain on

**4. The downsides of regional citrate anticoagulation** 

Fig. 2. Pre-dialyzer ionized calcium (iCa) concentration plotted against plasma citrate concentration.

During regional citrate anticoagulation, citrate enters the body in the form of both free citrate and calcium-citrate complexes. When this citrate is metabolized, each molecule yields three molecules of bicarbonate, which will have an impact on the acid-base status. Also, calcium is released from calcium-citrate complexes as they are metabolized, which impacts serum calcium concentration. The use of trisodium citrate or Acid Citrate Dextrose (ACD) solution further entails an additional sodium load to the patient that should be taken into account. In clinical practice, the dialysis prescriptions for regional citrate anticoagulation typically incorporate reduced sodium (by about 2 mmol/L) and bicarbonate (by about 5 mmol/L) concentrations. Magnesium concentration in the dialysate may be increased since citrate also complexes magnesium, leading to increased magnesium losses across the dialyzer.

Over the years, different algorithms for the administration of regional citrate anticoagulation have been suggested and studied, both for intermittent as well as continuous hemodialysis. These algorithms usually define blood and dialysate flow rates, the starting rates for citrate infusion and calcium substitution, rules on how these rates should be adjusted in case of iCa deviations from the specified circuit or systemic target ranges, time points for monitoring iCa, and downward adjustments for sodium and bicarbonate in the dialysate. Since citrate and calcium kinetics during dialysis depend on many factors, including the type of dialyzer used and the blood and dialysate flow rates, such algorithms generally are only applicable to the particular dialysis setting for which they have been validated. The purpose of all these algorithms is always to make the administration of regional citrate anticoagulation as safe and simple as possible, i.e. to minimize the risk for calcium or acid-base derangements, circuit clotting or other complications while requiring as little monitoring or intervention by the staff as possible.
