Haemodialysis Techniques and Adequacy 1

HAEMODIALYSIS TECHNIQUESAND ADEQUACY 1

SP410 COMBINING RENAL CELLS ANDMICRO- ANDNANOTECHNOLOGIES: A NEW ROUTE TO THEDEVELOPMENTOF BIOARTIFICIAL PLATFORMS FOR INVITROTESTING DRUG NEPHROTOXICITY

Anna Giovanna Sciancalepore1, Fabio Sallustio2,3,4, Salvatore Girardo5,6, LauraGioia Passione1,3,5, Andrea Camposeo1,5, Elisa Mele1, Mirella Di Lorenzo5,7,Vincenzo Costantino2, Francesco Paolo Schena2,4 and Dario Pisignano1,5,31Center for Biomolecular Nanotechnologies @unile, Arnesano (LE), Italy,2University of Bari, Bari, Italy, 3University of Salento, Lecce, Italy, 4ConsorzioCarso, Valenzano (Bari), Italy, 5National Nanotechnology Laboratory of ConsiglioNazionale Delle Ricerche, Arnesano (LE), Italy, 6Technische Universität Dresden,Dresden, Germany, 7Istituto Italiano Di Tecnologia, Genoa, Italy

Introduction and Aims: Different types of cells have been used for the development ofpotential bioartificial kidney devices. Recently, the reprogramming of adult cells toembryonic nephron progenitors (C.E.Hendry et al JASN 2013) and human embryonicstem cells differentiated to renal proximal tubular cells (K.Narayan et al KI 2013) havebeen used as potential building blocks for a bioartificial kidney. Here we propose thecombination of adult renal progenitor/stem cells with different microfabrication andnanofabrication technologies to develop miniaturized, bioartificial proximal tubule-likeplatforms, which are very promising tools for next-generation bio-analytic assays andfor studying the nephrotoxicity of drugs. The potentialities of these interdisciplinary,cross-cutting platforms for in-vitro testing of drugs are presented and discussed.Methods: Our class of devices is composed of overlapped elastomeric layers,embedding microfluidic connections, porous and functionalized membranes, andpolymeric valves, as well as suitable pumps to control all the involved flows, besidesliving cells. All the tested experimental geometries are designed and realized to mimicthe in vivo kidney structures, and specifically renal tubules. Employedmicrotechnologies include optical and soft lithography, and particular care is paid toensure the biocompatibility of all the involved device surfaces. In the devices, livingcells are placed in contact with functionalized membranes to tailor and control thetransport of solutes.Results: The basement membrane of tubules is composed by several extracellularmatrix proteins, we generally find that fibronectin promoted an enhanced proliferationof stem cells compared to other proteins. Various kinds of renal cells were used to testour devices. Then, the culture conditions and initial seeding concentration on-chipwere optimized up to reaching confluence of cells. Working conditions, operatingfluxes as well as fluidic connections were optimized to obtain functional bio-chips withpolarized ARPCs. Solute transport across membranes was studied in detail, and foundto be modulated by the embedded cells up to permeability values of the order of 0.5μm/s.Conclusions: Bioartificial proximal-tubule like device platforms represent aninteresting model for studying the nephrotoxicity of drugs by microfluidic approaches.The combination of cross-cutting technologies derived from complementarydisciplines will certainly constitute a strategic pathway to implement novel bio-assaysof remarkable nephrologic interest in the near future.

SP411 STARTING DIALYSIS ON ONCE-WEEKLY SCHEDULEWITHSOFT, PRE-DILUTION HFOR HDF CAN PRESERVE RESIDUALRENAL FUNCTION AND ALLOW LESS FREQUENTTREATMENT FOR YEARS, IN MANY ESRD PATIENTS

Francesco Gaetano Casino1, Salvatore Domenico Mostacci1, Maria Di Carlo1,Andrea Sabato1 and Clelia Procida11Nephrology Unit, Matera, Italy

Introduction and Aims: Based on empirical observations, we hypothesized that lessfrequent Haemodialysis (HD) could preserve residual renal function (RRF).Accordingly, for decades, we have been trying to start maintenance HD ononce-weekly schedule in almost every patient with significant RRF, soon increasing thefrequency of treatment in the case of deterioration of RRF and Urinary Output (UO)or clinical status. More recently we have decided to start all new patients ononce-weekly predilution HDF or HF. The aim of this study was to evaluate the impactof different initial dialysis schedules and/or modalities on RRF survival.Methods:We retrieved data of all ESRD patients started on HD at our Unit fromJanuary, 2000 to June, 2013, and followed-up for at least 6 months. The patients were

divided into 3 groups (G), based on the initial schedule (the one present 2 months afterthe start) and/or modality of treatment: standard HD, 3 sessions per week (3HD/w,G1); HD once or twice weekly, (1HD/w, G2), and pre-dilution HDF or HF once weekly(1HDF/w, G3). For the sake of simplicity, 500 mL/day was arbitrarily set as the criticallevel of UO, so that, the RRF survival time, to be used with Kaplan-Meyer (KM)analysis, was computed as the time interval between the date of the 1st dialysis and thatof the last measured UO equal or greater than 500 mL/day.Results:We found a total of 150 patients fulfilling the inclusion criteria. As shown intable 1, the main baseline patients' characteristics were similar for the 3 groups. TheKM analysis showed a significant (Log Rank, p<0.001) difference among groups, withan impressive 2-year cumulative RRF survival of 89% for G3, vs 27% for G1, and 63%for G2 patients. The 3-year survival rate for G3 patients remained stable at 89%.Conclusions: Due to the observational nature of the study, the above results onlysuggest that starting HD on a less frequent schedule could allow a long RRF survivaltime, further prolonged by using HDF or HF. Interestingly, similar data showing abeneficial effect of twice-weekly HD on RRF survival have recently been published for alarge group of Chineese patients. Such an effect is usually explained by less hypotensiveepisodes on 2HD/w vs 3HD/w patients. However, it is also possible that less frequentdialysis could be more biocompatible. Moreover, the associated higher BUN levelscould maintain a beneficial osmotic diuresis. A soft HDF or HF, using low volumeflows, ultrapure diialysate and very compatible membranes, would both avoid too lowBUN levels and increase biocompatibility. While waiting for evidence-based new guidelines for dialysis start, we believe that everyone could safely test our approach bystarting a few suitable patients on 1HDF or HF/w, being careful to increase thefrequency on the basis of the observed GFR and UO (or interdialysis weight gain)values as well as clinical status, paying a particular attention to the control ofbiochemistries, ECF volume and blood pressure. We would stress that using theUKM-based incremental approach is not recommended because, due to the theerroneous assumption of equivalency between renal and dialytic clearance, itoverestimates the dialysis needs, so that 1 HD/w would be nearly impossible. Moreover,as hypothesised above, a high efficient treatment could increase the RRF loss rate.

SP412 MIDDLEMOLECULE REMOVAL IN HDF COMPARISON OFMID- VERSUS POST-DILUTION (MIDEMM STUDY)

Caroline Créput1, Raymond Vanholder2, Jean Claude Stolear3, Gaëlle Lefrancois4,Melanie Hanoy5, Joelle Nortier6, Jacky Potier7, Luisa Sereni8 and For The MidemmStudy Group.1AURACentre Hemod Henri Kuntziger, Paris, France, 2University Hospital, Gent,Belgium, 3CHWaPi, Tournai, Belgium, 4ECHO, Nantes, France, 5CHU, Rouen,France, 6CUB Erasme, Bruxelles, Belgium, 7CH Public du Cotentin, Cherbourg,France, 8Bellco srl, Mirandola, Italy

Introduction and Aims: Online hemodiafiltration (HDF) with high-volumesubstitution fluid is an optimal way to remove uremic substances ranging widely inmolecular size from small to low molecular weight (MW). Post-dilution HDF is themost efficient infusion mode to obtain maximum clearance.Mid-dilution infusion is aninteresting alternative that represents simultaneous pre- and post- dilution infusionmodes and could be a highly effective technique to remove uremic toxins, avoiding thedisadvantages of pre- and post-dilutional modes.The aim of this multicentric study wasto compare mid-dilution (MID) with post-dilution HDF (POST) by evaluating theirefficiency in removing different middle MW (MMW) and protein bound uremictoxins.Methods:We performed a cross-over study including 158 Patients from 21 centres. Allpatients were randomized in two different groups for three months: group A (1 monthMID - 1 month POST -1 month MID) and group B (1 month POST- 1 month MID -1month POST). 64 patients were excluded from the final analysis because of incompletedata. The reduction rate (RR) of middle and protein bound molecules were centrallydetermined from serum samples as well as the second generation Daugirdas Kt/Vd.Albumin loss was carried out for both groups.Unpaired t student test was performedusing GraphPad prism version 4.00 for Windows.

SP411 Baseline patients characteristics and RRF survival by group

Ageyears

Weightkg

baselineGFRmL/min

baselineKt/V persession

deltaGFRml/min/y

mean RRFsurvivalmonths

2-y cumsurv%

G1 N=48 67±17 68±17 5.1±2.2 1.6±0.3 9±15 17±2 27±7G2 N=80 69±15 63±12 6.0±2.0 1.5±0.3 3±5 49±5 63±6G3 N=22 67±13 68±16 5.0±1.5 1.3±0.4 1±3 54±3 89±7

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

Nephrology Dialysis Transplantation 29 (Supplement 3): iii209–iii222, 2014doi:10.1093/ndt/gfu153

at Utrecht U

niversity on September 2, 2014

http://ndt.oxfordjournals.org/D

ownloaded from

http://ndt.oxfordjournals.org/

Results: The data of 94 (48 from group A and 46 from group B) patients (53M and41F) were fully analysed. The median age was 70 (27-92) years and dialysis vintage was47,2 (7,5÷454,6) months. No difference was found in the demographic characteristicsand treatment parameters. 164 MID sessions and 161 POST sessions were analysed. Astatistically significant difference in RR (%) was found for three MMWmolecules: β-2Microglobulin (β2M), Complement Factor D (CFD) and Retinol Binding protein(RBP). Values were 80,1±0,4 in POST vs 81,6±0,4 in MID (p=0,01) for β2M; 72,8±0,8in POST vs 76,4±0,6 in MID (p=0,0003) for CFD and 24,1±0,9 in POST vs 30,0±0,8 inMID (p=0,003) for RBPThe other investigated molecules, ADMA, Homocystein,Leptin and Myoglobin, shown a better MID RR but it is not statistically significant.Thereinfused volume was significantly higher in MID than in POST (average total volumeof 43,63 L in MID vs 20,96 L in POST), but also the amount of reinfused volume inMID exchanged in its post dilution stage (estimated around the 2/3 as shown inMaduell publication) is significantly higher (28,8 L in MID vs 20,96 L in POST); thiscould explain the depuration capability of MID respect POST for the MMWmolecules,indeed, was found a linear correlation (R2 0.83) between the delta differences in RR(RR Mid - RR Post) and MWof molecules (Figure 1). No significant differencesbetween MID - and POST-dilution were observed for small MWmolecules depuration(assessed by second generation daugirdas Kt/vd), neither for Albumin loss.Conclusions:MID is superior to remove MMWmolecules as compared to POST. Thisvery likely can be related to an higher total amount and efficiency of substituted volumeobtained in the MID group as compared to the POST group.

SP413 ECODIALYSIS: IS IT POSSIBLE TO DESIGN ANECO-FRIENDLY SYSTEM?

Martina Ferraresi1, Amina Pereno2, Marta Nazha1, Silvia Barbero2 and GiorginaB Piccoli11University of Torino, Turin, Italy, 2Politecnico of Turin, Turin, Italy

Introduction and Aims: Attention to the environmental impact is still limited inmedicine. Chronic Hemodialysis produces about 600000 tons of plastic wastes per year.The economic crisis and the awareness of the ecosystem induced to focus attention onthe lifespan of disposables,“from cradle to grave”. A new outlook is presently focussedon recycle, that is the subsequent start of new cycles leading to a “from cradle to cradle”model: a “new life” for the waste products (Fig 1). Aim of the study is an analysis of thedisposables employed in chronic hemodialysis, for identifying strategies limiting theenvironmental impact and containing the costs.Methods: An analysis of the disposables employed on dialysis and of their "finaldestiny" (the grave) was performed in 3 subsequent bicarbonate dialysis sessions with 3different dialysis machines. All disposables and packagings were photographed,classified, weighted and analyzed as for type of materials, possibility to recycle,contamination with blood or biological fluids.Results: Each dialysis session produces between 4 and 6 kg of wastes; it may be dividedinto about 2 Kg of residual fluids (to be discharged); 2 Kg of "contaminated" wastes (i.e.in contact with blood or fluids) and 2 kg of "non contaminated" wastes. Thedifferentiation is crucial, as the weight of contaminated waste products is the maindeterminant of disposal cost (approximately 2 Euro/kg in Italy). Furthermore, eachdialysis session produces between 0.9 and 1.4 kg of packaging (cardboard and plastic);this is usually discharged separately, but where this procedure is not followed, it addsconsiderably to the volume and weight of the final wastes.Therefore, a undifferentiatedwaste collection may produce over 6 kg of waste products per session; the cost (up to12-14 Euros) corresponds to 20-40% of the cost of the disposables. While all thecardboard and paper wastes are readily recyclable, the plastic wastes (noncontaminated) can theoretically enter a dedicated recycle process. In this regard, thewastes may be classified into "families" of different plastic materials, with differentcompatibility for joint recycling. However, in most of the cases the types of plasticcomponents are not identifiable and separable.Further problems are relatedwith:-Packaging oversize: the content of most of the packaging of dialysis materialsoccupies between 50 and 75% of the space, increasing costs (production, wastes,

transportation).-Emptying: there are no automated systems for emptying residualfluids after the dialysis session.-Difficult separation of materials: many packages arelaminated made of different components.-Difficult separation of contaminatedmaterial: there is no clear definition of "contaminated".Conclusions: Attention to the life cycle of the dialysis disposables may conjugate theattention to our planet, reducing the "mountain" of wastes produced every year; simpletasks, as careful emptying and differentiating between "contaminated" and "noncontaminated" wastes may lead to a 20% saving of the costs of a dialysis session.Cooperation with the Industry is needed for designing recycling strategies in keepingwith the modern "cradle to cradle" approach.

SP414 SURFACE, A PARAMETER TO CONSIDER IN HIGHCONVECTION VOLUME HDF

Alain Ficheux1, Nathalie Gayrard1, Flore Duranton1, Caroline Guzman1,Ilan Szwarc2, Johanna Bismuth -Mondolfo2, Philippe Brunet3, MarieFrançoise Servel2 and Angel Argilés1,21RD – Néphrologie and Université Montpellier 1, EA7288, Montpellier, France,2Néphrologie Dialyse St Guilhem, Centre de Dialyse de Sète, Sète, France,3Service de Néphrologie, Hôpital de La Conception – Université Aix-Marseille,Marseille, France

Introduction and Aims: Convection volume seems to be crucial to the survivalbenefits proposed for HDF. However, high convection requires increasingtransmembrane pressure (TMP) which in turn may change the membrane's behaviourand dialyser's performances.We wanted to characterise the influence of membranesurface area on the physics and on the removal performances of high convectionvolume on-line post-dilutional HDF.Methods: Twelve stable dialysis patients were successively treated with Amembris® 1.8m² and 2.3 m² dialysers, and two high convection flows, one (QUF-optimal) obtainedwhile maintaining the dialysis setting at the maximum in vivo global ultrafiltrationcoefficient (GKD-UF max) and the other one at the maximum convection flow(QUF-max) limited only by the European Best Practice Guidelines (EBPG) (<30%blood flow / 300 mmHg of TMP) for 1 week each. Continuous sampling of spentdialysate was performed in all dialysis sessions and total mass of urea, creatinine, andtotal proteins were measured. SDS-PAGE scanning of the removed proteins and ELISAmeasurements of β-2-microglobulin (B2M), retinol binding protein, lambda lightchains of immunoglobulins, α1-antitrypsin and albumin, were performed.Results: Increasing from QUF-optimal to QUF-max using the 1.8 m² dialyser resultedin frequent TMP alarms and only 33% of the sessions reached the prescribedconvection volume. Increasing the dialyser's surface to 2.3 m² significantly decreasedthe number of alarms and increased the number of sessions reaching the aimedconvection volume (100% at QUF-optimal and 79% at QUF-max). The total amountof urea removed was 545±43, 473±32 and 491±44,471±38 mmol/session in HDF withQUF-optimal and QUF-max respectively for the 1.8 and 2.3 m2 surface (NS). Thecorresponding Kt/V values were 1.77±0.05, 1.78±0.05 and 1.75±0.04, 1.75±0.05, (NS).Removal of low mol wt proteins (observed on SDS-PAGE pattern analysis) andparticularly B2M did not change in the 4 different conditions (274±35, 290±35, 266±24and 283±35 mg/session (NS)). High molecular weight proteins removal increased withconvection, notably for albumin (from 386±57 to 793±158 with 1.8 m² and from 559

SP412

SP413

Abstracts Nephrology Dialysis Transplantation

iii | Abstracts Volume 29 | Supplement 3 | May 2014

at Utrecht U



ownloaded from


±69 to 1052±178 mg/session with 2.3 m² (p<0.001). The highest albumin loss wasobserved with the larger dialyser at QUF-max (figure 2).Conclusions: Increasing membrane surface area diminished the number of alarmsallowing a more frequent accomplishment of the prescribed convection volumes.However, the use of larger dialysers in a QUF-max situation, results in an increasedalbumin loss suggesting that when large dialysers are used a QUF-optimal settingseems more appropriate.

SP415 SAFETY AND PERFORMANCE OF A HOME HAEMODIALYSISDEVICE: RESULTS FROM THE FIRST IN-HUMAN STUDYWITHA NOVEL HOME HD SYSTEM

Angelito Bernardo1, Jason Demers2, Audrey Hutchcraft1, Thomas C Marbury3,Marc Minkus1, Matthew Muller1, Ruth Stallard1 and Bruce Culleton11Baxter Healthcare, Deerfield, IL, 2Deka R & D Corp, Manchester, NH, 3OrlandoClinical Research Center, Orlando, FL

Introduction and Aims: Despite the clinical and humanistic benefits associated withhigh dose haemodialysis (short daily or frequent nocturnal HD), relatively few patientsreceive this therapy. Among the many stated barriers to the growth of high dose HDis the absence of a HD device designed for the patient as the user, a device withintrinsic safety features to mitigate risks associated with HD in the home environment,and a device with features that limit the burden on the patient when performingindependent HD.

Methods:We performed a first in-human, prospective, single arm clinical study with anovel HD device, Vivia (Baxter Healthcare, Deerfield, IL, USA). Unique features of theVivia Haemodialysis System, including pneumatic blood pumps, infusion-qualitydialysate, and extended use of the dialyser (2.1m2, polyethersulfone membrane) andblood set, were tested over 10 weeks (2-week stabilization period and 8-week evaluableperiod) at two clinical sites in the United States. Safety was assessed on all subjects whoused Vivia at least once. Urea clearance was assessed using weekly standard Kt/V,values transformed from second generation estimates for single pool Kt/V. Theassociation between fluid weight removed (as measured by the Vivia HD System) andweight change (determined using weigh scales) was measured for each treatment.Dialysate was sampled at a minimum of weekly for each subject throughout the study.Criteria for success was defined as a bacterial count of 0 CFU/mL and an endotoxinlevel of < 0.03 EU/mL, consistent with AAMI and ISO standards for dialysate forinfusion.Results: Twenty-two subjects (mean age 51 years, 45% F) received 4 treatments perweek for 10 weeks. Adverse events were similar to those expected for prevalent HDpatients. Hypotension was the most common adverse event (2 events per 100treatments). No deaths and no device-related serious adverse events occurred duringthe study. During the evaluable period, the mean (standard deviation) duration of eachtreatment was 3.8 (0.2) hours with a mean blood flow rate and dialysate flow rate of 357(21) ml/min and 395 (9) ml/min, respectively. The mean weekly standard Kt/V ureawas 2.97 (0.29). The association between fluid weight removed as measured by Viviaand weight change is shown in the figure (R2 0.97). The maximum number of usesfrom a single dialyser and blood set was 24. No decrement in urea clearance withmultiple uses of a dialyser was observed. Of 272 dialysate samples obtained on Viviadevices in use during the study, endotoxin levels were less than 0.03 EU/ml and72-hour culture results revealed no bacterial growth for all samples.Conclusions: This first in-human clinical study supports the safety and performance ofa novel HD system.

SP416 PILOT TRIALON IONIC STRENGTH HEMODIAFILTRATION, ANOVEL DIALYSIS TECHNIQUE FOR INCREASED PROTEINBOUND TOXIN REMOVAL

Detlef H Krieter1, Thomas Körner1, Eric Devine2, Marieke Rüth2,Joachim Jankowski3, Christoph Wanner1 and Horst-Dieter Lemke21University Hospital Würzburg, Würzburg, Germany, 2Excorlab GmbH,Obernburg, Germany, 3Charité - Universitätsmedizin, Berlin, Germany

Introduction and Aims: Protein bound uremic toxins (PBT) are difficult to remove byconventional hemodialysis (HD). Aim of the present study was to demonstrate theclinical feasibility of a modified hemodiafiltration technique increasing the ionicstrength in plasma to enhance the PBT removal.Methods: In a prospective, randomized, controlled trial enrolling 8 maintenancedialysis patients (NCT01923961), HD was compared with online predilutionhemodiafiltration (HDF) and predilution HDF using an infusion fluid with increasedNaCl concentration (HDFNaCl). Blood and dialysate flow rate (250 and 575 mL/min,resp.), treatment time (240 min), and high-flux dialyzer (PUREMA® H, 2.1 m²) werealways identical. In both HDF modes, the infusion flow rate was 125 mL/min. InHDFNaCl, the infusate Na+ was adjusted to approach 240 mmol/L in blood beforeentering the dialyzer and dialysate Na+ was set at the technically feasible minimum of130 mmol/L. Removal of free and total para-cresyl sulfate (pCS) and indoxyl sulfate(IS) was determined by plasma clearances (K), reduction ratio (RR), and mass indialysate. Hemocompatibility was assessed by the time courses of white blood cells,platelets, hemolysis (free hemoglobin, LDH), complement C5a, andthrombin-antithrombin III (TAT). Na+ was monitored in both arterial and venousblood.

SP414

SP414

SP415

Nephrology Dialysis Transplantation Abstracts

Volume 29 | Supplement 3 | May 2014 doi:10.1093/ndt/gfu153 | iii

at Utrecht U



ownloaded from


Results: All treatments were well tolerated without any adverse event. Compared toHD and HDF, the RRs of free and total PBT were highest in HDFNaCl (free pCS, 66.5±14.4, 72.0±14.6 vs. 74.0±13.8 %; total pCS, 43.0±9.4, 42.9±14.3 vs. 46.0±12.9 %; freeIS, 60.4±16.5, 68.4±8.2 vs. 72.6±6.1 %, (P=0.026); total IS, 47.8±10.3, 45.2±7.7 vs. 52.0±12.9 %). The same trend was seen for the PBT mass detected in dialysate (total pCS352±234, 415±299 vs. 419±228 μmol; total IS 302±172, 280±145 vs. 350±171 μmol).The values for K were highly variable. No differences between the treatment modeswere observed for the hemocompatibility parameters. In HDFNaCl, Na+ in bloodentering the dialyzer was 232±7 mmol/L. In arterial blood, it slightly increased overtime (0 min, 132±2 to 240 min, 136±3 mmol/L; P<0.001). Compared with HD, at 240min, arterial (133±2 vs.136±3 mmol/L; P=0.032) and venous (136±3 vs. 140±3 mmol/L; P=0.01) Na+ was higher in HDFNaCl.Conclusions: Increased ionic strength HDF may enhance the removal of PBT withoutclinically and biochemically overt adverse effects. As the Na+ balance during dialysis iscrucial, implementing effective HDFNaCl will require adjustment of the minimumdialysate Na+ setting, which was limited to 130 mmol/L with the standard online HDFequipment deployed.

SP417 THE CLOSED LOOP CONTROLOF BLOOD VOLUME (BV) INON-LINE HEMODIAFILTRATION (OL-HDF)

Alessandro Surace1, Paolo Rovatti1, Denis Steckiph2, Elena Mancini3 andAntonio Santoro31Gambro Dasco SpA, Modena, Italy, 2Gambro Hospal SpA, Bologna, Italy,3Policlinico Sant'Orsola-Malpighi, Bologna, Italy

Introduction and Aims: The closed loop control of BV (Hemocontrol) is a recognizedstrategy for reducing intradialysis hypotension. This system, till now, has been onlyused in conventional hemodialysis (HD). On the other hand many recent studiessuggest that Hemodiafiltration (HDF) may have additive advantages in term ofhemodynamic tolerance in comparison to HD. Thus the implementation of HC inpost-dilution OL-HDF (HDF+HC) could combine the benefits of HC with those of thehigh-volume convective treatments. In the present study we were aimed at verifyingperformance and safety of HDF+HC in terms of dialysis efficiency and sodium balance.Methods: The integration of HC system in OL-HDF treatments was tested from May2012 to May 2013 in a prospective randomized cross-over pilot study (SOCRATHE,NCT01582867). Six patients (4 male, 2 female, mean age 73 ± 12 years) were treated on2 different modalities, HD+HC (control) and HDF+HC (intervention). For eachmodality 6 Run-In treatments without HC were executed, followed by 12 treatments

with HC. Each patient acted as his/her own control. Pre and post session plasma Na+

concentrations (Nap, measured by ion selective sodium electrode), inter-dialytic weightgain (IDWG), blood pressures (BPs), dialysis dose (KT/V), relative BV changes (ΔBV%) and reported thirst scores (TS, Likert type scales from 5=never to 25=always thirsty)were collected. Eventual detrimental effects on treatment efficacy, due to the mutualinteraction of HC system with OL-HDF therapy have been assessed.Results: The main results are summarized in Table 1. The infusion volume achieved inHDF+HC was comparable with the OL-HDF one (20.6±2.8 vs 20.2±3.9 l, p=0.845).Conclusions: HDF+HC therapy resulted safe and at least with the same performancesof conventional HD+HC as concerns IDWG, BP values and ΔBV% (Figure 1). The HCsystem in HDF reached the desired final targets (ΔBV%, Total Weight Loss and patientsodium) with the same accuracy showed in HD. Dialysis dose was higher in HDF+HCtreatments.Plasma sodium concentration resulted slightly higher in HDF+HCtreatments (about 1 mmol/l for both initial and final Nap): however, this sodiumdifference is widely within the instrumentation accuracy and can be compensatedmodifying the HC equivalent conductivity (left unchanged during the study in boththe modality). Hence, there are no sufficient evidences to suppose sodium retentioneffect when combining HC and OL-HDF. Further studies will be needed in order to:• Better investigate the genesis of an eventual Na+ increment separating the effects ofHC and OL-HDF on sodium balance.

• Define the correct prescription set-up for HDF+HC treatments basing on patientsclinical conditions.

• Demonstrate if such a system is able to produce relevant clinical benefits in the longterm.

SP418 PRESCRIPTIONS OF DIALYSATE POTASSIUMCONCENTRATIONS DURING SHORT DAILY AND LONGNOCTURNAL (HIGH DOSE) HAEMODIALYSIS

John K Leypoldt1, Baris U Agar2, Angelito Bernardo1 and Bruce F Culleton11Baxter Healthcare Corporation, Deerfield, IL, 2Baxter Healthcare Corporation,Round Lake, IL

Introduction and Aims: Dialysate potassium concentrations are prescribed to provideadequate dialytic removal while maintaining serum potassium levels within normallimits, and such prescriptions during thrice weekly haemodialysis (HD) are basedlargely on clinical experience. The prescription of dialysate potassium concentrationsduring short daily and long nocturnal (high dose) HD are challenging due to limitedclinical experience with such modalities. Our aim was to provide a quantitativeapproach for prescribing dialysate potassium concentrations based on a model ofpotassium kinetics.Methods: Potassium kinetic parameters based on a pseudo one-compartment modelwere first determined during kinetic modeling sessions in 547 patients participating inthe HEMO Study. Those patients had a predialysis serum potassium concentration of5.25±0.95 mEq/L and were treated thrice weekly with blood flow rates of 343±61 mL/min, dialysate flow rates of 688±129 mL/min, dialyser potassium dialysances of 179±27mL/min, and treatment times of 206±29 min. Patients were then categorized based onthe prescribed dialysate potassium concentration as 1K (1.02±0.05 mEq/L, N=60), 2K(2.01±0.05 mEq/L, N=437) or 3K (3.01±0.07 mEq/L, N=50) with measuredintradialytic decreases in serum potassium concentration (ΔK) of 1.89±0.71 mEq/L for1K, 1.58±0.83 mEq/L for 2K, and 1.02±0.55 mEq/L for 3K patients. Dialysatepotassium concentrations were then prescribed for each patient during short daily andlong nocturnal HD based on the kinetic model to maintain the identical predialysisserum potassium level and dialytic removal of potassium as during thrice weekly HD.Short daily HD was assumed to be performed 6 times per week with treatment timesone-half those during thrice weekly HD (103±15 min), all other conditions identical.Long nocturnal HD was assumed to be performed 5 times per week with a treatmenttime of 420 min and reduced blood and dialysate flow rates, achieving a dialyserpotassium dialysance of 148±5 mL/min.Results: Prescribed dialysate potassium concentrations (Dialysate K in mEq/L) and ΔKin mEq/L for high dose HD are tabulated: These results suggest that dialysatepotassium concentrations during short daily HD should be 0.2-0.5 mEq/L higher thanduring thrice weekly HD and approximately equal to 4 mEq/L during long nocturnalHD. Under such conditions, the intradialytic decrease in serum potassiumconcentration will be reduced by more than one-half during short daily and byapproximately three-quarters during long nocturnal HD of that during thriceweekly HD.

SP417 Table 1

IDWG [Kg] Initial Nap[mmol/L] Final Nap [mmol/L] Pre- dialysis BP [mmHg] Post-dialysis BP [mmHg] KT/V Thirst Score

HD+HC 3,02 ± 0,47 134,8 ± 4,2 135,4 ± 2,2 Systolic (S): 130 ± 23Diastolic (D): 67 ± 18

S: 121± 13D: 67 ± 17

1.29 ± 0.20 7,3 ± 2,5

HDF+HC 3,07 ± 0,56 135,8 ± 3,9 136,5 ± 2,1 S: 128D: 67 ± 17

S: 122D: 65 ± 13

1.37 ± 0.25 9,0 ± 2,5

p-value (Paired t) p=0,503 p=0,114 p=0,028 S: p=0,580D: p=0,768

S: p=0,936D: p=0,489

p=0.020 p=0,049

SP417



at Utrecht U



ownloaded from


Conclusions:We conclude that dialysate potassium concentrations during high doseHD modalities can be quantitatively predicted using a potassium kinetic model. Highdose HD modalities may improve clinical outcomes by reducing intradialytic decreasesin serum potassium concentration.

SP419 EFFECTOF THE DIALYSATE BICARBONATECONCENTRATION ON CALCIUMMASS BALANCE USINGCONVENTIONAL HAEMODIALYSIS

Svetlana Vankova1 and Jan Havlin11BBraun Avitum, Prague 4, Czech Republic

Introduction and Aims: The calcium mass balance (CMB) in haemodialysis (HD)patients is related to extraosseal calcifications. The diffusion gradient between ionizedcalcium in inflow of dialysate and serum is considered to be the main driving force ofintradialytic CMB. However, the influence of pH and intradialytic alkalinisation,important factors affecting the level of ionized calcium, has not been studied withrespect to intradialytic calcium flux. The aim of this study is to evaluate the effect ofvarious dialysate bicarbonate and calcium concentrations on CMB.Methods:We measured CMB in 9 stable patients on HD with a mean age of 74.4 years.All patients underwent 4 HD treatments using a different dialysate concentration ofcalcium (DCa) 1.25 vs. 1.5 mmol/L and bicarbonate (DHCO3) 26 vs 32 mmol/L. Bloodwas sampled before and after the HD for total (tCa) and ionized (iCa) serum calcium,pH and bicarbonate. The CMB (calcium mass from the dialysate to the patient, withouteffect of ultrafiltration) was determined from the tCa levels in the dialyzer inlet andfrom the continuous partial waste dialysate collection (150ml/h) samples. Wecalculated the difference between post and pre-dialysis values (ΔtCa, ΔiCa and ΔHCO3) and the differences between inflow dialysate and pre-dialysis serum concentrations(diffusion gradients of total calcium (Diff_tCa), ionized calcium (Diff_iCa) andbicarbonate (Diff_HC03)). Other parameters were set as follows: session duration - 4hours, the filter - low-permeability polysulfone 1,5m2, the blood flow - 350 ml/min, thedialysate flow - 600 ml/min, the mean URR - 70.8% (± 5.7).Results: The mean diffusive CMB was positive with 1.5_DCa and it increasednon-significantly with higher DHCO3 (57.7 ± 262 vs 153.9 ± 232.6 mg), with1.25_DCa/32_HCO3 it was almost neutral (-12.8 ± 165.2 mg) and it was negative with1.25_DCa/26_HCO3 (-147.5 ± 191.7 mg). According to repeated measures ANOVAthere was not a significant effect of DHCO3 on CMB (p = 0.23) and there was asignificant effect of DCa on CMB (p = 0.01). Serum tCa levels increased significantlywhen 1.5_DCa was used, while they remained stable when 1.25_DCa was used. SerumiCa levels increased significantly for all combinations of DCa/DHCO3, except for thecombination 1.25_DCa/32_DHC03 (Table 1). We found no significant correlationbetween CMB and ΔtCa or ΔiCa, but there was a significant correlation between CMBand ΔHCO3 (p = 0,04, r = 0,344). Both, Diff_Ca and Diff_iCa, significantly correlatedwith CMB (p = 0,01, r = 0,415, resp. p = 0,026, r = 0,372), however we found thestrongest correlation for Diff_HCO3 (p = 0,004, r = 0,469).Conclusions: The diffusive CMB had a positive correlation with the bicarbonategradient between the inflow of the dialysate and the predialysis blood concentration, aswell as with the tCa and iCa diffusion gradient. Furthermore, we observed a significantpositive correlation between CMB and serum HCO3 change during HD, but notbetween intradialytic changes of tCa or iCa. Therefore, the alkalinisation effect on ahigher intradialytic calcium gain should always be considered when selecting dialysateconcentration of bicarbonate. CMB was not found to depend on the intradialytic tCaor iCa changes.

SP420 SELECTIVE IONMEASUREMENT IN SPENT DIALYSATEWITHLASER INDUCED BREAKDOWN SPECTROSCOPY

D J Klomp1, F Van Beijnum2, J P.R. Day2, F P Wieringa1 and J P Kooman31TNO, Eindhoven, Netherlands, 2TNO, Delft, Netherlands, 3Maastricht UniversityMedical Center, Maastricht, Netherlands

Introduction and Aims: Correct electrolyte balancing is imperative for optimallong-term haemodialysis process. Numerous complications can occur due toelectrolyte derangements. To avoid or minimize the effects of this unbalance inpatients, it is necessary to be able to monitor and adjust these individual ion balances,furthermore in-line monitoring would also facilitate patient specific treatment. Besidesnon-ion selective overall conductivity measurements in dialysate, no stable, in-lineion-selective measurement tool is currently commercially available. For in-linemeasurement of K+, Na+ and Ca2+, a LIBS system (Laser Induced BreakdownSpectroscopy) was developed to enable the measurement of these electrolytes in thespent dialysate flow. LIBS is an optical modality that resembles flame photometry.Instead of a flame, a short tightly focused laser pulse excites the atoms by inducing amicro-plasma in a sample. The characteristic atomic emission spectrum is measuredmaking the components clearly distinguishable. The optical nature of LIBS enablesnon-contact “through a window” sensing, thus offering inherent electrical andmicrobiological safety. The method also has good perspectives for miniaturization andcontinuous real time monitoring.Methods: A LIBS setup was developed capable of creating a plasma discharge with adiameter of 200 micrometer in flowing dialysate (500 ml/min) at a repetition rate of 20Hz by the application of 20 mJ laser pulses (1W optical power injection). The atomicemission spectra emitted by the formed plasma was measured with an echellespectrometer combined with a gated iCCD camera. Each measurement point consistedof 400 pulses (8 s sample rate with running average). The setup was calibrated with aconcentration series of individual electrolytes in water.Results: Accurate atomic spectra were measured with sharp and well-separated peaksfor K+ and Ca2+ and broad self-reversed peaks for Na+ at the expected wavelengths(see figure). The significant parameters from each concentration series had a linearresponse with respect to the concentrations of the electrolytes. Measurements on twodifferent clean dialysates had good correlations with the calibration lines.Conclusions: The LIBS system was shown to be capable of in-line measurement ofindividual K+, Na+ and Ca2+ concentrations. For future work, a mobile setup has beenconstructed to facilitate measurements in the haemodialysis clinic. Thesemeasurements are planned in 2014. Knowledge of the exact concentrations ofelectrolytes in spent dialysate opens perspectives for real-time, patient-specifictreatment. Further development towards a miniaturized version is being pursued forincorporation in a portable artificial kidney using sorbent-based dialysatereconditioning.

SP421 DIALYSATE REGENERATION BY ELECTRO-OXIDATIONCOMBINEDWITH ACTIVATED CARBON DOES NOT INCREASEOXIDATIVE STRESS OR ENDOTHELIAL CYTOTOXICITY

H Gremmels1, D H Hazenbrink1, F Simonis2, M LOtten1, M Wester1, W H Boer1,J A Joles1 and K G Gerritsen11UMC Utrecht, Utrecht, Netherlands, 2Nanodialysis BV, Oirschot, Netherlands

Introduction and Aims:Major challenge in the design of a wearable dialysis device isthe removal of urea, since the daily urea production is very high and removal byadsorption has proved difficult. Electro-oxidation (EO) seems attractive sinceelectrodes are durable, small and inexpensive. We achieved clinically relevant ureadegradation in blood (9.5±1.0mmol/h) using an EO unit in series with activated carbon(AC) for removal of chlorine by-products [submitted]. Major concern of EO is thattoxic, in particular oxidative, by-products are generated that are not removed by AC.Therefore we investigated the effect of EO in combination with AC on oxidative statusand endothelial cytotoxicity of dialysate and dialyzed uremic plasma.Methods: The EO-unit contains 10 graphite electrodes (70g, 585cm2). Uremic plasma(1L, urea 24-43mM), obtained from plasmapheresis of patients with kidney failure, wasdialyzed for 1h and dialysate was recirculated in counter current direction through theEO-unit in series with a degasser (containing 15g of AC) and an AC filter (50g). Threeseries of experiments were performed, in which recirculation with and without EOwere compared. Oxygen radical adsorbance capacity (ORAC) of EO-treated fluids wasassessed using fluorescein as probe, 2,2-azobis-2-methyl-propanimidamide,dihydrochloride as free radical generator and Trolox for calibration. Endothelial colonyforming cells isolated from umbilical cord blood from 3 donors were exposed to EOand AC-treated fluids. Induction of intracellular reactive oxygen species (ROS) wasevaluated using chloromethyl-dichlorodihydrofluorescein diacetate (CM-H2DCFDA).PrestoBlue, a resazurin-based reagent, and neutral red were used to assess cell viability.CM-H2DCFDA and viability data were normalized to PBS-treated controls. Electriccell-substrate impedance sensing (xCELLigence) was used to monitor real-time cellproliferation.Results: Urea degradation by EO was 7.2±1.2mmol/h. Treatment by EO plus AC hadno influence on the total anti-oxidant capacity of dialysate and uremic plasma ascompared to treatment by AC alone (Fig. 1A). No effects of EO and/or AC treateddialysate and plasma were observed on intracellular generation of ROS, cell viabilityand cell proliferation (Fig. 1B-E).

SP418

Short Daily HD Long Nocturnal HD

Patient Category Dialysate K ΔK Dialysate K ΔK1K 1.46±0.23 0.81±0.47 3.72±0.56 0.54±0.222K 2.37±0.24 0.75±0.46 4.10±0.68 0.44±0.233K 3.19±0.17 0.49±0.28 4.26±0.48 0.25±0.15

SP419 Table 1

N 91.5_DCa/32_DHCO3

1.5_DCa/26_DHCO3

1.25_DCa/32_DHCO3

1.25_DCa/26_HCO3

CMB (mg) 153.9 (± 232.6) 57.7 (± 262.9) -12.8 (± 165.2 -147.5 (± 191.7)ΔtCa (mmol/

L)0.350 (± 0.097) 0.352 (± 0.094) 0.023 (± 0.140) 0.077 (± 0.104)

ΔiCa (mmol/L)

0.148 (± 0.045) 0.178 (± 0.028) -0.005 (± 0.072) 0.038 (± 0.038)

ΔHCO3(mmol/L)

5.67 (± 1.79) 1.98 (± 1.66) 6.01 (± 1.91) 1.78 (± 1.32)



at Utrecht U



ownloaded from


Conclusions: Biocompatibility studies suggest that dialysate regeneration by EOcombined with AC does not increase oxidative status or endothelial cytotoxicity ofdialysate and dialyzed uremic plasma as compared to AC without EO. Extensivebiocompatibility studies and analysis of oxidation products are now warranted.

SP422 SPECTRALMEASUREMENTOF POLYVINYLPYROLIDONEELUTED FROM POLYSULFONEMEMBRANES

Koichi Umimoto1, Yoshimasa Shimamoto1, Kohei Mastushima1 andMasahiro Miyata11Osaka Electro-Communication University, Shijonawate, Japan

Introduction and Aims: Polysulfone (PS) dialysis membranes are made hydrophilic byblending polyvinylpyrolidone (PVP) and are well known to have excellentbiocompatibility in clinical use. However, PS membranes have some side effects, suchas anaphylaxis and skin lesions, which are supposedly caused by PVP. In recent years,it has been reported that PS dialysis membranes elute high levels of PVP after beingstored for long periods.A colorimetric assay called the Muller method is generally usedfor measuring PVP eluted from dialysis membranes. However, ultraviolet (UV)spectroscopy offers an alternative approach for clinical analysis. UV spectroscopy doesnot use reagents, thereby permitting real-time analysis. In this study, we investigateddirect measurement of PVP by an optical method and assessed the relation between thestorage period of PS membranes and the quantity of PVP eluted.Methods: The concentration of PVP·K-90 (Wako chemicals, Ltd.) was determined byspectral analysis from the absorption measured at 203 nm based on its spectral peak.The actual concentration of PVP·K-90 was also determined by the Muller method.Then 1.0 L of physiological saline was passed through the blood side of a PS dialysismembrane (PS1.6UW Fresenius) and the concentration of PVP was measured by theoptical method and the Muller method. Subsequently, PS dialysis membranes(RENAK-PS1.6 Kawasumi) stored for 6 months to 34 months were examined and PVPeluted from the membranes was measured.Results: The predicted concentrations of PVP·K90 closely matched the actualconcentrations and there was a significant correlation between the actual and predictedconcentrations of PVP·K90 (r=0.993, P<0.001). The spectrum of PVP eluted from thePS membrane was similar to that of PVP·K-90. There was a significant correlationbetween the predicted and actual concentrations of eluted PVP (r=0.965, P<0.01). Theamount of PVP eluted from dialysis membranes stored for long periods ranged from28.6 mg to 91.8 mg. There was a strong positive correlation between the storage periodand the amount of PVP eluted (n=6, rs=0.886).Conclusions: Although the precision of measurement decreases slightly at lowconcentrations, it is possible to determine the concentration of PVP eluted from a PSdialysis membrane based on the spectral data obtained by using a UVspectrophotometer. This indicates that optical measurement can be employed as amethod for real-time monitoring of PVP eluted during priming of the dialyzer. Inaddition, the quantity of PVP eluted from PS membranes increases in proportion tothe storage period.

SP423 CHARACTERIZATION OF PLATELETACTIVATION ANDCOAGULATION INDICES IN AN IN VITRO DIALYSIS MODEL

Matthew Muller1, Avinash Naik1, Sharon Pokropinski1, Shawn Bairstow1,Jessica Svatek1, Susan Young1, Richard Johnson1 and Angelito Bernardo21Baxter Healthcare Corporation, Round Lake, IL, 2Baxter Healthcare Corporation,Deerfield, IL

Introduction and Aims: The Vivia Haemodialysis System differs from traditionalhaemodialysis equipment in that it uses a pneumatically controlled diaphragm bloodpump and a rigid cassette in the blood flow path. During haemodialysis, blood ischronically exposed to the extracorporeal circuit (ECC) consisting of the blood flowpath and the dialyser. Blood components interact with the ECC potentially elicitingchanges in platelets and coagulation factors. Platelet activation has been identified as arisk factor for cardiovascular events. Therefore, it is of clinical relevance to characterizethe biocompatibility characteristics of the Vivia ECC.Methods: In vitro studies using the Vivia ECC were performed to analyzebiocompatibility during simulated dialysis in a perfusion model. Treatments wereperformed using freshly donated heparin-anti-coagulated human blood at a blood flowof 400 ml/min for 2 hours. Blood incubated in a polypropylene tube on a rocker at 37°C served as a control. Blood samples were removed, mixed with NaCitrate or EDTA at15, 30, 60, 90 and 120 minutes and processed for analysis of platelet activation (CD62)and counts, thrombin activation (prothrombin fragment F1+2) and heparinconcentration.Results: Platelet CD62 expression increased from an initial 1.4% to a modest 4.8% attwo hours, a level comparable to control values, which increased to 5.7%. Soluble CD62(sCD62) levels (mean ± standard deviation) also increased during the runs (from 26 ±8 to 52 ± 23 ng/ml) but did not differ from control values (26 ± 8 to 48 ± 7 ng/ml)either. As expected for an in vitromodel test system with a large surface to volume ratio(Frank et al Kidney International (2001) 60:1972), there was a decrease in plateletcount during the perfusion runs. Thrombin activation (F1+2 levels) was minimal forthe first 45 minutes of perfusion, but then increased over control values at subsequenttime points. The magnitude of the increase in F1+2 was consistent with the literaturewhen corrected for the higher surface to volume ratio of this model system. Theelevation in F1+2 values was also heparin-dependent since increasing heparin levelsresulted in a 73% and 82% reduction in F1+2 values at 2 and 3 U/ml, respectively.Contact activation was investigated as a possible mechanism for thrombin activation,but no evidence for FXIIa production was found in the plasma samples.Conclusions: Platelet activation remained low and stable throughout the treatment asassessed by CD62 expression. Thrombin activation was low initially, but increased afteran hour of perfusion to levels comparable to other published studies, an effect thatcould be largely reversed by increasing heparin levels. Therefore, the Vivia ECCshowed excellent biocompatibility with respect to platelet activation and coagulationindices.

SP424 COMPARISON OF UREA KINETIC MODELSWITH ON-LINECLEARANCEMONITOR. WHAT IS THE BEST TARGET VALUE?

Csaba Rikker1, Edina Juhász1, Renáta Gáspár1 and László Rosivall21Péterfy Hospital, Budapest, Hungary, 2Institute of Pathophysiology, HungarianAcademy of Sciences, Semmelweis University, Budapest, Hungary

Introduction and Aims: Since the secondary analysis of National Cooperative DialysisStudy data (1985) Kt/V has been used for the assessment of the adequacy of dialysis.For three treatments per week, the 2006 Kidney Disease Outcomes Quality Initiative(KDOQI) adequacy guidelines recommend a minimum single-pool Kt/V (spKt/V) of1.2. The European Best Practice Guidelines (EBPG) 2002 recommend a minimumequilibrated Kt/V (eKt/V) value of 1.2, corresponding to a minimum spKt/Vof 1.40.Besides several calculations of Kt/V using pre- and postdialysis urea, in the last decadesionic dialysance (ID) appeared as a method of the on-line monitoring of delivereddialysis in every treatment without blood sampling. The aim of our study is to compareID to three other calculations of Kt/V and determine the minimal value equivalentwith eKt/V prescribed by EBPG.Methods:We measured in one online haemodiafiltration of 127 not selected patientsthe adequacy of treatment by spKt/V using Fresenius 5008 online clearance monitor(OCM). Simultaneously we determined spKt/V using the Lowrie and adjustedDaugirdas equation. The eKt/V was calculated from the latter equation. In accordancewith EBPG we considered the value eKt/V ≥1.2 as reference. We evaluated thecorrelation of OCM with the above mentioned mathematical methods of urea kineticmodel, dry weight, ultrafiltration (UF), ratio of true- and false positive cases below thereference value as well as the sensitivity, specificity, positive and negative predictivevalues at OCM spKt/V target 1.4, 1.3 and 1.2.Results: Lowrie spKt/V; OCM spKt/V ; Daugirdas spKt/V and Daugirdas eKt/V (mean±SD) were: 1.36±0.18, 1.52±0.22, 1.58±0.21 and 1,38±0.18, below the target value(spKt/V <1.4, eKt/V <1.2) were: 54.33%, 26.77%, 14.96% and 11.81% of the casesrespectively. Correlation of OCM spKt/V with Lowrie spKt/V, Daugirdas spKt/V andDaugirdas eKt/V are: 0.61, 0.61 and 0.60. OCM spKt/V showed slightly butsignificantly lower values compared to Daugirdas spKt/V (p= 0.0004). We could notfind any statistical correlation between OCM spKt/V and UF, dry weight, as well as UFexpressed by percent of dry weight. At target value of OCM spKt/V 1.4, 1.3 and 1.2below the reference value (eKt/V <1,2) expressed in % we have found true positivecases: 9.4, 8.7 and 5.5; false positive cases: 17.3, 6.3 and 1.6, true negative cases: 70.1,81.1 and 86.6 and false negative cases: 3.1, 3.9 and 6.3, sensitivity: 0.75, 0.69 and 0.47,specificity: 0.80, 0.93 and 0.98, positive predictive value: 0.35, 0.58 and 0.78, negativepredictive value: 0.96, 0.95 and 0.93 respectively.Conclusions: OCM is an excellent tool for following the adequacy of dialysis in everytreatment. In accordance with several studies it slightly but significantly underestimatesthe value of Kt/V. Decreasing cut-off of OCM spKt/V from 1.4 to 1.3 the number offalse positive cases diminishes by 11%, while the number of false negative cases risesonly by 0.8 %. We suggest OCM spKt/V 1.4 as target value, nevertheless ≥1.3 can be

SP421



at Utrecht U



ownloaded from


enough for adequate dialysis. Further measurements are planned to verify the findingsin other settings as well.

SP425 ALBUMIN DIALYSIS USINGMOLECULAR ADSORBENTSRECIRCULATING SYSTEM IN SEVERE LIVER FAILURE:PREDICTIVE FACTORS FORCLINICALOUTCOME - A SINGLECENTER EXPERIENCE

Elena Rusu1, Diana Zilisteanu1, Sonia Balanica2, Camelia Achim1, Teodora Atasie2,Flavia Carstea2 and Mihai Voiculescu11“Carol Davila” University of Medicine and Pharmacy Bucharest, Bucharest,Romania, 2Fundeni Clinical Institute, Bucharest, Romania

Introduction and Aims: The Molecular Adsorbents Recirculating System (MARS) isan artificial liver support system that removes albumin-bound and water-soluble toxinsthat accumulate in liver failure, providing better conditions for liver recovery.Weanalyzed the prognostic factors for clinical outcome of patients with AoCLF treatedwith MARS in the Clinic of Internal Medicine and Nephrology in order to improve theMARS procedure indication and the selection of patients for therapy.Methods: Between January 2001 and August 2013 we treated 48 liver failure patients,to whom we performed 88 MARS sessions. The etiology of severe liver failure was:acute liver failure (n=11), acute on chronic liver failure (n=25), post livertransplantation graft failure (n=8), and post-hepatectomy liver failure (n=2). The meanage of the patients was 41.6 ± 18.2 years (interval 3-66 years).Results: Before starting MARS therapy 8 patients had sepsis, 19 patients presentedrenal impairment (9 patients with hepatorenal syndrome), 26 patients presentedhepatic encephalopathy grade II or higher.MARS therapy was safe and well tolerated.We noticed favorable clinical effects: improvement in general condition, neurologicalstatus, regression of jaundice and pruritus, improvement in renal function andhemodynamic status. We obtained statistically significant improvement of serum totaland direct bilirubin, serum creatinine, urea lactate. We also obtain a significantdecrease of TNF-alpha after MARS. Regarding the outcome of the patients, in ALFgroup, 6 patients (54.5%) completely recovered the hepatic function. In the AoCLFgroup 3 patients were bridged with liver transplantation and 7 patients recovered theirpre-decompensation status. The mean survival of the AoCLF patients (n=15) on theliver transplantation waiting list was 23.0 ± 6.9 days. The Kaplan-Meier and survivalmodeler analisys showed that the unfavorable prognosis factors for survival were: thepresence of sepsis and multiple organ failure, hepatic encephalopathy grade 2 orhigher, and renal dysfunction. We hadn’t found a relationship between age, etiology ofliver failure or the bilirubin level and the patient’survival.Conclusions: The most promising results were observed among ALF patients. For acuteon chronic liver failure patients MARS therapy could be an efficient solution forprolonging patient survival and a bridging method until liver transplantation will beachieved. We concluded that the predictive factors for patient survival are: sepsis,multiple organ failure, hepatic encephalopathy grade 2 or higher, and renal dysfunction.

SP426 STUDY COSTS ANTICOAGULATION SYSTEMIC CHANGE INHEMODIALISIS UNIT

Tania Monzon Vazquez1, Sandra Saiz Garcia2, Vijay Mathani2 andBeatriz Escamilla Cabrera21Hospital Quiron Tenerife, Santa Cruz de Tenerife, Spain, 2Hospital QuironTenerife, Santa Cruz De Tenerife, Spain

Introduction and Aims: Currently in Spain anticoagulation in haemodialysis situatedabout 50% use of each unfractionated heparin (UFH) and low molecular weightheparin (LMWH). A recent meta-analysis found no differences in efficacy between thetwo kinds of heparins. Use of LMWH has the advantage of the safe handling and thebiggest drawback was its high cost.The aims of this study is analyze the economicbenefits of systematic change in anticoagulation in our haemodialysis unit, from UFHto LMWH.Methods: Costs 6 months before and after the change in anticoagulation on August2012, with an average over the whole period of 63 patients were analyzed. Then theeconomic cost of the products is detailed.The use of sodium heparin, implies a costoverrun from the use of 10cc and 20cc syringe, 2 needles 18G and two sodium heparinvials per haemodialysis session.Results: In the first study period, 41.3% of patients were treated with LMWH(enoxaparin 20 mg were 12.7% and enoxaparin 40 mg were 28.6%) and 55.6% withUFH. In the second period, all patients were treated with LMWH with following

distribution: 44.5% enoxaparin 20 mg and 55.6% enoxaparin 40mg.We analized thecosts for each product by an average of 63 patients, with 3 sessions per week during 6months. Total cost of haemodialysis circuit anticoagulation was in the first period was6344,64€ in the first period (UFH 5115,6€ + enoxaparin 1229,04€) vs. 2870,64€ in thesecond period (enoxaparin 20 mg: 1315,44€ + enoxaparin 40 mg: 1555,2€).This changesaved € 3474 by semester.Conclusions:Main disadvantage of using LMWH was at its high cost, however in lastyears has increased the market price of the UFH. According to our study, LMWH useis less expensive than using UFH, without any evidence of increased thrombotic orhemorrhagic processes.

SP427 ACUTE HEMODYNAMIC RESPONSE AND UREMIC TOXINREMOVAL IN CONVENTIONAL AND EXTENDEDHEMODIALYSIS AND HEMODIAFILTRATION: A RANDOMIZEDCROSSOVER STUDY

Tom Cornelis1, Frank M Van Der Sande1, Sunny Eloot2, Eline Cardinaels1,Otto Bekers1, Jan Damoiseaux1, Karel M Leunissen1 and Jeroen Kooman11Maastricht University Medical Centre, Maastricht, The Netherlands, 2GhentUniversity Hospital, Ghent, Belgium

Introduction and Aims: Intensive hemodialysis (HD) has significant benefits. Aim ofthis study was to evaluate the acute effects of extended HD and hemodiafiltration(HDF) on hemodynamic response and solute removal.Methods: Thirteen conventional HD patients randomly completed a single study of4-hour HD (HD4), 4-hour HDF (HDF4), 8-hour HD (HD8) and 8-hour HDF(HDF8). Blood pressure (BP) and heart rate (HR), pulse wave analysis, cardiac output(CO), microvascular density by sublingual capillaroscopy, as well as relative bloodvolume (RBV) and thermal variables were measured. Clearance and removal of uremictoxins were also studied.Results: Long treatments showed more stability of peripheral systolic BP (mmHg)(drop during HD4 -21.7±15.6; HDF4 -23.3±20.8; HD8 -6.7±15.2, p=0.037 versus HD4and p=0.077 versus HDF4; HDF8 -0.5±14.4; p=0.004 versus HD4 and p=0.008 versusHDF4). A similar observation was found for peripheral diastolic BP and central BP. CO(L/min) remained more stable in extended sessions (drop in HD4 -1.4±1.5; HDF4 -1.6±1.0; HD8 -0.4±0.9, p=0.022 versus HDF4; HDF8 -0.5±0.8; p=0.062 versus HD4 andp=0.025 versus HDF4), in line with decreased RBV slope in long dialysis. Nodifferences in microvascular density were found. Energy transfer rates (W) werecomparable (HD4 13.3±4.7, HDF4 16.2±5.6, HD8 14.2±6.0, HDF8 14.5±4.3). Smallmolecule and phosphate removal were superior during long treatments. Beta2microglobulin and FGF23 reduction ratios were highest in HDF8.Conclusions: Treatment time, and not modality was the determinant for thehemodynamic response. HDF significantly improved removal of middle molecules,with superior results in extended HDF.

SP428 PREDIALYSIS SODIUM LEVELS AND CARDIOVASCULARSTABILITY IN HEMODIALYSIS

Eduardo Baamonde Laborda1, Elvira Bosch Benitez-Parodi1, German PerezSuarez1, Gloria Anton Perez1, Fatima Batista Garcia1, Mar Lago Alonso2 andCesar Garcia Canton21Avericum, Telde, Spain, 2Hospital Insular Gran Canaria, Las Palmas, Spain

Introduction and Aims: A relationship between pre-hemodialysis plasma sodium andmortality has been described. Lower sodium levels are associated with higher mortality.One of the hypothesized reasons for this finding is based on intra-hemodialysiscardiovascular stability.AIMS:to analyze the relationship between pre-dialysis plasmasodium levels and intra-hemodialysis hemodynamic stability in a group of patients onstandard hemodialysis and constant sodium bath.Methods: Retrospective analysis of 231 prevalent patients in HD (59.7% male; averageage 61.36 ± 13.8 years; time in HD 33.41 ± 17.7 months; 54.1% diabetic). Patients wereclassified into tertile groups according to their pre-HD sodium levels (average of 12determinations): Group 0 (G0) with pre-HD Na &lt 137,58 mEq/L (n=77); Group 1(G1) with pre-HD Na from 137,59 to139.15 mEq/L (n=77); Group 2 (G2) with pre-HDNa &gt 139,16 mEq/L (n=77). Plasma sodium was corrected for glucose according to:Corrected-Na (cNa) = Na(mEq/L)+0.016*(serum glucose (mg/dl)-100). Patients werefollowed up for two years or until they left the program (median number of sessions269). Maximum and minimum mean blood pressure (MBP) was measured in every

SP426 MATERIAL COSTANTICOAGULATION

KIND OF HEPARIN COST (€/UNIT)ENOXAPARIN 20 MG 0,63 €ENOXAPARIN 40 MG 0,60 €SODIUM HEPARIN 1% 0,95 €NEEDLE 18G 0,04 €SYRINGE 5CC 0,02 €SYRINGE 20CC 0,04 €

SP426 TOTAL COST STUDY PERIOD

SEMESTER COSTPERIOD 1

SEMESTER COSTPERIOD 2

ENOXAPARIN 20 MG 408,24 € 1315,44 €ENOXAPARIN 40 MG 820,8 € 1555,2 €SODIUM HEPARIN (MATERIAL

INCLUDED)5115,6 € 0 €

TOTAL COST 6344,64 € 2870,64 €



at Utrecht U



ownloaded from


session, as well as inter-dialysis weigth gain, ultrafiltration rate (UF), duration ofsession, dry weight and body mass index (BMI).A session was considered to be asession with hypotension when the difference between maximum and minimum MBPwas higher than 30% and the value of MBP fell below 70 mmHg.Parameters oflaboratory tests (hemoglobin, creatinine, albumin, calcium, phosphorus, ferritin andPTHi), HD adequacy: (Kt/V and URR) and antihypertensive treatment were analyzed.Results: HD sodium level was negatively correlated with UF rate (r: 0.325; p:0.000),percent weight gain (r: 0.279; p: 0.000) and percent of sessions with hypotension: (r:0.146; p: 0.027)Conclusions: Pre-dialysis plasma sodium was inversely related with cardiovascularstability. Patients with lower levels showed higher tendency to hypotension. Patientswith lower sodium levels showed higher percent inter-dialysis weight gain and higherUF rates, which could account for their higher tendency to intra-HD hypotension.

SP429 AN EVALUATION OF BIOELECTRICAL IMPEDANCESPECTROSCOPY, IN ORDER TOMEASURE AND COMPAREBODYWATER DISTRIBUTION IN BOTH HEALTHY PREGNANTWOMEN AND PREGNANTWOMEN ON DIALYSIS

Seiji Hashimoto1, Masahide Seki1, Maoka Tomochika1, Rie Yamamoto1,Nobuhiko Okamoto1, Akira Nishikawa1 and Takao Koike11NTT East Japan Sapporo Hospital, Sapporo, Japan

Introduction and Aims: Pregnant women on dialysis face great difficulty carrying ababy to term. Some of those challenges include but are not limited to miscarriages andstillbirths. One of the major factors complicating prenatal care is the difficulty for themto properly manage body fluids. In this study we examine and compared fluid volumeusing Bioelectrical Impedance Spectroscopy (BIA) in healthy pregnant women and twoof our successful pregnancies of women on dialysis.Methods: In our study we subjected 52 healthy pregnant women to test at variousstages of their respected pregnancies. We did this in order to get enough results and orinformation, so we could compare our results with our two successful dialysispregnancies. In both respective groups we measured fluid volume using BIA. In thecases of our dialysis patients we measured them before and after dialysis every week. Inaddition we also measured the hANP before and after dialysis.Results: In the case of our two dialysis subjects both the TBW (Total Body Water) andICW (Intracellular Water) ware measured pre-dialysis. Moreover, neither case showedany significant changes throughout both respective pregnancies. However pre-dialysis theECW (Extracellular Water) increased moderately during the course of both successfulpregnancies.In comparison, with our healthy pregnant subjects, our pre-dialysis subjectsTBW/FFM (body water rate) measurements showed a tendency to be lower pre-dialysis.However the difference was not that significant. Furthermore, measurements of theECW/FFM (the cell external solution rate) also showed same tendency to be lower thanthat of their healthy counterparts. However, in the latter half of the dialysis subjectspregnancy terms, the ECW/FFM showed a value close to our healthy pregnant subjects.Furthermore according to the ICW/ECW (the intracellular and extracellular water ratio),the value post-dialysis also showed no significant difference to that of their respectivehealthy subjects. As for the ECW/TBW (the edematous rate), there were also nosignificant differences found. In addition, the hANP in our post-dialysis cases remainedalmost unchanged regardless of the gestational age of about 50.Conclusions: The BIA is regarded as a very useful testing method during pregnancy.Moreover it is even effective in the case of pregnant women with dialysis.

SP430 CORRELATION BETWEEN SERUM SODIUMCONCENTRATION AND PLASMACONDUCTIVITY ATDIFFERENT STAGES OF THE HEMODIALYSIS SESSION

Enrico Ravagli1, Laura Maldini2, Fabio Badiali2, Claudia Perazzini1, Giulia Lanciotti1,Denis Steckiph3, Alessandro Surace4, Paolo Rovatti4, Stefano Severi1 andAngelo Rigotti2

1University of Bologna, Cesena, Italy, 2“Degli Infermi” Hospital, Rimini, Italy,3Gambro Hospal S.p.A, Bologna, Italy, 4Gambro Dasco S.p.A, Medolla, Italy

Introduction and Aims:Monitoring changes in the body sodium pool duringhemodialysis (HD) sessions would be very useful and clinically relevant. Because acontinuous and non-invasive measurement of sodium concentration in blood duringHD is not currently possible, kinetic models for plasma conductivity (Cpl) estimationare used, based on the relationship between conductivity and ionic concentrations inplasma.The aim of the study was to investigate the correlation between the patient’sserum sodium concentration ([Na]) and Cpl estimated by the Diascan biosensor, forthe purpose of non-invasive estimation of [Na], at different stages during the HDsession.Methods: 24 dialysis patients were selected and a total of 152 sessions were studied.Arterial blood samples were taken for the determination of [Na] by blood gas analysis(direct potentiometry with an ion-selective electrode). Within 10 minutes after everysampling, Cpl was estimated by the Diascan biosensor and recorded by the dialysismachine. The timing of the blood sampling was 5’, 35’, 95’, 155’ from session start (SS)and 25’ before session end (SE).Recorded data was processed as follows: linearregression between [Na] and Cpl was calculated with the least-squares method for thewhole dataset at each sampling time. Correlation coefficients were also calculated. The95% confidence interval was calculated on the difference between the measured [Na]and the value estimated by linear regression.Results: The correlation between [Na] and Cpl for the whole dataset was R=0.54(p<0.001). The time-specific correlations are reported in Fig. 1. All regressioncoefficients are significant with p<0.001 except for SE (p=0.003). Since the correlationof SE data was largely lower than the correlation at any other time during the session,the global correlation was recalculated excluding data from the last measurement,resulting in a value of R=0.60 (p<0.001, Fig. 1, bottom right).The 95% confidenceinterval of the difference between measured and estimated [Na] (calculated using theparameters for linear regression estimated from the whole dataset) was ±5 mmol/l.Conclusions: Our results show that the global correlation between [Na] and Cpl iscomparable with that reported in previous studies.The evaluation of correlation atdifferent times during the session shows that this relationship is valid for the wholetreatment time, except at SE. This lower correlation is not clearly explained, but it couldbe related to the variability in measurement timing due to the variability in sessionduration.Since correlation is maintained up to SE, Diascan measurements could beused for non-invasive [Na] estimation. Moreover, estimation at SS (where correlation ishigh) could also be used for individualization of dialysis prescription.

SP431 NOCTURNAL HAEMODIALYSIS WITH THE VIVIAHAEMODIALYSIS SYSTEM

Phil McFarlane1, Rosa Marticorena1, Niki Dacouris1, Robert Pauly2, Sergey Nikitin2,Michael Amdahl3, Angelito Bernardo3 and Bruce Culleton31St. Michael's Hospital, Toronto, ON, Canada, 2University of Alberta Hospital,Edmonton, AB, Canada, 3Baxter Healthcare, Deerfield, IL

Introduction and Aims: The Vivia Haemodialysis System (Baxter Healthcare,Deerfield, IL, USA) was designed for the patient as the primary operator in a homeenvironment. To reduce the burden associated with performing frequent and / or longhaemodialysis (HD) in the home, the Vivia HD System includes features not typicallyavailable in traditional HD devices. These unique features include (1) extended use ofthe dialyser (2.1m2 polyethersulfone membrane) and blood set, facilitated by in-situhot water disinfection between treatments; (2) generation of on-line infusible qualitydialysate to allow automated priming, rinseback, and hemodynamic support duringhypotensive episodes; and (3) a fully integrated access disconnect system to mitigatethe risk associated with venous access disconnections. Given these unique features, wesought to determine the safety and efficacy of the Vivia HD System in a controlledclinical environment.

SP428

SP430



at Utrecht U



ownloaded from


Methods: Prevalent in-centre nocturnal HD patients were enrolled in this prospective,single arm clinical study at two clinical sites in Canada. All subjects received nocturnalHD, three times per week, for six weeks. Safety was assessed in all subjects who usedthe Vivia HD System at least once. Urea clearance was assessed using secondgeneration estimates for single pool Kt/V. Anticoagulation was obtained withunfractionated heparin with adjustments to maintain intra-dialysis activated partialthromboplastin times (aPTT) levels between 1.5x and 2.0x pre-dialysis values. Theassociation between fluid weight removed (as measured by Vivia) and weight change(determined using weigh scales) was calculated for each treatment. Dialysate wassampled at a minimum of three times per subject throughout the study. Dialysatecriteria for success was defined as a bacterial count of 0 CFU/mL and an endotoxinlevel of < 0.03 EU/mL, consistent with AAMI and ISO standards for dialysate forinfusion.Results: Seventeen subjects (mean age 55 years, 35% F) were enrolled. Sixteen subjectsexperienced at least one adverse event. The adverse event profile was similar to theevent profile expected for prevalent HD patients. Hypotension was the most commonadverse event (5 events per 100 treatments). There were no device related seriousadverse events (i.e. hospitalizations or deaths) during the study. During the evaluableperiod, the mean (standard deviation) duration of each treatment was 7.0 (0.6) hourswith a blood flow rate and dialysate flow rate of 293 (23) ml/min and 318 (27) ml/min,respectively. The mean single pool Kt/V was 2.49 (95% confidence interval 2.02 to2.96). The association between fluid weight removed as measured by Vivia and weightchange was high (R2 0.97). The maximum number of uses from a single dialyser andblood set was 17; 10 subjects (59%) achieved 10 or more uses from the same dialyser.Heparin bolus and infusion rates were increased from baseline in the majority ofsubjects by ~ 50%. Of 90 dialysate samples obtained, two samples showed bacterialcounts of 0.002 CFU/mL, both secondary to environmental contamination. Threedialysate samples had endotoxin results out of range: (0.03 EU/mL, 0.035 EU/mL and0.035 EU/mL). All positive endotoxin samples were retested in quadruplicate and allmeasured less than 0.03 EU/mL. The positive samples were due to known variability ofthe testing method at the acceptance limit.Conclusions: This study confirms the safety and expected performance of the ViviaHD System for extended duration HD treatments.

SP432 MID-TERM EVALUATION OF CITRATEDIALYSATE-REINFUSATE IN ON-LINE POSTDILUTIONALHEMODIAFILTRATION

Giovanni Calabrese1, Domenico Mancuso1, Antonio Mazzotta1, Giuseppe Vagelli1,Chiara Balenzano1, Denis Steckiph2, Andrea Bertucci2, Mario Della Volpe1 andMarco Gonella11Nephrology and Dialysis Unit ASL AL, Casale Monferrato, Italy, 2Gambro-HospalSpA, Bologna, Italy

Introduction and Aims: During HDF even low acetate concentrations in thedialysate-reinfusate (dAC) result in a sharp increase of serum AC, which induces therelease of proinflammatory mediators. A citrate-based concentrate (dCIT) was recentlymarketed, but, since CIT binds Ca++and can change Ca mass balance, this concentrateis not yet widely used. The aim of the present study was to evaluate the effect ofreplacing dAC with dCIT upon Ca and PTH level during high-volume HDF online. Asecondary end point was to evaluate the best strategy aimed to avoid undesirablechanges on pre-dialysis PTH value.Methods: In 16 patients (pts) on long-term HDF (Polyamide m. 2.1 m2, Qb 400 andUF 94±10 ml/min), dAC (3.0 mmol/l) was substituted with dCIT (1 mmol/l, SelectbagCitrate, Gambro) on acid concentrate for one week, keeping unchanged for each pt theoral therapy and dCa (1.5 mmol/l) (Period 1). During the following 6 months offollow-up on CIT-HDF, the mean dose of CaCO3 and Calcitriol was increased (25 and46%, respectively) (Period 2). During the last 6 months (Period 3), dCa was keptunchanged in 3 pts and was increased in 13 (1.65 mmol/l) and oral therapy returned tobaseline prescription. The HD monitor (ARTIS, Gambro) as well as the otherhemodialysis prescriptions (Session length, Blood flow, Dialyzer) were kept constantthrough all the periods.At the baseline and at the end of each period, PTH, total andionized Calcium (tCa, iCa) were measured at pre and post-dialysis in the mid-weeksession. Moreover, in the same sessions, total convective volume was recorded.Results: At baseline (AC-HDF) the tCa level increased during the session (from 2.31±0.12 to 2.63±0.16 mM, p<0.0001), while it was stable during Period 1, but it increasedagain in Period 2 and 3 (Period 1: from 2.37±0.14 to 2.42±0.11 mM, p=0.12; Period 2:from 2.31±0.15 to 2.47±0.11 mM, p<0.01; Period 3: from 2.23±0.08 to 2.60±0.10 mM,p<0.0001). The iCa increased in Ac-HDF (from 1.13±0.05 to 1.22±0.03 mM,p<0.0001) and decreased during the Cit-HDF treatments, except the Period 3 (Period1: from 1.12±0.07 to 1.07±0.03 mM, p<0.01; Period 2: from 1.09±0.07 to 1.05±0.03mM, p<0.03; Period 3: from 1.14±0.07 to 1.12±0.09 mM, p<0.55). The PTHconsistently followed the iCa, decreasing in Ac-HDF (273±196 to 141±99 pg/ml,p<0.0001) and decreased during the other Cit-HDF treatments, except during Period 3(Period 1: from 298±174 to 358±178 pg/ml, p<0.01; Period 2: from 337±207 to 409±204 pg/ml, p<0.005; Period 3: from 246±223 to 189±130 pg/ml, p<0.01).Finally, theconvective volume, set automatically by TMP biofeedback (UltraControl, Gambro),increased by citrate introduction, although not significantly (Baseline: 24.7±2.6 L,Period 1: 26.0±2.6 L, Period 2: 26.1±3.9 L, Period 3: 26.0±2.6 L, p=0.06).Conclusions: The adequate dialysate Ca in extracorporeal dialysis is still debated andshould be chosen on the basis of clinical factors. If a negative Ca balance is unwantedon CIT-HDF, the increase of dCa of 0.15 mmol/l can be taken into account as first

approach and the increase of oral therapy (Calcium an/or Vit. D derivates) as asecond step.

SP433 USEFULNESS OF TOXINOMETERMINI TOMONITORDIALYSATE PURIFICATION IN CENTRAL DIALYSATEDELIVERY SYSTEM

Takayuki Uchida1, Katsunobu Ando1, Masaya Kofuji1, Tsukasa Higuchi1,Naoki Momose1, Kiyonori Ito1, Yuichirou Ueda1, Haruhisa Miyazawa1,Yoshio Kaku1, Aoi Nabata1, Taro Hoshino1, Honami Mori1, Izumi Yoshida1,Susumu Ookawara1 and Kaoru Tabei11Saitama Medical Center Jichi Medical University, Saitama, Japan

Introduction and Aims: Central dialysate delivery system (CDDS), which producesdialysate at a central supply equipment, and distributes it from central to dialysismachines, was major system for supplying dialysate in hemodialysis (HD) treatment inJapan. Dialysate purification is essential to use the high performance dialyzers in Japan.Therefore, we routinely measured endotoxin (ET) activity and cultivated bacteria intodialysate in our dialysis center to evaluate dialysate purification. In this study, we aimedto evaluate the usefulness of Toxinometer Mini (WAKO Pure Chemical IndustriesLtd), which is automated instrument that measures ET activity by simple, easy andcost-benefit operation.Methods: To confirm the accuracy of Toxinometer Mini measurement, the data of ETmeasured by Toxinometer Mini was compared with authorized clinicalexaminationlaboratory (SRL, Tokyo, Japan). Thereafter, we measured ET by Toxinometer Minievery months for 18 months. For evaluating the count of bacteria, membrane filtrationmethod (37mm Quality Monitor Nihon Pall Ltd) was applied for cultivating bacteria.Samples were collected every month at reverse osmosis system, central supplyequipment and on-line hemodiafiltration (HDF) machines, every six months at powderdissolution machine, and once a year at each HD machines.Results: There was a significant and positive linear correlation between ETconcentrations measured by Toxinometer Mini and authorized clinical examinationlaboratory (r=0.99, p<0.001).The Bland-Altman plots for the comparative analysisshowed good agreements between the two methods for both Toxinometer Mini andauthorized clinical examination laboratory(N=82). Lower limit of ET measurementreached 0.001EU/mL for both instruments,and these values were satisfied thesensitivity of ET measurement recommended by Japanese Society for Dialysis Therapy(JSDT) guideline. We had measured ET concentrations and bacterial cultures for 18months in CDDS and single patient dialysis machine (SPDM) (N=199). Only once ofthese samples was positive in ET measurement at SPDM , and also ,several times ofbacterial cultures were positive in CDDS and SPDM , however detected ETconcentrations and colonization of cultivated bacteria were below the standardrecommended by JSDT guideline. In comparison between costs in ET measurementusing Toxinometer Mini and using authorized clinical examination laboratory, theformer required only 8.5 euro whereas the latter required about 20 euro per sample.Conclusions: The present study indicated that the ET measurement usingToxinometer Mini would be effective for monitoring dialysate purification. Sufficientmanagement to dialysate purification could be performed by ET measurement usingToxinometer Mini and cultivation of bacteria in dialysate.

SP434 DETECTION OF ALBUMIN FRAGMENTS ON DIALYZERMEMBRANES

Koichi Umimoto1, Miyuki Suyama1, Yoshimasa Shimamoto1, Masahiro Miyata1and Aki Kamada11Osaka Electro-Communication University, Shijonawate, Japan

Introduction and Aims: In recent years, albumin fragments (Af) have been found inthe urine of patients with diabetic nephropathy and the clinical significance of Af hasbeen noted. When hemodialysis (HD) is performed, albumin is known to be adsorbedto the dialyzer membrane and it leaks into the dialysate with high-flux dialyzers. Wehave previously reported about analysis of protein adsorbed to dialyzer membranes.This time, we investigated adsorbed Af on dialyzer membranes and Af in theultrafiltration (UF) fluid after dialysis with high-flux dialyzers.Methods: After standard HD with a cellulose triacetate (CTA) membrane orpolymethylmethacrylate (PMMA) membrane, the dialyzers were rinsed with 0.9%NaCl solution. Then the membranes were cut into small fibers and the proteins wereeluted from each membrane. At 30 minutes after starting HD with a PMMAmembrane or a polyester polymer alloy (PEPA) membrane, UF fluid was collected.Then the samples were analyzed by dodecyl sulfate polyacrylamide gel electrophoresis(SDS-PAGE) and proteins were identified by western blotting (WB) using antihumanserum. In addition, quantitative image analysis was performed with a Gel Doc EZ (BioRad, Ltd.).Results: SDS-PAGE and WB showed the main band at 66 kilodaltons (KDa) after HDwith both CTA and PMMA membranes, and this was identified as albumin. Thinbands were also found in the vicinity of 20 KDa and 40 KDa after HD with bothmembranes and these were identified as Af. In the UF fluid, albumin was detected at 66KDa with both PMMA and PEPA membranes, and its quantity was 16.25×105and20.69×105 signal intensity·mm2, respectively. There was no other band with the PMMAmembrane except the band of albumin at 66 KDa, but thin bands of AF were found in



at Utrecht U



ownloaded from


the vicinity of 20 KDa with the PEPA membrane and the quantity of Af was 0.13×105

signal intensity·mm2.Conclusions: It is already known that serum proteins are adsorbed to dialyzermembranes and that albumin leaks into the dialysate with high-flux membranes. Theseresults confirm that Af is also adsorbed to dialyzer membranes and that a smallamount of Af leaks into the dialysate as well as albumin with PEPA membranes.

SP435 EVALUATION OF VOLUME OVERLOAD IN HEMODIALYSISPATIENTS USING BIOIMPEDANCE DEVICE

Rika Sakai1, Akihiro Minakawa1, Keiichi Fukudome1, Shuichi Hisanaga1,Tabito Ishihara2, Kazuhiro Yamada3, Shin Fukunaga3, Hiroko Inagaki4,Chihiro Tanaka4, Yuji Sato5 and Shoichi Fujimoto61Koga General Hospital, Miyazaki, Japan, 2Fujimoto Central Hospital, Miyazaki,Japan, 3Social Insurance Miyazaki Konan Hospital, Miyazaki, Japan, 4ChiyodaHospital, Miyazaki, Japan, 5Dialysis Division, University of Miyazaki Hospital,Miyazaki, Japan, 6Department of Hemovascular Medicine and Artificial Organs,Faculty of Medicine, University of Miyazaki, Miyazaki, Japan, Miyazaki, Japan

Introduction and Aims: Assessment of body fluid overload in hemodialysis patients isclinically important. However, objective methods that evaluate fluid over load arelimited.Methods: This was cross-sectional study assessing the usefulness of a newbioimpedance spectroscopy device to evaluate over hydration (OH) just before and justafter hemodialysis session in Japan. 147 patients (mean age 65.9 years, 57.8% men,31.3% diabetes was basal kidney disease, 19.7% comorbid with congestive heart failure,ischemic heart disease and valvular heart disease) were enrolled.Results: Pre-hemodialysis (HD) OH was ranged from +1.08 to +4.22 and post-HD OHfrom -1.02 to +2.10 (10thand 90th percentile, p<0.001 by Wilcoxon signed-rank test).Patients number was increased from 9.5% of pre-HD patients to 61.9% of post-HDpatients who were within normal range of OH (-1.1 to +1.1 L). Amount ofultrafiltration was significantly correlated with absolute change of OH between pre-HDand post-HD (r=0.508, p<0.01). Post-HD OH was positively correlated with post-HDplasma atrial natriuretic peptide(ANP) levels (r=0.505, p<0.01). By multivariate logisticregression analysis, higher ANP level (over than 100 pg/ml) was significantly associatedwith post-HD OH, age and CHF or valvular heart disease, and higher post-HD OHlevel (over than +1.1 L) was significantly associated with pre-HD OH and plasma ANP.Conclusions: In conclusion, new bioimpedance spectroscopy device is useful toevaluate over hydration in dialysis patients. These OH data suggested that a certainlevel of our patients might not be set ideal dry weight.

SP436 ON-LINE HEMODIAFILTRATION (OL-HDF): WHICHMODALITYFORWHICH BLOOD FLOW RATE?

Jacky Potier1, Julien Bouet1 and Guillaume Queffeulou11CHPC, Cherbourg, France

Introduction and Aims: Even if Post-HDF (POST) has become the reference, otherol-HDF modalities (HDF) differentiated by their infusion site (PRE, MIXED and MID)permit Middle Molecules (MM) to be optimally removed. Then, the absence ofvolumetric recommendation - as with POST- and generalization of HDF, whatevervascular access and blood flow rate (Qb) have led us to compare HDF from their MMremoval efficiency and to find the best match between Qb and HDF.Methods: 10 patients (pts); M=7; F=3; Age 74±10.9y; dialysed since 55.2±70munderwent 4 hours PRE, POST MIXED and ol-HDF sessions with 5008Cordiax(Fresenius), AutoSub+ and FXCordiax1000 dialyser. MID 4 hours sessions wereperformed with an OLPURMD220 (Bellco) dialyser and Total Convective Volume(Vconv) in MID was fixed to be equal to the one obtained in MIXED in the sameconditions. MM removal efficiency of beta2-microglobulin (β2m; 11.8kDa), myoglobin(Myo; 17.2kDa), Prolactin (PLT; 23kDa) and Orosomucoïd (ORO; 42kDa) wereevaluated by Reduction Ratios (RR(%) = (Cpre-CPost) / Cpre with Cpost corrected forhemoconcentration).Each pt was dialysed with each one of the 4 HDF modalities, butwas allocated to only one Qb Group (Qb): 250mL/min (Qb250) for 4 pts (3/4 with achronic jugular catheter); Qb300 for 3 pts and Qb350 for 3pts. Each pt was dialysed witheach one of the 4 HDF. Statistical analysis (StatView) was performed with Student'sunpaired test. Data are expressed by mean +/- SD and P<0.05 is considered as significant.Results: For POST andMIXEDmodalities, the lower Vconv (L) were obtained in Qb250group (Vconv POST = 16.3±1.7; p=0.001, Vconv MIXED = 28.7±2.6 (NS)), versusQb300 (V Conv POST: 24.7±1.3, VconvMIXED: 34.3±5.6) or Qb350 (Vconv POST:25.6±1.7, Vconv MIXED: 34.9±2.3). Mean Vconv PRE was 57.2±9.8 with no influence ofthe Qb.RRβ2m: Qb250 :79.2±2.8 ;Qb300 :81.7±3.3 ;Qb350 :82.3±5.9 . No significantdifference between Qb or between HDFmodalities in each Qb groupRRmyo: Qb250 :62.0±8.6 ;Qb300 :70.0±10.1 ;Qb350 :67.8±12.2 , p=0.032 between Qb250 and Qb300 . InQb300 group, MIXED and POST were more efficient for myoglobin removal (RR = 76.2±5.9 and 77.8±0.8) vs PRE(56.3±7.5).RRPLT: Qb250 :65.1±10.1 ;Qb300 :67.8±15.9 ;Qb350 :63.7±14.6 - No significant difference between Qb or between HDFmodalities ineach Qb group.RRORO: Qb250 :6.0±5.4 ;Qb300 :6.3±8.6 ;Qb350 :9.4±4.0 - No significantdifference between Qb. In Qb350 group, MIXED and POST were more efficient thanPRE for orosomucoid removal (RR MIXED = 11.5±1.6; p=0.01, RR POST = 10.31±3.2;p=0.049, RR PRE= 4.1±2.03)

Conclusions: Vconv seemed optimal from Qb300. Globally, in terms of MM removal,there was no difference between Qb, each one including all HDFmodalities. This findingargues for a universal use of HDF whatever Qb. In each Qb, there was only minordifferences between HDF and always between POST or MIXED versus PRE. For Qb250and Qb300 MIXED was always the more efficient and proportionally to the MWof themolecule. Finally we recommend all HDF modalities, but especially those with aPost-dilution participation if removal of toxins with a MWabove the β2m one is wished,as often actually recommended.

SP437 EFFICACY AND SAFETY OF TINZAPARIN ANTICOAGULATIONOF THE EXTRACORPOREAL CIRCUIT WITH A SINGLE BOLUSADMINISTRATION IN NOCTURNAL HOME HEMODIALYSIS

Robert Bell1, Linda Nolin1, Vincent Pichette1, Helene Provencher1,Caroline Lamarche1, Annie-Claire Nadeau-Fredette1, Georges Ouellet1,Martine Leblanc1, Sarah Bezzaoucha1, Yasmina Kouidmir1, Jeannine Kassis1,Marie-Louise Alonso1, Jean-Philippe Lafrance1 and Michel Vallee11Hopital Maisonneuve-Rosemont, Montreal, QC, Canada

Introduction and Aims: Tinzaparin, a low molecular weight heparin, has shownpractical benefits over unfractionated heparin for extracorporeal circuit (ECC)anticoagulation during in-center hemodialysis treatment. However, efficacy and safetyof anticoagulation with tinzaparin has not been substantiated in patients with extendeddialysis sessions, as in nocturnal home hemodialysis (NHD). In this regard, we presentour experience with tinzaparin, administrated as a single bolus injection into thearterial line of the ECC for patients requiring 8 hours of anticoagulation.Methods: Fifteen chronic dialysis patients admitted to our NHD program between2009 and 2012 had their first 8-hour in-center dialysis session receiving twice thetinzaparin dose that gave adequate anticoagulation for their usual 4-hour dialysistreatment. This represented a mean tinzaparin dose of 54± 12 and 110± 19 anti-Xaunits/kg for a 4-hour and 8-hour dialysis session, respectively. Safety and dose/responsewere assessed by measuring anti-Xa levels at time 0 and at 15, 30, 60, 120, 240, 360 and480 min after administration. Anticoagulation efficacy was evaluated visually, in 14 ofthe 15 patients, by assessing clot formation in both the dialyzer and the venous bubbletrap through a simple scoring system.Results: A mean peak plasma anti-Xa level of 1,35 ± 0,52 U/ml was reached within 15min and averaged 0,32 ± 0,16 U/ml by the end of the 8-hour session. Although therelationship between anti-Xa levels and clinical outcomes is unclear, usualrecommended therapeutic anti-Xa levels for once-daily treatment ranges from 1.0 and2.0 U/ml. There was no complete coagulation of the ECC in either the dialyzer or thevenous bubble trap. There were 7 moderate coagulation events: 1 in the dialyzer and 6in the venous bubble trap. There were no episodes of minor or major bleeding in any ofthe patients. Subsequent doses were adjusted to prevent further clotting. The meantinzaparin dose was 111± 20 anti-Xa units/kg before and 113±15 anti-Xa units/kg afterdose adjustment, p > 0,05.Conclusions: Our experience is the first to demonstrate that a single weight-basedintravenous bolus administration of tinzaparin in NHD is effective and safe.

SP438 MULTIPARAMETRIC ONLINE SENSORS FOR DIALYSISDEVICE

Julien Fils1 and Pascal Mailley21Metemis, Moirans, France, 2CEA-LETI, Grenoble, France

Introduction and Aims:Monitoring system for dialysis machine. To improve thequality of the dialysis treatment and to follow the health status of the patient duringthis treatment, the continuous monitoring of the blood composition in terms of ioncontains provides accurate information. Indeed, away from the patient staus, thismonitoring gives a clear view of the efficiency of the machine and allows anticipatingadjustments in the device setup.Methods: In the framework of the European FP7 project Nephron+, the partnerdeveloped a sensing platform (multisensory system and associated electronicmanagement) for the monitoring of endogenous ions including H+ (pH) calcium,sodium, potassium and chloride. It is worth to notice that this platform is not limitedto the aforementioned species and may include if required sensors dedicated to smallorganic species such as urea or glucose. The plateform sensors are embedded into theWearable Artificial Kidney Device developed within this NEPHRON+ project tomonitor the status of the device and the health parameters of the dialysate patient.Each of these sensors works on the potentiometry mode: briefly, the difference ofpotential between a reference electrode and a selective sensitive electrode is measured,its amplitude being proportional to the concentration of the target ion. Thepotetiometric electrodes, so called Ion Selective Electrodes (ISE) are designed in threelayers. The first one is a metallic electrode which is connected to the electronic board,then a liquid layer which acts as internal electrolyte (that contains a referenceconcentration of the target ion) between the metallic electrode and the sensitivemembrane used as third layer. The latter is chemically designed to be specific for oneion without interferences from the ionic fauna of the dialysate.These electrodes aredesigned to be low cost in order to allow their integration in one-shot systems. Theirsmall size, less than one 1mm square, enables their positioning in direct contact withthe dialysate within the fluidic sensing chamber which is placed directly on the main



at Utrecht U



ownloaded from


dialysate fluidic line. These sensors can be implemented in several part of the devicedue to their small form factor. Furthermore, the sensors have been proved to bemechanically stable (no leakage or drilling of the membrane) for at least one monthunder function. In addition, the membrane and the inner solution are biocompatibleaccording to the ISO standards (13995) and can be sterilized thanks to manufacturingconditions (sterile environment) or beta sterilization.The sensors deliver real timemonitoring of ion concentrations with a response time of few seconds. Thereby, thepotential measurements are sampling every second or more (according to the requireddynamic) and sent to the display or the main board of the device in real time.Results: The sensor platform has been tested by the company Nanodialysis andvalidated after more than one week of use in continuous within artificial dialysate. Eachsensor shown long lifetime (more than 2 months), and provided the required accuracyand precision. The sensors show good performances against interferences. At now,owing to their simple and efficient architecture, the sensors can be implemented withinan industrial process to produce them in a large scale.Conclusions: The potentiometric sensor platform, designed for the monitoring ofdialysis, has demonstrated its ability to provide continuous real time multiparametricinformation about the state of the patient and of the dialysis machine efficacy. The designof the sensors, especially the constitutive materials, provides biocompatibility and abilityto be sterilized. The transfer to an industrial company will allow providing these sensorplatforms or unitary sensors in large quantity to perform their CE marking.

SP439 COUPLED PLASMA-FILTRATION ADSORPTION (CPFA)REDUCES TUBULAR INJURY IN BILE-ASSOCIATED CASTNEPHROPATHY THROUGH DIRECTADSORPTION OFBILIRUBIN AND LIVER-TYPE FATTY ACID BINDING PROTEIN

Vincenzo Cantaluppi1, Davide Medica2, Alessandro D Quercia2,Sergio Dellepiane2, Silvia Ferrario2, Massimo Gai2, Gianluca Leonardi2,Cesare Guarena2, Marialuisa Caiazzo3 and Luigi Biancone21University of Torino, Torino, Italy, 2Nephrology, Dialysis and Kidney TransplantationUnit, Torino, Italy, 3Scientific Affairs, Bellco Srl, Mirandola (MO), Italy, Italy

Introduction and Aims: Bile-associated deposition of intraluminal casts and proximaltubular epithelial cell (TEC) apoptosis are possible causes of acute kidney injury (AKI)in the course of severe liver dysfunction. TEC injury is mediated by bilirubinadsorption through a mechanism dependent on the activity of the endocytic receptormegalin. Liver-type Fatty Acid Binding Protein (L-FABP) is a 15 KDa peptidebelonging to the free fatty acid family able to bind hydrophobic molecules includingbilirubin. During liver failure, the increase of L-FABP plasma levels is known toenhance bilirubin uptake with consequent damage of tubular cells. Indeed, L-FABP isable to bind to megalin mediating bilirubin uptake in TEC. The aim of this study wasto investigate the protective role of Coupled Plasma Filtration Adsorption (CPFA) onbile cast nephropathy through direct L-FABP and bilirubin resin adsorption.Methods:We reported the case of a kidney transplant patient who developed sepsis,AKI and liver dysfunction treated by CPFA. We evaluated plasma levels of bilirubinand L-FABP. Renal biopsies, urine sediment and NGAL were performed at differenttime points. In vitro, we tested: 1) static and dynamic adsorption of L-FABP topolystyrene resin; 2) pro-apoptotic effect (TUNEL) of patient’s plasma drawn beforeand after CPFA on human tubular epithelial cells (TEC): the role of L-FABP wasfurther confirmed in tubular cells engineered to knock-down megalin, the L-FABPreceptor, by small interfering RNA (siRNA).Results: A 50-year-old man was subjected to kidney transplantation with slow recoveryof graft function. Kidney biopsy revealed acute tubulo-interstitial and vascular rejectiontreated with thymoglobulin. He then developed septic shock for Legionella infectionwith multiple organ failures (serum creatinine 5.2 mg/dl and oliguria requiring dialysis;bilirubin 42 mg/dl with liver biopsy showing marked cholestasis; plasma L-FABP 52ng/ml). Urine analysis showed the presence of tubular cells, casts and intense positivityfor bilirubin: urine NGAL level was 356 ng/ml. A new kidney biopsy showing bile castnephropathy was performed. After CPFAwas started, we observed an increase of urineoutput, a decrease of bilirubin (<15 mg/dl), L-FABP (9 ng/ml) and urine NGAL (82ng/ml).In vitro, we observed that the polystyrene resin efficiently adsorbed L-FABP instatic and dynamic condition. When we incubated TEC with patient’s plasma, weobserved a cytotoxic and pro-apoptotic effect with activation of caspases. After CPFAtreatment, the pro-apoptotic activity of patient's plasma on TEC was significantlyreduced. In addition, plasma-induced TEC apoptosis was dependent on the presence ofmegalin, the L-FABP receptor: indeed, the pro-apoptotic effect of plasma wassignificantly reduced in TEC engineered to knock-down megalin by siRNA.Conclusions: CPFAmay have a protective role in bile-associated cast nephropathy andTEC apoptosis through the direct adsorption of bilirubin and L-FABP to the syntheticpolystyrene resin. CPFA may be exploited as therapeutic option to limit AKI duringsepsis and in bridge to transplantation of patients with end stage liver failure.

SP440 CARBON FOOTPRINTS OF AN IN-CENTRE HAEMODIALYSISDEVICE, THE NXSTAGE SYSTEM ONE AND THE VIVIAHAEMODIALYSIS SYSTEM

Margaret Enos1, Bruce Culleton1 and Derek Wiebenson21Baxter Healthcare Corporation, Deerfield, IL, 2Baxter Healthcare SA, Zurich,Switzerland

Introduction and Aims: Increasing haemodialysis (HD) duration and / or frequency isgaining popularity and has been shown to improve clinical and humanistic outcomes.Furthermore, there is an increasing interest to create a more sustainable therapy for thisgrowing patient population, as demonstrated by initiatives such as the National HealthService’s carbon reduction strategy in the United Kingdom, the Green NephrologyProgramme for sustainable kidney care, and Environmentally Preferable Purchasingpolicies adopted by hospitals and purchasing organizations. In order to furtherunderstand the potential environmental impact of a growing home HD patientpopulation, this study assessed the carbon footprints of a conventional HD (CHD)device (Arena, Baxter Healthcare, Deerfield, IL), used both in-centre and in the homesetting, and two Home HD machines (Vivia, Baxter Healthcare, Deerfield, IL andSystem One, NxStage, Lawrence, MA), including regimens of nocturnal HD, shortdaily HD and CHD for six countries (Canada, France, Germany, Sweden, the UnitedKingdom, and the United States).Methods: This study utilized the life cycle assessment method to calculate thecradle-to-grave greenhouse gas emissions including consumable supplies, energy andwater used during treatment, people transportation, and waste disposal. The frequencyand duration of the treatments assessed were 5 times a week, 7 hours per treatment fornocturnal HD; 6 times a week, 3 hours per treatment for short daily HD; and 3 times aweek, 4 hours per treatment for CHD.Results: The overall carbon footprint was greatest for the System One device whenused for short daily HD, followed by the same device when used for frequent nocturnalHD. Averaging the results across all six countries, the carbon footprint was mostdependent on the quantity and types of consumables used for each therapy (see Table).Consumables represented 83-87% of the carbon footprint for System One, 66-68% forArena in the home setting, and 43-51% for Vivia. Comparing the same therapy, theVivia HD System, in which the dialyser and blood set can be used multiple times, had a50-56% smaller carbon footprint than NxStage System One and a 42% smaller carbonfootprint than the Arena device. Either patient travel for a monthly doctor’s visit orwater usage during treatment made the smallest contribution (30%) after treatmentconsumables (>40%).Conclusions: These results suggest that maximizing the use of consumables orproducing consumables in a more environmentally responsible manner will have thegreatest impact in delivering sustainable HD therapies. Although high dose HD hasmore weekly treatment sessions, performing this therapy in the home with Viviagenerates fewer emissions than CHD performed less frequently in-centre.

SP441 IS THERE A RECOMMENDED CONVECTIVE VOLUME FORMIXED HEMODIFILTRATION?

Jacky Potier1, Mélanie Hanoy2 and Simon Duquennoy31CHPC, Cherbourg, France, 2CHU, Rouen, France, 3CHU, Caen, France

Introduction and Aims: Since ESHOL study, one considers a Convective Volume(Vconv) >= 23.1L as an objective but this goal is restricted to Post ol-HDF (POST).MIXED is currently proposed on 5008 Cordiax (Fresenius) with a complete automatedmode. The delivered Substitution Volume (VSubs) is then shared between Pre andPost-Dilution modalities. Observational data showed high consistency betweendelivered Vconv in POST - managed with AutoSub+ - and its corresponding PostVconv (= Post VSubs + UF) in MIXED.We tried to determine if this Volumeequivalence is correlated to a similar Middle Molecule (MM) solute removal, the aimbeing to find a volumetric recommendation also for MIXED.Methods: 21 patients (pts) from 3 centers; M=12; F=9; Age = 72.1±13.9 years; Weight= 73.1±14.5 Kg, dialyzed since 84.2±83.6 months, underwent one 4 hours dialysissession with POST (AutoSub+) and one MIXED session. All sessions were performedon 5008 Cordiax with FX Cordiax 1000 hemodiafilter. MM removal efficiency ofbeta2-microglobulin (β2m; 11.8kDa), Myoglobin (Myo; 17.2kDa), Prolactin(PLT;23kDa) and Orosomucoïd (ORO; 42kDa) was evaluated by Reduction Ratios (RR= (Cpre-CPost) / Cpre with Cpost corrected for hemoconcentration.)Each patient wastreated with only one of the 4 Qb groups (Qb): 250mL/min (Qb250) for 6 pts (1catheter); Qb300 for 5 pts (2 catheters), Qb350 for 5pts and Qb400 for 5pts. Statisticalanalysis (StatView) was performed with Student's unpaired test. Data are expressed bymean +/- SD. P<0.05 is considered as significant.Results: (Cf Table 1) Total Vconv (L) was proportional to Qb. It was higher in MIXED:34.5±5.1 (from 29.1±3.4 with Qb250 to 39.5±1.7 with Qb400). Post infusion ratio inMIXED reached 66 to 69%. and there was no significant difference between Post Vconv

SP440



at Utrecht U



ownloaded from


in the two modalities: Mean Post Vconv in MIXED = 24.4±3.5 (from 20.0±1.2 withQb250 to 28.7±1.3 with Qb400) and Mean Post VConv in POST = 25.2±3.6 (from 21.3±0.9 with Qb250 to 29.9±1.6 with Qb400), For each Qb, RR were not significantlydifferent between POST and MIXED, whatever the solute, even if Qb250 RR valueswere lower for β2M and Myo. For higher MW solutes, MIXED appeared better (butNS) and for PLT, RR was even inversely correlated to Qb in POST.Conclusions: 5008 Cordiax system delivers similar Post VConv - whatever Qb - inMIXED and POST modalities, probably linked to the new AutoSub+. This similarityallows us to compare the MM removal efficiency of two equivalent Post Vconv, the firstpre-diluted with MIXED and the second considered as the reference in terms ofmorbidity and mortality. As the MM removal is consistently identical, we suggest thevolumetric recommendations for MIXED should be similar to those of the POST (PostVconv>= 23.1L). Otherwise, the Pre-D contribution with MIXED could explain thetendency toward a better removal of the highest MW solutes that should be confirmedby a study with a greater number of patients.

SP442 SAFETY AND EFFICACYOF REGIONAL CITRATEANTICOAGULATION IN SUSTAINED LOWEFFICIENCYDIALYSIS

Wang Tingli1, Zhang Ling1, Shi Yunying1 and Fu Ping11West China Hospital of Sichuan University, Chengdu, China

Introduction and Aims: Regional citrate anticoagulation,which has been widely usedin CRRT, can reduce bleeding complications.However, the application of regionalcitrate anticoagulation and heparin has not been compared in SLED. In order toprovide evidence for clinical practice, this study evaluated the safety and efficiency ofregional citrate anticoagulation in SLED, and comparedits application with heparin.Methods:We prospectively enrolled 63 patients with AKI or ESRD in nephrologydivision, West China Hospital, Sichuan University, between November 2011 andOctober 2013. These patients were randomized to receive citrate with calcium-freedialysate or heparin with conventionaldialysate. All of the patients received SLEDtreatment by Fresenius4008sARrTplus dialyzer through either femoral or internaljugular venous catheter, with each session of treatment lasting for 8 hours.In thecitrate group, we pumped in 4% tri-sodium citrate solution through the arterial line at130ml/h and 10% calcium gluconate through the venous line at 40ml/h. We recordedsystemic citrate concentration at 2h and 5h, peripheral and post filter ionized calciumlevel at 0h, 2h and 5h respectively. In the heparin group, we injected 2000-2500 IUheparin in vein and followed with 200-250 IU per hour pumped in through the arterialline.The blood flow was at 150ml/min and the conventional dialysate at 200ml/min.PT and APTT was monitored at 0h and 2h in all of the patients.Results: Thirty patients in the citrate group underwent 56 sessions of SLED whilethirty three patients in the heparin group underwent 62 sessions. The average age ofthe two groups was 71.7 ± 10.0 and 70.1 ± 12.8 years old. Both groups containedpatients with heart failure and/or respiratory failure, septic shock, uremicencephalopathyand poor volume control. A total of 4 patients died duringhospitalizationof multiple organ failure, and the death was not related to SLED and itsanticoagulation. The other 59 patients (93.7%) was discharged after treatment orconverted to outpatient IHD.A patient in the heparin group had hematoma andbleeding at femoral vein catheter. In heparin group, PT and APTT at 2h wassignificantly higher than that in the citrate group(PT: 15.5 ± 2.0 vs. 12.3 ± 2.7s, P<0.001; APTT:56.0 ± 10.9 vs. 32.8 ± 6.1s, P<0.001). The Citrate group received calciumgluconate pump through the venous side, so there was no significant effect on ionizedcalcium in peripheral blood.Ionized calcium levels in peripheral blood at 0h, 2h, 5hwere1.22 ± 0.15 mmol / L, 1.21 ± 0.11 mmol / L and 1.25 ± 0.11 mmol / L(P = 0.76).Post-dialyzer ionized calcium at 2h and 5h was significantly lower than that in theperipheral blood (0.29 ± 0.05mmol / L vs. 0.31 ± 0.06 mmol / L, P<0.001). Throughclearance by the dialyzerand the body's own metabolism,citrate level in peripheralblood was significantly lower than the post-dialyzer level (P<0.001).Transmembranepressure 2h and 5h were 105.1 ± 14.8mmHg and 107.1± 15.5mmHgrespectively(P= 0.68).Conclusions: SLED under regional citrate anticoagulation is safe and effective. Citrateachieves satisfying regional anticoagulation effect without interfering systemic clottingfunction, and shows the similar anticoagulant effect to heparin without causing theriskof bleeding. There is no significant metabolic alkalosis and disturbance ofelectrolyte during and after SLED.It is more safe for the patients with high risk ofbleeding, and also provides clinicians with an alternative anticoagulant approach forSLED.

SP443 CITRIC ACID AND ACETATE FREE BASED DIALYSATE INONLINE HEMODIAFILTRATION POST-DILUTION ALLOWSHEPARIN FREE SESSIONS

Thibault Dolley-Hitze1, Didier Hamel1 and Marie-Laure Lombart11AUB Santé, Saint-Malo, France

Introduction and Aims: Up to now, heparin is the simplest mean to realise OnLineHemoDiaFiltration Post-Dilution (OL-HDF) without clotting however it inducesseveral adverse events (increased bleeding risk, bone metabolism disorders,dyslipidemia) and exposes nurses to patients’ blood. Citric acid based concentrate hasbeen yet studied in order to stop heparin injection during dialysis sessions (J. Aniort,Blood purif. 2012) but this dialysate contains a small proportion of acetate (citric acid0.8mM, acetate 0.3mM) with its own side effects (per-dialytic hypotension, cramps,inflammation). The aim of our study is to determine if citric acid and acetate free fluid(citric acid 1mM, acetate 0mM) is safe in HDF-OL PostD and can allow heparinremoval.Methods: All stable patients treated by OL-HDF with 3mM acetate on machinesaccepting the studied concentrate (Select Bag Citrate®, Gambro) were included. They allreceived citrate concentrate and have been randomly assigned to either the controlgroup (fraxiparin as usual) or study group (fraxiparin slow withdrawal, 0.1ml per weekuntil discontinuation) for a 3 months period. Dialysis Fluids contained 0.10 or0.15mEq/l calcium more than usual (1.60 or1.65mEq/l) in order to avoid hypocalcemiabut other dialysis parameters didn’t change. The blood calcium concentration (total orionized) has been checked weekly for 4 weeks. All the hemorrhagic or clotting eventswere thorough noted as well as online dialysance and convection volumes at eachsession. Statistical analyses were performed with Statview® software and non parametrictests were applied.Results: 17 patients were enrolled in the study, 9 were randomly assigned to the controlgroup and 8 to the study group (fraxiparin discontinuation). 589 sessions wereanalyzed, 306 in the control group and 283 in the study group. Total and ionizedcalcium concentration remained stable during the study period. In the study group, nomore clotting events were observed (13 in each group, p=0.94). Fraxiparin wasdiscontinued by 5 out 8 patients in the study group and 0 out 9 in the control group(p=0.009) with 0.075ml mean dose in study group vs 0.41 ml in control group(p=0.018). The dialysis dose did not differ among the two groups (Kt/V Daugirdas 2:control group 1.81 vs 1.72 study group, p=0.68 and On-line Kt 51.64 vs 50.53, p=0.40)as well as total convective volumes (26.45 l in control group vs 26.50 l in study group,p=0.94).Conclusions: Heparin free OL-HDF is achievable with the use of citrate based andacetate free concentrates. Dialysis parameters and performances remain stable.Nevertheless, not all the patients can benefit from this technique and the reasonsshould be determined by a larger trial.

SP444 EFFECTOF HOTWATER DISINFECTION ON DIALYSERSOLUTE CLEARANCESWITH EXTENDED USE IN A NEWHAEMODIALYSIS DEVICE

John K Leypoldt1, Angelito Bernardo1, Audrey M Hutchcraft2,Raymond Vanholder3 and Bruce F Culleton11Baxter Healthcare Corporation, Deerfield, IL, 2Baxter Healthcare Corporation,Round Lake, IL, 3University Hospital, Ghent, Belgium

Introduction and Aims: Conventional reuse of dialysers creates potential exposure tothe toxic effects of chemicals, and some chemical-based reuse methods do notmaintain dialyser solute clearances, depending on membrane and chemical. A newhaemodialysis (HD) device, Vivia (Baxter Healthcare, Deerfield, IL, USA), wasdesigned to permit extended or multiple use of the dialyser and blood tubing set byin-situ, non-chemical hot water disinfection, thereby eliminating chemical exposureand the risk of infections due to blood circuit disconnections. We aimed to evaluate theeffects of hot water disinfection on dialyser solute clearances by performing a clinicalstudy with Vivia to assess the effects of extended use on urea and β2-microglobulin(β2-M) clearances.Methods: A prospective, single arm clinical study was performed on 22 chronic HDpatients (12 M, 10F). Mean ± SD age of the patients was 50.9 ± 9.5 years; 16 patientswere black, 5 were white and 1 was a Taino Indian. Hot water disinfection (exposure to85ºC for 1 hour) was performed prior to every use of the dialyser, including first use.Over the course of this 8-week evaluation, HD treatments were performed with a

SP441 Table 1 RR Results (MIXED in the top line of each cell of the table)

RR% β2m Myo PLT ORO

Qb250 80.4±4.7 81.3±2.9 71.0±8.7 74.3±6.6 72.9±7.8 73.0±7.1 14.4±7.2 12.8±10.6Qb300 84.3±1.4 84.1±2.5 80.1±3.4 78.2±6.3 78.9±7.9 74.72±7.0 16.6±11.0 11.7±10.1Qb350 84.5±6.2 84.8±6.2 78.9±7.9 74.72±7.0 75.7±12.9 70.0±18.9 16.0±4.0 16.9±7.6Qb400 84.2±5.8 85.6±3.6 77.4±3.1 76.4±5.1 76.0±7.7 67.3±15.7 17.6±4.0 12.03±9.6

SP444

Dialyser Use Counts URR β2-M Clearance (mL/min)

1 69 ± 8 (171) 77 ± 31 (171)2-4 70 ± 7 (71) 79 ± 27 (73)5-10 73 ± 6 (34) 77 ± 20 (33)>10 74 ± 7 (21) 72 ± 15 (20)



at Utrecht U



ownloaded from


2.1 m2 polyethersulfone membrane dialyser for 3.9 ± 0.2 hours, 4 times per week.Dialysers were replaced if in-situ dialyser sodium clearance was <90% of the initialdialyser sodium clearance or if there were abnormalities in physical appearance.Pre-dialysis and post-dialysis blood samples were obtained at first use, second use andthen weekly; the urea reduction ratio (URR) and dialyser β2-M clearance werecalculated to assess extended use dialyzer performance.Results: Dialyser performance was assessed in 207 dialysers across 648 treatments.Thirty-nine percent of dialysers were used multiple times and 41% of treatments weredelivered by dialysers that achieved 5 or more uses. Abnormalities in physicalappearance, primarily residual clots in the venous header of the dialyser or the bloodset, most often terminated extended use. Overall, dialyser performance was evaluated ata blood flow rate of 357 ± 21 mL/min and a dialysate flow rate of 395 ± 9 mL/min.Performance results are tabulated as mean ± SD (Number of determinations): Post-hocrepeated measures analysis of URR and β2-M clearance on dialyser use count fordialysers used at least 10 times in 12 patients was also performed. In this analysis, URRwas independent of dialyser use count (P=0.7), but the estimated β2-M clearancedecreased (P=0.02) by 0.74 (standard error of 0.27) mL/min for each dialyser use.Conclusions: Hot water disinfection during extended use of large surface areapolyethersulfone dialysers in the Vivia device has no effect on dialysis adequacy asassessed by URR. Extended use decreases dialyser β2-M clearance; however, theseeffects are small and likely not clinically significant given the high baseline β2-Mclearance rates.

SP445 THE ROLE OF CONVECTION ON THE LONG-TERMVARIATIONS (Δ)OF SERUM BETA 2MYCROGLOBULIN (B2M),C-REACTIVE PROTEIN (CRP) CONCENTRATIONS, AND ESAREQUIREMENT (ΔESA) IN UREMIC PATIENTS TREATED BYPOST DILUTIONALON-LINE HDF

Ezio Movilli1, Corrado Camerini1, Paola Gaggia1, Roberto Zubani1, Paolo Feller1,Alessandra Pola1, Orsola Carli1, Chiara Salviani1, Chiara Manenti1 andGiovanni Cancarini11U.O. Nephrology, Spedali Civili and University of Brescia, Brescia, Italy

Introduction and Aims: Inflammation and increased ESA requirement are frequentlyassociated in patients on dialysis treatment. On-line Hemodiafiltration (Ol-HDF),putting together high levels of diffusion and convection could improve both conditions.However, it is still not known which depurative component predominate indetermining this result. Aim of the study: to evaluate the role of convection anddiffusion on ΔB2M, ΔhsCRP, and ΔESA in Ol-HDF.Methods: 30 patients, 26 men, age 57±13 years, dialytic vintage 12-108 months, wereswitched from conventional HD to post dilutional Ol-HDF. At 12 months the effect ofOl-HDF on ΔhsCRP, ΔB2M, and ΔESA (U/Kg/sett) were evaluated. Other variablesconsidered: Body weight (BW), serum albumin (sAlb), Hemoglobin (Hb), Kt/V. Irontherapy and ESAwere administered IV according to the K/DOQI guidelines in order tomaintain TSAT between 20-40%, serum ferritin between 150-500 ng/ml, Hb between11-12 g/dL. Qb, treatment time and Qd remained constant. Ol-HDF was performedutilizing High-flux membranes 1.9-2.1 sqm. Ultrapure dialysate and substitution fluidwas employed in both HDF and HD treatments. Data are expressed as mean±SD.Paired t test, Mann-Whitney U test, simple and multiple regression analysis wereemployed for statistical evaluation.Results: Total convective volume (TCV) was 21.8±1.7 l/session Significant reduction ofhsCRP: (from 5.3±7.5 to 2.1±2.7 mg/dl; p<0.01), B2M (from 29.0±14.4 to 21.3±12.3 mg/dl; p<0.0001) and ESA (from 92±6 to 57±35 U/Kg/week; p<0.008). No significantvariations of Kt/V, Hb, BW, sAlb. A significant inverse correlation was found betweenTCV and ΔB2M (r: 0.74; p<0.0001), and TCV and ΔhsCRP (r: 0.41; p<0.02), nocorrelation between TCV and ΔESA. No correlation was found between Kt/V and ΔB2M,ΔhsCRP, and ΔESA. Multiple regression analysis with ΔESA as dependent variableshowed ΔhsCRP as the only significantly associated independent factor (p<0.008).Conclusions:Ol-HDF induces a long-term significant reduction of B2M and hsCRPconcentrations.The entity of reduction is directly correlated to the amount of TCV. Theobserved reduction in ESA requirement is associated to the reduction of inflammationand seems to be independent from convection.

SP446 THE ROLE OF LUNG ULTRASOUNDS, BIOIMPEDANCEANALYSES AND NATRIURETIC PEPTIDES IN THEASSESSMENTOF EXTRAVASCULAR LUNGWATER INHEMODIALYSIS PATIENTS

Laura Bozzoli1, Elisa Colombini1, Guido Ricchiuti1, Giovanna Pisanu1,Luna Gargani2 and Carlo Donadio11University of Pisa, Pisa, Italy, 2CNR, Pisa, Italy

Introduction and Aims: Background: Lung ultrasound (LUS) allows a non-invasiveevaluation of extravascular lung water (EVLW) through the analysis of B-lines in heartfailure patients. Natriuretic peptides are also measured to evaluate heart failure. Morerecently LUS has been proposed to assess EVLW in maintenance hemodialysis (MHD)patients. Total body (TB) and segmental thoracic impedance analysis (BIA) canestimate TB and lung water compartments.Objectives: 1) To evaluate the correlationbetween B-lines and TB and segmental thoracic hydration estimated by means of BIA;

2) To compare LUS and BIA data with natriuretic peptides dynamic changes,echocardiographic findings and inferior vena cava (IVC) diameters; 3) To evaluate theusefulness of a comprehensive antero-lateral and posterior LUS scanning.Methods: Thirty-three adults MHD patients (24 M and 9F), were examined. LUS(Fig.1), TB and thoracic BIA, natriuretic peptides and physical examination wereperformed immediately before and after a dialytic session. Echocardiography and IVCanalysis were performed in the interdialytic period.Results: A higher number of B-lines was found in lateral and posterior lung segments(Fig.2). Also the dynamic changes in B-lines were not homogeneous. The number ofB-lines, and their decrease with HD, were correlated with TB and thoracic BIA.Thoracic and TB fluid compartments and their variations were correlated, suggestingthat the increase in lung fluids was driven by the increase in TB fluids. BNP andNT-proBNP, which were markedly increased due to severe renal dysfunction, were stillcorrelated with LVEF and with thoracic BIA. The residual number of B-lines after HDwas correlated with LVEF and with BNP. No correlation was found between IVC, leftatrium dimensions, diastolic dysfunction, physical examination and B-lines.Conclusions: A comprehensive LUS examination is advisable. LUS together with BIAand BNP, at the end of dialysis, may have a complementary role in the assessment ofpulmonary congestion and allow to unravel the pathogenesis of EVLW increase inMHD patients.

SP447 CLINICAL EVALUATION OF NEWHIGH CUTOFF MEMBRANE

Antonino Sidoti1, Maria L Lusini1, Marina Biagioli1, Paolo M Ghezzi2, Luisa Sereni2,Marialuisa Caiazzo2 and Giuseppe Palladino21Nephrology Unit, Poggibonsi, Italy, 2Bellco srl, Mirandola, Italy

Introduction and Aims: The uremic syndrome is characterized by the retention ofvarious solutes that would normally be excreted by the kidneys. Uremic toxinsrepresent an heterogeneous group of substances that includes organic compounds andpeptides both in their “native” and modified form, the latter by post-translational

SP446

SP446



at Utrecht U



ownloaded from


changes.Standard HD membranes are unable to depurate solutes with MW greaterthan 18 kDa;. The aim of the study was to evaluate the performance of a new, morepermeable dialytic membrane (Synclear 05) in terms of extraction capability ofmiddle-high MWmolecule useful for uremic toxin removal.Methods: Twenty ESRD patients (11 M), were enrolled for a prospective, crossoverstudy in order to compare the extraction capability of two membranes used in SUPRA(Synclear 02) and KIDNEY (Synclear 05) therapies.Serum samples and pre-cartridgeultrafiltrate (UF) were collected at the beginning and at the end of each middle weeksession Plasma and UF samples were used to determine, β-2 microglobulin (β2M), freelight chain (FLCs) κ and λ, Interleukin-6 (IL-6), α-1 acid glycoprotein (A1AG1),Albumin and Immunoglobulins G (IgG) levels. β2M, FLC κ, FLC λ, A1AG1, Albuminand IgG were determined by nephelometric assays (BNII, Siemens HealthcareDiagnostics, Tarrytown, NY, USA); IL-6, were evaluated by Solid Phase SandwichELISA (Quantikine ELISA kit, R&D System, Minneapolis, MN, USA).Results: Plasma and UF levels were evaluated both at the start and at the end of eachtreatment. Pre dialysis levels were not statistically different (data not shown). Theextraction capability was determined as percentage ratio between UF and plasmaconcentrations both at the start that at the end of session and then averaged.In order tocompare the extraction capability and, consequently, the Molecular weight Cut off(MWCO) of the two membranes, a wide range of MWmolecules were tested.Statistically significant differences between SUPRA and KIDNEY extraction capabilitieswere found for FLCs κ [23 KDa] (40 ± 11 vs 65 ± 23%, p<0,0001), IL-6 [24 KDa](49 ±10 vs 57 ± 12 %, p<0,05, FLCs λ [46 KDa] (19 ± 5 vs 28 ± 8 %, p<0,0001), A1AG1 [43KDa] (6 ± 2 vs 12 ± 5 %, p<0,0001), Albumin [66 KDa] (3 ± 1 vs 7 ± 3 %, p<0,0001)and IgG [150 KDa] (0 vs 2 ± 1, p<0,0001). In the figure are reported the sieving curvesof the two membranes calculated with the above data, compared with a standard HighFlux membrane.Conclusions: The results of this study demonstrate that, compared to Sync 02, Sync 05membrane offers a higher permeability to middle MWmolecules, and possess aMWCO higher than Sync 02 and the other standard HDF membranes.

SP448 EXAMINATION OF NEUTROPHIL INTERACTION ON NEWPOLYSULFONE (PS) MEMBRANE NV

Tadashi Tomo1, Kaede Ishida1 and Takeshi Nakata11Oita University Hospital, Yufu, Japan

Introduction and Aims: In hemodialysis, the platelet is thought to be activated by thecontact of an artificial material and blood, various factors including PDMP areproduced, and it leads to the arteriosclerosis progress through the interaction of theactivation of the white blood cell and the vascular endothelical cell. It has been reportedthat The new PS membrane “NV” has been reported to clinically control the plateletactivation by introducing new hydrophile polymer, and to have a high anti-thrombus.

We examined radical generation of neutrophle and the stimulation of cytokineproduction by the NV membrane in vitro.Methods: The conventional type of PS membrane “CX”made by Toray was assumedto be a contrast. New blood from healthy volunteer person was circulated for one hourat 37C by taking out the hollow fiber of CX or NV and making the pencil type dialyzer.After circulated, the radical generation of the neutrophilic leukocyte or the plasma wasmeasured by the chemiluminescence method. IL-6 and pentraxin3 of the plasma weremeasured with ELISA.Results: The amount of the radical production in NV (11±11 RLU) was significantlylower than in CX (22±22 RLU)(p=0.999). IL-6 production in NV (33±33 pg/mL) wasalso significantly lower than in CX (44±44 pg/mL) (p=0.04). The production ofpentraxin3 in NV (55±55 ng/mL) was also significantly lower than in CX (66±66 ng/mL) (p=0.03).Conclusions: The new membrane NV decreased the radical production of theneutrophilic leukocyte, and had the possibility that dialysis patient's inflammation canbe controlled by suppressing the cytokine production.

SP449 COMPARISON OF CONVECTIVE VOLUMES AND DIALYSISDOSE GENERATED BY DIFFERENTAUTOMATED DEVICES INONLINE HEMODIAFILTRATION POST-DILUTION

Didier Hamel1 and Thibault Dolley-Hitze11AUB Santé, Saint-Malo, France

Introduction and Aims: Since ESHOL study (Maduell F., JASN 2013) was published,OnLine HemoDialFiltration Post-Dilution (OL-HDF) has become the gold standard toperform chronic hemodialysis. Due to this study, total convective volume (=ultrafiltration rate + substitution volume) is also wished to reach 25L or more persession to improve patients’ survival. In our dialysis center, above 90% of the patientsbenefit from HDF realized with 4 types of HDF machines (AK-200 ULTRA, ARTIS,5008 and 5008-CORDIAX) for which automatic devices control substitution volumes(Ultra-control® for AK200-Ultra and ARTIS, Autosub® for 5008 and Autosub plus® for5008 CORDIAX). We wondered if these machines and devices could deliver the sameperformances, consequently we decided to conduct a retrospective and monocentricstudy to compare total convective volume and also dialysis dose.Methods: Univariate and multivariate analyses were conducted on patients who hadundergone at least ten 4-hour OL-HDF post-dilution sessions during the first 9months in 2013 and models used only the first 70 sessions. All patients had completedata for all covariates that were designated as either case-mix factors (age, gender) ordialysis-specific factors (weight, biological and dialysis parameters: substitutionvolume, ultra-filtration rate, on line dialysance (K and Kt) and biological Kt/Vrecorded on DIALOG7 Software) that hence were time-dependent. Associationsbetween total convective volumes or Kt parameters and device type were estimated byusing generalized linear regression (SAS PROCMixed; SAS Institute, Cary, NC) withan unstructured covariance structure. Before model development, we verified Gaussiandistributions of continuous variables. To verify model assumptions, we performedroutine regression diagnostics that assessed normality, linearity, homogeneity ofvariance, and influence. Model results are summarized by using parameter estimatesand 95% CIs. All P reported are 2 sided, and statistical significance is defined as P lessthan 0.05. All statistical analyses were performed using SAS (version 9.3).Results: 3446 sessions from 62 patients (40 male) aged 74 ± 13 years were analyzed(1997 sessions with 5008, 542 with 5008 CORDIAX, 707 with AK200 and 200 withARTIS); 4 different hemodiafilters have been used during the study period (Elisio®, TSSL®, VitaPES® and FX100®). The choice of filters did not depend on dialysis machine.Total convective volumes (least squared mean ± SEM) were 27.02 ± 0.24, 27.26 ± 0.25,26.58 ± 0.25 and 26.45 ± 0.29 for 5008, 5008 CORDIAX, AK200 and ARTISrespectively; Hochberg’s adjusted p-values were all statistically significant for pair-wisecomparisons except between AK200 UC ARTIS (p = 0.9889) and between CORDIAXand G5008 (p = 0.1252). Online Kt figures were quite similar with least squared mean± SEM being 58.61 ± 0.45, 58.08 ± 0.47, 52.86 ± 0.48 and 51.41 ± 0.54 for 5008, 5008CORDIAX, AK200 and ARTIS, respectively; all p-values statistically significant forpair-wise comparisons. Multivariate analyses confirmed these results when comparing5008 or 5008 CORDIAX to AK200 or ARTIS.Conclusions: These retrospective and monocentric datas show lower total convectivevolume for patients treated with Artis dialysis machine® AK-200 ultra® compared to5008® and 5008® Cordiax but all the convective volumes obtained were above 25l. Dialysis dose were also worse when treatment is delivered by Artis® or by AK-200®but remain above current recommendations.

SP447



at Utrecht U



ownloaded from


Documents

Haemodialysis Techniques and Adequacy 1