- Email: [email protected]
Microwave-assisted synthesis of alkyl cellulose in aqueous medium
Thrombosis Research 126 (2010) e428–e433
Contents lists available at ScienceDirect
Thrombosis Research j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / t h r o m r e s
Regular Article
Comparison of the fibrinogen Clauss assay and the fibrinogen PT derived method in patients with dysfibrinogenemia W. Miesbach a,⁎, J. Schenk b, S. Alesci a, E. Lindhoff-Last c a b c
Haemophilia Center, Medical Clinic III/Institute of Transfusion Medicine, Johann Wolfgang Goethe-University, Frankfurt am Main, Germany University Hospital Saarland, Institute of Transfusion Medicine, Germany Goethe University Frankfurt, Medical Clinic III/Department of Angiology, Germany
a r t i c l e
i n f o
Article history: Received 13 April 2010 Received in revised form 30 August 2010 Accepted 2 September 2010 Available online 13 October 2010 Keywords: Fibrinogen Clauss PT-derived fibrinogen Diagnosis Dysfibrinogenemia
a b s t r a c t Fibrinogen assays are an important screening tool for blood coagulation disorders. Although different methods are available, no consensus has been reached as to which method is preferable. In 27 patients with dysfibrinogenemia, plasma fibrinogen concentration was measured by Clauss and PT-derived methods on two fully automated coagulation analyzers utilizing different reagents. In addition, immunological and heat fibrinogen concentrations as well as global coagulation tests were measured. Results: The median fibrinogen determined by the Clauss assay was 0.40 g/l, with a range of 0.30–2.07 g/l (normal range: 2.67–4.37 g/l) and 0.60 g/l, with a range of 0.60–2.20 g/l (normal range: 1.5–4.5 g/l) using two different reagents. The median fibrinogen determined by the PT-derived method was 2.41 g/l, with a range of 0.97–4.87 g/l (normal range: 1.84–4.8 g/l) and 2.64 g/l, ranging from 1.38 to 4.39 g/l (normal range: 2.0–4.0 g/l) by the use of two different reagents. No correlation was found when comparing both methods using two reagents from different manufacturers. The PT-derived method “overestimated“ the fibrinogen by approximately five times the value measured by the Clauss assay. While fibrinogen measured by the PT-derived method correlated with fibrinogen antigen concentrations measured by the immunological fibrinogen (pb 0.002) or heat fibrinogen method (pb 0.002), fibrinogen measured by the Clauss method correlated with functional coagulation parameters, such as Reptilase Time (pb 0.002), Thrombin Time (pb 0.002) or Prothrombin Time (pb 0.02). Conclusion: Although many patients with dysfibinogenemia are asymptomatic, in case of bleeding, immediately diagnosis and treatment is warranted. The Clauss assay is the diagnostic tool of choice when diagnosing or treating patients with low fibrinogen levels. The use of the PT-derived method may potentially pose a greater risk to patients, as the plasma concentration may be erroneously reported as normal. © 2010 Elsevier Ltd. All rights reserved.
Introduction A number of methods are used to determine fibrinogen concentration in plasma, although no consensus has yet been reached as to which method is preferable. Several studies that have compared different methods used to determine fibrinogen concentration have been conducted. The currently available methods to measure plasma fibrinogen concentration include radial immunodiffusion [1], electro immunoassay, protein precipitation by physicochemical factors [2] and functional assays [3–6]. The gold standard of all methods is the so-called “clot-recovery method”, which is based upon the direct quantification of the generated fibrin in clots [7,8]. Although this method is highly
⁎ Corresponding author. Haemophilia Centre, Medical Clinic III /Institute of Transfusion Medicine, Johann Wolfgang Goethe-University, Theodor-Stern-Kai 7, 60590 Frankfurt am Main, Germany. Tel.: +49 69 6301 5051; fax: +49 69 6301 6738. E-mail address: [email protected] (W. Miesbach). 0049-3848/$ – see front matter © 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.thromres.2010.09.004
accurate and reproducible, it is time-consuming and, therefore, not suited to clinical applications. The widely used prothrombin-time- (PT-) derived method [4,5] has the advantage of providing both PT and fibrinogen levels estimates, with the latter at no cost. The PT-derived method is not a direct determination of plasma fibrinogen. Instead, this method estimates fibrinogen based upon absorbance changes during a prothrombin time assay. The magnitude of scattered light at 450 nm delta OD during the clotting process of PT on automated photo optical coagulometers [1,2] has been positively correlated with different fibrinogen concentrations [4,5]. The reference curve for the determination of fibrinogen is obtained by progressively diluting the same calibration plasma with a buffer solution. Given that this buffer may affect the rate and/or mode of fibrin fibrils formation, there is much concern with regard to the accuracy of the PT-derived method [9–11]. Another widely used method is the functional fibrinogen assay according to Clauss [3]. The Clauss fibrinogen assay is based upon the thrombin clotting time. Like the PT-derived method, it does not measure fibrinogen directly, but determines the time until detectable
W. Miesbach et al. / Thrombosis Research 126 (2010) e428–e433
clot formation. This assay measures the clotting time of citrate plasma after the addition of excess thrombin (50-100 NIH U/ml final concentration) [6]. There is a great deal of variability in the assay protocol itself, as well as in the source and composition of the reagents that are used. In addition, the assay results may be dependent on the ratio of high to low molecular weight fibrinogen [12,13] as well as the presence of fibrin(ogen) degradation products [14]. The precision of each of the methods in measuring the actual concentration of fibrinogen in plasma is currently unknown. The aim of our study was to investigate whether or not the PT-derived method can provide reliable results of fibrinogen even under more challenging conditions, such as lower reference levels in patients with dysfibrinogenemia and to compare these results to those of the Clauss assay. Patients and Methods Each patient's evaluation included the standard screening tests: Prothrombin time (PT), Reptilase time (RT) and Thrombin time (TT). Fibrinogen was assayed by several methods: the functional method according to Clauss [3], the immunological method by radioimmunodiffusion [1], the heat precipitable fibrinogen method (Schulz, 15) and the PT-derived method [3,4]. However, the heat method could only be performed in 24 patient's blood samples. Patients Twenty-seven patients were entered in the study, with nineteen females and eight males. The median age was 53 years, with a range from 22-77 years. We included all samples of patients with dysfibrinogenamia who were consecutively treated at our outpatient department. Diagnosis of dysfibrinogenamia was established in all patients by measuring several fibrinogen tests (reduced fibrinogen activity–antigen ratio, fibrinogen Clauss with reagents from IL). All patients had previous laboratory results that were consistent with a diagnosis of dysfibrinogenemia. None of the patients had hypofibrinogenemia. Two patients suffered from acquired dysfibrinogenemia due to liver disease. Other bleeding disorders, e.g. von Willebrand syndrome were excluded. None of the patients were treated with heparin. In 67% of the patients with inherited dysfibrinogenemia, a fibrinogen gene analysis (Dr. D. Galanakis, New York, Dr. C. Geisen, Frankfurt) confirmed the diagnosis. Methods Our study compared the PT-derived method with the Clauss fibrinogen assay on two automatic coagulation analyzers, which were the ACL-Top (Instrumentation Laboratory; Kirchheim, Germany) and the STA-R Evolution (Roche Diagnostics; Mannheim, Germany). The determination fibrinogen by the Clauss methods was performed using two different reagents, which were Fibrinogen C (Instrumentation Laboratory, coagulation analyzer: ACL-Top) and STA Fibrinogen (Roche Diagnostics, coagulation analyzer: STA-R-evolution). We used two different reagents and analyzers for each method to exclude possible unique influences from only one manufacturer. The normal range of fibrinogen Clauss has been established by the the manufactur's specifications. The determination of fibrinogen with the PT-derived method was conducted utilizing two different reagents, which were HemosIL PT-Fibrinogen HS PLUS (Instrumentation Laboratory, coagulation analyzer: ACL Top) and STA Neoplastin Plus (Roche Diagnostics, coagulation analyzer: STA-R-Evolution). The immunological fibrinogen was determined using NOR-Partigen-Fibrinogen (Siemens,Marburg, Germany). The heat fibrinogen was determined according to the Schulz method [15]. For determination of fibrinogen according to the heat method two microhematocrit tubes are filled with patients plasma. One is centrifuged
e429
and the total protein in the plasma is measured by refractometer. The second tube is heated at 56 C for 3 minutes, which precipitates the fibrinogen. The second tube is then centrifuged and the protein result read similarly. The protein result in the heated tube is subtracted from the result in the unheated tube; the difference is equivalent to the fibrinogen that was removed from the plasma in the second tube by heating and centrifuging. The pre-calibration of the Clauss fibrinogen assay (Roche Diagnostics and IL) has been determined with a secondary standard of the 98/612 International Standard established in 1999. When the plasma to be tested is 1:20 diluted, the STA Fibrinogen procedure for determination of Clauss fibrinogen has a linearity range of 1.5 - 9 g/l on STA analyzers and the IL assay has a linearity range of 1.5–6 g/L (dilution 1 : 10). Below these levels, there is a limitation concerning the linearity of the curves. We did not carry out an additional test to run samples lower than the detection limit. Data Analysis Statistical correlation between methods was performed by nonparametric analysis, using the Spearman-Rank-Correlation. This method of correlation was selected because comparisons could only be determined qualitatively and not quantitatively, since some of the measurements from the Clauss assay were below 0.6 g/l on the STA instrument and below 0.3 g/l on the ACL Top. For descriptive statistic, median mean and standard deviation of patient data were calculated. Results All 27 of the patients in this study had a prior diagnosis of dysfibrinogenemia. Among the patients, 25 suffered from a hereditary dysfibrinogenemia and 2 from an acquired dysfibrinogenemia due to liver disease (patient No.7 and No.27). Patients with acquired form of dysfibrinogenemia presented with clinical and laboratory symptoms of dysfibrinogenemia, suffered from liver disease and no fibrinogen gene mutation could be detected. Irrespective of abnormal laboratory results, 19 patients (70%) presented with at least one episode of undue hemorrhage, with easy bruising being the most common event. Other hemorrhagic events included post-surgical hemorrhage (n = 9), hemorrhage from gums (n = 4), menorrhagia (n = 3), and hemorrhage from gastro-intestinal sites (n = 1). Only 4 of the 27 patients (15%) were asymptomatic. Thirty-three percent of the patients (9/27) had a history of one or more arterial or venous thrombotic events. Four patients had a venous thrombosis, and one had thrombophlebitis. Four arterial thromboses occurred, including ischemic stroke (n = 3) and peripheral artery occlusion (n = 1) (Table 1). Two different reagents were used to obtain the measurements for this study. There was a significant correlation for the Clauss assay (rho = 0.85, p b 0.05) measured with the reagents by Instrumentation Laboratory and Roche Diagnostics and the PT-derived method (rho = 0.97, p b 0.05) measured with the reagents by Instrumentation Laboratory and Roche Diagnostics. The median fibrinogen determined by the Clauss assay using Fibrinogen C reagent was 0.40 g/l, with a range from 0.30 - 2.07 g/l (normal range: 2.67–4.37 g/l). The median fibrinogen determined by the Clauss assay using STA Fibrinogen was 0.6 g/l, with a range from 0.60–2.20 g/l (normal range: 1.5–4.5 g/l). The median fibrinogen determined by the PT-derived fibrinogen assay using HemosIL PT-Fibrinogen HS PLUS was 2.41 g/l, with a range from 0.97-4.87 g/l (normal range: 1.84–4.8 g/l). The median fibrinogen determined by the PT-derived fibrinogen assay using STA Neoplastin Plus was 2.64 g/l, with a range from 1.38 - 4.39 g/l (normal range: 2.0–4.0 g/l (Table 2). The median fibrinogen determined by the PT-derived fibrinogen method using either of the reagents was approximately five times higher than the median for fibrinogen determined by the Clauss assay.
e430
W. Miesbach et al. / Thrombosis Research 126 (2010) e428–e433
For the Roche Diagnostics and Instrumentation Laboratory reagents, 8/27 and 5/27, respectively, of the patients had levels above 3.0 g/l with the PT-derived method. The PT-derived method resulted in decreased levels of fibrinogen in only 15% of the patients (4/27) when determined with the coagulation analyzer and reagents from Roche Diagnostics, and in only 18% of the patients (5/27) when determined with the coagulation analyzer and reagents from IL. The median concentration for immunological fibrinogen was 2.74 g/l, with a range of 1.02 - 4.23 g/l (normal range: 2.05–4.39 g/l) and for heat fibrinogen was 2.50 g/l, with a range of 1.20 - 6.25 g/l (normal range: 2.0–4.0 g/l). In three patients (No. 15, 25 and 27), there were low fibrinogen levels even by using immunological method or heat fibrinogen. In two of them, diagnosis of dysfibrinogenemia could be confirmed by fibrinogen gene analysis. One patient had acquired dysfibrinogenemia. The determination of Prothrombin time (PT) resulted in a median of 14 sec. (12–21 sec., normal range: 12–14 sec.), Thrombin time (TT) resulted in a median of 35 sec. (17–42 sec., normal range: 16–20 sec.), of Reptilase time (RT) in a median of 120 sec. (20–120 sec., normal range: b20 sec.). No significant relationship was found when comparing both methods using two reagents from different manufacturers. The comparison of the measurements by Clauss assay and the PT-derived method with the reagents of Instrumentation Laboratory and Roche Diagnostics yielded a correlation of rho = 0.12 (p N 0.20) and rho = 0.17 (p N 0.20), respectively. Fibrinogen as determined by the Clauss method using either reagent correlated with screening coagulation parameter, including
Reptilase time (rho = - 0.66 and - 0.68, p b 0.002), Thrombin time (rho = - 0.69 and - 0.75, p b 0.002) or Prothrombin time (rho = 0.52 and 0.48, p b 0.01). Fibrinogen analyzed by the PT-derived method correlated with immunological fibrinogen levels (rho=0.86 and 0.82, pb 0.002) and heat fibrinogen levels (rho=0.91 and 0.92, pb 0.002) (Fig. 1). The PT derived fibrinogen reagent used.for this figure was the HemosIL PT-Fibrinogen HS PLUS (Instrumentation Laboratory, coagulation analyzer: ACL Top).
Discussion Although several studies have compared the Clauss fibrinogen assay with the PT-derived method, as well as the variability of fibrinogen assay results between different laboratories, most studies were conducted on healthy populations. Only a few of these studies included patients with coagulation disorders or patients receiving anticoagulant therapy. Our study is the first to compare the performance of different fibrinogen assays on a larger group of patients with dysfibrinogenemia. We were able to consistently demonstrate that fibrinogen levels in patients with dysfibrinogemia were significantly overestimated by the PT-derived methods, regardless of the testing equipment or the reagents. Measurement of fibrinogen by the functional Clauss method correlated with other functional coagulation assays, such as the Reptilase Time or Thrombin Time. Results of PT-derived fibrinogen, however, correlated with fibrinogen antigen concentrations that were determined by immunologic methods (radial immunodiffusion) and precipitation methods
Table 1 Demographic data and clinical symptoms of the included patients. Demographic data
Clinical symptoms
Patient
Age
Gender
Bleeding
Thrombosis
1 2 3 4
67 48 63 62
F F F F
Post-surgical hemorrhage, gum bleeding Easy bruising, post-surgical hemorrhage Easy brusising Easy bruising, gum bleeding
Stroke Peripheral artery occlusion
5 6
77 51
M F
7
70
M
8 9 10 11
22 62 67 60
M F M F
12
36
F
13 14 15 16 17 18
40 39 25 50 26 46
F F M F F M
19
74
M
20 21 22 23 24 25 26 27
73 31 25 62 55 53 72 48
F F F F M F F M
median Range
53 years 22–77 years
F: 67%, M: 33%
Stroke Post-surgical hemorrhage, gum bleeding Post-surgical hemorrhage, gastro-intestinal
Easy bruising Easy bruising, post-surgical hemorrhage, menorrhagia Easy bruising, post-surgical hemorrhage, menorrhagia Easy bruising, post-surgical hemorrhage Easy bruising Easy bruising, post-surgical hemorrhage
Deep vein thrombosis
Deep vein thrombosis Thrombophlebitis Deep vein thrombosis
Easy bruising, post-surgical hemorrhage, gum bleeding Easy bruising, post-surgical hemorrhage Post-surgical hemorrhage Easy bruising, post-surgical hemorrhage Easy bruising, menorrhagia Post-surgical hemorrhage Deep vein thrombosis Stroke Easy bruising, post-surgical hemorrhage 70%
33%
W. Miesbach et al. / Thrombosis Research 126 (2010) e428–e433
e431
Table 2 Two fibrinogen assay methods for diagnosis of dysfibrinogenemia: A comparison of two assay brands for each method. Prothrombin (PT) derived assay
Immun.
Heat
Global functional assay
Patient
Clauss assay STA FIB-C g/l
ACL-Top FIB-C g/l
STA Fib PTder g/l
ACL-Top Fib PTder g/l
g/l
g/l
PT sec
TT sec
RT Sec
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27
b0.60 b0.60 b0.60 b0.60 b0.60 b0.60 0.78 b0.60 b0.60 b0.60 b0.60 b0.60 b0.60 b0.60 0.92 0.77 b0.60 1.57 b0.60 0.73 b0.60 0.81 2.20 b0.60 1.97 0.70 1.43
b 0.30 0.35 0.39 b 0.30 0.37 0.34 0.71 b 0.30 b 0.30 0.37 b 0.30 0.34 0.35 0.40 1.00 0.50 0.41 1.08 0.37 0.49 0.50 0.57 1.08 0.42 2.07 0.51 1.55
2.12 2.64 3.04 2.58 2.91 2.79 1.89 1.50 2.58 2.46 2.23 2.15 2.68 2.76 1.38 3.46 2.81 4.39 3.21 3.19 2.12 3.82 4.10 2.46 2.51 3.28 1.80
1.97 2.52 2.63 2.41 2.56 2.32 1.51 1.34 2.35 2.36 1.97 2.09 2.50 2.68 0.97 3.55 2.51 4.87 2.81 2.99 1.80 3.87 3.63 2.06 2.14 3.16 1.70
1.40 3.12 3.12 1.40 3.12 4.08 1.30 1.02 2.50 2.50 1.92 2.15 3.26 3.12 1.11 3.66 2.74 4.23 3.12 3.00 1.40 3.94 Nd Nd Nd 2.74 Nd
1.50 3.00 3.00 1.80 3.00 3.00 1.20 1.20 1.80 2.40 1.50 2.40 2.50 3.70 120 4.30 2.40 5.60 3.70 6.25 1.80 5.00 3.70 3.00 1.80 6.25 1.40
13 14 13 14 13 13 12 18 14 15 15 14 14 14 12 19 13 12 14 13 12 14 14 12 21 14 12
28 38 35 38 37 35 25 41 42 37 36 36 37 35 22 33 Nd 32 35 28 28 33 Nd 35 18 36 17
80 N120 N120 N120 N120 N120 37 N120 N120 N120 N120 N120 N120 N120 22 108 N120 70 N120 N120 20 N120 37 N120 20 N120 20
2.64 1.38–4.39 2.0–4.0
2.41 0.97–4.87 1.84–4.8
2.74 1.02–4.23 2.05 –4.39
2.50 1.20–6.25 2.0–4.0
14 12–21 12–14
35 17–42 16–20
N120 20–N 120 b20
Median Range Normal range
0.60 0.60–2.20 1.5–4.5
0.40 0.30–2.07 2.67–4.37
(induced by heat according to Schulz). The median of the Clauss assay by IL was lower than the median of the Clauss assay by Roche Diagnostics because of the different detection limits of both tests. However, results outside the linear range are extrapolated and possibly do not accurately reflect the true plasma concentration.
Fig. 1. Correlation of the PT-derived (PT-der.) method with the immunological (imm.) method and the heat fibrinogen method.
Fibrinogen Gene
Asp 318 Gly Nd A-a-Arg-16-Cys g-Ala-357-Thr A-a-Arg-16-Cys A-a-Arg-16-Cys acquired g-Ala-357-Thr g-Ala-357-Thr A-a-Arg-16-Cys g-Ala-357-Thr A-a-Arg-16-Cys Nd A-a-Arg-16-Cys A-a-Arg-16-Cys Nd Nd A-a-Arg-16-Cys A-a-Arg-16-Cys Nd Cys65Tyr Nd Nd A-a-Arg-16-Cys A-a-Arg-16-Cys A-a-Arg-16-Cys acquired
A large study of a random population sample (n = 1373) of men and women aged 24-64 years of age [16] found a similar range and distribution for both the Clauss assay and the PT-derived method. This study compared three routine assays, which were the Clauss assay, the PT- derived clottable fibrinogen assay, and the immunonephelometric assay. The discrepancy in results between this study and our study, as well as some other studies may reflect a different patient population sample and/or different instruments, reagents and standards [17,18]. Interestingly, there was a significant lack of agreement in defining high and low thirds of the fibrinogen level, despite the very similar mean values and distributions between the Clauss and PT-derived assays in this study. This discrepancy has been noted in other studies and may reflect differences in the clotting process between these assays. In another study, a patient cohort consisting of 411 normal subjects, 50 orally anti-coagulated patients and 23 patients suspicious for dysfibrinogenemia was used to compare the fibrinogen levels determined by the Clauss assay [19]. At the same time, additional studies, including radial immunodiffusion and heat precipitation, were performed on those patients with suspicion of dysfibrinogenemia. The overall results of this study indicated that the PT-derived method resulted in values that were, on average, 24% higher than those of the Clauss method in the group of anti-coagulated patients when compared to the normal group. Of the 23 patients with a suspicion of dysfibrinogenemia, the diagnosis was confirmed in six patients. In this study, there was a difference between the values obtained by the Clauss method and the PT-derived method (mean fibrinogen by Clauss method: 0.76 g/l vs. mean fibrinogen by PT-derived method: 3.64 g/l), whereas the values obtained by the fibrinogen antigen with radial immunodiffusion and the heat precipitation assay were similar. This study concluded that the PT-derived method may have an additional diagnostic utility, in which it can be used to identify polymerization abnormalities when there is a discrepancy between both determinations.
e432
W. Miesbach et al. / Thrombosis Research 126 (2010) e428–e433
The diagnosis of dysfibrinogenemia might be suspected, and additional tests should be performed in these cases. Similar results were reported in an earlier study on a cohort that included 20 healthy individuals and 38 patients with various diagnoses [20]. All participants were tested with both the PT-derived and the Clauss method. The results demonstrated that PT-derived values were significantly higher than the Clauss measurements. The discrepancy was greater in certain groups, such as those patients receiving oral anticoagulants. In some cases involving patients with documented hypofibrinogenemia and/or dysfibrinogenemia, the PT-derived method determined that the fibrinogen levels were within the normal ranges, whereas the Clauss assay determined that the fibrinogen levels were low. Another study included samples from 242 normal and 44 orally anticoagulated subjects [21]. Although the fibrinogen determination was reliable and accurate with each method in normal subjects, there was a systematically overestimated fibrinogen level in plasma with high fibrinogen concentrations (N400 mg/dl) and in patients on vitamin K antagonist therapy. For patients on anticoagulant therapy, the PT-derived fibrinogen suffered from a distinct systematic error, with a progressive overestimation at INR values ≥1.25, which is a phenomenon that has been previously described [20]. These systematic errors were thought to be the result of an increased fibrin gel turbidity that was caused by the slower rate of fibrinogen cleavage. The results of various studies indicate that the PT-derived method may be safely used to determine fibrinogen when the main focus is to predict an increased risk of vascular complications that are associated with high plasma fibrinogen levels [22] or to explore the link between coagulation parameters and the risk of thrombotic events in large-scale studies. The heterogeneity of plasma fibrinogen may act as a contributing factor to the discrepancies that are seen between the different fibrinogen assays [23]. The results of different fibrinogen assays may provide contrasting, or even conflicting, information on clottability in different subjects, as shown by case control studies in acute myocardial infarction [24], previous myocardial infarction [24] and chronic peripheral artery disease [25]. The PT-derived assay can economically and easily be performed in laboratories with suitable instruments. Due to the many potential sources for discrepancies between the PT-derived and Clauss methods, extrapolation of the results must be performed with care and only in well characterized groups of patients and not in inter-group comparisons. The systematic comparison of PT-derived fibrinogen values with Clauss fibrinogen values could constitute a process validation tool and help to find abnormalities of fibrinogen polymerization. Although a few studies have included patients with dysfibrinogenemia, overall, fibrinogen in patients with high or low fibrinogen levels has been found to be more accurately determined by the Clauss method than the PT-derived method. Currently available methods for determination of fibrinogen also include enzyme immunoassays, as well as immunoturbidimetric assays. Both are in considerably more widespread use for measurement of immunological fibrinogen levels than radial immunodiffusion or electroimmunoassays. In our study, fibrinogen measured with the Clauss assay was almost always decreased. In contrast, fibrinogen measured with the PT-derived method yielded mild to moderately decreased levels of fibrinogen in 15% of the patients when reagents from Roche Diagnostics were used and in 18% of the patients when reagents from Instrumentation Laboratory. The PT-derived fibrinogen levels and the results of the immunoassay and the heat precipitation may reflect the true fibrinogen antigen level and correlate well. Some abnormal fibrinogens, however, may lead to abnormally translucent gels, causing an underestimation of fibrinogen levels in the PT-derived fibrinogen (fibrinogen Mannheim V, [26], among others). The functional assays as determined by the clotting time assay may reflect the individual bleeding risk in patients with dysfibfrinogenemia. Thus, in patients with bleeding tendencies, fibrinogen should be deter-
mined both by PT-derived and clotting time assay, in order not to miss functional defects of the fibrinogen molecule. Furthermore, diagnosis of dysfibrinogenemia may be carried out by the discrepancy of the clotting time assay and the PT-derived fibrinogen since PT-derived fibrinogen may be more convenient for many laboratories than an immunoassay for initial screening. These results clearly show that further work is needed to investigate the possible causes of discrepancies between different plasma fibrinogen assays and their potential clinical significance. In addition, more studies in both the general population and in patients with coagulopathies, such as dysfibrinogenemia, are needed to provide more data. Limitations of the study In most patients the Clauss method measured values below the detection limit. The aim of the study was to determine if PT-derived method provides reliable results in the lower range of fibrinogen and if this method correlates with the Clauss method. Therefore we did not retest the patient's plasma in lower dilution to get more exact results. By omitting results below the detection limit, the correlation coefficient between the two assays would be even worst. For the same reason, we did not add further measurements with assays of a high reproducibility, e.g. the determination of the clottable protein or the immunoturbidimetric method for the fibrinogen antigen. Conflict of interest statement The authors state no conflict of interest. References [1] Mancini G, Carbonara AO, Heremans JF. Immunochemical quantitation of antigens by single radial immunodiffusion. Immunochemistry 1965;2:235–54. [2] Desvignes P, Bonnet P. Direct determination of plasma fibrinogen levels by heat precipitation: a comparison of the techniques against clottable fibrinogen with spectophotometry and radial immunodiffusion. Clin Chim Acta 1981;110: 9–17. [3] Clauss A. Gerinnungsphysiologische Schnellmethode zur Bestimmung des Fibrinogens. Acta Haematol 1957;17:237–46. [4] Chantarangkul V, Tripodi A, Manucci PM. Evaluation of a fully automated centrifugal analyzer for performance of hemostasis tests. Clin Chem 1987;33: 1888–90. [5] Rossi E, Mondonico P, Lombardi A, Preda L. Method for the determination of functional (clottable) fibrinogen by the new family of ACL coagulometers. Thromb Res 1988;52:453–68. [6] Tan V, Doyle CJ, Budzynski AZ. Comparison of the kinetic fibrinogen assay with the Clauss method and the clot recovery method in plasma of patients with conditions affecting fibrinogen coagulability. Am J Clin Pathol 1995;104:455–62. [7] Jacobsson K. Studies on the determination of fibrinogen in human blood plasma. Scand J Clin Lab Invest 1955;7(suppl 14):1–54. [8] Gaffney PJ, Wong MY. Collaborative study of a propose international standard for plasma fibrinogen measurement. Thromb Haemost 1992;68:428–32. [9] Chantarangul V, Tripodi A, Manucci PM. Results of a collaborative study for fibrinogen measurement. Evidence that the use of a common calibrator improves interlaboratory agreement. Blood Coagul Fibrinolysis 1994;5:761–6. [10] Kitchen S, Jennings I, Preston FE. Comparison of fibrinogen determinations using a Clauss assay and two prothrombin time derived methods [abstract]. Thomb Haemost 1995;73:1245. [11] Palareti G, Maccaferri M, Manotti C. Fibrinogen assays: a collaborative study of six different methods. Clin Chem 1991;37:714–9. [12] Hoffmann JJML, Vijgen M, Niewenhuizen W. Comparison of the four fibrinogen assays during thrombolytic therapy. Fibrinolysis 1990;4(suppl 2):121–3. [13] Seifried E, Oethinger M, Tanswell P, Hoege-de Nobel H, Niewenhuizen W. Studies on the functionality of fibrinogen during rt-PA therapy: results of three different methods of fibrinogen determination. Blood Coagul Fibrinolysis 1992;3:81–7. [14] Siemens HJ, Wagner T. Clinical experiences of fibrinogen determination at ACL. Labor Med 1989;12:125–8. [15] Schulz FH. Eine einfache Bewertungsmethode von Leberparenchymschäden (volumetrische Fibrinbogenbestimmung). Acta Hepatol 1995;3:306–13. [16] Rumley A, Woodward M, Hoffmeister A, Koenig W, Lowe GDO. Comparison of plasma fibrinogen by Clauss, prothrombin time-derived, and immunonephelometric
W. Miesbach et al. / Thrombosis Research 126 (2010) e428–e433
[17] [18] [19]
[20] [21]
assays in a general population: implications for risk stratification by thirds of fibrinogen. Blood Coagul Fibrinolysis 2003;14:197–201. Mackie IJ, Lawrie AS, Howarth D, Martin J, Rumley A, Lowe GDO, et al. abstract. Blood Coagul Fibrinolysis 1999;10:S102. Chitolie A, Mackie U, Grant D, Hamilton JL, Machin SM. Inaccuracy of the ‘derived’ fibrinogen measurement. Blood Coagul Fibrinolysis 1994;5:955–7. Llamas P, Santos AB, Outeirino J, Soto C, Tomas JF. Diagnostic utility of comparing fibrinogen Clauss and prothrombin time derived method. Thromb Res 2004;114: 73–4. Chitolie A, Mackie IJ, Grant D, Hamilton JL, Machin SM. Inaccuracy of the ' derived' fibrinogen measurement. Blood Coagul Fibrinolysis 1994 Dec;5(6):955–7. De Cristofaro R, Landolfi R. Measurement of plasma fibrinogen concentration by the prothrombin-time-derived method: applicability and limitations. Blood Coagul Fibrinolysis 1998;9:251–9.
e433
[22] Heinrich J, Balleisen L, Schulte H, Assmann G, van de Loo J. Fibrinogen and factor VII in the prediction of coronary risk. Results from the PROCAM study in healthy men. Arterioscler Thromb 1994;14:54–9. [23] Niewenhuizen W. Biochemistry and measurement of fibrinogen. Eur Heart J 1995;16:A6–A10. [24] Rumley A, Lowe GDO, Woodward M, Morrison C, Niewenhuizen W. Comparison of four fibrinogen assays in a case-control study of previous myocardial infarction [abstract]. Thromb Haemost 1999;82(suppl):847. [25] Lowe GDO, Rumley A, Niewenhuizen W, Lee AJ, Smith FB, Fowkes FGR. abstract. Thromb Haemost 1999;82:274 (suppl). [26] Dempfle CE, George PM, Borggrefe M, Neumaier M, Brennan SO. Demonstration of heterodimeric fibrinogen molecules partially conjugated with albumin in a novel dysfibrinogen: fibrinogen Mannheim V. Thromb Haemost 2009 Jul;102(1):29–34.
Contents lists available at ScienceDirect
Thrombosis Research j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / t h r o m r e s
Regular Article
Comparison of the fibrinogen Clauss assay and the fibrinogen PT derived method in patients with dysfibrinogenemia W. Miesbach a,⁎, J. Schenk b, S. Alesci a, E. Lindhoff-Last c a b c
Haemophilia Center, Medical Clinic III/Institute of Transfusion Medicine, Johann Wolfgang Goethe-University, Frankfurt am Main, Germany University Hospital Saarland, Institute of Transfusion Medicine, Germany Goethe University Frankfurt, Medical Clinic III/Department of Angiology, Germany
a r t i c l e
i n f o
Article history: Received 13 April 2010 Received in revised form 30 August 2010 Accepted 2 September 2010 Available online 13 October 2010 Keywords: Fibrinogen Clauss PT-derived fibrinogen Diagnosis Dysfibrinogenemia
a b s t r a c t Fibrinogen assays are an important screening tool for blood coagulation disorders. Although different methods are available, no consensus has been reached as to which method is preferable. In 27 patients with dysfibrinogenemia, plasma fibrinogen concentration was measured by Clauss and PT-derived methods on two fully automated coagulation analyzers utilizing different reagents. In addition, immunological and heat fibrinogen concentrations as well as global coagulation tests were measured. Results: The median fibrinogen determined by the Clauss assay was 0.40 g/l, with a range of 0.30–2.07 g/l (normal range: 2.67–4.37 g/l) and 0.60 g/l, with a range of 0.60–2.20 g/l (normal range: 1.5–4.5 g/l) using two different reagents. The median fibrinogen determined by the PT-derived method was 2.41 g/l, with a range of 0.97–4.87 g/l (normal range: 1.84–4.8 g/l) and 2.64 g/l, ranging from 1.38 to 4.39 g/l (normal range: 2.0–4.0 g/l) by the use of two different reagents. No correlation was found when comparing both methods using two reagents from different manufacturers. The PT-derived method “overestimated“ the fibrinogen by approximately five times the value measured by the Clauss assay. While fibrinogen measured by the PT-derived method correlated with fibrinogen antigen concentrations measured by the immunological fibrinogen (pb 0.002) or heat fibrinogen method (pb 0.002), fibrinogen measured by the Clauss method correlated with functional coagulation parameters, such as Reptilase Time (pb 0.002), Thrombin Time (pb 0.002) or Prothrombin Time (pb 0.02). Conclusion: Although many patients with dysfibinogenemia are asymptomatic, in case of bleeding, immediately diagnosis and treatment is warranted. The Clauss assay is the diagnostic tool of choice when diagnosing or treating patients with low fibrinogen levels. The use of the PT-derived method may potentially pose a greater risk to patients, as the plasma concentration may be erroneously reported as normal. © 2010 Elsevier Ltd. All rights reserved.
Introduction A number of methods are used to determine fibrinogen concentration in plasma, although no consensus has yet been reached as to which method is preferable. Several studies that have compared different methods used to determine fibrinogen concentration have been conducted. The currently available methods to measure plasma fibrinogen concentration include radial immunodiffusion [1], electro immunoassay, protein precipitation by physicochemical factors [2] and functional assays [3–6]. The gold standard of all methods is the so-called “clot-recovery method”, which is based upon the direct quantification of the generated fibrin in clots [7,8]. Although this method is highly
⁎ Corresponding author. Haemophilia Centre, Medical Clinic III /Institute of Transfusion Medicine, Johann Wolfgang Goethe-University, Theodor-Stern-Kai 7, 60590 Frankfurt am Main, Germany. Tel.: +49 69 6301 5051; fax: +49 69 6301 6738. E-mail address: [email protected] (W. Miesbach). 0049-3848/$ – see front matter © 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.thromres.2010.09.004
accurate and reproducible, it is time-consuming and, therefore, not suited to clinical applications. The widely used prothrombin-time- (PT-) derived method [4,5] has the advantage of providing both PT and fibrinogen levels estimates, with the latter at no cost. The PT-derived method is not a direct determination of plasma fibrinogen. Instead, this method estimates fibrinogen based upon absorbance changes during a prothrombin time assay. The magnitude of scattered light at 450 nm delta OD during the clotting process of PT on automated photo optical coagulometers [1,2] has been positively correlated with different fibrinogen concentrations [4,5]. The reference curve for the determination of fibrinogen is obtained by progressively diluting the same calibration plasma with a buffer solution. Given that this buffer may affect the rate and/or mode of fibrin fibrils formation, there is much concern with regard to the accuracy of the PT-derived method [9–11]. Another widely used method is the functional fibrinogen assay according to Clauss [3]. The Clauss fibrinogen assay is based upon the thrombin clotting time. Like the PT-derived method, it does not measure fibrinogen directly, but determines the time until detectable
W. Miesbach et al. / Thrombosis Research 126 (2010) e428–e433
clot formation. This assay measures the clotting time of citrate plasma after the addition of excess thrombin (50-100 NIH U/ml final concentration) [6]. There is a great deal of variability in the assay protocol itself, as well as in the source and composition of the reagents that are used. In addition, the assay results may be dependent on the ratio of high to low molecular weight fibrinogen [12,13] as well as the presence of fibrin(ogen) degradation products [14]. The precision of each of the methods in measuring the actual concentration of fibrinogen in plasma is currently unknown. The aim of our study was to investigate whether or not the PT-derived method can provide reliable results of fibrinogen even under more challenging conditions, such as lower reference levels in patients with dysfibrinogenemia and to compare these results to those of the Clauss assay. Patients and Methods Each patient's evaluation included the standard screening tests: Prothrombin time (PT), Reptilase time (RT) and Thrombin time (TT). Fibrinogen was assayed by several methods: the functional method according to Clauss [3], the immunological method by radioimmunodiffusion [1], the heat precipitable fibrinogen method (Schulz, 15) and the PT-derived method [3,4]. However, the heat method could only be performed in 24 patient's blood samples. Patients Twenty-seven patients were entered in the study, with nineteen females and eight males. The median age was 53 years, with a range from 22-77 years. We included all samples of patients with dysfibrinogenamia who were consecutively treated at our outpatient department. Diagnosis of dysfibrinogenamia was established in all patients by measuring several fibrinogen tests (reduced fibrinogen activity–antigen ratio, fibrinogen Clauss with reagents from IL). All patients had previous laboratory results that were consistent with a diagnosis of dysfibrinogenemia. None of the patients had hypofibrinogenemia. Two patients suffered from acquired dysfibrinogenemia due to liver disease. Other bleeding disorders, e.g. von Willebrand syndrome were excluded. None of the patients were treated with heparin. In 67% of the patients with inherited dysfibrinogenemia, a fibrinogen gene analysis (Dr. D. Galanakis, New York, Dr. C. Geisen, Frankfurt) confirmed the diagnosis. Methods Our study compared the PT-derived method with the Clauss fibrinogen assay on two automatic coagulation analyzers, which were the ACL-Top (Instrumentation Laboratory; Kirchheim, Germany) and the STA-R Evolution (Roche Diagnostics; Mannheim, Germany). The determination fibrinogen by the Clauss methods was performed using two different reagents, which were Fibrinogen C (Instrumentation Laboratory, coagulation analyzer: ACL-Top) and STA Fibrinogen (Roche Diagnostics, coagulation analyzer: STA-R-evolution). We used two different reagents and analyzers for each method to exclude possible unique influences from only one manufacturer. The normal range of fibrinogen Clauss has been established by the the manufactur's specifications. The determination of fibrinogen with the PT-derived method was conducted utilizing two different reagents, which were HemosIL PT-Fibrinogen HS PLUS (Instrumentation Laboratory, coagulation analyzer: ACL Top) and STA Neoplastin Plus (Roche Diagnostics, coagulation analyzer: STA-R-Evolution). The immunological fibrinogen was determined using NOR-Partigen-Fibrinogen (Siemens,Marburg, Germany). The heat fibrinogen was determined according to the Schulz method [15]. For determination of fibrinogen according to the heat method two microhematocrit tubes are filled with patients plasma. One is centrifuged
e429
and the total protein in the plasma is measured by refractometer. The second tube is heated at 56 C for 3 minutes, which precipitates the fibrinogen. The second tube is then centrifuged and the protein result read similarly. The protein result in the heated tube is subtracted from the result in the unheated tube; the difference is equivalent to the fibrinogen that was removed from the plasma in the second tube by heating and centrifuging. The pre-calibration of the Clauss fibrinogen assay (Roche Diagnostics and IL) has been determined with a secondary standard of the 98/612 International Standard established in 1999. When the plasma to be tested is 1:20 diluted, the STA Fibrinogen procedure for determination of Clauss fibrinogen has a linearity range of 1.5 - 9 g/l on STA analyzers and the IL assay has a linearity range of 1.5–6 g/L (dilution 1 : 10). Below these levels, there is a limitation concerning the linearity of the curves. We did not carry out an additional test to run samples lower than the detection limit. Data Analysis Statistical correlation between methods was performed by nonparametric analysis, using the Spearman-Rank-Correlation. This method of correlation was selected because comparisons could only be determined qualitatively and not quantitatively, since some of the measurements from the Clauss assay were below 0.6 g/l on the STA instrument and below 0.3 g/l on the ACL Top. For descriptive statistic, median mean and standard deviation of patient data were calculated. Results All 27 of the patients in this study had a prior diagnosis of dysfibrinogenemia. Among the patients, 25 suffered from a hereditary dysfibrinogenemia and 2 from an acquired dysfibrinogenemia due to liver disease (patient No.7 and No.27). Patients with acquired form of dysfibrinogenemia presented with clinical and laboratory symptoms of dysfibrinogenemia, suffered from liver disease and no fibrinogen gene mutation could be detected. Irrespective of abnormal laboratory results, 19 patients (70%) presented with at least one episode of undue hemorrhage, with easy bruising being the most common event. Other hemorrhagic events included post-surgical hemorrhage (n = 9), hemorrhage from gums (n = 4), menorrhagia (n = 3), and hemorrhage from gastro-intestinal sites (n = 1). Only 4 of the 27 patients (15%) were asymptomatic. Thirty-three percent of the patients (9/27) had a history of one or more arterial or venous thrombotic events. Four patients had a venous thrombosis, and one had thrombophlebitis. Four arterial thromboses occurred, including ischemic stroke (n = 3) and peripheral artery occlusion (n = 1) (Table 1). Two different reagents were used to obtain the measurements for this study. There was a significant correlation for the Clauss assay (rho = 0.85, p b 0.05) measured with the reagents by Instrumentation Laboratory and Roche Diagnostics and the PT-derived method (rho = 0.97, p b 0.05) measured with the reagents by Instrumentation Laboratory and Roche Diagnostics. The median fibrinogen determined by the Clauss assay using Fibrinogen C reagent was 0.40 g/l, with a range from 0.30 - 2.07 g/l (normal range: 2.67–4.37 g/l). The median fibrinogen determined by the Clauss assay using STA Fibrinogen was 0.6 g/l, with a range from 0.60–2.20 g/l (normal range: 1.5–4.5 g/l). The median fibrinogen determined by the PT-derived fibrinogen assay using HemosIL PT-Fibrinogen HS PLUS was 2.41 g/l, with a range from 0.97-4.87 g/l (normal range: 1.84–4.8 g/l). The median fibrinogen determined by the PT-derived fibrinogen assay using STA Neoplastin Plus was 2.64 g/l, with a range from 1.38 - 4.39 g/l (normal range: 2.0–4.0 g/l (Table 2). The median fibrinogen determined by the PT-derived fibrinogen method using either of the reagents was approximately five times higher than the median for fibrinogen determined by the Clauss assay.
e430
W. Miesbach et al. / Thrombosis Research 126 (2010) e428–e433
For the Roche Diagnostics and Instrumentation Laboratory reagents, 8/27 and 5/27, respectively, of the patients had levels above 3.0 g/l with the PT-derived method. The PT-derived method resulted in decreased levels of fibrinogen in only 15% of the patients (4/27) when determined with the coagulation analyzer and reagents from Roche Diagnostics, and in only 18% of the patients (5/27) when determined with the coagulation analyzer and reagents from IL. The median concentration for immunological fibrinogen was 2.74 g/l, with a range of 1.02 - 4.23 g/l (normal range: 2.05–4.39 g/l) and for heat fibrinogen was 2.50 g/l, with a range of 1.20 - 6.25 g/l (normal range: 2.0–4.0 g/l). In three patients (No. 15, 25 and 27), there were low fibrinogen levels even by using immunological method or heat fibrinogen. In two of them, diagnosis of dysfibrinogenemia could be confirmed by fibrinogen gene analysis. One patient had acquired dysfibrinogenemia. The determination of Prothrombin time (PT) resulted in a median of 14 sec. (12–21 sec., normal range: 12–14 sec.), Thrombin time (TT) resulted in a median of 35 sec. (17–42 sec., normal range: 16–20 sec.), of Reptilase time (RT) in a median of 120 sec. (20–120 sec., normal range: b20 sec.). No significant relationship was found when comparing both methods using two reagents from different manufacturers. The comparison of the measurements by Clauss assay and the PT-derived method with the reagents of Instrumentation Laboratory and Roche Diagnostics yielded a correlation of rho = 0.12 (p N 0.20) and rho = 0.17 (p N 0.20), respectively. Fibrinogen as determined by the Clauss method using either reagent correlated with screening coagulation parameter, including
Reptilase time (rho = - 0.66 and - 0.68, p b 0.002), Thrombin time (rho = - 0.69 and - 0.75, p b 0.002) or Prothrombin time (rho = 0.52 and 0.48, p b 0.01). Fibrinogen analyzed by the PT-derived method correlated with immunological fibrinogen levels (rho=0.86 and 0.82, pb 0.002) and heat fibrinogen levels (rho=0.91 and 0.92, pb 0.002) (Fig. 1). The PT derived fibrinogen reagent used.for this figure was the HemosIL PT-Fibrinogen HS PLUS (Instrumentation Laboratory, coagulation analyzer: ACL Top).
Discussion Although several studies have compared the Clauss fibrinogen assay with the PT-derived method, as well as the variability of fibrinogen assay results between different laboratories, most studies were conducted on healthy populations. Only a few of these studies included patients with coagulation disorders or patients receiving anticoagulant therapy. Our study is the first to compare the performance of different fibrinogen assays on a larger group of patients with dysfibrinogenemia. We were able to consistently demonstrate that fibrinogen levels in patients with dysfibrinogemia were significantly overestimated by the PT-derived methods, regardless of the testing equipment or the reagents. Measurement of fibrinogen by the functional Clauss method correlated with other functional coagulation assays, such as the Reptilase Time or Thrombin Time. Results of PT-derived fibrinogen, however, correlated with fibrinogen antigen concentrations that were determined by immunologic methods (radial immunodiffusion) and precipitation methods
Table 1 Demographic data and clinical symptoms of the included patients. Demographic data
Clinical symptoms
Patient
Age
Gender
Bleeding
Thrombosis
1 2 3 4
67 48 63 62
F F F F
Post-surgical hemorrhage, gum bleeding Easy bruising, post-surgical hemorrhage Easy brusising Easy bruising, gum bleeding
Stroke Peripheral artery occlusion
5 6
77 51
M F
7
70
M
8 9 10 11
22 62 67 60
M F M F
12
36
F
13 14 15 16 17 18
40 39 25 50 26 46
F F M F F M
19
74
M
20 21 22 23 24 25 26 27
73 31 25 62 55 53 72 48
F F F F M F F M
median Range
53 years 22–77 years
F: 67%, M: 33%
Stroke Post-surgical hemorrhage, gum bleeding Post-surgical hemorrhage, gastro-intestinal
Easy bruising Easy bruising, post-surgical hemorrhage, menorrhagia Easy bruising, post-surgical hemorrhage, menorrhagia Easy bruising, post-surgical hemorrhage Easy bruising Easy bruising, post-surgical hemorrhage
Deep vein thrombosis
Deep vein thrombosis Thrombophlebitis Deep vein thrombosis
Easy bruising, post-surgical hemorrhage, gum bleeding Easy bruising, post-surgical hemorrhage Post-surgical hemorrhage Easy bruising, post-surgical hemorrhage Easy bruising, menorrhagia Post-surgical hemorrhage Deep vein thrombosis Stroke Easy bruising, post-surgical hemorrhage 70%
33%
W. Miesbach et al. / Thrombosis Research 126 (2010) e428–e433
e431
Table 2 Two fibrinogen assay methods for diagnosis of dysfibrinogenemia: A comparison of two assay brands for each method. Prothrombin (PT) derived assay
Immun.
Heat
Global functional assay
Patient
Clauss assay STA FIB-C g/l
ACL-Top FIB-C g/l
STA Fib PTder g/l
ACL-Top Fib PTder g/l
g/l
g/l
PT sec
TT sec
RT Sec
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27
b0.60 b0.60 b0.60 b0.60 b0.60 b0.60 0.78 b0.60 b0.60 b0.60 b0.60 b0.60 b0.60 b0.60 0.92 0.77 b0.60 1.57 b0.60 0.73 b0.60 0.81 2.20 b0.60 1.97 0.70 1.43
b 0.30 0.35 0.39 b 0.30 0.37 0.34 0.71 b 0.30 b 0.30 0.37 b 0.30 0.34 0.35 0.40 1.00 0.50 0.41 1.08 0.37 0.49 0.50 0.57 1.08 0.42 2.07 0.51 1.55
2.12 2.64 3.04 2.58 2.91 2.79 1.89 1.50 2.58 2.46 2.23 2.15 2.68 2.76 1.38 3.46 2.81 4.39 3.21 3.19 2.12 3.82 4.10 2.46 2.51 3.28 1.80
1.97 2.52 2.63 2.41 2.56 2.32 1.51 1.34 2.35 2.36 1.97 2.09 2.50 2.68 0.97 3.55 2.51 4.87 2.81 2.99 1.80 3.87 3.63 2.06 2.14 3.16 1.70
1.40 3.12 3.12 1.40 3.12 4.08 1.30 1.02 2.50 2.50 1.92 2.15 3.26 3.12 1.11 3.66 2.74 4.23 3.12 3.00 1.40 3.94 Nd Nd Nd 2.74 Nd
1.50 3.00 3.00 1.80 3.00 3.00 1.20 1.20 1.80 2.40 1.50 2.40 2.50 3.70 120 4.30 2.40 5.60 3.70 6.25 1.80 5.00 3.70 3.00 1.80 6.25 1.40
13 14 13 14 13 13 12 18 14 15 15 14 14 14 12 19 13 12 14 13 12 14 14 12 21 14 12
28 38 35 38 37 35 25 41 42 37 36 36 37 35 22 33 Nd 32 35 28 28 33 Nd 35 18 36 17
80 N120 N120 N120 N120 N120 37 N120 N120 N120 N120 N120 N120 N120 22 108 N120 70 N120 N120 20 N120 37 N120 20 N120 20
2.64 1.38–4.39 2.0–4.0
2.41 0.97–4.87 1.84–4.8
2.74 1.02–4.23 2.05 –4.39
2.50 1.20–6.25 2.0–4.0
14 12–21 12–14
35 17–42 16–20
N120 20–N 120 b20
Median Range Normal range
0.60 0.60–2.20 1.5–4.5
0.40 0.30–2.07 2.67–4.37
(induced by heat according to Schulz). The median of the Clauss assay by IL was lower than the median of the Clauss assay by Roche Diagnostics because of the different detection limits of both tests. However, results outside the linear range are extrapolated and possibly do not accurately reflect the true plasma concentration.
Fig. 1. Correlation of the PT-derived (PT-der.) method with the immunological (imm.) method and the heat fibrinogen method.
Fibrinogen Gene
Asp 318 Gly Nd A-a-Arg-16-Cys g-Ala-357-Thr A-a-Arg-16-Cys A-a-Arg-16-Cys acquired g-Ala-357-Thr g-Ala-357-Thr A-a-Arg-16-Cys g-Ala-357-Thr A-a-Arg-16-Cys Nd A-a-Arg-16-Cys A-a-Arg-16-Cys Nd Nd A-a-Arg-16-Cys A-a-Arg-16-Cys Nd Cys65Tyr Nd Nd A-a-Arg-16-Cys A-a-Arg-16-Cys A-a-Arg-16-Cys acquired
A large study of a random population sample (n = 1373) of men and women aged 24-64 years of age [16] found a similar range and distribution for both the Clauss assay and the PT-derived method. This study compared three routine assays, which were the Clauss assay, the PT- derived clottable fibrinogen assay, and the immunonephelometric assay. The discrepancy in results between this study and our study, as well as some other studies may reflect a different patient population sample and/or different instruments, reagents and standards [17,18]. Interestingly, there was a significant lack of agreement in defining high and low thirds of the fibrinogen level, despite the very similar mean values and distributions between the Clauss and PT-derived assays in this study. This discrepancy has been noted in other studies and may reflect differences in the clotting process between these assays. In another study, a patient cohort consisting of 411 normal subjects, 50 orally anti-coagulated patients and 23 patients suspicious for dysfibrinogenemia was used to compare the fibrinogen levels determined by the Clauss assay [19]. At the same time, additional studies, including radial immunodiffusion and heat precipitation, were performed on those patients with suspicion of dysfibrinogenemia. The overall results of this study indicated that the PT-derived method resulted in values that were, on average, 24% higher than those of the Clauss method in the group of anti-coagulated patients when compared to the normal group. Of the 23 patients with a suspicion of dysfibrinogenemia, the diagnosis was confirmed in six patients. In this study, there was a difference between the values obtained by the Clauss method and the PT-derived method (mean fibrinogen by Clauss method: 0.76 g/l vs. mean fibrinogen by PT-derived method: 3.64 g/l), whereas the values obtained by the fibrinogen antigen with radial immunodiffusion and the heat precipitation assay were similar. This study concluded that the PT-derived method may have an additional diagnostic utility, in which it can be used to identify polymerization abnormalities when there is a discrepancy between both determinations.
e432
W. Miesbach et al. / Thrombosis Research 126 (2010) e428–e433
The diagnosis of dysfibrinogenemia might be suspected, and additional tests should be performed in these cases. Similar results were reported in an earlier study on a cohort that included 20 healthy individuals and 38 patients with various diagnoses [20]. All participants were tested with both the PT-derived and the Clauss method. The results demonstrated that PT-derived values were significantly higher than the Clauss measurements. The discrepancy was greater in certain groups, such as those patients receiving oral anticoagulants. In some cases involving patients with documented hypofibrinogenemia and/or dysfibrinogenemia, the PT-derived method determined that the fibrinogen levels were within the normal ranges, whereas the Clauss assay determined that the fibrinogen levels were low. Another study included samples from 242 normal and 44 orally anticoagulated subjects [21]. Although the fibrinogen determination was reliable and accurate with each method in normal subjects, there was a systematically overestimated fibrinogen level in plasma with high fibrinogen concentrations (N400 mg/dl) and in patients on vitamin K antagonist therapy. For patients on anticoagulant therapy, the PT-derived fibrinogen suffered from a distinct systematic error, with a progressive overestimation at INR values ≥1.25, which is a phenomenon that has been previously described [20]. These systematic errors were thought to be the result of an increased fibrin gel turbidity that was caused by the slower rate of fibrinogen cleavage. The results of various studies indicate that the PT-derived method may be safely used to determine fibrinogen when the main focus is to predict an increased risk of vascular complications that are associated with high plasma fibrinogen levels [22] or to explore the link between coagulation parameters and the risk of thrombotic events in large-scale studies. The heterogeneity of plasma fibrinogen may act as a contributing factor to the discrepancies that are seen between the different fibrinogen assays [23]. The results of different fibrinogen assays may provide contrasting, or even conflicting, information on clottability in different subjects, as shown by case control studies in acute myocardial infarction [24], previous myocardial infarction [24] and chronic peripheral artery disease [25]. The PT-derived assay can economically and easily be performed in laboratories with suitable instruments. Due to the many potential sources for discrepancies between the PT-derived and Clauss methods, extrapolation of the results must be performed with care and only in well characterized groups of patients and not in inter-group comparisons. The systematic comparison of PT-derived fibrinogen values with Clauss fibrinogen values could constitute a process validation tool and help to find abnormalities of fibrinogen polymerization. Although a few studies have included patients with dysfibrinogenemia, overall, fibrinogen in patients with high or low fibrinogen levels has been found to be more accurately determined by the Clauss method than the PT-derived method. Currently available methods for determination of fibrinogen also include enzyme immunoassays, as well as immunoturbidimetric assays. Both are in considerably more widespread use for measurement of immunological fibrinogen levels than radial immunodiffusion or electroimmunoassays. In our study, fibrinogen measured with the Clauss assay was almost always decreased. In contrast, fibrinogen measured with the PT-derived method yielded mild to moderately decreased levels of fibrinogen in 15% of the patients when reagents from Roche Diagnostics were used and in 18% of the patients when reagents from Instrumentation Laboratory. The PT-derived fibrinogen levels and the results of the immunoassay and the heat precipitation may reflect the true fibrinogen antigen level and correlate well. Some abnormal fibrinogens, however, may lead to abnormally translucent gels, causing an underestimation of fibrinogen levels in the PT-derived fibrinogen (fibrinogen Mannheim V, [26], among others). The functional assays as determined by the clotting time assay may reflect the individual bleeding risk in patients with dysfibfrinogenemia. Thus, in patients with bleeding tendencies, fibrinogen should be deter-
mined both by PT-derived and clotting time assay, in order not to miss functional defects of the fibrinogen molecule. Furthermore, diagnosis of dysfibrinogenemia may be carried out by the discrepancy of the clotting time assay and the PT-derived fibrinogen since PT-derived fibrinogen may be more convenient for many laboratories than an immunoassay for initial screening. These results clearly show that further work is needed to investigate the possible causes of discrepancies between different plasma fibrinogen assays and their potential clinical significance. In addition, more studies in both the general population and in patients with coagulopathies, such as dysfibrinogenemia, are needed to provide more data. Limitations of the study In most patients the Clauss method measured values below the detection limit. The aim of the study was to determine if PT-derived method provides reliable results in the lower range of fibrinogen and if this method correlates with the Clauss method. Therefore we did not retest the patient's plasma in lower dilution to get more exact results. By omitting results below the detection limit, the correlation coefficient between the two assays would be even worst. For the same reason, we did not add further measurements with assays of a high reproducibility, e.g. the determination of the clottable protein or the immunoturbidimetric method for the fibrinogen antigen. Conflict of interest statement The authors state no conflict of interest. References [1] Mancini G, Carbonara AO, Heremans JF. Immunochemical quantitation of antigens by single radial immunodiffusion. Immunochemistry 1965;2:235–54. [2] Desvignes P, Bonnet P. Direct determination of plasma fibrinogen levels by heat precipitation: a comparison of the techniques against clottable fibrinogen with spectophotometry and radial immunodiffusion. Clin Chim Acta 1981;110: 9–17. [3] Clauss A. Gerinnungsphysiologische Schnellmethode zur Bestimmung des Fibrinogens. Acta Haematol 1957;17:237–46. [4] Chantarangkul V, Tripodi A, Manucci PM. Evaluation of a fully automated centrifugal analyzer for performance of hemostasis tests. Clin Chem 1987;33: 1888–90. [5] Rossi E, Mondonico P, Lombardi A, Preda L. Method for the determination of functional (clottable) fibrinogen by the new family of ACL coagulometers. Thromb Res 1988;52:453–68. [6] Tan V, Doyle CJ, Budzynski AZ. Comparison of the kinetic fibrinogen assay with the Clauss method and the clot recovery method in plasma of patients with conditions affecting fibrinogen coagulability. Am J Clin Pathol 1995;104:455–62. [7] Jacobsson K. Studies on the determination of fibrinogen in human blood plasma. Scand J Clin Lab Invest 1955;7(suppl 14):1–54. [8] Gaffney PJ, Wong MY. Collaborative study of a propose international standard for plasma fibrinogen measurement. Thromb Haemost 1992;68:428–32. [9] Chantarangul V, Tripodi A, Manucci PM. Results of a collaborative study for fibrinogen measurement. Evidence that the use of a common calibrator improves interlaboratory agreement. Blood Coagul Fibrinolysis 1994;5:761–6. [10] Kitchen S, Jennings I, Preston FE. Comparison of fibrinogen determinations using a Clauss assay and two prothrombin time derived methods [abstract]. Thomb Haemost 1995;73:1245. [11] Palareti G, Maccaferri M, Manotti C. Fibrinogen assays: a collaborative study of six different methods. Clin Chem 1991;37:714–9. [12] Hoffmann JJML, Vijgen M, Niewenhuizen W. Comparison of the four fibrinogen assays during thrombolytic therapy. Fibrinolysis 1990;4(suppl 2):121–3. [13] Seifried E, Oethinger M, Tanswell P, Hoege-de Nobel H, Niewenhuizen W. Studies on the functionality of fibrinogen during rt-PA therapy: results of three different methods of fibrinogen determination. Blood Coagul Fibrinolysis 1992;3:81–7. [14] Siemens HJ, Wagner T. Clinical experiences of fibrinogen determination at ACL. Labor Med 1989;12:125–8. [15] Schulz FH. Eine einfache Bewertungsmethode von Leberparenchymschäden (volumetrische Fibrinbogenbestimmung). Acta Hepatol 1995;3:306–13. [16] Rumley A, Woodward M, Hoffmeister A, Koenig W, Lowe GDO. Comparison of plasma fibrinogen by Clauss, prothrombin time-derived, and immunonephelometric
W. Miesbach et al. / Thrombosis Research 126 (2010) e428–e433
[17] [18] [19]
[20] [21]
assays in a general population: implications for risk stratification by thirds of fibrinogen. Blood Coagul Fibrinolysis 2003;14:197–201. Mackie IJ, Lawrie AS, Howarth D, Martin J, Rumley A, Lowe GDO, et al. abstract. Blood Coagul Fibrinolysis 1999;10:S102. Chitolie A, Mackie U, Grant D, Hamilton JL, Machin SM. Inaccuracy of the ‘derived’ fibrinogen measurement. Blood Coagul Fibrinolysis 1994;5:955–7. Llamas P, Santos AB, Outeirino J, Soto C, Tomas JF. Diagnostic utility of comparing fibrinogen Clauss and prothrombin time derived method. Thromb Res 2004;114: 73–4. Chitolie A, Mackie IJ, Grant D, Hamilton JL, Machin SM. Inaccuracy of the ' derived' fibrinogen measurement. Blood Coagul Fibrinolysis 1994 Dec;5(6):955–7. De Cristofaro R, Landolfi R. Measurement of plasma fibrinogen concentration by the prothrombin-time-derived method: applicability and limitations. Blood Coagul Fibrinolysis 1998;9:251–9.
e433
[22] Heinrich J, Balleisen L, Schulte H, Assmann G, van de Loo J. Fibrinogen and factor VII in the prediction of coronary risk. Results from the PROCAM study in healthy men. Arterioscler Thromb 1994;14:54–9. [23] Niewenhuizen W. Biochemistry and measurement of fibrinogen. Eur Heart J 1995;16:A6–A10. [24] Rumley A, Lowe GDO, Woodward M, Morrison C, Niewenhuizen W. Comparison of four fibrinogen assays in a case-control study of previous myocardial infarction [abstract]. Thromb Haemost 1999;82(suppl):847. [25] Lowe GDO, Rumley A, Niewenhuizen W, Lee AJ, Smith FB, Fowkes FGR. abstract. Thromb Haemost 1999;82:274 (suppl). [26] Dempfle CE, George PM, Borggrefe M, Neumaier M, Brennan SO. Demonstration of heterodimeric fibrinogen molecules partially conjugated with albumin in a novel dysfibrinogen: fibrinogen Mannheim V. Thromb Haemost 2009 Jul;102(1):29–34.