Fibrinolysis

6 August 2005
11:00 to 14:30
Harbourside Meeting Room 4
Sydney Convention and Exhibition Centre

Chair: O. Matsuo, Japan
Co-chairs: C. Dempfle, Germany; D. Hendriks, Belgium; C. Longstaff, UK; M. Nesheim, Canada

TAFI/proCPU

 Dr Willemse presented a new kinetic assay for measuring proCPU/TAFI levels in plasma. He first gave an introduction about currently available assays and pointed out the major pitfalls of antigen and activity-based assays. The novel assay he presented is based on the quantification of arginine-cleaved from hippuryl-L-arginine by CPU using 3 coupling enzymes (arginine kinase, pyruvate kinase and lactate dehydrogenase) finally leading to the consumption of NADH which can be followed continuously at 340 nm. The assay shows excellent correlation with the HPLA-assisted reference assay and has a high precision. Compared with HPLC, the assay is much easier to perform and it allows a much faster determination of proCPU concentrations. This, combined with the broad linear range of proCPU determination (100-2400U/L) makes it a useful tool for sensitive screening of clinical samples. Because the assay measures the cleaved arginine it can be used with all kinds of C-terminal arginine-containing substrates and can be used as a useful tool for screening different synthetic and physiological CPU substrates.

In discussion the problem of availability of one of the enzymes needed for the coupled assay was highlighted since it must be purified and is not commercially available.

Standardization of fibrinolytic factors

Calibration of SSC Plasmas

Dr. Longstaff reported on the calibration of SSC plasma #2 and #3 to measure t-PA antigen, PAI-1 antigen and activity. Eight labs joined in the collaboration study to assess the feasibility of calibration. GCV results for t-PA antigen in SSC #2 and #3 were variable, and a little less so for 94/730. However, there was some improvement in variability after the removal of outlier and 94/730 was close to the expected value of 25 ng/ml. Normalisation of SSC2 and 3 using 94/730 as calibrator improved the spread of data and suggests the preparation will be useful as a standard for tPA antigen assays.

Table 1 Summary of potencies
t-PA Antigen (ng/ml) [Expected plasma quoted values <10, or 1 – 12, or 95% <9 ng/ml; 94/730 25 ng/ml]

*One statistical outlier removed

With the same aim, PAI-1 antigen was compared with 92/651, SSC #2 and #3. The mean value in 92/654 was 73.5 ng/ml, though the expected value was 250 ng/ml. SSC #2 and #3 gave consistent and reasonable results. Labs 1 and 6, using the same kit, gave consistently low results for all samples (see table 2). Results from this small study do not allow us to assign a potency for PAI-1 antigen in ng/ml in plasma with confidence.

Table 2 Summary of results
Summary of PAI-1 Antigen after removing labs 1 and 6 [ Normal: 4 -43 ng/ml]

In the study to calibrate the SSC plasmas for PAI-1 activity, laboratories split into 2 groups reporting activity in ng/ml or inhibitor units/ml. These results could not be easily compared. Both groups showed large variability, especially the ng/ml group. Further work is needed before standardization of PAI-1 activity can be attempted.

During the discussion with the audience it was generally agreed that some problems with 92/654 may be due to the recombinant nature of the PAI-1 used to spike the plasma in this International Standard. It was suggested that a new standard of plasma containing a high level of (native) PAI-1 would be useful, especially for the diagnosis of elevated PAI-1 in patient samples.

Standardization of methodology for plasminogen activator activity.

C Longstaff gave a final report on the methodology study for determination of thrombolytic potency of plasminogen activators. The aim of the study was to investigate the feasibility of a proposed assay for the determination of absolute enzyme activity is SI units (pM/s plasmin production), as an alternative to International Units (IU). This approach would fulfill some recommendations for assay methods and allows different thrombolytics to be compared, which is not the case currently as IU are different for current plasminogen activator IS. A very detailed assay protocol was agreed before the study and all participants were provided with all critical reagents to measure streptokinase, tPA and uPA activity. In spite of this there was a wide spread of final results returned for absolute enzyme activity. Means of all assay results gave satisfactory dose response curves for all activators. Some improvement in variability (expressed as %GCV) was observed if streptokinase was used as a standard for the other plasminogen activators, but variability was still quite large, around 35%. The conclusions of the study were that absolute determination of enzyme activity is very difficult. It is also problematic to provide a new method to laboratories and expect them to perform the method well without training. The traditional method of calibrating International Standards by recruiting as many labs as possible and allowing them to use familiar methods is a pragmatic approach that will be difficult to change in favor of a specific assay if this involves much complexity.

Standardization problems with recombinant streptokinase

Colin Longstaff reported on observations with potency determinations of native and recombinant streptokinase using the current 3 rd International Standard (IS) for Streptokinase. The international standards for streptokinase have been used successfully for more than 40 years and international collaborative studies show excellent agreement between consecutive standards. Many companies around the world manufacture streptokinase to treat the global epidemic of cardiovascular disease seen in developing countries and in Eastern Europe. Some of these products are recombinant (rec) and some of these rec products do not behave well against the 3 rd IS for streptokinase in different assay formats. For example, using two standardized methods, (1) without fibrin and (2) with fibrin the potency of two recombinant streptokinase products available in India was measured relative to the 3 rd IS for Streptokinase. The ratio of potency for normal, native streptokinase is 1.0 comparing these assay methods, but for one rec streptokinase the ratio was 0.3 and for a second rec streptokinase was 1.5. Thus the inclusion of fibrin in the assay can dramatically affect the potency of the product and the dose given to patients. This is especially important since different pharmacopeias recommend assay methods, without fibrin (eg European and British Pharmacopoeia) or with fibrin (Indian Pharmacopoeia). Changing the assay format could result in lethal doses of streptokinase being given to patients. Further work is needed to determine the cause and possible solutions to these problems, which may include provision of additional standards or pharmacopoeial methods.

D-dimer

 Assay of D-dimer in multicenter trial I. Jennings

The measurement of D-dimer is used for diagnosis of DIC, monitoring the treatment of DIC, diagnosis of DVT/PE and the prediction of recurrence The multicenter trial for the assay of D-dimer was performed (431 centers participated). For this trial, 4 D-dimer samples were prepared; high D-dimer level (pool 1: ~1000ng/ml), low D-dimer level (pool 2: ~300 ng/ml), mixture of equal quantities of pool 1 and pool 2 (pool 3) and pool of plasma with ~300 ng/ml of D-dimmer (pool 4).

This trail demonstrated the large variation in results between reagent groups. The comparison of findings in different centers was difficult. The calibration curve was made by plotting overall median D-dimer level versus individual laboratory D-dimer levels for low, mix and high sample. Although the calibration curves could be constructed, comparison of results only valid where linearity was good and data was not extrapolated. Difference may exist between different samples that preclude valid comparison between methods.

Harmonization of D-Dimer assays: Results of the FACT4 study C. Dempfle

At first, Dr. Dempfle presented the consensus of D-Dimer antigen. Monoclonal antibodies used in D-dimer antigen assays should display minimal cross-reactivity with fibrinogen, (monomeric) fragment D from fibrinogen, non-crosslinked fibrin, other proteolytic fragments of fibrinogen or non-crosslinked fibrin.

In clinical plasma samples, fibrin fragment D-dimer represents only a portion of the total D-dimer antigen. A major portion of D-dimer antigen in clinical plasma sample has a higher molecular weight than fibrin fragment D-dimer. Based on these findings, fibrin fragment D-dimer is not a primary candidate for a calibrator. Distribution of fibrin compounds and matrix should closely match with clinical plasma samples. Since D-dimer antigen is not a homogeneous and monoclonal antibodies against D-dimer antigen react with different antigenic sites of the D-dimer antigen structure, a primary reference standard cannot be formulated. Therefore, pooled patient plasma samples could be used for harmonization of D-dimer antigen assays. The pools should contain a variety of clinical plasma samples including the target groups DVT, PE and DIC. Different responses of D-dimer antigen assasys in different concentration ranges preclude the use of simple conversion factors.

For D-dimer assay, a common calibrator should be need. Fibrin fragment D-dimer is not suitable for D-dimer assay. One standard is sufficient if the standard is a pooled plasma with a large number of donors. It is necessary to evaluate assays individually for determination of the cut off for DVT exclusion. Dr. Dempfle proposed the preparation and validation of a lyophilized reference preparation based on pooled human plasma. Dr. Dempfle also recommended the definition and validation of the procedure for calibration of the reference preparation, and the clinical evaluation of the calibrator.

Problems with the D-Dimer Assay J. Olson

In reporting the D-Dimer level, two units are usually used. One is D-Dimer Units (D-DU), another one is fibrinogen equivalent unit (FEU). Surprisingly, 1 ng/ml (D-DU) of D-Dimer equals to 2 ng/ml (FEU) of D-Dimer. Data from a study in the U.S.A. indicated nine commercial kits for D-Dimer are available, six recommends reporting FEU and three recommends reporting D-DU. Among all methods for reporting the quantitative D-Dimer, there is wide variation in the type and magnitude of units reported.

Nearly 40% of laboratories are converting the analyzed units and reporting in unit other than those recommended by the manufacturer. Many laboratories are unclear about which type of units they are reporting. This is the major problem with the D-dimer assay performance in the U.S.A.

General discussion and activity for next term

Since we have still several questions and issues to be clarified, we would like to keep the studies on TAFI, Standard of fibrinolytic factors and D-Dimer. Requests to the audience were made to suggest new issues that might be appropriate for the committee.