Hemodialysis treatment requires anticoagulation, usually with unfractionated or low molecular weight heparin to prevent thrombosis of the dialyzer and the extracorporeal circuit (dialyzer and blood lines or cassette system). In clinical practice, it is not unusual to perform hemodialysis for patients with active bleeding or increased bleeding risk conditions when heparin anticoagulation is contraindicated. Various solutions have been attempted to prevent clotting of the extracorporeal circuit.

Heparin-free dialysis using intermittent saline flushes is widely adopted as the method of choice for patients at risk of bleeding 1 2 and is currently recommended by the European Best Practice Guidelines in hemodialysis (2002) 3 , although the level of evidence is somewhat weak since no randomized controlled studies evaluating this method have been reported to date 4 5 6 7 8 9 10 11 . Heparin-free treatment with on-line predilution rather than regular saline flushes is also a technique used at some dialysis units. Saline infusion is far from optimal for several reasons, including an increased volume load that would need to be removed with dialysis and an added logistic burden on dialysis nurses, owing to the need for close one-to-one nursing in cases of intermittent boluses 1 . In addition, this technique is still associated with clotting risks (15% to 35%, depending on the literature) which, in addition to blood loss, may reduce the efficiency of a dialysis session that has to be stopped frequently or prematurely. Regional (heparin administration into the arterial line and protamine into the venous line) or tight heparinization (use of minimal dose of heparin) is not recommended for patients with active bleeding or at risk of bleeding 3 . Indeed, although these methods lower the risk of bleeding compared to the standard method, there is still a notable risk of bleeding (5% to 50%) 12 .

The use of regional citrate anticoagulation (RCA) is limited by the need for additional equipment (additional pumps for citrate and calcium), the potential risk of deterioration of electrolyte and acid–base equilibrium (hypocalcemia, hyponatremia) requiring close monitoring of electrolytes 12 , and a need for clotting time evaluation since the risk of bleeding is reduced but may still occur. Recently, the feasibility of anticoagulating the extracorporeal circuit during hemodialysis using a simple citrate-enriched dialysate was evaluated in a pilot (no prespecified sample size calculation or primary endpoint), monocentric, prospective, randomized, crossover study of 24 patients with high risk of bleeding 13 . For anticoagulation of the extracorporeal circuit, one treatment used the citrate-enriched dialysate (citrate group), while the other treatment used conventional saline flushing (saline group). With either method, a heparinized, saline-rinsed polyamide dialyzer was used and no heparin was administered during dialysis. Ninety-two per cent (22 out of 24) and 100% of patients tolerated the procedure well in the citrate group and saline group, respectively. Eight per cent (two out of 24) of the treatments in each group had to be abandoned because of clotting in the extracorporeal circuit, while significantly less thrombus formation in the venous air traps was detected in the citrate group 13 . Other alternatives such as prostacyclin are costly and require close hemodynamic monitoring 3 . In a pilot (without sample size calculation or prespecified primary endpoint), monocentric, randomized, controlled trial, Yixiong et al. recently compared heparin-free saline flushes with low-dose argatroban (a direct thrombin inhibitor) in a total of 80 hemodialysis sessions performed in 52 patients with high risk of bleeding. They observed a three-fold decrease in major clotting with argatroban (P <0.05), without safety issues 14 .

Another alternative for heparin-free hemodialysis is to coat heparin on the hemodialyzer hollow fibers 1 12 . Stamatiadis et al. reported the results of 16,954 sessions performed with patients bleeding or at risk of bleeding (15,730 retrospectively and 1,224 prospectively collected), with a method combining a priming of two types of hollow fiber wet dialyzers (ethylene vinyl alcohol and polyethylene glycol-coated membranes with 1 liter of saline containing 5,000 IU of unfractionated heparin). After initiation of the treatment, except for hourly rinsing with 50 ml normal saline to inspect the circuit (100 ml for a 3-hour session), no other special nursing attendance was applied. Cumulative failure of treatment, as defined by clotting of the extracorporeal circuit requiring termination of the procedure or replacement of the clotted part, was not greater than 5% 15 .

Gambro (Lund, Sweden) have for several years developed coated hemodialyzers such as Nephral ST that are equipped with a membrane, the surface of which is treated with polyethyleneimine (PEI) allowing extemporaneous coating with heparin when the hemodialyzer is primed with saline solution containing heparin. Several clinical studies have been performed with Nephral ST to demonstrate the possibility of decreasing systemic heparin doses or of even performing hemodialysis treatments without heparin 16 17 18 . In a study aimed at evaluating the AN69 ST membrane in heparin-free dialysis in patients at risk of bleeding (n = 68) less than 2% massive clotting were observed (6 out of 331 dialysis sessions) 17 . Evenepoel et al. performed a monocenter, prospective, randomized trial including 33 hemodialysis patients at high risk of bleeding, in whom regional anticoagulation was achieved by means of either AN69 ST (Nephral ST; 11 patients, 31 sessions), RCA-Ca0 (11 patients, 32 sessions) or RCA-Ca3.0 (11 patients, 30 sessions). Clotting phenomena necessitating premature termination of the dialysis session were encountered in 39%, 13% and 0% using AN69 ST, RCA-Ca3.0 and RCA-Ca0, respectively (P <0.005). The authors concluded to the superiority of citrate regional anticoagulation 19 . In a bicentric, Austrian, randomized, crossover trial, Kodras et al. treated 10 patients receiving oral anticoagulants with AN69 ST and FX 100 (a polysulfone membrane) for 1 week (total 30 sessions of heparin-free dialysis, with only priming of the dialyzer and no systemic heparinization). Changes of clotting markers, occurrences of complete thrombosis and Kt/V were similar with both membranes 20 . More recently, a prospective, randomized, crossover study examined 10 stable patients during intermittent hemodialysis with: 1. regular saline flushes of extracorporeal circuit; 2. RCA; and 3. AN69 ST membrane after extracorporeal circuit priming. All 10 procedures with RCA were successfully completed after 4 hours, whereas six out of 10 procedures with saline flushes and five out of 10 procedures with AN69 ST were terminated prematurely because of clotting (P <0.05). Due to insufficient statistical power of the negative results, the number of incomplete procedures did not allow the authors to directly compare saline flushes and AN69 ST 21 .

A new dialyzer, Evodial (Gambro), which is an upgrade of the Nephral ST dialyzer, is grafted with unfractionated heparin during the manufacturing process. In vitro and in vivo data requested for CE marking have shown the stability of heparin grafting. Several clinical studies have been performed with Evodial 22 23 . A 30% heparinization reduction with Evodial dialyzers led to an improvement in oxidative stress, thereby testifying an effective biocompatibility 22 . In a study performed in 45 chronic dialysis patients, Kessler et al. have shown that the systemic heparin dose, regardless of heparin type (unfractionated heparin or low molecular weight heparin), can be reduced by 45% ± 13% without any coagulation issues, which was defined as very early clotting signs such as quality of rinse-back in the circuit (dialyzer and blood lines) 18 .

The present HepZero study hypothesis is that in patients requiring heparin-free dialysis, the heparin-free treatment with Evodial can be easily performed (without saline flushes or blood predilution) and is at least not inferior and may be superior to the standard care heparin-free treatment in terms of clotting, as assessed for the first time by an international, multicenter, randomized, controlled, open-label trial with two parallel groups.

Anticoagulation for chronic dialysis patients with contraindications to heparin administration is challenging. The regimens used as alternatives, for example anticoagulation-free dialysis, regional heparinization with protamine, direct thrombin inhibitors, prostacyclin or RCA, encounter considerable limitations. Current guidelines state that in patients with increased bleeding risks, strategies that can induce systemic anticoagulation should be avoided. Treatment strategies that avoid this include no use of anticoagulants with regular flushing or RCA 3 . To date, the use of RCA is restricted to specialized units, since its application is cumbersome 26 , because it usually requires additional pumps for citrate and calcium infusions, which are not provided by standard dialysis machines, as well as the need for careful monitoring of electrolytes. These issues not only increase the complexity of the RCA procedure but also the likelihood of complications, the most dangerous being systemic hypocalcemia, since it can cause life-threatening arrhythmias 30 . Therefore, heparin-free dialysis using saline infusion is currently de facto considered as the gold standard in patients with high risk of bleeding 1 2 and serves as the control intervention in the present trial. For obvious logistical and technical reasons, the study is conducted as open-label and the primary endpoint could not be assessed blindly. Indeed, neither the predilution process nor the saline flushes can be masked and the dialyzers can easily be differentiated by the nursing staff (different housings, different membrane colors and transparency), since they must be carefully examined during the dialysis treatment and they cannot be completely covered by a label aimed at preventing any differentiation between the two groups. In order to minimize potential bias due to the open-label design, it was decided to have the study primary endpoint: 1. evaluated by the most widespread semi-quantitative clotting scale used in research 13 14 19 20 25 26 ; 2. independently rated by two observers; in case of discordance between the two observers or in case of premature session stoppage (grade 4), the final adjudication must be made by a third authorized and trained person (the principal investigator or registered co-investigators); and 3. training and certification of nurses and investigators with regard to grading.

Several protocols of heparin-free dialysis are currently being used in routine practice, which may include saline flushes delivered at frequent intervals, requiring close monitoring by dialysis staff, or predilution, where a continuous infusion of saline is run to the dialyzer; the former procedure being the most commonly used. Taking into account this heterogeneity and in order to increase the external validity of the study results, it was decided to allow both types of saline infusions, which reflect real-life situations; however, accompanied with guidelines relative to the volume and output of saline infusion, in order to minimize the heterogeneity within the control group. To the best of our knowledge, the HepZero study is the first international, multicenter, randomized, controlled study aimed at comparing two techniques of heparin-free dialysis, including its gold standard, in two parallel groups, in an open-label design. It is sufficiently powered to first evaluate the non-inferiority of Evodial (which may allow a decrease in dialysis staff workload during heparin-free dialysis sessions) then its potential superiority versus the standard care of saline infusions, and may therefore have major clinical implications for patient care.

Trial status at the time of submission the 15^th January 2013: 245 patients recruited.

DSMB: Data and Safety Monitoring Board; PEI: Polyethyleneimine; RCA: Regional citrate anticoagulation; UF: Ultrafiltration.

The authors declare that they have no competing interests relevant to the present manuscript, besides the disclosures presented below.

The authors are HepZero steering committee members as well as the study statistician (RF). They conceived of the study, participated in its design and coordination, and helped to draft the manuscript. All authors read and approved the final manuscript.