Vitamin E–Coated and Heparin-Coated Dialyzer Membranes for Heparin-Free Hemodialysis: A Multicenter, Randomized, Crossover Trial

Original Investigation Vitamin E–Coated and Heparin-Coated Dialyzer Membranes for Heparin-Free Hemodialysis: A Multicenter, Randomized, Crossover Trial Mohamed Shariful Islam, MBBS,1 Zarih Alcheikh Hassan, MD,2 Florence Chalmin, MD,1 Sandor Vido, MD,1 Mohamed Berrada, MD,1 David Verhelst, MD,2 Patrick Donnadieu, MD,2 Olivier Moranne, MD, PhD,1 and Vincent L.M. Esnault, MD, PhD 1,3 Background: Hemodialysis requires effective anticoagulation to avoid blood circuit clotting. In patients at high risk for bleeding, several alternative methods have been developed. Study Design: Multicenter, prospective, randomized, crossover study evaluating the noninferiority of vitamin E–coated compared with heparin-coated dialyzers in a 4-hour heparin-free hemodialysis strategy. Settings & Participants: 32 adult long-term hemodialysis patients from 2 French hemodialysis units with well-functioning fistulas or double-lumen catheters. Intervention: Patients were randomly allocated to a first period using either vitamin E– or heparin-coated dialyzers. After a washout period of 2 hemodialysis sessions, each patient was switched to the alternative dialyzer for a second period. Each study period started with 2 hemodialysis sessions with reduced heparin dose (50% and 25% of usual heparin dose, respectively, for sessions 1 and 2) followed by 2 heparin-free sessions. Outcomes: The primary end point was the percentage of successful study periods, defined as no circuit-clotting event leading to premature interruption of any of the 4 dialysis sessions. Secondary end points included total number and cumulative duration of hemodialysis sessions without clotting, number of saline solution flushes, dialysis circuit bubble trap status and dialyzer membrane status by visual inspection, and dialysis adequacy. Results: The percentage of success with vitamin E–coated dialyzers (25/32 study periods [78%]) was not inferior to that with heparin-coated dialyzers (26/32 study periods [81%]). Visual inspection showed equal numbers of clean dialysis circuit bubble traps (vitamin E–coated, 34/121; heparin-coated, 32/120), whereas clean fiber bundles were more frequently noted with the vitamin E–coated compared with heparin-coated dialyzers (25/121 vs 2/120; P 5 0.002). Limitations: Results may not extrapolate to critically ill patients. Differences in dialyzer transparency may account for visual inspection scores. Conclusions: The success rate of 4-hour heparin-free hemodialysis sessions is lower than that previously claimed in uncontrolled studies. Vitamin E–coated and heparin-coated dialyzers exposed patients to similar and unacceptable high failure rates. Further studies are required to improve heparin-free hemodialysis. Am J Kidney Dis. 68(5):752-762. ª 2016 by the National Kidney Foundation, Inc. INDEX WORDS: Heparin-free hemodialysis; vitamin E-coated dialyzer; heparin-coated dialyzer; blood circuit clotting; fibrin ring; anticoagulation; saline flushes; end-stage renal disease (ESRD); randomized controlled trial; non-inferiority trial.

H

emodialysis requires effective anticoagulation to prevent circuit and dialyzer clotting,1-4 usually using continuous or intermittent administration of unfractionated heparin to maintain the activated cephalin time level at 80% of baseline.1 Inadequate From the 1Nephrology Department, Pasteur University Hospital of Nice, Nice; 2Nephrology Department, General Hospital of Avignon, Avignon; and 3Nice Sophia-Antipolis University, Nice, France. Received May 13, 2015. Accepted in revised form May 8, 2016. Originally published online June 23, 2016. Trial registration: www.ClinicalTrials.gov; study number: NCT01221337. Address correspondence to Mohamed Shariful Islam, MBBS, Nephrology Department, Pasteur University Hospital, 30 avenue de la Voie Romaine, 06001 Nice, France. E-mail: [email protected]  2016 by the National Kidney Foundation, Inc. 0272-6386 http://dx.doi.org/10.1053/j.ajkd.2016.05.013 752

anticoagulation may result in circuit clotting, with increased treatment cost, nurses’ workload, and blood transfusion requirement.1 It also decreases dialysis quality due to the clotting of multiple fibers, as well as protein cake formation inside the hemodialyzer.4-8 Systemic anticoagulation reduces clotting risk but also increases bleeding risk, especially in patients with kidney failure. This is due to numerous disease-related clotting abnormalities9,10 or to concomitant therapies (eg, antiplatelet agents, vitamin K antagonist, and catheter locking with heparin11). Heparin has numerous potential adverse effects: bleeding in cases of injury, prolonged and laborious compression of the needle puncture of fistulas, bioincompatibility with degranulation of neutrophils and platelets,12 pruritic allergic reactions, and heparin-induced thrombocytopenia.13 In order to reduce clotting and avoid bleeding events, several anticoagulation procedures have been developed: Am J Kidney Dis. 2016;68(5):752-762

Dialyzer Membrane for Heparin-Free Hemodialysis

regional heparin-protamine,14 minimal dose heparin,15 low-molecular-weight heparins,16-18 short dialysis,19 circuit rinsing by repeated saline solution flushes,20,21 heparin-primed hemodialyzers,21-24 regional citrate anticoagulation,25-30 citrate dialysate,31-33 predilution hemodiafiltration,34 human albumin-coated dialyzer,35 heparin-coated membrane,36 and vitamin E–coated membrane.37,38 Few randomized studies are available20,28,30,33,35,39,40 and evidence supporting heparin-free hemodialysis procedures is mainly based on observational or uncontrolled studies at risk of bias.24,38,41 None of the randomized controlled trials showed full efficacy.42 Heparin-coated and vitamin E–coated dialyzers seem to be 2 promising dialyzers, with vitamin E– coated dialyzers being of particular interest for patients with heparin-induced thrombocytopenia, for whom the heparin-coated dialyzers might be contraindicated.43 The aim of this randomized, multicenter, prospective, crossover study was to show the noninferiority of a vitamin E–coated dialyzer compared to a heparin-coated dialyzer in terms of extracorporeal circuit clotting events during heparin-free hemodialysis sessions.

METHODS Patient Selection Criteria Inclusion criteria were as follows: (1) 18 years or older; (2) long-term hemodialysis for at least 3 months, with bipunctured well-functioning fistulas (n 5 29) or long-term double-lumen tunneled catheters (n 5 3); (3) minimum blood flow rate of 300 mL/min; (4) mean of last 3 predialysis systolic blood pressures $ 110 mm Hg; (5) affiliation to French health care social security system; and (6) ability to understand and adhere to the protocol design. Patients’ usual anticoagulation was lowmolecular-weight heparin. Exclusion criteria were as follows: severe inflammatory disorders with C-reactive protein level . 30 mg/L, documented active external hemorrhage, severe anemia (hemoglobin , 10 g/dL) with or without blood transfusion requirement, documented thrombophilia, acute septic state, active diseases (dysglobulinemia, vasculitis, and human immunodeficiency virus [HIV] infection), vitamin K antagonists, history of one or more hypotensive event requiring a nurse’s intervention during the week preceding the study inclusion, on a kidney transplant waiting list, pregnant or breast-feeding, and protected by law on account of disabled status.

Membrane Characteristics Vitamin E–coated dialyzers consist of a polysulfone membrane

with a surface area of 2.1 m2 and ultrafiltration coefficient of 81 mL/h/mm Hg (Hemotech). Heparin-coated dialyzers consist of a polyacrylonitrile membrane with a surface area of 2.2 m2 and ultrafiltration coefficient of 65 mL/h/mm Hg (Hospal). Both dialyzers are high-permeability biocompatible membranes.

dialysis with either the vitamin E–coated dialyzer initially followed by the heparin-coated dialyzer or to undergo dialysis with the heparin-coated dialyzer initially, followed by the vitamin E– coated dialyzer (Fig 1). An independent clinical research assistant of our clinical research department in the university hospital of Nice, having no connection with the selected patients, performed randomization and dialyzer allocation providing a code number to identify each patient. The randomly assigned dialyzers were not concealed during the sessions and the dialyzer name was apparent on visual inspection. Study periods were separated by a washout period of 2 hemodialysis sessions with the patient’s usual heparin dose and dialyzer. During each study period, the usual heparin dose was reduced by 50% during the first hemodialysis session and by 25% for the second, followed by 2 successive heparin-free hemodialysis sessions (Table 1). This strategy of reducing heparin dose progressively allowed us to eliminate the membrane tolerance factor before starting the no-heparin sessions. The blood level was kept at the top of the dialysis circuit bubble trap when starting each session and lowered every hour in order to avoid fibrin ring formation. In cases of elevated venous pressure (.200 mm Hg), 200-mL saline solution flushes were permitted. Session interruption was decided and considered as a clotting event in the case of: (1) circuit venous pressure increase . 300 mm Hg without vascular access problem or mechanical obstruction not resolved by saline solution flushes and/or (2) transmembrane pressure . 100 mm Hg. In some cases, and when possible, the blood flow rate was slightly reduced to prevent higher circuit venous pressure in order to continue running the session. Any patient with a clotting event immediately entered the washout period. For these patients, the washout period could encompass more than the planned 2 sessions before switching to the second treatment period with the other dialyzer starting on the following Friday or Saturday (Table 1).

Enrollment (n=64)

Refused to participate (n=11)

Excluded for not meeting inclusion criteria (n=21)

Randomization (n=32)

Group A Vitamin E-coated (n=16)

Group B Heparin-coated (n=16)

Language barrier (n=3)

Total drop out due to clotting events (n=3)

Total drop out due to clotting events (n=2)

Wash out period (n=16)

Wash out period (n=16)

Cross-over to Heparin-coated (n=16)

Cross-over to Vitamin E-coated (n=16)

Total drop out due to clotting events (n=4)

Total drop out due to clotting events (n=4)

Study Design and Procedures This was an open-label randomized crossover study. The crossover design consisted of 2 study periods of 4 hemodialysis sessions (each session was 4 hours) starting on a Friday or Saturday according to the patient’s current dialysis schedule. Patients were randomly assigned 1:1 using balanced blocks of 4 to undergo Am J Kidney Dis. 2016;68(5):752-762

Figure 1. Study flow chart. 753

Islam et al Table 1. Study Periods Period 1: Vitamin E– (n 5 16) or Heparin-Coated (n 5 16) Dialyzer

Session Day Heparina Visit type

1 Fri/Sat 50% Exam

2 Mon/Tue 25% Exam

3 Wed/ Thu None Exam; blood test

4 Fri/Sat None Exam

Washout, Usual Dialyzer

5 6 Mon/Tue Wed/Thu 100% Washout period

Period 2: Heparin- (n 5 16) or Vitamin E2Coated (n 5 16) Dialyzer

7 Fri/Sat 50% Exam

8 Mon/Tue 25% Exam

9 Wed/Thu None Exam; blood test

10 Fri/Sat None Exam

Abbreviation: Exam, clinical examination. a Percentages are relative to usual dose.

Study End Points The primary end point was the percentage of successful study periods defined as no circuit-clotting event leading to premature interruption of any of the 4 dialysis sessions. After a failed session, subsequent sessions were imputed to failure. Secondary end points were as follows: (1) the number of heparin-sparing or heparin-free hemodialysis sessions without circuit clotting leading to premature end of the session, (2) the number of minutes across the 4 heparin-sparing or heparin-free hemodialysis sessions without or before circuit clotting leading to premature end of the session, (3) the number of saline solution flushes per study period and the number of patients requiring saline solution flushes, (4) grades given to the dialysis circuit bubble trap status by visual inspection, (5) grades given to the dialyzer membrane status by visual inspection, (6) dialysis adequacy (urea and creatinine reduction rate and equilibrated Kt/V [Daugirdas equation]), and (7) hemoglobin levels. Visual inspections of the dialysis circuit bubble trap and dialyzers were done at the end of the session. Dialysis circuit bubble trap status was graded by visual inspection as follows: 1 5 clean bubble trap, 2 5 fibrin ring, 3 5 nonobstructive clot, and 4 5 total circuit clotting. Dialyzer membrane status was graded by visual inspection as follows: 1 5 dialyzer has clean fiber bundles, 2 5 ,5% of fibers clotting, 3 5 multiple fibers clotting, and 4 5 total clotting of the dialyzer. Investigators were instructed to look for the fiber network by inspecting the tail of the dialyzers and not the lateral side to minimize bias based on whether the clotted fiber was near the edge of the membrane. Grades were assigned based on a consensus including opinions of physicians, nurses, and auxiliary staff. Blood samples were taken before and after the third hemodialysis session of each study period to measure blood electrolytes, serum creatinine, and urea. Hemoglobin levels, blood cell counts, albumin levels, C-reactive protein levels, activated cephalin time, international normalized ratio, and D-dimer were assessed only before this hemodialysis session. At the end of each hemodialysis session, a mixture of taurolidine and heparin was used as a catheter lock. Blood (5 mL) was withdrawn at the beginning of each hemodialysis session to eliminate the lock solution.

According to the results, the conclusions could be as follows: the vitamin E–coated dialyzer is inferior, not inferior but not superior, or superior to the heparin-coated dialyzer. At inclusion, characteristics of the 2 groups were compared. Quantitative values were compared with the Mann-Whitney test, and qualitative values, with the Fischer test. After completion of the study, the primary outcome (qualitative, ie, no clotting leading to premature interruption of the study period) and secondary outcomes (ie, total number of heparin-sparing or heparin-free hemodialysis sessions and total minutes across the 4 sessions of each period without or before massive circuit coagulation events, total number of saline solution flushes, number of patients requiring saline solution flushes, dialysis circuit bubble trap, and dialyzer visual inspection scores) were tested with a Dunnet c2 test for qualitative values and Student test for quantitative values.44,45 According to a recent publication, the success of heparin-free hemodialysis with a heparin-coated dialyzer is w90% during a single hemodialysis session.41 Thus, the success rate for 2 hemodialysis sessions can be considered as 80%. In this crossover study, the number of patients needed to test the noninferiority of a vitamin E–coated versus a heparin-coated dialyzer with a 20% risk b, a 12% accepted difference, and unilateral hypothesis was 32 (16 by group).46 All analyses and the power calculations were done using SAS software, version 9.1 (SAS Institute Inc).

Ethics After giving full oral and written information about study objectives and potential risks and benefits, patients gave their written agreement to participate. A reflection period of 7 days was given to each patient after signature. Patients could refuse to participate or withdraw their consents at any time during the clinical research period. Approval of the research ethics committee (Sud Mediterranée V, Nice) was obtained under reference number 10-043 on August 19, 2010 (registration number 2010-A00779-30), and the study was conducted in full compliance with current French and European Commission regulations.

RESULTS

Statistics

Study Flow and Patient Characteristics

The main criterion of efficacy was the success rate of a heparinfree hemodialysis strategy with vitamin E–coated compared with heparin-coated dialyzers. When the vitamin E–coated and heparincoated dialyzer group–period interaction (ie, carryover effect) was ruled out, the treatment effect of each dialyzer was tested with aggregated data for both periods. The analysis followed a 2-step procedure, and the probability that the efficacy of a vitamin E–coated dialyzer could be inferior to that of a heparin-coated dialyzer was tested first. If this probability was rejected at the 5% significance level on a unilateral test (ie, if noninferiority was demonstrated), superiority would be tested.

Thirty-two patients (mean age, 68 6 18 [standard deviation] years; 63% men) were included in this study. In November 2010 to March 2011, the university hospital of Nice recruited 24 patients, and the general hospital of Avignon, 8 patients. All 32 patients completed the study according to the protocol within the predetermined protocol period. A field clinical assistant was responsible for data collection on a printed chart, which was verified by

754

Am J Kidney Dis. 2016;68(5):752-762

Dialyzer Membrane for Heparin-Free Hemodialysis

the principal investigator, and then the data were entered in the database for computer analysis. A biostatistician controlled the data set and performed the statistical analyses. As shown in Table 2, baseline characteristics by randomly assigned groups, vitamin E–coated dialyzer first versus heparin-coated dialyzer first, respectively, were similar in terms of age (72 vs 65 years), male preponderance (75% vs 50%), mean weight (66.5 vs 67.2 kg), percentage of fistulas (94% vs 88%), and usual average heparin dose.

for vitamin E–coated dialyzers and 81% (26 of 32 study periods) for heparin-coated dialyzers (P 5 0.8). Comparing vitamin E2 and heparin-coated dialyzers, the secondary outcomes of number of successful heparin-sparing/free hemodialysis sessions (114 of 128 [89%] and 115 of 128 [90%], respectively; P 5 0.9) and cumulative durations across 4 hemodialysis sessions without clotting events (892 and 891 minutes per patient, respectively; P 5 0.8) were similar (Table 3). Comparable numbers of saline solution flushes were used with either dialyzer. Similarly, numbers of patients requiring saline solution flushes were not significantly different (Table 3). Visual inspection showed comparable numbers of clean dialysis circuit bubble traps (vitamin E–coated: 34 of 121; heparin-coated, 32 of 120; Table 4). Despite transparency differences between the 2 dialyzers, visual inspection of membrane status (Table 4) suggested superiority of the vitamin E– coated dialyzer compared to the heparin-coated dialyzer (grade 1 [dialyzer has clean fiber bundles] in 25 of 121 vs 2 of 120, respectively; P 5 0.002; Figs 2 and 3 show examples of the 4 grades assigned during visual inspection of the dialysis circuit status and dialyzer membrane status, respectively).

Primary and Secondary Outcomes

Other Outcomes

No period effect was detected (P 5 0.2), so the 2 study periods were combined for each dialyzer type in calculation of the primary and secondary outcomes. As shown in Table 3, the percentage of successful study periods (defined as no circuit-clotting event leading to premature interruption of any of the 4 dialysis sessions) was 78% (25 of 32 study periods)

Transmembrane pressure levels were within normal ranges in both dialyzers, excluding periods of intradialyzer fibrin clotting events (Table 3). Mean blood flow rates were similar in both dialyzers (vitamin E–coated, 340 mL/min; heparincoated, 344 mL/min; Table 3). Nevertheless, blood flow rates were lower during sessions complicated

Table 2. Baseline Characteristics by Dialyzer Randomization Group

Age, y Male sex Mean weight, kg Fistula Mean heparin dose, U

Vitamin E–Coated First (n 5 16)

Heparin-Coated First (n 5 16)

P

72 6 15 12 (75) 66.5 6 10 15 (94) 4,375 6 1,088

65 6 22 8 (50) 67.2 6 13 14 (88) 4,375 6 1,500

0.3 0.3 0.9 0.9 0.5

Note: Values for categorical variables are given as number (percentage); values for continuous variables, as mean 6 standard deviation.

Table 3. Selected Outcomes

Primary outcome No. of successfulb study periods Secondary outcomesc Total no. of sessions without clotting Total cumulative time before or without massive circuit clotting events, min/patient Total no. of saline flushes per study period Total no. of patients who needed saline flushes Other outcomes Mean transmembrane pressure, mm Hg Mean blood flow rate, mL/min

Vitamin E–Coated (n 5 32)a

Heparin-Coated (n 5 32)a

P

25/32 (78%)

26/32 (81%)

0.8

114/128 (89%) 892

115/128 (90%) 891

0.9 0.8

26 19 (59%)

25 20 (63%)

0.6 0.6

23.5 6 20 340 6 26

24.2 6 21 344 6 23

0.9 0.8

Note: Values for categorical variables are given as n/N or number (percentage); values for continuous variables, as mean 6 standard deviation. a Due to crossover design, all 32 participants in the study were studied with both dialyzers. b Defined as no circuit-clotting event leading to a premature interruption of any of the 4 dialysis sessions in that study period. c Excluding visual inspection results, which are shown in Table 4, and equilibrated Kt/V and hemoglobin level, which are given in Table 5. Am J Kidney Dis. 2016;68(5):752-762

755

Islam et al Table 4. Subjective Evaluation of Dialysis Circuit Bubble Trap Status and Dialyzer Membrane Status During 4 Hemodialysis Sessions

Subjective Evaluation by Visual Inspection

Vitamin E–Coated (n 5 121 sessions)

Heparin-Coated (n 5 120 sessions)

Dialysis circuit bubble trap status Grade 1: Clean dialysis circuit bubble trap Grade 2: Fibrin ring Grade 3: Nonobstructive clot Grade 4: Total circuit clotting

34 19 63 5

(28) (16) (52) (4)

32 26 57 5

(27) (22) (47) (4)

Dialyzer membrane status Grade 1: Dialyzer has clean fiber bundles (white/cream color) Grade 2: ,5 of fibers clotting Grade 3: Multiple fibers clotting (red or rose-colored dialyzer) Grade 4: Total clotting of dialyzer

25 72 22 2

(21) (59) (18) (2)

2 30 83 5

(2) (25) (69) (4)

P

0.2

0.002

Note: Values are given as number (percentage). Photographs with examples of each grade are provided in Figs 2 and 3.

with clotting events (Fig 4A and B). However, the between-dialyzer difference was not statistically significant. Figure 5 shows the success rate (no circuit-clotting event leading to premature interruption of a session), but by session rather than overall across all 4 sessions of a given study period. Success rates for the noheparin sessions (sessions 3 and 4) were nominally lower than rates for heparin-sparing sessions (sessions 1 and 2).

both dialyzer groups. Blood D-dimer levels were higher than normal (.500 mg/L) in both dialyzer groups (vitamin E–coated, 1,171 mg/L; heparincoated, 1,190 mg/L), but the standard deviation was broad and there was no correlation with clotting events. Activated cephalin times and international normalized ratios were similar at the start of the first no-heparin session in each study period. This excludes possible effects of other anticoagulant agents and any concomitant bleeding disorders.

Laboratory Values

Protocol Deviations

Overall, C-reactive protein levels were not different between groups (Table 5). Other laboratory parameters, such as serum albumin and hemoglobin levels, as well as dialysis efficacy parameters, were similar in

One patient refused to complete the last session. Because this was the second session of the heparinfree treatment period, we imputed the last session for this patient to success because the preceding

Figure 2. Examples of different grades by visual inspection of dialysis circuit bubble trap status (grade 1: clean bubble trap; grade 2: fibrin ring; grade 3: nonobstructive clot; grade 4: total circuit clotting). 756

Am J Kidney Dis. 2016;68(5):752-762

Dialyzer Membrane for Heparin-Free Hemodialysis

Figure 3. Examples of different grades by visual inspection of dialyzer membrane status for (top images) vitamin E–coated dialyzers and (bottom images) heparin-coated dialyzers. (Grade 1: dialyzer has clean fiber bundles; 2: ,5% of fibers clotting; grade 3: multiple fibers clotting; 4: total clotting of dialyzer).

Am J Kidney Dis. 2016;68(5):752-762

757

Islam et al A

360

dialyzer. However, the first session of the second period also ended with a single needle and premature circuit clotting. Two patients reluctant to modify their usual treatment duration were allowed shorter than planned sessions (3 hours 30 minutes and 3 hours 45 minutes instead of 4 hours). These 2 patients had these same durations of dialysis sessions in both treatment periods with both dialyzers.

Blood flow rate in vitamin E-coated dialyzers

350

Qb ml/min

340 330 320 310 300 290

Qb_T0h

Qb_T1h

Qb_T2h

Qb_T3h b T

Qb_Tfin

349

350

349

349

342

320

340

340

292

330

Vitamin E-coated / no circuit clotting Vitamin E-coated / circuit clotting

B

360

Blood flow rate in heparin-coated group

Qb ml/min

350 340 330 320 310 300 290 Heparin-coated / no circuit clotting Heparin-coated / circuit clotting

Qb_T0h

Qb_T1h

Qb_T2h

Qb_T3h

Qb_Tfin

351

352

349

348

344

348

347

347

347

344

Figure 4. Circuit-clotting events as a function of mean blood flow rate (Qb; in milliliters per minute) for (A) vitamin E–coated and (B) heparin-coated dialyzers.

session was successful. Another patient with fistula dysfunction due to a puncture problem continued the sessions with a single needle and experienced a clotting event leading to premature interruption during the first session with the 50% heparin dose. Such an event occurred again during the second session with the 25% heparin dose. According to the study design, this patient entered the washout period of 4 sessions with a single needle to allow fistula recovery and use of a double needle after switching to the other

81% 78%

Session 4 (no heparin)

84% 88%

Session 3 (no heparin) Session 2 (25% heparin)

97% 94%

Session 1 (50% heparin)

97% 97%

0%

20%

Heparin-coated

40%

60%

80%

100%

120%

Vitamin E-coated

Figure 5. Per-session success rate. Success rate is defined as no circuit-clotting event leading to a premature end to the dialysis session. After a failed session, subsequent sessions were imputed to failure. 758

DISCUSSION In this multicenter, prospective, randomized, crossover study, we showed that a vitamin E–coated dialyzer was not inferior to a heparin-coated dialyzer for the primary outcome, the percentage of successful study periods (defined as no circuit-clotting event leading to premature interruption of any of the 4 dialysis sessions). The success rate found in our study was lower for both dialyzers than that reported by the manufacturers because claimed success rates for specific membrane dialyzers for heparin-free hemodialysis were as high as 97% in previously published uncontrolled studies.24,38,41 In our randomized controlled trial, we had unacceptably high failure rates with vitamin E–coated and heparin-coated dialyzers (22% and 19%, respectively) in our attempt to decrease heparin doses and then perform 2 consecutive heparin-free 4-hour hemodialysis sessions. A recent prospective, multicenter, international, open-label, randomized controlled study, carried out after our trial in November 2011 to February 2013, showed a better success rate with heparin-coated dialyzers than with standard of care.39 However, the success rate with heparin-coated dialyzers was only 68.5%. Another randomized controlled trial showed success rates comparable to those found in our study; that study found success rates of 85% for the AN69ST membrane versus 81% for the polysulfone membrane in patients with a stable condition, of which a few were taking vitamin K antagonists.40 However, in that study, only 3-hour heparin-free hemodialysis sessions were performed. A recent prospective randomized crossover trial showed that heparin-free hemodialysis is feasible in patients taking an oral anticoagulant with minimal level of inflammation and adequate blood flow, with no advantage of heparin-coated membranes over polysulfone membranes.47 Specific membrane dialyzers, such as vitamin E–coated and heparincoated, may only allow lowering of the heparin dose with putative clinical benefit.38,41 Vitamin E–coated dialyzers have been claimed to have antioxidant and biocompatible properties with long-term beneficial effects38,48-50 on inflammatory processes, platelet aggregation, and erythropoiesis.38,50 In heparin-free hemodialysis, this dialyzer may also be an acceptable alternative to treat patients Am J Kidney Dis. 2016;68(5):752-762

Dialyzer Membrane for Heparin-Free Hemodialysis Table 5. Blood Tests Before and After the First No-Heparin Session of Each Treatment Period

CRP, mg/L Albumin, g/L Hemoglobin, g/dL Rate of urea reduction, % Rate of creatinine reduction, % Equilibrated Kt/Va (Daugirdas equation) D-Dimer, mg/L Activated cephalin time, s International normalized ratio

Vitamin E–Coated (n 5 32)a

Heparin-Coated (n 5 32)a

P

8.18 (0.30-39) 38.7 (31-45) 11.20 6 1.40 77 6 5 71 6 5 1.48 6 0.30 1,171 6 1,100 35 6 4.37 1.04 6 0.06

12 (0.30-41) 38.4 (30-46) 11.20 6 1.50 75 6 8 68 6 7 1.43 6 0.30 1,190 6 1,113 35 6 3.89 1.04 6 0.06

0.9 0.8 0.9 0.9 0.9 0.07 0.9 0.9 0.9

Note: Values are given as mean 6 standard deviation or median (minimum-maximum). Abbreviation: CRP, C-reactive protein. a Due to crossover design, all 32 participants in the study were studied with both dialyzers.

with heparin-induced thrombocytopenia who cannot be exposed to heparin-coated dialyzers. Using visual inspection of dialyzer membrane status, we observed less clotting with the vitamin E–coated dialyzer than with the heparin-coated dialyzer. In our short-term study, we did not detect differences in nutritional parameters, hemoglobin levels, and adequacy of dialysis between both dialyzers. Saline solution flushes are commonly considered as an appropriate method of blood dilution to avoid intradialyzer fibrin clots. This method has been frequently used in association with specific membrane technologies.21 However, it does not reduce the risk for clotting.51,52 Furthermore, during saline solution flushes, we observed sudden elevations in venous pressures in the blood circuit linked to the dislocation of the fibrin ring in the dialysis circuit bubble trap followed by circuit clotting. In these patients, the circuit clotting was not due to the dialyzer clotting because transmembrane pressures remained within normal ranges throughout the sessions with both dialyzers (23.5 vs 24.2 mm Hg for vitamin E–coated and heparin-coated dialyzers, respectively). Alternative anticoagulation modalities have been tested. Regional extracorporal circuit anticoagulation with heparin followed by protamine neutralization has been abandoned because of a high risk for delayed bleeding because heparin has a longer elimination half-life than protamine.14 Lowering the standard continuous dose of heparin15 or using low-molecularweight heparins16 does not reduce rates of clotting or bleeding events.24 Citrate regional anticoagulation and citric acid dialysate yield better results to decrease dialyzer clotting during heparin-free hemodialysis, especially in sustained low-efficiency dialysis in critically ill patients.28,53-57 Nevertheless, these methods require close monitoring of serum electrolyte levels, especially calcium, and expose patients to metabolic disorders and frequent nurse interventions.29,53,55,57-59 Am J Kidney Dis. 2016;68(5):752-762

The main limitation of the present study is the inclusion of select stable patients with well-functioning fistulas and double-lumen catheters, thus favoring a better success rate. We excluded single-needle dialysis, 1-lumen central venous catheters, and critically ill patients with hemodynamic instability and acute inflammatory and infectious disorders. Therefore, our results cannot be extrapolated to critically ill patients at high bleeding risk who may require heparin-free hemodialysis.60 The only randomized trial in critically ill patients at high bleeding risk that we are aware of failed to recruit the required number of patients and was stopped prematurely: only 39 unstable patients performed 1 hemodialysis session with AN69ST and 1 session with polysulfone dialyzers, with no significant differences between groups.61 Inflammation and fistula and catheter dysfunction with low extracorporeal blood flow are common causes of circuit clotting.51-62 In our study, one patient had a fistula dysfunction after randomization and could be dialyzed with only a single needle. He experienced massive clotting of the extracorporeal circuit before hour 3 of the 50% heparin hemodialysis session with both dialyzers. High blood flow may be an important factor of circuit survival during heparin-free hemodialysis that has not yet been evaluated by randomized controlled studies.4,8,51 In our study, mean blood flow rates were similar and high (340 vs 344 mL/min with vitamin E– coated and heparin-coated dialyzers, respectively). Nevertheless, patients with circuit clotting events had nominally lower blood flow rates (Fig 4A and B). High blood flow rates diminish the time of blood exposure in extracorporal circuits, thus lowering clotting events. Undetected blood flow reductions occur during continuous venovenous hemofiltration, which is strongly correlated with filter life span.62 Further research is needed to improve the success rate of heparin-free hemodialysis, especially in critically ill patients with high bleeding risk and 759

Islam et al

inflammatory states. Most previous studies did not take into account mechanical factors such as vascular access dysfunctions, blood flow rates, length of the extra corporal circuit, the air-blood contact in dialysis circuit bubble traps, and number of dialysis circuit bubble traps, especially in single-needle circuits. All these factors may have a substantial influence on the number of clotting events. The disappointing results of the previous heparin-free hemodialysis trials may be due to unexplored mechanical factors. In conclusion, both vitamin E–coated and heparincoated dialyzers expose patients to an unacceptably high rate of failure of 4-hour heparin-free hemodialysis sessions, and vitamin E–coated dialyzers are not inferior to heparin-coated dialyzers. Success rates in this prospective randomized trial were lower than previously claimed in uncontrolled studies. Several mechanical factors and a combined strategy with citric acid dialysate should be considered in future trials, rather than focusing on only specific membrane technologies, in order to improve heparin-free hemodialysis strategies.

ACKNOWLEDGEMENTS Preliminary data for this study were presented as a poster at the 2011 ASN Kidney Week meeting, Philadelphia, PA, November 8-13, 2011. We thank Mr Jean-Charles Kerihuel, our English language medical writer; Ms Vanina Oliveri and a second clinical research assistant; and our nursing staff for participation in this study. Support: This work was promoted by the clinical research department of Nice University Hospital with financial support by an unrestricted grant from Hemotech Pharmaceutical. The sponsor had no role in study design or collection, analysis, and interpretation of data; it was involved in neither writing the report nor the decision to submit the report for publication. Financial Disclosure: The authors declare that they have no other relevant financial interests. The hemodialysis center of Nice, as well as that of Avignon, is a usual business partner of Hemotech and Hospal. Both may occasionally and partially support travel and hotel expenses when medical staffs attend scientific meetings. Contributions: Research idea: MSI; study design: MSI, VLME, OM; study monitoring and supervision: MSI, FC, SV, MB (University Hospital of Nice) and ZAH, DV, PD (General Hospital of Avignon); supervision of data collection: MSI (Nice), ZAH (Avignon); statistical analysis: OM; data analysis/interpretation: MSI, VLME, OM. Each author contributed important intellectual content during manuscript drafting or revision and accepts accountability for the overall work by ensuring that questions pertaining to the accuracy or integrity of any portion of the work are appropriately investigated and resolved. MSI takes responsibility that this study has been reported honestly, accurately, and transparently; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned and registered have been explained. Peer Review: Evaluated by 2 external peer reviewers, a Statistical Editor, a Co-Editor, and the Editor-in-Chief.

REFERENCES 1. Fischer KG. Essentials of anticoagulation in hemodialysis. Hemodial Int. 2007;11(2):178-189. 760

2. Suranyi M, Chow JS. Review: anticoagulation for haemodialysis. Nephrology (Carlton). 2010;15(4):386-392. 3. Murray PT, Reddy BV, Grossman EJ, et al. A prospective comparison of three argatroban treatment regimens during hemodialysis in end-stage renal disease. Kidney Int. 2004;66(6): 2446-2453. 4. Joannidis M, Oudemans-van Straaten HM. Clinical review: patency of the circuit in continuous renal replacement therapy. Crit Care. 2007;11(4):218. 5. Hofbauer R, Moser D, Frass M, et al. Effect of anticoagulation on blood membrane interactions during hemodialysis. Kidney Int. 1999;56(4):1578-1583. 6. Frank RD, Weber J, Dresbach H, et al. Role of contact system activation in hemodialyzer-induced thrombogenicity. Kidney Int. 2001;60(5):1972-1981. 7. Romao JE Jr, Fadil MA, Sabbaga E, et al. Haemodialysis without anticoagulant: haemostasis parameters, fibrinogen kinetic, and dialysis efficiency. Nephrol Dial Transplant. 1997;12(1): 106-110. 8. Keller F, Seemann J, Preuschof L, et al. Risk factors of system clotting in heparin-free haemodialysis. Nephrol Dial Transplant. 1990;5(9):802-807. 9. Lohr JW, Schwab SJ. Minimizing hemorrhagic complications in dialysis patients. J Am Soc Nephrol. 1991;2(5):961-975. 10. Nampoory MR, Das KC, Johny KV, et al. Hypercoagulability, a serious problem in patients with ESRD on maintenance hemodialysis, and its correction after kidney transplantation. Am J Kidney Dis. 2003;42(4):797-805. 11. McGill RL, Blas A, Bialkin S, Sandroni SE, Marcus RJ. Clinical consequences of heparin-free hemodialysis. Hemodial Int. 2005;9(4):393-398. 12. Gritters M, Borgdorff P, Grooteman MP, et al. Platelet activation in clinical haemodialysis: low-molecular-weight heparins as a major contributor to bio-incompatibility? Nephrol Dial Transplant. 2008;23(9):2911-2917. 13. Shen JI, Winkelmayer WC. Use and safety of unfractionated heparin for anticoagulation during maintenance hemodialysis. Am J Kidney Dis. 2012;60(3):473-486. 14. Blaufox MD, Hampers CL, Merrill JP. Rebound anticoagulation occurring after regional heparinization for hemodialysis. Trans Am Soc Artif Intern Organs. 1966;12: 207-209. 15. Swartz RD, Port FK. Preventing hemorrhage in high-risk hemodialysis: regional versus low-dose heparin. Kidney Int. 1979;16(4):513-518. 16. Schrader J, Stibbe W, Armstrong VW, et al. Comparison of low molecular weight heparin to standard heparin in hemodialysis/ hemofiltration. Kidney Int. 1988;33(4):890-896. 17. Lim W, Cook DJ, Crowther MA. Safety and efficacy of low molecular weight heparins for hemodialysis in patients with endstage renal failure: a meta-analysis of randomized trials. J Am Soc Nephrol. 2004;15(12):3192-3206. 18. Davenport A. Review article: low-molecular-weight heparin as an alternative anticoagulant to unfractionated heparin for routine outpatient haemodialysis treatments. Nephrology. 2009;14(5):455-461. 19. Brunet P, Simon N, Opris A, et al. Pharmacodynamics of unfractionated heparin during and after a hemodialysis session. Am J Kidney Dis. 2008;51(5):789-795. 20. Ramesh Prasad GV, Palevsky PM, Burr R, et al. Factors affecting system clotting in continuous renal replacement therapy: results of a randomized, controlled trial. Clin Nephrol. 2000;53(1): 55-60. Am J Kidney Dis. 2016;68(5):752-762

Dialyzer Membrane for Heparin-Free Hemodialysis 21. Richtrova P, Opatrny K Jr, Vit L, et al. The AN69 ST haemodialysis membrane under conditions of two different extracorporeal circuit rinse protocols a comparison of thrombogenicity parameters. Nephrol Dial Transplant. 2007;22(10): 2978-2984. 22. Lavaud S, Canivet E, Wuillai A, et al. Optimal anticoagulation strategy in haemodialysis with heparin-coated polyacrylonitrile membrane. Nephrol Dial Transplant. 2003;18(10): 2097-2104. 23. Chanard J, Lavaud S, Maheut H, et al. The clinical evaluation of low-dose heparin in haemodialysis: a prospective study using the heparin-coated AN69 ST membrane. Nephrol Dial Transplant. 2008;23(6):2003-2009. 24. Kim YG. Anticoagulation during haemodialysis in patients at high-risk of bleeding. Nephrology (Carlton). 2003;8(suppl): S23-S27. 25. Oudemans-van Straaten HM. Citrate anticoagulation for continuous renal replacement therapy in the critically ill. Blood Purif. 2010;29(2):191-196. 26. Gritters M, Grooteman MP, Schoorl M, et al. Citrate anticoagulation abolishes degranulation of polymorphonuclear cells and platelets and reduces oxidative stress during haemodialysis. Nephrol Dial Transplant. 2006;21(1):153-159. 27. See KC, Lee M, Mukhopadhyay A. Delay in onset of metabolic alkalosis during regional citrate anti-coagulation in continuous renal replacement therapy with calcium-free replacement solution. Ann Card Anaesth. 2009;12(2):122-126. 28. Evenepoel P, Dejagere T, Verhamme P, et al. Heparincoated polyacrylonitrile membrane versus regional citrate anticoagulation: a prospective randomized study of 2 anticoagulation strategies in patients at risk of bleeding. Am J Kidney Dis. 2007;49(5):642-649. 29. Kozik-Jaromin J, Nier V, Heemann U, et al. Citrate pharmacokinetics and calcium levels during high-flux dialysis with regional citrate anticoagulation. Nephrol Dial Transplant. 2009;24(7):2244-2251. 30. Fealy N, Baldwin I, Johnstone M, et al. A pilot randomized controlled crossover study comparing regional heparinization to regional citrate anticoagulation for continuous venovenous hemofiltration. Int J Artif Organs. 2007;30(4):301-307. 31. Ahmad S, Callan R, Cole JJ, et al. Dialysate made from dry chemicals using citric acid increases dialysis dose. Am J Kidney Dis. 2000;35(3):493-499. 32. Gabutti L, Lucchini B, Marone C, et al. Citrate- vs. acetatebased dialysate in bicarbonate haemodialysis: consequences on haemodynamics, coagulation, acid-base status, and electrolytes. BMC Nephrol. 2009;10:7. 33. Kossmann RJ, Gonzales A, Callan R, et al. Increased efficiency of hemodialysis with citrate dialysate: a prospective controlled study. Clin J Am Soc Nephrol. 2009;4(9):1459-1464. 34. van der Voort PH, Gerritsen RT, Kuiper MA, et al. Filter run time in CVVH: pre- versus post-dilution and nadroparin versus regional heparin-protamine anticoagulation. Blood Purif. 2005;23(3):175-180. 35. Wu X, Chen H. Clinical study of heparin-free hemodialysis with the inside of hollow fibers in dialyzer coated by human albumins. Nephron. 2002;92(4):925-928. 36. Morena M, Jaussent I, Chalabi L, et al. Biocompatibility of heparin-grafted hemodialysis membranes: impact on monocyte chemoattractant protein-1 circulating level and oxidative status. Hemodial Int. 2010;14(4):403-410. 37. Sasaki M. Development of vitamin E-modified polysulfone membrane dialyzers. J Artif Organs. 2006;9(1):50-60.

Am J Kidney Dis. 2016;68(5):752-762

38. Huraib S, Tanimu D, Shaheen F, et al. Effect of vitaminE-modified dialysers on dialyser clotting, erythropoietin and heparin dosage: a comparative crossover study. Am J Nephrol. 2000;20(5):364-368. 39. Laville M, Marc Dorval M, Fort Ros J, et al. Results of the HepZero study comparing heparin-grafted membrane and standard care show that heparin-grafted dialyzer is safe and easy to use for heparin-free dialysis. Kidney Int. 2014;86: 1260-1267. 40. Brunet P, Frances J, Vacher-Coponat H, et al. Hemodialysis without heparin: a randomized, controlled, crossover study of two dialysis membranes (AN69ST and polysulfone F60). Int J Artif Organs. 2011;34(12):1165-1171. 41. Nilsson LG, Bosch J. Managing dialysis with low systemic heparinisation – an observational survey on the usage of the new Evodial dialyzer. http://gambramedical.ru/assets/files/Evodial% 20PMCF%20final%20white%20paper_G.pdf. Accessed June 21, 2016. 42. Meijers BKI, Poesen R, Evenepoel P. Heparin-coated dialyzer membranes: is non-inferiority good enough? Kidney Int. 2014;86:1084-1086. 43. Matsuo T, Wanaka K. Heparin-induced thrombocytopenia and hemodialysis. J Blood Disord Transfus. 2011;suppl 2: 2155-9864. 44. Dunnett CW, Gent M. Significance testing to establish equivalence between treatments, with special reference to data in the form of 2X2 tables. Biometrics. 1977;33(4):593-602. 45. Schuirmann DJ. A comparison of the two one-sided tests procedure and the power approach for assessing the equivalence of average bioavailability. J Pharmacokinet Biopharm. 1987;15(6): 657-680. 46. Tubert-Bitter P, Manfredi R, Lellouch J, Begaud B. Sample size calculations for risk equivalence testing in pharmacoepidemiology. J Clin Epidemiol. 2000;53:1268-1274. 47. Krummel T, Scheidt E, Borni-Duval C, et al. Haemodialysis in patients treated with oral anticoagulation: should we heparinize? Nephrol Dial Transplant. 2014;29(4):906-913. 48. Yang CC, Hsu SP, Wu MS, et al. Effects of vitamin C infusion and vitamin E-coated membrane on hemodialysisinduced oxidative stress. Kidney Int. 2006;69(4):706-714. 49. Morimoto H, Nakao K, Fukuoka K, et al. Long-term use of vitamin E-coated polysulfone membrane reduces oxidative stress markers in haemodialysis patients. Nephrol Dial Transplant. 2005;20(12):2775-2782. 50. Cruz DN, De Cal M, Garzotto F, et al. Effect of vitamin Ecoated dialysis membranes on anemia in patients with chronic kidney disease: an Italian multicenter study. Int J Artif Organs. 2008;31(6):545-552. 51. Caruana RJ, Raja RM, Bush JV, et al. Heparin free dialysis: comparative data and results in high risk patients. Kidney Int. 1987;31(6):1351-1355. 52. Sagedal S, Hartmann A, Osnes K, et al. Intermittent saline flushes during haemodialysis do not alleviate coagulation and clot formation in stable patients receiving reduced doses of dalteparin. Nephrol Dial Transplant. 2006;21(2):444-449. 53. Brophy PD, Somers MJ, Baum MA, et al. Multi-centre evaluation of anticoagulation in patients receiving continuous renal replacement therapy (CRRT). Nephrol Dial Transplant. 2005;20(7):1416-1421. 54. Kutsogiannis DJ, Gibney RT, Stollery D, et al. Regional citrate versus systemic heparin anticoagulation for continuous renal replacement in critically ill patients. Kidney Int. 2005;67(6): 2361-2367.

761

Islam et al 55. Morabito S, Pistolesi V, Tritapepe L, Fiaccadori E. Regional citrate anticoagulation RRTs in critically ill patients with AKI. Clin J Am Soc Nephrol. 2014;9:2173-2188. 56. Sands JJ, Kotanko P, Segal JH, et al. Effects of citrate acid concentrate on heparin N requirements and hemodialysis adequacy: a multicenter, prospective, noninferiority trial. Blood Purif. 2012;33(1-3):199-204. 57. Clark JA, Schulman G, Golper TA. Safety and efficacy of regional citrate anticoagulation during 8-hour sustained lowefficiency dialysis. Clin J Am Soc Nephrol. 2008;3(3):736-742. 58. Kreuzer M, Bonzel KE, Buscher R, et al. Regional citrate anticoagulation is safe in intermittent high-flux haemodialysis treatment of children and adolescents with an increased risk of bleeding. Nephrol Dial Transplant. 2010;25(10):3337-3342.

762

59. Davenport A, Tolwani A. Citrate anticoagulation for continuous renal replacement therapy (CRRT) in patients with acute kidney injury admitted to the intensive care unit. Nephrol Dial Transplant. 2009;2(6):439-447. 60. Amanzadeh J, Reilly RF Jr. Anticoagulation and continuous renal replacement therapy. Semin Dial. 2006;19(4):311-316. 61. Lefevre LM, Dugard A, Thaler F, Robert R, Mentec H, Boulain T. Incidence de coagulation du circuit extra-corporel lors de dialyse par rinçage ou par filtre AN69ST sans héparines systémiques chez des patients à risque hémorragique élevé en réanimation. Réanimation. 2009;18S:S17-S18. 62. Baldwin I, Bellomo R, Koch B. Blood flow reductions during continuous renal replacement therapy and circuit life. Intensive Care Med. 2004;30(11):2074-2079.

Am J Kidney Dis. 2016;68(5):752-762