Beyond the Randomized Clinical Trial: Citrate for Continuous Renal Replacement Therapy in Clinical Practice

Original Paper Received: May 8, 2013 Accepted: August 23, 2013 Published online: November 19, 2013 Beyond the Randomized Clinical Trial: Citrate for Continuous Renal Replacement Therapy in Clinical Practice L. Tovey H. Dickie S. Gangi M. Terblanche C. McKenzie R. Beale D. Treacher M. Ostermann Department of Critical Care, King s College London, King s Health Partners, Guy s & St Thomas Foundation Hospital, London, UK Key Words Citrate Acute kidney injury Continuous renal replacement therapy Anticoagulation Abstract Background: Premature circuit clotting is a major problem during continuous renal replacement therapy (CRRT). Six randomized controlled trials confirmed that regional anticoagulation with citrate is superior to heparin. Our objective was to compare circuit patency with citrate, heparin and epoprostenol in routine clinical practice. Methods: We retrospectively analysed data on circuit patency of all circuits used in a single centre between September 2008 and August 2009. We differentiated between premature filter clotting, elective discontinuation and waste. Results: 309 patients were treated with CRRT (n = 2,059 circuits). The mean age was 65.7; 63.8% were male. The methods to maintain circuit patency were unfractionated heparin (42.3%), epoprostenol (23.0%), citrate (14.7%), combinations of different anticoagulants (14.6%) and no anticoagulation (4.7%). Premature clotting was the most common reason for circuit discontinuation among circuits anticoagulated with heparin, epoprostenol or combinations of different anticoagulants (59 62%). Among circuits anticoagulated with citrate the main reason for discontinuation was elective (61%). Hazard regression analysis confirmed significantly better circuit survival with citrate. Changing from heparin to citrate decreased the risk of premature circuit clotting by 75.8%. Conclusion: In routine clinical practice, regional anticoagulation with citrate is associated with significantly better circuit patency than heparin or epoprostenol. 2013 S. Karger AG, Basel Introduction The practice of renal replacement therapy (RRT) of critically ill patients in the intensive care unit varies worldwide with differences in the main providers (i.e. nephrology versus critical care team, nurses versus doctors), preferred modalities and type of anticoagulation [1]. Unfractionated heparin is the most commonly used anticoagulant during continuous RRT (CRRT) mainly as a result of familiarity, low cost and ease of administration and monitoring. The obvious drawback is the risk of bleeding. An alternative is citrate which offers regional anticoagulation without systemic anticoagulation. Citrate acts by chelating calcium and therefore inhibits the clotting cascade at several levels. The principle of regional anticoagulation is based on pre-filter infusion of citrate aiming for an ionized Ca concentration [Ca i ] of <0.35 mmol/l, and calcium replacement either post-filter or via a separate line to correct for calcium losses [2]. The advantages are E-Mail karger@karger.com www.karger.com/nec 2013 S. Karger AG, Basel 1660 2110/13/1242 0119$38.00/0 Marlies Ostermann, PhD, MD, FRCP Department of Critical Care Medicine King s College London, Guy s & St Thomas Foundation Hospital London SE1 7EH (UK) E-Mail Marlies.Ostermann @ gstt.nhs.uk

better circuit patency and fewer bleeding complications. In addition, citrate serves as a source for the production of bicarbonate. Different methods of citrate anticoagulation have been developed for continuous haemodialysis, haemofiltration and haemodiafiltration. Some protocols use a fixed dose of citrate in relation to blood flow [3]. Others measure postfilter [Ca i ] and adjust the citrate infusion accordingly, a technique which is more complicated but optimizes anticoagulation [4]. Despite the fact that two meta-analyses including six different randomized controlled trials (RCTs) concluded that regional anticoagulation with citrate was superior to heparin [4, 5], it is not standard care yet. The most common concerns relate to the complexity of the technique, the risk of serious metabolic complications, concern about citrate accumulation in patients with liver disease and potentially higher financial costs compared to heparin. It is not uncommon that the beneficial effect of a therapy is confirmed through well-conducted RCTs but its adoption into routine clinical practice is slow [6]. This is particularly relevant for complex therapies which may be effective in the controlled research setting but are difficult to implement in day-to-day practice where less experienced healthcare workers are faced with different pressures and constraints. For this reason it is important to evaluate a proven therapy and to assess its efficacy in routine clinical care. In early 2008, we were the first intensive care unit in the UK to introduce regional anticoagulation with citrate. At that time, a decision was made to limit the use of citrate to patients with either contraindications to unfractionated heparin or repeated episodes of premature filter clotting on heparin. The objective of the following analysis was to evaluate the performance of citrate anticoagulation in routine clinical practice outside the research setting by comparing circuit patency with citrate, heparin and epoprostenol anticoagulation. M e t h o d s Setting Guy s & St Thomas NHS Foundation Hospital is a tertiary care centre with 49 critical care beds where CRRT can be provided. The critical care team is responsible for the initiation and provision of RRT. CR RT P ro t o c o l In 2008 2009, we used continuous veno-venous haemofiltration (CVVH) and continuous veno-venous haemodialysis (CVVHD) for patients with severe acute kidney injury. Our firstline anticoagulant was unfractionated heparin which was routinely administered via the circuit, unless the patient needed systemic heparin for other medical indications. As described previously, our heparin algorithm does not have a target-activated partial thromboplastin time ratio (APTTr) but instead allows the nursing staff to adjust the dose of heparin as guided by the condition of the patient and RRT performance provided the systemic APTTr is kept 2 to prevent over-anticoagulation [7]. We performed CVVH in pre- and post-dilution mode targeting a filtration fraction <20%. If circuits clotted prematurely, the proportion of pre-dilution was gradually increased, sometimes up to 100%. Our citrate-based CRRT protocol consists of CVVHD and is based on the algorithm described by Morgera et al. [8]. The treatment is nurse-led with close supervision by specialist critical care nurses and the medical team. We use MultiFiltrate machines, Ultraflux AV1000S hemofilters (polysulfone, 1.8 m 2, cutoff 30 kda) and Ci-Ca CVVHD 1000 MultiFiltrate circuits (Fresenius Medical Care, Bad Homburg, Germany). The dialysis solution is calciumfree Ci-Ca Dialysate K2 (Na + 133 mmol/l, K + 2.0 mmol/l, Mg 2+ 0.75 mmol/l, Cl 116.5 mmol/l, HCO 3 20 mmol/l). 4% trisodium citrate and CaCl (100 mmol/l) solutions are used for citrate and CaCl 2 administration, respectively. Predefined blood flow and dialysis solution rates are prescribed according to the patient s estimated ideal body weight in a 3: 1 ratio. Infusion rates are adjusted according to post-filter and systemic [Ca i ] aiming for post-filter [Ca i ] 0.25 0.35 mmol/l and systemic [Ca i ] 1.12 1.20 mmol/l. In 2008 2009, circuits were routinely changed after 72 h if no circuit failure occurred. For patients with contraindications to heparin and citrate, we used systemic epoprostenol infusion at a rate of 5 ng/kg/min, to be increased to 10 ng/kg/min if necessary to prevent premature clotting. An a ly s i s We retrospectively analysed data on circuit patency of all circuits used during a 12-month period between September 2008 and August 2009. The type of anticoagulant and the reasons for circuit discontinuation were recorded. We differentiated between premature filter clotting, elective discontinuation (i.e. surgery, diagnostic procedures, recovery of renal function, change of line) and waste (i.e. human error, machine errors). For the statistical analysis we compared circuits anticoagulated with heparin (via circuit or systemically), citrate, epoprostenol and combinations of different anticoagulants and recorded mean values and standard deviation (SD). Heparin was used as the reference category. Proportional hazard Cox model was used to evaluate the effect of anticoagulant type on time of clotting. We modelled competing risks using a latent survivor time approach [9]. Statistical analysis was performed using Stata SE version 11.0 (STAT Corp., USA). E t h i c s The need for individual informed consent was waived because this was a retrospective analysis of data collected prospectively for routine care, and there was no breach of privacy or anonymity (UK National Research Ethics Service). R e s u l t s Between September 2008 and August 2009, 309 patients were treated with CRRT of whom 21 patients had two separate episodes of acute kidney injury where CRRT 120 Tovey/Dickie/Gangi/Terblanche/ McKenzie/Beale/Treacher/Ostermann

was required (n = 2,059 circuits). Their mean age was 65.7, and 63.8% were male. Different methods were used to keep the circuits patent: unfractionated heparin (42.3% of circuits), epoprostenol (23.0%), citrate (14.7%), combinations of different anticoagulants (14.6%) and no anticoagulation (4.7%). Data was missing for 12 circuits. The mean duration of circuit patency was 19.4 h (SD 17.8) with heparin, 34.7 h (SD 28.3) with citrate and 18.8 h (SD 16.0) with epoprostenol. In patients treated with unfractionated heparin via the circuit, the mean APTTr was 1.82 (SD 0.28). Premature clotting was the most common reason for circuit discontinuation among circuits anticoagulated with heparin, epoprostenol or combinations of different anticoagulants ( fig. 1 ). Among circuits anticoagulated with citrate, discontinuation was more likely to be elective (line changes, scans, etc.). A total of 108 circuits (5.2%) were discontinued for other reasons (i.e. access problems, machine failure, human error). Hazard regression analysis of the remaining 1,939 circuits showed significantly better circuit survival with citrate ( table 1 ; fig. 2 ). Changing from heparin to citrate decreased the risk of premature circuit clotting by 76.5%. There was no significant difference between heparin and epoprostenol. D i s c u s s i o n % Premature clotting Elective discontinuation Waste 70 60 50 40 30 20 10 0 Heparin Epoprostenol Citrate Combination Rx Fig. 1. Reasons for circuit discontinuation. Survival Heparin 1.0 Epoprostenol Citrate 0.8 Other combinations 0.6 0.4 Color version available online Several RCTs and meta-analyses have shown that citrate achieves good circuit patency and is superior to heparin [3 6, 10 15]. RCTs are regarded as the gold standard for establishing causation between a therapy and an outcome. In a well-conducted RCT, the treatment that is found to have a beneficial effect is usually considered efficacious. Yet, even when an RCT demonstrates benefit for a particular therapy, its role in routine clinical practice is not always certain and incorporation within the larger healthcare environment may be hampered by several factors, including complexity, manpower and financial constraints [6]. Our observational data demonstrate that citrate remains superior to heparin in routine clinical practice when administered and monitored by trained critical care nursing staff. In our opinion, this is an important result, especially in light of the fact that both, the Renal Association UK and the Kidney Disease Improving Global Outcomes (KDIGO) group include regional anticoagulation with citrate in their official guidelines [16, 17]. In fact, the 0.2 0 Fig. 2. Circuit survival per anticoagulant. Table 1. Risk of premature circuit clotting AnticoagulantPredictor variable hazard rate 1 95% CI p value Heparin Citrate Epoprostenol Other 0 12 24 36 48 60 72 Hours 1 0.235 1.083 0.568 0.184 0.299 0.900 1.302 0.466 0.692 <0.001 0.396 <0.001 Cox proportional hazard regression. 1 Hazard rate per 1 unit change, as appropriate for each variable. Citrate for Continuous RRT in Clinical Practice 121

most recent KDIGO guideline recommends citrate as the first-line anticoagulant for routine CRRT [16]. Circuit patency is critical when administering CRRT. Premature filter clotting can be a major problem and result in inadequate delivery of RRT, increased blood loss, financial costs and waste of valuable nursing time. Our data confirm a significantly lower risk of filter clotting with citrate. However, as demonstrated, premature filter clotting cannot be completely avoided even with citrate. A variety of studies using different protocols found diverging circuit survival times [10 15]. In most studies, filter life was better when using citrate compared to heparin [11, 13 15] while others found no difference [10, 12]. It is obvious that the efficacy of regional anticoagulation with citrate depends on meticulous attention to detail and adherence to the protocol. Failure to achieve post-filter [Ca i ] <0.35 mmol/l will contribute to inadequate anticoagulation within the filter and premature filter clotting. Undoubtedly, the success of citrate-based anticoagulation ultimately relies on an understanding of the principles of the technique and close attention to detail. However, premature circuit clotting can occur despite optimal postfilter ionized calcium concentrations. The mechanisms are not fully understood and require further research. It is important to acknowledge some limitations of our analysis. Firstly, we report the data obtained in a single centre. It is possible that our positive experience with citrate is due to the fact that our staff were well trained before the technique was introduced in our department. However, we acknowledge that our mean filter running time of 32 h was lower compared to more recent data reported by Morgera et al. [18] who observed a median circuit survival of 61.5 h (interquartile range 34.5 81.8) in patients treated with citrate-based CVVHD. Similarly, we observed that 34% of filters anticoagulated with citrate clotted prematurely compared to 5% in the study by Morgera et al. [18]. Although this may be related to a limited experience during the first year of us using citrate, we note that our data are better compared to results by Fealy et al. [mean circuit life with citrate 16.3 h (SD 2.4)] and similar to findings by Oudemans-van Straaten et al. [mean circuit patency 28.5 h (SD 8.8)] [4, 19, 20]. Secondly, at the time of the study period (2008 2009), citrate was reserved for patients who either had contraindications to heparin or had had several episodes of premature circuit clotting with heparin. As a result, the groups were different. Patients treated with heparin, citrate and epoprostenol were heterogenous and varied in their comorbidity and risk of bleeding. This is also the reason why we were not able to make valid comments related to the use of blood products, metabolic complications and financial costs. Thirdly, due to limited data at the time, we did not use citrate in patients with liver disease. We are aware of recent studies showing that regional anticoagulation with citrate is possible in liver patients and our practice has changed accordingly since 2009 [21, 22]. Fourthly, our heparin algorithm only includes a maximum APTTr value but not a target value [7]. The aim of this strategy is to achieve circuit patency without unnecessary or excessive patient anticoagulation and is in line with expert opinion and current practice in the UK [23, 24]. However, we cannot exclude that circuits anticoagulated with heparin would have lasted longer if the dose of heparin had been increased. Finally, the majority of non-citrate circuits (>95%) were run in CVVH mode while patients on citrate received CVVHD. We acknowledge that this may have affected our results since haemofiltration requires higher blood flows and is associated with more hemoconcentration within the filter as a consequence of ultrafiltration [25]. In conclusion, our data confirm that citrate performs well in routine clinical practice and offers the advantage of reduced risk of clotting compared to heparin or epoprostenol. References 1 Uchino S, Bellomo R, Morimatsu H, Morgera S, Schetz M, Tan I, Bouman C, Macedo E, Gibney N, Tolwani A, Oudemans-van Straaten H, Ronco C, Kellum JA: Continuous renal replacement therapy: a worldwide practice survey. The beginning and ending supportive therapy for the kidney (BEST kidney) investigators. Intensive Care Med 2007; 33: 1563 1570. 2 Oudemans-van Straaten HM, Ostermann M: Bench-to-bedside review: citrate for continuous renal replacement therapy, from science to practice. Crit Care 2012; 16: 249. 3 Hetzel GR, Schmitz M, Wissing H, Ries W, Schott G, Heering PJ, Isgro F, Kribben A, Himmele R, Grabensee B, Rump LC: Regional citrate versus systemic heparin for anticoagulation in critically ill patients on continuous venovenous haemofiltration: a prospective randomized multicentre trial. Nephrol Dial Transplant 2011; 26: 232 239. 4 Zhang Z, Hongying N: Efficacy and safety of regional citrate anticoagulation in critically ill patients undergoing continuous renal replacement therapy. Intensive Care Med 2012; 38: 20 28. 5 Wu MY, Hsu YH, Bai CH, Lin YF, Wu CH, Tam KW: Regional citrate versus heparin anticoagulation therapy: a meta-analysis of randomized controlled trials. Am J Kidney Dis 2012; 59: 810 818. 6 Nallamothu BK, Hayward RA, Bates ER: Beyond the randomized clinical trial: the role of effectiveness studies in evaluating cardiovascular therapies. Circulation 2008; 118: 1294 1303. 122 Tovey/Dickie/Gangi/Terblanche/ McKenzie/Beale/Treacher/Ostermann

7 Ostermann M, Dickie H, Tovey L, Treacher D: Heparin algorithm for anticoagulation during continuous renal replacement therapy. Crit Care 2010; 14: 419. 8 Morgera S, Schneider M, Slowinski T, Vargas- Hein O, Zuckermann-Becker H, Peters H, Kindgen-Milles D, Neumayer HH: A safe citrate anticoagulation protocol with variable treatment efficacy and excellent control of the acid-base status. Crit Care Med 2009; 37: 2018 2024. 9 Hougaard P: Frailty models for survival data. Lifetime Data Anal 1995; 1: 255 273. 10 Oudemans-van Straaten H: Citrate anticoagulation therapy in the critically ill. Blood Purif 2010; 29: 191 196. 11 Kutsogiannis DJ, Gibney RT, Stollery D, Gao J: Regional citrate versus systemic heparin anticoagulation in critically ill patients. Kidney Int 2005; 67: 2361 2367. 12 Betjes MG, van Oosterom D, van Agteren M, van de Wetering J: Regional citrate versus heparin anticoagulation during venovenous hemofiltration in patients at low risk for bleeding: similar hemofilter survival but significantly less bleeding. J Nephrol 2007; 20: 602 608. 13 Morgera S, Scholle C, Voss C, Haase M, Vargas-Hein O, Krausch D, Meizer C, Rosseau S, Zuckermann-Becker H, Neumayer HH: Metabolic complications during regional citrate anticoagulation in continuous veno-venous hemodialysis: single-center experience. Nephron Clin Pract 2004; 97:c131 c136. 14 Durao MS, Monte JC, Batista MC, Oliveira M, Iizuka IJ, Santos BF, Pereira VG, Cendoroglo M, Santos OF: The use of regional citrate anticoagulation in acute kidney injury. Crit Care Med 2008; 36: 3024 3029. 15 Monchi M, Berghmans D, Ledoux D, Canivet JL, Dubois B, Damas P: Citrate versus heparin for anticoagulation in continuous venovenous hemofiltration: a prospective randomized study. Intensive Care Med 2004; 30: 260 265. 16 Clinical Practice Guidelines for Acute Kidney Injury 2012. http://www.kdigo.org/clinical_ practice_guidelines/aki.php. 17 UK Renal Association: Clinical Practice Guidelines Acute Kidney Injury. http://www. renal.org/clinical/guidelinessection/ AcuteKidneyInjury.aspx. 18 Morgera S, Schneider M, Slowinski T, Vargas- Hein O, Zuckermann-Becker H, Peters H, Kindgen-Milles D, Neumayer HH: A safe citrate anticoagulation protocol with variable treatment efficacy and excellent control of the acidbase status. Crit Care Med 2009; 37: 2018 2024. 19 Oudemans-Van Straaten HM, Bosman RJ, Koopmans M, van der Voort PH, Wester JP, van der Spoel JI, Dijksman LM, Zandstra DF: Citrate anticoagulation for continuous venovenous hemofiltration. Crit Care Med 2009; 37: 545 552. 20 Fealy N, Baldwin I, Johnstone M, Egi M, Bellomo R: A pilot randomized controlled crossover study comparing regional heparinization to regional citrate anticoagulation for continuous venovenous hemofiltration. Int J Artif Organs 2007; 30: 301 307. 21 Schultheiss C, Saugel B, Philip V, Thies P, Noe S, Mayr U, Haller B, Einwachter H, Schmid RM, Huber W: Continuous venovenous hemodialysis with regional citrate anticoagulation in patients with liver failure: a prospective observational study. Critical Care 2012; 16:R162. 22 Patel S, Wendon J: Regional citrate anticoagulation in patients with liver failure time for a rethink? Crit Care 2012; 16: 153. 23 Bellomo R, Baldwin I: Anticoagulation; in Kellum J, Bellomo R, Ronco C (eds): Continuous Renal Replacement Therapy. Oxford, Oxford University Press, 2010, pp 135 140. 24 Gatward JJ, Gibbon GJ, Wrathall G, Padkin A: Renal replacement therapy for acute renal failure: a survey of practice in adult intensive care units in the United Kingdom. Anaesthesia 2008; 63: 959 966. 25 Ricci Z, Ronco C, Bachetoni A, D Amico G, Rossi S, Alessandri E, Rocco M, Pietropaoli P: Solute removal during continuous renal replacement therapy in critically ill patients: convection versus diffusion. Crit Care 2006; 10:R67. Citrate for Continuous RRT in Clinical Practice 123