key: cord-1039484-yocfl17l
authors: Khadzhynov, Dmytro; von dem Berge, Uwe; Muench, Frédéric; Karaivanov, Stoyan; Koerner, Roland; Kruse, Jan M.; Zickler, Daniel; Budde, Klemens; Eckardt, Kai-Uwe; Lehner, Lukas J.
title: Efficacy and complications of regional citrate anticoagulation during continuous renal replacement therapy in critically ill patients with COVID-19
date: 2021-11-09
journal: J Crit Care
DOI: 10.1016/j.jcrc.2021.10.010
sha: 26a2df013de4a5456d8e4db098c98c0a50dc5b3c
doc_id: 1039484
cord_uid: yocfl17l

BACKGROUND: We compared filter survival and citrate-induced complications during continuous renal replacement therapy (CRRT) with regional citrate anticoagulation (RCA) in COVID-19 and Non-COVID-19 patients. METHODS: In this retrospective study we included all consecutive adult patients (n = 97) treated with RCA-CRRT. Efficacy and complications of RCA-CRRT were compared between COVID-19 and Non-COVID-19 patients. RESULTS: Mean filter run-time was significantly higher in COVID-19 patients compared to Non-COVID-19 patients (68.4 (95%CI 67.0–69.9) vs. 65.2 (95%CI 63.2–67.2) hours, respectively; log-rank 0.014). COVID-19 patients showed significantly higher activated partial thromboplastin time (aPTT) throughout the CRRT due to intensified systemic anticoagulation compared to Non-COVID-19 patients (54 (IQR 45–61) vs. 47 (IQR 41–58) seconds, respectively; p < 0.001). A significantly higher incidence of metabolic alkalosis, hypercalcemia and hypernatremia, consistent with reduced filter patency and citrate overload, was observed in COVID-19 patients compared to Non-COVID-19 patients (19.1% vs. 12.7%, respectively; p = 0.04). These metabolic disarrangements were resistant to per-protocol adjustments and disappeared after replacement of the CRRT-filter. CONCLUSIONS: RCA-CRRT in COVID-19 patients with intensified systemic anticoagulation provides an adequate filter lifespan. However, close monitoring of the acid-base balance appears warranted, as these patients tend to develop reduced filter patency leading to a higher incidence of citrate overload and metabolic disturbances. TRIAL REGISTRATION (LOCAL AUTHORITY): EA1/285/20 (Ethikkommission der Charité - Universitätsmedizin Berlin); date of registration 08.10.2020.

The coronavirus disease 2019 (COVID-19) pandemic has created a surge in patients requiring intensive care worldwide [1] . A significant proportion of critically ill patients with severe respiratory distress syndrome due to SARS-CoV2 infection develop AKI and require renal replacement therapy (RRT) [2] [3] [4] [5] , compromising the capacity for delivering RRT in hotspots of disease outbreak [6] [7] [8] . Additionally, the incidence of filter clotting during RRT was reported to be very high despite the use of systemic anticoagulation, presumably due to an incompletely understood coagulopathy in patients with COVID-19 [9] [10] [11] . In combination with the rapid increase in the number of patients with AKI requiring CRRT the high incidence of unexpected treatment discontinuations due to shortened filter life-span may enhance the workload of medical staff and aggravate shortages of dialysis filters or other supplies. Consequently, there have been concerns about the optimal treatment strategy of COVID-19 patients requiring CRRT [6] [7] [8] 12] . Regional citrate anticoagulation for CRRT is generally recommended [13] . However, until now there is almost no evidence supporting the use of RCA during CRRT in COVID-19 patients with AKI [7] . In addition, nothing is known about the incidence and character of citrate related acid-base and electrolyte disturbances during RCA-CRRT in patients with COVID-19 as compared to other ICU patients. Regardless of therapy mode, in all RCA protocols citrate is partially infused into the patient and subsequently metabolized, potentially leading to alkalization [14] . In order to avoid development of metabolic alkalosis during treatment, sufficient RCA protocols usually include a dialysis solution with reduced sodium and bicarbonate concentration (Fig. 1A) [15, 16] . Due to increased coagulation tendency COVID-19 patients may be prone to an early loss of clearance at the filter level (so called filter-clogging), resulting in impaired citrate elimination and consequently excessive citrate load with metabolic alkalosis, hypernatremia and hypercalcemia. Strict protocol application normally allows to avoid those complications by adapting (reducing) the blood-to-dialysis flow ratio. However, in terms of irreversible filter patency loss due to filter-clogging, those metabolic complications could be resistant to per protocol adaptations and rapid replacement of the filter might be necessary (Fig. 1B) .

In light of these considerations the aim of the actual study was 1) to determine the filter lifetime and 2) the frequency, extend, and timecourse of acid-base and electrolyte disturbances indicative of reduced filter patency during CRRT with RCA in critically ill COVID-19 patients in comparison with Non-COVID-19 patients.

This retrospective study was conducted at the Charité Berlin, the largest university tertiary center in Germany. We included all consecutive adult patients with symptomatic COVID-19 disease and a positive reverse-transcriptase polymerase chain reaction (PCR) test result from an oropharyngeal swap, sputum, or bronchioalveolar lavage fluid, and AKI treated with the RCA-CRRT between March 1st, 2020 and May 31st, 2020. Patients were recruited from 7 different ICUs. During the study period in the first wave of the COVID-19 pandemic, the ICUs of our university hospital were separated into COVID-19 and Non-COVID-19 units. In both categories of ICUs diagnosis of AKI and indication for CRRT were determined in interdisciplinary consultation between intensivists and the renal consult service. There was no difference in the CRRT approach, materials used or medical staff involved in providing CRRT between COVID-19 and Non-COVID-19 patients. The study was approved by the local ethics review committee (Ethikkommission der Charité -Universitaetsmedizin Berlin; EA1/285/ 20) and was performed in accordance with the Declaration of Helsinki. The need for patient's informed consent was waived by the local ethics and data safety committees due to retrospective character of the study and anonymization of all data sets before analysis.

We used a previously published weight-adapted protocol for CVVHD [17] with RCA in all patients requiring CRRT and all patients received the same default settings at therapy start, aiming a therapy dose of 20 to 25 mL/kg/h. Briefly, CVVHD was conducted using high-flux dialyzers (AV1000, Fresenius Medical Care (FMC), Bad Homburg, Germany), calcium-free dialysis solution (Ci-Ca dialysate, FMC) and 4% trisodium citrate solution (FMC) with a starting dose of 4 mmol/L blood. Calcium substitution was started in a dose of 1.7 mmol per liter effluent flow. The initial ratio of blood flow was set accordingly to the chosen dialysate flow to maintain a ratio of 3:1. According to the published algorithm [17] , dialysate flow was increased to a lower ratio of blood to dialysate flow in case of a high blood bicarbonate concentration and vice versa in case of a low blood bicarbonate concentration. Filters were changed routinely after 72 h of treatment, as indicated by the manufacturer, with allowance of a tolerance ±12 h for practical reasons. A detailed description of this RCA-CVVHD protocol is presented in the Supplement part (Supplemental Data File -Supplement 1). The CRRT in patients with ECMO therapy was performed through a separate central vascular catheter and were not connected to the ECMO circuit.

Routine check of metabolic and electrolyte parameters as well as postfilter ionized calcium measurement for anticoagulation monitoring was scheduled 3 times daily or more often if clinically required. The Schematic view of an extracorporeal circuit with regional citrate administration in CVVHD mode and usual per protocol adjustments during metabolic alkalosis. Citrate solution (often as a trisodium citrate) is administered in the arterial line of the CRRT circuit and forms citrate-calcium complexes (CCC, orange circles), which are largely removed from the blood through the CRRT filter and effluent. CCCs which are not removed from the extracorporeal circulation return to the patient's blood and are further metabolized, producing bicarbonate and releasing calcium and sodium ions (A). In case of a metabolic alkalosis standard blood-to-dialysis ratio (3:1) is reduced by increasing the dialysate flow to enhance CCC removal (B, normal filter patency). In case of normal filter sieving patency, adjustment of blood-to-dialysis ratio leads to a decrease of systemic sodium load and bicarbonate generation. In terms of filter clogging the adjustment of the blood-to-dialysis is ineffective, as the sieving patency of the filter is reduced (B, filter clogging). Thus, filter clogging leads to metabolic alkalosis, hypernatremia and hypercalcemia, owing to increased CCC-substrate for bicarbonate generation and increased sodium and calcium load. Characteristically increasing systemic ionized calcium levels, resulting in lower calcium substitution rates are an early indicator of filter clogging. monitoring was started immediately after initiation of CRRT and every 8 h thereafter. Systemic (typically from arterial line) and postfilter blood samples were analyzed on an automated blood gas analyzer (ABL900, Radiometer, Copenhagen, Denmark).

Filter clotting was defined as filter loss unrelated to protocol violation, technical failures or access problems. Filter-clogging was defined in case of: 1) metabolic complications consistent with citrate overload (metabolic alkalosis, hypercalcemia (or reduced calcium infusion rate) and increasing sodium concentration), 2) those metabolic derangements were resistant to per protocol described adjustment algorithm and 3) those metabolic derangements were transient after the dialysis filter replacement. Electronic medical records of patients were thoroughly screened for documented increase of transmembrane pressure (TMP).

The following clinical data were extracted from electronic patient data management system: age, gender, reason for admission on ICU, acute physiology and chronic health evaluation score II (APACHE II) at admission, Sepsis-related Organ Failure Assessment score (SOFA) at admission, length of ICU stay, CRRT parameters (duration of RCA-CVVHD, reason for circuit discontinuation, filter run-time and flow rates of blood, citrate-, dialysate-, and calcium-solutions) as well as ICU mortality. The observation period was limited to ICU treatment.

Baseline demographics and laboratory data were described as mean ± standard deviation (SD) for normally distributed parameters and as median (interquartile range, i.e., 25-75 percentile range) for non-normally distributed values. Statistical significance of the differences between groups for continuous variables was tested using student's t-test or Mann-Whitney-U test in case of a non-parametric distribution. The Chi-squared test was used to assess the statistical difference between the groups for categorical variables. The over-all filter survival and cumulative risk of filter-clogging were calculated by using the Kaplan-Meier-Method. Significance level was set to α < 0.05. All analyses were conducted using SPSS 23 for Windows (SPSS Inc., IBM, Armonk, USA).

During the study period there were 44 consecutive patients with COVID-19 and 53 Non-COVID-19 patients treated with RCA-CVVHD in the participating ICUs. Baseline patient characteristics and medical history of both COVID-19 and Non-COVID-19 patients are summarized in Table 1 . Baseline characteristics were similar between both groups, with the exception of a significantly higher proportion of patients with hypertension and malignancies and a higher proportion of patients treated with extracorporeal membrane oxygenation in the COVID-19 group. All included patients have received systemic anticoagulation with unfractionated heparin. There was no significant difference in the number of patients treated with aspirin between the groups (13.6% in COVID-19 vs 13.2% in Non-COVID-19 patients; p = 0.968). Platelet count was significantly higher in COVID-19 patients compared with Non-COVID-19 patients (190 /nL vs. 157 /nL, respectively; p < 0.001).

There was no significant difference in the ICU mortality rates between the groups (48% in COVID-19 patients vs 59% in Non-COVID-19 patients; p = 0.29). The median length of ICU stay was significantly higher in COVID-19 patients compared to Non-COVID-19 patients (42 days (IQR 22-69) vs 24 days (IQR 9-55), respectively; p = 0.045). Clinical outcomes of patients as well as CRRT outcomes are summarized in Table 2 . 

There was a significantly higher incidence of combined metabolic disarrangements (metabolic alkalosis, hypercalcemia and increasing sodium concentration) consistent with reduced filter patency and resistant to per protocol described adjustment algorithm, defined as filter clogging, during RCA in COVID-19 patients (19.1% vs. 12.7% filters in Non-COVID-19 patients, p = 0.04). No records of elevated TMP were found in EMR of patients with filter clogging. The proportion of patients with more than one filter clogging episode was significantly higher in COVID-19 patients (52% vs. 23% patients in the Non-COVID-19 group, p = 0.002; Table 2 ). Characteristic metabolic disarrangements for filter clogging and the resistance to protocol-based CRRT adjustments as well as they time-course after filter replacement are shown in Fig. 3 . The cumulative incidence of the filter clogging was time-dependent (Fig. 4) . In the COVID-19 group significantly higher levels of fibrinogen and reduced levels of D-dimers were found in patients without filter clogging compared with patients with impaired filter permeability (Table 4 ).

To the best of our knowledge, this is the first study comparing filter survival and patency during RCA-CRRT in COVID-19 and Non-COVID-19

patients. The study revealed three main findings. First, a RCA protocol with the initial citrate dose of 4 mmol/L blood offers an effective anticoagulation strategy for CVVHD in COVID-19 patients with intensified anticoagulation, resulting in a mean circuit life-span of almost 70 h. Second, despite initial concerns, the lifespan of CVVHD filters in COVID-19 patients in present study was not only non-inferior but, surprisingly, even superior compared to Non-COVID-19 patients. This unexpected finding can be partially explained by the fact that early during the evolution of the COVID-19 pandemic recommendations to start intensified anticoagulation in severely ill COVID-19 patients were implemented in clinical practice. Accordingly, in our study COVID-19 patients had significantly higher aPTT throughout the CVVHD treatment compared to Non-COVID-19 patients. Third, even in the presence of systemic anticoagulation we found evidence for enhanced filter clogging with reduced filter patency and consequent citrate overload during RCA-CVVHD in COVID-19 patients.

We are not aware of other studies comparing filter patency of RCA-CRRT between COVID-19 and Non-COVID-19 patients. Moreover, a systematic analysis of citrate-induced complications during RCA-CRRT in COVID-19 patients is lacking. The available evidence is solely based on a limited number of literature with small sample size and significant patient-and RRT-level heterogenity, comparing the anticoagulations strtegies during different modalities of RRT [18] [19] [20] . A recent retrospective and small study compared different anticoagulation strategies during RRT in COVID-19 patients [20] . Arnold et al. reported a significantly longer circuit lifespan of CRRT in the RCA group compared to patients treated with UFH alone, partially supporting the findings of the present study. Another retrospective study compared the anticoagulation strategies during CRRT [19] . In line with our results, Shankaranarayanan et al. showed that most superior filter life was reached during combined regional citrate and systemic heparin anticoagulation. A prospective trial comparing regional citrate anticoagulation with systemic heparin anticoagulation during CRRT in patients with COVID-19 is ongoing and should provide results at the end of 2021 [21] . Limited evidence shows a higher anticoagulation efficacy of direct thrombin inhibitor, argatroban in patients with COVID-19 treated with CRRT [19] . However, until now, argatroban should be used at most as an escalation therapy in patients with frequent and unexplained circuit clotting. The routine use of argatroban in COVID-19 patients treated with CRRT cannot be unequivocally recommended without further studies on safety and efficacy in this population.

Significantly higher frequency of combined metabolic disarrangements (metabolic alkalosis, hypercalcemia and increasing sodium concentration) consistent with reduced filter patency and citrate overload during RCA-CVVHD in COVID-19 patients could be explained by increased filter clogging, a well-known term from the early times of dialysis research [22] . This normally benign complication results in the reduced permeability of the filter. However, it gains apparent clinical significance when unphysiological solutions, such as citrate acid, are infused to the extracorporeal circuit. Filter sieving patency especially in terms of RCA-CRRT plays a crucial role regardless of the protocol used. All these protocols require pre-filter administration of a citrate solution, mostly in form of trisodium citrate. During treatment, after infusion of citrate solution into the extracorporeal circuit and formation of calcium-citrate complexes (CCC), a significant amount (approximately 50%) of CCCs is eliminated into the dialysate [16] . The clearance of CCC should be maintained as high as possible to reduce the administration of (sodium-) citrate to the patient. CCCs, which are not removed through the dialysis filter, return to the systemic circulation and are further metabolized via the citric acid cycle, producing bicarbonate and releasing sodium and calcium. Thus, RCA leads to plasma alkalization due to bicarbonate generation and increases sodium load [16] . In order to avoid development of metabolic alkalosis during treatment, a sufficient RCA protocol has to compensate for alkalization of the blood from infused citrate by applying an adapted dialysis solution with the reduced bicarbonate and sodium concentration [23] [24] [25] [26] . In case of the intact permeability and sieving patency of the hemofilter, quick correction of the metabolic alkalosis using the described RCA-CVVHD protocol could be simply reached by reducing the blood to dialysate ratio [15] . In case of reduced filter permeability due to early filter clogging citrate delivery increases and citrate overload might occur [16] , resulting in metabolic alkalosis, hypernatremia and hypercalcemia. Clearly, those metabolic disarrangements are resistant to per protocol prescribed adjustments of the blood to dialysate ratio as the sieving of CCC is significantly reduced. Rapid filter replacement in this case is the only available option [16] . Not surprisingly, results of the present study showed that the cumulative incidence of filter clogging is increasing with the increasing treatment time, especially after 30 h of filter run-time. This somewhat trivial finding is of pronounced importance because it underscores the importance of careful metabolic monitoring of filter patency and the efficacy of protocol adjustments, particularly for CRRT filters running longer than 48 h. The mechanisms leading to increase filter clogging in COVID-19 although not completely clear, are in line with the COVID-19 induced coagulopathy. Recently, several groups reported the association of thromboembolic complications with impaired fibrinolysis in critically ill COVID-19 patients [27, 28] . However, the extent of the influence of hypofibrinolysis as an important component of COVID-19 induced coagulopathy on increased filter clogging remains to be investigated.

Despite the fact that there is a generally accepted indication for therapeutic systemic anticoagulation in severely ill COVID-19 patients, we believe RCA should be the preferred option for CRRT in this group of patients, regardless of the systemic anticoagulation. According to present data, this strategy allows to reach a superior filter life span without serious adverse events. Further randomized controlled studies on the safety and efficacy of this anticoagulation approach in COVID-19 population are warranted.

The present study has some limitations. Although it includes patients with a wide spectrum of comorbidities and a variety of causes of critical illness in the Non-COVID-19 group, it focusses on patients with a high disease severity in a tertiary care center. The generalizability to other settings and patients with a less severe course and earlier stages of the disease needs to be tested. Second, due to the retrospective nature unidentified confounding factors might have been missed. Nevertheless, the relatively high number of patients and circuits used increases the validity of the comparison between COVID-19 and Non-COVID-19 patients. Moreover, there was no difference in the CRRT approach, materials or involved medical team for providing CRRT between COVID-19 positive and Non-COVID-19 patients, minimizing potential bias by diagnosis. Thirdly, we used a standardized RCA-Protocol with 4-mmol citrate per 1 L blood flow and CVVHD as the CRRT-Mode; different protocols and especially different CRRT filters may show different performance. However, protocols using reduced citrate dose or those with a different CRRT-mode have been shown to be less effective in Non-COVID-19 patients [29, 30] . At last, values of TMP were not available due to the retrospective character of the study, however all medical charts were screened for records of TMP elevation in circuits with filter clogging. 

Despite the initial concerns, CRRT combined with RCA and intensified systemic anticoagulation allows to achieve an effective filter lifespan in COVID-19 patients. However, due to significantly higher incidence of filter clogging leading to citrate overload and consecutive metabolic alkalosis, increasing sodium concentration and hypercalcemia, special attention should be paid to acid-base balance of COVID-19 patients on RCA-CRRT especially in CRRT-filters running for more than 48 h, as they tend to develop reduced filter patency, which is resistant to CRRT-protocol adjustments.

Ethical approval and consent to participate Ethical approval was granted by the local ethics review committee (Ethikkommission der Charité -Universitätsmedizin Berlin; EA1/285/ 20) on the 22nd October 2020, retrospectively registered. The study was performed in accordance with the Declaration of Helsinki. According to the local ethics review committee and data safety board no consent to participate was required due to the retrospective and anonymous character of this single center study.

An interactive web-based dashboard to track COVID-19 in real time

AKI in hospitalized patients with COVID-19

Disaster response to the COVID-19 pandemic for patients with kidney disease in New York City

AKI treated with renal replacement therapy in critically ill patients with COVID-19

High rates of long-term renal recovery in survivors of COVID-19-associated acute kidney injury requiring kidney replacement therapy

Impending shortages of kidney replacement therapy for COVID-19 patients

COVID-19-associated acute kidney injury: consensus report of the 25th acute disease quality initiative (ADQI) workgroup

Estimating shortages in capacity to deliver continuous kidney replacement therapy during the COVID-19 pandemic in the United States

High prevalence of early asymptomatic venous thromboembolism in anticoagulated COVID-19 patients hospitalized in general wards

High incidence of venous thromboembolic events in anticoagulated severe COVID-19 patients

High risk of thrombosis in patients with severe SARS-CoV-2 infection: a multicenter prospective cohort study

Preparedness of kidney replacement therapy in the critically ill during COVID-19 surge

Acute kidney injury in critically ill patients with COVID-19

Regional citrate anticoagulation for continuous renal replacement therapy

Treatment of metabolic alkalosis during continuous renal replacement therapy with regional citrate anticoagulation

Complications of regional citrate anticoagulation: accumulation or overload? Crit Care

A safe citrate anticoagulation protocol with variable treatment efficacy and excellent control of the acid-base status

Dialysis filter life, anticoagulation, and inflammation in COVID-19 and acute kidney injury

Anticoagulation strategies and filter life in COVID-19 patients receiving continuous renal replacement therapy: a single-center experience

Comparison of different anticoagulation strategies for renal replacement therapy in critically ill patients with COVID-19: a cohort study

Cov-hep study: heparin in standard anticoagulation based on citrate for continuous veno-venous hemodialysis in patients with COVID-19: a structured summary of a study protocol for a randomized controlled trial

Study of membrane permeability in haemodialysis

Regional citrate anticoagulation for continuous arteriovenous hemodialysis in critically ill patients

Regional citrate anticoagulation in continuous hemodialysis-acid-base and electrolyte balance at an increased dose of dialysis

Metabolic complications during regional citrate anticoagulation in continuous venovenous hemodialysis: single-center experience

Safety and efficacy of regional citrate anticoagulation in continuous venovenous hemodialysis in the presence of liver failure: the liver citrate anticoagulation threshold (L-CAT) observational study

Thromboembolic complications in critically ill COVID-19 patients are associated with impaired fibrinolysis

Impaired fibrinolysis in critically ill COVID-19 patients

Citrate anticoagulation for continuous venovenous hemofiltration

Effect of regional citrate anticoagulation vs systemic heparin anticoagulation during continuous kidney replacement therapy on Dialysis filter life span and mortality among critically ill patients with acute kidney injury: a randomized clinical trial

Besides our obligation to thank all patients included in the study, we are devoting this manuscript with deep sadness and mourning to our beloved friend and colleague, Dr. Torsten Slowinski, whose premature passing during the preparation of the manuscript left an empty space in our hearts.

Not applicable.

The data that support the fndings of this study are available from the corresponding author upon reasonable request.

D.K., T.S., D.Z. and R.K. have received funds for speaking at symposia organized on behalf of Fresenius Medical Care, Germany. The other authors here have nothing to declare.

The study was funded through the department funds.

Supplementary data to this article can be found online at https://doi. org/10.1016/j.jcrc.2021.10.010.