Artifi( irrl O,ynm
lh(h):553-55X. Blackwell Scientific Publication5. Inc.. Boston C 1992 lnlernational Society for Artificial Organs
Adsorption of Nafamostat Mesilate by Hemodialysis Membranes Oshi Inagaki, Yoshihiko Nishian, Ryotaro Iwaki, Kiyohiko Nakagawa, Yoshihiro Takamitsu, and Yosikazu Fujita DtJprrrtmrnt of Kidney cind Dialysis, Hyogo College of Medicine, Nishinomiyci, Jripan
Abstract: ‘The adsorption of the anticoagulant nafamostat mesilate (FUT-175) by five different hemodialysis membrane\ was studied in vivo and in vitro. In vivo, FUTI75 was adsorbed strongly by a polyacrylonitrile (AN691 membrane and slightly by another polyacrylonitrile (JPAN) membrane but not by Cuprophan (CU), hemophan (HE), or polymethylmethacrylate (PMMA) membranes during hemodialysis performed in 4 patients in whom FUT-175 was used as an anticoagulant. Only during hemodialysis using the AN69 membrane did FUT-175 not induce ;I significant prolongation of celite-activated coagulation time. In vitro studies showed that FUT-I75 was
adsorbed by the AN69, J-PAN, and PMMA membranes but not by the CU and HE membranes. Methylene blue, a dye that possesses a cationic portion in its chemical structure, stained AN69, J-PAN, and PMMA membranes. Since FUT-175 also possesses a cationic portion, we conclude that FUT-175 is adsorbed by negatively charged membranes via an ionic bond and is unsuited for use as an anticoagulant in hemodialysis using an AN69 membrane because of that membrane’s marked capacity t o adsorb dyeFUT- 175. Key Words: Anionic dye-Cationic Hemodialysis membrane-Membrane adsorptionMembrane charge-Nafamostat mesilate.
Nafamostat mesilate (FUT-175) is a protease inhibitor that strongly inhibits the activities of various coagulation enzymes ( I ) . Because of its short halflife, this agent is regarded as a useful regional anticoagulant in the hemodialysis of patients with a bleeding tendency (2). The use of FUT-175 as a regional anticoagulant for plasma exchange ( 3 ) and plasma perfusion has also been reported (4). However, in our experience FUT- I75 lacked sufficient anticoagulation effects for the performance of hemodialysis using polyacrylontrile (AN69) membranes. The AN69 membrane has a marked ability to adsorb various substances present in the blood such as vancomycin ( S ) , complement proteins (6), p-thromboglobulin (7), and pz-rnicroglobulin (8). Studies have been conducted on the charge on the dialysis membrane (9,lO). It has been shown that the AN69 membrane has a negative charge on its surface ( I I , 12). Our study was designed to investigate the influence
of the membrane charge on the adsorption of FUT175 using the AN69 and other dialysis membranes. MATERIALS AND METHODS Materials Five different hollow-fiber dialyzers were tested in this study: Filtral-I2 (1.15 m’; Hospal Medical Corp., Tokyo) using an AN69 membrane, PAN13DX (1.3 m’; Asahi Medical Co., Ltd. ,Tokyo) using a different polyacrylonitrile polymer (J-PAN), B21.0 (1.0 m’; Toray Medical Ltd., Tokyo) using a polymethylmethacrylate (PMMA) membrane; RAIOH (1.0 m2; Kawasumi Laboratories Inc., Tokyo) using Cuprophan (CU) membrane, and MA- IOH (1 .O m’; Kawasumi Laboratories Inc., Tokyo) using a hemophan (HE) membrane. A commercial polyvinyl-chloride tubing set (LAPH- 14; Kawasumi Laboratories Inc., Tokyo) was used as the connecting line. FUT-175 (molecular weight 540) was obtained from the Torii Pharmaceutical Co. Ltd., Tokyo (Fig. I).
Received March 1992; revised June 1992.
In vivo studies Four patients who were undergoing maintenance hemodialysis participated in the in vivo study. Each
Addre55 correspondence and reprint requests to Dr. 0. Inagaki
at the Department of Kidney and Dialysis, Hyogo College of Medicine, Mukogawa-cho 1 - 1 , Nishinomiya 663, Japan
553
554
0.tNAGAKI ET AL.
FIG. 1. Chemical structure of nafamostat mesilate (FUG-175).
was dialyzed five times with five different dialyzers in random order. Prior to each hemodialysis, the dialyzer and tubing set were rinsed with 1,000 ml of saline without heparin or FUT-175. Bicarbonate was used as the dialysate buffer in all cases. The flow rate of blood was adjusted to 175 ml/min and that of the dialysate to 500 mllmin. Immediately after beginning the hemodialysis, FUT-175 was infused continuously at a rate of 40 mg/h. Samples for determination of FUT-175 and celite-activated coagulation time (ACT) were obtained from the circuit at the inlet and the outlet of the dialyzer and from the dialysate at the outlet of the dialyzer. During hemodialysis using the AN69 membrane, a bolus injection of heparin (2,000 IU) was administered after sampling at the 60 min test period to prevent coagulation in the circuit.
In vitro membrane binding of FUT-175 The five different dialyzers and the tubing sets used in this study were initially rinsed with 1,000 ml of saline. The dialysate space of each device was filled with saline. Four dialyzers were used to test each type of dialysis membrane. The FUT-175 solution (40 pg/ml) diluted with saline was maintained at 37°C and routed through the circuit at a rate of 175 mlimin aided by a pump. Samples for the measurement of FUT-175 were taken from the circuit at the inlet and the outlet of the dialyzer. In vitro membrane staining Methylene blue (MB) was used as the cationic dye, and orange I1 (OR) was used as the anionic dye (Fig. 2 ) . MB was first dissolved in ethanol and then diluted to a concentration of 2.5% with distilled water. OR was dissolved to a concentration of 10% with distilled water. Each dialyzer device and tubing set was initially rinsed with 1,000 ml of distilled water. Then, 200 ml of each staining solution was maintained at 37°C and circulated in the circuit. After 20 min, the circuit was rinsed with 800 ml of 1% acetic acid in the case of MB and 800 ml of distilled water in the case of OR to wash out any residual dyes. Each type of dialysis membrane was
pc o% N /H
NHz
2CH$03H
tested in each of four dialyzers. Finally, we evduated the extent of staining of each membrane. Measurements The concentration of FUT-175 in saline was measured by ultraviolet absorbance at 240 nm. The concentration of FUT-175 in the blood and the dialysate fluid was determined by high-performance liquid chromatography (13). ACT was measured with an automatic device (Hemochron; International Technidyne Corp., NJ, U.S.A.). All values were given as the mean -t SD. The significance of the differences was analyzed by Student’s t test. RESULTS
In vivo studies The plasma concentration of FUT-175 was decreased in the outlet of the dialyzer in all membranes as compared with the inlet concentration of FUT175 (Fig. 3). Plasma concentrations of FUT-175 were very low in the outlet of the dialyzer with the J-PAN membrane 5 min after starting treatment and in the AN69 membrane at all test periods. Furthermore, dialysate concentration of FUT-175 at the outlet of
Methylene Blue
Orange II N a 0 3 S 0 N= N
D
W FIG. 2. Chemical structures of methylene blue and orange II.
FUT-1 75 A N D HEMODIAL YSIS MEMBRANES cu
555 J-PAN
AN69
.
1
FIG. 3. Changes in the concentration of FUT-175 in plasma and dialysate in vivo. Value in plasmaat the inlet of the dialyzer, open box; value in plasma at the outlet of the dialyzer, dotted box; value in dialysate at the outlet of the dialyzer, black box. Values are the mean 5 SD of 4 patients. Asterisks indicate statistical significance (p < 0.05) compared with inlei value of FUT-175.
2
5
60
5
180
60
1EO
60
180
5
60
180
5
60 180 bin)
FUT-175 was detected at the outlet of the dialyzer. The maximum adsorbed doses of FUT-175, calculated as the difference between its concentration at the inlet and the outlet, was about 3 mg each for the H E and CU membrane dialyzers, about 100 mg each for the PMMA and J-PAN membrane dialyzers and more than 400 mg for the AN69 membrane dialyzer.
the dialyzer was very low in the J-PAN membrane at 5 min and in the AN69 membrane at all test periods. In the HE, CU, PMMA, and J-PAN membranes, the levels of ACT at the inlet and outlet were prolonged at each test period as compared with the predialysis values (Fig. 4). In the AN69 membrane, the outlet levels of ACT were not prolonged at 5 and 60 min as compared with predialysis values, but they were prolonged at 180 min because of the heparin usage. In vitro membrane binding of FUT-175 In the experiment using the H E and CU membranes, the concentration of FUT-175 at the outlet of the dialyzer increased immediately after starting the experiment and reached to each inlet level (Fig. 5). In the J-PAN and PMMA membranes, the concentration of FUT-175 increased slowly. In the AN69 membrane, no FUT-175 was detected 20 min after beginning the experiment. This study was continued longer than 20 min only with the AN69 membrane. Three hours after beginning the experiment,
In vitro membrane staining Staining with the cationic dye MB was observed with the PMMA, J-PAN, and AN69 membranes but not with the H E and CU membranes. Only the HE membrane stained strongly by OR, an anionic dye (Table 1). DISCUSSION The AN69 membrane possesses good biocompatibility, and it is highly permeable to p,-microglobulin (14,15). Its benefits in preventing dialysis-related amyloidosis have been described (16,17). However, anaphylactoid reactions have been observed during
cu
FIG. 4. Changes in celite-activated coagulation time in vivo. Predialysis value, black box; value at the inlet of the dialyzer, open box; value at the outlet of dialyzer, dotted box. Values are the mean 5 SD of 4 patients. Asterisks indicate statistical significance (p < 0.05) compared with predialysis value.
5
v
J-PAN
PMMA
AN69
r
i
; 2
300
$ E
; 2" 4
"
*OD
100
0 5
60
180
5
60
180
5
60
180
5
6 0 180
5
60
180
Ininl
A r t f Organs, Vol. 16, No. 6 . 1992
0. INAGAKI ET A L .
5.56
L
I
I
I
2
5
8
I
12
I
20
(rnin)
FIG. 5. Changes in FUT-175 concentration at the outlet of the dialyzer in vitro. Values are the mean of 4 experiments.
hemodialysis with this membrane in patients who were receiving angiotensin-converting enzyme inhibitors (18). It is considered that these anaphylactoid reactions may be due to the negative charge on the membrane surface of AN69 (18). According to recent observations (5-8), the AN69 membrane has a marked capacity to adsorb various substances in the blood. The present in vivo study showed that plasma concentrations of FUT- 175 were markedly decreased at the outlet of the AN69 membrane dialyzer as compared with the inlet levels. Concentrations of FUT-175 in the dialysate were also very low at the outlet of the AN69 membrane dialyzers. In fundamental experiments in rats (19,20), FUT-175 was shown to be rapidly carried into tissues after intravenous administration and degraded into inactive substances in the tissues. On the other hand, an in vitro study shows that FUT-175 is stable after 30 min of incubation with plasma at 37°C (21). It is thought that FUT-175 is not degraded in the circuit between the inlet and outlet of the dialyzer. We therefore considered that FUT-175 is eliminated from the circuit mainly via membrane adsorption during hemoTABLE 1. Results of stuining ,four dialysis membranes with methylene blue und orunge II Membrane
HE
cu PMMA J-PAN AN69
Methylene blue
Orange I1
Unstained Unstained Blue Blue Blue
Orange Unstained Slightly stained Unstained Unstained
ArrifOrgnrls, Vol. 16, No. 6 , 1992
dialysis using AN69 membranes. As a result, FLIT175 failed to induce a significant anticoagulant effect during such hemodialysis. We also evaluated the charge of the dialysis membranes by staining them with cationic and anionic dyes. The AN69 membrane was stained with the cationic dye but not the anionic dye. While there are several staining theories, it is considered that cationic and anionic dyes stain the fibers mainly by creating an ionic bond between the dye and fiber. We used acetic acid to wash out residual MB in the test circuit because we could wash it out easier with acetic acid than with distilled water. Residual OR in the test circuit was washed out easily with distilled water. Fthomaneck et al. (12) described the possibility of detecting the membrane change by staining with cationic and anionic dyes. This relatively simple method is suitable for detecting the charges on the dialysis membrane and for evaluating the interaction between the dialysis membranes and charged substances in the blood. As with the cationic dye, FUT-175 also possesses an amine group in its chemical structure (Figs. 1 and 2). Thus, it is considered that FUT-175 is also cationic. The AN69 membrane has a negatively charged portion that originates in the sulfone group of its polymer structure (Fig. 6). Therefore, it is possible that FUT-175 is adsorbed on the AN69 membrane via an ionic bond. The B2 type of PMMA membrane also has a sulfone group that is not present in the polymer structure of the BK type of PMMA membrane (Fig. 6). The J-PAN membrane has a carboxyl group that also possesses an anionic quality similar to a sulfone group (Fig. 6). In the present study, the B2 type of PMMA and J-PAN membranes were stained by the cationic dye and adsorbed FUT-175 as was also observed with the AN69 membrane in vitro. Therefore, it is possible that FUT-175 is adsorbed on the B2 type of PMMA and J-PAN membranes via an ionic bond. However, the B2 type of PMMA and J-PAN membranes did not strongly adsorb FUT-175 in vivo. Our present in vitro results showed that the maximum adsorption doses of FUT-175 in the PMMA and J-PAN membrane dialyzers wcxe smaller than that found in the AN69 membrane dialyzer. Furthermore, it is known that the protein layers are built up on the dialysis membrane in clinical use (22), and the properties of dialysis membranes are modified by these protein layers (23). Therefore, the difference between in vivo and in vitro results for FUT-175 adsorption with the PMMA and J-PAN membranes may have been caused by the accumulation of a protein layer on these membrane surfaces.
55 7
FUT-175 A N D HEMODIALYSIS MEMBRANES J-PAN
PMMA (BK) CH3
I
FIG. 6. Chemical structures of dialysis membranes.
C0-0CH3
L
. ...
I
CO-OCH3
S03Na
cu
HE
-
CHzOH
CHzOH
H
-
. . . . ._ CH2-C I -.
H
CHzOH
L
'CzHdf."i
H
OH
OH
CHzOH n
'C2H5
As reported by Schmitt et al. (lo), the HE mem-
brane possesses a diethylaminoethyl group as a positively charged portion in its chemical structure (Fig. 6). In our study, the H E membrane was stained by OR but not MB and did not adsorb FUT-175. These results suggest that this membrane does not adsorb the positively charged FUT-175 because of its positive charge. Furthermore, the CU membrane did not stain with either MB or OR and did not adsorb FUT175. I t is therefore considered that the CU membrane carries a weaker charge on its surface. In conclusion, our results demonstrate that the FUT- I75 was adsorbed to the negatively charged dialysis membrane via an ionic bond. This suggests that other charged substances in the blood may also be adsorbed by either positively or negatively charged dialysis membranes. However, our results did not directly demonstrate adsorbed FUT- 175 on the dialysis membranes. Further studies are therefore required to define the direct demonstration of substances adsorbed on differing dialysis membranes via an ionic bond.
REFERENCES I . Hitomi Y, Ikari N, Fujii S . Inhibitory effect of a new synthetic protease inhibitor (FUT-175) on the coagulation system. Hurmatosfasis 1985;15:164-8. 2. Akizawa T , Kotaoka T, Sato M, Koshikawa S, Hirasawa Y, Kazama M, Mimura N , Ota K. Comparative clinical trial of
3.
4.
5.
6.
7. 8.
9.
10. 11.
12.
regional anticoagulation for hemodialysis. Trans A m Soc Artiflntern Organs 1988;34:176-8. Hiraishi M, Yarnazaki Z, lchikawa K , Kanai F, Idezuki Y, Onishi T, Takahama T, Inoue N. Plasma collection using nafamostat mesilate and dipyridamole as an anticoagulant. Int J Artif Organs 1988;11:212-6. Yamazaki Z, Hiraishi M, Kanai F, Takahama T, ldezuki Y , lnoue N . Pharrnacodynamics of FUT-175 anticoagulant in adsorbent plasma perfusion. Trans A m Soc Artif Intern Organs 1989;35:567-9. Torras J , Cao C , Rivas MC, Can0 M, Fernandez E , Montoliu J . Pharmacokinetics of vancomycin in patients undergoing hemodialysis with polyacrylonitrile. Clin Nephrol 1991; 36135-41. Cheung AK. Adsorption of unactivated complement proteins by hemodialysis membranes. Am J Kidney Dis 1989; 14~472-7. Adler AJ, Berlyne GM. P-Thrornboglobulin and platelet factor-4 levels during hemodialysis with polyacrylonitrile. ASAZO J 1981;4:100-2. Zingraff J , Beyne P, Urena P, Uzan M, Man N K , DescampsLatscha B, Drueck T. Influence of hemodialysis membranes on ,&-microglobulin kinetics: in vivo and in vitro studies. Nephrol Dial Transplant 1988;3:284-90. Ono, T, Iwamoto N , Kataoka H, Taniguti Y , Kimura M, Kunitomo T. Anionic polymethyl methacrylate membrane for rapid dialysis. Trans A m Soc Artif Intern Organs 1982;28:38-42. Schmitt S , Holtz M, Klinkmann H, Esther G, Courtney JM. Heparin binding and release properties of DEAE cellulose membranes. Biomaterials 1983;4;309- 13. Matsuda T. Biological responses at non-physiological interfaces and molecular design of biocornpatible surfaces. Nephrol Dial Transplanf 1989;4(SuppI):60-6. Fthomaneck U , Vienken J , Waldschlager U, Diamantoglou M, Schutt W, Falkenhagen D, Klinkmann H. Detection of charges and their distribution on dialysis membranes with Artif Organs, Vol. 16. N o . 6 , 1992
558
0. INAGAKI ET AL.
cationic and anionic dyes using confocal laser scanning microscopy. Int J Artif Organs 1991;14:686-90. 13. Marunaka T,Maniwa M, Matsushima E, Yoshida K, Azuma R, Kurotori M, Komatsu S. High-performance liquid chromatographic determination of 6-amidino-2-naphtyl[4-(4,5-dihydro-lH-imidazol-2-yl)amino] benzoate dimethanesulphonate and its metabolites in biological fluids. J Chromarogr 1988;433:177-86. 14. Shono T,Inagaki 0, Nagasaka H, Fujita Y. Evaluation of protein-permeable membranes: biocompatibility and clinical performance. Jpn J Artif Organs 1988;17:124-7 (in Japanese). 15. Seyfert UT, Helmling E, Hauk W, Skroch D, Albert W. Comparison of blood biocompatibility during hemodialysis with cuprophane and polyacrylonitrile membranes. Nephrol Dial Transplant 1991 ;6:428-34. 16. Hauglustaine D,Waer M, Michielsen P, Goebels J , Vandeputte M. Hemodialysis membranes, serum &-microglobulin, and dialysis amyloidosis. Lancet 1986;l:121 1-2. 17. Van Ypersele de Strihou C, Jadoul M, Malghem J, Maldaugue B, Jamart J . Effect of dialysis membrane and patient’s age on signs of dialysis-related amyloidosis. Kidney In? 1991;39:1012-9.
Art$Organs. Vol. 16, N o . 6, 1992
18. Tielemans C, Madhoun P, Lenaers M, Schandene L, Goldman M, Vanherweghem JL. Anaphylactoid reactions during hemodialysis and AN69 membranes in patients receiving ACE inhibitors. Kidney Znr 1990;38:982-4. 19. Nanpo T,Ohtsuki T, Jin Y, Shibuya M, Sasaki H, Kurumi M. Pharmacokinetic studies of FUT-175(1).Blood level profiles, tissue distribution, metabolism and excretion in rats after intravenous administration. Kiso to Rinsho 1984;18:3971-92 (in Japanese). 20. Esumi Y, Takaichi M, Washino T, Jin Y, Kakurai Y, Shibuya M, Sasaki H, Kurumi M. Pharmacokinetic studies of FUT175 (6). Blood level profiles, distribution, metabolism and excretion in rats after constant-rate infusion. Kiso to Rin:rho 1984;18:4050-66(in Japanese). 21. Ino Y, Suzuki K, Sat0 T, Iwaki M. Comparative studies of nafamostat mesilate and various serine protease inhibitors in vitro. Foliu Pharmacol Japon 1986;88:449-55(in Japanese). 22. Kulik EA, Kalinin ID, Sevastianov VI. The heterogeneity of proteinisurface interactions and structural alterations of adsorbed albumin and immunoglobulin G . Artif Organs 199I ;I5:386-9I . 23. Sharma CP, Chandy T. The effect of antihypertensive drrigs on protein adsorption, platelet adhesion, and blood coagulation toward an artificial surface. Artiforgans 1989;13:219-28.
lh(h):553-55X. Blackwell Scientific Publication5. Inc.. Boston C 1992 lnlernational Society for Artificial Organs
Adsorption of Nafamostat Mesilate by Hemodialysis Membranes Oshi Inagaki, Yoshihiko Nishian, Ryotaro Iwaki, Kiyohiko Nakagawa, Yoshihiro Takamitsu, and Yosikazu Fujita DtJprrrtmrnt of Kidney cind Dialysis, Hyogo College of Medicine, Nishinomiyci, Jripan
Abstract: ‘The adsorption of the anticoagulant nafamostat mesilate (FUT-175) by five different hemodialysis membrane\ was studied in vivo and in vitro. In vivo, FUTI75 was adsorbed strongly by a polyacrylonitrile (AN691 membrane and slightly by another polyacrylonitrile (JPAN) membrane but not by Cuprophan (CU), hemophan (HE), or polymethylmethacrylate (PMMA) membranes during hemodialysis performed in 4 patients in whom FUT-175 was used as an anticoagulant. Only during hemodialysis using the AN69 membrane did FUT-175 not induce ;I significant prolongation of celite-activated coagulation time. In vitro studies showed that FUT-I75 was
adsorbed by the AN69, J-PAN, and PMMA membranes but not by the CU and HE membranes. Methylene blue, a dye that possesses a cationic portion in its chemical structure, stained AN69, J-PAN, and PMMA membranes. Since FUT-175 also possesses a cationic portion, we conclude that FUT-175 is adsorbed by negatively charged membranes via an ionic bond and is unsuited for use as an anticoagulant in hemodialysis using an AN69 membrane because of that membrane’s marked capacity t o adsorb dyeFUT- 175. Key Words: Anionic dye-Cationic Hemodialysis membrane-Membrane adsorptionMembrane charge-Nafamostat mesilate.
Nafamostat mesilate (FUT-175) is a protease inhibitor that strongly inhibits the activities of various coagulation enzymes ( I ) . Because of its short halflife, this agent is regarded as a useful regional anticoagulant in the hemodialysis of patients with a bleeding tendency (2). The use of FUT-175 as a regional anticoagulant for plasma exchange ( 3 ) and plasma perfusion has also been reported (4). However, in our experience FUT- I75 lacked sufficient anticoagulation effects for the performance of hemodialysis using polyacrylontrile (AN69) membranes. The AN69 membrane has a marked ability to adsorb various substances present in the blood such as vancomycin ( S ) , complement proteins (6), p-thromboglobulin (7), and pz-rnicroglobulin (8). Studies have been conducted on the charge on the dialysis membrane (9,lO). It has been shown that the AN69 membrane has a negative charge on its surface ( I I , 12). Our study was designed to investigate the influence
of the membrane charge on the adsorption of FUT175 using the AN69 and other dialysis membranes. MATERIALS AND METHODS Materials Five different hollow-fiber dialyzers were tested in this study: Filtral-I2 (1.15 m’; Hospal Medical Corp., Tokyo) using an AN69 membrane, PAN13DX (1.3 m’; Asahi Medical Co., Ltd. ,Tokyo) using a different polyacrylonitrile polymer (J-PAN), B21.0 (1.0 m’; Toray Medical Ltd., Tokyo) using a polymethylmethacrylate (PMMA) membrane; RAIOH (1.0 m2; Kawasumi Laboratories Inc., Tokyo) using Cuprophan (CU) membrane, and MA- IOH (1 .O m’; Kawasumi Laboratories Inc., Tokyo) using a hemophan (HE) membrane. A commercial polyvinyl-chloride tubing set (LAPH- 14; Kawasumi Laboratories Inc., Tokyo) was used as the connecting line. FUT-175 (molecular weight 540) was obtained from the Torii Pharmaceutical Co. Ltd., Tokyo (Fig. I).
Received March 1992; revised June 1992.
In vivo studies Four patients who were undergoing maintenance hemodialysis participated in the in vivo study. Each
Addre55 correspondence and reprint requests to Dr. 0. Inagaki
at the Department of Kidney and Dialysis, Hyogo College of Medicine, Mukogawa-cho 1 - 1 , Nishinomiya 663, Japan
553
554
0.tNAGAKI ET AL.
FIG. 1. Chemical structure of nafamostat mesilate (FUG-175).
was dialyzed five times with five different dialyzers in random order. Prior to each hemodialysis, the dialyzer and tubing set were rinsed with 1,000 ml of saline without heparin or FUT-175. Bicarbonate was used as the dialysate buffer in all cases. The flow rate of blood was adjusted to 175 ml/min and that of the dialysate to 500 mllmin. Immediately after beginning the hemodialysis, FUT-175 was infused continuously at a rate of 40 mg/h. Samples for determination of FUT-175 and celite-activated coagulation time (ACT) were obtained from the circuit at the inlet and the outlet of the dialyzer and from the dialysate at the outlet of the dialyzer. During hemodialysis using the AN69 membrane, a bolus injection of heparin (2,000 IU) was administered after sampling at the 60 min test period to prevent coagulation in the circuit.
In vitro membrane binding of FUT-175 The five different dialyzers and the tubing sets used in this study were initially rinsed with 1,000 ml of saline. The dialysate space of each device was filled with saline. Four dialyzers were used to test each type of dialysis membrane. The FUT-175 solution (40 pg/ml) diluted with saline was maintained at 37°C and routed through the circuit at a rate of 175 mlimin aided by a pump. Samples for the measurement of FUT-175 were taken from the circuit at the inlet and the outlet of the dialyzer. In vitro membrane staining Methylene blue (MB) was used as the cationic dye, and orange I1 (OR) was used as the anionic dye (Fig. 2 ) . MB was first dissolved in ethanol and then diluted to a concentration of 2.5% with distilled water. OR was dissolved to a concentration of 10% with distilled water. Each dialyzer device and tubing set was initially rinsed with 1,000 ml of distilled water. Then, 200 ml of each staining solution was maintained at 37°C and circulated in the circuit. After 20 min, the circuit was rinsed with 800 ml of 1% acetic acid in the case of MB and 800 ml of distilled water in the case of OR to wash out any residual dyes. Each type of dialysis membrane was
pc o% N /H
NHz
2CH$03H
tested in each of four dialyzers. Finally, we evduated the extent of staining of each membrane. Measurements The concentration of FUT-175 in saline was measured by ultraviolet absorbance at 240 nm. The concentration of FUT-175 in the blood and the dialysate fluid was determined by high-performance liquid chromatography (13). ACT was measured with an automatic device (Hemochron; International Technidyne Corp., NJ, U.S.A.). All values were given as the mean -t SD. The significance of the differences was analyzed by Student’s t test. RESULTS
In vivo studies The plasma concentration of FUT-175 was decreased in the outlet of the dialyzer in all membranes as compared with the inlet concentration of FUT175 (Fig. 3). Plasma concentrations of FUT-175 were very low in the outlet of the dialyzer with the J-PAN membrane 5 min after starting treatment and in the AN69 membrane at all test periods. Furthermore, dialysate concentration of FUT-175 at the outlet of
Methylene Blue
Orange II N a 0 3 S 0 N= N
D
W FIG. 2. Chemical structures of methylene blue and orange II.
FUT-1 75 A N D HEMODIAL YSIS MEMBRANES cu
555 J-PAN
AN69
.
1
FIG. 3. Changes in the concentration of FUT-175 in plasma and dialysate in vivo. Value in plasmaat the inlet of the dialyzer, open box; value in plasma at the outlet of the dialyzer, dotted box; value in dialysate at the outlet of the dialyzer, black box. Values are the mean 5 SD of 4 patients. Asterisks indicate statistical significance (p < 0.05) compared with inlei value of FUT-175.
2
5
60
5
180
60
1EO
60
180
5
60
180
5
60 180 bin)
FUT-175 was detected at the outlet of the dialyzer. The maximum adsorbed doses of FUT-175, calculated as the difference between its concentration at the inlet and the outlet, was about 3 mg each for the H E and CU membrane dialyzers, about 100 mg each for the PMMA and J-PAN membrane dialyzers and more than 400 mg for the AN69 membrane dialyzer.
the dialyzer was very low in the J-PAN membrane at 5 min and in the AN69 membrane at all test periods. In the HE, CU, PMMA, and J-PAN membranes, the levels of ACT at the inlet and outlet were prolonged at each test period as compared with the predialysis values (Fig. 4). In the AN69 membrane, the outlet levels of ACT were not prolonged at 5 and 60 min as compared with predialysis values, but they were prolonged at 180 min because of the heparin usage. In vitro membrane binding of FUT-175 In the experiment using the H E and CU membranes, the concentration of FUT-175 at the outlet of the dialyzer increased immediately after starting the experiment and reached to each inlet level (Fig. 5). In the J-PAN and PMMA membranes, the concentration of FUT-175 increased slowly. In the AN69 membrane, no FUT-175 was detected 20 min after beginning the experiment. This study was continued longer than 20 min only with the AN69 membrane. Three hours after beginning the experiment,
In vitro membrane staining Staining with the cationic dye MB was observed with the PMMA, J-PAN, and AN69 membranes but not with the H E and CU membranes. Only the HE membrane stained strongly by OR, an anionic dye (Table 1). DISCUSSION The AN69 membrane possesses good biocompatibility, and it is highly permeable to p,-microglobulin (14,15). Its benefits in preventing dialysis-related amyloidosis have been described (16,17). However, anaphylactoid reactions have been observed during
cu
FIG. 4. Changes in celite-activated coagulation time in vivo. Predialysis value, black box; value at the inlet of the dialyzer, open box; value at the outlet of dialyzer, dotted box. Values are the mean 5 SD of 4 patients. Asterisks indicate statistical significance (p < 0.05) compared with predialysis value.
5
v
J-PAN
PMMA
AN69
r
i
; 2
300
$ E
; 2" 4
"
*OD
100
0 5
60
180
5
60
180
5
60
180
5
6 0 180
5
60
180
Ininl
A r t f Organs, Vol. 16, No. 6 . 1992
0. INAGAKI ET A L .
5.56
L
I
I
I
2
5
8
I
12
I
20
(rnin)
FIG. 5. Changes in FUT-175 concentration at the outlet of the dialyzer in vitro. Values are the mean of 4 experiments.
hemodialysis with this membrane in patients who were receiving angiotensin-converting enzyme inhibitors (18). It is considered that these anaphylactoid reactions may be due to the negative charge on the membrane surface of AN69 (18). According to recent observations (5-8), the AN69 membrane has a marked capacity to adsorb various substances in the blood. The present in vivo study showed that plasma concentrations of FUT- 175 were markedly decreased at the outlet of the AN69 membrane dialyzer as compared with the inlet levels. Concentrations of FUT-175 in the dialysate were also very low at the outlet of the AN69 membrane dialyzers. In fundamental experiments in rats (19,20), FUT-175 was shown to be rapidly carried into tissues after intravenous administration and degraded into inactive substances in the tissues. On the other hand, an in vitro study shows that FUT-175 is stable after 30 min of incubation with plasma at 37°C (21). It is thought that FUT-175 is not degraded in the circuit between the inlet and outlet of the dialyzer. We therefore considered that FUT-175 is eliminated from the circuit mainly via membrane adsorption during hemoTABLE 1. Results of stuining ,four dialysis membranes with methylene blue und orunge II Membrane
HE
cu PMMA J-PAN AN69
Methylene blue
Orange I1
Unstained Unstained Blue Blue Blue
Orange Unstained Slightly stained Unstained Unstained
ArrifOrgnrls, Vol. 16, No. 6 , 1992
dialysis using AN69 membranes. As a result, FLIT175 failed to induce a significant anticoagulant effect during such hemodialysis. We also evaluated the charge of the dialysis membranes by staining them with cationic and anionic dyes. The AN69 membrane was stained with the cationic dye but not the anionic dye. While there are several staining theories, it is considered that cationic and anionic dyes stain the fibers mainly by creating an ionic bond between the dye and fiber. We used acetic acid to wash out residual MB in the test circuit because we could wash it out easier with acetic acid than with distilled water. Residual OR in the test circuit was washed out easily with distilled water. Fthomaneck et al. (12) described the possibility of detecting the membrane change by staining with cationic and anionic dyes. This relatively simple method is suitable for detecting the charges on the dialysis membrane and for evaluating the interaction between the dialysis membranes and charged substances in the blood. As with the cationic dye, FUT-175 also possesses an amine group in its chemical structure (Figs. 1 and 2). Thus, it is considered that FUT-175 is also cationic. The AN69 membrane has a negatively charged portion that originates in the sulfone group of its polymer structure (Fig. 6). Therefore, it is possible that FUT-175 is adsorbed on the AN69 membrane via an ionic bond. The B2 type of PMMA membrane also has a sulfone group that is not present in the polymer structure of the BK type of PMMA membrane (Fig. 6). The J-PAN membrane has a carboxyl group that also possesses an anionic quality similar to a sulfone group (Fig. 6). In the present study, the B2 type of PMMA and J-PAN membranes were stained by the cationic dye and adsorbed FUT-175 as was also observed with the AN69 membrane in vitro. Therefore, it is possible that FUT-175 is adsorbed on the B2 type of PMMA and J-PAN membranes via an ionic bond. However, the B2 type of PMMA and J-PAN membranes did not strongly adsorb FUT-175 in vivo. Our present in vitro results showed that the maximum adsorption doses of FUT-175 in the PMMA and J-PAN membrane dialyzers wcxe smaller than that found in the AN69 membrane dialyzer. Furthermore, it is known that the protein layers are built up on the dialysis membrane in clinical use (22), and the properties of dialysis membranes are modified by these protein layers (23). Therefore, the difference between in vivo and in vitro results for FUT-175 adsorption with the PMMA and J-PAN membranes may have been caused by the accumulation of a protein layer on these membrane surfaces.
55 7
FUT-175 A N D HEMODIALYSIS MEMBRANES J-PAN
PMMA (BK) CH3
I
FIG. 6. Chemical structures of dialysis membranes.
C0-0CH3
L
. ...
I
CO-OCH3
S03Na
cu
HE
-
CHzOH
CHzOH
H
-
. . . . ._ CH2-C I -.
H
CHzOH
L
'CzHdf."i
H
OH
OH
CHzOH n
'C2H5
As reported by Schmitt et al. (lo), the HE mem-
brane possesses a diethylaminoethyl group as a positively charged portion in its chemical structure (Fig. 6). In our study, the H E membrane was stained by OR but not MB and did not adsorb FUT-175. These results suggest that this membrane does not adsorb the positively charged FUT-175 because of its positive charge. Furthermore, the CU membrane did not stain with either MB or OR and did not adsorb FUT175. I t is therefore considered that the CU membrane carries a weaker charge on its surface. In conclusion, our results demonstrate that the FUT- I75 was adsorbed to the negatively charged dialysis membrane via an ionic bond. This suggests that other charged substances in the blood may also be adsorbed by either positively or negatively charged dialysis membranes. However, our results did not directly demonstrate adsorbed FUT- 175 on the dialysis membranes. Further studies are therefore required to define the direct demonstration of substances adsorbed on differing dialysis membranes via an ionic bond.
REFERENCES I . Hitomi Y, Ikari N, Fujii S . Inhibitory effect of a new synthetic protease inhibitor (FUT-175) on the coagulation system. Hurmatosfasis 1985;15:164-8. 2. Akizawa T , Kotaoka T, Sato M, Koshikawa S, Hirasawa Y, Kazama M, Mimura N , Ota K. Comparative clinical trial of
3.
4.
5.
6.
7. 8.
9.
10. 11.
12.
regional anticoagulation for hemodialysis. Trans A m Soc Artiflntern Organs 1988;34:176-8. Hiraishi M, Yarnazaki Z, lchikawa K , Kanai F, Idezuki Y, Onishi T, Takahama T, Inoue N. Plasma collection using nafamostat mesilate and dipyridamole as an anticoagulant. Int J Artif Organs 1988;11:212-6. Yamazaki Z, Hiraishi M, Kanai F, Takahama T, ldezuki Y , lnoue N . Pharrnacodynamics of FUT-175 anticoagulant in adsorbent plasma perfusion. Trans A m Soc Artif Intern Organs 1989;35:567-9. Torras J , Cao C , Rivas MC, Can0 M, Fernandez E , Montoliu J . Pharmacokinetics of vancomycin in patients undergoing hemodialysis with polyacrylonitrile. Clin Nephrol 1991; 36135-41. Cheung AK. Adsorption of unactivated complement proteins by hemodialysis membranes. Am J Kidney Dis 1989; 14~472-7. Adler AJ, Berlyne GM. P-Thrornboglobulin and platelet factor-4 levels during hemodialysis with polyacrylonitrile. ASAZO J 1981;4:100-2. Zingraff J , Beyne P, Urena P, Uzan M, Man N K , DescampsLatscha B, Drueck T. Influence of hemodialysis membranes on ,&-microglobulin kinetics: in vivo and in vitro studies. Nephrol Dial Transplant 1988;3:284-90. Ono, T, Iwamoto N , Kataoka H, Taniguti Y , Kimura M, Kunitomo T. Anionic polymethyl methacrylate membrane for rapid dialysis. Trans A m Soc Artif Intern Organs 1982;28:38-42. Schmitt S , Holtz M, Klinkmann H, Esther G, Courtney JM. Heparin binding and release properties of DEAE cellulose membranes. Biomaterials 1983;4;309- 13. Matsuda T. Biological responses at non-physiological interfaces and molecular design of biocornpatible surfaces. Nephrol Dial Transplanf 1989;4(SuppI):60-6. Fthomaneck U , Vienken J , Waldschlager U, Diamantoglou M, Schutt W, Falkenhagen D, Klinkmann H. Detection of charges and their distribution on dialysis membranes with Artif Organs, Vol. 16. N o . 6 , 1992
558
0. INAGAKI ET AL.
cationic and anionic dyes using confocal laser scanning microscopy. Int J Artif Organs 1991;14:686-90. 13. Marunaka T,Maniwa M, Matsushima E, Yoshida K, Azuma R, Kurotori M, Komatsu S. High-performance liquid chromatographic determination of 6-amidino-2-naphtyl[4-(4,5-dihydro-lH-imidazol-2-yl)amino] benzoate dimethanesulphonate and its metabolites in biological fluids. J Chromarogr 1988;433:177-86. 14. Shono T,Inagaki 0, Nagasaka H, Fujita Y. Evaluation of protein-permeable membranes: biocompatibility and clinical performance. Jpn J Artif Organs 1988;17:124-7 (in Japanese). 15. Seyfert UT, Helmling E, Hauk W, Skroch D, Albert W. Comparison of blood biocompatibility during hemodialysis with cuprophane and polyacrylonitrile membranes. Nephrol Dial Transplant 1991 ;6:428-34. 16. Hauglustaine D,Waer M, Michielsen P, Goebels J , Vandeputte M. Hemodialysis membranes, serum &-microglobulin, and dialysis amyloidosis. Lancet 1986;l:121 1-2. 17. Van Ypersele de Strihou C, Jadoul M, Malghem J, Maldaugue B, Jamart J . Effect of dialysis membrane and patient’s age on signs of dialysis-related amyloidosis. Kidney In? 1991;39:1012-9.
Art$Organs. Vol. 16, N o . 6, 1992
18. Tielemans C, Madhoun P, Lenaers M, Schandene L, Goldman M, Vanherweghem JL. Anaphylactoid reactions during hemodialysis and AN69 membranes in patients receiving ACE inhibitors. Kidney Znr 1990;38:982-4. 19. Nanpo T,Ohtsuki T, Jin Y, Shibuya M, Sasaki H, Kurumi M. Pharmacokinetic studies of FUT-175(1).Blood level profiles, tissue distribution, metabolism and excretion in rats after intravenous administration. Kiso to Rinsho 1984;18:3971-92 (in Japanese). 20. Esumi Y, Takaichi M, Washino T, Jin Y, Kakurai Y, Shibuya M, Sasaki H, Kurumi M. Pharmacokinetic studies of FUT175 (6). Blood level profiles, distribution, metabolism and excretion in rats after constant-rate infusion. Kiso to Rin:rho 1984;18:4050-66(in Japanese). 21. Ino Y, Suzuki K, Sat0 T, Iwaki M. Comparative studies of nafamostat mesilate and various serine protease inhibitors in vitro. Foliu Pharmacol Japon 1986;88:449-55(in Japanese). 22. Kulik EA, Kalinin ID, Sevastianov VI. The heterogeneity of proteinisurface interactions and structural alterations of adsorbed albumin and immunoglobulin G . Artif Organs 199I ;I5:386-9I . 23. Sharma CP, Chandy T. The effect of antihypertensive drrigs on protein adsorption, platelet adhesion, and blood coagulation toward an artificial surface. Artiforgans 1989;13:219-28.
