BRIEF COMMUNICATION

Internal filtration in dialyzers with different membranepermeabilities

Yuichi Sato • Kenjiro Kimura • Tatsuya Chikaraishi

Received: 23 October 2009 / Accepted: 23 March 2010 / Published online: 5 June 2010

� The Japanese Society for Artificial Organs 2010

Abstract Over the last decade, hemodialysis with

enhanced internal filtration (IF) has been investigated as an

alternative to conventional dialysis. Several factors affect

IF, including the geometry and permeability of hollow-

fiber dialyzers. Although various studies have been

performed, the association between IF and membrane

permeability has not been fully examined because of the

difficulty in measuring IF. Therefore, in this study, we set

up an experimental circuit and attempted to directly mea-

sure IF as well as membrane permeability in five dialyzers.

In the circuit, we placed two dialyzers of the same type in

series, and a special sampling port between them, thereby

making it possible to determine IF by measuring the extent

to which blood was concentrated between the two dialyz-

ers. We showed that a significant amount of IF occurred in

this tandem-dialyzer circuit, ranging from 23.5 to 100 ml/min,

which increased linearly with increasing membrane per-

meability. We also showed that membrane permeability

was reduced in the first dialyzer to a greater extent than in

the second one after four hours of circulation, suggesting

that filtration caused substantial membrane fouling. In this

study we practically demonstrated that membrane perme-

ability is highly relevant to the phenomenon of IF.

Keywords Convective solute removal �Internal filtration � Dialysis

Introduction

Internal filtration (IF) in a hemodialyzer has been vigor-

ously investigated because hemodialysis with enhanced IF

dialyzers may serve as a clinical alternative to conventional

dialysis [1–6]. Previous studies have identified several

factors affecting IF, including the geometry and perme-

ability of hollow-fiber dialyzers. In theory, hollow fibers

with greater membrane permeability induce larger amounts

of IF, yet the relationship between them has not been fully

examined because of the difficulty involved in measuring

IF. In this study, we set up an experimental circuit

involving two dialyzers in series with a blood sampling

port between them. The circuit enabled us to directly

quantify IF by determining the extent to which blood was

concentrated by the IF phenomenon. We report the amount

of IF in several dialyzers with different membrane per-

meabilities but with almost identical geometries.

Materials and methods

Membrane permeability in dialyzers

We employed five cellulose triacetate dialyzers, FB-150G,

FB-150U, FB-150F, FB-150UH, and FB-150FH (Nipro,

Osaka, Japan). These dialyzers have nearly identical

geometries in their hollow fibers, as shown in Table 1. We

determined the pre-circulation ultrafiltration rate (UFR) of

each dialyzer as a marker of membrane permeability. We

placed each dialyzer in a circuit filled with reverse osmosis

Y. Sato (&) � T. Chikaraishi (&)

Department of Urology,

St. Marianna University School of Medicine,

2-16-1 Sugao, Miyamae-ku, Kawasaki 216-8511, Japan

e-mail: sato@marianna-u.ac.jp

T. Chikaraishi

e-mail: chik@marianna-u.ac.jp

K. Kimura

Department of Hypertension and Nephrology,

St. Marianna University School of Medicine, Kawasaki, Japan

123

J Artif Organs (2010) 13:113–116

DOI 10.1007/s10047-010-0506-z

(RO) water. While performing straightforward ultrafiltra-

tion at three different rates (Qf = 900, 1,800, 2,700 ml/h),

with two dialysate ports open to the atmosphere, we

measured the blood-side and dialysate-side hydraulic

pressures by a manometer, before and after the dialyzer, to

calculate the transmembrane pressure (TMP). We then

determined UFR using the following equation:

UFR (ml=h mmHg) ¼ Qf (ml/h)=TMP (mmHg)

TMP : ðPBinþ PBoutÞ=2 � ðPDinþ PDoutÞ=2;

PBin: blood-side pressure at the blood inlet, PBout: blood-

side pressure at the blood outlet, PDin: dialysate-side

pressure at the dialysate inlet, PDout: dialysate-side pres-

sure at the dialysate outlet.

After obtaining three sets of data for each dialyzer, UFR

was determined by a linear approximation using Microsoft

Excel X for Mac.

Determination of IF in the experimental circuit

To quantify the amount of IF in the dialyzers, we set up a

special circuit containing two dialyzers of identical type

that were placed in series with a blood sampling port

between them. There we circulated bovine blood at

200 ml/min from a 4-l reservoir and delivered dialysate at a

rate of 500 ml/min in the countercurrent manner with zero

net filtration. Anticoagulation was achieved by adding

heparin at 1,000 unit/h, and the hematocrit of bovine blood

was adjusted to 30–34% using normal saline.

In this tandem-dialyzer circuit, where the effective

length of the hollow fibers is doubled, the resultant increase

in TMP induces greater amounts of IF. We assume that a

substantial amount of filtration occurs in the first dialyzer,

which is counterbalanced by the same amount of backfil-

tration in the second one. Owing to the special blood

sampling port we provided, we were able to measure how

much the blood was concentrated between the two dia-

lyzers, thereby quantifying IF. After circulating for an

hour, we drew blood and determined the amount of IF (QIF)

using the following equation:

QIF ¼ 200ð1� HctPre=HctMidÞ;

Table 1 Characteristics of five

hollow-fiber dialyzers and

membrane permeabilities before

circulation, represented by UFR

Length

(mm)

Diameter

(lm)

Density

(%)

Number Thickness

(lm)

UFR

(ml/h mmHg)

FB-150G 227 200 50 10,300 15 35

FB-150U 227 200 50 10,300 15 116

FB-150F 227 200 50 10,300 15 160

FB-150UH 227 200 50 10,300 15 288

FB-150FH 227 185 47 11,000 15 322

where HctPre: Hct before the first dialyzer, HctMid: Hct

between the two dialyzers.

Change in UFR after circulation

After circulating for 4 h, we measured the post-circulation

UFRs in the first and second dialyzers in the same manner

as the pre-circulation UFRs were measured in a circuit

filled with RO water.

Results

The membrane permeabilities in five dialyzers, represented

by the pre-circulation UFRs, are shown in Table 1. In the

tandem-dialyzer circuit, the hematocrit value substantially

increased between the two dialyzers. In FB-150F, for

instance, it increased from 33 to 42%, and thus the amount

of IF was calculated to be 43 ml/min. As shown in Fig. 1,

IF increased with increasing membrane permeability.

In three types of dialyzer (FB-150U, FB-150F, FB-

150UH), circulation continued for 4 h, although we found

varying degrees of residual blood inside the dialyzers upon

completion. We measured the post-circulation UFRs in the

first and second dialyzers. In FB-150U, the UFRs were

52.9 and 68.6 ml/h mmHg; in FB-150F, they were 69.2

and 105.9 ml/h mmHg; in FB-150UH, they were 116.1 and

153.2 ml/h mmHg. In all three dialyzers, the post-circula-

tion UFRs were lower than the pre-circulation UFRs. In

addition, the post-circulation UFR was lower in the first

dialyzer than in the second one for all three types of

dialyzer. In the remaining two types (FB-150G and

FB-150FH), circulation did not last longer than 3 h because

of a remarkable increase in the venous pressure due to

blood clotting inside the dialyzer.

Discussion

It has been suggested that convective treatment such as

hemodiafiltration is associated with improved conditions of

114 J Artif Organs (2010) 13:113–116

123

patients with end-stage renal disease [7–12]. However,

such treatment has not been widely practiced because it

requires additional processes and equipment, especially in

the preparation of substitution fluids. Over the past decade

hemodialysis with enhanced IF has been investigated [1–6]

because it can be an efficient and readily applicable alter-

native in the treatment of dialysis patients.

Theoretical analysis has shown that the structure and

permeability of hollow fibers play crucial roles in pro-

ducing IF [3]. By increasing the total length, or decreasing

the inner diameter, the blood-side pressure drops rapidly

from the inlet to the outlet, thus increasing the local

transmembrane pressure, leading to an enhancement of IF.

In this study, we intended to focus on the relation between

IF and membrane permeability. Although membrane per-

meability, in theory, is highly relevant to IF, data on it are

scarce, mainly because it is difficult to directly measure IF.

Previously, some methods have been introduced to deter-

mine IF [5, 6], although they are somewhat cumbersome or

inaccurate. In order to directly measure IF and investigate

the association between IF and membrane permeability, we

set up a special circuit containing two dialyzers in series

with a blood sampling port between them. Using this

experimental circuit, we were able to measure IF by

determining the extent to which blood was concentrated

between the two dialyzers, which was a phenomenon

caused by IF.

We found that the amount of IF increased with

increasing measured UFR. It was noted that one dialyzer

(FB-150FH) seemed to have disproportionately larger

amounts of IF (Fig. 1), which could be explained by the

slightly narrower fiber (185 lm) of this specific dialyzer,

which further enhanced IF compared with its 200-micron

counterparts. In three types of dialyzer (FB-150U, FB-

150F, FB-15UH), circulation continued for more than 4 h

in the tandem circuit. In all of the dialyzers, the post-

circulation UFR was lower than the pre-circulation UFR,

which indicates that protein adhesion to the membrane

(i.e., membrane fouling) during the circulation has a neg-

ative impact on UFR. In addition, post-circulation UFR

was decreased in the first dialyzer, which suggests that IF

in the first dialyzer causes more membrane fouling than in

the second dialyzer, where backfiltration mainly occurred.

This result is in accordance with the results of Yamamoto

et al. [2], who found that membrane fouling caused by IF in

a dialyzer tends to occur near the dialysate outlet port.

There are obvious limitations to the present study. Dif-

ferent bovine blood samples were used in each experiment,

with only the hematocrit values adjusted; a blood sampling

port with a short tubing between the two dialyzers could

modify the hydrodynamics in the circuit, which potentially

affects QIF. Moreover, although QIF was simulated to take

the largest value near the center of a dialyzer [3], it might

not necessarily do so between the two dialyzers of this

experimental set-up.

Taking all of these shortcomings into account, the

results are admittedly qualitative rather than quantitative.

The experimental setting is different from a clinical one.

However, given the difficulty involved in quantifying IF,

we believe that our direct measurements of IF have

advanced knowledge in this research field.

Conclusion

We have demonstrated a strong correlation between IF and

membrane permeability; the measured IF increased linearly

with increasing membrane permeability in five different

dialyzers. After the circulation, UFR was decreased,

especially in the first dialyzer, suggesting that the IF caused

significant membrane fouling, which led to decreased water

permeability of the membrane.

References

1. Mineshima M, Ishimori I, Sakiyama R. Validity of internal

filtration-enhanced hemodialysis as a new hemodiafiltration

therapy. Blood Purif. 2009;27:33–7.

2. Yamamoto K, Hiwatari M, Kohori F, Sakai K, Fukuda M,

Hiyoshi T. Membrane fouling and dialysate flow pattern in

an internal-filtration enhancing dialyzer. J Artif Organs.

2005;8:198–205.

3. Mineshima M, Ishimori I, Ishida K, Hoshino T, Kaneko I, Sato Y,

Agishi T, Tamamura N, Sakurai H, Masuda T, Hattori H. Effects

of internal filtration on the solute removal efficiency of a dialyzer.

ASAIO J. 2000;46:456–60.

4. Dellanna F, Wuepper A, Baldamus CA. Internal filtration—

advantage in haemodialysis? Nephrol Dial Transpl. 1996;11

(Suppl 2):83–6.

5. Ronco C, Brendolan A, Feriani M, Milan M, Conz P, Lupi A,

Berto P, Bettini M, La Greca G. A new scintigraphic method to

80

100

120

20

40

60

QIF

(m

l/min

)

00 50 100 150 200 250 300 350

UFR (ml/hr mmHg)

Fig. 1 Relationship between membrane permeability and internal

filtration in five dialyzers (FB-150G, FB-150U, FB-150F, FB-150UH,

FB-150FH, from left to right)

J Artif Organs (2010) 13:113–116 115

123

characterize ultrafiltration in hollow fiber dialyzers. Kidney Int.

1992;41:1383–93.

6. Sato Y, Mineshima M, Ishimori I, Kaneko I, Akiba T, Teraoka S.

Effect of hollow fiber length on solute removal and quantification

of internal filtration rate by Doppler ultrasound. Int J Artif

Organs. 2003;26:129–34.

7. Leypoldt JK, Cheung AK, Carroll CE, Stannard DC, Pereira BJ,

Agodoa LY, Port FK. Effect of dialysis membranes and middle

molecule removal on chronic hemodialysis patient survival. Am J

Kidney Dis. 1999;33:349–55.

8. Kuchle C, Fricke H, Held E, Schiffl H. High-flux hemodialysis

postpones clinical manifestation of dialysis-related amyloidosis.

Am J Nephrol. 1996;16:484–8.

9. Koda Y, Nishi S, Miyazaki S, Haginoshita S, Sakurabayashi T,

Suzuki M, Sakai S, Yuasa Y, Hirasawa Y, Nishi T. Switch from

conventional to high-flux membrane reduces the risk of carpal

tunnel syndrome and mortality of hemodialysis patients. Kidney

Int. 1997;52:1096–101.

10. Locatelli F, Marcelli D, Conte F, Limido A, Malberti F, Spotti D.

Comparison of mortality in ESRD patients on convective and

diffusive extracorporeal treatments. Kidney Int. 1999;55:286–93.

11. Locatelli F, Manzoni C, Di Filippo S. The importance of con-

vective transport. Kidney Int. 2002;61:s115–20.

12. Canaud B, Bragg-Gresham JL, Marshall MR, Desmeules S,

Gillespie BW, Depner T, Klassen P, Port FK. Mortality risk for

patients receiving hemodiafiltration versus hemodialysis: Euro-

pean results from the DOPPS. Kidney Int. 2006;69:2087–93.

116 J Artif Organs (2010) 13:113–116

123