ASAlO RENAL ABSTRACTS

DETERMINANTS OF VASCULAR ACCESS FLOW (Q4).A, T.

Lubkowski, A. Yu, S. Frinak . University of West Virginia, Morgantown,
WV & Henry Ford Hospital, Detroit, MI.
Absolute value of QAand AQAover time are major determinants of access
patency and predict the likelihood of access events.(JASN 1999). However
Q A in stable accesses may change in response to systemic hemodynamics.
We examined the effect of mean arterial pressure (MAP), cardiac output
(CO), and segmental resistances (R) on Access Flow. Q A and CO (Limin)
were determined by TransonicTM ultrasound dilution. Static intra-access
pressures (mm Hg) at the arterial segment (AS) and venous segment (VS)
were determined with the access unoccluded and occluded (0).During ac-
cess occlusion. the AS pressure was equated to arterial pressure (MAPo)
and compared to the cuff (C) pressure (MAPc) while the VS pressure re-
flected venous pressure (VP). Total and segmental vascular resistances (mm
Hg-miniL) were calculated as APIQ. We studied 59 AV grafts and 36 native
fistulae (AVF) with measurements on 2 or more occasions in 44 grafts and
25 AVF. Results are mean 1 SEM with p <0.05 accepted as significant.
Systemic and access pressures are shown in table below:
Parameter MAPc MAPo AS VS VP
Access type Catheters Obstruction Surg Remo Prior Buried in-situ
Graft(I55) 10411* 99*2 58-t2* 4011* 1210.5 (n) n (%) eq ved Haemo DaysJAv.
AVF (78) 90+2 95i2 3112 253 1010.8 n n % (Range)
Bland-Atman plot of the difference between MAPc and MAPo vs. their
mean showed no effects in either access type. Flow and resistances were. BuriedTC 18(33) 12 2 42 46.85
Parameter QA CO Access AS vs (54) -
(7 152)
Access type (R) (R) (R) MPC (39) 4(10) 2 1 33 88.15
Graft(155) 0.9110.03 5.1*0.1 I1515 54&* 37+3*
(17- 448)
AVF (78) 0.91+0.05 5.510.2 11016 87i6 16*3
Non-access systemic R was lower in patients with AVF that those with TC (88) I5 (17) 2 5 48 None
grafts (2611 vs. 3OkI). Multivariate regression showed that in grafts Q A
correlated positively with MAP and CO and negatively with VP, AS R and
systemic R. In AVF, only the effects of CO and AS R were significant.
Conclusions: Hemodynamics of grafts &AVF differ. Cardiac output and
the arterial segment resistance are the chief determinants of Q A in both ac-
cess types. Increases in Q , with MAP in grafts need to be considered when
evaluating AQ, as part of the trend analysis for detecting dysfunction.

ACCESS SURVEILLANCE (AS) AND PREVENTION OF EXTRA- TO INTRACELLULAR FLUID RATIO IN

THROMBOSIS IN CHRONIC HEMODIALYSIS NORMOHYDRATED AND OVERHYDRATED SUBJECTS
PATIENTS
Access clotting is the leading cause of morbidity and Bioimpedance analysis is a potential technique to measure fluid
hospitalization in ESRD patients. Without the benefit of distributions in the body. A method of whole body bioimpedance
equipment to measure dynamic flows, we employed clinical
(WBIA) based dry weight estimation for dialysis patients is shown,
criteria to determine when radiographic intervention (RI) was
which accounts for the individual nutrition and training status of each
required to prevent thrombosis. We prospectively studied 14
patients (pts.) over one year who fulfilled the criteria of : high patient.
venous pressure (HVP) 120 mmHg at Qb=200 using 15 g. Extracellular and intracellular fluid volumes (ECF and ICF) of healthy
needle; swollen arm (SA); reduced blood flow (BF) and difficult controls (n=34, age 25-44 y) were measured with a multifrequency
cannulation (DC). There were 25 RI in these 14 patients : H W WBIA analyzer (4200, XITRON Technologies Inc.). ECFACF was
accounted for 14 RI; SA in eight studies; reduced BF in two variable, but correlated with the deviation of measured TBW from
studies and one study for DC. Results showed that venous TBW according to the Watson equation. Hereby the deviation of
stenosis (VS) requiring angioplasty was found in 23 of the 25 bioimpedance based TBW and Watson-TBW is affected by the
studies. Stents were placed in one patient. Two studies were
individual nutrition and training status. After establishing this
negative for VS. There were only two thrombosis in these 14
correlation for healthy individuals, the ECMCF ratio of chronic
patients (14.2 thrombosis per 100 pt. years). By comparison, 23
other patients not in the AS study had 40 episodes of thrombosis hemodialysis patients (n=31, age 31-82y) was measured and compared
during the same period (173 thrombosis per 100 pt. years.). We to the curve of the healthy population. Pre-dialysis patients who were
conclude that HVP and prompt RI is a cost effective and efficient considered clinically overhydrated showed higher ECF/ICF ratios than
method to prevent access thrombosis in these patients until a healthy controls with comparable nutrition and training status.
more easily readable method is found. Conclusion: For dry weight assessment by use of bioimpedance
analysis it is essential to account for the individual nutrition and
training status of each patient.
A. Siddiaui and C.R. Schleifer; Division of Nephrology,
Lankenau Hospital and Renal Care Group East, Wynnewood, PA W. Kleinekofort, C. Rode, M. K r h e r , V. Wizemann*
Fresenius Medical Care AG, Bad Homburg, Germany
'Georg Haas Dialysezentrum, Giessen, Germany

206
 ASAIO RENAL ABSTRACTS 207

IMPROVED STATIC PRESSURE MONITORING FOR CLINICAL RESULTS OF ISOPROPYL ALCOHOL USE AS AN
ANTIMICROBIAL IRRICANT FOR USE IN A NEW
DETECTION AND CLASSIFICATIONOF ACCESS STENOSIS
SUBCUTANEOUS VASCULAR ACCESS DEVICE FOR
HEMODIALYSIS. John Moran, C. David Finch, Tewksbury, MA
According to the DOQl guidelines for vascular access, prospective
Preventmg vascular access complications remains one of the greatest
monitoring of static venous dialysis pressures can be used to detect challenges for dialysis providers and clinicians. A new subcutaneous
outflow stenoses. However, with this method it is not possible to access device (Lifesite@Dialysis Access System, Vasca, Tewksbury.
identify stenoses which are localized between the arterial and venous MA) has been designed to offer an alternative. The device consists of a
titanium alloy valve and attached silicone cannula. It is lmplanted
dialysis needle. below the clavicle dunng an mpatient or outpatient procedure, with the
9 dialysis patients (7 fistulae, 2 grafts) with severely stenotic access single lumen cannula placed in the central venous circulation for
hemodialysis (HD). Two systems are implanted, one for draw and one
were studied. Both arterial and venous static pressures were monitored for return. The LifeSite utilizes an mternal pinch clamp mechanism that
and corrected for hydrostatic offsets. Height differences between access is actuated with a 14-ga dialysis needle. When the needle is mserted,
site and fluid level in the extracorporeal circuit were found to cause a the internal metal pinch clamp opens, allowing fluid flow; when the
needle is removed, the mechanical pinch clamp closes and flow stops.
pressure change of 0,75 mm Hg per cm. Patients with venous outflow Access to the Lifesite System is obtained utilizing the buttonhole
stenoses showed static arterial and venous pressure ratios >0,5.in case technique (insertion of the needle in the same location each tune). The
advantages of the buttonhole technique mclude easier and quicker
of intra-access stenosis between the needles the venous pressure ratio
cannulation, less pain and patient anxiety, reduced scarring, and the
was normal (<0.5), but the arterial pressure ratio was increased >OS. ability to dismfect the valve, smus pocket and smus tract with an
After static access pressure measurement all patients were referred to antmicrobial irrigation solution. Problems typically associated with
implantable devices for HD relate to the presence of foreign bodies
angiography, with subsequent angioplasty of stenotic lesions or which must be accessed transcutaneously 3 timesiweek or more. and
complete access revision. Following the intervention. access pressures the resulting mherent risk of infection. The design of the LifeSite
System allows for antmicrobial irrigation utilizing a 25-ga needle. This
returned to normal values.
slze needle does not actuate the system's internal metal pinch clamp,
Conclusion: Arterial and venous access pressure normalized by mean ensurmg that antimicrobial solution does not enter the circulation. We
blood pressure detects venous stenosis as well as stenosis between the report our experience utilmng up to Iml of Isopropyl Alcohol (IPA)
pre- and post-dialysis. With 27 patients totaling 1997 days of use, we
arterial and venous dialysis needle. To minimize the rate of access report no device-related infections. Patient tolerance has been excellent,
thrombosis both arterial and venous pressure should be monitored. with only one m 28 patients unwillmg to continue IPA irrigation. We
fmd the ability to utilize an antirmcrobial solution a superior design
characteristic of the LifeSite Device. IPA provides excellent results as
W. Kleinekofort, C. Rode, M. KrZmer. V. Wizemann' an inexpensive, safe and effective antirmcrobial agent.
Fresenius Medical Care AG, Bad Homburg, Germany
'Georg Haas Dialysezentrum, Giessen, Germany

COMPLICATIONS OF SECONDARY ON-LINE MEMBRANE IN-VZTRO EVALUATION OF HEMOLYSIS IN A NOVEL

PLASMAPHERESIS SUBCUTANEOUS ACCESS SYSTEM. John Moran, M.V.
The purpose of this retrospective study was to evaluate complications and Kameneva, B.M. Repko, H.S. Borovetz, P.F. Marad, Tewksbury, MA
adverse events associated with secondary on-line membrane plasmapheresis
(MP), such as plasma filtration (PF) and cryofiltration (CF). We performed The purpose of this study was to measure the degree of hemolysis, if
over 1,750 MP in 64 patients over the last 12 years in our institution. This is any, with the Lifesite@ Hemodialysis Access System (Vasca, Inc.,
the most MP procedures reported in the US. Twenty-five patients received Tewksbury, MA). The LifeSite System incorporates a smooth flow path
550 PF using ethylene vinyl alcohol (EVAL) hollow fiber membrane with that turns through a right angle. The LifeSite System was tested with a
pore sizes 0.01 to 0.04 pm to remove macromolecules with different 14-ga AV fistula needle as well as a 14-ga right-angled needle. The
molecular weights such as IgG, IgA, IgM, immune complexes, or LDL- right-angled needle offers ergonomic benefits over a standard AV
cholesterol. Thuty-nine patients received 1,200 CF with acrylic co-polymer fistula needle and is a desirable alternative.
(Venapor pleat) membrane with pore size 4.3 pm to remove cryoglobulin The LifeSite was evaluated with a standard 14-ga AV Fistula needle
which precipitated at 4'C temperature. and 14-ga right-angled needle and the results compared to a 10F Tesio-
Two patients on PF and 3 patients on CF experienced anaphylactoid Cathw (MedComp) and a 15-ga AV Fistula Needle Set (JMS CO.,
reaction while on concomitant ACE-inhibitor (ACE-I) therapy. LTD). Hemolysis was evaluated In-viho using fresh bovine blood
Anaphylactoid reactions were mild and not life-threatening. The subsequent pumping through a circulahng loop with the test device incorporated.
reactions were avoided by either discontinuing or temporarily holding the The blood was recirculated for 2 hours at 300 and 450 mlimin, which
ACE-I prmr to procedure. Some patients on concomitant ACE-I and MP did is the equivalent of 33 hours of hemodialysis treatment for a patient
not develop anaphylactoid reaction. Determination of vasodilatory with a 5-liter blood volume. Sublethal blood damage was examined by
metabolites, 6-keto PGFla, PGD-M and methyl histamine in plasma and using the hemorheological assay.
2,3 dinor 6-keto PGFla and methyl histamine in urine showed they were The testing demonstrated that the LifeSite System, when used m
only marginally elevated. Two patients developed sepsis, later found to be conjunction with a 14-ga AV fistula needle or a 14-ga right-angled
unrelated to MP. Other complications include vascular access, hypotension, needle induced the smallest changes in hemorheological parameters,
citrate-toxicity, filter plugging, and increased transmembrane pressure and produced less hemolysis, than either the 1OF Tesio-Cath or the 15-
(TMP) requiring backwash (if EVAL). MP requires less blood flow than PE ga AV Fistula Needle Set. The degree of hemolysis was not clinically
and in most cases, antecubital venous access is sufficient, and there is no significant with any of the tested devices.
need for replacement fluids such as FFP, albumin, and hetastarch. MP is a
closed procedure and the chance of contamination and infection is less than
PE. There is also less citrate required.
In conclusion membrane plasmapheresis is safe and has less
complications regarding vascular access, citrate toxicity, and infection
compared to plasma exchange. Side effects from replacement fluid such as
allergic reaction to FFP. albumin. and hetastarch is avoided.
SimY GA, siami FS '
Vanderbilt University Depaxtment of Medicme, Department. of Veterans
Affairs Medical Cent&, Nashville, Tennessee, USA
 208 ASAIO RENAL ABSTRACTS

ON-LINE TEMPERATURE MONITORING IN CONTINUOUS PREDICTIVE VALUE OF VENOUS PRESSURE FOR

HEMODIALYSIS (CVVHD): A FEASIBILITY STUDY. G m THROMBOSIS: STATIC, SLOW OR HIGH
Shahriar Rahmati, Laura Rosales, Claudio Ronco, Richard Amerlmg, FLOW.
Allen Kaufman. Alan Dubrow, Nathan W. Levm, MD Monitoring outlet venous pressure (OP) in arteriovenous
Renal Research Institute, New York, NY, USA
Thermal Balance in patients undergoing CVVHD has never been grafts (AVG) has been recommended to improve graft sur-
studied. The blood temperature monitor (BTM) is a device Implemented vival. Corrected intra-access pressure (IP) to mean blood
in the Fresenius 2008H machine, allowing for the calculation of access pressure (MAP) ratio has been considered to be superior
recirculation, thermal energy balance and body temperature control. BTM to dynamic pressure (QB 200-250 d m i n ) in predicting
has only been used at blood and dialysate flows typical for conventional
hemodialysis. In this study we evaluate the possible use of the BTM in
AVG malfunction. More recently, monitoring OP at
CVVHD. The study was conducted in vitro using a modified 2008H higher QB(4OOmVmin) has been reported to be better
machine. The machine was calibrated at 37.0"C with dialysate flows (Qd) predictor of access failure (AJKD 34: 212, 1999).
of 500 and 100 mumin. Accuracy of temperature (T) setting was then However, the IP/MAP ratio were not compared to
checked at Qd of 300,200,100 mumin. Results displayed the best OP at varying QB. This study evaluates OP at QBOf 0,
accuracy at 200 mumin. At 100 d m i n wide fluctuations could be 50, 250 and 400 mumin. Pts were followed for 6
observed (f12%). At 300 ml'min, actual T never reached the set T
months and 41 pts without thrombosis (GI) were com-
(average delta S -0 6°C). BTM could not be used in T control mode
because of delay in the effects of thermal boll utilized for recirculation pared to 24 who developed thrombosis (GII) OP (mmHg) at
measurement. T variations in blood occurred beyond the detection time of varying QB(Mean SEW are: *P<0.05 (GI VS GII)
the BTM (4 minutes). BTM could calculate automatically the total energy QB 0 50 250 400
balance durmg a CVVHD session (8 hours measurement time) and it GI 24B* 4 2 s* 12W4 2236
could be used in thermal enerzy control mode (TECM) to maintain a
constant flux of energy across the hemodialysis membrane with values
GII 3633 56k4 1489 23433
varying from - 100 to + 100 Kjih). In conclusion, BTM can be used in Corrected IPiMAP was 0.41 vs 0.53 at zero QBand
CVVHD to measure thermal energy balance and to control the flux of O P W was 0.42 vs 0.53 at Q ~ 5 for 0 GI and GII respect-
energy delivered or withdrawn from the patient during treatment. This ively (P<0.05).O P W at higher QBwere not predictive.
features of the BTM is of great interest for the critically ill patient. These data suggest that OP/MAP at low QBmay be superior to
Pahents may frequently be byperthermic and may requlre a progressive
cooling to improve hemodynamic stability and prevent thermal damage.
OP at higher QBin predicting graft thrombosis. Determination
On the other hand, when large volumes of fluid are exchanged, the risk of of OP/MAP at QBof 50 d m i n may be easier than at zero and
a negative thermal balance is highly enhanced. For all these aspects. the does not require any special equipment or corrections.
use of BTM m the acute renal failure should be investigated and future C. Shah, R Raja, Albert Einstein Medical Center,
application of these monitors in CRRT machines may be advantageous. Philadelphia,PA 19141

OBSERVATION OF PD CATHETER NEAR THE

TITANIUM ADAPTOR
Shieevuki Kushihata', Akira Hirabayasil ,
Yoshitaka Sekiguchi', Noriaki Yorioka'
1.Hiroshima General Hospital, Hiroshima, Japan
2.Second Department of Internal Medicine,
Hiroshima University School of Medicine

Perforation of PD catheters near the titanium

adapter is one complication of long term PD . There
are various opinions about the cause of such
perforation, but it is unclear, and optimum
treatment has not been established.
We performed electron microscopy of PD
catheters removed from the 5 patient who stopped
the peritoneum dialysis after over 7 year.
Deterioration of t h e silicon catheter material itself
was not found, but there were small scratches a t
the point of contact with t h e titanium adapter.
This suggested that traction on t h e catheter could
cause damage leading t o perforation near the
1 Flow velocity Qb = 250 mllrnin Qb = 350 rnllrnm titanium a d a p t e r .
Therefore, if we find PD catheter perforation
Average I 0 51 1 0.51 1 0 13 1 073
near the titanium adapter,the catheter should be
Central region I 0 55 1 0.58 1 0 77 1 079
Penpheral region 1 0 48 I 045 ] 0 68 I 066 cut off near the perforation and titanium
adapter should be changed. Also, patients should
We may conclude that the cartridge design IS adequate and no
be instructed to carefully protect the area around
channelling effects could be detected m the blood compartment The
flow distnbution is slightly affected by changes in flow rate and the titanium adapter with gauze as well to properly
hematocrit showmg an optimal utilization of the available surface for fix the catheter.
molecule adsorption
 ASAlO RENAL ABSTRACTS 209

ANTIBODIES TO PROTHROMBIN AND FACTOR V CORRELATE EARLY APPLICATION OF ENDOTOXIN ADSORPTION

WITH PTFE GRAFT THROMBOSIS (PMX-20R) TO HIGH RISK PATIENTS WITH SEPSIS
By Jeffrey J. Sands, Sybil Nudo and Thomas L. Ortel
We have described a relationship between antibodies to topical Tamotsu T o i i a Yuichi Sato, Naoko Yoshimura, Akihitc
bovine thrombin preparations (BTP) and PTFE graft thrombosis. Sannomiya, Kazuhiko Tsuji, Tomonori Kawase, Ken Utsumi
BTP can induce antibodies to multiple coagulation proteins, including Tohru Murakami, Ichiro Koyama, Ichiro Nakajima, Shohe
prothrombin (FII), factor V (N),and pz-glycoproteinI (PzGPI). These Fuchinoue, Tetsuzo Agishi
antibodies are associated with an increased thrombotic risk. Department of Surgery 111, Tokyo Women’s Medical
88 hemodialysis patients (44 AV fistulas [AVF], 28 PTFE grafts University, Tokyo, Japan.
[PTFE], 16 cuffed catheters [CC]) were studied for the presence of PURPOSE: Despite the many advances in the treatment of
antibodies to human and bovine FII, bovine FV (bFV), and human sepsis, mortality rates in septic patients remain high. We
P2GPI. Antibody levels were measured by ELISA. These results have been treating septic patients, especially in high risk
were correlated with the patients’ access history. patients, using a polymyxin B immobilized fiber column
51 patients (58%) had elevated antibodies to at least 1 protein. 41 (PMX-20R). The aim of this study was to evaluate the
patients had elevated anti-human FII antibodies (17 PTFE [60.7%]; effect of direct hemoperfusion (DHP) using PMX-20R on
14 AVF [31.8%; p<0.05 cornpared to PTFE]; 10 CC [62.5%]); 17 had sepsis after emergent surgery.
elevated anti-bFV antibodies (7 PTFE [25%], 5 AVF [ I 1.4%], 5 CC PATIENTS: Between January 1999 and December 1999,
[31.3%]), and 8 had elevated antibodies to PzGPl(2 PTFE [7.1%], 5 five patients, aged 55 to 79 years ( 6 9 t lo), who had sepsis
AVF [I 1.4%], 1 CC [6.3%]). Patients with PTFE grafts had elevated due to peritonitis were treated with DHP after emergenl
antibodies most frequently (21 [75%] vs. 19 AVF [43%; p<O.Ol
compared to PTFE] and IICC [68.8%;p=O.O8]). 26 of 27 patients
surgery at our institute. Four of the 5 patients had been on
(96.3%) with antibodies to BTP had elevated antibodies to one or dialysis treatment due to chronic renal failure, and one
more of these proteins compared to 25 patients without anti-BTP patient had received renal transplantation 3 weeks before.
antibodies (41%; p<O.Ol). 12 of 13 patients with PTFE grafts and The 5 patients included 2 with colonic perforation, 2 with
elevated antibodies (92.3%) had a previous thrombosis, compared to ruptured suture after colectomy, 1 with duodenal
9 of 15 without antibodies (60%; p<0.05). The number of prior perforation. Acute physiology and chronic health evaluation
thromboses and mean thrombosis rates were also higher in PTFE (APACHE) I1 score ranged from 12 to 20 (17 = 3).
patients with antibodies (1.24 vs. 0.14 prior thromboses, pe0.01; RESULTS: DHP using PMX-20R, for 2 hours at a flow
42.67 vs. 6.44 thromboses/lOO patient years, p<0.05). Six of 7 rate of 100 mumin, was performed twice in each patient
patients (85.7%) with elevated anti-bFV antibodies had a previous within 3 days after surgery. After completion of DHP,
thrombosis compared to 7 of 21 without antibodies (33.3%; ~ ~ 0 . 0 5 ) . circulatory state and oxygenation index (from 331 t 8 6 to
In conclusion, hemodialysis patients with PTFE grafts have an 3 9 0 5 101) were improved in all patients. Four patients
increased incidence of elevated antibodies to FII, FV, and pzGPI.
These antibody positive patients had a higher incidence of graft survived, one died due to multiple organ failure (MOF).
thrombosis. Further studies are necessary to see if limiting exposure CONCLUSIONS: Our results suggest that early
to BTP will decrease the incidence of these antibodies and PTFE application of PMX-20R may prevent MOF and improve the
graft thrombosis. patient survival with sepsis after surgery.

CLTNlCAL APPLICATION OF THE ENHANCED INTERNAL SELF-EXPANDING NITINOL STENTS FOR URETER;
FILTRATION DIALYZER EXPERIMENTAL STUDY
As shortcomings in conventional hemodialysis have become The use of stents to maintain the lumen is widelyspread
apparent, hemodiafiltration (HDF) seems to be an option. However, in many fields including coronary artery, bronchus, biliary tract
HDF has not gained a wide acceptance because of its complexity and and esophagus. The aim of this study was to evaluate the
expense. Meanwhile, OUT previous analysis in dialyzer modification
showed that a reduction in the fiber diameter as well as an increase in efficacy and safety of self-expanding stents in the canine ureter.
the fiber length and density ratio increase the total water flux across Actively expanding nitinol Accuflex stents (diameter 8-10 mm,
the membrane (i.e. internal filtration), thus improving convective length 4-6cm, Boston Scientific Corp., MA, USA) were used
elimination of high molecular weight substances and achieving HDF- for this experiment. Under general anesthesia, through a
like treatment without necessitating reinfusion fluid. midline-abdominal incision, stents were inserted directly into
In order to evaluate the effects of the modification, four types of the ureter through a small incision in 8 mongrel dogs (20-26
celliilose triacetate dialyzers with different fiber diameters, lengths kg). Each animal had unilateral stent insertion. The observation
and density ratios (as below) were manufactured by NISSHO (Osaka, period was 6, 23 days, 2, 5 (n=2), 6 (n=2) and 7 months. The
Japan) and cmplvycd in four slablc cnd-slagc rcnal discasc palicnts.
specimens were evaluated macroscopically and microscopically.
They received dialysis sessions with one of these dialyzers once
weekly for 12 weeks. Contamination in dialysate was carefully Macroscopically, all stents were patent. No urine leakage was
monitored. No adverse effects such as extracapillary leakage or observed in any of the animals. One dog at 5 months and
excessive intracapillw clotting were observed. No significant another at 6 months had hydronephritis with
increase in B2-microgloblin (B2M) clearance was seen. but a tendency hyperepithelialization inside the stent, while all others had no
was nolcd that B2M clcarancc incrcascd grcally in Ihc dialymn with dilatation of the renal pelvis. On day 6, the inserted stent was
175pm of fiber diameter than in those with 200 pm on increasing the widely patent with clots in the mesh. A smooth shiny lumen
b l d tlow through the dialyzers. was obtained in one dog at 5 months and another at 6 months.
Although ultrapure water is necessary, the enhanced internal- Microscopically, at 6 days, denuded epithelium and bleeding in
filtration dialyzer is readily applicable without any further equipment the ureteral wall were observed. At 5 months, epithelialization
and can be an altematne to conventional hemodralj7er.
of the lumen was complete and the inserted stent was
incorporated in the wall. It was concluded that self-expanding
nitinol stents can be used for ureteral stenosis.
Tomizawa Y, Goya N, Nishino S, Ishikawa N, Toma H,
LDj I 200 I 216 I 61 -!I Nishida H, Endo M and Koyanagi H
Mineshima M”, Tshimori I”, Agishi T”,Masuda T”, Hattori Tokyo Women’s Medical University, Tokyo, Japan
H“
1) Kidney Center, Tokyo Women’s Medid University
2) NISSHO
210 ASAIO RENAL ABSTRACTS

PERITONEAL ULTRAFILTRATION(UF)WITH EXTRANEALB PRETRANSPLANT PROGNOSTIC TESTING OF DAMAGED

FOR OVERHYDRATION DUE TO CONGESTIVE HEART KIDNEYS DURING EMS PERFUSION
Bart M Stubenitsky, Maurits H Booster, Lauren Brasile, Dorian
FAILURE (CHF) AND NEPHROTIC SYNDROME Araneda, Carl E Haisch and Gauke Kootstra
Drama M (MD), Mesquita M (MD), Wens R (MD), Paciorkowski F 'Surgery, azM, Maastricht, Netherlands
(MD), Keller J (RN), Dumortier F (RN), Rossez N (RN) CHU Purpose- Prognostic testing represents the basis for expansion of the
Brugmann, Brussels, Belgium donor pool with ischemically damaged kidneys. We evaluated if actual
Congestive heart failure unresponsive to maximal drug restoration of renal metabolism by ex vivo warm perfusion could be
used to predict the status of an organ before transplantation. Using a
treatment has been known to benefit from UF canine autotransplantation model, we messed the potential of
We wish to report our experience with pentoneal UF using prognostic testing during ex vivo warm perfusion with EMS technology.
ExtranealB Methods- Kidneys were subjected to 30 minutes of warm ischemia
Eight patients with severe CHF ( " H A stages 111 in 3 and IV in 5 (WI) followed by 24 hours of static storage in ViaSpan at 4C. Following
WI and static storage the kidneys were transitioned to three hours of
with ejection fractions under 30% in all), and one with diabetic EMS perfusion at 30C. During this period parameters indicative of renal
nephrotic syndrome, unresponsive to dietary and drug treatments metabolism and function were employed to predict outcomes
(ACE-I, diuretics, digitalis, vasodilators) were offered to try prospectively. Parameters included measures of innate metabolic
ExtranealB therapy after informed consent Three of them had normal capacity, perfusion characteristics, oxidative metabolism and the
renal function while six had various degrees of renal impairment condition and barrier function of the vasculature. A Viabilty Index (VI)
was calculated as the sum of the four parameters mentioned above.
(creatinine clearances ranging from 25 to 60 mlimin ) The causes of Results- A VI in the range of 2.5 to 2.9 was associated with mild
CHF were diabetic and ischemic cardiomyopathies in 1 and 7 acute tubular necrosis (ATN) (3 dogs); mean peak serum creatinine
patients, respectively Depending on the level of renal function (< or value was 3.5mgidL and mean time to normalization ofthe serum
2 50 ml/min) and on the degree of overhydration, one or two 2 L chemistries was 7.3 days. Similarly, a V I in the range of 2.2 to 2.45
was associated with a moderate ATN (4 dogs); mean peak serum
exchanges were applied wlth dwells of 24 or 12 hrs, respectively creatinine value was 5.6mgIdL and a mean time to normalization of the
Resulting daily ultrafiltration vaned from 750 to 1400 ml, achieving a serum chemistries was 14.3 days. When the calculated VI was in the
weight loss of 4 to 6 3 kg during the first week Thereafter, UF was range of 1.5 to 2.00, the ATN was severe. In the two dogs with a severe
adapted to the desired weight loss (maximum 20 kg over 1 month) ATN, the mean peak serum creatinine was I0.3mg/dL.
In one case the calculated VI was negative. The dog was symptomatic
No significant hypotension occurred despite a basal BP no higher than of uremia and was euthanized on day 7 post-transplant with a serum
115160 mm Hg Four patients progressed to more severe renal failure creatinine of 1Z.Smg/dL.
needing 3 or more glucose exchanges, 3 patients did regain enough The results indicate the possibility of utilizing warm temperature
compensation to stop UF, and one received a successfull heart perfusion to evaluate kidneys prior to transplantation. The Vl used was
transplantation after 29 months There were 4 episodes of peritonitis able to identify and classify the severity of the ATN and the occurrence
of primary non-function, offering a sensitive assay for prospective organ
over 100 patients-months Hospitalisation rate was reduced from testing.
I Simonth in the 6 months before UF to 0 2/month over the mean 12
months of follow-up There were no side effects of ExtranealB
In conclusion, UF with ExtranealB is practical (1 or 2 manipu-
lations per day), efficient and well-tolerated It can improve possibly
survival and certainly quality of life in patients with severe CHF or
nephrotic syndrome

EMS PERFUSION OF ISCHEMICALLY DAMAGED KlDNEYS PRIOR NONINVASIVE CONTINUOUS PHOTOMETRIC UREA
TO TRASSPL.4NTATlON REDL'CES REPEWUSION INJLRY DETERMINATION: A NEW TECHNIQUE. Robert R. Steuer and
Songbiao Zhang. In-Line Diagnostics Corp., Riverdale. Utah USA.
RLl Stubenitsky', MI1 Booster'. L Brasile', D Araneda'. CE Haisch'
and CI Kootstra'
'Surgery. arM.Maastricht, Netherlands: '. Breonics. Inc., Schenectady. To provide a real-time hernodialysis end point monitor, a multiple-
SY.United States; and 'Surgery, ECU. Greenville. NC, United States wavelength+& unique optical technique for noninvasive continuous
determination of Urea in a flowing bath is described. Two wavelengths
Purpose- Ischemia, both warm and cold, inhibit cellular metabolism. corresponding to Urea (L,)and water (h) absorptionbands were selected.
This inhibition ofmetabolism has been postulated to influence tne The transmittance (%TJ at h,was used as a reference to compensate the
se\enty of the repertbsion injury. Reestablishing renal metabolism prior baseline absorption from the bath while %Tuat L, represented the peak
to imphtation could potentially amcliorate posnransplant graft absorption of Urea. First, transninanceat these two wavelengths in a non-
function. Using a canine autotransplantation model combining both Urea circulating bath was recorded continuously through 1.2 mm chambers.
warm and cold ischemia the impact of EMS perfusion was evaluated Then small increments of a Urea stock solution were added to the bath at
given time intervals to achieve different Urea concentrations, prea].
.Methods- kidneys were exposed to 30 minutes of warm ischemia
followed b) 24 hours static storags in Viaspan at 4OC. Kidneys in the Stepwise changes in %T were monitored as [urea] changed. The
logarithmic ratio of these %T values was directly related to [urea].
control group were reimplantated after the period of hypothemic
preservation: kidneys in the experimental group were transitioned to 3 The slope and R' of the %T ratio in the concentrationrange of 0 0.4 -
hour5 of EMS perfusion at 30C prior to reimplantation. Contralateral g/dL were - 0.024 and 0.9955. The standard deviation of the points from
nephrectomy was perfomed before reperfusion of the preserved kidne). the regression line in ratio units was equivalent to 8.3 m@dL Urea detection
Posnransplant renal function was assessed by serum creatinine and sensitivity limit. While further improvements in stability and precision are
sun ival. still ongoing. it can be concluded that prea]can be determined
noninvasively and continuously in real time by using this new multiple
Results- EMS perfused kidncjs had lower 24 hours posmransplant wavelength optical technique.
serum crcatinine values than control kidneys (mean of3.1 mg/dL vs 4.0
mg dL). 'lhe sun,ival rate tor EMS perfused kidneys was 90 % (9 10) vs
73 90( 8 ~I I ) of the control kidneys
l'hese results pro\ ide evidence that ex VIVO restoration of cellular
metabolism reduces damage seen upon actual reperfusion tiom the cold.
thereby improving posnransplant graft function and survival.
 ASAIO RENAL ABSTRACTS 21 1

MEASURING VASCULAR ACCESS FLOW RATES (Qa) CLINICAL EXPERIENCE WITH ARTIFICIAL LIVER SUPPORT
WITHOUT LINE REVERSALS: A 2-STEP METHOD. IN CHRONIC LIVER FAILURE WITH ENCEPHALOPATHY
Robert R. Steuer and SongbiaoZhang. In-Line Diagnostics Corp., L. Kramer, A. Gendo, G. Funk, C. Madl, D. Fakenhagen', A. Gangl.
Riverdale, Utah,USA. Dept. of Medicine IV, Univ. of Vienna, Austria; 'Center of Biomedical
Technology, Danube-University, Krems, Austria.
Measurement of Q. b y saline dilution techniques can be used to To clarify the role of extracorporeal detoxification in advanced cirr-
assess vascular access function, but it is time consuming, requires hosis with hepatic encephalopathy (HE), we tested the hypothesis that a
line reversals and is inconvenient to perform during routine six-hour episode of extracorporeal detoxification but not conventional
hemodialysis. We propose a new approach without line reversals for treatment would lead to clinical and neurophysiological improvement.
measuring Q,. A 30-ml saline bolus is injected into the dialyzer After obtaining ethics committee approval, 35 patients with cirrhosis
venous line from which cardiac output, Qh, is determined. T h q the and HE grades I1 to IV not improving with standard treatment have
been enrolled. HE was assessed clinically and by sensory evoked
previously described AH technique is performed with Crit-Line by potentials, an objective parameter of cerebral function. In study A, 20
changing ultra filtration rate, Qf, &om 0 to 30 mumin for 4 minutes, cirrhotic patients were randomized to six hours of ongoing medical
but in the normal h e configuration. A differential equation solution treatment or, in addition, sorbent suspension dialysis (BioLogic-DTB).
for Q. was obtained as a function of Qh, Qf, and AHnomd(measured In study B, the newly-developed FPSA (fractionated plasma adsorption
at 60 seconds). This was determined retrospectively against a Q. system, Falkenhagen et al., Int J Artif Org 1999;23:81) has been
reference in 47 patients. evaluated similarly involving 10 treated and 5 control patients so far.
Results: with a Q h measurement error off 7%. Q. was determined Groups were comparable at baseline. Clinical grade of HE
to within f 225 &,in., R' = 0.805. These data suggest that reliable (means1SEM) did not change after sorbent dialysis (2.9i0.2 at baseline
routine measurement of Qa,without line reversals, can be done with vs. 2.910.4 following treatment, p=NS) but improved after FPSA
this new approach. Therefore, c a r d a c pulmonary recirculation, (3.5i0.3 vs. 2.410.5, p=0.04). There was no change in conventionally
treated patients (2.4i0.4 vs. 2.510.7, p=NS). Cerebral function as
CPR can be determined as well. measured by sensory evoked potentials improved after sorbent dialysis
am (N70 peak latency, 12817 vs. 1 I0110 ms, P<O.O5) and FPSA treatment
, ,
(1291-10 vs. lI41l1, p=0.05). Again, there was no change in standard
treatment (N70: 114i9 vs. 113i7 ms, P=NS). Bilirubin increased
(+23%) after sorbent dialysis and decreased (-24%) after FPSA. Plate-
lets dropped by 65% after sorbent dialysis and by 19% after FPSA.
In summary, these controlled studies suggest that neurophysiological
dysfunction in severe HE can be rapidly improved by extracorporeal
detoxification. The clinical stage of HE was ameliorated in the FPSA
group only. This suggests additional benefit from removal of albumin-
bound toxins. If biocompatibility can be improved, extracorporeal
detoxification may become a defmite therapeutic option in patients
with cirrhosis and severe encephalopathy.

HEMODIALYSIS (HD) AGAINST LIPID (LPD) AND EFFECT OF VIT-E MODIFIED DIALYZERS MEMBRANE
ALBUMIN (ALB) FOR HEPATORENALFAILURE (HRF). ON HUMAN GRANULOCYTES ADHESION MOLECULES
N.L.Manohar, St.Vincent's Med Ctr, Staten Island, New York EXPRESSION DURING HEMODIALYSIS (HD) PROCEDURE
Chronic renal failure induces a clinical state of
Both fat soluble & protein bound toxins accumulate in HRF
immunodeficiency, that is also depend on widely spectrum of dialysis
pts. HD against LPD (food grade soy bean oil) has been used membranes use in hernodialysis patients. Dialysis with complement-
to remove LPD soluble drugs such as Glutethimide in the activating cellulose membranes is associated with leukopenia which
1960's. Hence HD against LPD might benifit pts with HRF. has been related to an increase in adhesion molecule expression on the
A 52 yr old female with ethanol abuse, cirrhosis, worsening surface of circulating leukocytes. The expression of CD45+. CD4+8+.
hepatic & renal h c t i o n & increasing ascites & edema was CD4+8. CD18+1 Ib+, CD18+1 Ib-. and CD18-1 Ib+- adhesion
subjected to 2 high flux & one low flux HD sessions, without molecule on granulocytes was evaluated during HD in 10 patients
using flow cytometric analysis. The study protocol included the
drop in either serum bilirubin (SB) or international normalized measurement of molecules expression using Clirans RSI S(Terumo)
ratio (INR).Four sessions of ALB-HD on high flux dialyzer, cellulose membrane at the time"0" "phase-I", and after 10 dialysis
(1000 cc of 5 % ALB, single pass, in the middle of a conven- session "phase 2". After of 2 weeks of adaptation period with Vitamin
tional 4 hour HD), lead to only - 1-2 % intra dialytic drop in E modified dialysis membrane (ExcerfaneR. Terumo) we started to 3rd
SB. In the inter dialytic period, there was further 3 4 mg YO phase of our study, and finally to 4'hphase after 10 session of Exc. R
drop in SB. One session of HD against LPD (500 cc of membrane using. Results of adhesion molecules expression (%) pre-
emulsified soy bean oil of intra-venous hyperalimentation *, and post-** HD period are presented as below:
HDmemb. CD45+ CD4+8- CD4+ CD18+ CD18+ CD18-
quality) plus 750 cc of ALB, lead to an inter dialytic drop of Adh. Mol. 8+ Ilb+ Ilb- Ilb+
SB by 9.2 mg %. The highest SB of 30.8 mg YOdropped to 11.2 *RSIS-phasel 91.556 3.7_+7 35.5+8 17.2k12 50.6-6 0.22M.I
mg??at the time of discharge home. The highest INR of 8.8 **RS15-phasel 9 7 . 4 s 3 . 7 ~ 49.79 16.5511 72.1513 0.59B.6
*RS15-phase 2 9 5 . 9 s 1 7 . 6 3 42.89 l5.9&8 65.3?14 0.33M.4
dropped only transiently with fresh frozen plasma (FFP),but **RS15-phase2 94.1+3 13.1k6 44.758 14.5% 55.3516 0.51B.3
after the LPD plus ALB-HD, the INR fell to 2.5 without further *Exc. R-phase3 91.154 9.1510 42.4L5 28.7k18 46.7-4 0.57M.S
FFP. The Ascites and edema improved with ultra filtration on **Exc.R-phase3 90.6M I1.8i7 5 0 . 7 9 18.8514 52.1520 0.89io.9
HD and the patient could ambulate again. The grade 1 hepatic *Exc. R-phase4 92.19 15.1+7 34.9+8 38.6+17 34.1i15 0.46M.3
coma resolved by discharge. **Exc.R-phase4 90.754 12.957 4 1 . 1 9 26.2513 4 3 . 1 s 0.44M.3
LPD as emulsified soy bean oil is about 10 times less Significant decreases in the pre-post HD values of CD45+.
and C D 18+1 I b-adhesion molecules expression during session with
expensive than ALB and may be a more efficient binder of fat Vit.E modified Excerfane membrane in comparison to classical
soluble toxins than ALB-HD alone in HRF. cellulose has been observed.
W. Zatuska', A. Ksiqek'. J. Rolinski'
Dept. ofNephrol.', and Immunolog): Med. Academy, Lublin, Poland
 212 ASAIO RENAL ABSTRACTS

EsIlMATION OF DLALYSIS DOSE USING I(TN INDEX WHERE THE REDYB RECIRCULATING DIALYSIS SORBENT
V IS ESTIMATED BY MULTWREQUENCY BIOIMPEDANCE SYSTEM: STRATEGIES TO INCREASE CLEARANCES.
SPECTROSCOPY @JS) VERSUS DLF'FERENT FORMULAS FOR The standard REDY dialysate flow of 250 d m i n limits the
DETERMININGOF KTN. clearance of small solutes. To increase clearances, we investigated:
The most extended index for quantification of dialysis 1) recirculating the dialysate. The inlet and outlet of the malysate
dose is Kt/V calculated by urea kinetic modeling (UKM). side of the dialyzer were connected to a pump and the dialysate
Based on principle, that volume of urea distribution (V) is equal was recirculated at a flow rate of 250 d r m n thereby resulting in a
to post dialysis total body water (TBW) in human objects, we flow of 500 d m i n through the halyzer (recirc), 2) two dialyzers
meaured regulalry TBW using whole body bioimpedance (BIS) were connected in parallel and 3) two dialyzers in series. The
technique. BIS analysis has been validated using multifrequency control was the standard system. The REDY machine was prepared
BIS approach (4200 Hydra, Analyzer. Xitron, San Diego, CA. with bicarbonate dialysate (Kit 2). The "blood was the same
U.S.A.). The purpose of our study was to evaluate of Kt/V index lalysate to which had been added 100 mg/d of urea-N and 15
in 321 hemodialysis (HD) patients where V is measured by BIS mgidl of creatinine. The dialyzers used were Baxter CA 170. The
technique (Kt/Vsls), versus different formulas based on UKM blood and dialysate were warmed to 37 degrees and both flows
calculation: single pool spKt/V, 2"d generation Daugirdas were adjusted to 250 d m i n . Pre and post blood pumps were
adjusted to revented ultrafiltration. The results were as follows:
formula (Kt/Vbup) = -LnBUN,,JBUN,,,-(O.O08*T~)-UFIW,,,,
and equilibrated Daugirdas-Schneditz formula (eKt/V) =
Variable Average Clearances &
spKtN-0.6 x WV + 0.03; where BUN,,, and BUN,,= pre. and
ContTol 164k7.9 133f15
post HD blood urea nitrogen concentration, UF4trafiltration
Recirc 138k7.6 124f15 <0.001 0.292
rate. Td=dialysis time, and Wpos,=postHD weight.
Parallel 169k6.4 141f4.9 0.229 0.256
Kt/VB,s spKt/V KtlVD,,, eKt/V
Series 20258.4 175k9.1 <0.001 <0.001
meanSD 1.3310.29 1.04i0.32 1.05i0.34 0.9210.28
P. p< 0.001 p< 0.001 p< 0.001 The change in flow from countercwent to diaiysate recirculation
gave a lower clearance even though the dialysate flow rate was
r 0.991 0.999 0.992
doubled whch possibly increased &A. Doubling the area using 2
BIS technique can be reccomended for accurate and non dialyzers in parallel was no better than the control Evidently
invasive measurement of TBW=V. but K~/VBIShighly dividing the flows to give 125 mumin in each reduced clearances.
overestimated of eKt/V as "gold standard". because of Double the area in series gave significantly higher clearances. T h ~ s
incorrect manufacture prescription of dialyzer clereance. indicates that a single long dialyzer may be the optimum dialyzer.
W. Zatuska, A. Ksiqzek. T. Maiecka
Department of Nephrology, Medical Academy, Lublin, Poland M. Roberts, S.W. Wong, E. Dinovo, N. Yanagawa and D.B.N.
Lee. Greater Los Angeles Healthcare System. North Hills, CA

T Cells of Dialysis And Praedialysis Patients Are Undergoing HEMODIALYSIS OF THE NORMAL GOAT.
CD9YFas) Mediated Apoptosis Ing T, Thomhill J, King W, Lascio M, Driscoll M, Schroeder C,
J.Ankersmit'. B.Moser', LTeufel', M.Schuste?, G.Wieselthaler'. M. Prakash D, Kuna P, Kiellstrand C. Aksys Ltd and Hines VA, IL.
Grimm', G. Bolz-Nicoulescu', S. Itescu2, E.WoIner' Dept of CT Surg Pre-clinical studies of hemodialysis equipment is increasingly
Vienna, Austria', Columbia University New York, Transplant requested by governments. We studied the suitability of normal goats
Immunology to test a dialysis machine with 10 hemodialyses using 1.8 m2
Purposeof Study: T cells of Left Ventricular Assist Device polysulfone dialyzers.
(LVAD) recipients are akrrantly activated via the CD95 pathway with Methods: Blood-access was by per-cutaneous jugular vein
high levels of apoptosis in vivo and low lymphocyte counts. Since LVAD
recipients and Dialysis patients are reported to have high incidence of catheters, (CV-C), normal chemistries and deficiencies checked by
infections we sought to investigate whether similar activation pathways pre-dialysis electrolyte tests, hemolysis by free Hgb, clearances by
are induced. Methods: Thus we explored 20 hedialysis (PD) and 20 arterio-venous BUN differences and urea-volume (V-U) and urea
Dialysis (D) patients. 10 Healthy Humans (HI+) served as controls. By dynamics and transport by pre-and post- and rebound BUN and total
FACS analysis we stained T cells for CD95. HLA DR expression. and dialysate collection.
utilized HIV batch analysis to evaluate absolute counts of subpopulations. Results: The goats weighed 61 and 51 kg. Prominent jugular veins
For detecting apoptosis we utilized annexin V mABs, an AB binding to made CV-C easy to place, the long, mobile neck made them difficult
phosphatidylserine,a phospholipid present on cell membranes undergoing to position. Anti-coagulation was with sc. Heparin, oral aspirin and
apoptosis. Results: D vs PD patients demonstrated in a cross warfarin. The latter is useless in ruminants. No goat bled from the
sectional analysis signifcantly lower CD4 T cells (mean 420/mm3 vs anticoagulation. 4 CV-Cs were needed the first 15 days before CV-C
668/mm3,p<0.05). In conmt CD8 T cells showed no difference success. There was no evidence of clinically important hemolysis.
(334/mm3 vs 363/mm3, p<NS). This reduction of lymphocyte levels was Urea- or PO4 deficits did not develop. V-U was 55% 19%SD. of
accompanied with high expression of CD95(Fas) of CD3 cells on both body-weight, rebound difficult to estimate at low BUN values,
groups studied (mean 72.2% vs 80% - HH 33.5%). HLA-DR expression achieved clearances in agreement with predicted and
as sign of T cells activation was not increased in both groups studied achievedprescribed KtN was 1 .&0.24 SD. Hematology, liver-
(mean 10% vs 10%). As CD95 triggering is initiatmg programmed cell enzymes and electrolytes remained stable. The goats were easy to
death, we compared the levels of annexin V D and PD patients
handle.
demonstrated a 7 fold increase in annexin binding in comparison to HH
(72% vs 74% vs 10%). Togehter, these results suggest that dialysis and Conclusion: Goats are ideal animals for dialysis experiments. They
predialysis patients are susceptible to undergo AICD (Activation Induced are docile, friendly and easy to handle. Have weight in line with
Cell Death) and are likely to be responsible for decreased T lymphocyte humans, favorable venous anatomy but require x-ray for catheter
counts in the patients evaluated. This increase of this apoptotic pathway position . They do not develop deficiencies by intense dialysis. For
might suggest another mechanism of immunosuppresion besides uremic maximum precision urea-determinations should be calibrated for low
sideproducts and might suggest an additional explaination for the values.
increased incidence of systemic infections in the patient populations
studied.
 ASAIO RENAL ABSTRACTS 213

Total Body Water (TBW) Content in Chronic Maintenance CORRELATES OF SURVIVAL IN HEMODIALYSIS (HD)
Hemodialysis (HD) Patients: Comparison to the National PATIENTS (PTS): 12 YEARS OF FOLLOW-UP
Health and Nutrition ExaminationSurvey 111 (NHANES Ill) Disappointingly poor survival during treatment for end stage renal disease
(ESRD) in the United States has been attributed to various causes including
An accurate definition of TBW in HD Patients is a requisite inadequate dialysis and malnutrition. The objective of this study was to
for precise quantitation of the amount of dialysis delivered, urea identify the factors that are associated with long-term survival on HD. We
kinetic modeling, estimation of body cell mass, and dry weight have prospectively examined enrollment demographic, and visceral and
monitoring. Routine clinical methods are based on age, sex, height somatic nutritional markers in 530 HD pts from 1987 onwards (monitored
and weight. We evaluated TBW values obtained with various through October, 1999). For the prealbumin study, 126 pts were enrolled
formulae (Watson, Hume, Tzamaloukas, Randall, % body weight, from June 1991. Pre-albumin was measured by rate nephelometry on a
and adjusted % body weight) to direct measurements using Beckman Array Protein System. Serum intact PTH was measured by
bioelectrical impedance (BEI) in 154 HD patients (age: 62 f 16 y; radioimmunoassay. Mean age was 60116 (SD) y s ; female, 54%; Afiican
46% female; 32% black). Bias confidence intervals for TBW were American, 57%; diabetic, 47%. Enrollment levels of nutritional markers
lowest with the Watson (2.0 L), Hume (2.4L) and Tzamaloukas were significantly lower in diabetics compared to non-diabetics (P<O.OOl).
(2.4L) equations (Bland-Altman method). Direct BE1 resistance Over 12 years of observation, the observed survival (Kaplan Meier) of
diabetic pts was significantly lower than that of non-diabetics (P<O.OOOl).
(R) and reactance (Xc) measurements in HD patients were also
Survival was positively influenced by enrollment levels of biochemical
compared to those obtained in the general population (NHANES
markers of nutrition such as albumin (P<O.OOl), prealbumin (P<O.OOl) and
111). To minimize bias between HD patients and NHANES 111
creatinine (P<O.OOl) as well as PTH (P=0.002) (Univariate Cox's
subjects, R and Xc values were normalized by height (H), and proportional hazard model). Over 12 years of observation, the survival of
comparisons made only within 40 to 90 years of age. HD pts with lower enrollment levels of albumin (P<O.OOOl), creatinine
Female (P<O.OOOl) and PTH (P=0.009) was significantly lower compared to those
HD NHANES HD NHANES with higher levels of these markers (Kaplan Meier method). After
WH: 288561 277f24 354f52 372+29 adjustment for confounding variables the results were similar. Over 8 years
XcIH: 2459 36+3 28+9 46+4 of observation, pts with higher levels of enrollment prealbumin had
*
Vector: 295 63 279 5 24 355 f 52 375 f 30 significantly (P=O.008) increased survival compared to those with lower
Angle: 37+7 5 8 ~ 8 36f7 56+7 prealbumin levels. After adjustment for confounding variables the results
HD patients have vectors of normal magnitude but lower angle, that were similar (P=0.025). Serum PTH and serum prealbumin levels correlated
although fall within the 75-95% tolerance limits, imply higher TBW positively with other nutritional markers. After longer follow-up and with
and decreased body cell mass than in the general population of more enrolled pts, this study c o n f m s OUT findings regarding the prognostic
similar age (NHANES Ill). importance of enrollment nutritional markers in HD pts. Diabetic HD pts
had poorer nutritional status, which may be partly responsible for the excess
Francis Dumler. Division of Nephrology, William Beaumont mortality seen in this population.
Hospital, Royal Oak, Michigan, U.S.A. M.M. Avram. N. Mittman, R. Sreedhara, A. Henry, J.Chattopadhyay.
Avram Center for Kidney Diseases, The Long Island College Hospital,
Brooklyn, New York.

EXPERIMENTAL AND CLINICAL STUDY FOR B1- THE ASSOCIATION OF DIALYSIS DOSE WITH MORTALITY
DIRECTIONAL PERITONEAL DIALYSIS (BPD) IN HEMODIALYSIS (HD) PTS (PTS): A 12-YEAR STUDY
BPD was introduccd t o cnhance solutc rcmoval cfficicncy in Previously we reported that HD treatment dose by urea reduction ratio
conventional PD thcrapies such as CAPD. In thc BPD: wc (URR) is an important and independent determinant of more than 7 years
inscrt a conventional P D cathetcr into intrapcritoncal space. survival in 253 HD pts. In the present work we have studied the effect of
through u.hich a ccrtain amount of pcritoncal dialysatc is dialysis dose measured by URR with larger number of pts and extended
introduced. With a recersiblc pump, a prcsct volumc of thc follow-up period (more than 12 years). The objective of this study was to
pcritoncal dialysatc rcciprocates bctwccn thc peritoncal determine whether dialysis treatment dose as measured by URR is an
cavity and a dialysatc rcscrvoir? giving thc system a drain independent predictor of long-term mortality in HD pts. We enrolled 404
phasc and an infusion phasc altcmativcly. I n thc drain phasc. HD pts treated at the Long Island College Hospital from January, 1987 and
a part of thc dwclt dialysatc is drained through the PD followed them up to June, 1999. Mean age was 60*16 (SD) y s ; female,
cathctcr at 100 mbmin (Ql11t) and cntcrs thc rcscrvoir. In thc 55%; African American, 56%; diabetic, 47%. Mean cumulative URR was
infusion phasc. the dialysatc in thc rcscrvoir is rcturncd to thc 62% i 6.3 over the 12-year study period. Mean URR was higher in females
(65% vs. 60% males, P<O.OOOI), and non-diabetics (63% vs. 62% diabetics,
pcritoncal carit!' at 300 ml!min (QII.;). In both phascs. thc
P<0.05). Over a period of 12 years, mean cumulative URR correlated
add-on dial),zcr purifics thc pcritoncal dial!salc con ti nu all!^. positively with hematocrit ( ~ 0 . 3 3 ,P<0.05) and ferritin ( ~ 0 . 3 2 ,Pc0.05)
Thc dialysatc flow ratc (OD) from thc machinc MYIS 100 and negatively with body weight (F-0.33, P<0.05), creatinine (I=-0.18,
ml/min. In this study. solute removal characteristics of BPD P<0.05)and anion gap (I=-0.31, P<0.05). Observed 12 years cumulative
were cxamincd in a caninc and a clinical stud!. Solutc survival of pts stratified by three groups of URR was determined by Kaplan
rcduction ratc in BPD with a dialysatc rcscrvoir volume ( I n) Meier method. Pts with URR 60%-65% had the best survival. Interestingly,
of 200 ml: 350 ml, 500 ml was compared u%h that in the survival rate of pts with URR>65% (highest URR group) was
convcntional PD (CPD, 120 min dwelling s 2) and tidal PD significantly lower than pts with URR 60%-65%. Expected survival,
(TPD, 70 min d u d l i n g x 12, 50 c/o of tidal volumc ratio) ovcr adjusting for confounding variables, yielded similar results. Mean serum
thc samc period o f t i m e (740 min). Thc urca rcduction riltc in creatinine level was significantly lower in pts with highest URR (>65%)
BPD, CPD and T P D wcrc 20.0*7.2'% (n=16). 3 5 3 . 6 % group (9.9 vs. 11.5 for <60% and 11.4for 60%-65%, P<O.OOOl). In Cox's
(n=7). and 1 7 . 3 4 . 7 % (n=h), respcctivcl! . Thc crcatininc proportional hazards model, when we adjusted for mean creatinine, the
rcduction ratc in BPD, IPD and TPD wcrc 2 1 . 0 ~ 7 . 0 % relative risk of mortality for pts with URR>65% became significantly lower
(n=16), 2 . 1 ~ 4 . 1 % (n=7), and 1 7 . 2 ~ 6 2 %(n=6). In the compared to those with URR<60%, and similar to those with URR 60%-
clinical stud),. BPD and C P D (180 min s 2) trcatmcnts u'crc 65%. This suggests that higher mortality in pts with URR>65% may be due
applied to 7 renal failurc patients. Highcr solutc rcmocal to poor nutritional status, particularly lower muscle mass. Dialysis treatment
cfficicncy u'as shown in BPD than in CPD trcatrncnts. The dose in terms of cumulative URR can independently predict more than 12
urea rcduction rate was obtaincd as 9.720.4% (n=8) in BPD, years survival in HD pts.
N. Mittman, R. Sreedhara, K.K.00, J.Chattopadhyay, M.M. Avram. Avram
shighcr (P<O.OI) than thc 1.0~2.8 5 in CPD (n=10).
Center for Kidney Diseases, The Long Island College Hospital, Brooklyn,
M.. S. Suzuki. 1'. Sato, I. Ishimori. K.Ishicla. Y. New York.
Okuda. I. Kancko and T. Agishi, Kidney Ccntcr, Toky)
Women's Mcdical Univcrsity, Tokyo: Japan
214 ASAIO RENAL ABSTRACTS

PATIENT SAFETY TECHNOLOGY FOR MICROSPHERES DECLINING INCIDENCE OF END-STAGE RENAL DISEASE
ADSORBENT SUSPENSION-TECHNOLOGIES (ESRD) COMPLICATING HUMAN IMMUNODEFICIENCY
Adsorption Technology, based on microsphere technology using particles VIRUS (HIV) INFECTION. T.K.S. Rao, E.A. Friedman. S U N Y ,
smaller than 5- pm, kept in suspension is accompanied by different Health Science Center at Brooklyn, Brooklyn, New York.
problems, especially related to patients safety. On the other side the use of W e initially described a syndrome of HIV associated renal disorders
adsorbent particles in a range between 1-5 pm offers tremendous
advantages especially related to efficiency and flexibility. Aim of the in 1983, three years following identification of A D S as the result of
study The aim of the investigations was to develop safety systems in order infection with HIV. HIV-associated nephropathy (HIVAN)
to avoid the entrance of microadsorbents in the patient for the MDS characterized by massive proteinuria and focal and segmental
(microspheres based detoxification system)-technology which was glomemlosclerosis, in our early experience, typically led to ESRD in
developed by our group and can be used for any type of specific weeks to months. Since 1983, we have maintained a registry of all
adsorption-therapy using adsorbent-microparticles in a range between 1-5 patients with HN infection who develop either potentially reversible
pm. The safety system, therefore, should detect microparticles in the blood acute renal failure (ARF)and/or ESRD. Tabulated below are the total
circuit in concentration in a range of 10.' in order to ensure safety for number of ESRD patients who were begun on maintenance
patients. Method: For this purpose we add particles marked with a special
fluorescence dye in order to get an emission of higher wavelength (525 nm) hemodialysis at a large municipal hospital in inner city New York
after excitation by (490 nm). The measurement of the emission has been ESRD
done using a photomultiplier tube in combination with a lock in-amplifier to (ALL) D.M. -
HTN HIVAN
pick the very low signal out of noise. The spectral selectivity was reached 1983 -85 303 (27) (24) (61
by using a special arrangement of optical filters for excitation and for 1986-YO 436 (25) (21) (21)
emission to prevent the influence of cross light. Results: The use of the 1991 -9.5 440 (28) (29) (24)
newly developed sensor system enables the evaluation of particle 1996-99 271 (33) (34) (12)
concentrations for particles smaller than 5 pm in concentration until 10'' in Number within ( ) represents 8 of total ESRD patients for that period
blood circuits blood flow: (50-500 mumin.) using conventionally only D.M. = Diabetes mellitus, HTN = Hypertension
slightly modified blood lines with absolute safety. Conclusion: The new Starting about 1994, highly active antiretroviral therapy (HAART)
particle sensor based on fluorescence microparticles (<5 pm) enables a high and other prophylactic regimen were introduced in the management of
effective detection of microparticles in extracorporeal blood purification
system based on MDS-technology or other membrane based HlV pts. Between 1986 through 1995, HIVAN accounted for 22.5%
microadsorbent-systems. Therefore a clinical use of the MDS-Systems can of all causes of ESRD. But over the past 4 years (1996 - 99), there has
now be prepared for the next time. The new technology also offers been a significant decline to 12% in the number of pts with new onset
possibilities to control membrane filtration processes, especially in aqueous of ESRD secondary to HIVAN. W e speculate that modem
systems, getting a detection rate of l o 8 management of HIV infected patients may be responsible for this
D. Falkenhagen, J. Hamnann*, A. Schrefl, W. Strobl, F. Aussenegg**, A. decline in the incidence of ESRD secondary t o HIVAN. A similar
Leitner** C.Doppler Institute for Adsorption Technology in Medicine * decrease was also noted in ARF in pts with HIV disease at our hospital
Centre of Biomedical Technology, Danube University Krems, Austria **
Institute for experimental physics, Karl-Franzens University Graz, Austria over the same observation period.

DECLINING INCIDENCEOF ACUTE RENAL FAILURE (ARF) INCREASED DIALYSIS DOSE IMPROVES
IN HUMAN IMMUNODEFICIENCYVIRUS (HIV) INFECTION: RESPONSE T O RECOMBINANT ERYT'HROPOIETIN
AN IMPACT OF THERAPY? T.K.S. Rao, E.A. Friedman. SUNY, (EPO) IN PATIENTS RECEIVING HEMODIALYSIS
Health Science Center at Brooklyn, Brooklyn, New Yoric. Onyekachi Ifudu, J o s e p h Feldman, Leila J. Macey,
HIV infection was identified in 1980 and its renal manifestations
potentially leading to ESRD were recognized in 1983. W e main a
Eli A. Friedman; SUNY-HSC, Brooklyn, USA.
registry of all patients with HN infection who develop either The relation between dialysis adequacy and anemia was
omitted in the recent guidelines on anemia management in
potentially reversible acute renal failure (ARF) andor end stage renal
ESRD. We studied 60 hemodialysis patients on a fixed
disease (ESRD). For the purposes of this retrospective study, we
dose of EPO to determine the effect of increased dialysis
define severe ARF as the abrupt loss of renal function inducing a rise
dose on hematocrit.
in those patients in whom the serumcreatinine (Scr) concentration rises
Patients were randomly assigned to receive increased
to >6 mgldl, a n d o r requires dialysis support. The annual incidence of
dialysis dose (FSO dialyzer; 4.5 hours thrice weekly;n=30)
severe ARF for each 5 or 6 year period from 1983 through 1999 is
or standard therapy (MCA 160 dialyzer, 4 hours thrice
tabulated below:
weekly:n=30)and hematocrit was measured weekly for 12
YEARS ARF ANNUAL ARF ANNUAL weeks.
(All) Incidence HIVPts Incidence With adjustment for EPO dose, the odds of achieving a
1983- 1988 259 43 84 ( 3 2 % ) 14 hematocrit o f 33 ?hover three consecutive weeks was 3.8
1989- 1993 216 43 72 ( 3 3 8 ) 14 times more likely among study subjects than the control
1994 - 1999 203 34 33 ( 1 6 % ) 6.5 subjects (95%CI=1.01-14.19; P = 0.04).
() number within represents % of ARF from all causes A second outcome used was the percentage of all follow-
Over the first 11 years of study (1983-93), while the incidence of up weeks during which patients attained a hematocrit of
severe ARF from all causes was 4 3 cases per year, HlV patients 33%. This occurred during 30% of the control patients'
represented 33.5% (14 per year) of the total. Over the last 5 years follow-up weeks compared to 45% of the follow-up weeks
(1994-99), while there was an overall slight decrease in the annual for the study subjects. This difference was statistically
incidence of ARF from all causes (34 cases per year), there has been a significant (P = 0.02) after adjustment for EPO dose. The
more than 50% decline in HIV pts. During the past 5 years, only 1 6 8 estimated mean percent of such visits after adjusting for
of all ARF cases were seen in HIV pts, for an annual incidence of 6.5. EPO dose was 26% for controls and 49vo for the study
Highly active antiretroviral therapy (HAART) and other regimens for subjects.
HIV infection introduced about 1994 may have caused the We conclude that anemia improves with increased
renoprotective benefit detected. W e noted a similar decrease in the dialysis dose in EPO-treated hemodialysis patients.
number of new onset ESRD cases ifi pts with HIV disease over the
same observation period.
 ASAIO RENAL ABSTRACTS 215

ADJUSTMENT TO DIALYSIS RELATES TO RACE AND A NEW METHOD FOR CONTROLLING ULTRAFILTRATION
ANXIETYDEPRESSION SCORES BUT NOT TO INCOME OR RATE IN CONVENTIONAL CONTINUOUS ARTEROVENOUS
HEMOFILTERATION (CAVH). Moro 0. Salifu, MD and Eli A.
EDUCATION IN INNER CITY HEMODlALYSIS PATIENTS Friedman, MD, Renal Dis. Division, SUNY HSC at Brooklyn, N.Y.
MS Markell, A Ehrlikh, J Sherman, J Cheung, H Surya, CD Brown, Purpose: In conventional CAVH, the ultrafiltrate (UF) volume is
EA Friedman; SUNY Health Sci Ctr at Brooklyn, NY controlled imprecisely with an UF clamp, which is labor intensive and
Hemodialysis necessitates a disruption in routine lifestyle which requires frequent adjustment to avoid excessive fluid removal. We present a
may create anxietyidepression. A cohort of 20 randomly chosen simple method for precisely controlling the UF volume.
patients dialyzed in-center at an inner-city unit were evaluated using Method: 7 consecutive CAVH’s were performed in 4 patients with
Beck’s depression scale, a standard anxiety scale, and a newly massive edema. A standard circuit was accomplished in each case by using
blood tubing sets and an HF400 obtained from Renal Systems. A 3-way
created form designed to measure adjustment to dialysis (ADJUST), stopcock was connected to the dialysate port at the venous end. The air
which includes 11 statements regarding loss of control, chamber of a standard intravenous infusion tubing was cut out and the
frustration(boredom), and physical symptoms (pain, muscle tubing itself connected to the stopcock and secured safely with a tape. The
discomfort). Patients rate each statement (from 1, “doesn‘t bother tubing was flushed to remove air bubbles and connected to a regular
me at all” to 5,”I can barely stand it”) and the scores are summed. intravenous infusion pump (UF pump). The patient side of the tubing was
Depression and anxiety scores were highly correlated ( ~ 0 . 9 , inserted into the tubing of a drainage bag and secured with tape. The desired
p<O.OOOl), and both correlated with ADJUST (FO.5, piO.05: ~ 0 . 8 , amount of fluid losshour was estimated based on the patients weight gain
and hemodynamic parameters. This amount was dialed into the UF pump,
p<O.OOOl) and were inversely correlated with age of onset of renal which then pumped UF into the drainage bag at a constant rate. Electrolytes
disease (F-0.5, p<O.O5, F-0.5, p<O.O5), but not education or income were replaced using the predilution method. No additional replacement
level. There were 14 black and 6 non-black patients. Black subjects fluid was required since the ultrafitration rate was constant and
had significantly lower anxiety (9.511.8 vs. 21.3i7.4, p<O.O5) and predetermined. Additional adjustments were made as necessary. No heparin
better ADJUST scores (22.212 vs. 3 1.6*4.6, p<O.O5), and lower was used. Treatments ranged from 27-78 hrs.
depression scores (7.5*1.8 vs. 14.8*4.1, p=0.07), hut did not differ Results: The target fluid removal was achieved in all treatments and the
UF rate required only daily adjustments for total fluid intake. The average
from non-blacks for age (53.4*3.1 vs. 50.5i5.9 yrs), education or
time to reporting a problem by the intensive care nurse was 30 hours and
income level, or time on dialysis. the average time to filter clotting was 38 hrs.
We conclude, in our population I . Poor adjustment to dialysis may Conclusion: Our modification offers an effective. safe. less laborious
relate to underlying anxiety/depression; 2. Patients who develop alternative method to more expensive equipment for CAVH. Predilution
renal disease at an earlier age report more anxietyldepression; 3. may be effective in patients in whom heparin is contraindicated. Tubing for
Black patients have better adjustment, and less anxiety/depression, UF designed as above may be more effective than tubing with graded
irrespective of income or education. The underlying reasons for this clamps. Until such tubing is commercially available, the above technique
affords constant rate ultrafilteration.
finding and whether it contributes to the better survival of blacks on
hemodialysis remain to be investigated.

DIURNAL VARIATION IN BIOAVAILABILITY OF IRON TRANSFORMATION OF A MUNICIPAL HOSPITAL

AND THE RESPONSE TO RECOMBINANT HEMODIALYSIS UNIT TO A “STABILIZING” UNIT. Andrew
ERYTHROPOIETIN(EP0) I N HEMODIALYSIS PATIENTS Peter Lundin. III, Benzena Dosunmu, Nathan Joseph, Yalem
O, J Uribarri, I Ra1wani.V Vlacich, K Reydel, G Woredekal, Eli A. Friedman. S U N Y , Health Science Center at
Delosreyes,& EA Friedman: SUNY-HSC a t Brooklyn, Mount Brooklyn, Brooklyn, New York.
Sinai Medical Center, NY, Nephrology Foundation of Maintenance hemodialysis in ambulatory patients is performed in
Brooklyn. either a hospital based or free-standing dialysis facility (DF). The
Impaired utilization of storage iron for erythropoiesis is a pattern of patient flow to DFs is minimally understood. We examined
major problem in patients with ESRD. In health, a diurnal the demographics, dialysis quality, and outcome from 1989-1999 in a
variation in t h e s e r u m level of transferrin saturation may municipal hospital 16 station DF with a three-shift, 6 day-a-week, 72
translate into greater erythropoiesis in t h e evening. patient capacity (1 1,OOO annual dialyses). Newly referred patients are
Iron, now administered intravenously in most dialysis derived from an in-hospital “acute” dialysis unit while over capacity
patients is initially taken u p and processed by macrophages patients are transferred to community DFs.
before transfer to transferrin. This extra step in iron From 1989 to 1999, mean patient age increased from 41.7 yrs to 50.9
metabolism is potentially disruptive to the diurnal variation yrs as the number of treated patients fell progressively from 149/yr to
in iron bioavailability. 109/yr reflecting a decrease in newly treated patients from 57/yr to
We-studied 309 hernodialysis patients for 3 months to 33/yr. Only 1 to 4 ptdyr received a kidney transplant. Stabilization
determine whether patients dialyzed in t h e evening (4* and retention of the cohort on dialysis reflected improved dialysis
Shift) have a better response to EPO (higher hematocrit) adequacy: urea reduction rate rose from 57.9% in 1989 to 65.9% in
than those dialyzed early in the morning (1‘‘ Shift). 1999 while hematocrit and serum albumin level improved from 24.3%
1st Shift,,.,,, 4thShift,,.,,, P to 30.9% and 3.7 g/dl to 3.9 g/dl respectively.
HCT(%) 364.4 374.6 NS Annual deaths fell from 24 in 1989 (1 1 in HIV+ pts) to 10 in 1999 (1
EPO Dose(U/Ka 654.5 7327 NS in an HIV+ pt). A key factor contributing to better outcome was
Urn(%) 724.8 73d.9 NS subsidence of mortality due to AIDS in pts with HIV-associated ESRD.
lV Iron(mg) 221239 176~27 NS Remarkably, only 1 death was recorded in an HIV+ hemodialysis
Transferrin Sat(%) 35& 3323.8 NS patient over the past three years. Transfers to community DFs which
Linear regression analysis with achieved hematocrit as t h e ranged from 24 to 49 annually over the decade fell to a low of 15 in
outcome variable showed that the time of day a patient 1999. Our findings affirm the linkage between dialysis adequacy and
receives dialysis (P=0.77) has no effect o n their response to survival and underscore the favorable impact of present antiviral
EPO. r q m e n s on the course of HIV infection. Accordingly, the main
We conclude that there is n o diurnal variation in t h e mission of our DF has been transformed from that of a way station to
response to EPO in patients o n hemodialysis.
that of a stabilizing and often permanent treatment resource.
216 ASAIO RENAL ABSTRACTS

UREA DISTRIBUTION VOLUME CAN BE DETERMINED FROM

UREA MASS AND REMOVAL RATE
Stemby J, Persson R, Wmgren K, Alquist Hegbrant M
A new method for the determination of the urea distribution volume
(V*), based on a pre-dialysis blood urea sample and an analysis of the
dialysate urea concentration profile provided by GAMBRO on-line urea
monitor DQM 200, has been reported previously (Sternby, JASN, 9:2118-
2123, 1998). We carried out a clinical study to compare the new method
with a reference method for urea distribution volume determination (V.J.
Moreover, the body water volume given by Deuterium (D,O) dilution
technique (Vo20)was measured in parallel in a subgroup of the patients.
We studied 15 clinically stable dialysis patients (1 female) ofthe average
age 69 'i. 20 years. 29 regular hemodialysis treatments were performed.
Body water volume (Vo20)determinations were conducted for 8 patients.
V,, is based on the relation between the blood urea change (Cbt - C,)
during the dialysis and the total removed urea (TRU) according to
V,pTRU/(Cb,-Cw), corrected for UF and urea generation. TRU was given
as the product of the urea concentration and the volume of the collection of
all effluent dialysate during a treatment.
NIAGARA@ n=14 OPTI-FLOW@ "=IT In the new method the KN (the log-slope of the urea profile) and the
SetOb 300 350 400 450 300 350 400 450 urea removal rate (Q, * C,) are determmed. Since the KN reflects the
non reversed with no R n=l I n 4 relative change of the body urea, the urea mass present in the body at any
dumtion of insertion 40213 days 113-92 days
299116 342+24 382+41417+61
given time during the treatment (mJ can be calculated as m, =Qd * C, /
RealQb 312+11 353tlO 393+12 424+16
R 0 0 0 0 0 0 0 0 KN. The urea distribution volume by the new method is given as V*=md
nos reversed with R n=2 n 4 C,,, where m, is the initial urea mass of the patient.
durarion of insertion 3027 days 105272 days The new method for urea distribution volume (mean = 37.7 L f 6.5)
RealQb 31816 370t221 4 1 W 5 449t20 302+4 349t2 375t19 426t17 showed good correlation ( ~ 0 . 9 5 n=27/29)
, to the reference method (mean
R 12t4 1011 1517 16t7 6+7 6k6 717 716
-d -=I n=5
= 37.2 L i 5.7). The body water volume (D,O) was higher (mean = 44.7 L
134278 days f 5.7 or 56.1% f 4.1 of the body weight) than the urea distribution volume
duration of insertion 26 days
RealQb 326 370 398 446 303116 346+19384+23 420128 by the new method (mean = 41.3 L f 4.5 or 51.8% f 2.7 of the body
R 0 0 4 6 23+10 22C8 2316 26t4 weight), but with a good correlation ( ~ 0 . 9 2 n=7/8).
,
Conclusion: The urea distribution volume measured by the the new
method agrees well with the reference method. The body water volume by
Deuterium dilution was 8% higher than the urea distribution volume by the
new method.

PRELIMINARY EXPERIENCE WITH A CUFFED ePTFE GRAFT FOR DETERMINATION OF EXTRA RENAL CLEARANCE AND
HEMODIALYSIS VASCULAR ACCESS GENERATION RATE OF 0,-MICROGLOBULIN IN HEMODIALYSIS
Scott L. Nvberg, Christopher Hughes, Margaret Benda, E l i b e t h Garr, PATIENTS USING A KINETIC MODEL
Craig Hocum, James McCarthy, Sylvester Sterioff, Mark Stegall. Divisions
of Transplantation and Nephrology, Mayo Clinic, Rochester, MN. Introduction: Beta-2-microglobulin (l3,-M) has been proposed as a
marker of middle molecules to assess the efficacy of dialysis. Only a few
Background- Thrombosis is the leading complication of synthetic data have been published about extra renal clearance and generation rate of
hemodialysis grafts and stenosis at the graft-vein anastomosis is a l3,-microglobulin which are necessary for calculation of total clearance and
contributing factor in many cases. A new cuffed ePTFE graft (VENAFLO, mass removal of l3,-M in hemodialysis patients. Here we have developed a
Impra, Inc.) was developed to decrease the incidence of outflow stenosis simple method to derive extra renal clearance and generation rate of l3,-M
and improve long-term graft patency. from measuring the pre-and post-dialysis blood concentrations of l3,-M
Purpose- The purpose of this study was to determine the one-year graft using kinetic modelling.
patency and blood flow rates in a new cuffed ePTFE graft. Methods: Ten stable hernodialysis patients are included in this study.
Methods- A pilot study was conducted on twelve (7 male, 5 female) Pre- and post-dialysis concentrations of B1-M were measured during
consecutive patients 36 to 76 yr of age (mean=65yr). Seven were high risk dialysis with low flux dialysers(F6 HPS) and after ten days switching to
due to a prior history of clotted hemodialysis accesses (1-6, mean=3.3). high flux dialysers(F60S). With a validated two pool model the generation
Placement sites included forearm (2), upper arm (S), and leg (2). The rate of a,-M can be determined if extra renal clearance is known.
outflow vein diameter ranged from 3 - 8 mm (mean=6.8mm). Blood flow Assuming the generation rate of O,-M to be constant in each patient, the
rates were determined by ultrasound dilution technique (HDOI Transonic computer reiterated the calculation of extra renal clearance until the
System) at 3 month intervals. calculated generation rate was equal both for the low flux and the high flux
Results- Two patients died (malignancy, myocardial infarction) with patent dialyser.
grafts during the first year. One patient moved and was lost to follow-up. Results: Extra renal clearance was found to be between 2.0 and 4.17
One-year patency of the remaining cuffed ePTFE grafts was 89% (8/9). ml/min(average value 3.2 mlimin). Generation rate was found in a rather
Only one graft (high risk, 6 prior grafts) was lost to thrombosis during the narrow range between 1.63 and 2.53 mgkglday(average value 1.99
fvst year of follow-up; outflow stenosis was ruled out as its cause of mgkglday). No correlation was found between extra renal clearance and
generation rate.
thrombosis. When f m t measured 3 months after p ft placement, blood
Conclusions: With this simple method extra renal clearance and
flow rates ranged from 620 to 2083 mL/min (mean=1225mL/min). No
generation rate of a,-M can be determined by means of switching
differences in flow rates were observed between 3 months and 12 months
hemodialysis patients from impermeable to permeable membranes.
(mean=l143mL/min) in the 8 grafts available for comparison.
Conclusion- Stable blood flow rates and excellent one-year graft patency
X.O.Xul, N.Grune?, A.Al-Bashi?, Ch.Trutt-Ibing', H.Melze?,
were achieved with a cuffed ePTFE graft, even in high risk patients with W.Fassbinder', S.Stille?, H.Mann2
prior history of graft thrombosis. A randomized, prospective trial has been 1 Renal division, Renji hospital, Shanghai second medical university,
initiated to compare the long-term patency of cuffed and traditional (non- China, 2 Dialysis Center Aachen, 3 Medical clinic Ill, Fulda, Germany
cuffed) ePTFE grafts for hemodialysis access.
Study Sponsor- Mayo Foundation; no industrial support was received.
 ASAIO RENAL ABSTRACTS 217

DEGLYCEROLIZMG RBCs: CENTRIFUGATION vs. FILTRATION SPECT VISUALISATION OF FLOW THROUGH

J. Draheim, H. Reddy, D. Rudolph, R. Coughlin ARTIFICIAL ORGANS
SymBiotech Inc., Wallingford CT 06492 A measuring set up has been designed to visualise the flow profie in
a hollow fiber dialyser (Fresenius F6) and an artificial lung (Cobe
This spdy compared a standard centrifugation based method for Optima) by means of SPECT techniques.
deglycerolizing cxyopreserved RBCs and a proprietary, automated method A computer controlled injection system using an elemonic valve
based upon hollow fiber membrane (HFM) filtration. was developed. Optimisation of the injected volume (radioactive
solution of 99mTc) and the speed of injection was necessary to get
The starting material for each type of process (centrifugation or filtration) minimal flow perturbation and a plausible image acquisition. Images are
was packed RBCs that had been cryopreserved in glycerol, stored frozen, taken by 2 fxed positioned gamma cameras resulting in 2D pictures.
and thawed following the guidelines in the AABB Technical Manual Using water instead of blood, the flow rates needed to be adjusted to
(1999). Ten units of thawed RBCs were deglycerolized by the American account the difference in viscosity (factor 2.5). To visualise the flow
Red Cross using a standard centrifugatiodwashing process. Another ten (1 OOmVmin) in the blood compartment of the dialyser, transport through
units of thawed RBCs were deglycerolized using a new, automated HFM the membrane is avoided by using high molecular weight radioactive
filtratiodwashing process under development at SymBiotech, Inc. In the tracers (60kD). Between successive measurements, a rinsing time of 5
SymBiotech process, RBCs recirculate through the lumens of a HFM minutes is taken into account to remove radioactivity. Because of the
cartridge, while glycerol-rich extracellular fluids are removed by filtration, hydrophobic membrane and the higher flow rate (2.6Vmin) in an
and replaced by saline solutions. A programmable logic controller artificial lung, high molecular weight tracers and rinsing time are
automates the timing, sequence, and magnitude of all flow rates. unneceskq.
The quality of the fmal deglycerolized RBCs (dRBCs) produced by each The concentration profile in a hollow fiber dialyser shows a lower
method were assessed by measuring sterility, total RBC hemoglobin velocity in the centre. The remarkable activity accumulation in the
(Hb,,), total RBC count, product volume, osmotic fragility, ATP/Hb, and headers could be due to the adhesion of radioactive tracers to the plastic
DPG/Hb. Extracellular fluids (excf) were examined to determine tube. The 2D imaging of the artificial lung shows very clearly the
[glycerol],, osmolality, [Hb],, [LDH],, and [KfInd. The average concentration distribution over time and can be used for understanding
values of total Hb,, total RBC count, DPG/Hb, [glycerol],, [Hb],, the complex flow profile.
[LDH],,, and [K'],, measured for each method were not significantly Because of the axisymmetric shape of the blood compartment in a
different at a 99% confidence level. The products from the filtration dialyser, the 2D images can be used to evaluate optimal flow
methodhad significantly (99% confidence level) higher average ATPMb distribution of current and future dialysers. To study in greater detail the
and larger volume than did the products from centrifugation. flow through an artificial lung, a 3D image should be created using
rotating gamma cameras.

S.,
P. Dierickx, L. Bouwens, B. Cuvelier, D. De Wachter,
R. Dierckx, P. Verdonck
IBITECH, Ghent University, Gent

DEVELOPMENT OF MEMBRANES FOR BIOARTIFICIAL IN VITRO PERMEABILITY OF DIALYSERS

ORGANS An in vitro set up has been designed to study the hydraulic water
Th. Groth', F. Fey-Lamprech?, B. Seifert', G. Mihanetzis3, G. permeability of hollow fiber dialysers for forward and reverse
Malsch', W. Albrecht', U. Gross2, Y. Missirlis3, D. Paul' ultrafiltration flow (UF). The measurements are done using new
'GKSS Research Center, Kantstr. 55, D-14513 Teltow, 'Free samples and dialysers with a deposited protein layer.
University Berlin, D-12200 Berlin, Germany, University of The ultrafiltration coefficient KUF (mVh.mmHg) is derived from
PatraS, E o n - P a m , Greece
volumetrical flow measurements and transmembrane pressure
Membranes for bioartificial organs (BAO) supporting or
(TMP) measured with fluid filled pressure transducers. The tests are
replacing the function of kidney have to support the attachment
performed using reverse osmose (RO) water or bicarbonate dialysate
and function of kidney epithelial cells. If the blood to be
at room temperature. The protein layer on the membrane is induced
detoxified contacts the membrane directly, the polymer
membrane must be both tissue and hem0 compatible. It was the in vivo or in vitro by circulating human plasma through the dialysers
aim of this work to develop membranes that support the during 20 minutes. TMP is corrected for colloid oncotic pressure.
attachment and fhnction of kidney epithelial cells having a good Three samples of dialysers (Fresenius F6, F8, F60 and F80) have
hernocompatibility of the blood contacting side. been evaluated.
Copolymers of polyacrylnitrile (PAN) were synthesized KUF is almost independent of TMP up to 600mmHg (low flux)
differing in their content of co-monomers to change the and 60mmHg (high flux) for forward and reverse flow. Using
wettability of membranes or to introduce different types of dialysate instead of RO water decreases the permeability of the
functionalities. Flat membranes prepared from these polymers. membrane (30% in F6 and 17% in F60 and F80). A protein layer on
were tested by interaction with 3T3 fibroblasts to rule out any
toxicity and tissue-incompatibility. Investigations with kidney
epithelial cells estimated the attachment, morphology and
function of these cells on the different membranes.
Haemocompatibility of the different polymers was studied testing
blood clotting, complement activation and platelet adhesion/
activation under static and flow conditions on the membranes.
...Fibroblasts attached and proliferated particularly well on PAN
with high content of amine-containing co-monomers. Studies
1
the membrane decreases KUF with 21-26% (in vitro) and 15% (in
vivo) in low flux and 53-68% (in vitro) in high flux dialysers.
Backfiltration renders a 75% lower KUF for F80.

F80
7.39k0.16
7.58 f 0.13

197+4
5.49 k 0.22
5.97 f 0.06

92.1 f4.3
6.34 f 0.08
6.38 k0.03

with kidney epithelial cells revealed a different pattern of tissue For the usual TMP range, KUF can be considered insensitive to
compatibility with best growth and epithelial barrier function on TMP with forward and reverse flow. Because of a lower
membranes with moderate content of amine containing and backfiltration coefficient, the risk of pyrogenic reactions by non-
hydrophilic co-monomers. Blood compatibility was found to be pure dialysate could be overestimated.
good on sulfonated and hydrophilic PAN. In conclusion of the S . ,
D. De Wachter, J. Vienken*, R. Pohlmeier', P. Verdonck
investigation, PAN membranes with hydrophilic co-monomers Hydraulics Laboratory, IBITECH, Ghent University, Gent
seem to be useful for the application in a bioartificial kidney. 'Fresenius Medical Care, Germany
Supported by European Commission - grant BE97-4329.
218 ASAIO RENAL ABSTRACTS

COST OF DAILY DIALYSIS IN THE U.S.A. VASOMOTION MODEL FOR EXPLANATION OF UREA
Robers S. Aksys Ltd. REBOUND AFTER HEMODIALYSIS (HD) (Mathematical model).
Purpose V. Kislukhin, Transonic System, Ithaca, NY.
Urea rebound is usually explained as urea entering the circulation iiom
Daily hemodialysis (DHD) is thought the ”best“treatment
poorly perfused tissue after HD. The purpose of the following mathematical
for patients with ESRD. While doubling the frequency of the model is to describe the observed urea rebound by phenomena of
treatment is necessarily more expensive than 3 times a week vasomotion. This phenomena is related to the fact that a significant part (up
dialysis, better general health may offset those costs by to 80 -90%, depending on the organ) of the microvessels may be closed
reducing global health care expenditures for DHD patients. while others are perfused. If vasomotion is slow (the same microvessels are
Methods- perfused) then the byproducts of metabolism accumulate in tissue around
Proiect HOPE studied this by reviewing all available closed microvessels. The probability schema of these effects can be
described as shown below:
literature about the cost of DH% and applied an economic openvessel - a - open vessel
model that simulated costs of daily versus conventional, 3 ----I3 P Y
times a week, HD. Dialysis Consulting Group costed out closed v e s s G - closed vessel
resources used in a short daily and a long nightly HD program a =probability to be open and remain open
in the US. Costs were drawn from facilities’ books, reviewed p= to be open and become closed a+ p = 1
onsite & categorized for both programs. v = probability to be closed and remain closed
p =to be closed and become open p+ v = l
Results
The vasomotion model that describes the process of urea accumulation
The literature review and model estimated that-at a national between HD sessions and urea removal during HD includes generation of
level, DHD would cost a facility from $39 to $73 extra per urea, q, distribution volume of urea, V, and clearance of urea during I D, K.
week and patient. Site visits estimated those extra costs to be The model shows that during HD, concentration of urea in the perfused
$213 to $345 in 2 programs. tissue approaches q V K and concentration of urea in tissue around closed
Depending on several variables, however, societal savings microvessels approaches q(VK +Tc) where Tc is the average time when
from DHD could range from $150 to $640 million per year. closed microvessels will be closed (Tc = Up) and directly related to the
value of the urea rebound. This model also shows that at the same level of
Conclusion
blood flow, the amount of urea retained in the tissue can vary significantiy
Different methodologies give rise to widely divergent cost due to vasomotion. Simple back extrapolation of urea changes between the
estimates but independent of this, the global cost of DHD may sessions to the end of the previous session provides an estimate of the Tc
be lower than that of the standard.3 times a week drill with its value.
unacceptably high morbidity and mortality costs. We can conclude that this mathematical model based on vasomotion can be
a useful tool for urea modeling and can explain observed clinical
phenomena.

WHAT “BLOOD VOLUME” DO WE DIALYZE?

M Kiaii, C Kianfar, P Heideneheim, Rh4 Lindsay
Optimal Dialysis Research Unit and London Health Sciences Centre,
London, Ontario, Canada
Blood volume changes [BV] during hemodialysis m] can be
done by optical monitoring of hematocrit [hct] eg by Crit-Line
monitor. The hct differs in different parts of the body. The arterial
hct is established after blood is mixed in the right heart. Only in the
cardiopulmonarycircuit [CP] is the hct uniform and representative of
the value measured in the arterial HD line. OBV by Cnt-line may
only measure OCP volume. The “central blood volume” [CBV] is
defined as the volume of blood in the heart, great vessels and lung
capillaries and can be measured by indicator dilution methods during
dialysis [Transonic HD monitor] from the cardiac output [CO] and
the indicator transit time. Changes in OBV and CBV should be
similar. The interrelationships of OBV and CBV plus those of CO
were tested during HD following periods of UF and then subsequent
vascular compartment refilling [R]. UF sufficient to cause OBV of -
6.7%[*2.4;n=21] was associated with a reduction in CBV by 0.1L
[*0.2] [relative fall 3.75%] and in CO of 0.94 L h i n [ i l . l ] [relative Time 1300 1330 1500 1700
fall 10.5%]. The changes in these parameters all significantly BUN 87 80 78 71
correlated [mnin 0.46, p m i n 0.0121. Following R to a OBV CREAT 12.9 12.2 11.6 11.3
increase of 2.4% [* 1.3, n=ll] CBV increased by 0.3 1 L [* 0.391, a GO2 18 19 21 23
relative change of +22% and CO by 1.07 L h i n [*1.61] a change of
+26%. The CBV and CO changes correlated [1=0.86, p=O.OOl] but
all correlations with OBV were lost . These data suggest OBV and
CBV refer to different “blood volumes” with changes moving in the
Same direction but at different rates. Changes in CO are closer
related to changes in CBV than 0 BV.
 ASAIO RENAL ABSTRACTS 219

RESISTANCE OF AV SHUNTS: IS IT CONSTANT DURING Direct Bioimpedance Parameters in Chronic Hemodialysis

HEMODIALYSIS? T. DeDner, V. Kislukhin, N. Krivitski University of (HD) Patients in Reference to the National Health and Nutrition
California, Davis, Transonic Systems Inc. Examination Survey 111 (NHANES 111)
Resistance (R) to flow in the AV shunt is usually considered to be
fixed and does not change during hemodialysis (HD). This means The nutritional status of HD patients is a significant
that R = (MAP - CVP)/Qa = constant (Eq.1). MAP is the patient's determinant of morbidity and mortality. Malnutritionsurveillance is
mean arterial pressure, CVP is the central venous pressure, and Qa usually.done using anthropometric techniques, dietary recall, and
is access blood flow. In clinical practice, only a surrogate of surrogate biochemical analyses. Bioelectrical impedance (BEI)
resistance (Rs) can be measured during HD (Rs = MAP/Qa). If a analysis allows direct measurement of total body resistance (R)
significant change in Rs is found during HD, and R in Eq.1 is and reactance (Xc) with subsequent calculation of total body water
assumed to be constant, then the CVP must also change. To (TBW) and body cell mass (BCM). More important, R and XC BE1
investigate the correctness of this hypothesis, paired measurements measurements have been obtained in the general population as
of Qa, cardiac output (CO) (HDOI, Transonic Systems Inc.), and part of NHANES 111. We have compared BE1 values in 154 HD
MAP were performed in 53 patients with grafts and in 27 patients with patients (age: 62 f 16 y; 46% female; 32% black) to those obtained
native fistulas at the beginning and at the end of the HD session. in NHANES 111. To reduce bias in age differences between study
The resulting data was separated into three groups according to groups, comparisons were limited to the 40-70years of age range.
the hypothetical change in CVP required to keep R constant during Female
dialysis: 1)unchanged: 0 2 5 mm Hg, n = 37;2)increased: 15 10 NHANES N
-!J NHANES !$2
mm Hg. n = 20;3)decreased: -18 12 mm Hg, n = 23.Neither Qa R (ohms): 477+41 494511 575+65 6045142
nor the type of shunt (graft vs. fistula) had an apparent effect on Rs. Xc (ohms): 63580 45+18 73f70 49+17
The change in access flow during dialysis correlated poorly with the R/H (ohms/m): 272 24 279 f 63 351 i 45 379 90 +
change in MAP (r = 0.37,p = 0.001,n = 80). In a subset of patients Xc/H (ohms/m): 40f14 38518 23+7 22+8
with heart problems, the correlation was better (r = 0.60,p < 0.001,n 0 P<O.OOOl when compared to HD.
= 41).CO correlated poorly with changes in access flow (r = 0.18,p = When compared by percentiles, only Xc parameters deviated
0,11 n = 77).but in diabetic patients the correlation was better (r =
I
significantly from NHANES 111 in HD patients by 65,44,35,24 and
0.52, p < 0.05, n = 20).
The'data in group 2 suggests that if resistance remained constant 4% at the 1 Oth, 25". 50m,75'hand 90" percentile respectively.
during hernodialysis CVP must have increased on average by 15 mm
Hg during the session. It is difficult clinically to accept such a Overall, HD patients have normal height adjusted R values and
hypothetical change considering that fluid removal during the session smal1er.X~parameters. These results are consistent with greater
was 3.5 0.7liters in this group. The correlation of Qa changes with TBW and lower BCM content in HD patients than in the age and
various cardiovascular pathology shows that these relationships are sex matched NHANES 111 population.
more complicated. It is more likely that in addition to the CVP
changes during HD, some patients experience vasoconstrictionthat Francis Dumler. Division of Nephrology, William Beaumont
increases vascular resistance in the AV shunt pathway. Hospital, Royal Oak, Michigan, U.S.A.

SOLUTE MASS TRANSFER AREA COEFFICIENTS IN CARDIAC OUTPUT IS DEPENDENT O N HEMODIALYSIS

A PEFUTONEAL DIALYSIS POPULATION ACCESS FLOW. S. Pandeva and R.M. Lindsay
The mass transfer properties of the peritoneal membrane are Th e long-term impact of hemodialysis access flow (Qa) o n cardiac
best characterized by solute mass transfer area coefficients output (CO) and cardiac function is not known. We have previously
(KBD). In this study we examined ,KBD values for urea, reported the correlation between Q a and CO during hemodialysis (HD)
creatinine and glucose in 109 peritoneal dialysis patients. The (ASAIO J 1999), but did not observe any decline in CO with transient
following procedure was employed for the 145 test exchanges. (1 minute) fistula occlusion. We had the opportunity to study the
Samples were drawn from the peritoneum for solute concen- effects of altering Q a o n CO in 2 patients using the Transonic HDOl
tration measurement immediately following infusion of 2L of Hemodialysis Monitor [Transonic Systems Inc.].
dialysate and at 15.30.60. 120, 180 and 240 min dwell time. We performed serial measurements of CO and Q a in a patient who
The volume drained at the end of the 240 min exchange was successively had one, two, and then one hemodialysis access. At
measured and a blood sample taken for plasma solute concen- baseline in November 1997 a left upper arm arterio-venous fistula
tration measurement. KBD values were extracted from these [LAAVF] was in place. A right upper arm synthetic gr& was inserted
data employing Garred's simple model for peritoneal mass January 1998, and the LAAVF was ligated in Sep 1998. The baseline
transfer. CO was 6.8 L/min with a Q a of 1.4 L/min. Following the insertion of
Mean KBD results *standard deviation were: the second access, the CO reached a maximum of 8.4 L/min with a
Urea *
23.6 10.0 a m i n (range = 10.4- 77.0) combined Q a of 4.1 L/min. Following the ligation of the first access,
Creatinine 10.4 2 7.0 mUmin (range = 2.2- 35.0) the Qa was 1.6 L/min with a CO was 7.5 L/min and 6.4 L/min at 1
Glucose 11.5 f 3.2 mUmin (range = 4.4- 30.3) week and 1 year post-ligation, respectively.
Four patients had 4 KBD evaluations spaced over time periods In a second patient with a Q a of 2.2 L/min in whom the diagnosis of
ranging from 18.4 to 42.4 months. The repeated K B D values for high-output cardiac failure was suspected, the LAAVF was occluded
all patients were consistent and did qot demonstrate any long for 15 minutes (min) during right heart catheterization. The baseline, 8
term change in peritoneal mass transfer capacity. min, and 15 min CO were 7.73, 5.96, and 5.71 L/min, respectively.
Plotting urea KBD versus glucose KBD demonstrated a Th e decline in CO was essentially equal to the Q a at baseline. The Q a
proportional relationship between the. two: during HD post-catheterization remained unchanged.
UreaKBD 1.78 glUCOSeKBD (R=O.76) CO significantly increases with the insertion of a hemodialysis access
and decreases with removal of the access. CO also decreases with
The proportionality coefficient of 1.78 is close to the 1.73
constant predicted from diffusion theory (KBD inversely occlusion of the fistula. Th e impact of high Q a on cardiac function
proportional to square rmt of molecular weight). The linear remains to be determined. Th e decision to band or ligate an access in
relationships found when creatinine KBD was plotted versus those suspected of having high-output cardiac failure should be
KeD for urea and glucose (R = 0.73, 0.88) had an unexpected individualized.
negative intercept. Further studies are needed to determine the impact of having a
markedly elevated Qa and the long-term implications of hemodialysis
These results indicate the variability of KBDamong peritoneal
dialysis patients and may be useful for kinetic modeling . access on cardiac function.
U Garred', W Turek', A Slingeneyer'.
Lakehead University', Lapeyronie Hospital'.
220 A S 1 0 RENAL ABSTRACTS

PRELIMINARY RESULTS OF A PHASE I TRIAL EVALUATING A NON- TRANSPORT CHARACTERISTICS OF DIFFERENT

CELL BASED EXTRACORPOREAL HEPATIC SUPPORT DEVICE HEMODIALYSIS MEMBRANES
Awad SS. Swaniker F. Alarcon W. Posner S. Haft J. Bartlett RH
Previous studies have demonstrated the importance of both
BACKGROUND: Toxins such as ammonia, aromatic ammo acids, free fatty acids and difhsive and convective (sieving) transport on the clearance of
bilirubin have been implicated as the cause of hepatic encephalopathy and cerebral clinically-significant middle molecules during hemodialysis.
edema in patients with acute liver falure. We have previously reported clearance of
these toxins, using a non-cell based extracorporeal hepatic support device (ECHS) based However, there have been few quantitative analyses of the
on the MARS concept HYPOTHESIS. Continuous Venn-venous hernodiafiltration intrinsic transport properties of available dialysis membranes.
with alburmn dialysis would: 1) decrease elevated levels of hepatic toxins, 2) reverse the In vitro experiments were performed using cellulose acetate
rauo of branched cham to aromatic amino acids (Fischer rauo), 3) reverse hepatic (CA 110 - Baxter), cellulose triacetate (CT110 - Baxter),
encephalopathy, 4) m a n t a n stable h e m o d y n m c s . METHODS: Patients with acute
hepatic failure who were UNOS Status I, IIA or 7 were placed on continuous veno-
polysulfone (F60 - Fresenius), and acrylonitrile copolymer
venous hemodiafiltrauon with countercurrent dialysis using a 10% albumin solution (AN69 - Hospal) dialyzers. Clearance and sieving data were
(ECHS). Mean arterial blood pressure (MAP), heart rate, SVO2, ICP, and hepatic obtained with urea, vitamin B,,, and dextrans as model solutes.
encephalopathy score were recorded at baseline and throughout treatment. Blood Simulated dialysis was done using donated human plasma.
samples were analyzed for total bilirubin. ammoma, aromatic and branched cham m n o Results were analyzed using membrane transport theory.
acids, lactor 7, and free fatty acids. Comparisons were made using Student’s t test
RESULTS: To date Six patients have been enrolled into this phase I study (Status I: Data for the CA, CT, and AN69 dialyzers were consistent
n=3, Status I I A 3 7 . ~ 3 ) During
. treatment there was no sigillficant change in MAP, with the homogeneous structure of these membranes. The
heart rate. or S V 0 2 when compared to baseline (p=NS) and there were no adverse effective pore size of the AN69 membrane was 508 larger than
mechanical effects There was a marked decrease in the ammonia in 5 of six patients that for the CT and 6 times larger than that for the CA. This
and a decrease in the total billrubin in 3 of six patients Also. there was a reversal of the
Fischer ratio in 4 of five patients. There was a trend of increasing hctor 7 levels in all results in much higher clearance for the AN69, particularly for
six patients. (Table I ) Three of six patients had a significant improvement in their large molecular weight solutes. Results for the F60 suggest a
hepatic encephalopathy score from HES=4 to HES=O (Status I: 1 of 3, Status 1 1 A 3 7 2 distinct 2-layer membrane structure. Quantitative estimates of
of 3). Of the three Status I patients, one recovered native hepatic function and one was the pore size of each layer were obtained using hydrodynamic
successfully bridged to transplantation. CONCLUSION: Preiimnary results suggest
that this ECHS device which utrlizes selective hemodiafiltration with alburmn dialysis is
analysis. The tight skin provides a large resistance to
effective in clearing some of the toxins that have been implicated as the cause of hepatic convective transport but has minimal effect on diffusion.
encephalopathy and cerebral edema There was an associated decrease i n the hepatic Exposure to plasma causes a significant reduction in
encephalopathy score with minimal hemodynamic effects As our experience increases, clearance for large solutes because of the resistance provided
the efficacy of this device as a bridge to transplantation will be better understood. by a protein layer on the membrane surface. The properties of
Table 1 PRE POST this layer depend upon the membrane and dialysis conditions.
UKOS
SWLVS
NH3 T
mllr
Fador
7
Fucher
Ratio
I113 T.
mnh
Factor
7
Fiwher
muo
A tighter protein layer formed on the F60 membrane, causing a
Paemti IIX-11 1 211 I? 07 12 1 S O j l 25 greater reduction in solute clearance than seen with either the
187 256 9 0s 41 ? I S 1 4 20
Patient2
Piben13
llX-17
1 145 229 13 06 IS? 239 I5 06 AN69 or CT dialyzers. The relative reduction in clearance was
Pltienfl
Pltirnl5
IIA-17
I
93
188
232
It1
8
3
10
pen4
5s
lii
3 0 3 1 8
11s s
I2
pen4
also greater at high ultrafiltration rates. The implications of
hucm6 I 188 17 I3 I? 27 62 41 2: these results to clinical dialysis are discussed.

S. Morti and A.L. Zydney, Department of Chemical

Engineering, University of Delaware, Newark, DE

EFFECT OF UREMIC TOXINS ON RED BLOOD CELL MODELING OF DYNAMIC CHANGES IN BLOOD VOLUME
MEMBRANE PERMEABILITY AND SEGMENTAL EXTRACELLULAR VOLUME DURING
Previous studies have identified specific inhibition of HEMODIALYSIS
membrane transport systems in uremia. For example, the red Previous studies have shown that blood volume (BV) measurements
blood cell (RBC) permeability to uric acid is reduced by more can be used to estimate vascular refilling from the interstitial
than a factor of 3 in dialysis patients, with a smaller reduction compartment during hemodialysis (HD). However, BV changes
for creatinine. However, the identity of the responsible uremic cannot predict extracellular volume (ECV) conditions. A further
toxins is unknown. parameter is needed to define ECV. A four-compartment model (BV,
This study examined the effect of several uremic toxins on Virunk,V,, and V,,,) that can be parameterized by segmental
RBC permeability to uric acid and creatinine. Normal RBC bioimpedance analysis (SBIA) is used to develop a relationship
from healthy donors were washed and rapidly re-suspended in
between BV and ECV (see Fig.1). kl
plasma or buffer containing uric acid or creatinine. The RBC
suspension was placed in a stirred ultrafiltration cell, with cell-
and protein-free samples collected through a semipermeable
membrane. Data were obtained in the presence/absence of
FHTE Fig.1

several uremic toxins. Limited studies were performed using UF

chromatographically-purified plasma ultrafiltrate fractions
obtained from hemodialysis patients. Intravascular refilling is postulated to occur by fluid transfer from
Xanthme and hypoxanthine caused more than a 2.5-fold two main compartments (central [trunk] and peripheral [arm and
reduction in RBC permeability to uric acid, but had no effect leg]). Intercompartment transfers are modeled as volume driven,
on creatinine. This behavior was consistent with data for the assuming relatively fixed compartmental compliances. A group of
chromatographically-purified uremic ultrafiltrate fractions differential equations are used to predict the fluid shifts during HD.
containing these solutes. The different behavior for uric acid SBIA is used to detect ECV changes in the trunk and the periphery.
and creatinine is due to the different transport systems used by Ten patients were studied during HD: ECV in the arm, trunk and leg
these solutes. p-Cresol caused a 70% reduction in RBC were measured by SBIA and BV was measured by an ultrasound
permeability to uric acid and a 50% reduction in permeability technique. Results indicate that refilling from the arms and legs may
to creatinine, consistent with the transport inhibition seen with have an exponential trend. A transfer coefficient (k, k l and k2) can
the lipophilic chromatographic fractions. These results provide be varied to obtain a best fit to the data. The point at which the
new insights into the uremic syndrome and have implications refilling curve tends to plateau seems to describe the best fit for dry
for uric acid and creatinine clearance during hemodialysis. body weight. A similar relationship is not seen in the refilling curve
for the central compartment.
S. Morti and A.L. Zydney, Department of Chemical F . ,E.F. Leonard*, C. Ronco, and N. W. Levin, Renal Research
Engineering, University of Delaware, Newark, Delaware Institute & Beth Israel Medical Center, *Columbia University, NY
R. De Smet and R. Vanholder, Department of Internal
Medicine, Nephrology Section, University Hospital, Gent,
Belgium
 ASAIO RENAL ABSTRACTS 22 1

ESTIMATION OF DRY BODY WEIGHT BY SEGMENTAL DIALYZER FLUX CAN ALTER THE DEPENDENCE OF MASS
BIOIMPEDANCE ANALYSIS DURING HEMODIALYSIS TRANSFER-AREA COEFFICIENT (m) ON DIALYSATE
FLOW RATE
A new concept of estimation of dry weight (DW) for The increase in urea (Ur) KJ when increasing dialysate flow rate
hemodialysis (HD) patients is proposed. DW is defined as the is dialyzer-dependent; however, the parameters that govern this
state of body hydration in which refilling rate from arm and leg dependence have not been studied extensively. We evaluated the
tends to dissociate from that from the trunk at the end of HD. effects of blood and dialysate flow rate on Ur and creatinine (Cr)
The study investigated the relative value of measurement of KJ values for low flux (Polyflux 14L) and high flux (Polyflux 14s)
changes in extracellular fluid volume (ECV), relative blood dialyzers containing polyamide membranes of identical surface area
volume (RBV) and systolic blood pressure (SBP) in assessment (N=S each). KJ values, calculated from the mean of blood and
of dry weight. Eleven patients were studied during regular 22 dialysate side clearances, were determined in vitro at zero net
HD sessions. ECV regional refilling was measured by ultrafiltration for 3 different blood (Qb) and dialysate (Qd) flow rate
segmental bioimpedance analysis (SBIA). RBV and SBP were combinations. All values are reported in ml/min, and KJ values for
measured at IOmin, intervals. Patients were divided into two each dialvzer and solute are reuorted as mean*SEM:
groups according to the presence of hypotension at the end of
&
-A
HD. (Grl n= 6, pts with hypotension; Gr2 n=16, pts without QdQ 14L-Ur 14L-Cr 14s-Ur
hypotension). Difference between estimated DW and post-HD 300/SOO 69W20 46W10 67W20 S2W10
(AW) and the changes between the final and immediately 4SO/SOO 67W10 4SW10 70W20 S3W10
previous measurement are indicated as ASBP, ARBV, AECVA,,, 450/800 72W10* 49&10** 80W20** 61W10**
AECVT,,~ and AECVL,, respectively. (*p<O.OS & **p<O.Ol, higher than that at Qd=SOO mumin)
G AW ASBP ARBV AECVk, AECV,. AECV,, KJ increased with increasing Q d but not increasing Qb; the
I 0.24k0.6 -15.21210 -1.8720.7 -0.46tl.7 -1.7820.9 -0.5320.33 increases in Ur and Cr KJI values were higher for the high flux
2 0.98t1 -1.87k13 -0.6k2.6 -3.56k5.3 -435k5.7 -2.6424.64 (16ilY0 and lSil%) than the low flux dialyzer (6i2Y0 and 7*2%).
We conclude that increasing Q d increases both Ur and Cr KJ values
The results indicate that ASBP was significantly greater and A and that this effect is more substantial for dialyzers containing high
ECVA, and AECVLe9were significantly less in group 1 patients. flux polyamide membranes. Whether this phenomenon is due to
This limitation of refilling from the periphery to the trunk in enhanced internal filtration for dialyzers containing high flux
group1 patients may indicate when dry weight is approached. membranes requires further study. This study suggests that increases
Development of the SBIA technique could provide a new in smaH solute clearances, and therefore dialysis dose, when
approach to the determination of dry weight in dialysis patients. increasing Q d can be more substantial for dialyzers containing high
F., C. Ronco, D. Schneditz, L. De Simone, and N. W. Levin, flux membranes.
Renal Research Institute & Beth Israel Medical Center, NY J.K. Leypoldt & A.K. Cheung. VA Medical Center & University of
Utah

EXAMINATION OF NEW SUPPLEMENTAL FLUID SETIWG METHOD IN DEVELOPING CATXETER-BASED SYSTEM TO MEASURE BLOOD
FLOW IN IIEMODIALYSIS GRAFTS DURING ANGIOPLASTY PROCEDURE
DOUBLE FILTR4TION PLASMAF’HERESIS (DFPP) D. Starostin, N. Krivitski, D. Gerardini, T. Vesely;
Transonic Systems, Inc., Malinkrodt Radiology Institute.
Masaki Kimikawq Kei Eguchi, Micbio Mineshima, Satoshi Teraoka, Telsuzo Agishi Effective angioplasty extends the potency of the A-V
vascular access. The purpose of the study was to develop
Kidney Center, Tokyo Women’s Medical University, Japan blood flow measurement technology that can give on-time
estimation for results of angioplasty procedure to
(Purpose) We tried establishment of effective and safety treatment method by appropriate ensure high quality of the intervention.
Methodology: Bench experiments and preliminary
supplemental fluid setting in double filtration plasmapheresis (DFPP). clinical trials were performed. Intravascular Blood Flow
Monitor (IBF’M) uses electrical impedance dilution
(Methods) Patients of pre and post kidney transplanlation received DFPP treatmentS under technique. To perform the measurements, 5F catheter with
electrodes was inserted into the graft directed to the
continvous blmd volume (BV) monitoring. The Crit-Line insmment was used as a BV venous anastomosis. 5% NaCl solution was infused at the
rate of 10 ml/min for 6-8 sec through the side arm of
change monitor. We estimated the relation with patient’s BV change rate and quantity of the catheter introducer. The changes in impedance were
recorded by the IBEM. Conditioned signals were further
supplemental flui4 albumin concentration of supplemental fluid and serum albumin digitized and sent to a PC via a serial port for
automatic data interpretation.
concentration before DFPP. Furthermore from these results, we established new Results. The two main sources of errors discovered on
the bench were bad mixing and penetration of the
supplemental fluid settings methods with a condition of BV decrease rate less than 10 % electrical field out of the vessel. When electrical
field was mostly located inside the vascular access,
and IgG removal rate 70 %. And then we applied it to clinic. correlation between dilution and volumetric measurements
was r-0.98 (n=32). In clinical trials reproducibility
(Results) Loss of semm albumin was recognized accompanied with removal of globulin (relative difference between two consecutive
measurements performed within 3 minutes) was 4 . 2 ? 4 . 3 %
and decrease of BV due to hrpoproteinemia was observed Decrease of BV was
(n=70). Pacing catheters allowed by the FDA for clinical
remarkable when the volume of supplemental fluid was large, albumin wncentration of
trials. gave satisfactory relative measurements, but
failed to give high absolute accuracy due to penetration
supplemental fluid was low and serum albumin concentration before treatment was hi&.
of the electrical field out of the vessel.
Bench experiments and clinical trials demonstrated
New supplemental fluid setting methods wa established from these parameters and was
that catheter-based system allows reliable determination
of intra-access blood flow changes during the
applied to clinic. BV decrease rate was 10.0 21 6.0 % and IgG removal rate was 69.6 z
angioplasty. Further development is now devoted to
improving the absolute accuracy of technology by
8.9 90. optimization of the catheter configuration.

(Conclusion) It seems to be possible that we can enforce an effective and safety treatment

with a little BV change by the new established supplemental fluid setting methods.
222 ,45210 RENAL ABSTRACTS

COMPARATIVE EFFECTS O F ANTI-THYMOCME GLOBULIN CONCENTRATED CITRATE (23%) FOR CATHETER LOCK
(ATGAM) VS INTERLEUKIN-2RECEPTOR MONOCLONAL SOLUTION
ANTIBODY BASED INDUCTION THERAPY ON CADAVERIC RENAL Ash!.'Mankus .', RA'.', Suvon JM'.', Smeltzer B',Ing T"
TRANSPLANT OUTCOME. Sunder M. Lal, Nilakshi Gupta, Gilbert 'Greater Lafayette Health Services, Inc.
Ross, Jr., Departments of Medicine, Anesthesiology, Surgery. 2ArnettClinic, Lafayette, Indiana
University of Missouri Health Science Center, Columbia, Missouri. 'HemoCleanse, Inc., West Lafayette, Indiana
Mondlor polyclonal antibody based induction therapy was introduced T.Jniversityof Illinois, Chicago, Illinois
to delay the introduction of cyclosporine in the setting of ischemidand o The standard anticoagulant "lock" for cuffed central venous dialysis
reperfusion injury to the allograft kidney. Recently newer agents like catheters is 5-10,000 units of heparin in each lumen. Sodium citrate is an
Mycophenolate Mofetil (MMF), Rapamycin. and the chimeric interleukin anticoagulant with intrinsic antibacterial activity (at hypertonic
2 receptor monoclonal antibody ((IL]-2Rmab} basiliximab and concentrations), and only transient anticoagulant effects if accidentally
Daclizumab), have been used for induction therapy. infused to the patient. Prior studies of citrate as catheter lock solution
We retrospectively, compared the safety and efficacy (decreased utilized low citrate concentrations (14%)in combination with high
incidence of rejection and the need for dialysis) of induction treatment concentrations (27 mg/ml) of gentamicin. We performed a prospective
with ATGAM (Group I) v s IL-2Rmab (Group II) in the recipients of study comparing concentrated citrate lock solution to heparin in a dialysis
cadaveric renal transplant (CRT). All patients received standard doses unit with SO-60% of patients having chronic central venous catheters (40
of Azathioprine or MMF. and prednisone. All patients received either catheters total, mostly Ash Split Cath" with some Tesio' and Hickmano).
cyclosporine (2mglkg) or FK (0.05mglkg) as a single daily dose. Heparin was the standard catheter lock. In 341011th intervals we
Standard d o s e s of CYA or FK were introduced when creatinine (Scr) implemented: 10% citrate with 3 m g h l Gentamicin@,20% citrate with 3
decreased to ~2.5mgldL. mg/ml Gentamicin, 47% citrate alone, heparin, and 23% citrate alone.
Both groups were comparable with regards to the age, sex, HLA The incidence of bacteremia in all patients in the unit was calculated and
matches, cold ischemic times, and panel reactive antibody titen. Both compared to the baseline incidence during use of heparin. In patients with
ATGAM and the ILZRmab were administered per standard protocol. bacteremia, citrate catheter lock was continued during treatment, and the
Results: Mean " SD. Five patients in Group I and two patients in Grou catheter was not routinely removed. Incidence of bacteremia was 4.6% of
II needed hemodialysis in the first 72 hours. Acute allograft rejection patients!month with heparin, and decreased to 1.9% with 10% and 20%
w a s s e e n in six patients (Group I) and in two patients in Group II citrate with gentamicin, and to 0% with 47% citrate alone (P<O.OOI).
(P=NS.) The incidence of bacteremia increased on return to heparin (2.7%). then
Group I (ATGAM): 1 Month 3 Months 6 Months decreased monthly during use of 23% citrate (2.1% average, P<0.05).
Scr(mg/dL. n=15) 3.3 2.1
" 2.6"l.l 2.3" 1.0
Group II (IL-ZRmab): 1 Month
Use of urokmase for occluded catheters significantly decreased with
3 Months 6 Months
Scr(mg/dL,n=13) 1.9 " 1.3' 1.6 " 0.4' 1.7 " 0.4' implementation of citrate for catheter lock (P=0.02), and use of tPA was
*p<0.5between Groups. infrequent after urokinase became unavailable. Catheter survival in the unit
Compared to Group I, Il-2Rmab treated patients showed (1) lower for Split Cath catheters was 83% at 1 year. Side effects were nonexistent
incidence of allograft rejection, (2) lower incidence of ATN, decreasing with 23% citrate lock and blood entry to catheters was rare. Concentrated
need for HD (3)significantly lower mean Scr, and (4) w a s well tolerated citrate is an effective catheter lock solution that provides prolonged central
venous catheter use with diminution in catheter-related infections and
occlusion by clot.

ACCESS RECIRCULATION AND THE USE OF VENOVENOUS The Duo Split Acute Hemodialysis Catheter: a Two-Limbed Acute
DIALYSIS CATHETERS Hemodialysis Catheter with Novel Technique for Insertion
m, Kianfar C, Lindsay RM Mankus RA', Ash'.' , JM', Tim Schweikert'
Sutton
Optimal Dialysis Research Unit, The University of Western Ontario,
London, Canada 'Arnett Clinic, Lafayette, IN, 'HemoCleanse Inc, West Lafayette, IN;
Access recirculation (AR) is important as it can diminish the adequacy of 'MedCdmp Inc, Harleysville, PA
dialysis. It is believed that AR is common in veno-venous catheters. Acute hemodialysis catheters are designed to be placed over a guidewire
However, the frequency of AR with the use of such catheters has not been into the central veins by Seldinger technique, and therefore are rigid with a
established. pointed tip. The guidewire technique avoids the need for tunneling the
We studied 24 patients with catheters (19 Perm-caths, and S temporary catheter and threading the catheter tip through a thin-walled "sheath" with
catheters). AR was measured by the ultrasound dilution (Transonic) risk of blood loss or air embolus. Acute catheters generally provide lower
technique. Qb was increased to maximum rate (venous pressure < 250 and more irregular blood flow than chronic tunneled catheters such as the
mmHg, arterial pressure > -250 mmHg). Qb ranged from 300 - 450 mumin Ash Split Cath, and average duration of use is limited to a only a few days
(average 420 mumin). With catheters in nonnal position, 2 people had non- due to risks of infection or vein irritation. The Duo Split is a flexible acute
significant AR (4%), 3 had significant AR (>5%), and the rest had 0 catheter with two separate intravascular limbs, each with a blunt cylindrical
recirculation. AR ranged from 7% - 5 1% (average 18%) with catheters in tip and multiple side-holes similar to the Split Cath, but without a cuff and
reverse position. subcutaneous tunnel. A single-slotted "Quill" guide surrounds the catheter
In conclusion, AR is rarely present when veno-venous catheters are used
during placement, with a tapered tip that follows the guidewire into the
in the normal position. If AR is present, it usually implies that there is
vein. Simple traction on the Quill causes the tapered tip and the slot to
catheter malfunction such as thrombosis. Reversing the lines results in
some recirculation, varying with the individual, and the Qb. Therefore, to expand around the catheter, leaving the Duo Split in place within the vein.
ensure optimal dialysis adequacy, care must be taken in using veno-venous DuoSplit catheters were placed into the femoral vein (4) and IJ vein (2) of 6
catheters in reversed position, and maximal possible Qb should be used patients with acute renal failure. During dialysis blood flow rates (BFRs) of
with the lines in the normal position. 200-300 mumin were obtained (average 229 mumin) with average venous
pressures of 150 mm and average arterial pressures of 192mm.Flow rates
and pressures were similar to conventional acute catheters placed mostly in
the IJ vein. Duration of Duo Split catheter use was longer at 6.2 days
(maximum 11 days) versus 4.0 days for conventional catheters. There were
no complications of clotting, septicemia, exit infection or loss of blood
flow. The DuoSplit catheter in the femoral or IJ vein provides adequate
blood flow rate for treatment of acute renal failure, and may allow longer
catheter use than with conventional acute catheters..
 ASAlO RENAL ABSTRACTS 223

LIVER DIALYSIS IN TREATMENT OF HEPATIC FAILURE AND SURVIVAL AND HYDRAULIC FUNCTION O F THE ASH SPLIT
HEPATORENAL FAILURE: RANDOMIZED CLINICAL TRIALS CATWMHEMODIALYSIS CATHETER
AND CLINICAL EXPERIENCE Stephen'R. Ash',2, Rita A. Mankus', James M. Sutton'
T Kuczek4, DE Blake2, CH Gingrich' 'Arnett Clinic and GLHS, Lafayette, IN; 'Purdue University, W. Lafayette,
Purdue University. 'Comparative Medicine, 'Dept of Statistics; IN; 'HemoCleanse, W. Lafayette, IN
'HemoCleanse, Inc., W. Lafayette, IN; 'Amett Clinic, Lafayette, IN
The Liver Dialysis Unit is a currently marketed artificial liver that employs The Ash Split CathTMis a dual-lumen tunneled hemodialysis catheter with
hemodiabsorption (dialysis of blood against powdered sorbent) to remove a single double-D transcutaneous portion connecting to multi-holed
the numerous small molecular weight toxins of hepatic failure. A cylindrical tips in a central vein. Blood enters and exits from the entire
randomized, prospectively controlled study of Liver Dialysis in treatment of circumference of the tips, similar to Tesio catheters, but insertion is through
hepatic failure was performed in 5 centers. Entry criteria were: decrease in a single venous puncture. Since 1998, 125 Split Cath catheters have been
hepatic function associated with Stage 2-4 encephalopathy, renal failure or placed in ESRD patients in our practice, mostly by Nephrologists, with
respiratory failure allowed, but not already on dialysis or C W H Another average follow-up of 1-18 months (mean 8). Citrate has been the most
28 patients were treated in crossover studies. . Most patients had stage IV common catheter lock (10-47%, mostly 23%). Only a few catheters have
encephalopathy, renal insufficiency, and respiratory insufficiency or failure. been removed due to septicemia, exit site infection, or loss of blood flow
Of the 75 patients, 32 had fulminant hepatic failure (FHF), and 43 had rate, and interventions have been rare. Lifetable analysis indicates 83%
acute-on-chronic failure. Liver Dialysis (6 hrs, 3-5 days) significantly one-year catheter survival. Average pressures and flows were
improved neurologic (70%) and physiologic status (72%) of all patients automatically determined during one month of dialysis procedures. For
vs. control groups (27% and 10% respectively). Liver Dialysis both arterial side and venous side, the Split Cath provided the same flow
significantly improved patient outcome (improvement of condition for rates and hydraulic resistance as needles in a graft or fistula (example,
transplant or recovery of liver function) for patients with A-on-C (57%) vs. graph below) Tesio catheters were similar, hut flow rates for Hickman
control patients (36%), though not for FHF patients. In 32 patients with catheters were lower. Recirculation percentage was rarely above zero for
hepatic and renal failure on entry, Liver Dialysis significantly improved the Split Cath. The fistula is the best chronic access method for ESRD
patient outcome (41%) vs. control patients (0%). Since market patients. For patients in whom fistulas are not workable, catheters such as
introduction, more than 50 patients, mostly A-on-C have been treated in the Split Cath can provide longevity and hydraulic function at least equal to
many hospitals for hepatic failure, with 58% positive outcome, confming A-V grafts.
the above studies. The suspension is easily modified by adding dry
powdered chemicals to the sorbent bag. Branched chain amino acids
added to the sorbent result in a greater increase in Fischer's ratio during
treatment, and new ammonium sorbents increase the removal of
ammonium. Clinical effects of Liver Dialysis in treatment of hepatic failure
may be further hastened and augmented by these improvements.
0 100 200 300 400
MUMIN

EVERY-OTHER NIGHT HEMODIALYSIS (QOHD) WITH PLATE SILICONE CUFFED DUAL LUMEN CATHETERS FOR
DIALYZER AS BLOOD PUMP AND SINGLE-LUMEN ACCESS JBMODIALYSIS VASCULAR ACCESS: A TWO YEARS
Ash'.' ,DJ', Harker K', Truitt RBI,Korkor A'
Carr EXPERIENCE.
'HemoCleanse, Inc., W. Lafayette, IN; 2Greater Lafayette Health Systems, Casani A, Andriani F, Betti GC, Szorenji E, Palla R
Lafayette, IN; 'Purity Dialysis Centers, Waukesha, WI Nepbrology and Dialysis Unit. City Hospital. 54100 Massa Italy
Extended duration dialysis (8 hours, 3 or more times per week) can
diminish morbidity and mortality of hemodialysis. What is needed is a We present our two years experience using the Permcath catheter,
hemodialysis system with simple setup and breakdown, intrinsic safety, and Quinton InsUument.The percutaneous introduction was performed by a
automaticity so it can operate overnight unattended and without bothersome nephrologist under ultrasonic real-time guidance, always in internal
right jugular vein.
alarms. The HHD is a dialysis system with pressure-controlled blood
This catheter was placed in a total of 81 patients for plasmapheresis
pumping; expansion of plate dialyzer membranes draws blood through a 10
(7 pts) and ESRD pts : f i s t choice ( 21 pts), fistula maturation (43 pts),
French single-lumen access and compression returns the blood. Direct
exhausted vascular access (10 pts).
blood flow measurement and automated fluid boluses and fmal rinse make In the 21 first choice patients the median life of the catheter was 5
the HHD considerably easier to use than standard hemodialysis machines. months: 7 patients dead with the functioning catheter. one patient was
The HHD was approved for market in treatment of renal failure using a transplanted. In remaining 13 patients the mean survival of functioning
Sorbsystem column, and clearances were only slightly less than standard catheter was 11 months. In the fistula's maturation group the median
dialysis. A clinical trial is beginning with the HHD in providing 8-hour permanence of the functioningcatheter was 3 months. In the exhausted
QOHD in-the home and in a nursing home. For this trial, dialysate is vascular access group the mean life of catheter was 6 months: 6 patients
created by adding dry salts and acid to 100 liters of purified water produced dead and one patient was transplanted with still functioning catheter.
by an RO system actuated by household water pressure. All blood-side and The 3 catheters Still in use are functioning from a mean of 7 months.
dialysate-side components are discarded after each treatment, except the The blood flow was 300-350 d m i n in all cases and the recirculation
100 liter dialysate tank (air-dried after each treatment). The study will calculated by the DOQI guidelines never was over 10%
determine: time needed for setup and breakdown of the machine, ability of Insertion complications: one hemotorax and 8 cases (9% for 90
the HHD to perform all functions of the dialysis procedure, and chemical insertions) of subcutaneous hemorragic soffision
and clinical effects of the treatments. The HHD may allow QOHD to be We observed 14 episodes of catheter failure starting treatment over
routinely performed unattended after machine setup by the patient, a 5050 treatments (0.3itreatment) Urokinase instillation completely
partner, or a dialysis technician. The resulting dialysis treatment should not recovered the normal function in 11 cases In 3 cases the substitution of
only be highly chemically effective, but economically practical. catheter was needed.
Five patients (6%) experienced episodes of fever and gram-positive
bacteriemia requiring catheter removal (without further
complications).
In our clinical experience Permcath catheters is an efficient and safety
intermediate-duration vascular access for hemodialysis. prolonged
apheresis treatment and to allow maturation of A-V fistulas, and it is a
good long-term access in patients with exhausted vascular access.
224 ASAIO RENAL ABSTRACTS

C R E A T E SUPPLEMENTATION IN HEMODIALYSIS HAEMOFILTRATION WITHOUT ANTICOAGULATION OR SALINE

FLUSH IS A SAFE AND PRACTICAL APPROACH IN SELECTED
PATIENTS. Steven Kirschbaum and B m Kirschbam. Virginia CRITICALLY ILL PATIENTS WITH ACUTE RENAL FAILURE
Commonwealth University, Richmond, VA.
Han Khim TAN', Ian BALDWIN-, Rinaldo BELLOMO"
End stage renal failure patients receiving hernodialysis frequently
complain of fatigue, muscle aches, weakness, lack of endurance, .
* Department of Renal Medicine, Singapore General Hospitai
Departmentof Intensive Care, Austin and Repatriation Medical
Centre, Heidelberg, Melbourne, Victoria, Australia
and cramps. Several studies have suggested that creatine
supplements can improve peak performance and stamina in healthy Continuous venovenous haemofiltration (CWH) is established
athletes without enough conversion of creatine t o creatinine t o treatment for acute renal failure (ARF), particularly in critically ill
patients with multi-organ failure (MOF). Various technical factors
raise the blood creatinine concentration. We conducted a study t o impact on its successful delivery. One such factor is
determine the safety and potential benefit of low dose creatine as anticoagulation. This is often complicated by haemorrhage in
high risk patients and can cause heparin-induced
a low cost dietary supplement. Six patients were enrolled. They thrombocytopaenia (HIT) when this is the anticoagulant used.
were evaluated by psychological testingusing the Kidney Disease This was a prospective study of 12 critically ill patients admitted
to the ICU of the Austin and Repatriation Medical Centre with
Quality of Life Short Form, stationary bicycle riding, and blood MOF and ARF (mean APACHE II and SAPS II scores: 28.3 +3.4, I-
tests before, during and for three weeks after taking2 gcreatine 60.5 +I- 5.9). All of them required CWH for renal replacement
therapy. Mechanical ventilation and vasopressor support were
per day for 1 month. The way in which the patients viewed needed in 58 % and 75 % of the cohort. The overall mortality was
themselves and their health did not change significantly 42 %. The mean duration of ICU stay was 10.3 +/- 1.6 days, of
which 8.2 +/- q.4 were days during which CWH was used.
throughout the 8 weeks of the study. Four of six patients Some of the patients were at high bleeding risk because they
improved their exercise performance. Of the two who did not were in the immediate post-operative period. Others were
coagulopathic (aPTT 61.3 +I- 7.0 sec, INR 1.9 +I- 0.2) andl or
improve, one took only 9 of the intended 34 doses of creatine and thrombocytopaenic (126500 +I- 28 730/mm') at baseline. During
the other took creatine sporadically and not according t o protocol. the course of CWH, they remained coagulopathic andl or
thrombocytopaenic p P l T 48 +3 I-sec, INR 1.8 +I- 0.1 and platelet
Electrolytes, BUN, creatinine, phosphate, liver hnction studies, 72 600 +I- 82391 mm ). The peak serum urea and creatinine prior
and hematologic profiles were not sipftcantly altered by creatine. to CWH commencement were 19.6 +/- 3.8 mmoll L and 264 +49 I-
No patient demonstrated intolerance to the supplement. We mmoU L. Adequate uraemic control was evidenced by the mean
serum urea and creatinine levels during CWH therapy of 46.0 +/-
conclude that 2 g creatine daily given to dialysis patients had no 1.1 mmoU L and 198 +12 I-mmoU L.
important adverse effects, and for some, appeared to help their The mean circuit lifespan using pre-dilution CWH and
polyacrylonitrile AN 69 hollow-fibre haemofilter, with neither
fatigue and improve their physical endurance. Based of these anticoagulation nor saline flush was 32.0 +I- 5.4 h, of which 43 .X
preliminary findings, a larger study of longer duration seems lasted more than 30 h. Haemofllter lifespan was not significantly
correlated with any of the standard coagulation indices.
justified. However, there was an inverse correlation approaching
statistical significance between platelet count and haemofilter
lifespan ( ~ 0 .0 5 4 3 ) .
Conclusion: Predilution CWH with neither anticoagulation nor
saline flush in critically ill patients at high bleeding risk is a safe
and practical approach.

Pressurized Dialysate and Ultrafiltration Management : PRINCIPLE FOR A CHRONIC PERCUTANEOUS LEAD
a new concept for Short Daily (Home)Hemodialysis. Objective of this study is to explore the feasibility of a new
J. Traeger*, R. Galland*, C. Gharib**. P. Barthez**, L. Juillard****,M. design principle for percutaneous leads.
Ladle****, D. Raclet*. C. Mas***.R. Hadden***. A percutaneous lead is designed for an enduring connection
* AURAL. Lyon ** Physiology lab. C1. Bernard University Rockefeller, between the external world and the inner space of the body.
Lyon *** FLAVIN LTD **** H6pital E. Hemot. Lyon This is of importance for patients, who rely on such a
Purpose of study :
connection in form of a tube for a diagnosis or a therapy.
Ultrafiltrationrate is determined by transmembrane pressure. Such an application is, for example, the ambulant peritoneal
which is classically controlled either by flow measurement or by dialysis. Another application is the drive line tube of an
volumetric control. A new technique is described that directly manages mechanical cardiac assist system. All these applications are
dialysate pressure and so the transmenbrane pressure. endangered by inflammation, infection, marsupialization and
extrusion. Often percutaneous-leads become so severely
Methods : infected, that they have to be removed completely. These
Dialysate pressurisation is used to control both dialysate flow and trans- problems arise, because the infection can creep along the
membrane pressure. The following equation can be used to calculate the artificial material of the tube. Once inside the body, the
required pressures to give a desired dialysate flow and ultr'atiltrate : T = infection is difficult to fight. The new principle counteracts
Pd + Pb - (UFKUF)where T tank pressure. Pd = pressure to generate the the migration of the infection by a growing sleeve. This
dialysate flow. Pb = pressure in the blood compartment of the dialyser, sleeve separates the tube from the surrounding skin. The
UF = ultraliltrate. KUF = dialyser UF coefficient growing sleeve imitates one prominent feature of natural
percutaneous structures such as fingernails: Growth from the
Results inside of the body to the outside.
This is achieved by pumping liquid silicone rubber though
Tank Pressure in an extrusion port inside of the body. The water vapor within
40mmHgunits 9.1 11 13.6 15.1 16 16.2 18.7 the body solidifies the liquid silicone, which then becomes
Dialysate 128,X 154.8 181,2 210,5 205 208.8 234 the sleeve. The silicone rubber is selected for the attachment
UF 12.6 22.2 19.6 45 20.8 11.6 19.3
YOerror UF 1% -13% -6% -11% 0% -12% 3%
of cells. The bacterial biofilm, which tends to form at the
location, where the sleeve exits the skin, can by this action be
This has been done using a vely simple pre-prototype dewce. Further. this pushed out of the body.
device was used to conduct three days of hernodialysis on an aneplmc The technical feasibility has been shown in several lab
pig. In five hours of use over three days. the ultraliltrate was found to be models. Major problem has been the design of the port, and
accurate to 0.4 %, and dialysate accurate to between 0.8 %n to 2.7 % of the selection and modification of the silicone rubber to ensure
target volumes. the production of a straight and leak free sleeve.
Affeld K., Bagheri, M. -
Charite, Biofluidmechanics Laboratory. ..Humboldt University
Berlin, Berlin, Germany
 ASAIO RENAL ABSTRACTS 225

MULTIORGAN SUBSTITUTION WITH THE HOCHFLElSCH@ DO DIFFERENT DlALYTlC TECHNIQUES SHOW

BlONlClZERIN A HOME SElTING. Full J. Fraude, Sum Grand DIFFERENT ATHEROSCLEROTIC AND ANTIOXIDANT
Hoax, Earnest T. Trye. Departments of Biomedical Engineering ACTIVITIES?
and Medicine. Tertiary Medical Center and Hospice, Ground
Zero, New Mexico. (Introduced by EA Friedman) To compare the chronic effect of several dialytic techniques (BHD,
Reimbursement limitations imposed by HMOs necessitate AFB, HDF, PFD) on atherosclerosis and on antioxidant activity, two
innovative restructuringof medical services. Following rejection different indexes were created. The first (atherosclerotic index = Al)
is formed using the sum of 3 plasma substances: MDA, HCY and
of payment for in-hospital long-term organ substitution, we
CYS (malondialdehyde, homocysteine, cysteine). The second
effected substantial cost savings and staff reduction with (antioxidant activity index = AOAI) is the sum of 5 erythrocyte (E)
improved outcome by resort to home treatment. parameters: E-GSH, GPx, CAT, SOD, GR (=E-glutathione, E-
We studied 86 consecutive patients (50 men, 36 women) with glutathione peroxidase, E-catalase. E-superoxide dismutase, E-
multiorganfailure age 79.8 yrs (51 to 107 yrs) in an ICU because glutathione reductase).
of: head trauma 21 (24.4%), CVA 31 (36%), medical or surgical These indexes were calculated in 20 controls and in 51 chronic HD
error 24 (28%), unknown 10 (11.6%). All were intubated, 57 patients. before, during and after the first session of the week. All
(66%)were unresponsiveto pain and none were orientedto time patients had been treated with a given technique for at least 12
or place. Each subject had at least three failed systems months.
In the table, values before (B) and after (A) session are shown.
including respiratory, cardiovascular, and renal failure. Patients
Al AOAI AI+AOAI
were continuouslyattached to a Bionicizer (previouslydescribed
compact ventricular assist pump +flat plate oxygenator+ hollow (")[ B A I B A I B A ]
fiber hemodialyzer). Membrane surface area of the HochfIeischCO
CONTROLS (20) 204 - 2381 - 2585 -
BHD (35) 554 337 1029 1756 1583 2093
Bionicizer is 2.3 M2(approximatesalveolar + glomerular surface AFB (5) 542 292 781 1792 1323 2084
areas in a 70 kg man).
HDF (') 493 289 1109 1497 1602 1786
Preparatory Bionicizer treatments were sustained for 48 hrs PFD
and continuedfrom thrice daily to alternate days for 7 days. After
(4) 504 254 1099 2219 1603 2473
HD-MEAN (5') 523 293 1021 1816 1544 2109
one week, 73 (84%) patients were alert and lucid, 10 (11.6%)
continued moribund, and 3 (3.4%) died. Home Bionicizer
treatments were successfully performed by 43 (50%) patients
assisted by a family member. All of 4 regional HMOs agreed to
fund home but not facility Bionicizer treatments.
We conclude: 1. In advanced multiorgan failure, novei
reimbursement solutions may be feasible. 2.lrreversible
multiorgan failure need not obstruct full rehabilitation.

PLASMAPHERESIS (PL) IN THE RX OF SEVERE

HYPERLIPIDEMIA (LIPIDEMIA), PANCREATITIS,
(PANCR) AND ADULT RESPIRATORY DISTRESS
SYNDROME (ARDS) DURING PREGNANCY (PREG)
Yudis. M.. Ghantous, V., Sirota. R.. Stein,
H., Snipes, E.. Gronich, J., Collins. D.,
Snyder, R.
We have recently seen a patient (PT)
present w/severe PANCR and hypertrigly-
ceridemia (hyper-TG) who developed ARDS. PL
was given to reduce her lipidemia. Her
course slowly improved regarding her PANCR
and ARDS.
Our PT is a 26-yo female with Type V
lipidemia who presented 23 weeks pregnant
with severe PANCR, hyper-TG, and ARDS.
Lipitor had been recently d/c'd. BUN 5,
creatinine 0.6, cholesterol 555, trigly-
cerides 3,000, lipase 3,075. Because of the
massive lipidemia, PANCR, and early ARDS, PL
was initiated.
Pre-PL#l Pre-PL#2 1 day p PL#
Cholesterol 555 189 111
Triglycerides 3,000 738 262
PL was felt indicated to reduce the
massive lipidemia. A major technical problem
during the'PL was the machine shutting down
probably due to the lipemic plasma. Clinical
improvement followed the PL. Anti-lipidemic
and dietary therapy, as well as
consideration for repeat PL will be
necessary during the remainder of her PREG.