Citrate Anticoagulation for CKRT with Liver Failure

Since Mehta and colleagues' first description in 1990,¹ the use of regional citrate anticoagulation for continuous KRT (CKRT) has become more widespread. Subsequent clinical trials demonstrated greater safety and efficacy (a “win–win”) for citrate anticoagulation compared with more standard heparin-based protocols.² Indeed, regional citrate anticoagulation represents an elegant and highly effective approach to extracorporeal anticoagulation that does not confer any higher bleeding risk to the patient. As such, in 2012, the first iteration of the Kidney Disease Improving Global Outcomes AKI guidelines recommended regional citrate anticoagulation as the first-line option for anticoagulation in patients undergoing CKRT.³

Yet, important barriers continue to limit the broader implementation of citrate anticoagulation. Most regional citrate anticoagulation protocols result in some delivery of citrate to the patient and therefore rely on endogenous metabolism (primarily by the liver) of citrate to bicarbonate. For this reason, all previous randomized trials of citrate anticoagulation have excluded patients with severe liver disease. Furthermore, the Kidney Disease Improving Global Outcomes guidelines identified severe liver disease as a contraindication to citrate anticoagulation because of the risk of reduced citrate metabolism resulting in citrate accumulation and toxicity. Nevertheless, some observational studies have suggested that citrate can be safely applied in this population.⁴

In this issue of CJASN, Bai and colleagues attempt to shed further light on this area by reporting on the results of the first randomized controlled trial of citrate anticoagulation for CKRT in patients with liver failure and higher bleeding risk.⁵ In their study, 89 patients with liver failure (acute, chronic, or acute-on-chronic) were randomized to undergo CKRT with either regional citrate anticoagulation or no anticoagulation. The primary outcome was CKRT filter failure within 72 hours, and this occurred significantly more frequently in the no anticoagulation group compared with the regional citrate anticoagulation group (56% versus 27%, P = 0.003). This is a clinically important finding, as increased CKRT interruptions lead to reduced delivered dose of dialysis and the potential for greater blood loss in the circuit. The increased nursing workload associated with filter changes can also have an effect on overall patient care by distracting from other potential nursing care needs. This study therefore extends the known filter patency benefits of citrate anticoagulation to the liver failure population.

However, efficacy is not the primary concern regarding regional citrate anticoagulation in liver failure but rather safety related to citrate accumulation and toxicity. Clinically, the first potential sign of citrate toxicity is ionized hypocalcemia due to excess in vivo citrate binding of calcium in the setting of impaired citrate metabolism. Bai and colleagues observed a markedly higher incidence of severe hypocalcemia (defined as serum ionized calcium <0.9 mmol/L) in the citrate arm compared with controls (77% versus 13%), although they reported no adverse outcomes related to the hypocalcemia and there was no difference in the overall 28-day mortality. With subsequent protocol-based adjustments in calcium administration, all patients were reported to have improvement in their systemic ionized calcium levels. A second sign of citrate toxicity then becomes the elevation of the total:ionized calcium ratio, with a value >2.5 commonly used as a cutoff. In this trial, 57% of patients in the citrate group exceeded this limit compared with just 7% in the control group. Taken together, these findings seem to reinforce the concerns for liver failure being a contraindication to citrate administration. Yet, it is also notable that in most cases, the abnormalities were resolved with adjustments to the citrate protocol, such as with increased calcium or reduced citrate administration. So perhaps, the question is not whether liver failure is truly a contraindication to citrate anticoagulation, but rather, what adjustments to citrate protocols are necessary to accommodate patients with liver failure?

The first step in solving this puzzle is to fully account for citrate kinetics and recognize that absent citrate metabolism only leads to citrate accumulation in combination with the specific features of some regional citrate anticoagulation prescriptions. Theoretically, citrate accumulation can be avoided by achieving “truly regional” citrate anticoagulation where the circuit return blood has only a negligible citrate level (<2 mM) and the single-pass filter removal of citrate (citrate extraction ratio, or ECit) exceeds 70%. Under these conditions, since citrate does not reach the patient, there is no potential for citrate accumulation, regardless of the ability to metabolize it. We have developed such a regional citrate anticoagulation protocol for CKRT at our institution that is routinely used in patients with liver failure as well as others with severely impaired citrate metabolism, such as those in severe shock.⁶

How can we be sure that the small amount of citrate in the return blood will not accumulate to toxic levels? Comprehensive kinetic modeling of citrate fluxes during CKRT–regional citrate anticoagulation allows us to simulate systemic citrate levels as a function of time and separate the combined effects of citrate metabolism and the design of a specific CRKT-regional citrate anticoagulation prescription. Clinical studies of citrate kinetics in critically ill patients have supported this modeling approach.^7,8 In CKRT, steady-state solute levels will develop over time and will be equal to solute generation rate divided by solute clearance (combined filter and endogenous). In the case of citrate, generation (i.e., delivery via CKRT circuit) is determined by blood flow (QB) x citrate concentration ratio (Rcit)×(1-Ecit), and filter clearance is determined by QB×(1-Hct/100)×Ecit.⁶ To provide effective anticoagulation, particularly prefilter, Rcit must be in the 3.5–6 mmol/L range. Therefore, the most relevant parameter to limit systemic citrate accumulation is Ecit, and kinetic modeling can be used to maximize Ecit to prescribe regional citrate anticoagulation with minimized citrate delivery, maximized citrate clearance, and no risk of systemic citrate accumulation.

To illustrate this, Figure 1 shows the effect of different regional citrate anticoagulation protocols^6,9 on the risk of citrate accumulation and toxicity. The solid citrate curve represents a standard regional citrate anticoagulation protocol⁹ with expected Ecit of 40% and resultant citrate delivery of 20.5 mmol/h with filter clearance 2.5 L/h. The dashed and dotted citrate curves represent a “shock” protocol with two different QB, 100 and 150 ml/min and effluent flow 4 and 6 L/h (approximately 67% of QB), with Ecit 76 and 72%, citrate delivery approximately 7 and 10.5 mmol/h, and filter clearance 3.5 and 4.9 L/h, respectively. In the first 12 hours of modeling, body clearance keeps the systemic citrate level at or below 1.2 mM for all protocols regardless of Ecit. This is why virtually all published regional citrate anticoagulation protocols avoid citrate toxicity in most patients. The steady-state citrate level develops in 2–3 hours, as was similarly observed in the clinical trial by Bai and colleagues. At 12 hours, body metabolism of citrate is abruptly lost, for instance due to warm liver ischemia during a cardiac arrest. In the next 12–24 hours, the standard regional citrate anticoagulation protocol results in severe citrate accumulation, whereas the two “shock” prescriptions (with Ecit >70%) show maximum systemic citrate levels not permitted to exceed 2.5 mM.

Figure 1.

Citrate accumulation on the basis of systemic clearance of citrate and regional citrate anticoagulation protocol. Adapted from ref. 6 with permission. CKRT, continuous KRT.

In summary, the clinical trial from Bai and colleagues confirmed the benefits of citrate anticoagulation on filter life in patients with liver failure undergoing CKRT but also confirmed that citrate accumulation is a real concern in this population. However, with appropriate adjustments—ideally proactively, using citrate kinetic principles—the benefits of regional citrate anticoagulation can be realized while minimizing the risks even in patients with severely impaired citrate metabolism.

Disclosures

M. Heung reports consultancy for CardioSounds Inc., Potrero Inc., and Wolters Kluwer (Lexicomp); research funding from Astute Medical Inc., CardioSounds Inc., and Spectral Medical Inc.; and advisory or leadership role for Advances in Chronic Kidney Disease: Associate Editor/Editorial board. B. Szamosfalvi reports research funding from Renal Research Institute.

Funding

None.

Author Contributions

Conceptualization: Michael Heung, Balazs Szamosfalvi.

Writing – original draft: Michael Heung, Balazs Szamosfalvi.

Writing – review & editing: Michael Heung, Balazs Szamosfalvi.