THROMBOSIS RESEARCH 21; 207-213, 1981 0049-3848/81/010207-05$02.00/O Printed in the USA. Copyright (c) 1981 Pergamon Press Ltd. All rights reserved.
SIMPLIFIED PROCEDURE FOR THE ASSAY OF FIBRINOPEPTIDE A IN PLASMA Harry van Hulsteijn, Ernest [email protected]
and Rogier Bertina Thrombosis and Haemostasis Research Unit, University Hospital Leiden, Rijnsburgerweg 10, 2333 AA LEIDEN, The Netherlands
(Received 11.9.1980; in revised form 9.12.1980. Accepted by Editor J.J. Sixma)
INTRODUCTION Since the development by Nossel in 1971 of a sensitive and specific radioimmunoassay (RIA) for human Fibrinopeptide A (FPA) in plasma (1) the clinical application of the FPA assay for example as an aid to diagnose ongoing thrombosis has been limited because of the long duration of the various procedures, the intricate processing of the samples and the limited number of samples that can be processed per assay (2-9). Therefore we introduced a modification in the FPA assay by omitting the 24 h dialysis step after ethanol precipitation, which allowed the rapid and accurate determination of a large number of samples. Moreover the present assay is simple to perform and needs less blood, which makes it a useful instrument for routine clinical use.
MATERIALS Standard FPA, desaminotyrosyl-FPA and rabbit anti-human FPA serum were purchased from IMCO Corporation Ltd. (Stockholm, Sweden). Na -lz51 was used from the Radiochemical Centre (Amersham, England). Ovalbumin was obtained from Koch-Light Laboratories Ltd. (Colnbrook, England). Heparinum sodium (Thromboliquine) was purchased from Organon (Oss, The Netherlands), aprotinin (Trasy101) from Bayer (Leverkusen, Western Germany) and charcoal (Norit; carbo adsorbens N. Ph. VI) from N.V. Algemeene Norit Maatschappij (Amsterdam, The Netherlands). METHODS Iodination of desaminotyrosyl - FPA Iodination of desaminotyrosyl - FPA was performed as described in the IMCOprotocol using the chloramine-T method (10). Iodinated FPA was separated from free iodide by passage through a Sephadex G-15 column equilibrated with 0.05 M Tris, 0.10 M NaCl, pH 7.5 and pretreated with 10 mg ovalbumin. Key-words : Fibrinopeptide A, radioimmunoassay, fibrinogen, ethanol. 207
RAPID FIBRINOPEPTIDE A ASSAY
Preparation of plasma After a flawless venepuncture with a 1.0 mm Wassermann needle 2.25 ml blood was collected into a calibrated polystyrene tube (Sarstedt Benelux B.V., Bladel, The Netherlands) containing 0.25 ml 0.15 M NaCl with 250 IU of heparin and 250 KIU of aprotinin. Raising the concentration of heparin and aprotinin 5 fold as well as adding ethylene diaminetetraacetic acid (EDTA) to the anticoagulant mixture did not change the results of the FPA assay. The content of the tube was mixed thoroughly, placed immediately in icewater and stored up to a maximum period of 1 h. Hereafter the blood was centrifuged for 30 min at 2000 g and 4' C. About 1.2 ml of the supernatant was stored in a plastic tube at -70' C prior to FPA determination. Separation of FPA from fibrinogen in plasma The separation of FPA from fibrinogen in plasma was performed both according to the IMCO protocol (ethanol precipitation and dialysis) and by ethanol precipitation alone. The patients plasma was thawed at 37' C and kept at 0' C for at least 15 min. Next, 1 ml plasma was mixed with 1 ml ethanol (> 99.8%) of 0' C. The mixture was kept at 0' C for 30 min. After centrifugation for 20 min at 2000 g and 4' C the tube was kept again at 0' C for 30 min followed by a second centrifugation (20 min, 2000 g, 4' C). 1 ml of the supernatant was removed and mixed thoroughly in a polystyrene tube with 1 ml of 0.05 M Tris, 0.10 M NaCl, pH 8.5 (0' C). The mixture was kept at 0' C and processed immediately. Radioimmunoassay of FPA The radioimmunoassay of FPA was performed essentially according to the IMCO protocol: incubation of test sample with FPA-antiserum for a period of 18 h at 4' C, addition of '251-FPA and incubation for 1 h at 4' C, and the separation of free and bound 1251-FPA by ovalbumin coated charcoal. One main modification was introduced: since by omitting the dialysis step the concentration of ethanol in the test sample is 25%, the final concentration of ethanol in the standard dilutions of FPA was also brought to 25%. We examined the possibility to reduce the 18 h incubation period of the test sample with the FPA-antiserum to only 2 h at room temperature. All patient samples were assayed in two dilutions each in duplicate. The buffer used for dilution of the samples was 0.05 M Tris, 0.10 M NaCl, 0.1% ovalbumin, 25% ethanol, pH 8.5.
RESULTS Standard curve In 12 consecutive standard curves obtained within a period of three months with the same '251-FPA and FPA-antiserum concentrations, the mean amount of FPA causing 50% inhibition of binding was 0.15 ng/tube and the mean amount of FPA, that caused 20% inhibition of binding was 0.058 ng/tube (fig. 1). The standard curve obtained with the standard FPA from IMCO has been compared with a curve, obtaining by using standard FPA purchased from SchwarzMann (Orangeburg, N.Y., U.S.A.) generously provided by Dr. J.A. van Mourik from the Central Laboratory of the Netherlands Red Cross Blood Transfusion Service (Amsterdam, The Netherlands). There was no difference between the two standard FPA preparations with respect to the effectiveness by which they compete with the 12'I-FPA for binding to the anti-FPA antibodies. Although identical results were obtained by reducing the 18 h incubation of
RAPID FIBRINOPEPTIDE A ASSAY
the test sample with the FPA-antiserum to 2 h at room temperature we used for practical reasons the 18 h period. The lower detection limit of the assay is 0.06 ng FPA/tube which would correspond to 0.24 ng FPA/ml plasma. During a 3 month period the percentage of 1251-FPA bound to the FPA-antiserum in the absence of added FPA was reduced from 50 to 35%.
/ 1 / 1
FIG. 1 Standard curve for the radioimmunoassay of FPA. The drawn curve represents the mean of 12 separate curves. SD's are indicated by vertical bars. Effectiveness of ethanol in separating FPA from fibrinogen Firstly a possible co-precipitation of FPA with fibrinogen by ethanol was studied by adding 12'I-FPA to ten test samples before the ethanol precipitation. More than 90% of 12'I-FPA was recovered in the supernatant. This percentage proved to be independent of the FPA level in plasma (O-20 ng/ml). After a tenfold concentration of the ethanol supernatant (lyophilization)no significant amount of fibrinogen could be detected by rocket immunoelectrophoresis, indicating that more than 99.9% of the fibrinogen had been removed. To determine the actual amount of fibrinogen left in the ethanol supernatant FPA concentrations were measured before and after treatment of the supernatant with bovine thrombin (1 NIH unit/ml; 45 min at 37’ C) (Table 1).
TABLE 1 Effectiveness of ethanol in separating FPA from fibrinogen in plasma Fibrinogen in plasma
FPA in twice diluted ethanol supernatant before thrombin after thrombin
RAPID FIBRINOPEPTIDE A ASSAY
The analysis of the ethanol supernatant of sample 1 (normal plasma with normal FPA level; plasma fibrinogen 780 mg%) indicates that ethanol precipitation removed more than 99.99% of the fibrinogen. In this case it is not clear whether the actual observed FPA concentration (0.045 pmol/ml) is due to free FPA or to cross-reaction of the anti-FPA antibody with fibrinogen. In the latter case the affinity of the antibody for binding of FPA would be about 6.5 times higher than that for binding of fibrinogen. From the analysis of the ethanol supernatant of sample 2(D1C plasma with elevated FPA level; plasma fibrinogen 140 mg%) it is clear that even in the most unfavourablecase residual fibrinogen never can account for the elevated FPA concentrations in this case.
FIG. 2 Recovery ci'added standard FPA in the FPA assay. Variable amounts of FPA (O-40 ng) were added both to plasma of a healthy individual (O-O) and to DIC plasma (0-O). The dotted lines represent 100% recovery. 0
FPA added lng /ml plasma)
Recovery of FPA The amount of FPA recovered in our assay procedure was determined by adding increasing concentrations of standard FPA (O-40 ng/ml) to plasma of a healthy individual (containing 0.60 ng FPA/ml) before the plasma was processed (see fig. 2). To determine the influence of increased plasma levels of FPA on the recovery, the experiment was repeated with plasma of a patient suffering from diffuse intravascular coagulation, which contained 19.70 ng FPA/ml. From fig. 2 it can be concluded that in both cases there was a constant recovery of the added FPA of about 100%. Table 2 compares results obtained with 6 patient plasmas (with elevated FPA levels) both after ethanol precipitation alone and after ethanol precipitation plus dialysis (24 h; IMCO protocol).
RAPID FIBRINOPEPTIDE A ASSAY
FPA levels measured in 6 patient plasmas using both ethanol precipitation in combination with dialysis and ethanol precipitation alone Ethanol precipitation
Ethanol precipitation plus dialysis
FPA in ethanol supernatant (ng/2ml)
FPA inside dialysis tube (ng/2ml)
3.52 2.96 3.60
0.88 0.90 1.80
4.86 4.40 3.86
5.12 4.93 4.48
5.40 3.46 4.34
5.78 3.80 5.10
FPA outside dialysis tube (ng/8ml)
It is clear that after a dialysis period of 24 h not in all cases an equilibrium distribution has been reached; although the total amount of FPA is rather independent of the actual method used, it appears less suitable to use only the FPA level observed in the outer compartment because of the variable degree to which equilibrium has been reached in the different dialysis procedures.
Stability of FPA in plasma FPA levels determined in plasma samples immediately after venepuncture did not differ from FPA levels determined after storage of the plasma for four months at -70' C.
Reproducibility and accuracy of the FPA assay The reproducibility of the FPA assay was tested as follows: in four different plasmas the FPA concentration was measured on four occasions (mean FPA levels ranging from 0.32 to 8.72 ng/ml); the mean inter-assay coefficient of variation was 5.90%. The accuracy of the FPA assay was tested by examining three different plasmas five times in the same assay (mean FPA levels ranging from 0.44 - 1.84 ng/ml); the mean intra-assay coefficient of variation was 2.47%.
FPA levels in plasma of healthy individuals and patients With the before described assay we determined FPA levels in the plasma of healthy individuals and of different patient groups (Table 3). The normal mean FPA value is 0.68 2 0.41 n&ml (range 0.24 - 1.88) which corresponds to 0.44 -+ 0.26 pmol/ml (range 0.15 - 1.22).
RAF'IDFIBRINOPEPTIDE A ASSAY
TABLE 3 FPA levels in plasma of healthy individuals and different patient groups FPA (ng/ml) Healthy individuals (n=40)' Deep venous thrombosis (n=15)' Pulmonary embolism (n=15J3 Diffuse intravascular coagulation (n=5J4 Postoperative patients (n=3j5
0.68 + 0.41 4.70 z 1.65 4.92 + 1.52 22.00 +- 7.22 143.40 5 43.20
(I): 20 males (age 21-45; mean 30 yrs), 20 females (age 19-51; mean 28 yrsl, no use of oral contraceptives. (2): Diagnosed by ascending phlebography. (3): Diagnosis based on the presence of one or more well-defined lobar or segmental defects on a 6-view perfusion lungscan with normal findings on the chest X-ray in the corresponding area. (4): Diagnosis based on a thrombocyte count < 100.lOg/l, plasma fibrinogen < 150 mg%, plasma antithrombin III antigen < 70% ( Laurel11 and serum Fibrin (ogen) degradation products > 80pg/ml (Laurell). (5): Venepuncture performed within 6 h postoperatively.
DISCUSSION Since the introduction of a RIA for the determination of FPA in plasma (I) various methods have been used for the separation of FPA from fibrinogen in plasma (2-9) because of the possibility of cross-reactivity of fibrinogen (degradation products) with the anti-FPA antibody (2,3,11). In this paper we introduce an important simplification in the FPA assay by omitting the 24 h dialysis procedure after ethanol precipitation; in our experience ethanol precipitation alone is sufficient to achieve a satisfactory separation of FPA from fibrinogen in plasma, as more than 90% of ('251)-FPA is recovered in the supernatant after ethanol precipitation (see also fig. 2) while the fibrinogen is removed with more than 99.99% efficiency. Further it was demonstrated that residual fibrinogen in the ethanol supernatant is too low to result in elevated FPA levels. Moreover the results of Table 2 indicate that at elevated plasma FPA levels most of the FPA like material is in a low molecular weight form because it passes the dialysis membrane. This rather points to FPA than to fibrinogen (degradation products). In our opinion? there is no need for the ethanol evaporation procedure described by Harenberg et al (91, because no important interference of the ethanol was found either with the antigen-antibody reaction or with the adsorption properties of the charcoal. The time necessary to perform the assay can further be shortened by reduction of the incubation time of the test sample with the FPA antiserum from 18 h at 4' C to 2 h at room temperature. This may be important in particular clinical situations such as a suspicion of thrombosis in a patient having contraindications against anticoagulant therapy. The results of FPA determinations in a group of 40 healthy individuals (mean FPA level 0.68 + 0.41 ng/ml) are in agreement with those mentioned by other investigators (2-91, although there is a big difference in reported recoveries varying from 25% (3) to 110% (5). Plasma FPA concentrations showed to be elevated in conditions, in which an
RAPID FIBRINOPEPTIDE A ASSAY
enhanced thrombin generation is to be expected, such as deep venous thrombosis, pulmonary embolism, diffuse intravascular coagulation and in the early postoperative period. However, it should be reminded that raised FPA levels are not necessarily caused by an increased thrombin generation (12). In our opinion the present modified assay of FPA fulfills the necessary requirements for routine clinical use.
ACKNOWLEDGEMENT We thank Mrs. Susan Leenheer for typing the manuscript.
REFERENCES 1. NOSSEL, H.L., YOUNGER, L.R., WILNER, G.D., PROCUPEZ, T., CANFIELD, R.E. and BUTLER JR., V.P. Radioimmunoassay of human fibrinopeptide A. Proc. Nat. Acad. Sci. U.S.A., 68, 2350-2353, 1971. 2. NOSSEL, H.L., YUDELMAN, I., CANFIELD, R.E., BUTLER JR, V.P., SPANONDIS, K., WILNER, G.D. and QURESHI, G.D. Measurement of fibrinopeptide A in human blood. J. Clin. Invest., 54, 43-53, 1974. 3. GERRITS, W.B.J., FLIER, 0.Th.N. and VAN DER MEER, J. Fibrinopeptide A immunoreactivity in human plasma. Thromb. Res., 2, 197-212, 1974. 4. BUDZYNSKI, A.Z. and MARDER, V.J. Determination of human fibrinopeptide A by radioimmunoassay in purified systems and in the blood. Thromb..Diath. Haemorrh., 34, 709-717, 1975. 5. CRONLUND, M., HARDIN, J., BURTON, J., LEE, L., HABER, E. and BLOCH, K.J. Fibrinopeptide A in plasma of normal subjects and patients with disseminated intravascular coagulation and systemic lupus erythematosus. J. Clin. Invest., 58, 142-151, 1976. 6. KOCKUM, C. Radioimmunoassay of fibrinopeptide A - Clinical applications. Thromb. Res., 8, 225-236, 1976. 7. HOFMANN, V. and STRAUB, P.W. A radioimmunoassay technique for the rapid measurement of human fibrinopeptide A. Thromb. Res., 11, 171-181, 1977. 8. YUDELMAN, I.M., NOSSEL, H.L., KAPLAN, K.L. and HIRSCH, J. Plasma fibrinopeptide A levels in symptomatic venous thromboembolism. Blood, 5l_, 11891195, 1978. 9. HARENBERG, J., HEPP, G. and SCHMIDT-GAYK, H. Fibrinopeptide A in human plasma-evaluation of a new radioimmunoassay technique on microliter-plates. Thromb. Res., 3, 513-522, 1979. 10.
HUNTER, H.M. and GREENWOOD, F.C. Preparation of iodine-131 labelled human growth hormone of high specific activity. Nature, 194, 495-496, 1962.
11. BUDZYNSKI, A.Z., MARDER, V.J. and SHERRY, S. Reaction of plasmic degradation products of fibrinogen in the radioimmunoassay of human fibrinopeptide A. Blood, 45, 757-767, 1975. 12. PEUSCHER, F.W., AKEN, W.G. VAN, FLIER, O.Th.N., STOEPMAN-VAN DALEN, E.A., CREMER-GOOTE, Th.M. and MOURIK, J.A. VAN. Effect of anticoagulant treatment measured by fibrinopeptide A (FPA) in patients with venous thromboembolism. Thromb. Res., 2, 33-43, 1979.