THROMBOSIS RESEARCH 58; 273-281,199O 0049-3848190 $3.00 + .OO Printed in the USA. Copyright (c) 1990 Pergamon Press pk. All rights reserved.
THE SimpliRED
D DIMER TEST: A NOVEL ASSAY FOR THE DETECTION OF CROSSLINKED FIBRIN DEGRADATIONPRODUCTS IN WHOLEBLOOD
, D.B. Rylattl , P.G. Bundesenl M.A. John’, M.J. Elms*, E.J. O’Reillyl and C.J. Hillyard’. I AGEN BIOMEDICAL LIMITED, 11 Durbell Street, P.O. Box 391, Acacia Ridge, Qld, Australia, 4110. 2 Department of Haematology, Royal Brisbane Hospital, Herston Road, Brisbane, Qld, Australia, 4029.
(Received
15.11.1989;
accepted
in revised form 15.2.1989
by Editor F.J. Morgan)
ABSTRACT A new system for the detection of fibrin degradation products in whole blood has been developed. The test provides a clearly visible agglutination of the patient’s red blood cells in the presence of elevated levels of the crosslinked fibrin derivative, D dimer. The test, which uses a bispecific reagent prepared from Fab’ fragments of One monoclonal antibodies, gives a positive result in l-2 minutes. monoclonal antibody (RAT-lC3/86) was raised against human red blood cells, and the second (DD-3B6/22) was specific to the crosslinked fibrin derivative, D dimer. Addition of the bispecific reagent to a drop of patient’s whole blood resulted in red blood cell agglutination when elevated levels of D dimer were present in the Clinical trials showed sensitivity equivalent to that of sample. current commercial tests. Samples from patients with thrombotic disease states as well as normals were examined. The test was compared with commercial latex agglutination and enzyme immunoassay systems and showed good correlation with the presence of elevated levels of crosslinked fibrin degradation products. This technology represents an advance which allows rapid “on the spot” whole blood analysis, for the diagnosis of thrombotic disorders.
INTRODUCTION The determination of fibrin/fibrinogen breakdown products (FDP) present in More recently, serum is important in the diagnosis of thrombotic disorders. the determination of crosslinked fibrin degradation products (XDP) has proved Rapid and is replacing the FDP assay. specific for fibrin breakdown, methods employed in the detection of FDP have included agglutination inhibition assays, using either tanned red blood cells (1) or latex particles Key words:
Bispecific
antibody,
whole
blood 273
immunoassay,
D dimer
274
THE SimpliRED D DIMER TEST
Vol. 58, No. 3
Staphylococcal clumping was an early system coated with fibrinogen (2). utilizing the characteristic clumping of some strains of staphylococci in the Improved agglutination assays based upon presence of fibrinogen (3). antibody coated latex particles have produced a rapid diagnostic system (4all these methods require the processing of the blood sample to ~~odu~~w~~~~~a or serum. We have developed a new system for the detection of circulating antigen or The test indicates a positive result by visible antibody in whole blood (7). The utilization of agglutination of the patient’s own red blood cells. endogenous red blood cells removes the need for extensive sample preparation, producing a unique test which is rapid and simple. The system uses a monoclonal antibody produced against red blood cells, which is conjugated to either an antigen or antibody. This reagent links the red blood cells and the circulating antibody or antigen, resulting in agglutination of the cells. The assay of XDP uses a monoclonal antibody to the crosslinked derivative D dimer, previously employed in both a quantitative enzyme immunoassay (8) and a semi-quantitative latex agglutination test (6). Although the latex test is accurate and rapid, it still requires sample preparation as the The SimpliRED D dimer initial step, confining the assay to the laboratory. test enables patient assessment in situations where laboratory facilities are not available.
METHODS Preparation of human XDP: Crosslinked fibrin clots were prepared and lysed as described by Masci et al (8). The lysate from 500 mL of plasma was concentrated by pressure dialysis to a volume of 20 mL (762 mg) and solid urea was added to give a final concentration of’3 M. XDP was purified by Sephacryl S300 gel chromatography at a flow rate of 1.0 mL/min on a 2 x 100 cm column in a buffer containing 3 M urea and 0.1 M sodium bicarbonate pH 8.4. Fractions containing XDP (59 mL; 480 mg) were pooled and dialyzed first against 100 volumes of phosphate buffered saline pH 7.4, and finally 50 volumes of a phosphate buffered saline containing 50% glycerol and 0.01% sodium azide. This preparation contains 189K Da dimers and higher molecular weight aggregates (ie Two D dimers with MWapprox. 400K Da by HPLC). The XDP preparation was stored in aliquots at -2OOC. Production of monoclonal antibodies: Two monoclonal antibodies were utilized in the development of this red blood cell agglutination system. The antibody DD-3B6/22, specific for human D dimer, was produced from the immunization of BALB/c mice with XDP, as described previously (9). The second antibody, RATlC3/86, was the result of immunization of BALB/c mice with washed human red blood cells as described by Rylatt et al (7). Two screening procedures selected the antibody which recognized the membrane antigen, glycophorin A, without causing agglutination of the cells. Additional assays ensured that the antibody would react with erythrocytes from all patients, regardless of blood group. Production of conjugation of Brennan et al acetate buffer to digest each
bispecific reaqent: The test reagent was produced by Fab’ fragments using a modification of the method described by (10). The antibody solutions were equilibrated into an 0.1 M pH 3.5, containing 0.07 M NaCl. A 1% pepsin solution was used intact monoclonal antibody to produce the F(ab’)?: fragments.
Vol. 58, No. 3
THE SimpliRED D DIMER TEST
275
The pepsin and antibody were combined in a 1:lOO ratio, digested at 37OC, then stopped by adjusting the pH to 8.0 with 1.5 M Tris buffer. Digestion times were 15 min for DD-3B6/22 and 45 min for RAT-lC3/86. The digests were purified by gel filtration on a Sephacryl S300 column equilibrated in 100 mM phosphate buffer, pH 6.8. The resulting F(ab’)n pool was concentrated by ultrafiltration to give a concentration of approximately 3 mg/mL. The F(ab’)z fragments were subsequently split into Fab’ fragments by reduction with 1 mM mercaptoethylamine in 100 mM phosphate buffer pH 6.8 containing 10 mM sodium arsenite and 1 mM EDTA. The time of exposure to the reducing agent varied for each antibody. DD-3B6/22 F(ab’)z was allowed to reduce for 2 hr at room temperature (RT), while RAT-lC3/86 reduced for 3 hr at 37OC. The hinge region sulfydryl groups on the DD-3B6/22 Fab’ species were blocked by the addition of 5 mM DTNB and mixed at RT for 3 hr. The RATlC3/86 Fab’ species were prepared only as required and left in the unblocked The preparations were again purified by gel filtration on a reduced form. The two Fab’ species were mixed together at RT for 2 Sephacryl S300 column. hr, which resulted in recombination of the fragments to produce the final bispecific antibody at an average yield of 50%. The reaction was monitored by HPLC analysis on a Waters 300SW column to ensure optimum production of product. Once achieved, the bispecific antibody was purified by gel filtration on a Sephacryl S300 column. The resulting pool was diluted in 100 mM phosphate buffer, pH 6.8 to give the desired sensitivity. The final formulation also included 5 g/L BSA, and an unrelated monoclonal antibody of similar isotype, to block any non-specific reaction. Agglutination test vrocedure: The test procedure required 10 uL of whole blood (taken in the form of a finaer nrick or henarinized blood sample) which The slide was was combined on a slide with 25 pL of-bispecific-reagent. rocked for l-2 minutes and a positive or negative agglutination reaction recorded. The semi-quantitative Latex aqqlutination procedure: test Dimertest latex was used as a comparative assay. 10 pL sample of plasma was mixed with 25 uL of latex agglutination rocked for 3 minutes (6) . The subsequent
latex agglutination On a glass slide, solution and gently was then recorded.
a
The quantitative two-site enzyme immunoassay Enzyme immunoassay procedure: Dimertest EIA was used to determine the D dimer levels of all samples The assay as outlined by Rylatt et al (9), uses DD-3B6/22 as the examined. capture antibody, and a panspecific antibody in the final labeling step. Sensitivity: Undiluted SimpliRED D dimer test reagent was examined against to which D dimer antigen had been added in group 0 negative whole blood, The lowest D dimer concentration which still produced a serial dilutions. positive agglutination pattern was recorded as the sensitivity limit. The SimpliRED D dimer test reagent was diluted to give a Stabilitv trials: A1iquot.s of the test sensitivity limit of approximately 0.25 mg/L D dimer. over a 28 day reagent were placed at 37OC, RI’, 4OC and -20°C and examined The reagent was examined against group 0 negative whole blood period. samples to which D dimer antigen had been added to give final concentrations A control of unspiked blood of 2.50, 1.25, 0.63,0.31, 0.16 and 0.08 mg/L. The sample with the lowest D dimer concentration still was always included. resulting in a positive agglutination pattern was recorded as the sensitivity of the reagent. Clinical
trials:
Two trials
were
conducted
during
the
examination
of
this
276
THE SimpliRED
D DIMER TEST
Vol. 58, No. 3
bispecific reagent. A preliminary evaluation involved the examination of Heparinized randomly selected hospital patients and normal individuals. samples were provided by the staff at Princess Alexandra Hospital, Brisbane, latex agglutination and red blood and tested in-house by enzsyme immunoassay, cell agglutination. A series of normals were also tested by both heparinized The bispecific reagent was adjusted to a and finger prick blood samples. sensitivity of approximately 0.20 mg/L, above which agglutination would occur. An independent clinical trial was undertaken by one of us (MJE). A selection of normal and thrombotic disease patients were examined by enzyme immunoassay, latex agglutination and red blood cell agglutination. For this trial, the sensitivity of the bispecific reagent was adjusted to 0.25 mg/L The anticoagulant such that blood from normal subjects did not agglutinate. used was sodium citrate. RESULTS Sensitivity: The sensitivity of this system was seen to be comparable to that of the enzyme immunoassay, with the lower limit of detection being 0.06 was consistent between batches. No prozone mg/L XDP. This sensitivity effect was observed in any experiments. The highest antigen concentration examined was 1.2 x 103 mg/L, with agglutination proportional to concentration. Figure 1 shows the agglutination patterns obtained with dilutions of antigen, compared with a negative reaction. The agglutination is clearly discernable. Our experience has been that the agglutination of the red blood cells is easier to read than that of latex particles.
Fiq. 1 Agglutination of red blood cells by the bispecific reagent in the From left to right the D dimer presence of crosslinked fibrin derivatives. concentrations were 1.75, 0.44, 0.11 and 0 mg/L (as determined by DIMERTEST EIA) .
Vol. 58, No. 3
277
THE SimpliRED D DIMER TEST
Stability: An accelerated stability trial of the bispecific antibody was set The reagent samples stored between -2O*C and UP at -20°C, 4OC, RT, and 37’C. RT remained stable throughout the 28 days, while those stored at 37OC started In an accelerated stability test a week at to lose activity after 10 days. 37OC is considered to represent stability for one year at 4OC. Clinical trials: Preliminary trials were carried out on whole blood samples from 67 hospital patients (at the Princess Alexandra Hospital in Brisbane) as well as a series of normals (Table 1). Of the 30 normals, 22 were examined by both finger prick and heparinized venous sampling. All samples showed The average concentration of D dimer for samples negative negative results. on the SimpliRED test was 0.10 mg/L, while those positive had an average concentration of 0.63 mg/L, showing a significant variation between positive and negative. The correlation was good between all detection systems. Twenty-six samples were recorded as positive on both latex and SimpliRED tests. While 12 were positive on the SimpliRED test and negative on latex, none of these were diagnosed as thrombotics, and therefore indicated that the Adjustment of the sensitivity of the bispecific reagent was too sensitive. bispecific reagent to the top of the normal range was made prior to the subsequent independent clinical trial, to match that of Dimertest Latex.
TABLE 1 Prelim #inary
EIA D Dimer
Cl i nical
Cone
(mg/L)
Latex SimpliRED
-ve
Evaluation*
of Bispecif
ic
-ve
Reagent
Latex
SimpliRED
+ve
+ve
Simpl iRED -ve
Simp 1iRED +ve
co.20
51
7Y
0
0
0.20-0.50
6
8
0
8
jo.50
1
2
* Total
number
of samples
18
was 97
The trial conducted at Royal Brisbane Hospital examined 101 normal, 18 intravascular coagulation (DIG), and Pulmonary embolism (PE), 32 Disseminated The results were 21 Deep venous thrombosis (DVT) whole blood samples. comparable with the latex assay and showed no false diagnosis by the The average D dimer concentration for normals was SimpliRED D dimer test. 0.09 mg/L, while PE samples showed 0.80 mg/L, and both the DIG’s and DVT’s averaged 0.96 mg/L (Figure 2).
278
THE SimpliRED D DIMER TEST
Vol. 58, No. 3
??
1 ??
?? ??
f
Fiq. 2 The examination of the SimpliRED D dimer test for the diagnosis of thrombotic disease states, (N: Normals, PE: Pulmonary embolism, DIC: DVT: Deep venous thrombosis). A Disseminated intravascular coagulation, is positive agglutination of red cells is indicated by ??, while a negative shown by ??.
DISCUSSION The time required for immunoassays has decreased with advances in technology, however the rate limiting step in most systems has been sample preparation. By using the patient’s whole blood immediately upon sampling, time and sample handling are significantly reduced. The production of a bispecific antibody enabled two separate antigenic sites (e g red blood cell membrane and D dimerl to be recognized by the one reagent component. A multivalent antigen acts as a bridge between two bispecific antibodies, and agglutination of the red cells results. In the case of fibrin degradation analysis, the crosslinked subunits enable formation of such a matrix. Where the antigenic site is not multivalent, the system can be modified to a two-site assay. Two different bispecific antibodies can be prepared from monoclonal antibodies against separate antigenic sites on the antigen. The reagent would then contain a mixture of these bispecific antibodies. Alternatively, an inhibition assay, as described by Rylatt et al (71, also evaluates antigens with unique epitopes. Preliminary correlation observed.
evaluation studies of the XDP bispecific reagent showed good between the EIA D dimer values and the agglutination patterns As a finger prick was all that was required to sample the
Vol. 58, No. 3
THE SimpliRED D DIMER TEST
279
Patient’s Status, the system was considered ideal for use outside the laboratory. A high level of sensitivity was seen, with the undiluted reagent detecting a concentration of 0.06 mg/L D dimer, comparable with enzyme immunoassays. For use as a rapid diagnostic tool the reagent can be set to a sensitivity where normal individuals show no agglutination. For our preliminary evaluation, the bispecific reagent was set at 0.20 mg/L as an approximate value for normal levels. This proved to be more sensitive than required in the clinical situation. Twelve patients showed positive agglutination on the SimpliRED test while negative with latex. None of these patients had thrombotic disorders. At the other end of the scale, the SimpliRED D dimer test missed two samples showing elevated D dimer levels of 0.52 and 0.88 mg/L. Again, in neither case was thrombosis diagnosed. The major difference between these systems and the Dimertest EIA is the form of the antigen detected. In the clinical situation, high molecular weight derivatives are the predominant fibrin breakdown product (ll-14), and it is these species the latex and SimpliRED assays recognise. Earlier work indicated that the antibody DD-3B6/22 is specific to a single site on the D dimer component of XDP preparations, close to the crosslinking site (15). Consequently, both the Dimertest latex and the SimpliRED D dimer test require the higher molecular weight forms to provide the two sites necessary to bridge the red cells. The Dimertest EIA uses two separate antibodies each This enables detection of D seeing different sites on the D dimer molecule. dimer in 189K Da and higher molecular weight forms. With normals who have will occur despite a high levels of the single dimer form, no agglutination In patients suffering thrombosis, the concentration of high EIA reading. crosslinked derivatives is significantly higher than the dimer form, and agglutination will occur. The second trial was directed specifically at the diagnosis of thrombotic disease. The XDP levels of patients with thrombotic disease were hence a rapid yes/no answer is possible. significantly higher than normals, The bispecific antibody was able to diagnose correctly all hospital Patients examined, and in the hospital laboratory the test reagent was considered to the degree of agglutination is be rapid, simple and accurate. As seen, proportional to the amount of XDP present. The time in which the agglutination occurs also is dependent upon the concentration of antigen. the more rapid the agglutination. If a more The higher the concentration, the system could be used in accurate assessment of XDP levels were required, and adapted to equipment such as conjunction with a standard antigen series, a laser or kinetic reader, for calibration of the Pattern. A semiquantitative assessment could be made by simply diluting the test reagent. Using a standard antigen concentration as a guide, an estimate of the antigen level could be made. The whole blood assay presented in this paper rewires no centrifuge or and the result is read in two specialized equipment, is simple to perform, The agglutination pattern is easy to assess and therefore suitable minutes. For these reasons it is considered for personnel with limited training. ideal for doctor’s surgery or emergency room analysis, where rapid diagnosis or for field studies and Third World applications, where no is advantageous, equipment is available. ACKNOWLEDGEMENTS We would Hospital, enabling
like to thank Tony Badrick and staff at the Princess Alexandra Brisbane, for their assistance in the suPPlY of Patient samples, our preliminary testing of the bispecific antibody.
280
THE SimpliRED
Vol. 58, No. 3
D DIMER TEST
REFERENCES 1. MERSKEY, C., LALEZARI, P. and JOHNSON, A.J. method for measuring fibrinolytic split products Exv. Biol. Med., 131, 871-875, 1969. 2. ALLINGTON, M.J. Detection latex clumping method. Stand.
of fibrin(ogen1 J. Haem., l&
A rapid, simple in human serum.
sensitive Proc. SOC.
degradation products 115-119, 1971,
by a
3. HAWIGER, J., NIEWIAROWSKI, S., GUREWICH, v. and THOMAS, D.P. Measurement of fibrinogen and fibrin degradation products in serum by staphylococcal clumping test. J. Lab. Clin. Med., 75, 93-108, 1970. 4. ELLMAN, L., CARVALHO, A. and COLMAN, R.W. The Thrombo-Wellcotest N. Eng. screening test for disseminated intravascular coagulation. 288, 633-634, 1973. 5. ELMS, M., BUNCE, I., P.P. and WHITAKER, A.N. products in Plasma using Clin. Path., 25, 360-364,
J.
as a Med.,
BUNDESEN, P.G., RYLATT, D.B., WEBBER, A.J., MASCI, Rapid detection of cross-linked fibrin degradation monoclonal antibody coated latex particles. Am. J. 1986.
6. HILLYARD, C.J., BLAKE, A.S., WILSON, K., RYLATT, D.B., MILES, S., BUNCH, R ., ELMS, M.J., BARNES, A. and BUNDESEN, P.G. A Latex Agglutination Assay for D Dimer: Evaluation and Application to the Diagnosis of Thrombotic Clin. Chem., 33 (101, 1837-1840, Disease. 1987. 7. RYLATT, D.B., KEMP, B.E., BUNDESEN, P.G., JOHN, M.A., O'REILLY,E.J., COTTIS, L.E., MILES, S.J., KHAN, J.M., DINH, D.P., STAPLETON, D. and HILLYARD, C.J. A rapid whole blood immunoassay system. Aust. Med. J., 152, 75-77, 1990. WHITAKER, A.N., ELMS, M.J., MASCI, P.P., BUNDESEN, P.G., RYLATT, D.B., 8. WEBBER, A.J. and BUNCE, I.H. Measurement of cross linked derivatives in using monoclonal antibodies. J. Clin. Pathol., 37, plasma : an immunoassay 882-887, 1984. RYLATT, D.B., BLAKE, A.S., COTTIS, L.E., MASSINGHAM,D.A., FLETCHER, 9. WHITAKER, A.N., ELMS, M., BUNCE, I., WEBBER, A.J., WYATT, W.A., MASCI, P.P., An immunoassay for human D dimer using monoclonal D. and BUNDESEN, P.G. Thromb. Res., 31, 767-778, 1983. antibodies. of bispecific BRENNAN, M., DAVISON, P.F. and PAULUS, H. Preparation 10. antibodies by chemical recombination of monoclonal immunoglobulin G1 fragments. Science, 229, 81-83, 1985. 11. GAFFNEY, P.J. The occurrence and clinical relevance in blood. Ann. New York Acad. SC., 408, 407-425, 1983.
of fibrin
fragments
12. WHITAKER, A.N., ELMS, M.J., MASCI, P.P., BUNDESEN, P.G., RYLATT, D.B., WEBBER, A.J. and BUNCE, I.H. Measurement of cross linked fibrin derivatives an immunoassay using monoclonal antibodies. J. Clin. Pathol., 37, in plasma: 882-887, 1984. CREIGHTON, L.J., 13. GAFFNEY, P.J., CALLUS, M. and THORPE, R. Monoclonal antibodies to crosslinked fibrin degradation products (XL-FDP). II.
Vol. 58, No. 3
Evaluation 1988.
THE SimpliRED D DIMER TEST
in
a variety
of clinical
conditions.
Br.
J.
14. CARR,. J.M., McKINNES, M. and McDONAGH, J. Diagnosis intravascular coagulation. Role of D dimer. Am. J. Clin. 1989. 15. GREENBERG, C.S., dimer levels with the Am. J. Clin. Pathol.,
281
Haem.,
68,
91-96,
of disseminated Path., 91, 280-287,
DEVINE, D.V. and McCRAE. Measurement use of a monoclonal antibody coupled 81, 94-100, 1987.
of plasma fibrin to latex beads.
II
THE SimpliRED
D DIMER TEST: A NOVEL ASSAY FOR THE DETECTION OF CROSSLINKED FIBRIN DEGRADATIONPRODUCTS IN WHOLEBLOOD
, D.B. Rylattl , P.G. Bundesenl M.A. John’, M.J. Elms*, E.J. O’Reillyl and C.J. Hillyard’. I AGEN BIOMEDICAL LIMITED, 11 Durbell Street, P.O. Box 391, Acacia Ridge, Qld, Australia, 4110. 2 Department of Haematology, Royal Brisbane Hospital, Herston Road, Brisbane, Qld, Australia, 4029.
(Received
15.11.1989;
accepted
in revised form 15.2.1989
by Editor F.J. Morgan)
ABSTRACT A new system for the detection of fibrin degradation products in whole blood has been developed. The test provides a clearly visible agglutination of the patient’s red blood cells in the presence of elevated levels of the crosslinked fibrin derivative, D dimer. The test, which uses a bispecific reagent prepared from Fab’ fragments of One monoclonal antibodies, gives a positive result in l-2 minutes. monoclonal antibody (RAT-lC3/86) was raised against human red blood cells, and the second (DD-3B6/22) was specific to the crosslinked fibrin derivative, D dimer. Addition of the bispecific reagent to a drop of patient’s whole blood resulted in red blood cell agglutination when elevated levels of D dimer were present in the Clinical trials showed sensitivity equivalent to that of sample. current commercial tests. Samples from patients with thrombotic disease states as well as normals were examined. The test was compared with commercial latex agglutination and enzyme immunoassay systems and showed good correlation with the presence of elevated levels of crosslinked fibrin degradation products. This technology represents an advance which allows rapid “on the spot” whole blood analysis, for the diagnosis of thrombotic disorders.
INTRODUCTION The determination of fibrin/fibrinogen breakdown products (FDP) present in More recently, serum is important in the diagnosis of thrombotic disorders. the determination of crosslinked fibrin degradation products (XDP) has proved Rapid and is replacing the FDP assay. specific for fibrin breakdown, methods employed in the detection of FDP have included agglutination inhibition assays, using either tanned red blood cells (1) or latex particles Key words:
Bispecific
antibody,
whole
blood 273
immunoassay,
D dimer
274
THE SimpliRED D DIMER TEST
Vol. 58, No. 3
Staphylococcal clumping was an early system coated with fibrinogen (2). utilizing the characteristic clumping of some strains of staphylococci in the Improved agglutination assays based upon presence of fibrinogen (3). antibody coated latex particles have produced a rapid diagnostic system (4all these methods require the processing of the blood sample to ~~odu~~w~~~~~a or serum. We have developed a new system for the detection of circulating antigen or The test indicates a positive result by visible antibody in whole blood (7). The utilization of agglutination of the patient’s own red blood cells. endogenous red blood cells removes the need for extensive sample preparation, producing a unique test which is rapid and simple. The system uses a monoclonal antibody produced against red blood cells, which is conjugated to either an antigen or antibody. This reagent links the red blood cells and the circulating antibody or antigen, resulting in agglutination of the cells. The assay of XDP uses a monoclonal antibody to the crosslinked derivative D dimer, previously employed in both a quantitative enzyme immunoassay (8) and a semi-quantitative latex agglutination test (6). Although the latex test is accurate and rapid, it still requires sample preparation as the The SimpliRED D dimer initial step, confining the assay to the laboratory. test enables patient assessment in situations where laboratory facilities are not available.
METHODS Preparation of human XDP: Crosslinked fibrin clots were prepared and lysed as described by Masci et al (8). The lysate from 500 mL of plasma was concentrated by pressure dialysis to a volume of 20 mL (762 mg) and solid urea was added to give a final concentration of’3 M. XDP was purified by Sephacryl S300 gel chromatography at a flow rate of 1.0 mL/min on a 2 x 100 cm column in a buffer containing 3 M urea and 0.1 M sodium bicarbonate pH 8.4. Fractions containing XDP (59 mL; 480 mg) were pooled and dialyzed first against 100 volumes of phosphate buffered saline pH 7.4, and finally 50 volumes of a phosphate buffered saline containing 50% glycerol and 0.01% sodium azide. This preparation contains 189K Da dimers and higher molecular weight aggregates (ie Two D dimers with MWapprox. 400K Da by HPLC). The XDP preparation was stored in aliquots at -2OOC. Production of monoclonal antibodies: Two monoclonal antibodies were utilized in the development of this red blood cell agglutination system. The antibody DD-3B6/22, specific for human D dimer, was produced from the immunization of BALB/c mice with XDP, as described previously (9). The second antibody, RATlC3/86, was the result of immunization of BALB/c mice with washed human red blood cells as described by Rylatt et al (7). Two screening procedures selected the antibody which recognized the membrane antigen, glycophorin A, without causing agglutination of the cells. Additional assays ensured that the antibody would react with erythrocytes from all patients, regardless of blood group. Production of conjugation of Brennan et al acetate buffer to digest each
bispecific reaqent: The test reagent was produced by Fab’ fragments using a modification of the method described by (10). The antibody solutions were equilibrated into an 0.1 M pH 3.5, containing 0.07 M NaCl. A 1% pepsin solution was used intact monoclonal antibody to produce the F(ab’)?: fragments.
Vol. 58, No. 3
THE SimpliRED D DIMER TEST
275
The pepsin and antibody were combined in a 1:lOO ratio, digested at 37OC, then stopped by adjusting the pH to 8.0 with 1.5 M Tris buffer. Digestion times were 15 min for DD-3B6/22 and 45 min for RAT-lC3/86. The digests were purified by gel filtration on a Sephacryl S300 column equilibrated in 100 mM phosphate buffer, pH 6.8. The resulting F(ab’)n pool was concentrated by ultrafiltration to give a concentration of approximately 3 mg/mL. The F(ab’)z fragments were subsequently split into Fab’ fragments by reduction with 1 mM mercaptoethylamine in 100 mM phosphate buffer pH 6.8 containing 10 mM sodium arsenite and 1 mM EDTA. The time of exposure to the reducing agent varied for each antibody. DD-3B6/22 F(ab’)z was allowed to reduce for 2 hr at room temperature (RT), while RAT-lC3/86 reduced for 3 hr at 37OC. The hinge region sulfydryl groups on the DD-3B6/22 Fab’ species were blocked by the addition of 5 mM DTNB and mixed at RT for 3 hr. The RATlC3/86 Fab’ species were prepared only as required and left in the unblocked The preparations were again purified by gel filtration on a reduced form. The two Fab’ species were mixed together at RT for 2 Sephacryl S300 column. hr, which resulted in recombination of the fragments to produce the final bispecific antibody at an average yield of 50%. The reaction was monitored by HPLC analysis on a Waters 300SW column to ensure optimum production of product. Once achieved, the bispecific antibody was purified by gel filtration on a Sephacryl S300 column. The resulting pool was diluted in 100 mM phosphate buffer, pH 6.8 to give the desired sensitivity. The final formulation also included 5 g/L BSA, and an unrelated monoclonal antibody of similar isotype, to block any non-specific reaction. Agglutination test vrocedure: The test procedure required 10 uL of whole blood (taken in the form of a finaer nrick or henarinized blood sample) which The slide was was combined on a slide with 25 pL of-bispecific-reagent. rocked for l-2 minutes and a positive or negative agglutination reaction recorded. The semi-quantitative Latex aqqlutination procedure: test Dimertest latex was used as a comparative assay. 10 pL sample of plasma was mixed with 25 uL of latex agglutination rocked for 3 minutes (6) . The subsequent
latex agglutination On a glass slide, solution and gently was then recorded.
a
The quantitative two-site enzyme immunoassay Enzyme immunoassay procedure: Dimertest EIA was used to determine the D dimer levels of all samples The assay as outlined by Rylatt et al (9), uses DD-3B6/22 as the examined. capture antibody, and a panspecific antibody in the final labeling step. Sensitivity: Undiluted SimpliRED D dimer test reagent was examined against to which D dimer antigen had been added in group 0 negative whole blood, The lowest D dimer concentration which still produced a serial dilutions. positive agglutination pattern was recorded as the sensitivity limit. The SimpliRED D dimer test reagent was diluted to give a Stabilitv trials: A1iquot.s of the test sensitivity limit of approximately 0.25 mg/L D dimer. over a 28 day reagent were placed at 37OC, RI’, 4OC and -20°C and examined The reagent was examined against group 0 negative whole blood period. samples to which D dimer antigen had been added to give final concentrations A control of unspiked blood of 2.50, 1.25, 0.63,0.31, 0.16 and 0.08 mg/L. The sample with the lowest D dimer concentration still was always included. resulting in a positive agglutination pattern was recorded as the sensitivity of the reagent. Clinical
trials:
Two trials
were
conducted
during
the
examination
of
this
276
THE SimpliRED
D DIMER TEST
Vol. 58, No. 3
bispecific reagent. A preliminary evaluation involved the examination of Heparinized randomly selected hospital patients and normal individuals. samples were provided by the staff at Princess Alexandra Hospital, Brisbane, latex agglutination and red blood and tested in-house by enzsyme immunoassay, cell agglutination. A series of normals were also tested by both heparinized The bispecific reagent was adjusted to a and finger prick blood samples. sensitivity of approximately 0.20 mg/L, above which agglutination would occur. An independent clinical trial was undertaken by one of us (MJE). A selection of normal and thrombotic disease patients were examined by enzyme immunoassay, latex agglutination and red blood cell agglutination. For this trial, the sensitivity of the bispecific reagent was adjusted to 0.25 mg/L The anticoagulant such that blood from normal subjects did not agglutinate. used was sodium citrate. RESULTS Sensitivity: The sensitivity of this system was seen to be comparable to that of the enzyme immunoassay, with the lower limit of detection being 0.06 was consistent between batches. No prozone mg/L XDP. This sensitivity effect was observed in any experiments. The highest antigen concentration examined was 1.2 x 103 mg/L, with agglutination proportional to concentration. Figure 1 shows the agglutination patterns obtained with dilutions of antigen, compared with a negative reaction. The agglutination is clearly discernable. Our experience has been that the agglutination of the red blood cells is easier to read than that of latex particles.
Fiq. 1 Agglutination of red blood cells by the bispecific reagent in the From left to right the D dimer presence of crosslinked fibrin derivatives. concentrations were 1.75, 0.44, 0.11 and 0 mg/L (as determined by DIMERTEST EIA) .
Vol. 58, No. 3
277
THE SimpliRED D DIMER TEST
Stability: An accelerated stability trial of the bispecific antibody was set The reagent samples stored between -2O*C and UP at -20°C, 4OC, RT, and 37’C. RT remained stable throughout the 28 days, while those stored at 37OC started In an accelerated stability test a week at to lose activity after 10 days. 37OC is considered to represent stability for one year at 4OC. Clinical trials: Preliminary trials were carried out on whole blood samples from 67 hospital patients (at the Princess Alexandra Hospital in Brisbane) as well as a series of normals (Table 1). Of the 30 normals, 22 were examined by both finger prick and heparinized venous sampling. All samples showed The average concentration of D dimer for samples negative negative results. on the SimpliRED test was 0.10 mg/L, while those positive had an average concentration of 0.63 mg/L, showing a significant variation between positive and negative. The correlation was good between all detection systems. Twenty-six samples were recorded as positive on both latex and SimpliRED tests. While 12 were positive on the SimpliRED test and negative on latex, none of these were diagnosed as thrombotics, and therefore indicated that the Adjustment of the sensitivity of the bispecific reagent was too sensitive. bispecific reagent to the top of the normal range was made prior to the subsequent independent clinical trial, to match that of Dimertest Latex.
TABLE 1 Prelim #inary
EIA D Dimer
Cl i nical
Cone
(mg/L)
Latex SimpliRED
-ve
Evaluation*
of Bispecif
ic
-ve
Reagent
Latex
SimpliRED
+ve
+ve
Simpl iRED -ve
Simp 1iRED +ve
co.20
51
7Y
0
0
0.20-0.50
6
8
0
8
jo.50
1
2
* Total
number
of samples
18
was 97
The trial conducted at Royal Brisbane Hospital examined 101 normal, 18 intravascular coagulation (DIG), and Pulmonary embolism (PE), 32 Disseminated The results were 21 Deep venous thrombosis (DVT) whole blood samples. comparable with the latex assay and showed no false diagnosis by the The average D dimer concentration for normals was SimpliRED D dimer test. 0.09 mg/L, while PE samples showed 0.80 mg/L, and both the DIG’s and DVT’s averaged 0.96 mg/L (Figure 2).
278
THE SimpliRED D DIMER TEST
Vol. 58, No. 3
??
1 ??
?? ??
f
Fiq. 2 The examination of the SimpliRED D dimer test for the diagnosis of thrombotic disease states, (N: Normals, PE: Pulmonary embolism, DIC: DVT: Deep venous thrombosis). A Disseminated intravascular coagulation, is positive agglutination of red cells is indicated by ??, while a negative shown by ??.
DISCUSSION The time required for immunoassays has decreased with advances in technology, however the rate limiting step in most systems has been sample preparation. By using the patient’s whole blood immediately upon sampling, time and sample handling are significantly reduced. The production of a bispecific antibody enabled two separate antigenic sites (e g red blood cell membrane and D dimerl to be recognized by the one reagent component. A multivalent antigen acts as a bridge between two bispecific antibodies, and agglutination of the red cells results. In the case of fibrin degradation analysis, the crosslinked subunits enable formation of such a matrix. Where the antigenic site is not multivalent, the system can be modified to a two-site assay. Two different bispecific antibodies can be prepared from monoclonal antibodies against separate antigenic sites on the antigen. The reagent would then contain a mixture of these bispecific antibodies. Alternatively, an inhibition assay, as described by Rylatt et al (71, also evaluates antigens with unique epitopes. Preliminary correlation observed.
evaluation studies of the XDP bispecific reagent showed good between the EIA D dimer values and the agglutination patterns As a finger prick was all that was required to sample the
Vol. 58, No. 3
THE SimpliRED D DIMER TEST
279
Patient’s Status, the system was considered ideal for use outside the laboratory. A high level of sensitivity was seen, with the undiluted reagent detecting a concentration of 0.06 mg/L D dimer, comparable with enzyme immunoassays. For use as a rapid diagnostic tool the reagent can be set to a sensitivity where normal individuals show no agglutination. For our preliminary evaluation, the bispecific reagent was set at 0.20 mg/L as an approximate value for normal levels. This proved to be more sensitive than required in the clinical situation. Twelve patients showed positive agglutination on the SimpliRED test while negative with latex. None of these patients had thrombotic disorders. At the other end of the scale, the SimpliRED D dimer test missed two samples showing elevated D dimer levels of 0.52 and 0.88 mg/L. Again, in neither case was thrombosis diagnosed. The major difference between these systems and the Dimertest EIA is the form of the antigen detected. In the clinical situation, high molecular weight derivatives are the predominant fibrin breakdown product (ll-14), and it is these species the latex and SimpliRED assays recognise. Earlier work indicated that the antibody DD-3B6/22 is specific to a single site on the D dimer component of XDP preparations, close to the crosslinking site (15). Consequently, both the Dimertest latex and the SimpliRED D dimer test require the higher molecular weight forms to provide the two sites necessary to bridge the red cells. The Dimertest EIA uses two separate antibodies each This enables detection of D seeing different sites on the D dimer molecule. dimer in 189K Da and higher molecular weight forms. With normals who have will occur despite a high levels of the single dimer form, no agglutination In patients suffering thrombosis, the concentration of high EIA reading. crosslinked derivatives is significantly higher than the dimer form, and agglutination will occur. The second trial was directed specifically at the diagnosis of thrombotic disease. The XDP levels of patients with thrombotic disease were hence a rapid yes/no answer is possible. significantly higher than normals, The bispecific antibody was able to diagnose correctly all hospital Patients examined, and in the hospital laboratory the test reagent was considered to the degree of agglutination is be rapid, simple and accurate. As seen, proportional to the amount of XDP present. The time in which the agglutination occurs also is dependent upon the concentration of antigen. the more rapid the agglutination. If a more The higher the concentration, the system could be used in accurate assessment of XDP levels were required, and adapted to equipment such as conjunction with a standard antigen series, a laser or kinetic reader, for calibration of the Pattern. A semiquantitative assessment could be made by simply diluting the test reagent. Using a standard antigen concentration as a guide, an estimate of the antigen level could be made. The whole blood assay presented in this paper rewires no centrifuge or and the result is read in two specialized equipment, is simple to perform, The agglutination pattern is easy to assess and therefore suitable minutes. For these reasons it is considered for personnel with limited training. ideal for doctor’s surgery or emergency room analysis, where rapid diagnosis or for field studies and Third World applications, where no is advantageous, equipment is available. ACKNOWLEDGEMENTS We would Hospital, enabling
like to thank Tony Badrick and staff at the Princess Alexandra Brisbane, for their assistance in the suPPlY of Patient samples, our preliminary testing of the bispecific antibody.
280
THE SimpliRED
Vol. 58, No. 3
D DIMER TEST
REFERENCES 1. MERSKEY, C., LALEZARI, P. and JOHNSON, A.J. method for measuring fibrinolytic split products Exv. Biol. Med., 131, 871-875, 1969. 2. ALLINGTON, M.J. Detection latex clumping method. Stand.
of fibrin(ogen1 J. Haem., l&
A rapid, simple in human serum.
sensitive Proc. SOC.
degradation products 115-119, 1971,
by a
3. HAWIGER, J., NIEWIAROWSKI, S., GUREWICH, v. and THOMAS, D.P. Measurement of fibrinogen and fibrin degradation products in serum by staphylococcal clumping test. J. Lab. Clin. Med., 75, 93-108, 1970. 4. ELLMAN, L., CARVALHO, A. and COLMAN, R.W. The Thrombo-Wellcotest N. Eng. screening test for disseminated intravascular coagulation. 288, 633-634, 1973. 5. ELMS, M., BUNCE, I., P.P. and WHITAKER, A.N. products in Plasma using Clin. Path., 25, 360-364,
J.
as a Med.,
BUNDESEN, P.G., RYLATT, D.B., WEBBER, A.J., MASCI, Rapid detection of cross-linked fibrin degradation monoclonal antibody coated latex particles. Am. J. 1986.
6. HILLYARD, C.J., BLAKE, A.S., WILSON, K., RYLATT, D.B., MILES, S., BUNCH, R ., ELMS, M.J., BARNES, A. and BUNDESEN, P.G. A Latex Agglutination Assay for D Dimer: Evaluation and Application to the Diagnosis of Thrombotic Clin. Chem., 33 (101, 1837-1840, Disease. 1987. 7. RYLATT, D.B., KEMP, B.E., BUNDESEN, P.G., JOHN, M.A., O'REILLY,E.J., COTTIS, L.E., MILES, S.J., KHAN, J.M., DINH, D.P., STAPLETON, D. and HILLYARD, C.J. A rapid whole blood immunoassay system. Aust. Med. J., 152, 75-77, 1990. WHITAKER, A.N., ELMS, M.J., MASCI, P.P., BUNDESEN, P.G., RYLATT, D.B., 8. WEBBER, A.J. and BUNCE, I.H. Measurement of cross linked derivatives in using monoclonal antibodies. J. Clin. Pathol., 37, plasma : an immunoassay 882-887, 1984. RYLATT, D.B., BLAKE, A.S., COTTIS, L.E., MASSINGHAM,D.A., FLETCHER, 9. WHITAKER, A.N., ELMS, M., BUNCE, I., WEBBER, A.J., WYATT, W.A., MASCI, P.P., An immunoassay for human D dimer using monoclonal D. and BUNDESEN, P.G. Thromb. Res., 31, 767-778, 1983. antibodies. of bispecific BRENNAN, M., DAVISON, P.F. and PAULUS, H. Preparation 10. antibodies by chemical recombination of monoclonal immunoglobulin G1 fragments. Science, 229, 81-83, 1985. 11. GAFFNEY, P.J. The occurrence and clinical relevance in blood. Ann. New York Acad. SC., 408, 407-425, 1983.
of fibrin
fragments
12. WHITAKER, A.N., ELMS, M.J., MASCI, P.P., BUNDESEN, P.G., RYLATT, D.B., WEBBER, A.J. and BUNCE, I.H. Measurement of cross linked fibrin derivatives an immunoassay using monoclonal antibodies. J. Clin. Pathol., 37, in plasma: 882-887, 1984. CREIGHTON, L.J., 13. GAFFNEY, P.J., CALLUS, M. and THORPE, R. Monoclonal antibodies to crosslinked fibrin degradation products (XL-FDP). II.
Vol. 58, No. 3
Evaluation 1988.
THE SimpliRED D DIMER TEST
in
a variety
of clinical
conditions.
Br.
J.
14. CARR,. J.M., McKINNES, M. and McDONAGH, J. Diagnosis intravascular coagulation. Role of D dimer. Am. J. Clin. 1989. 15. GREENBERG, C.S., dimer levels with the Am. J. Clin. Pathol.,
281
Haem.,
68,
91-96,
of disseminated Path., 91, 280-287,
DEVINE, D.V. and McCRAE. Measurement use of a monoclonal antibody coupled 81, 94-100, 1987.
of plasma fibrin to latex beads.
II
