performance properties of modern locksolutions: … · taurolock™, killing curves, and schilcher...

PERFORMANCE PROPERTIES OF

MODERN LOCKSOLUTIONS:

COMPARISON OF

CITRATE SOLUTIONS

WITH

TAUROLOCK™/TAUROLOCK™-HEP500

AS LOCK SOLUTIONS

IN

CATHETERS AND PORTS Status April 2015

TAUROLOCK™ vs. HIGLY CONCENTRATED CITRATE ( ≥ 30%)

LOCK SOLUTIONS ___________________________________________________________________________________________________________

Status April 2015 Page 1 of 2

PERFORMANCE PROPERTIES OF MODERN LOCK SOLUTIONS (HIGHLY CONCENTRATED CITRATE VS TAUROLOCK™ )

A) Effective Prevention of Infections B) Maintenance of the Access Device’s Patency C) Safe Application (Even in Case of incorrect Application)

Ad A) Effective Prevention of Infections The German Dialysis Standard 2014 recommends the use of highly concentrated citrate (30% or 45%) or taurolidine-citrate solutions since they are antimicrobial, and the rate of catheter-related bacteraemias can be reduced. Although the bactericidal (respectively bacteriostatic) activity of highly concentrated citrate solutions is in vitro much weaker and more specific to single organisms than TauroLock™ (see attachment, brochure TauroLock™, killing curves, and Schilcher 2015) it seems to be efficient enough to reduce catheter related infections in clinical practice. Ad B) Maintenance of the Access Device´s Patency In a study comprising 232 patients, Power et al. (see attachment) showed that the application of a citrate block (46,7%) required an increased use of urokinase which was twice as high compared to the heparin block. This might be due to the reduction of the locking volume on account of patients’ complaints which then resulted in a reduced filling of the catheter tip. The high concentration of citrate may also lead to protein precipitation with concomitant blockage of the catheter (Schilcher et al, 2012). Alternatively it should be discussed whether the lack of heparin results in an increased adhesion of fibrin or fibrous material to the inner lumen. This hypothesis is supported by the findings of Solomon et al., 2010 (see attachment) which confirm the results of Power regarding the increased use of urokinase with heparin-free taurolidine-citrate (TauroLock™) in dialysis patients. The addition of 500 IU/mL of heparin (TauroLock™-Hep500) reduced the need of thrombolytic intervention to a level comparable to heparin 5000 IU/mL. Apparently – at least in dialysis – the use of small concentrations of heparin in lock solutions provides patency of the device. Ad C) Safe Application (even in case of inaccurate application) c1) Cardiac Risks: Highly concentrated citrate solutions used as catheter lock solutions are banned in the USA (see attachment, FDA warning letter). In a clinical unit in the Netherlands, the use of highly concentrated citrate (30%) as a lock solution was responsible for two cardiac arrests in the same patient within 24 hours. The over-instillation though had just been ≤ 1 ml per lumen (Punt et al., see attachment). Therefore, highly concentrated citrate used as a catheter block is only recommended by the manufacturer if the exact catheter volume is known. These risks resulted in a limited use of highly concentrated citrate only by trained personnel only according to the German Dialysis Standard 2014. c2) Embolic Risks: Willicombe et al., 2009 (attachment) reported embolic complications while using highly concentrated citrate (46,7%) as a lock solution. There were eight cases of pulmonal or cerebral embolic events, three of which are well documented. Schilcher et al., 2012 (see attachment) reported the interaction between whole blood and citrate and detected protein precipitations during their interaction in vitro at levels of >12% citrate. In vivo, protein precipitates were found in all ten catheters applied with 20% and 46,7% citrate lock solution, but not when filled with 4% and 10% citrate. The authors also showed that these protein precipitates blocked a 20 µm mesh filter resembling small vessels. Therefore, hypertonic citrate locks are potentially dangerous and may be the decisive cause for reported embolic events and the increased use of Urokinase to deblock the catheter.

TAUROLOCK™ vs. HIGLY CONCENTRATED CITRATE ( ≥ 30%)

LOCK SOLUTIONS ___________________________________________________________________________________________________________

Status April 2015 Page 2 of 2

German Dialysis Standard 2014

Deutsche Gesellschaft für Nephrologie

in Cooperation with the Verband Deutsche Nierenzentren e.V.

and with the Gesellschaft für Pädiatrische Nephrologie (GPN)

Excerpt

B14.2 Prevention of Infection in Central Venous Catheters

… Between dialysis treatments the central venous access device may be blocked using a diluted heparin solution. Heparin, however, does not have any antibacterial properties. Antibacterial lock solutions should therefore be preferred, which reduce the rate of catheter-related bacteraemias considerably. The use of antibiotics can not be recommended due to the potential development of resistance. Alternatively, highly concentrated citrate (30% or 45%) or taurolidine-citrate solutions can be used. Due to the risk of severe cardiac arrhythmias highly concentrated citrate must be strictly administered by trained staff according to the instructions of the manufacturer. B16.3 Hygiene measures regarding exceptional pathogens

Note to Chapter B16.3 of the German Dialysis Standard: The commission for hospital hygiene and infection prevention at the Robert-Koch-Institut (KRINKO) recommends the application of the Hygieneguideline 2008 (Bundesgesundheitsblatt 2014, 57:696, 719, 721). Thereby the relevant paragraph that is now identically included in the German Dialysis Standard 2014 becomes binding according to the national law of infection protection (Infektionsschutzgesetz §23 (3)).

AntimicrobiAl cAtheter lock system to provide pAtency And infection control

0123

Parenteral nUtrItIOn

Gabe

6,2 1,6

Chambrier

6,58 1,09

Cullis

5,71 0,43

Infections per 1000 catheter daysHeparin taurolock™

taurolock™ prevents catheter infections:

OnCOlOGY

Schröder

1,4 0,4*

Simon

2,3 0,5


Querfeld

1,3 0,3

*Tau

roLo

ck™

-HEP

100

DIalYsIs

Solomon

3,25 1,22*


Betjes

2,1 0

Allon

5,6 0,5

Taylor

5,2 0,6

*Tau

roLo

ck™

-HEP

500

Prophylaxis against catheter related bloodstream infections: Central venous catheters (CVC) are used as short or long term vascular access devices in hemodialysis, on-cology, ICU and total parenteral nutrition. High risks for CVC malfunction are catheter related infections (CrI). these infections may be triggered by microbial colo-nization of the catheter and the microorganisms can spread from here to the bloodstream. CrI may develop septic symptoms which require the immediate removal of the catheter.

taurolock™ catheter lock solutions do not contain an-tibiotics and were developed for prophylactic use. they reduce catheter related infections significantly (~ 90%).

the combination of citrate (4%) with (cyclo)-taurolidine and heparin/urokinase has excellent anticoagulative and anti-microbial properties also against resistant mi-croorganisms like Mrsa und Vre.

therefore taurolock™ is recommended in different guidelines such as the Hygiene Guidelines completing the German Dialysis standard, the guidelines from the German society of applied Hygiene in Dialysis and the evidence-based recommendations of the German society for Paediatric Oncology and Hematology (GPOH).

Prophylaxis against biolog ical occlusion in the catheter:the taurolock™ Catheter lock system contains a threefold prophylaxis against occlusion in the catheter: all locking so-lutions contain 4% citrate as anticoagu-lant. this concentration removes calcium safely and effectively from the clotting cascade.

the optional use of low concentrated heparin supports an additional anti-coagulative effect via binding to an-tithrombin. the prophylactic use of taurolock™-U25.000 (which contains 25.000 IU of urokinase) achieves the best prophylaxis against occlusion by preven-tion of biological clotting.

the decision which locking solution is most adequate depends on the individual patient situation. the alternative use of different locking solutions in the same catheter (e.g. taurolock™-HeP500, taurolock™-U25.000) is possible.

30% Citrate

10 lo

g cf

u/m

L

time in days

0 1 2 3 4 24

10 lo

g cf

u/m

L

time in hours

legend

*detection limit(10 cfu/ml)

Heparin

10 lo

g cf

u/m

L

0 1 2 3 4 24

time in hours

10 lo

g cf

u/m

L

0 1 2 3 4 24

time in hours

46,7% Citrate

10 lo

g cf

u/m

L

time in days

If used prophylactically, taurolock™ prevents the de-velopment of a biofilm on the surface of the catheter lumen:

TauroLock™5 months implanted –

No colonization

Heparin Lock – 7 months implanted – S. epidermidis biofilm

covers surface completely

taurolock™ is bactericidal and fungicidal within 2 hours:

Clearly superior in comparison to the activity of Citrate and Heparin:

0123

Instillation of taurolock™

Follow the manufacturer’s instructions that accompany the particular vascular access product utilized. specific catheter lock volumes are associated with each device.

1. Flush the device with 10 ml of saline.

2. Withdraw taurolock™ from the container using an appropriate syringe.

3. Instill taurolock™ slowly (not more than 1 ml per second, infants and children less than two years of age not more than 1 ml per 5 second) into the access device in a quan-tity sufficient to fill the lumen completely. Consult the manufacturer’s instructions for the specific fill volume or specify fill volume during implantation. the volume has to be strictly respected. taurolock™ will remain inside the access device until the next treat-ment.

4. If aspiration of taurolock™ is needed and possible, it should be withdrawn from the port/catheter and discarded prior to initiation of next treatment.

5. Flush the device with 10 ml of saline.

taurolock™ is safe:the concentration of 4% citrate in taurolock™ is safe and efficient - according to the recommendation of the FDa (ref.: FDa Warning letter, april 2000).

no hypocalcaemic effects are observed in contrast to highly concentrated citrate solutions (30% resp. 46,7%) e.g. arrhythmia, cardiac arrest*, emboli**, tingling fingers and metallic taste***.

taurolock™ is biocompatible and non toxic.

In contrast to highly concentrated citrate there is no protein precipitation if using taurolock™****.

* Punt, C.D., Boer, W.e. Cardiac arrest following injection of concentrated trisodium citrate, Clinical nephrology, 2008, 69: 117-118.

** Willicombe, M.K., Vernon, K., Davenport, a. embolic Complications From Central Venous Hemodialysis Catheters; Used With Hypertonic Citrate locking solutions, american Journal of Kidney Diseases, 2010, 55: pp 348 - 351.

*** Polaschegg, H.-D., sodemann, K. risks related to catheter locking solutions containing concentrated citrate, nephrol. Dial. transplant. 2003, 18: 2688-2690.

**** schilcher, G. Polaschegg H.D. et al. Hypertonic trisodium Citrate Induces Protein Precipitation in Hemodialysis Catheters, selected asn Meeting abstracts, 2011

Product selection for application

Dialysis

Oncology

Parenteral Nutrition

Product

Manufacturer:

tauroPharm GmbHJägerstraße 5aD-97297 Waldbüttelbrunntel.: +49 931 304299-0Fax: +49 931 304299-29 ISO13485

85100/08/14

taurolock™ catheter lock solutions are available in different containers:

Ampoule (10 x 3 mL)

Ampoule (10 x 5 mL)

Vial (100 x 10 mL)

Vial (5 x 5 mL)

Product

WWW.TAUROLOCK.COM

1. GUIDELINES AND RECOMMENDATIONS1.1. Hygiene Guidelines Completing ‘GERMAN DIALYSIS STANDARD 2006‘ German Workgroup for Clinical Nephrology in Cooperation with the Verband Deutsche Nierenzentren der

DD nÄ e.V. and the Society for Pediatric Nephrology: Hygiene Guidelines 2008 completing the German Dialysis Standard 2006.1.2. Diagnosis, prevention and treatment of haemodialysis catheter-related bloodstream infections (CRBSI): a position statement of European Renal Best Practice (ERBP)

R. Vanholder, B. Canaud, R. Fluck, M. Jadoul, L. Labriola, A. Marti-Monros, J. Tordoir, W. Van Biesen, NDT Plus (2010) 3: 234-2461.3. Vascular Access for Haemodialysis Renal Association (United Kingdom), R. Fluck, M. Kumwenda (2011)1.4. Guidelines for Applied Hygiene in Dialysis German Society for Applied Hygiene in Dialysis e.V., DGaHD 2013, Chapter 9.5.8 Lock Solutions for CVC and Port-Systems, 3rd Edition1.5. National Kidney Foundation (NKF): KDOQI Guidelines KDOQI Guideline, Guideline 7, update 2006.1.6. Evidence-based Recommendations for the Use of Permanent CVADs in Paediatrics. A. Simon, Society of Paediatric Hematology and Oncology, GPOH 2013, 3rd Edition1.7. S3 – Guideline of the German Society for Nutrional Medicine (DGEM) in Cooperation with GESKES and AKE: Nutritional support in the homecare and out patient sectors

S.C. Bischoff et al., Aktuel Ernährungsmed, 2013 e101 – e1541.8. Guidelines for the Prevention of Intravascular Catheter-related Infections, 2011 CDC, Center of Disease Control, USA, 20112. PUBLICATIONS: PROPHYLAXIS OF INFECTION IN DIALYSIS2.1. A Meta-analysis of Hemodialysis Catheter Locking Solutions in the Prevention of Catheter-Related Infection

Y. Jaffer, N. M. Selby, M. W. Taal, R. J. Fluck, and C. W. McIntyre, Am J Kidney Dis 51:233-241.2.2. Prevention of dialysis catheter-related sepsis with a citrate–taurolidine-containing lock solution M. G. H. Betjes and M. van Agteren, Nephrol Dial Transplant, 2004, 19:1546-

1551. Department of Internal Medicine, Division of Nephrology, Erasmus Medical Center, Dijkzigt Rotterdam, 2.3. Observational Study of Need for Thrombolytic Therapy and Incidence of Bacteremia using Taurolidine-Citrate-Heparin (TCH), Taurolidine-Citrate (TC) and Heparin Catheter

Locks in Patients Treated with Hemodialysis L. R. Solomon, J. S. Cheesbrough, R. Bhargava, N. Mitsides, M. Heap, G. Green, P. Diggle, Sem Dial 2011.2.4. A Randomized Double-Blind Controlled Trial of Taurolidine-Citrate Catheter Locks (vs. Heparin (5000 IU/mL) for the Prevention of Bacteremia in Patients Treated With He-

modialysis L. R. Solomon, J. S. Cheesbrough, L. Ebah, T. Al-Sayed, M. Heap, N. Millband, D. Waterhouse, S. Mitra, A. Curry, R. Saxena, R. Bhat, M. Schulz, P. Diggle, American Journal of Kidney Desease, Vol 55, No 6 (June), 2010; pp 1060 - 1068.

2.5. Prophylaxis against Dialysis CatheterRelated Bacteraemia with a Novel Antimicrobial Lock Solution M. Allon, Clin. Infect Dis 2003, 36:1539-1544.2.6. Dialysis Catheter-Related Bacteraemia: Treatment and Prophylaxis M. Allon, MD, American Journal of Kidney Diseases, 2004, 44, 779-791.2.7. Two Years´ Experience with Dialock and CLS™ (A New Antimicrobial Taurolidine-Citrate Lock Solution) K. Sodemann, H.-D. Polaschegg, B. Feldmer Blood Purif 2001;19:251-254.2.8. The Targeted Use of TauroLock in Reduction of Episodes of Line Sepsis in the „High Risk“ Haemodialysis Population

M. A. Vernon, J. Goddard, Poster during British Renal 2006, Harrogate P205 (RA6432)2.9. Preventing infections of central venous catheters with a taurolidine/citrate solution

O. Kramenko, Western Galilee Hospital, Nahariya, Israel, Presentation at EDTNA/ERCA Congress 2006, Madrid.2.10. A New Haemodialysis Catheter-Locking Agent reduces infections in Haemodialysis Patients C. Taylor, J. Cahill, M. Gerrish, J. Little, Journal of renal Care 34 (3), 116-120.2.11. Approaches to Prolong the Use of Uncuffed Hemodialysis Catheters: Results of a Randomized Trial V. Filiopoulos, D. Hadjiyannakos, I. Koutis, S. Trompouki, T. Micha, D.

Lazarou, D. Vlassopoulos; Department of Nephrology, Am J Nephrol 2011; 33:260-268.3. PUBLICATIONS: PROPHYLAXIS OF INFECTION IN ONCOLOGY3.1. Central Venous Catheters and Catheter Locks in Children with Cancer: a Prospective Randomized Trial of Taurolidine versus Heparin

M. Møller Handrup, J. Kjølseth Møller, H. Schrøder, Pediatric Blood Cancer DOI 10.1002/pbc3.2. Randomized controlled trial of taurolidine citrate versus heparin as catheter lock solution in paediatric patients with haematological malignancies

M. J. Dümichen, K. Seeger, H. N. Lode, J. S. Kühl, W. Ebell, P. Degenhardt, M. Singer, C. Geffers, U. Querfeld, J. Hospital Inf., 80 (2012) 304-309.3.3. Taurolidine-citrate lock solution (TauroLock) significantly reduces CVAD-associated gram-positive infections in paediatric cancer patients

A. Simon, R. A. Ammann, G. Wiszniewsky, U. Bode, G. Fleischhack, M. M. Besuden, BMC Infectious Diseases 2008, 8:102.3.4. Treatment of long-term catheter-related bloodstream infections with a taurolidine block: a single cancer center experience

G.-M. Haag, A.-K. Berger, D. Jäger, J. Vasc. Access 2011, DOI: 10.5301/JVA.2011.6265.3.5. Taurolidine is effective in the treatment of central venous catheter-related bloodstream infections in cancer patients

M. Koldehoff, J. L. Zakrzewski, Int. J. Antimicrobial Agents 24 (2004), 491-495.3.6. First Report World-Wide of Clinical Use of Taurolidine – 4% Citrate Catheter Lock Solution To Treat an Intravascular Catheter Colonised With a Mycobacteria; With a Highly

Successful Outcome T. A. Collyns, et al, Leeds Teaching Hospitals NHS Trust, Leeds, United Kingdom, Posterpresentation 47th ICAAC, Chicago.4. PUBLICATIONS: PROPHYLAXIS OF INFECTION IN PARENTERAL NUTRITION4.1. Significant Reduction in Central Venous Catheter–related Bloodstream Infections in Children on HPN After Starting Treatment With Taurolidine Line Lock

H.-P. Chu, J. Brind, R. Tomar, S. Hill, JPGN 55 (2012) 403-407.4.2. Taurolidine Lock is highly effective in preventing catheter-related bloodstream infections in patients on home parenteral nutrition: A heparin-controlled prospective trial

T. M. Bisseling, M. C. Willems, M. W. Versleijen, J. C. Hendriks, R. K. Vissers, G. J .Wanten, Clinical Nutrition 2010; 29; 464-468.4.3. Taurolidine lock solution in the secondary prevention of central venous catheter-associated bloodstream infection in home parenteral nutrition patients

A. Touré, M. Lauverjat, C. Peraldi, M. Boncompain-Gerard, P. Gelas, D. Barnoud, C. Chambrier, Clinical Nutrition (2012), doi:10.1016/j.clnu.2012.01.001.4.4. Study on the effectiveness of taurolidine lock in the secondary prevention of infections related to central venous lines in home parenteral nutrition

E. Lerebours, F. Joly, C. Chambrier, S. Schneider, D. Seguy, E. Fontaine, L. Armengol-Debeir, J. Blot, H. Roth; Nutrition clinique et métabolisme 27 (2013) O534.5. Efficacy of Taurolidine in prevention of catheter related bloodstream infections in patients on Home Parenteral Nutrition

A. Al-Amin, J. Sarveswaran, J. Wood, C. Donnellan, D. Burke. Br. J. Surgery 2012;99 83-222.4.6. Taurolidine Lock – Experience from the West of Scotland P. S. Cullis, R. F. McKee, Clinical Nutrition (2011), doi;10.1016/j.clnu.2010.12.0084.7. Effectiveness of TauroLock™ in preventing recurrent catheter-related bloodstream infections in patients on home parenteral nutrition

A. Taniguchi, J. Eastwood, A. Davidson, J. Nightingale, S.M. Gabe, Proceedings of the Nutrition Society (2009), 68 (OCE1), E585. PROVIDE PATENCY TO ACCESS DEVICES BY USING UROKINASE5.1. German Guideline for Access Devices in Haemodialysis (extract) – Recommendation of German Expert Panel

M. Hollenbeck, V. Mickley, J. Brunkwall, H. Daum, P. Haage, J. Ranft, R. Schindler, P. Thon, D. Vorwerk, Nephrologe 2009, 4, 158-176.5.2. National Kidney Foundation, KDOQI Guidelines 2000, Guidelines for Vascular Access, guideline 6, Table III-2. Protocols for Urokinase Administration 5.3. Prophylactic Urokinase in the Management of Long-Term Venous Access Devices in Children: A Children‘s Oncology Group Study

P. W. Dillon, G. R. Jones, H. A. Bagnall-Reeb, J. D. Buckley, E. S. Wiener, G. M. Haase, J Clin Oncology, 2004 (22), 2718-2723. 5.4. Review and update of the use of urokinase in the prevention and management of CVAD-related complications in paediatric oncology patients

A. Simon, U. Bode, K. Lieber, K. Beutel and G. Fleischhack, American Journal of Infection Control, Vol. 36 (1), 2008, 54-58.6. PUBLICATIONS ON ANTIBACTERIAL ACTIVITY OF TAUROLOCK - PREVENTION OF BIOFILM6.1. Antimicrobial Activity of a Novel Catheter Lock Solution

C. B. Shah, M. W. Mittelman, J. W. Costerton, S. Parenteau, M. Pelak, R. Arsenault, L. A. Mermel, Antimicrob. Agents Chemother. 2002, 46; 1674-16796.2. Activities of Taurolidine In Vitro and in Experimental Enterococcal Endocarditis C. Torres-Viera, C. Thauvin-Eliopoulos, M. Souli, P. DeGirolami, M. G. Farris, C. B. Wennersten,

R. D. Sofia, G. M. Eliopoulos, Antimicrobrial Agents and Chemotherapy 2000, 44; 1720-1724.

Nephrol Dial Transplant (2014) 0: 1–6doi: 10.1093/ndt/gft527

Original Article

Loss of antimicrobial effect of trisodium citrate dueto ‘lock’ spillage from haemodialysis catheters

Gernot Schilcher1, Daniel Schneditz2, Werner Ribitsch1, Joerg H. Horina1, Martin Hoenigl3,

Thomas Valentin3, Alexander R. Rosenkranz1 and Robert Krause3

1Clinical Division of Nephrology, Department of Internal Medicine, Medical University of Graz, Graz, Austria, 2Institute of Physiology, Medical

University of Graz, Graz, Austria and 3Section of Infectious Diseases and Tropical Medicine, Department of Internal Medicine, Medical

University of Graz, Graz, Austria

Correspondence and offprint requests to: Gernot Schilcher; E-mail: [email protected]

ABSTRACT

Background. Due to its reported antimicrobial effects, hyper-tonic citrate (46.7%) is a widely used catheter lock solution,but following instillation, citrate inevitably spills into the sys-temic circulation. This process is mainly driven by hydrauliceffects during instillation and density differences betweenblood and lock solution. Hence, in haemodialysis catheters,intra-luminal citrate concentration ranges from 0% (at the tipin catheters with side holes), 3% (between the side holes andthe highest point of the catheter) to 46.7% (at the Luer end)with possible differences in antimicrobial effects. We investi-gated in vitro the antimicrobial effect of pure citrate 46.7%,citrate 46.7% diluted with saline and blood to a net concen-tration of 3% (=citrate 3%), and of citrate-free blood, simulat-ing in vivo conditions in different catheter sections.Methods. Time–kill studies measuring the antimicrobial effectof citrate 46.7%, citrate 3% and citrate-free blood were per-formed with overnight cultures of Escherichia coli (E. coli) andStaphylococcus aureus (S. aureus).Results. Citrate 46.7% reduced the number of E. coli by 2 logunits but after 24 h, 106 CFU/mL were still present. Citrate 3%and citrate-free blood had no antimicrobial effect on E. coli.Citrate 46.7%, citrate 3% and citrate-free blood had scarceantimicrobial effect on S. aureus within 24 h.Conclusions. Spillage of catheter lock solution leading toreduced intra-luminal citrate concentrations considerablyreduces the antimicrobial effect of citrate 46.7% on E. coli. Asnone of the solutions tested had relevant antimicrobial effecton S. aureus, the antimicrobial effect of 46.7% citrate lock sol-ution in vivo has to be seriously questioned.

Keywords: catheter-related bloodstream infection, centralvenous catheter, lock solution, lock spillage, trisodium citrate

INTRODUCTION

Central venous catheters (CVCs) provide reliable venousaccess for haemodialysis and are still in common use despitethe emphatic recommendations in national guidelines favour-ing arteriovenous fistulae [1]. Anticoagulative, fibrinolytic orantimicrobial catheter lock solutions are used as prophylaxisor therapy to maintain the intra-luminal patency of CVCs inan attempt to avoid thrombosis and infection [2, 3]. Trisodiumcitrate (30–46.7%) is widely used and considered effective inreducing catheter-related bloodstream infections (CRBSI) inhaemodialysis patients [4, 5]. However, the use of hypertoniccitrate remains controversial due to reported adverse eventssuch as a decrease of ionized calcium due to chelation, whichcan cause cardiac arrhythmia, or pulmonary embolisms whenplasma proteins precipitate in the lumens of the CVC [6–11].

Instillation of the listed filling volume of lock solutions,such as citrate 46.7%, is generally believed to ‘lock’ the CVC,suggesting that the entire injected volume remains inside theCVC. Recent studies report, however, that ∼20% of the cath-eter locking volume leaks into systemic circulation at the timeof instillation [12]. This spillage during instillation is a physicalconsequence of a parabolic flow profile within cylindricaltubes, such as CVCs. Furthermore, gravity-induced loss oflock solution into the systemic circulation should be con-sidered when lock solutions with different densities comparedwith blood (e.g. trisodium citrate 46.7%) are used [13–15].

© The Author 2014. Published by Oxford University Press onbehalf of ERA-EDTA. All rights reserved.

1

NDT Advance Access published February 9, 2014 at U

niversitatsbibliothek Wuerzburg on February 18, 2014

http://ndt.oxfordjournals.org/D

ownloaded from

http://ndt.oxfordjournals.org/http://ndt.oxfordjournals.org/

Finally, in catheters with side holes, the lock solution is com-pletely washed out of the tip region immediately after instilla-tion, representing additional loss of locking anticoagulant[16]. The aforementioned flow and exchange phenomenacause inhomogeneous intra-luminal dilution of lock solutionwhich, depending on the individual design, divides the cath-eter into sections with different citrate concentrations. Theantimicrobial effect of citrate 46.7% can therefore be expectedto vary among these catheter sections. Previous in vitro studiesusing pure citrate lock solution showed an antimicrobial effectfor concentrated citrate [17]. In vivo studies, however, reportedconflicting results regarding the efficacy of concentrated citratein reducing CRBSI in haemodialysis patients [4, 18].

The objective of this in vitro study was to investigate theantimicrobial effect of pure citrate 46.7%, citrate 46.7% dilutedwith blood and saline to a net concentration of 3% (=citrate3%), and of citrate-free blood, simulating in vivo conditions indifferent sections of haemodialysis catheters ‘locked’ withcitrate 46.7%.

MATERIALS AND METHODS

The methodological approach is based on recent reports re-garding mechanisms and quantification of the lock spillagephenomenon [12, 13, 16, 19]. In brief, physical effects follow-ing instillation of trisodium citrate 46.7% into the lumens of acatheter dilute the lock solution and divide the catheter intodifferent sections corresponding to the particular citrate con-centrations (Figure 1). As the citrate concentration within the‘connector section’ remains nearly unaffected by dilution, puretrisodium citrate 46.7% was used to characterize the antimi-crobial effect in this catheter section. The residual trisodiumcitrate concentration of ∼3% within the ‘core section’ arisesfrom two successive physical effects. Firstly, the saline used toflush and clear the catheter of blood is not completely replacedwhen a Newtonian lock solution (aqueous solution) is used.This is due to the parabolic flow profile and incomplete salineremoval from the catheter wall during injection. The magni-tude of this saline dilution is in the range of 20% regardless ofinjection time and patient position (injection spillage). Sec-ondly, the density difference between trisodium citrate 46.7%and blood promotes gravity-induced seepage of citrate out ofthe catheter, which is accompanied by blood inflow into thecatheter. The end point of this process is reached after 20 min,resulting in a final citrate concentration of ∼3% (gravitationalspillage). The exchange of trisodium citrate 46.7% againstblood continues up to the highest point of the catheter, i.e. thevertex, which is usually the venous insertion point in patientswith jugular or subclavian catheters. The section boundaries inour model were set accordingly. To prepare the citrate 3% testsolution, pure trisodium citrate 46.7% was consecutivelydiluted with saline and with blood. A mixture of four partscitrate 46.7% with one part saline 0.9%, corresponding to theinitial dilution effect of 20% (injection spillage), was the sourceof further dilution with blood (gravitational spillage) to a netconcentration of 3% trisodium citrate. In haemodialysis cath-eters with side holes, the region between the tip and the most

distal side hole, the ‘tip section’, contains citrate-free blood fol-lowing complete wash-out of lock solution within secondsafter instillation. The antimicrobial effect of pure citrate 46.7%,citrate 3% and citrate-free blood thus was investigated bytime–kill studies adopted from previous literature [17]. Briefly,Staphylococcus aureus ATCC 29213 and Escherichia coli ATCC35218 were cultured overnight in brain heart infusion brothand diluted with each of the test solutions to a final concen-tration of 108 CFU/mL. At the start of the microbiologicaltests and at 1, 3 and 24 h, 100 µL of the suspensions wereplated on chocolate agar (S. aureus) or MacConkey blood agarplates (E. coli) (Biomerieux, Marcy l’Etoile, France) and incu-bated for 24 h at 37°C in ambient air. Colonies were thencounted and recalculated to CFU/mL. All tests were performedin duplicate. Saline 0.9% (Fresenius Kabi AG, Bad Homburg,Germany) and blood from a healthy adult male donor,

F IGURE 1 . Section model of catheters used with trisodium citrate46.7%. According to the intra-luminal citrate concentration, the cath-eter can be divided into different sections (dotted lines). ‘a’ denotesthe highest point (vertex) of the catheter representing a borderbetween the sections, usually the venous insertion point in patientswith subclavian or jugular catheters. Under specific influences (cath-eter compression, certain patient positions or open clamps), thisborder may be lateralized towards the Luer end. The connectorsection (orange colour) contains nearly pure citrate 46.7%. The coresection (yellow colour) contains a mixture of saline, blood and citrate46.7%, resulting in an effective citrate concentration of ∼3%. The tipsection (red colour), only present in catheters with side holes, con-tains citrate-free blood due to the immediate wash-out effect of locksolution after instillation.

ORIG

INALARTIC

LE

G. Schilcher et al. 2

at Universitatsbibliothek W

uerzburg on February 18, 2014http://ndt.oxfordjournals.org/

Dow

nloaded from


provided via sterile venipuncture technique into blood collec-tion tubes (Vacuette®, Greiner Bio-One, Austria), was used fordilution of trisodium citrate 46.7% (Department of HospitalPharmacy, Medical University of Graz, Austria). Written in-formed consent was obtained from the blood donor. Haema-tocrit, albumin and total protein of the blood sample weremeasured. The study was approved by the Ethics Committeeof the Medical University of Graz, Austria.

RESULTS

Compared with baseline, the pure citrate 46.7% lock solutionreduced the number of E. coli by 2 log units but, after 24-hincubation, 106 CFU/mL were still present (Figure 2). Citrate3% and citrate-free blood had no antimicrobial effect on E. coli(Figure 2). Citrate 46.7%, citrate 3% and citrate-free bloodbarely reduced the number of S. aureus within 24 h of incu-bation (Figure 3). Sample characteristics of blood used forcitrate dilution were haematocrit 0.42, albumin 4.5 g/dL andtotal protein 6.5 g/dL.

DISCUSSION

The current standard for in vitro evaluation of the antimicro-bial effect of lock solutions focuses solely on the pure lock sol-ution [20, 21]. This might be inadequate when Newtonianfluids (aqueous solutions) and fluids that differ from blood interms of fluid density such as hypertonic citrate are investi-gated. Physical phenomena, i.e. sequential dilutions withsaline (injection spillage) and blood (gravitational spillage), aswell as the wash-out effect in catheters with side holes, con-siderably reduce the intra-luminal citrate concentration. ForNewtonian fluids, standard tests may only be valid for the

region between the Luer connector and the highest point ofthe catheter (vertex), where the lock solution is conserved innearly its original concentration. Numerous factors such aslock spillage should be considered to better translate in vitrotests into general practice recommendations.

Recently, Polaschegg [19] published a method for the quan-titative measurement of catheter lock spillage. When citrate46.7% was used as locking anticoagulant, within 20 min acitrate concentration approaching 3% was found in the cathetersection containing a blood/trisodium citrate 46.7% mixture.These results are reasonable due to the high-density differencebetween trisodium citrate 46.7% (1.24 g/cm3) and blood (∼1.05g/cm3) [10, 22]. Even a very low-density difference of 0.0094(0.9%) causes leakage of lock solution in exchange againstblood within 60 min under the influence of gravity (gravita-tional spillage) [13]. Subsequently, we developed a distributedmodel of intra-luminal citrate concentrations in standardhaemodialysis catheters filled with citrate 46.7% (Figure 1).Similarly, in femoral catheters, gravity forces citrate 46.7% toleak out in any patient position, supine or head down tilt, withthe tip of the catheter lowered relative to its insertion point intothe vein. Hence, the same gravitational spillage process, lockdilution with blood, as occurs in jugular or subclavian catheters,can be assumed. The only exception might be patients in per-manent upright position with a femoral catheter.

In previous in vitro studies, pure citrate 30% reduced thenumber of E. coli by 3 log units whereas no antimicrobialeffect was detected in S. aureus time–kill studies [17]. In ourstudy, we obtained comparable results with a 2-log unitreduction of E. coli but nearly unchanged colony counts ofS. aureus after 24 h of incubation in pure citrate 46.7%(Figures 2 and 3). These results are, however, only representa-tive for the connector section. Neither citrate 3%, representingintra-luminal conditions within the core section, nor citrate-free blood, representing the tip section in catheters with side

F IGURE 2 . Time–kill curves of E. coli suspended in citrate 46.7% (solid line with circles), citrate 3% (dashed line with squares) and bloodwithout citrate (dotted line with triangles) after 0, 1, 3 and 24 h of incubation.

ORIG

INALARTIC

LE

L o s s o f a n t i m i c r o b i a l e f f e c t o f c i t r a t e l o c k3



Dow

nloaded from


holes, reduced E. coli or S. aureus colony counts (Figures 2and 3). From a clinical point of view, the antimicrobial effectof trisodium citrate catheter lock solution may therefore beconsidered insufficient. Staphylococcus aureus is one of theleading causes of CRBSIs in haemodialysis patients; however,based on this study, its intra-luminal catheter colonization isaffected neither by citrate 3% (core section) nor by citrate46.7% (connector section). Although citrate 46.7% had someeffect on E. coli in the connector section only, intra-luminalcatheter eradication cannot be expected from our experimentsand previous data [17]. In addition, bacterial load in CRBSIreaches 106 CFU/mL or more and so is, in fact, much higherthan in previous in vitro tests [17, 23]. The catheter lock spil-lage phenomenon resulting in lower intra-luminal citrate con-centrations will abolish any antimicrobial effect of citrate46.7% on non-adherent (planktonic) organisms freely floatingin liquid medium in proximal catheter sections, as shown byour experiments. Conflicting clinical data about the efficacy ofconcentrated citrate in reducing CRBSI in haemodialysispatients have been reported. A randomized, controlled trial byWeijmer et al. [4] comparing citrate 30% with heparin 5%showed a significant reduction of the CRBSI rate attributed tocitrate use. That trial, however, included a heterogeneousgroup of patients with both acute and chronic renal failureusing uncuffed and cuffed CVCs. In an observational retro-spective analysis, there was a decrease in CRBSIs due to S.aureus after the standard locking anticoagulant was changedfrom heparin 5000 IU/mL to citrate 46.7% [5]. However, anantimicrobial effect against planktonic S. aureus is absentthroughout all intra-luminal catheter sections, as shown byour data and previous studies [17]. It is noteworthy thatCRBSI is caused only by spread of planktonic bacteria fromthe biofilms on CVCs into the bloodstream [24]. Power et al.[18] reported a randomized, controlled trial comparing citrate46.7% against heparin 5% in a homogeneous cohort of chronic

haemodialysis patients with a single type of cuffed jugularcatheters. There was no significant difference in rates of exit-site infection or CRBSI between the two groups. The authorsconcluded that the use of citrate 46.7% to prevent CRBSI wasnot justified. Interestingly, a wide range of baseline CRBSIrates from

techniques. In catheters filled with vancomycin, there was adecreasing gradient along the catheter lumens [35]. Sequentialintra-luminal dilution of antibiotic lock solutions with saline(injection spillage) and blood (gravitational spillage) maypromote failure of antibiotic lock therapy and even inducebacterial resistance in prophylactic use. The blood influx intothe catheter reduces the antibacterial effect of certain anti-biotics due to plasma protein binding [36]. Biocompatiblethixotropic gels rather than purely Newtonian fluids can be ex-pected to completely lock the vascular access without spillageand might be a future direction for antimicrobial lock sol-utions but they have not yet been tested.

Our study has some limitations. Firstly, the spillage duringinjection also causes saline dilution within the connectorsection; however, the very small amount of dilution was nottaken into account and can be neglected for in vitro testing.Secondly, to quantify the antimicrobial effect within the coresection, time–kill studies were conducted with a single citrateconcentration. Actually, due to the parabolic flow profileduring lock instillation (instillation spillage) as well as duringblood inflow into the catheter (gravitational spillage), the dilu-tional effects are not completely uniform throughout thewhole catheter length. The citrate concentration is slightlyhigher at the most distal point compared with the most proxi-mal point of the core section. Nevertheless, using citrate 3% isa reasonable compromise. Thirdly, also in terms of furtherphysical influences, in vivo conditions are not completelycomparable with in vitro conditions. Due to catheter com-pression, certain patient positions or opened catheter clampslock spillage might exceed the in vitro quantification andshorten the connector section resulting in further reduction ofantimicrobial effect [10, 19]. Finally, in vitro quantification ofantimicrobial activities inevitably requires some dilution of thestudied lock solution. Thus, in vitro results cannot be comple-tely extrapolated to in vivo settings.

In conclusion, spillage of catheter lock solution leading toreduced intra-luminal citrate concentrations considerablyreduces the antimicrobial effect of citrate 46.7% on E. coli.

As there was no relevant antimicrobial effect on S. aureus,even with citrate 46.7%, its antimicrobial potential to preventinfections in dialysis patients has to be seriously questioned.Furthermore, as inevitable lock spillage alters the antimicro-bial effect of a given lock solution in general, additional invitro tests accounting for dilution effects might better charac-terize the ‘real’ antimicrobial properties of lock solutions.

CONFLICT OF INTEREST STATEMENT

None declared.

REFERENCES

1. Rayner HC, Pisoni RL. The increasing use of hemodialysis catheters: evi-dence from the DOPPS on its significance and ways to reverse it. SeminDial 2010; 23: 6–10

2. Niyyar VD. Catheter dysfunction: the role of lock solutions. Semin Dial2012; 25: 693–699

3. Saxena AK, Panhotra BR. Prevention of catheter-related bloodstream in-fections: an appraisal of developments in designing an infection-resistant‘dream dialysis-catheter’. Nephrology 2005; 10: 240–248

4. Weijmer MC, van den Dorpel MA, Van de Ven PJ et al. Randomized,clinical trial comparison of trisodium citrate 30% and heparin as catheter-locking solution in hemodialysis patients. J Am Soc Nephrol 2005; 16:2769–2777

5. Winnett G, Nolan J, Miller M et al. Trisodium citrate 46.7% selectivelyand safely reduces staphylococcal catheter-related bacteraemia. NephrolDial Transplant 2008; 23: 3592–3598

6. FDA: Food and Drug Administration. (2000) FDA issues warning ontricitrasol dialysis catheter anticoagulant. FDA Talk Paper, 14. T00–T16,Food and Drug Administration.

7. Polaschegg HD, Sodemann K. Risks related to catheter locking solutions con-taining concentrated citrate. Nephrol Dial Transplant 2003; 18: 2688–2690

8. Punt CD, Boer WE. Cardiac arrest following injection of concentrated tri-sodium citrate. Clin Nephrol 2008; 69: 317–318

9. Willicombe MK, Vernon K, Davenport A. Embolic complications fromcentral venous hemodialysis catheters used with hypertonic citrate lockingsolution. Am J Kidney Dis 2010; 55: 348–351

10. Schilcher G, Scharnagl H, Horina JH et al. Trisodium citrate inducedprotein precipitation in haemodialysis catheters might cause pulmonaryembolism. Nephrol Dial Transplant 2012; 27: 2953–2957

11. Davenport A. Why do hypertonic citrate locks lead to dialysis cathetermalfunction; more than a weighty problem? Nephrol Dial Transplant2012; 27: 2621–2624

12. Polaschegg HD. Catheter locking-solution spillage: theory and experimen-tal verification. Blood Purif 2008; 26: 255–260

13. Polaschegg HD. Loss of catheter locking solution caused by fluid density.ASAIO J 2005; 51: 230–235

14. Doorenbos CJ, Van den Elsen-Hutten M, Heuven MJ et al. Estimation oftrisodium citrate (Citra-Lock) remaining in central venous catheters afterthe interdialytic interval. Nephrol Dial Transplant 2006; 21: 543–545

15. Schilcher G, Schlagenhauf A, Schneditz D et al. Ethanol causes proteinprecipitation—new safety issues for catheter locking techniques. PLoSOne 2013; 8: e84869

16. Polaschegg HD. Measuring lock solution spillage in-vivo versus in-vitro.Nephrol Dial Transplant 2009; 24: 2294–2295; author reply 2295

17. Weijmer MC, Debets-Ossenkopp YJ, Van De Vondervoort FJ et al.Superior antimicrobial activity of trisodium citrate over heparin forcatheter locking. Nephrol Dial Transplant 2002; 17: 2189–2195

18. Power A, Duncan N, Singh SK et al. Sodium citrate versus heparin cath-eter locks for cuffed central venous catheters: a single-center randomizedcontrolled trial. Am J Kidney Dis 2009; 53: 1034–1041

19. Polaschegg HD. A novel method for measuring catheter lock spillage: anin vitro study. Artif Organs 2012; 36: 824–829

20. Raad I, Hanna H, Jiang Y et al. Comparative activities of daptomycin, line-zolid, and tigecycline against catheter-related methicillin-resistant Staphy-lococcus bacteremic isolates embedded in biofilm. Antimicrob AgentsChemother 2007; 51: 1656–1660

21. Sherertz RJ, Boger MS, Collins CA et al. Comparative in vitro efficacies ofvarious catheter lock solutions. Antimicrob Agents Chemother 2006; 50:1865–1868

22. Schneditz D, Heimel H, Stabinger H. Sound speed, density and total proteinconcentration of blood. J Clin Chem Clin Biochem 1989; 27: 803–806

23. Brun-Buisson C, Abrouk F, Legrand P et al. Diagnosis of central venouscatheter-related sepsis. Critical level of quantitative tip cultures. ArchIntern Med 1987; 147: 873–877

24. Costerton JW, Stewart PS, Greenberg EP. Bacterial biofilms: a commoncause of persistent infections. Science 1999; 284: 1318–1322

25. Landry DL, Braden GL, Gobeille SL et al. Emergence of gentamicin-resist-ant bacteremia in hemodialysis patients receiving gentamicin lock catheterprophylaxis. Clin J Am Soc Nephrol 2010; 5: 1799–1804

26. Steczko J, Ash SR, Nivens DE et al. Microbial inactivation properties of anew antimicrobial/antithrombotic catheter lock solution (citrate/methyl-ene blue/parabens). Nephrol Dial Transplant 2009; 24: 1937–1945

27. Grudzinski L, Quinan P, Kwok S et al. Sodium citrate 4% locking solutionfor central venous dialysis catheters—an effective, more cost-efficientalternative to heparin. Nephrol Dial Transplant 2007; 22: 471–476

ORIG

INALARTIC

LE

L o s s o f a n t i m i c r o b i a l e f f e c t o f c i t r a t e l o c k5



Dow

nloaded from


28. Silva J, Antunes J, Carvalho T et al. Efficacy of preventing hemodialysiscatheter infections with citrate lock. Hemodial Int 2012; 16: 545–552

29. Patel PR, Yi SH, Booth S et al. Bloodstream infection rates in outpatienthemodialysis facilities participating in a collaborative prevention effort: aquality improvement report. Am J Kidney Dis 2013; 62: 322–330

30. Saxena AK, Panhotra BR. Haemodialysis catheter-related bloodstream in-fections: current treatment options and strategies for prevention. SwissMed Wkly 2005; 135: 127–138

31. Shanks RM, Sargent JL, Martinez RM et al. Catheter lock solutions influ-ence staphylococcal biofilm formation on abiotic surfaces. Nephrol DialTransplant 2006; 21: 2247–2255

32. Raad II, Fang X, Keutgen XM et al. The role of chelators in preventingbiofilm formation and catheter-related bloodstream infections. Curr OpinInfect Dis 2008; 21: 385–392

33. Abraham NM, Lamlertthon S, Fowler VG et al. Chelating agents exert dis-tinct effects on biofilm formation in Staphylococcus aureus depending onstrain background: role for clumping factor B. J Med Microbiol 2012; 61:1062–1070

34. Ramanathan V, Riosa S, Al-Sharif AH et al. Characteristics of biofilm ontunneled cuffed hemodialysis catheters in the presence and absence ofclinical infection. Am J Kidney Dis 2012; 60: 976–982

35. Soriano A, Bregada E, Marques JM et al. Decreasing gradient of antibioticconcentration in the lumen of catheters locked with vancomycin. Eur JClin Microbiol Infect Dis 2007; 26: 659–661

36. Merrikin DJ, Briant J, Rolinson GN. Effect of protein binding on antibioticactivity in vivo. J Antimicrob Chemother 1983; 11: 233–238

Received for publication: 2.9.2013; Accepted in revised form: 21.12.2013

ORIG

INALARTIC

LE

G. Schilcher et al. 6



Dow

nloaded from


CwEnldvis

Kp

J1

1

Sodium Citrate Versus Heparin Catheter Locks for Cuffed CentralVenous Catheters: A Single-Center Randomized Controlled Trial

Albert Power, MBBChir, MRCP, Neill Duncan, MBBS, MRCP, Seema K. Singh, MSc,Wendy Brown, RGN, Elizabeth Dalby, RGN, Claire Edwards, RGN, Kathleen Lynch, RGN,

Virginia Prout, RGN, Tom Cairns, MBBS, BA, Megan Griffith, FRCP, PhD,Adam McLean, FRCP, DPhil, Andrew Palmer, FRCP, and David Taube, FRCP

Background: Sodium citrate has antibacterial and anticoagulant properties that are confined to thecatheter when used as a catheter lock. Studies of its use as a catheter lock have suggested its efficacy inpreventing infection and bleeding complications compared with sodium heparin.

Study Design: Open-label randomized controlled trial of 2 catheter locks to examine the hypothesisthat sodium citrate catheter locks will reduce catheter-related bacteremia and exit-site infection.

Settings & Participants: 232 consenting long-term hemodialysis patients in 4 satellite dialysis unitsto a large dialysis program with protocolized treatment and targets. All patients were using twin-cathetersingle-lumen Tesio-Caths (MedComp, Harleysville, PA).

Intervention: 6 months’ use of 46.7% sodium citrate (citrate) or 5% heparin (heparin) lockedpostdialysis in the dead space of the central venous catheter.

Outcomes & Measurements: Primary end point of catheter-related bacteremia and exit-site infection.Secondary end points of catheter thrombosis defined by the use of urokinase lock and infusion, new catheterinsertion, catheter-related admission, blood transfusions, parenteral iron, and erythropoietin requirements.

Results: Catheter-related bacteremia did not differ in the 2 groups, with an incidence of 0.7events/1,000 catheter-days. There was no significant difference in rates of exit-site infection (0.7 versus0.5 events/1,000 catheter-days; P � 0.5). The secondary end point of catheter thrombosis defined bythe use of a urokinase lock was significantly more common in the citrate group, with an incidence of 8versus 4.3/1,000 catheter-days (P � 0.001). Other secondary end points did not differ. Citrate treatmentwas curtailed compared with heparin because of a greater incidence of adverse events, with a meantreatment duration before withdrawal of 4.8 � 2.0 versus 5.7 � 1.2 months, respectively (P � 0.001).

Limitations: Low baseline catheter-related bacteremia and exit-site infection event rates may haveunderpowered this study. High adverse-event rates may have been related to high-concentration citratethat led to increased overspill and reduction in lock volume. This may also explain the increased rates ofthrombosis in this group.

Conclusion: Widespread and long-term use of 46.7% citrate catheter locks with Tesio-Cath access isnot justified by this study.Am J Kidney Dis 53:1034-1041. © 2009 by the National Kidney Foundation, Inc.

INDEX WORDS: Hemodialysis; heparin; sodium citrate; catheter lock.

suPdw

vtas

entral venous catheters (CVCs) provide vas-cular access for the majority of patients

ho are incident on hemodialysis therapy inurope and the United States and form perma-ent access for a significant proportion of preva-ent patients.1 However, the relative hazard ofeath is 1.5 for CVCs compared with arterio-enous fistulae (AVFs),2 and the relative risk ofnfection was 7.6 for a CVC versus an AVF in 1tudy.3 When an opportunity exists, an AVF

From the Haemodialysis Research Group, Imperial Collegeidney and Transplant Institute, West London Renal and Trans-lant Centre, Hammersmith Hospital, London, UK.

Received August 20, 2008. Accepted in revised formanuary 13, 2009. Originally published online as doi:0.1053/j.ajkd.2009.01.259 on April 27, 2009.

Address correspondence to Albert Power, MBBChir,

American Journal of K034

hould be formed. We have shown that long-termse of a Tesio-Cath (TC; MedComp, Harleysville,A) was a reasonable substitute providing goodialysis adequacy with low complication rateshen an AVF could not be formed.4

It is accepted practice to lock CVCs with aolume of sodium heparin (heparin) restricted tohe lumen of the catheter to prevent thrombosisnd maintain catheter patency between dialysisessions. Sodium citrate locks (citrate) were used

RCP, West London Renal and Transplant Centre, Ham-ersmith Hospital, DuCane Rd, London W12 0HS, Unitedingdom. E-mail: [email protected]© 2009 by the National Kidney Foundation, Inc.0272-6386/09/5306-0017$36.00/0

MmK

doi:10.1053/j.ajkd.2009.01.259

idney Diseases, Vol 53, No 6 (June), 2009: pp 1034-1041

herdeisHighlight

herdeisHighlight

herdeisHighlight

herdeisHighlight

herdeisHighlight

herdeisHighlight

mailto:[email protected]

irdr

atmismbicuaiictp

stgcacoeu

thwclfpad

vapditha

awt7wir

A

ate

daipTnaaedaprmpir

T

tisbtivtocs

rwtardoevwsw

Citrate vs. Heparin Locks for Cuffed Central Venous Catheters in Hemodialysis 1035

nitially as an alternative for patients with hepa-in sensitivity.5 However, in vitro, citrate has aemonstrated antibacterial effect, whereas hepa-in has none.6

Overspill of 15% to 30% of the catheter lockfter injection of a volume of solution equal tohe dead space of the CVC specified by theanufacturer is recognized from in vitro stud-

es.7,8 Inadvertent spillage of catheter lock andystemic anticoagulation is not desirable for he-odialysis patients who are at heightened risk of

leeding from medical interventions and gastro-ntestinal causes. This bleeding tendency may beompounded by uremic platelet dysfunction andse of antiplatelet agents. Citrate has an antico-gulant action confined to the catheter becausets action is overwhelmed by the concentration ofonized calcium in the systemic circulation.9 Inontrast, small amounts of heparin have a sys-emic anticoagulant effect because of its greaterotency.10

To date, there have been mainly small andhort-term studies suggesting the benefits of ci-rate over heparin catheter locks on theserounds.11-14 One randomized controlled trialomparing 30% citrate with 5% heparin showedsignificant decrease in infective and bleeding

omplications, but the trial was in a mixed groupf patients with both acute renal failure andnd-stage renal failure (ESRF) with use of bothncuffed and cuffed CVCs.15

Our study is the first randomized controlledrial to compare the use of 46.7% citrate and 5%eparin catheter locks exclusively in patientsith ESRF and a single type of cuffed twin-

atheter single-lumen CVC, the TC, intended asong-term vascular access. We compare the ef-ects of these catheter locks on the primary endoint of catheter-related infection and the second-ry end points of measures of catheter patency,ialysis efficacy, and bleeding.

METHODSA total of 232 hemodialysis patients with internal jugular

enous TCs (Bio-Flex TC) gave consent and were randomlyssigned to open-label treatment for 6 months (Fig 1). Allatients who had been on dialysis therapy for longer than 90ays were eligible. Patients with a bleeding diathesis, anntervention, or pathological state within 3 months of entryhat would heighten the risk of bleeding and those withypocalcemia were excluded from the study. Patients were

ssigned to treatment arms by using single random-number r

llocation. However, randomization was unintentionallyeighted by a systematic error: odd random numbers from 0

o 9, and including 0, were allocated to citrate (ie, 0, 1, 3, 5,, and 9), and even random numbers from 0 to 9 excluding 0ere allocated to heparin (ie, 2, 4, 6, and 8). This resulted in

ncreased treatment allocation to citrate: 132 patients wereandomly assigned to citrate, and 100, to heparin.

dministrationof Catheter Locks andTCCare

TCs were manipulated by using strict aseptic technique,nd catheter locks were instilled slowly in a volume equal tohe dead space of the TC to minimize overspill at the end ofach dialysis session. Each administration was registered.

Catheter locks were evacuated at the beginning of the nextialysis session. An additional 20 mL of blood was evacu-ted before venesection for blood tests to avoid contaminat-ng test samples. This blood was then instilled back to theatient after tests had been performed to prevent wastage.he exit site was cleaned at each dialysis session with sterileormal saline followed by chlorhexidine solution, 4%, andir dried before application of a bio-occlusive dressing. Nontimicrobial ointment was applied to the exit site. Adverseffects were systematically assessed and recorded at eachialysis session. Patients were asked to report any symptomnd were not led by questioning to report any symptom inarticular during the course of the study. However, as aesult of the consent process, patients were aware that citrateight induce digital and facial paresthesia. Digital and facial

aresthesias with the use of citrate were considered anndication of overspill, and the catheter lock volume waseduced by 0.1 mL at the next dialysis treatment.

C Infection

Dialysis patients with pyrexia, defined as a tympanicemperature of 38°C or greater with or without a systemicnflammatory response were investigated for a TC-relatedource of infection by means of exit-site swabs and multiplelood cultures before starting antibiotic therapy. Antibioticherapy starts preempted microbiological confirmation ofnfection and followed an initial protocol of intravenousancomycin, 500 mg, after dialysis as Gram-positive coverailored to maintain trough levels greater than 10 mg/L andral ciprofloxacin, 250 mg, twice daily as gram-negativeover. Antibiotics were continued according to culture re-ults and antibiotic sensitivity for a minimum of 2 weeks.

Exit-site swabs were used if there were exudates or crust,edness, or induration at the exit site. Exit-site infectionsere treated initially with oral clarithromycin, 250 mg,

wice daily and oral ciprofloxacin, 250 mg, twice daily, anddjustment was made according to response and cultureesults for a minimum of 2 weeks. Tunnel infections wereefined as pain, redness, or induration along the subcutane-us course of the catheter with or without exudates at thexit site. Tunnel infections were treated from the outset withancomycin and ciprofloxacin, and these were adjustedhen culture results were available with the addition of a

econd appropriate oral antibiotic for a minimum of 4eeks.Patients with pyrexia showing a systemic inflammatory

esponse, relative hypotension determined from the indi-

vtaiptc

T

Sit3mdxnds1il

wnu

DV

sMebsMatb

P

0a

Power et al1036

idual patient’s usual range for blood pressure, or persistentunnel infection were admitted to the hospital. Bacteremialone did not qualify the patient for admission. If TC-relatednfection resulted in hypotension requiring inotrope support,ersistent bacteremia despite antibiotics, or a tunnel infec-ion for longer than 3 days, the TC was removed and a newatheter was sited.

CDysfunction

Target blood flow for the TC was 350 mL/min or greater.uboptimal blood flow less than 250 mL/min and/or decreas-

ng dialysis adequacy was used as a marker for TC dysfunc-ion. Single-pool Kt/V (spKt/V) was measured monthly, and

sequential monthly decreases in spKt/V, irrespective ofagnitude, defined decreasing dialysis adequacy. Catheter

isplacement or kinking was excluded by means of a chest-ray. Each affected patient then had 5,000 units of uroki-ase locked into each catheter of the TC for 2 hours, andialysis was reattempted on an outpatient basis. If thistrategy failed, patients were admitted to the ward for a2-hour intraluminal infusion of 12,500 units of urokinasento each catheter of the TC, as previously described in the

iterature.16 If this final strategy was unsuccessful, the TC r

as removed and replaced using the over-the-wire tech-ique. Oral antiplatelet or anticoagulant agents were notsed with the intent to improve blood flow rate.

ialysis Adequacy andHematologic andBiochemicalariables

All patients had thrice-weekly dialysis using low-fluxynthetic AM-BIO-1000Wet hemodialyzers (Asahi Kaseiedical Europe GmbH, Frankfurt, Germany). Dialysis ad-

quacy was measured by using spKt/V on a monthly basisy means of the Daugirdas method.17 Postdialytic urea wasampled after 1 minute by using the slow-flow method.easurements of spKt/V were subject to monthly consultant

udit, and the dialysis prescription was adjusted to achieve aarget spKt/V of 1.6 or greater. Routine hematology andiochemistry tests were performed monthly.

ower, Statistics, andEthical Permission

The bacteremia rate at our center was established to be.68 events/1,000 catheter-days through our participation inCenters for Disease Control and Prevention study.18 With

Figure 1. Trial flow dia-gram.

ecruitment of 100 patients to each arm for this 6-month

tc0rtwE0o

tmiafi

tM5hWer

9csTbacsc

pel6iF

rrRwa

NMAE

BDDTMBV

c

DMCNCEU1BMMBM

g


rial, it was 97% powered to show a 75% decrease inatheter-related bacteremia, the primary end point (� �.05). This was the magnitude of decrease in bacteremiaates seen in the previous randomized controlled trial.15 Therial adhered to the Declaration of Helsinki principles andas approved by the St Mary’s NHS Trust Local Researchthics Committee, trial number EC03.106R&D 03/FA/05E. An abstract of the trial has been published previ-usly.19

Parametric data were analyzed by using Student t test, andimeline incidence data were analyzed by using a Poissonodel. Kaplan-Meier survival analysis was performed on an

ntention-to-treat basis for patient survival, TC survival, anddverse-event–free survival. Statistical significance was de-ned as P � 0.05.

Table 1. Baseline Characteristics ofStudy Participants

Citrate Heparin P

o. of patients 132 100en/women 73:59 59:41 0.6ge (y) 63 � 14 62 � 13 0.5thnicityWhite 62 (47) 44 (44) 0.7South Asian 38 (29) 32 (32) 0.7Afro-Caribbean 29 (22) 20 (20) 0.7Asian 3 (2) 4 (4) 0.5

ody mass index (kg/m2) 25 � 6 26 � 6 0.3iabetes 56 (42) 44 (44) 0.8ialysis vintage (mo) 37 � 40 34 � 28 0.5esio-Cath vintage (mo) 19 � 14 20 � 17 0.4ean single-pool Kt/V 1.7 � 0.3 1.7 � 0.3 0.6lood flow rate (mL/min) 350 � 40 353 � 33 0.6enous pressure (mm Hg) 247 � 45 243 � 43 0.5

Note: Values expressed as mean � SD or number (per-ent) unless noted otherwise.

Table 2. S

eaths (/1,000 patient-y)ean single-pool Kt/V over 6 moatheter-related admissions (/1,000 catheter-d)ew catheters (/1,000 catheter-d)atheter-related bacteremia (/1,000 catheter-d)xit-site infections (/1,000 catheter-d)rokinase locks (/1,000 catheter-d)2-h urokinase infusions (/1,000 catheter-d)leeding complicationsean hemoglobin over 6 mo (g/dL)ean ferritin over 6 mo (ng/mL)lood transfusions (/1,000 catheter-d)ean intravenous erythropoietin dosage (U/wk)

Note: Values expressed as mean � SD and event rate (9

/L, �10. No conversion is necessary for ferritin in ng/mL and �g/L

RESULTS

One hundred thirty-two patients were adminis-ered 46.7% citrate catheter locks (DuraLock C;

edComp) and 100 patients were administered% heparin catheter locks (Monoparin sodiumeparin, 5,000 IU/mL; CP Pharmaceuticals,rexham, UK). There were no significant differ-

nces between treatment groups at the time ofandomization (Table 1).

Cumulative patient survival at 6 months was5% for both groups (log-rank test, P � 0.9)ensored for drop-out defined by change in dialy-is modality, transplantation, or use of an AVF.here were 5 deaths in the citrate group (2 fromronchopneumonia, 2 from sudden cardiac death,nd 1 from nonhemorrhagic cerebrovascular ac-ident) and 5 deaths in the heparin group (4 fromepsis unrelated to the CVC, and 1 from suddenardiac death; Table 2).

There were no significant differences in therimary end point of CVC-related bacteremia orxit-site infection rates (Table 2). Rates of cumu-ative survival free of CVC-related bacteremia atmonths were 91% in the citrate group and 89%

n the heparin group (log-rank test, P � 0.7;ig 2).In regard to secondary end points, CVC-

elated admission rates and new CVC insertionates were not significantly different (Table 2).ates of cumulative catheter survival at 6 monthsith optimal flow were 95% in the citrate group

nd 100% in the heparin group (log-rank test,

utcomes

itrate Heparin P

1-222) 102 (34-246) 0.97 � 0.2 1.7 � 0.2 0.9.56-1.5) 0.7 (0.36-1.2) 0.4.22-0.89) 0.7 (0.36-1.2) 0.4.40-1.2) 0.7 (0.36-1.2) 0.9.36-1.2) 0.5 (0.24-1.0) 0.5.6-9.2) 4.3 (3.4-5.4) �0.001.03-0.46) 0 0.10 0

0 � 0.3 12.1 � 0.3 0.40 � 9 543 � 9 0.2.44-1.3) 0.52 (0.2-1.0) 0.34 � 55 8,070 � 57 0.5

fidence interval). Conversion factor: hemoglobin in g/dL to

tudy O

C

94 (31.

0.9 (00.5 (00.7 (00.7 (0

8 (60.2 (0

12.51

0.8 (08,53

5% con

.

Puns(slfe

A

mae

(fp1eosc20wc(

ilcdqdhch

lso

o

(

Power et al1038

� 0.01; Fig 3). There was significantly greaterse of urokinase locks in the citrate group, butot 12-hour urokinase infusions (Table 2). Dialy-is adequacy was equivalent for both groupsTable 2). No significant clinical bleeding epi-odes were seen. Mean hemoglobin and ferritinevels were not statistically different. Blood trans-usion and erythropoietin requirements werequivalent for the 2 groups (Table 2).

dverse Effects

There were no withdrawals from heparin treat-ent because of adverse effects. Citrate was

ssociated with more reported side effects. Sev-nty-one of 132 patients who received citrate

0

.2

.4

.6

.8

1

Cum

. Sur

viva

l

0 1 2 3 4 5 6 7Time [months]

Cum

ulat

ive

stud

y su

rviv

al w

ith o

ptim

al c

athe

ter f

low

0

.2

.4

.6

.8

1

Cum

ulat

ive

CR

B-fre

e su

rviv

al

0 1 2 3 4 5 6 7Time [months]

Figure 2. Cumulative catheter-related bacteremiaCRB) survival. Sodium citrate, blue line; heparin, red line.

Figure 3. Cumulative study survival censored for sub-ptimal flow. Sodium citrate, blue line; heparin, red line.

dh

3.72 events/1,000 catheter-days) had side ef-ects (P � 0.001). These resolved in 24 of 132atients (34%) with dose reduction. Twenty of32 patients (15%) receiving citrate withdrewarly from the study because of adverse effectsf metallic taste and facial and/or digital paresthe-ia. As a result, citrate treatment was curtailedompared with heparin (mean follow-up, 4.8 �.0 versus 5.7 � 1.2 months, respectively; P �.001). Cumulative study survival rates afterithdrawal for adverse effects were 85% for

itrate versus 100% for heparin after 6 monthsP � 0.02; Fig 4).

DISCUSSION

Citrate was not associated with a reduction innfection. There was increased use of thrombo-ytic therapy to restore functional patency in theitrate arm, but no differences in CVC survival,ialysis adequacy, hospital admissions, or conse-uences of bleeding. This study is the first ran-omized controlled trial to compare citrate witheparin as catheter locks for a single twin-atheter system in a large number of prevalentemodialysis patients.The only randomized controlled trial pub-

ished to date, by Weijmer et al,15 analyzed aimilar number of patients (n � 291), althoughnly 98 had tunneled cuffed catheters (5 were

0

.2

.4

.6

.8

1

Cum

. Sur

viva

l

0 1 2 3 4 5 6 7Time [months]

Cum

ulat

ive

stud

y su

rviv

al w

ithou

t adv

erse

eve

nts

Figure 4. Cumulative study survival censored for with-

rawal through adverse effects. Sodium citrate, blue line;eparin, red line.

Tedwml3meviwd(wOos

dpiecttpbtavtdtiwaqjh

C4ycrcrareTcrpcreNe

lhwteoc1ssttdvcctbflW

BSHW

GL


Cs and the remainder were dual-lumen cath-ters). Patient demographics was considerablyifferent in their study: 30% were on twice-eekly hemodialysis therapy, 25% were on treat-ent for acute renal failure rather than estab-

ished on long-term hemodialysis therapy, and0% were on oral anticoagulation therapy (cou-arins). In the study by Weijmer et al,15 all

nrolled patients had a newly inserted catheterersus 2 patients in our study with a newlynserted TC. Mean catheter vintage in our studyas 24 months. Patients were of a youngerialysis vintage in the study by Weijmer et al15

mean, 1.2 years on dialysis therapy) comparedith our study (mean, 3 years on dialysis therapy).ther studies to date are heterogeneous in termsf design: type of study, concentration of lockingolutions, and type of catheter studied (Table 3).

In vitro studies had shown that citrate has aose-dependent antibacterial effect that sur-asses heparin related to the chelation of calciumons and not the osmolality of the solution.6 Asht al11 showed that the bactericidal effect ofitrate in vitro was manifest only with concentra-ions greater than 23%. Shanks et al20 showedhat citrate at a concentration of 2% or greaterrevented the in vitro formation of bacterialiofilm by staphylococci on a number of syn-hetic surfaces. This may help explain in part thentimicrobial effect of citrate catheter locks. Inivo, Ash et al11 noted a significant decrease inhe incidence of catheter-related bacteremia withiffering concentrations of citrate in a prospec-ive cohort study of patients with ESRF: thencidence of symptomatic bacteremia was 4.13%ith heparin, 1.79% with 23% citrate (P � 0.05),

nd 0% with 47% citrate (P � 0.05). Subse-uently, the randomized controlled trial by Wei-mer et al15 compared 30% citrate versus 5%eparin and showed a significant decrease in

Table 3. Summary of Studies Comparin

Reference No. of Patients Citrate (%) H

uturovic et al,12 1998 30 4tas et al,13 2001 11 30endrickx et al,14 2001 19 5eijmer et al,15 2005 291 30

rudzinski et al,23 2007 189 4

ok et al,21 2007 129 4 5

VC-related bacteremia with citrate (1.1 versus.1 events/1,000 catheter-days; P � 0.001). Anal-sis of the 98 patients with tunneled cuffedatheters in this study shows a decrease in CVC-elated bacteremia from 4.2 to 0.8 events/1,000atheter-days with the use of 30% citrate. Theseesults were used to power our study. However,t our center, we had a much lower bacteremiaate in the control group (0.7 events/1,000 cath-ter-days) than in the study by Weijmer et al.15

his could be the result of catheter type and ouratheter care protocols and is likely to haveeduced the impact of citrate in our study com-ared with others. A recent Canadian prospectiveohort study reported less catheter-related bacte-emia in the 4% citrate group, but a change inxit-site care may have confounded the data.21

o beneficial effect of citrate on the incidence ofxit-site infections has been shown.14,22

Small controlled studies have reported equiva-ent short-term patency rates with citrate versuseparin. Buturovic et al12 showed in 30 patientsith ESRF that 4% citrate catheter locks in

emporary single-lumen CVCs maintained anquivalent duration of access patency as heparinr polygeline. Stas et al13 compared 30% citrateatheter locks with 5% heparin catheter locks in1 patients with ESRF and cuffed double- andingle-lumen CVCs using a crossover study de-ign. There were no CVC occlusions, no need forhrombolytics, and no significant differences inhrombus formation for either type of lock. Hen-rickx et al14 prospectively evaluated 5% citrateersus 5% heparin in 19 patients with ESRF withuffed single-lumen CVCs and found a signifi-antly larger number of dialysis sessions withhrombus formation in those treated with citrate,ut no differences in dialysis adequacy, bloodow, number of occlusions, or urokinase use.eijmer et al15 found no difference in the num-

te and Heparin Catheter Locks to Date

) Outcome

No differenceNo differenceNo difference in use of thrombolytic or infectionNo difference in use of thrombolytic, reduced central

venous catheter–associated bacteremia withcitrate

No difference in use of thrombolytic

g Citra

eparin (%

5555

10

Reduced use of thrombolytic with citrate

bacseTctevcMeutsaIt5Clttaata5wowp

ilrwttoohsamcawt

bvpsuuCoStifosfsoapbcrcpcarphpcecsFe

£hScs

cai

Power et al1040

er of catheters removed because of poor flow ormount of urokinase used in their trial of 30%itrate versus 5% heparin catheter locks. Wehow significantly greater use of urokinase cath-ter locks in patients treated with 46.7% citrate.hree of 132 patients in the citrate group ac-ounted for 32% of the urokinase locks used, andhis conferred statistical significance to the differ-nce between groups. The reduction in lockolume because of side effects may have in-reased the rate of thrombosis in the citrate arm.ore recently, Grudzinski et al23 found no differ-

nce in the rate of flow-related catheter change,se of thrombolytic therapy (alteplase), and bac-eremias between 4% citrate catheter locks ver-us 10% heparin catheter locks in a retrospectivenalysis of their Canadian hemodialysis cohort.n an accompanying Canadian prospective longi-udinal cohort study comparing 4% citrate versus% heparin catheter locks, there were fewerVC exchanges and a longer time to thrombo-

ytic use and catheter change in the citrate-reated group.21 It is worth noting that althoughhese 2 studies looked at patients on the hemodi-lysis program, ie, not treated for reversiblecute renal failure and with tunneled CVCs,here was heterogeneity in the type of CVC used,nd this may have influenced results. More than0% of patients in the study by Grudzinski et al23

ere on warfarin therapy, with greater than 80%f these to maintain access patency. It is not clearhether patients were on anticoagulant or anti-latelet treatment in the study by Lok et al.21

There were no changes in markers of bleedingn our study; patient hemoglobin levels, ferritinevels, and blood transfusion and erythropoietinequirements were similar. No bleeding eventsere recorded in either group. This is in contrast

o Weijmer et al,15 who found less bleeding inhe 30% citrate group. Despite evidence of moreverspill in our study with greater concentrationsf citrate, the anticoagulant effect of citrate wouldave been overwhelmed in the circulation. Thetudy by Weijmer et15 included patients withcute renal failure and uncuffed catheters whoight have been at greater risk of bleeding,

ompounded by the combination of oral antico-gulation in 30%. Only 1 patient in our studyas concurrently on warfarin therapy (heparin-

reated group).

Citrate was associated with a significant num-er of early withdrawals (15%) because of ad-erse effects, greater than the 10% quoted in arior study using 47% citrate.11 Patients in thattudy received 47% citrate for only 3 months in anit with a 10% prevalence of TCs. In a studysing a dual-lumen cuffed catheter (AshSplitath; MedComp, Harleysville, PA), remarkably,nly 1 of 207 patients reported side-effects.24

tudies show systemic leakage of 15% to 30% ofhe lock solution in a variety of catheters.7,11 Ann vitro study showed that the majority leakedrom side holes on the catheter tip.8 The numberf symptoms associated with citrate in our studyuggests a significant degree of systemic leakagerom venous catheters. This may relate to theide holes in TCs. Also, this may relate to our usef the highest concentration of citrate clinicallyvailable, which is more dense and thereby morerone to overspill. Overspill also may be causedy variable intraluminal volume because TCsan be cut to a desired length after insertion andepair. Fill volumes are clearly marked on theatheter. As in the study by Weijmer et al,15

atients were asked to report symptoms afteratheter locking in a systematic way in our study,lthough reporting bias may have influencedesults as a consequence of the informed consentrocess for the trial. Concern about 47% citratead been raised after a report of a hemodialysisatient who received a total of 10 mL of 47%itrate through a CVC and subsequently experi-nced cardiac arrest.25 There were no adverseardiovascular events in the citrate group in ourtudy. High-dose citrate is not approved by theood and Drug Administration for use as a cath-ter-locking solution in the United States.

The financial cost of citrate catheter locks at8.34/lock was 16 times greater than that foreparin, at £0.5/lock, in the course of this trial.ince the conclusion of the trial, the cost of theitrate lock has decreased to £2.65/lock, with noignificant change in the cost of heparin.

Based on the lack of a significant effect onatheter-related infection and the greater rate ofdverse effects, widespread use of 46.7% citrates not justified by this study.

ACKNOWLEDGEMENTSSupport: None.

Financial Disclosure: None.

aD

vpE2

Am9

pr

ks

vcp

s2

le3

2ss2

fi1

s3

hvs

di

e1

cce

R3s

ntp

me

hwr

rlN

Gf2

Th4

tS2

dch

ccT

iP


REFERENCES1. Pisoni RL, Young EW, Dykstra DM, et al: Vascular

ccess use in Europe and the United States: Results from theOPPS. Kidney Int 61:305-316, 20022. Astor BC, Eustace JA, Powe NR, et al: Type of

ascular access and survival among incident hemodialysisatients: The Choices for Healthy Outcomes in Caring forSRD (CHOICE) Study. J Am Soc Nephrol 16:1449-1455,0053. Hoen B, Paul-DauphinA, Hestin D, et al: EPIBACDIAL:multicenter prospective study of risk factors for bacterae-ia in chronic hemodialysis patients. J Am Soc Nephrol

:869-876, 19984. Duncan NDC, Singh S, Cairns TDH, et al: Tesio-Caths

rovide effective and safe long-term vascular access. Neph-ol Dial Transplant 19:2816-2822, 2004

5. Purchase L, Gault MH: Hemodialysis with a Permcathept open with streptokinase and later citrate in a heparin-ensitive patient. Nephron 58:119-120, 1991

6. Weijmer MC, Debets-Ossenkopp YJ, van de Vonder-oort FJ, et al: Superior antimicrobial activity of trisodiumitrate over heparin for catheter locking. Nephrol Dial Trans-lant 17:2189-2195, 20027. Polaschegg HD, Shah C: Overspill of catheter locking

olution: Safety and efficacy aspects. ASAIO J 49:713-715,0038. Sungur M, Eryuksel E, Yavas S, et al: Exit of catheter

ock solutions from double lumen acute haemodialysis cath-ters—an in vitro study. Nephrol Dial Transplant 22:3533-537, 20079. McGill RL, Smith BC, Tucker Freedy H, et al: Use of

3.35% tri-sodium citrate as a dialysis catheter lockingolution after “heparin-free” hemodialysis. Abstract pre-ented at: 33rd Annual Meeting of the ASN. October 13-16,000, Toronto, Canada10. Bayes B, Bonal J, Romero R: Sodium citrate for

lling haemodialysis catheters. Nephrol Dial Transplant4:2532-2533, 199911. Ash SR, Mankus RA, Sutton JM, et al: Concentrated

odium citrate (23%) for catheter lock. Hemodial Int 4:22-1, 200012. Buturovic J, Ponikvar R, Kandus A, et al: Filling

emodialysis catheters in the interdialytic period: Heparinersus citrate versus polygeline: A prospective randomisedtudy. Artif Organs 22:945-947, 1998

13. Stas KJF, Vanwalleghem J, De Moor B, et al: Triso-ium citrate 30% vs heparin 5% as catheter lock in the

nterdialytic period in twin- or double-lumen dialysis cath- H

ters for intermittent haemodialysis. Nephrol Dial Transplant6:1521-1522, 200114. Hendrickx L, Kuypers D, Evenepoel P, et al: A

omparative prospective study on the use of low concentrateitrate lock versus heparin lock in permanent dialysis cath-ters. Int J Artif Organs 24:208-211, 2001

15. Weijmer MC, van den Dorpel MA, van de Ven, et al:andomized, clinical trial comparison of trisodium citrate0% and heparin as catheter-locking solution in hemodialy-is patients. J Am Soc Nephrol 16:2769-2777, 2005

16. Webb A, Abdalla M, Russell D: A protocol of uroki-ase infusion and warfarin for the management of thehrombosed haemodialysis catheter. Nephrol Dial Trans-lant 16:2075-2078, 200117. Daugirdas JT: Second generation logarithmic esti-ates of single-pool variable volume Kt/V: An analysis of

rror. J Am Soc Nephrol 4:1205-1213, 199318. Duncan N, McLean AG, Cairns T, et al: Low rates of

ospital admission and access-related infection associatedith Tesio-Caths in a prospective analysis. J Am Soc Neph-

ol 14:241A, 2003 (abstr)19. Duncan N, Singh S, Amao M, et al: A single centre

andomised control trial of sodium citrate versus heparinine locks for cuffed central venous catheters. J Am Socephrol 16:451A, 2005 (abstr)20. Shanks R, Sargent J, Martinez R, Graber M, O’Toole

: Catheter lock solutions influence staphylococcal biofilmormation on abiotic surfaces. Nephrol Dial Transplant 21:247-2255, 200621. Lok C, Appleton D, Bhola C, Khoo B, Richardson R:

risodium citrate 4%—An alternative to heparin capping ofaemodialysis catheters. Nephrol Dial Transplant 22:477-83, 200722. MacRae JM, Dojcinovic I, Djurdjev O, et al: Ci-

rate 4% versus Heparin and the Reduction of Thrombosistudy (CHARTS). Clin J Am Soc Nephrol 3:369-374,00823. Grudzinski L, Quinan P, Kwok S, Pierratos A: So-

ium citrate 4% locking solution for central venous dialysisatheters—An effective, more cost-effective alternative toeparin. Nephrol Dial Transplant 22:471-476, 200724. Winnett G, Nolan J, Miller M, Ashman N: Trisodium

itrate (TSC) 46.7% selectively and safely reduces staphylo-occal catheter-related bacteraemia (CRB). Nephrol Dialransplant 23:3592-3598, 200825. US Food and Drug Administration: FDA Issues Warn-

ng on triCitrasol Dialysis Catheter Anticoagulant. FDA Talkaper T00-16. Rockville, MD, US Department of Health and

uman Services, April 14, 2000

Nephrol Dial Transplant (2012) 0: 1–5

doi: 10.1093/ndt/gfs048

Original Article

Trisodium citrate induced protein precipitation in haemodialysiscatheters might cause pulmonary embolism

Gernot Schilcher1, Hubert Scharnagl2, Joerg H. Horina1, Werner Ribitsch1, Alexander R. Rosenkranz1,Tatjana Stojakovic2 and Hans-Dietrich Polaschegg1,3

1Clinical Division of Nephrology, Department of Internal Medicine, Medical University of Graz, Graz, Austria, 2Clinical Institute ofMedical and Chemical Laboratory Diagnostics, Medical University of Graz, Graz, Austria and 3Medical Devices Consultant,Köstenberg, Austria

Correspondence and offprint requests to: Gernot Schilcher; E-mail: [email protected]

AbstractBackground. The locking anticoagulant plays a decisiverole in the patency of central venous catheters (CVCs) usedfor haemodialysis. During injection, the hydraulic effectsinevitably cause lock solution to spill into the systemic cir-culation. Density differences between whole blood (WB)and the lock solution cause further gravity-induced seepageof lock solution. This is followed by an influx of WB into thecatheter, also described for trisodium citrate, which is acommon agent for serum protein precipitation. Emboliccomplications from haemodialysis catheters locked with hy-pertonic trisodium citrate have been reported. We aimed toinvestigate protein precipitation in trisodium citrate lockedcatheters as a possible cause of pulmonary embolisms.Methods. In vitro, WB and trisodium citrate (concentra-tions ranging from 4.7 to 46.7%) mixtures in a ratio of 1:4were used to assess protein precipitation. Additionally,WB/trisodium citrate mixture was pumped through a 20-lm mesh filter, simulating pulmonary vessels, and filtratepressure was measured. In vivo, listed filling volumes ofhaemodialysis catheters locked with trisodium citrate 4%(n ¼ 10), 10% (n ¼ 10), 20% (n ¼ 10) or 46.7% (n ¼ 10)were aspirated and then analysed for protein precipitation.Results. In vitro, protein precipitation capable of causingfilter occlusion was observed in test solutions containingtrisodium citrate above a concentration of 12%. In vivo,protein precipitation was detected in all samples from theCVCs filled with trisodium citrate 46.7% (n¼ 10) and 20%(n¼ 10). In contrast, there were no signs of precipitation insamples from the catheters filled with trisodium citrate 4%(n ¼ 10) or 10% (n ¼ 10).Conclusions. Our in vitro results demonstrate that proteinprecipitates inside haemodialysis catheters when trisodiumcitrate is used above the concentrations of 12%. Precipi-tated protein may have contributed to the pathophysiologyof reported embolisms from haemodialysis catheters filledwith hypertonic trisodium citrate. Based on our findings,we suggest that trisodium citrate lock solution up to theconcentration of 10% can be used safely.

Keywords: central venous catheter; haemodialysis access; lock spillage;protein precipitation; trisodium citrate lock solution

Introduction

Central venous catheter (CVC) use, representing a majorvascular access modality for haemodialysis, is steadily in-creasing despite recommendations favouring arteriovenousfistulae [1, 2]. Maintaining the intraluminal patency of aCVC requires instillation of a prophylactic locking antico-agulant, such as heparin [3]. Trisodium citrate (citrate) iscurrently used worldwide and considered a safe alternativeto heparin [4, 5]. However, the use of hypertonic citrateremains controversial [6–8].

Although leakage of catheter lock solutions into the sys-temic circulation of ~20–25% has been repeatedly demon-strated [9, 10], in the European Union lock solutions areregarded as a ‘medical device’ rather than a systemic drug[11]. The spillage during instillation of the listed fillingvolume into the CVC is a consequence of laminar flowdistribution within the catheter and thus cannot be avoided.Further loss of lock solution due to gravity should be con-sidered if lock solutions (e.g. citrate 30 or 46.7%) with adensity higher than blood are used [12–14]. In this case, thementioned seepage is followed by a reverse whole blood(WB) flow into the CVC and has been observed in vivo[13].

Hence, this WB remains inside the CVC while the cath-eter is not in use. This is of utmost importance if the in-stilled lock solution has potential protein precipitatingeffects. Since the 19th century, ‘salting out’ of plasma pro-teins using high concentrations of salts (e.g. citrate) hasbeen a common technique to precipitate a target protein[15, 16]. Reported embolisms from CVCs used with hyper-tonic citrate locking solution might be due to serum proteinprecipitation in the CVC [17].

The objective of this study was therefore to investigatewhether there might be protein precipitation in citrate-locked CVCs. In vitro test solutions consisting of WB and

� The Author 2012. Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved.For Permissions, please e-mail: [email protected]

NDT Advance Access published March 30, 2012 at U

niversitatsbibliothek Wuerzburg on A

pril 18, 2012http://ndt.oxfordjournals.org/

Dow

nloaded from

herdeisHighlight

herdeisHighlight

herdeisHighlight

herdeisHighlight

herdeisHighlight

herdeisHighlight

herdeisHighlight

herdeisHighlight

herdeisHighlight

herdeisHighlight

herdeisHighlight

herdeisHighlight

herdeisHighlight

herdeisHighlight

performance properties of modern locksolutions: … · taurolock™, killing curves, and schilcher...

Documents