Antibodies Against the Fibrin β-Chain Amino-Terminus Detect Active Canine Venous Thrombi
Background This study was performed to determine whether antibodies against the amino-terminus of the β-chain of fibrin (anti-β) could noninvasively distinguish actively enlarging thrombi from thrombi stabilized with anticoagulants.
Methods and Results Dogs with unilateral femoral vein thrombi were allocated into three groups: (1) no anticoagulation, (2) intravenous heparin maintained in the “therapeutic” range (0.2 to 0.5 U/mL plasma), and (3) “excess” heparin, maintained at >1.0 U/mL plasma. Thrombolysis was suppressed with tranexamic acid. 111In-labeled anti-β was infused, and gamma scans of the legs were performed at regular intervals for 24 hours. Scans were interpreted in a blinded fashion. In addition, for each scan, the number of gamma counts from the femoral area on the thrombosed side was compared with the contralateral side. Clot/blood isotope density was determined postmortem. Leg thrombi in the no-anticoagulation group were 100% detectable, mean (±SD) relative count in the thrombosed femoral area was 186% (±30%) of the contralateral side, and clot/blood ratio was 14.7 (±2.0). Thrombi in the therapeutic heparin group were only 75% detectable, relative counts in the thrombosed femoral areas decreased to 125% (±20%), and clot/blood ratio declined to 11.3 (±3.5). In the “excess heparin” group, leg thrombi were only 50% detectable, the thrombosed femoral area had relative counts of 118%±17%, and the clot/blood ratio fell to 7.8±1.9.
Conclusions Radiolabeled anti-β noninvasively distinguishes propagating thrombi from those stabilized by anticoagulants. They may be useful for detecting thrombosis clinically as well as for monitoring the efficacy of anticoagulation.
The diagnosis of deep venous thrombosis and pulmonary embolism on clinical grounds is problematic.1 Several noninvasive tests have been developed to facilitate the diagnosis. Each test, however, has significant limitations in sensitivity and specificity. Studies to evaluate the accuracy of ventilation/perfusion scans, impedance plethysmography, and duplex ultrasound have demonstrated their limitations and interobserver variability. These difficulties reflect the fact that the tests depend on blood flow changes, which can be mimicked by other diseases.
Radiolabeled monoclonal antibodies specific for thrombus-associated antigens may enable thrombi to be imaged noninvasively with nuclear medicine techniques. One such antibody, raised against the amino-terminus on the β-chain of fibrin, binds with high avidity to fibrin monomers but not to fibrinogen.2 3 When radiolabeled, it allows gamma camera imaging of thrombi induced by thrombogenic metal coils placed in the veins of animals.4 5 6 7 8 9 However, when the animals were anticoagulated with heparin, the images provided by the radiolabeled anti-β antibodies were less able to detect the thrombi.8 Similar results were observed in patients with deep vein thrombosis who were anticoagulated.10 11 12 13
The diminished reliability of radiolabeled anti-β antibody scans during anticoagulation has been attributed to the unopposed effect of fibrinolysis, causing the loss of amino-terminus antigenic sites8 and cleavage of antigen/antibody complexes.10 We have previously demonstrated, however, that the amino-termini of the β-chains are rapidly enveloped during normal fibrin polymer organization independent of fibrinolysis.14 Anticoagulation, by inhibiting the formation of new fibrin molecules, merely prevents the ongoing incorporation of the transiently exposed targets of the anti-β antibodies into preexisting thrombi. It follows, then, that anti-β antibodies are highly specific for actively propagating venous thrombi.
Binding of antibody to the β-chain amino-terminus of fibrin occurs during the intermediate state between the addition of new monomers to the polymer and the disappearance of binding sites as lateral associations occur among polymers.15 16 17 18 19 We hypothesize that anticoagulation will interfere with thrombus imaging by preventing monomer formation, while still allowing lateral associations to occur among preformed polymers.
The present study was undertaken to determine the effect of anticoagulation on our ability to image in vivo thrombosis in dogs with a radiolabeled anti-β antibody. We used tranexamic acid to completely inhibit plasmin-mediated fibrinolysis20 and strictly controlled the degree of anticoagulation with intravenous heparin.21
Previous in vivo studies of anti-fibrin antibodies have used thrombogenic coils to induce venous thrombi. However, the coils themselves may traumatize the walls of the extremity veins,5 and the anti-fibrin antibodies may bind to the injured tissue itself.22 In the present study, thrombi were formed in the leg veins without endothelial damage by infusion of thrombin into vein segments that had been occluded by intravascular balloons.
Anti-β and monoclonal nonspecific antibodies, prepared and labeled with 111In as previously described,19 were provided by Hybritech Inc. Human fibrinogen (>95% clottable) was purchased from Calbiochem Corp. Topical thrombin (bovine) was obtained from Armour Pharmaceuticals. Tranexamic acid was from Sigma Chemical Co. Unfractionated porcine heparin was obtained from Elkins-Sinn, Inc. Activated partial thromboplastin times were measured with Dade Actin FS Activated PTT Reagent, purchased from Baxter Diagnostics.
Healthy mongrel dogs (weight, 20 to 35 kg) were anesthetized with pentobarbital 30 mg/kg IV (Abbott Laboratories), intubated, and mechanically ventilated to maintain arterial blood gases within normal limits. Baseline lung perfusion scans were obtained. The animals received tranexamic acid 110 mg/kg IV at this time and every 6 hours thereafter to completely inhibit fibrinolysis.20
Baseline blood samples were collected in 0.1 volume of 3.2% (wt/vol) sodium citrate, and the plasma was obtained by centrifugation at 1600g for 10 minutes. Plasma (100 μL) was incubated with 100 μL aPTT reagent for 10 minutes. The plasma activated partial thromboplastin time (aPTT) was then determined in duplicate with a Lancer Coagulyzer II (Sherwood Medical) by addition of 100 μL 0.02 mol/L CaCl2.
The aPTT assay was standardized to correspond to plasma heparin levels, determined both by ex vivo heparin addition23 and by protamine titration.24 For the first method, baseline plasma samples from each animal were incubated for 1 hour with small volumes of heparin (1000 U/mL) to produce incremental heparin concentrations between 0.05 and 0.60 U/mL plasma, and the corresponding aPTTs were determined. A linear standard curve was obtained by plotting the log(sample aPTT/baseline aPTT) versus the plasma heparin concentration.25 Plasma heparin values during systemic in vivo anticoagulation were then extrapolated from the standard curve by use of the ratio of the treatment aPTT to the pretreatment aPTT. The validity of this method was confirmed in one animal, which had received a one-time intravenous bolus of heparin (300 U/kg), by protamine titration of blood samples collected every 30 minutes for 4 hours, as follows: plasma aPTTs were determined and compared with the baseline aPTT drawn before heparinization. Next, 100-μL aliquots of the samples were incubated with equal volumes of protamine sulfate (1.18 to 23.52 mg/mL in Veronal buffer, pH 7.35).24 Thrombin (100 μL, 3 U/mL in Veronal buffer) was then added. The clotting time was recorded as the number of seconds after thrombin administration necessary for a gelatinous connection to become visible between the fluid and a stirring rod. For each plasma sample, the heparin level was calculated from the lowest protamine concentration necessary to return the clotting time to baseline, assuming that 1.0 mg protamine inactivates 85 U heparin.24
There were closely correlating linear relationships between the log(sample aPTT/baseline aPTT) and the plasma heparin levels determined by ex vivo heparinization (r=.99; slope, 0.91; 95% CI, 0.85 to 0.97) and by protamine titration (r=.99; slope, 0.95; 95% CI, 0.80 to 1.10).
Double balloon catheters (Fig 1⇓) were advanced via hind-leg saphenous veins to the femoral veins on each side. Contrast venograms were performed to confirm the absence of thrombi in each femoral vein. The balloons were then inflated, creating a sealed chamber within the veins. Through a port between the double balloons, 200 U thrombin was injected into each venous lumen to induce thrombosis. After 1 hour, the balloons were deflated, and the induced thrombi were aged in situ for an additional 3 hours. Venograms were repeated to confirm the presence of thrombus.
For each animal, the balloon catheter was removed from one leg, and that thrombus was embolized by passive leg motion. The balloon catheter in the femoral vein on the other side was left in position to prevent embolization of the contralateral thrombus. Venograms were again performed on both legs.
Allocation of Groups
Three groups of dogs (three animals per group) were assigned to receive radiolabeled anti-β antibodies. Animals in the “no-anticoagulation” group received no heparin. Each animal in the “therapeutic heparin” group initially received an intravenous heparin bolus of 90 U/kg. Thereafter, intravenous heparin was continuously infused at initial rates of 30 U · kg−1 · h−1, and the doses were adjusted hourly to maintain aPTT values corresponding to between 0.2 and 0.5 U heparin/mL plasma. Each animal in the “excess heparin” group received an intravenous heparin bolus of 300 U/kg, followed by a continuous infusion of heparin at 90 U · kg−1 · h−1. The heparin dose was adjusted to keep the plasma heparin level >1.0 U/mL (corresponding to aPTT values greater than 10 times control).25 This degree of anticoagulation has been associated with complete suppression of thrombus propagation in previous canine experiments.26 One additional animal (“interrupted heparin”) was treated similarly to the excess heparin group for the first 8 hours of the study. The heparin infusion was then stopped for 8 hours, after which the aPTT had returned to baseline. The animal was then given another heparin bolus (300 U/kg), and the infusion was continued at its previous rate for the next 8 hours.
For comparative purposes, a negative control group was allocated, which consisted of two nonheparinized animals that received radiolabeled nonspecific antibodies.
Radiolabeled Antibody Scans
One hour after the appropriate heparin therapy was begun, 600 μCi of 111In-labeled antibodies (≈500 μg) was injected via a foreleg vein. A Dyna Camera 4 gamma camera (Picker) recorded 5-minute images of both hind legs at baseline and at regular intervals thereafter for the next 24 hours (Fig 2⇓).
To obtain an objective measure of focal antibody uptake, standard region-of-interest boxes, corresponding to the femoral areas on either leg, were isolated on the gamma camera images with an A2 Clinical Imaging System (Medical Data Systems). For each 5-minute scan, the number of gamma counts from the femoral vein area of the thrombosed leg was compared with those from the embolized leg.
Blood samples (2 mL) were collected at various times after antibody infusion and counted for 1 minute in a Gamma 4000 counting spectrometer (Beckman Instruments Inc) on an emission window calibrated for 111In. Blood clearance times were calculated from the reduction in isotope densities (gamma counts · min−1 · g blood−1) over time. Cumulative urine samples were collected, and total urinary output was recorded. Urine samples (1 mL) were counted as above to determine urinary and nonurinary clearance of the antibodies.
Postmortem Analysis of Thromboemboli
At the conclusion of the study (24 hours after antibody infusion), the animals were euthanized with pentobarbital. An autopsy was performed to define the location of venous thrombi and pulmonary emboli. The venous thrombi and pulmonary emboli were harvested, weighed, and gamma counted. Isotope densities (gamma counts · min−1 · g tissue−1) of the thrombi and emboli were compared with the densities in the final blood samples and expressed as clot/blood gamma emission ratios. The mean clot/blood ratios for thrombi and emboli were calculated for each experimental group. Tissue/blood ratios were also determined from samples of the following tissues: venous walls adjacent to the thrombi, normal venous walls, liver, spleen, heart, muscle, lung parenchyma, and kidney.
This protocol was approved by the University of California at San Diego Animal Subjects Committee, in compliance with federal regulations regarding the care and use of laboratory animals.
Interpretation of Scans
At the conclusion of the study, photographs of each of the leg scans, including the positive and negative controls, were randomly mixed, then interpreted separately by four readers who were unaware of the location of the thrombi, the group to which the animals had been allocated, or the timing of the scans. For each scan, the readers were asked to identify whether one femoral area was clearly more highlighted than the contralateral area. If they could not with certainty identify an area of focal isotope uptake on a scan, they were instructed to indicate “none.” Within each time point, the detectability of leg thrombi was calculated as No. of correct detections/No. of veins containing thrombi×No. of readers, while the rate of spurious detections was calculated as No. of spurious detections/(No. of veins not containing thrombi)×(No. of readers).
Focal uptake of the labeled antibodies was quantified by use of identical region-of-interest boxes over the thrombosed and contralateral femoral veins. For each scan, the increase in antibody uptake was calculated as counts from thrombosed femoral vein/counts from contralateral femoral vein.
Results are expressed as mean±SD, unless otherwise specified. Thrombus masses and clot/blood isotope density ratios among groups were compared by one-way ANOVA and the Tukey-Kramer multiple comparisons test. To compare thrombus detectabilities among groups during the last 12 hours of the study, a Fisher’s exact test was performed using the number of true-positives and false-negatives as categorical variables. Statistical tests were performed by use of the Instat software package, version 2.04a (GraphPad Software).
In the therapeutic heparin group, after equilibrium from the initial heparin bolus, plasma heparin levels remained between 0.2 and 0.5 U/mL (Fig 3⇓). In the excess heparin groups, plasma heparin levels were consistently >1.0 U/mL (results not shown).
In the animal whose heparin was interrupted, the heparin level was >1.0 U/mL until hour 8 and returned to 0 by hour 16 (Fig 4⇓). After the dog received an additional heparin bolus and the continuous infusion was restarted, the heparin level returned to >1.0 U/mL for the remainder of the study.
The scan readers had been instructed to have a high threshold for interpreting a scan as positive. Subsequently, the mean rate of spurious detections in any of the anti-β antibody scans was only 2.9% (95% CI, 5.0% to 0.8%) and did not differ among groups.
Leg scans of a subject in the no-anticoagulation group, taken at 4, 12, and 24 hours after antibody infusion, are shown in Fig 5a⇓. Scans were suggestive at 4 hours (thrombus detectability, 75%), and thrombi became 100% detectable by 12 hours (Fig 6⇓). Correspondingly, gamma emissions in the thrombosed femoral areas rose to 1.86±0.30 times those in the contralateral areas by 24 hours (Fig 7⇓).
Leg scans from the therapeutic heparin group (Fig 5b⇑) demonstrated a marked reduction in antibody uptake by the thrombus. Leg thrombi scanned at 12, 16, 20, and 24 hours were significantly less detectable than in the no-anticoagulation group (P<.0001). Twenty-four hours after antibody infusion, the detectability of the leg thrombi in the therapeutic heparin group was only 75% (Fig 6⇑). Although the thrombosed femoral area accumulated more gamma emissions than the contralateral area (ratio, 1.25±0.20), the increase was significantly less than that observed with the no-anticoagulation group (P<.05) (Fig 7⇑).
The detectability of the thrombi for the excess heparin group (Fig 5c⇑), scanned at 12, 16, 20, and 24 hours, was significantly less than for either the no-anticoagulation group (P<.0001) or the therapeutic heparin group (P=.02). The thrombi scanned after 24 hours were only 50% detectable (Fig 6⇑). The gamma emissions in the thrombosed femoral area were somewhat increased over the contralateral emissions (ratio, 1.18±0.17) (Fig 7⇑). The ratio was significantly less than for the no-anticoagulation group (P<.05). The ratio was also less than the one for the therapeutic heparin group, but the difference did not achieve statistical significance (P>.05).
Leg scans at 8, 12, and 24 hours from the animal whose heparin infusion was interrupted are shown in Fig 5d⇑. The thrombus became 100% detectable after the heparin infusion was stopped and remained at that level throughout the remainder of the study (Fig 6⇑). The gamma emissions in the thrombosed femoral area continued to rise for 12 more hours, despite the resumption of the heparin infusion (Fig 7⇑). At 24 hours, the thrombosed femoral area had 1.98 times the counts in the contralateral area.
Leg thrombi in the negative control group had a detectability of 0% throughout the study. For this group, the counts in the thrombosed femoral area were not significantly increased over those in the contralateral area.
Thrombus Masses and Isotope Densities
The mean masses and clot/blood gamma emission ratios for leg thrombi from each group are shown in the Table⇓. The mean thrombus mass in the no-anticoagulation group was 2.1±0.8 g. The mean thrombus masses in both anticoagulated groups were roughly half this size (therapeutic heparin group, 1.0±0.3 g; excess heparin group, 1.0±0.7 g). In the animal whose heparin was interrupted, the mass of the thrombus was 1.5 g, intermediate between the no-anticoagulation and therapeutic/excess heparin groups. The thrombi in the nonheparinized negative control group had a mean mass of 2.1±0.5 g.
The clot/blood ratio for the no-anticoagulation group was 14.7±2.0, which was consistent with the high detectability of the leg thrombi. The ratio for the therapeutic heparin group was 11.3±3.5. The excess heparin group had a mean clot/blood ratio of only 7.8±1.9. However, the animal whose heparin had been interrupted had a clot/blood ratio of 19.9. Although there was a trend toward lower clot/blood ratios with higher degrees of anticoagulation, none of the differences among groups achieved statistical significance. The clot/blood ratio in the negative control group was 1.9±0.8.
In animals receiving anti-β antibodies, the isotope density of endothelium adjacent to the thrombi was higher than that for blood, indicating some mural fibrin deposition. However, for all groups, the endothelium/blood ratios were <25% of the clot/blood ratios described above. The tissue/blood ratios for samples of spleen, heart, muscle, lung parenchyma, and kidney were all <1.0. Isotope densities in the liver were greater than those in blood, which is consistent with the relatively high nonrenal clearance of the antibodies. The mean liver/blood isotope density ratio, which did not differ among groups, was 1.9±0.9.
Urinary and total blood clearance of the anti-β antibodies did not statistically differ among the groups. The mean blood clearance half-life was 21±3 hours. Of the total amount cleared, only 22.4±7.7% could be accounted for by excretion into the urine. Gamma scans and gamma emission counting of tissue samples confirmed that a significant portion of the nonrenally cleared isotope was taken up by the liver.
The radiolabeled anti-β antibodies used in the present study allowed timely noninvasive identification of leg thrombi under conditions of normal coagulation. In nonheparinized animals, most of the gamma scans were suggestive by 4 hours, and all thrombi were detectable by 12 hours. The accurate interpretation of the leg scans corresponded to a higher number of gamma counts in thrombosed femoral areas, as well as high clot/blood ratios in the postmortem specimens.
Therapeutic anticoagulation coincident with suppression of thrombolysis decreased the detectability of the thrombi scanned at 24 hours to 75%, with an associated relative decrease in the counts in thrombosed femoral areas. This therapeutic heparin group also demonstrated both decreased thrombus size and lower clot/blood ratios, either (or both) of which could have been responsible for the decreased antibody accumulation. However, when heparin levels were in excess, the clot/blood ratios decreased further, whereas the clot masses were unchanged. Leg thrombi scanned at 24 hours in the excess heparin group were only 50% detectable for the presence of thrombi.
Heparin thus reduced thrombus labeling and reduced the detectability of leg thrombi. Although it has no direct effect on antibody binding,19 heparin inhibits the addition of new fibrin molecules to an existing thrombus. In the animal for which the heparin infusion was interrupted, a small amount of new fibrin was apparently deposited on the existing thrombus and resulted in a dramatic increase in the clot/blood ratio and a correspondingly rapid increase in its detectability.
The divergence between the two potential methods of interpreting the scans, by blinded readers and by comparison of femoral area gamma counts, is noteworthy. Readers were much more accurate in detecting thrombi in the therapeutic heparin group than in the excess heparin group. However, the difference between the groups was not reflected by a statistically significant decrease in the relative counts over the thrombosed femoral areas. The superior performance of the subjective image interpretations appears to reflect the ability of the readers to identify focal increases in isotope density within the femoral areas, which they correctly interpreted as venous thrombi.
Both methods of scan interpretation, however, were able to distinguish anticoagulated from nonanticoagulated thrombi. For the objective method, a ≥25% increase in relative counts in the thrombosed leg by 8 hours identified nonanticoagulated animals with a sensitivity of 93.3% and a specificity of 75.0%. Interestingly, the subjective method yielded identical sensitivity and specificity when positive scans were defined as those in which at least three of the four readers correctly located thrombi.
The inverse relationship between the degree of anticoagulation and the clot/blood ratios did not achieve statistical significance in these experiments. This may be due to the small number of animals in each group. In addition, the use of clot/blood ratios to compare antibody binding among groups introduces a possible bias in favor of higher ratios for smaller thrombi. Smaller thrombi, such as those observed in the anticoagulated groups, have larger surface areas relative to their masses. Therefore, the fibrin on the surface of these thrombi may have been exposed to more circulating antibodies than in the nonanticoagulated animals. This might result in higher clot/blood ratios for the anticoagulated thrombi. Also, the greater relative surface area in the smaller thrombi could have introduced a measurement error in favor of higher clot/blood ratios because of more water evaporation between the time of collection and weighing of very small specimens. These hypotheses are consistent with the observation that, within each animal, pulmonary emboli were smaller than deep venous thrombi and had higher clot/blood ratios (results not shown). Both of the potential biases would have reduced the observed difference in clot/blood ratios between normal and anticoagulated animals.
The specificity of the anti-β antibody for newly formed fibrin suggests several potential clinical uses. Scans after radiolabeled anti-β antibody injection would complement other techniques for the noninvasive diagnosis of deep venous thrombosis. In particular, this approach could help distinguish acute thrombi from other lesions, such as chronic organized thrombi. Chronic thrombi have similar appearances on current noninvasive tests but have quite different clinical implications.
In addition to the diagnosis of acute thrombi, the ability to identify ongoing fibrin deposition may help clarify the controversies such as the optimal duration of anticoagulation therapy for acute thrombosis.27 28 29 For example, after a 3-month course of anticoagulant therapy, a patient being treated for deep venous thrombosis could have medications held and a leg scan performed several hours later. A focal light-up would indicate that thrombosis is continuing and longer-term therapy is indicated. A prospective clinical study correlating the noninvasive detection of active thrombosis with radiolabeled anti-β antibodies to the development of recurrent deep venous thrombosis and pulmonary embolism seems warranted by the results of the in vivo experiments reported here.
Finally, the ability to determine whether preexisting thrombi have stopped propagating may be useful for the standardization of newer anticoagulant agents. Several anticoagulant alternatives to standard heparin are being developed for the treatment of thromboembolism. Unfortunately, in vitro tests of the anticoagulant potencies of the newer agents do not directly predict their in vivo antithrombotic effect.30 The antithrombotic effects of newer anticoagulants are currently determined by their ability to inhibit the accretion of iodinated exogenous fibrinogen onto preformed thrombi in animal subjects.30 Radiolabeled anti-β antibodies have several advantages over iodinated fibrinogen for measuring the antithrombotic potencies of anticoagulants. First, although exogenous fibrinogen may simulate the constituent of fibrinogen circulating in plasma, it does not simulate the fibrinogen contained within platelets, which is expressed on the platelet membrane after stimulation by thrombin.31 Anti-β antibodies will, however, bind to newly formed fibrin from any source (ie, platelets or plasma). Second, antibodies can be labeled with high-energy isotopes such as indium or technetium without losing immunoreactivity,8 which allows more reliable detection and avoids the complication of dehalogenation seen with iodinated proteins. Finally, because the degree of thrombus propagation is measured noninvasively, the same method may be used to evaluate the antithrombotic effect of anticoagulant drugs in humans, whereas radiolabeled fibrinogen is no longer available for clinical use. For instance, we are currently using the technique described in this paper to standardize the antithrombotic effect of different doses of low-molecular-weight heparin in dogs.
The detection of active thrombosis by anti-β antibodies also may make clinical trials of anticoagulant treatment more practical. Treatment failures are currently defined as the development of serious complications such as recurrent deep venous thrombosis or pulmonary embolism during long-term follow-up.32 Because these complications are difficult to diagnose and may take some time before becoming clinically apparent, demonstrating a treatment effect requires large numbers of patients to undergo repeated expensive testing. However, if the demonstration of active thrombosis during therapy as described in this study were shown to predict these complications, anticoagulant efficacy could be determined during therapy, before the development of serious complications.
The pulmonary emboli formed during this study were labeled at least as well as the leg thrombi in terms of their clot/blood ratios but could not be noninvasively imaged (results not shown). Although the lung parenchyma had isotope densities considerably lower than blood, the greater mass of adjacent blood pool in the vascular lung tissue prevented the signal from the relatively small emboli from being accurately detected by gamma scanning, even at postmortem study of the removed lungs. Methods to minimize the effect of background activity, such as single photon emission computerized tomography, may improve imaging of pulmonary emboli in vivo and are currently being explored in our laboratory.
This study was aided by a grant from The American Lung Association of California Research Program. The authors would like to thank the members of the UCSD Pulmonary Vascular group, Drs William Auger, Richard Channick, and Peter Fedullo, and Dr William Ashburn from the UCSD Division of Nuclear Medicine for reviewing the leg scans in a blinded fashion.
Reprint requests to Timothy A. Morris, MD, Assistant Professor of Medicine, UCSD Medical Center, 200 W Arbor Dr, San Diego, CA 92103-8372.
- Received January 30, 1997.
- Revision received June 9, 1997.
- Accepted June 19, 1997.
- Copyright © 1997 by American Heart Association
Hui KY, Haber E, Matsueda GR. Monoclonal antibodies to a synthetic fibrin-like peptide bind to human fibrin but not fibrinogen. Science. 1983;222:1129-1132.
Kudryk B, Rohoza A, Ahadi M, Chin J, Wiebe ME. Specificity of a monoclonal antibody for the NH2-terminal region of fibrin. Mol Immunol. 1992;21:89-94.
Rosebrough SF, Grossman ZD, McAfee JG, Kudryk BJ, Subramanian G, Ritter-Hrncirik CA, Witanowski LS, Yillapaugh-Fay G, Urritia E, Zapf-Longo C. Thrombus imaging with indium-111 and iodine-131-labeled fibrin-specific monoclonal antibody and its F(ab′)2 and Fab fragments. J Nucl Med. 1988;29:1212-1222.
Cerqueira MD, Stratton JR, Vracko R, Schaible TF, Ritchie JL. Noninvasive arterial thrombus imaging with 99-m technetium monoclonal antifibrin antibody. Circulation. 1992;85:298-304.
Rosebrough SF, McAfee JG, Grossman ZD, Kudryk BJ, Ritter-Hrncirik CA, Witanowski LS, Maley BL, Bertrand EA, Gagne GM. Thrombus imaging: a comparison of radiolabeled GC4 and T2G1s fibrin-specific monoclonal antibodies. J Nucl Med. 1990;31:1048-1054.
Knight LC, Maurer AH, Ammar IA, Shealy DJ, Mattis JA. Evaluation of indium-111-labeled anti-fibrin antibody for imaging vascular thrombi. J Nucl Med.. 1988;29:494-502.
Alavi A, Gupta N, Palevsky HI, Kelly MA, Jatlow AD, Byar AA, Berger HJ. Detection of thrombophlebitis with In-111-labeled anti-fibrin antibody: preliminary results. Cancer Res. 1990;50(suppl):958s-961s.
Lusiani L, Zanco P, Visona A, Breggion G, Pagnan A, Ferlin G. Immunoscintigraphic detection of venous thrombosis of the lower extremities by means of human antifibrin monoclonal antibodies labeled with In-111. Angiology. 1989;40:671-677.
Morris TA, Marsh JJ, Moser KM. Anti-fibrin antibodies bind monomeric, but not polymerized fibrin. Am Rev Respir Dis. 1993;147:A998. Abstract.
Laudano AP, Doolittle RF. Influence of calcium ion on the binding of fibrin amino terminal peptides to fibrinogen. Science. 1981;212:457-459.
Hantgan RR, Francis CW, Scheraga HA, Marder VJ. Fibrinogen structure and function. In: Colman RW, Hirsh J, Marder VJ, Salzman EW, eds. Hemostasis and Thrombosis: Basic Principles and Clinical Practice. Philadelphia, Pa: JB Lippincott Co; 1919:269-283.
Marsh JJ, Konopka R, Lang IM, Wang H, Pedersen C, Chiles P, Reilly CF, Moser KM. Suppression of thrombolysis in a canine model of pulmonary embolism. Circulation. 1994;90:3091-3097.
Moser KM, Spragg RG, Bender F, Konopka R, Hartman MT, Fedullo P. Study of factors that may condition scintigraphic detection of venous thrombi and pulmonary emboli with indium-111-labeled platelets. J Nucl Med. 1980;21:1051-1058.
Hirano T, Tomiyoshi K, Watanabe N, Tateno M, Oriuch N, Inoue T, Endo K. Technetium-99 m-labeled anti-fibrin monoclonal antibody accumulation in an inflammatory focus. J Nucl Med. 1992;33:1181-1182.
Simmons A. Technical Hematology. Philadelphia, Pa: JB Lippincott Co; 1980:291-322.
Fedullo PF, Moser KM, Moser KS, Konopka R, Hartman MT. Indium-111-labeled platelets: effect of heparin on uptake by venous thrombi and relationship to the activated partial thromboplastin time. Circulation. 1982;66:632-637.
Morgenstern A, Ruf A, Patscheke H. Ultrastructure of the interaction between human platelets and polymerizing fibrin within the first minutes of clot formation. Blood Coagul Fibrinolysis. 1991;1:543-546.
Hirsch J, Levine MN. Low molecular weight heparin. Blood. 1992;79:1-17.