Owing to our web site crash in November, all November posts were reposted in December, which turned out to be an uncharacteristically active month. Please return to the site if you have any comments to append to our posts, it is now safe to use.
Dave Easler contacted participant Dave McGlasson through the Medlab_L list to ask if he recommends to not use a POC INR method for patients bridging from LMWH to Coumadin. Dave gained the impression that the recommended method for testing INRs in this class of patients is with a laboratory analyzer rather than POC.
McGlasson answered, when bridging from heparin or a DTI to Coumadin the recommended test is the chromogenic factor X (ten, CFX). It is impervious to the other anticoagulants. POC PT/INR assays are clottable assays that can be affected by heparin or DTIs. Measure the CFX until it gets to be <30% activity. The CFX is just as easy to perform as an anti-Xa assay.
Julie Richards asked if there is any utility to doing bleeding times to assess platelet function in uremic patients/dialysis patients?
In George’s lost November 23 post he reluctantly agreed that the BT could provide some diagnostic or monitoring support in uremia. In uremia, a series of vaguely-defined toxins coat or modify the platelet membrane and suppress adhesion. The reduced platelet function may be indicated by a prolonged BT, which becomes shortened upon treatment with desmopressin (DDAVP) or renal dialysis. Soyoral YU, Demir C, Begenik H, et al. Skin bleeding time for the evaluation of uremic platelet dysfunction and effect of dialysis. Clin Appl Thromb Hemost. 201218:185–8 concludes that the BT is shortened in dialysis, however it provides no evidence that the BT actually predicts bleeding.
Several studies in the 1980s and 90s, for instance, Lind SE. The bleeding time does not predict surgical bleeding. Blood 1991;77; 2547–62, confirm that the BT generates only a 50% PPI for bleeding, the same as a coin toss, and on that basis most experts have recommended the assay be discontinued. The in vitro alternatives include platelet aggregometry, the Siemens PFA-100, the Accumetrics Ultegra, or the Thromboelastograph, all of which offer better predictive capabilities and reproducibility than the BT.
While platelet function testing offers clinical value in uremia, it provides little support subsequent to renal dialysis because patient platelets become activated as they are exposed to extracorporeal surfaces. Researchers may use flow cytometry to illustrate platelet function but there is presently no clinical application.
To generalize, George recommended one of the in vitro platelet function tests to predict uremic bleeding, avoiding the template bleeding time, a test that now belongs to the ages.
Shirlyn McKenzie and George developed this summary for the clinical applications of anticoagulant therapy:
- PCI: use IV anti-GP IIb/IIIa agent: tirofiban, abciximab, or eptifibatide. Follow up with UFH acutely, switch to anti-platelet drugs aspirin, 81 mg/day and clopidogrel 75 mg/day. Continue clopidogrel 6 months to 2 years, aspirin indefinitely. Fibrinolytic therapy raises bleeding risk during the procedure
- CABG: use high-dose UFH during surgery, monitor with the ACT, then switch to anti-platelet drugs as for PCI, same dose and duration.
- Chronic atrial fibrillation: Coumadin, start at 5 mg/d (2 mg/d if over 70), then adjust dosage to INR, use indefinitely to reduce risk of stroke. Substitute new oral anticoagulants dabigatran (Pradaxa), rivaroxaban (Xarelto), or apixaban (Eliquis) if insurance pays.
- DVT/PE prevention after orthopedic or neurological surgery or other situations associated with high risk of thrombosis such as medical condition with prolonged immobilization: Coumadin for three to six months or LMWH 14 days. Rivaroxaban is recently cleared for this purpose.
- DVT/PE treatment: Treat acutely with UFH while inpatient, switch to Coumadin at least six months, longer if risks remain high, or if not tolerated, LMWH. Rivaroxaban is now cleared for therapy.
- Holbrook A, Schulman S, Witt DM, et al. Evidence-based management of anticoagulant therapy: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines. Chest. 2012;141(2 Suppl):152S–84S. Review.
Tania Puro asks to explain why the PTT and PT are normal in dysfibrinogenemia.
George answers that dysfibrinogenemia is a qualitative or functional fibrinogen deficiency associated with moderate to severe liver disease. The liver produces normal or even elevated fibrinogen levels, but the molecule is coated with excess sialic acid and functions poorly. Liver disease is associated with soft tissue (anatomic) bleeding owing to deficiencies of several coagulation factors and to dysfibrinogenemia. The thrombin and reptilase clotting time are prolonged, and clot-based Clauss fibrinogen assay results are reduced. Because dysfibrinogenemia is a functional disorder, fibrinogen assays that rely on nephelometry, such as estimates provided by coagulometers in the performance of a PT, may be normal or elevated.
PT and PTT reagents are moderately insensitive to hypofibrinogenemia and dysfibrinogenemia, and only become prolonged when fibrinogen concentration or functional levels are below 100 mg/dL, which explains why the PT and PTT may be normal in dysfibrinogenemia.
Mohamed Bashir asked, what is the guideline for thrombophilia testing, how to test AT, PC, and PS?
George responded that thrombophilia profiles are developed locally to meet population needs, however most include AT, PC, and PS activity, APCR, prothrombin G20120A mutation, and LA testing. Some include ACA, anti-b-2-GP 1, fasting homocysteine, and coagulation factor VIII.
If the patient has experienced a recent thrombotic event or is currently on Coumadin, the tests that remain reliable are the FVL mutation, which substitutes for the APCR, prothrombin G20210A, ACA, anti-b-2-GP 1, and fasting homocysteine. The AT, PC, PS, APCR, and LA are reliable only when the subject is not receiving Coumadin and has not had a recent thrombotic event.
Owing to the false positive rates of most assays, we avoid screening healthy individuals. Thrombophilia profiles are chosen for patients who have experienced an unprovoked thrombotic event, an event in an unusual location or before age 60, or who have first-degree relatives with thrombotic risk factors. A positive APCR ratio is confirmed using the FVL mutation, and low AT, PC, or PS values are repeated after 12 weeks for confirmation. Low AT, PC, and PS activity assays are then followed up using AT, PC, and PS antigen assays to distinguish qualitative from quantitative deficiencies.
Ann McConnell asked, why are factor activity assays reported in the range of 50%–150%? “I am having a difficult time wrapping my mind around the concept that something can be more than 100% ‘active.’”
George responded that in fact the correct way to express factor activity is in international units, with 1 IU = the mean factor activity of normal plasma. Thus 150% would be 1.5 units. From a pure mathematical standpoint, percent is an incorrect expression, however, we’ve been using it to express factor activity since the mists of time, so, given normal human inertia, we probably won’t change!
Mahnaz Sairi asked, what is the highest limit we can report for FVIII? And how we determine this? Is it by clinical significance or by instrument linearity? Also, should we still do LA testing if both the PT and PTT are normal?
George suggests to begin with a calibrator of known factor activity such as CRYOcheck™ Normal Reference Plasma and prepare several ascending dilutions using the required buffer. Compute the anticipated activity for each dilution, then assay each dilution and prepare a linear graph of anticipated activity versus assay values. By inspection you will find where linearity is lost at the top and bottom of the curve; select the points just inboard to establish your limits.
The limits of linearity vary among instruments and reagents, however, when preparing dilutions, just as a rule of thumb, develop a range of 1%–150%. Some instruments use two curves, the second of which is designed to provide linearity at the low range, for instance, 0.1%–10%, in this case it is necessary to test linearity in both ranges by preparing and testing the appropriate calibrator dilutions. The instrument automatically selects dilutions and curve based upon the initial assay.
Specimens that are out of linearity, for instance, FVIII activity levels over 150%, are diluted, dilutions are re-assayed, and the result of the assay is multiplied by the dilution factor. High-end coagulometers automatically make the appropriate dilutions, and it is not unusual for FVIII levels to exceed 150%, as FVIII is an acute-phase reactant. Because chronically elevated FVIII is a thrombosis risk factor, laboratory directors may choose to report levels as high as 250%.
For the second question, the PTT reagents employed to monitor UFH or to screen for coagulopathies are formulated with high phospholipid concentrations, rendering them relatively insensitive to LA. These are employed to reduce the frequency of detection of transient, clinically innocent LAs. See Fritsma GA, Dembitzer FR, Randhawa A, et al. Recommendations for appropriate activated partial thromboplastin time reagent selection and utilization. Am J Clin Pathol 2012;137:904–8. If there is clinical suspicion of an LA, for instance, an unprovoked thrombotic event, it is necessary to test using low-phospholipid PTT reagents such as PTT-LA, and to test in parallel using a separate platform such as CRYOcheck™ LA Check™ (dRVVT screening reagent).
Lisa Ledford asked, Would you be willing to share your ∂-check values for PT/INR, PTT, fibrinogen, and D-dimer?
George checked with Warren Varden at UAB, who doesn’t use ∂-checks. Our special coagulation staff carefully reviews results and communicates with our inpatient and outpatient units and are able to discern when there is a change in therapy or a faulty specimen collection.
Joanna Carroll is running automated D-dimers in a clinic at a veterinary school and recently got a result of 0.00 ng/mL. Have you seen a value reported this low? Would it be better to report out the number as less than our lower limit of the reference range?
George has never seen a D-dimer value of 0.0 ng/mL, and assumes this was performed on a human. He suggests reporting as less than the lower limit; particularly as a result this low would have no clinical significance.
Added 12.7.12: Dave McGlasson has worked with a number of animal specimens, and says the monoclonal antibodies in the quantitative D-dimer kits are specific for humans and do not detect animal D-dimers.
Anti-Xa Calibrators (December 6)
Scott Miller wrote: We will soon be validating our PTT heparin therapeutic range by using a comparison between an anti-Xa assay and PTTs run with our new lot of PTT reagent. I have always understood that when calibrating an anti-Xa assay for this procedure, one should always make the calibrator by dilution of the same type of heparin used at the institution. Yet the kit we received from IL includes a box of WHO-certified calibrators. Should these be used instead of the homemade stuff?
George definitely recommends using the provided calibrators provided. Not only are they traceable to WHO international UFH and LMWH standards, they also enable you to construct a hybrid curve, one that may be used to report both UFH and LMWH results from the anti-Xa assay without have to change curves. Plus, home-brew curves tend to flatten out at the top end. George also suggests using the anti-Xa directly to report heparin results.
Kelly Townsend received a letter stating that BD Vacutainer is recalling a lot of their 2.7 mL sodium citrate tubes due to lithium heparin contamination. Has anyone seen issues with this and how are folks investigating?
Elaine Benoit sent an FDA alert regarding a heparin label change: “This label change will require manufacturers of Heparin Lock Flush Solution, USP and Heparin Sodium Injection, USP to clearly state the strength of the entire container of the medication followed by how much of the medication is in 1 milliliter (mL). These modifications will eliminate the need for health care professionals to calculate the total amount of heparin medication in a product containing more than 1 mL, thereby reducing the risk of miscalculations that may result in medication errors.”
Jasmina Ahluwalia asked if there is any advantage of an LIA assay over ELISA for (VWF:Ag).
George replied the two methods have similar sensitivity, specificity, linearity, and CV% characteristics, however the LIA is usually designed for automated applications, where the analytical time is often less than 10 minutes. The ELISA is usually performed manually, or on an automated fluid manager, and requires approximately 2.5 hours to complete. Your choice depends on how your laboratory is equipped.
In 2009 Dave McGlasson and George published McGlasson DL, Fritsma GA, Whole blood platelet aggregometry and platelet function testing. Semin Thromb Hemost 2009;35:168–80. This article, the second-most cited in the 2009 Seminars in Thrombosis and Hemostasis, reminded George of our Quick Question about the Ivy bleeding time posted in early November, and subsequently lost when our site crashed in late November. While our goal in posting the Quick Question was to reemphasize the futility of the bleeding time, research led to a lengthy and engrossing article by medical historian Dr. Patricia Spain Ward, 1931–1995, “Who Will Bell the Cat? Andrew C. Ivy and Krebiozen.” The article was published in the Bulletin of the History of Medicine 1984;58:28–52. George prepared a brief summary of the bleeding time test, Ivy’s role in its development, and his subsequent fall from grace when he promoted a false cancer cure, attached to this post.
This month’s Fritsma Factor features a number of December announcements about the efficacy and safety of dabigatran, rivaroxaban, and apixaban, which was released by the FDA on December 28 for prevention of stroke in Afib.