Presenter: John D. Olson, MD, PhD
Panelists: Dorothy Adcock Funk, MD, Kandice Kottke-Marchant, MD, PhD, Marisa B. Marques, MD
Moderator: George A. Fritsma
Precision BioLogic Laboratory Medicine Roundtable, June 20 and 21, 2008, Dartmouth, Nova Scotia
First, I’d like to thank Wendy Porteous, Steve Duff, Michael Scott, and all the Precision BioLogic Inc. participants who organized the June round-table discussion. We had the opportunity to share a number of productive ideas.
As past chair and a member of the College of American Pathologists (CAP) Coagulation Resource Committee my colleagues and I have had the opportunity to review and analyze assay quantitative D-Dimer proficiency data from CAP 2001 and 2004 surveys and presented that data at the ISTH in 2005. Publication is planned for this and other related D-Dimer data. However, what was learned is a bit alarming and is preliminarily presented here.
The quantitative D-dimer assay is primarily employed to evaluate or monitor patients with possible DIC and to exclude venous thromboembolism (VTE) in patients whose history and physical examination yields a low pre-test probability of disease. A D-Dimer result below a threshold, typically about 250 ng/mL when expressed as D-dimer units (DDU) or 500 ng/mL in fibrinogen equivalent units (FEU), is strong evidence against VTE and eliminates the necessity for imaging studies.
Though the majority of our laboratories provide stable automated quantitative D-dimer assays, the CAP 2001 and 2004 proficiency testing surveys revealed a mean CV of 28% across participating laboratories. Analysis of these data and a subsequent questionnaire reveal…
In the USA, laboratories report quantitative D-Dimer using immunoassay methods, some automated and some manual. Some may share an anti-D-Dimer monoclonal antibody (MAB), however manufacturers have produced several MABs of varying specificity and affinity. MAB variation may contribute to inter-laboratory variation.
D-Dimer molecular structure is not homogeneous. The range of assay MABs bind fibrin degradation products of varying molecular weight, identifying them all as containing D-Dimer.
There is no international D-Dimer calibrator.
Confusion regarding the type of units (FEU or DDU) can create erroneous laboratory conclusions. FEU and DDU results are based on the relative molecular weight of D-Dimer Vs. fibrinogen and the FEU value is double the DDU. Owing to unit confusion, some of laboratories may generate FEU results but publish a DDU threshold, while others report DDUs but use an FEU threshold. In the second instance VTE may be erroneously excluded because the threshold is too high. In fact, 39% of US laboratories report a VTE exclusion threshold that exceeds the manufacturer’s recommendation.
Many laboratories do not know how their VTE threshold was generated, how it was chosen or whether it reflects their clinical population. Most accept and publish the manufacturer’s threshold; about 30% go to original literature or perform a local analysis. Many published thresholds, distributor thresholds, or local thresholds are based upon small, non-robust samples. Some thresholds actually reside within the statistically generated reference range, confusing clinical decision-makers.
Laboratory practitioners mathematically convert final FEU or DDU results to a variety of magnitude of units: ng/mL, ug/mL, mg/dL, mg/L, etc. and a few don’t know what unit is being used. Mathematical errors may contribute to consistent reporting errors. In some cases laboratories compute by hand, in others, conversion formulas are electronically inserted in laboratory or hospital information system.
Some laboratories attempt to exclude VTE using semiquantitative assays, for instance, card-based latex immunoassays. These lack sensitivity and are only designed to diagnose disseminated intravascular coagulation (DIC).
There’s a separate problem not directly associated with the CAP surveys. The quantitative D-dimer assay has no valid DIC diagnostic threshold. DIC generally generates marked elevations, little evidence is available to physicians whose patients results come back at, for instance, 1000 to 2000 ug/mL FEUs. Clinicians may be unaware that D-Dimers rise in any inflammatory condition and levels like these are common among inpatients.
The D-Dimer assay reporting potentially creates a missed-diagnosis crisis. There is a need for an international D-Dimer standard to normalize assay results, for a single reporting convention, and provide correct electronic formulas that enable laboratories to report consistent results. Laboratories are cautioned to immediately review their reporting method, mathematical formulas, and VTE threshold value.