Effect of Hemodialysis on Protein C, Protein S, and Antithrombin III Levels

Effect of Hemodialysis on Protein C, Protein S, and Antithrombin III Levels Kar-Neng Lai, MD, FRCP, Jane A. Yin, MSc, Patrick M.P. Yuen, MD, FRCP(C), and Philip K.T. Li, MBBS, MRCP • We studied the effect of hemodialysis on natural coagulation inhibitors including protein C (PC), protein S (PS), and antithrombin III (AT III), as well as the correlations between the antigen level (immunological activity) and functional activity of individual coagulation inhibitor. Plasma AT III, PS, and PC were measured in 20 uremic patients on maintenance hemodialysis immediately before, during, and after dialysis treatment. These values were compared with those obtained from 20 matched healthy controls. Plasma PC and total PS antigen levels were measured by enzyme immunoassay. The plasma AT III antigen level was determined by the Laurell rocket immunoelectrophoresis. Functional activities of PC and AT III were determined by the amidolytic method. Free PS antigen level was quantitated by measuring the free PS-related antigen after the sample was treated with polyethylene glycol to remove the C4b-binding protein. Uremic patients on maintenance hemodialysis had a higher total PS antigen level, but a lower free PS antigen level compared with the controls. Both the antigen level and functional activity of AT III in uremic patients were significantly lower than those of controls. Their predialysis plasma PC antigen level and functional activity were not different from those of normal controls. A significant correlation between the antigen level and functional actvity of PC, PS, and AT III was demonstrated in healthy controls, but not in hemodialysis patients. No significant change in the level of AT III or PS was observed with hemodialysis, but a progressive increase of functional activity of PC was documented with hemodialysis. Furthermore, the coefficient of correlation between the antigen level and functional activity of PC improved significantly with dialysis treatment. Our study did not show a reduction of plasma natural coagulation inhibitors with hemodialysis. On the contrary, the risk of thrombosis may be reduced as PC levels increase with dialysis treatment, possibly due to the removal of an inhibitory substance on PC activity. © 1991 by the National Kidney Foundaton, Inc. INDEX WORDS: Protein S; protein C; antithrombin III; hemodialysis.

P

ATIENTS undergoing regular hemodialysis are subject to a special risk of thrombotic complications. A decrease in plasma antithrombin III (AT III) levels during dialysis has been suggested to be associated with the genesis of this complication,12 although other investigators have failed to confirm this observation. 3.4 Recently, two new vitamin K-dependent plasma proteins involved in blood coagulation were described 56 and given the names protein C (PC) and protein S (PS). In vivo, PC is activated by thrombin and the endothelial cell cofactor thrombomodulin. 7 Activated PC inhibits thrombin formation by preventing the activation of factors V and VIII. 8 Clinical evidence of PC involvement in the regulation of coagulation comes from the obFrom the Departments of Medicine and Pediatrics, Prince of Wales Hospital, The Chinese University of Hong Kong, Shatin, Hong Kong. Supported by the Mr and Mrs L. C. Liu research grant (Hong Kong). Diagnostica Stago supplied the antibodies. Address reprint requests to Kar-Neng Lai, MD, FRCP, Department of Medicine, Prince of Wales Hospital, Shatin, Hong Kong. © 1991 by the National Kidney Foundation, Inc. 0272-6386/91/1701-0007$3.00/0

38

servation that congenital deficiency in PC leads to an overt thrombotic state. 9 Furthermore, acquired defects in several clinical situations predispose these patients to thrombosis. IO PS serves as a cofactor for the anticoagulant effect of activated PC,6 and is essential for binding activated C to the platelet ll and endothelial cell surfaces 12 on which the anticoagulant activity is expressed. Similarly, patients with congenital or acquired PS deficiency are also subject to recurrent venous thrombosis.I314 The information of the effect of hemodialysis on these natural anticoagulants is either controversial (as in AT III), limited (as in PC), or unavailable (as in PC). This report describes our studies on the effect of hemodialysis on these proteins. The correlations between the antigen level and functional activity of individual natural coagulation inhibitor were also examined, as reduced levels of functional PC, despite normal plasma antigen level, have been reported in uremic patients. 15 .16 PATIENTS AND METHODS Plasma AT III, PS, and PC were measured in 20 patients with chronic renal failure on regular hemodialysis (nine men and II women, aged 35 ± 7 years; duration of maintenance dialysis,

American Journal of Kidney Diseases, Vol XVII, No 1 (January), 1991: pp 38-42

COAGULATION INHIBITOR LEVELS IN HEMODIALYSIS 40 ± 3 1 m onths). The causes o frenal failu re included chronic glomerulonephritis ( 16), polycystic kidney disease (I ) , and unknown etiology (3). Their reg ular medications included aluminum hydroxide, calcium supplements, and vitamins. Eight patients received antihypertensive treatment (calcium channel blockers) and none of the female patients was on estrogen therapy. All patients on hemodialysis had arteriovenous fi stulae and were dialyzed 4 hours, three times a week , using either a cellulose or cuprophan hollow-fiber dialyzer. Intravenous heparin was administered in a loading dose of 2,000 U followed by 800 to 1,500 U h ourly to maintain adequate anticoagul ation throughout the dialy sis (activated p artial thromboplastin time 1.5 times the baseline value) . Plasma AT III , PS, PC, albumin , and platelet count were immediately determined at I minute after the initial intravenous bolus o f h eparin prior t o the start of dialysis , 2 hours after starting dialysis , and immediatel y after completing dialy sis treatment. Follow ing a clean v enipuncture , blood sa mples for AT III , PS, and PC assay were collected into anticoagulant mixtures in a 9: I r atio in plastic tubes containing 3.8 % tri sodium citrate. They were placed in crushed ice immediately after coll ection and centrifuged at 3,0 00 rpm for 20 minutes at 4 °C in a r efri gerated centrifuge. Plasma collected w as aliquoted for immediate storage at - 70 °C until assay. Samples were collected over a period of 2 month s and assayed in a single batch. All plasma specimens were thawed at 37 °C for15 minutes before assay. Plasma total PS and PC-related antigens (immunological activities) w ere measured by the sandwich technique of enzyme immunoassay (ELISA) using commerci al assay kits fro m Diagnostica Stago (Asnieres, France) and following the manu facturer's recommendations. The micro ELISA plate was coated with specific F(ab ') fragments to either PC or PS and the enzymatic activity was obtained by using Ortho Phenyline Diamine with hydrogen peroxide as substrate and I mollL HCI to block the reaction. Absorbance of each we ll was determined at 490 nm using a MR600 Microplate reader (Dynatech , A lexandria, VA) . The plasma AT III antigen level (i mmunol ogical activity) was determined b y the Laurell rocket immunoelectrophores isl1 using assay kits f rom Diagnostica Stago. The agarose gel containing the specific antiserum was run at 4 mAl cm for 4 hours under moderate refrigerati on. The gel was soaked overnight in saline, followed b y distilled water, and then dried. The prec ipitation peaks were visualized by 0.4 % Coomassie blue staining. Functional activity of PC and AT III was measured by the amidolytic method using assay kits from Diagnostica Stago. AT III was determined byassaying the remaining thrombin activ ity by its amidolytic activ ity on the synthetic chromogenic sub-

39

strate CBS 34.47 after all avail able AT III activity was neutral ized . PC was activated in the presence of the specific activator, 18 and the activated PC was quantified by its amidasic activity on the synthetic chromogenic substrate CBS 65.25. The reaction was stopped with acetic acid, and the released chromophore, paranitroaniline , was m easured at 405 nm. Free PS antigen was quantitated by measuring the free PS-related antigens after the sample was treated with polyethylene glycol to remove the C4b-binding protein . 19 The results were expressed as percentage o f r feerence plasma obtained from Diagnostica Stago. Healthy hospital staffs and medical students (age- and sex-matched) were used as normal controls. None of the control subjects was o n estrogen preparation. The data (expressed as the mean ± SD) were a nalyzed u isng Wilcoxon signed-rank test, rank-sum test, and linear regression analysis when appropriated. All P values quoted are two-tailed .

RESULTS

The predialysis plasma AT III, PS , and PC levels of the 20 hemodialysis patients are presented in Table I. No correlation between any of these measurements of AT III, PS, or PC with standard coagulation profiles such as prothrombin time or partial thromboplastin time measured prior to dialysis was demonstrated. The changes of the plasma concentration of these natural coagulation inhibitors with hemodialysis treatment are shown in Table 2. A significant increase of the plasma albumin concentration was demonstrated with hemodialysis treatment (P < 0 .05) , but the mean platelet counts measured before dialysis (175,000 ± 67,000/ ilL) were not significantly different from the platelet counts determined after dialysis (181 ,000 ± 89,000/IlL) . The mean predialysis AT III antigen level and functional activity of our 20 hemodialysis patients were significantly lower than that of the normal controls (P < 0.01) . The plasma AT III antigen level and functional activity of these uremic patients did not change with dialysis treatment. The predialysis plasma total PS antigen level (105 % ± 13 %) was higher than that of the normal

Table 1. Plasma Antigenic and Functional Activities of AT III and PC, and Total and Free PS Antigens in Control Group and Uremic Patients (Predialysis Values)

Controls (n = 20) Patients (n = 20) P value (control v patient)

PC

PC

Total

Free

Antigen Level

Functional Activity

Antigen

Antigen

87 97

± ± NS

23 16

73 79

± ± NS

17 23

PS

81 105

± ±

11 13

<0.01

NOTE . Results are expressed as percentage of reference plasma.

PS

87 76

± ±

18 19

<0.05

AT III Antigen Level 88 76

±6 ± 12

<0.01

AT III Functional Activity 114 79

± ±

16 20

<0.01

LAI ET AL

40 Table 2.

Effect of Hemodialysis on the Antigenic and Functional Activities of AT III, PS, and PC

PC Antigen Level Predialysis (A) 97 ± 16 96 ± 14 2 hours after (B) start of dialysis 101 ± 14 Postdialysis (C) Significance levels (Wilcoxon signed-rank test) AvB NS AvC NS BvC <0.05

PC Functional Activity

Total PS Antigen

Free PS Antigen

AT III Antigen Level

AT III Functional Activity

Plasma Albumin

79 ± 23 82 ± 23

105 ± 13 105 ± 11

76 ± 19 78 ± 17

76 ± 12 77 ± 12

79 ± 20 82 ± 18

36.5 ± 2.4 37.9 ± 2.1

87 ± 23

109 ± 11

80 ± 22

77 ± 13

82 ± 19

39 .9 ± 2.3

NS <0 .02 <0.01

NS NS <0.05

NS NS NS

NS NS NS

NS NS NS

NS <0.05 NS

(giL)

NOTE. Results are expressed as percentage of reference plasma.

controls (P < 0.01) , but the free PS antigen level in uremic patients was significantly lower than that of controls (P < 0.05). A progressive increase of total and free PS antigen was observed with hemodialysis treatment, yet a significant difference was only demonstrated between total PS antigen level measured at 2 hours after the start of dialysis and immediately after dialysis therapy. No difference in predialysis plasma PC antigen level and functional activity was demonstrated between the dialysis patients and normal controls. With dialysis therapy, a progressive and significant increase of PC functional activity was observed . AT III, PS , and PC were not detected in the outflow dialysate. The correlation coefficients between the antigen levels and functional activities (or free antigen in PS) of these natural coagulation inhibitors in the hemodialysis patients before and after dialysis therapy and those of control subjects are tabulated in Table 3. There was good correlation between the antigenic and functional activities of these natural coagulation inhibitors in healthy controls, but not in the uremic patients before dialysis treatment. The predialysis coefficient of correlation Table 3. Correlation Coefficients (Linear Regression Model) of the Antigenic and Functional Activities of the Three Natural Coagulation Inhibitors

Uremic patients Predialysis Postdialysis Controls

PC

PS

AT III

0.43 (NS) 0.64 (P < 0 0 . 1) 0.84 (P < 0.001)

0.44 (NS) 0.43 (NS) 0.60 (P < 0.02)

0.43 (NS) 0.37 (NS) 0.58 (P < 0 0. 2)

between the antigenic and functional activities of PC in the uremic patients was only 0.43, but a highly significant linear correlation was demonstrated after dialysis treatment (P < 0.01). DISCUSSION

A high incidence of cerebrovascular accident leading to a 10% to 12 % mortality rate in both adult and pediatric dialysis populations 20 and increased thrombosis of vascular access in patients on maintenance dialysis21raise the potential risk of thrombosis in uremic patients on dialysis. A decrease in plasma AT III levels during dialysis has been suggested to be associated with the risk of thrombogenesis , ·2 but other investigators have failed to confirm this observation. 3.4.22 The controversies related to previous studies in plasma AT III may arise from (1) differences in assay methods, and (2) the possible discrepancy between the antigenic and functional activities of these natural coagulation inhibitors in uremia. In the previous studies, the AT III functional activity was assayed in the presence of heparin23 and such assay could be affected by the heparin infused during hemodialysis . Our AT III functional assay was less likely to be affected by heparin infused during dialysis, as the thrombin activity was measured by its amidolytic activity on a chromogenic substrate after all available AT III activity was neutralized . We have shown a good correlation between the antigenic and functional activities of these proteins in healthy controls, but not in uremic patients on maintenance hemodialysis, suggesting the functional activities of these proteins are affected by uremia. Our present study showed that the predialysis (or interdialytic) plasma levels of these natural coagulant inhibitors in hemodialysis patients were

41

COAGULATION INHIBITOR LEVELS IN HEMODIALYSIS

different from those of healthy controls. Both the antigenic and functional activities of AT III in uremic patients were significantly lower than those of controls. Low plasma PC activity had been speculated to contribute to thrombotic complications in uremia,23 but we did not demonstrate a difference in PC activity between our patients and healthy controls. PS exists in plasma in two forms and only free PS is active as a cofactor for the anticoagulant effect of activated Pc. 24 Apparently, increased levels of PS recently have been observed in patients on continuous ambulatory peritoneal dialysis,25 but the plasma PS levels in hemodialysis patients have not been studied previously. As observed in patients with nephrotic syndrome, 24 uremic patients on regular hemodialysis had a higher total PS antigen level but a lower free PS antigen level compared with controls. In nephrotic syndrome, decreased plasma free PS antigen has been speculated to be caused by selective urinary loss of free PS,24 although some investigators fail to support this theory.26 Loss of free PS antigen during hemodialysis is unlikely, as neither of these natural coagulation inhibitors was detected in the dialysate. Other explanations, including an elevation of C4b-binding protein levels that favor complex formation 24 and a possible shift from active to inactive forms (such as free to bound PS), 26 remain speculative. Both elevation and reduction of plasma AT III and PC levels have been reported with hemodialysis treatment.2.4.16.23.27.28 Postdialysis elevation of these proteins has been suggested to result from a combination of factors, namely, (I) removal of dialyzable substance with inhibitory effect of these proteins l6 ; (2) release of AT III with damage of platelets and blood vessels 4; and (3) increased plasma protein concentration as a result of water depletion during dialysis. 3 On the contrary, postdialysis reduction of AT III and PC has been suggested to result from increased consumption of AT III by infused heparin or by thrombin generated through platelet activation, which could then lead to PC consumption. 222 We have observed no significant change in antigenic and functional activities of AT III or PS with hemodialysis. Nevertheless, a progressive increase of functional activity of PC was documented with dialysis. Increased plasma protein concentration as a result of water depletion during dialysis may not be significant, as there was no simultaneous elevation of AT III or PS. Increased release of PC from damaged plate-

lets and blood vessels was less likely, as damaged platelets released both AT III and PC. Hence, the remaining possible mechanism appears to be the removal of a previously unknown PC inhibitor in plasma from uremic patients, which could be a causal factor in atherosclerotic and thrombotic complications affecting a high proportion of these patients. 16 Similar to the findings of Sorensen et al,28 we have demonstrated an improvement of correlation coefficient between the antigenic and functional activities of PC with dialysis therapy. However, this finding was not observed with plasma AT III or PS. In conclusion, our study suggests hemodialysis does not increase the thrombotic risk in uremic patients, as the dialysis does not reduce the plasma natural coagulation inhibitors. Thrombotic risk may be reduced as the plasma PC levels increase with dialysis, possibly due to the removal of an inhibitory substance on PC functional activity. REFERENCES I. Jorgensen KA, Stoffersen E: Heparin-induced antithrombin III deficiency. Lancet 2:1231-1232, 1977 2. Brandt P, Jespersen J, Sorensen LH: Antithrombin III and platelets in haemodialysis patients. Nephron 28: 1-3, 1981 3. Jorgensen M, Eriksen HO, Tranebjaerg L: Plasma antithrombin III concentration in patients on regular haemodialysis treatment. Nephron 40:22-24, 1985 4. Woo HT, Wei SS, Lee EJC, et al: Effects of haemodialysis and peritoneal dialysis on antithrombin III and platelets. Nephron 40:25-28, 1985 5. Stenflo J: A new vitamin K-dependent protein: Purification from bovine plasma and preliminary characterization. J Bioi Chern 251:355-363, 1976 6. Walker FJ: Regulation of activated protein C by a new protein: A possible function for bovine protein S. J Bioi Chern 255:5521-5524, 1980 7. Esmon NT, Owen WG, Esmon CT: Isolation of membrane bound cofactor for thrombin-catalyzed activation of protein C. J Bioi Chern 257:859-864, 1982 8. Marlar RA, Kleiss AJ, Griffin JH: Human protein C: Inactivation of factors V and VIII in plasma by the activated molecule. Ann NY Acad Sci 370:303-310, 1981 9. Gonzalez R, Alberca I, Sala N, et al: Protein C deficiency. Response to Danazol and DDAVP. Thromb Haemost 53:320-322, 1985 10. Mannucci PM, Vigano S: Deficiencies of protein C, an inhibitor of blood coagulation. Lancet 2:463-466, 1982 11. Harris KW, Esmon CT: Protein S is required for bovine platelets to support activated protein C binding and activity. J Bioi Chern 260:2007-2010, 1985 12. Stern DM, Nawroth Pp, Harris K, et al: Cultured bovine aortic endothelial cells promote activated protein C-protein Smediated inactivation of factor Va. J Bioi Chern 261:713-871, 1986 13. Comp PC, Esmon CT: Recurrent venous thromboem-

42 bolism in patients with a partial deficiency of protein S. N Engl 1 Med 311:1525-1528,1984

14. Vigano-D'Angelo S, D'Angelo A, Kaufman CE lr, et al: Protein S deficiency occurs in the nephrotic syndrome. Ann Intern Med 107:42-47, 1987 15. Sorensen Pl, Nielsen AH, Knudson F, et al: Defective protein C in uraemia. Blood Purif 5:29-32, 1987 16. Mehta RL, Scott G, Sloand lA, et al: Skin necrosis associated with acquired protein C deficiency in patients with renal failure and calciphylaxis. Am 1 Med 88:252-257, 1990 17. Laurell CB: Quantitative estimation of proteins by electrophoresis in antibody containing agarose gel. Protein BioI Fluids 14:499-512, 1966 18. Francis RB lr, Seyfert U: Rapid amidolytic assay of protein C in whole plasma using an activator from the venom of Agkistroden Contortrix. Am 1 Clin Pathol 87:619-625, 1987 19. Comp PC, Doray D, Patton D, et al: An abnormal plasma distribution of protein S occurs in functional proteins deficiency. Blood 67:504-508, 1986 20. Brunner FP, Wing Al, Dykes SR, et al: International review of renal replacement therapy: Strategies and results in Maher IF (ed): Replacement of Renal Function by Dialysis (ed

LAI ET AL

3). Dordrecht, The Netherlands, Kluwer, 1989, pp 693-719 21. Connolly JE, Brownell DA, Levine EF, et al: Complication of renal dialysis access procedures. Arch Surg 119: 13251328, 1984 22. Marciniak L, Gockerman lP: Heparin-induced decrease in circulating antithrombin III. Lancet 2:581-584, 1977 23. Sorensen Pl, Knudsen F, Nielsen AH, et al: Protein C activity in renal disease. Thromb Res 38:243-249, 1985 24. Vigano-D' Angelo S, D'Angelo A, Kaufman CE, et al: Protein S deficiency occurs in the nephrotic syndrome. Ann Intern Med 107:42-47, 1987 25. Vaziri ND, Shah GM, Winer RL, et al: Coagulation cascade, fibrinolytic system, antithrombin III, protein C and protein S in patients maintained on continuous ambulatory peritoneal dialysis. Thromb Res 53:173-180, 1989 26. Gouault-Heilmann M, Gadelha-Parente T, Levent M, et al: Total and free protein S in nephrotic syndrome. Thromb Res 49:37-42, 1988 27. Alegre A, Vicente V, Gonzalez R, et al: Effect of hemodialysis on protein C levels. Nephron 46:386-387, 1987 28. Sorensen Pl, Knudsen F, Nielsen AH, et al: Low protein C in hemodialysis patients. Blood Purif 2:93-97, 1984