Rebecca Jones asks, “When patients go to general surgery, how much platelet inhibition is acceptable for procedures to be performed? We have been doing a calculation giving surgeons percent inhibition, but are being asked to switch to the PRU.
George presumed Rebecca uses the VerifyNow P2Y12 cartridge for monitoring Plavix before and after surgery and reporting the result in PRUs. Your surgeons should understand result is qualitative, indicating only that the assay is positive or negative for Plavix’s antiplatelet effect. They should not interpret either aggregometry results or the PRU results as indicating a degree of antiplatelet effect, they should just think “all or nothing.” Reference given.
ASCLS Consumer Web Forum, from a primary care physician: “Does delayed fibrinolysis in a known FVL mutation patient reduce the value of a negative D-dimer test?”
George’s initial response: Researchers (referenced) found, “Patients with FVL mutation displayed higher levels of D-dimer and fibrinogen-fibrin degradation products in plasma after 24 hours. Patients with FVL generate higher levels of soluble fibrin, which may serve as cofactor in tissue plasminogen activator-induced plasminogen activation, leading to a more sustained activation of fibrinolysis with production of more fibrinogen- and fibrin-degradation products.” George concluded that a negative D-dimer in a FVL patient remains effective in ruling out VTE.
The questioner wrote in follow-up: “I have more than a professional interest in FVL, as I am a 70-YO FVL heterozygous patient with a past history of 3 major DVTs and 1 PE 28 years ago. I have been on Coumadin for 28 years. Two recent episodes of potential DVTs were ruled out with negative D-dimer quantitative tests. However I now absolutely have a moderate superficial lower leg thrombophlebitis, and 2 negative D-dimer tests performed on day 2 and 6, with another planned for day 11. I realize that both my age (old), and my Coumadin use may reduce the value of the D-dimer test. FVL persons have an exaggerated risk of DVT and PE. Therefore we make up a larger percentage of persons who may need and receive D-dimer testing than the percentage of FVL in the population. Thank you for continuing your search for the answer to my question.”
A PubMed search led to this statement: “In a geriatric population, conventional ELISA D-dimer is a good marker to exclude PE but, due to the co-morbid conditions, only a few patients presented with D-dimer values less than 500 ng/mL.”
George received four expert follow-up comments:
From Dr Emmanuel J Favaloro: I would suspect it feasible that it would slow the development, but a positive D-dimer would eventuate, and in the time course of development, a DVT should show up as a positive D-dimer in FVL positive patients to a similar level as that in FVL negative patients. I suspect it would be difficult to show a statistical difference in D-dimer levels post DVT/PE in FVL positive vs negative patients—the number of patients required would be far too large.
From Dave McGlasson: One of the issues I could see as a problem with the D-dimer comparison is the issue of age and what is considered normal with different age groups and whether normal specimens are collected on in-patient populations and out-patient subjects.
From Dorothy M. (Adcock) Funk, MD: I wonder if the clot burden in a superficial thrombophlebitis is sufficient to always cause the D-dimer to elevate above a given cut off? I think not. This of course also depends on the age of the clot compared to when testing was performed. Further, D-dimer is not FDA approved to exclude superficial thrombosis.
From John Olson, MD: Production of D-dimer from fresh clots falls off after about 2 weeks so the duration of the thrombus is also a consideration. Whether superficial phlebitis will elevate the D-dimer is a function of the amount of fresh clot there is to lyse, thus I expect the dimer response would be variable.
July 8 comment from the questioner: Thank you and your four experts for your thoughtful comments. At day 11 the D-Dimer Quantitative remains negative, and the superficial thrombosis (estimated volume 10-20 cc) has not propagated (INR 1.8 on day one, 3.1 on day 11), and is resolving on appropriate care (soaks, increased Coumadin, and reasonable activities). I continue to question the value of a negative D-dimer Quantitative test in FVL patients.
And finally, a July 12 comment from “Jlow:” The use of D-dimer assays to exclude VTE is well established, although this is not the case with all D-dimer assays since many manufacturers seem to rely on comparisons with the “gold standard” Vidas assay. The 2011 CLSI guideline H59-A is helpful in determining whether your D-dimer assay can be used for ‘exclusion of VTE” together with a clinical risk score or is an “aid to diagnosis of VTE.” There is nowhere near the same amount of data for D-dimer to be used to rule out superficial thrombosis because the appropriate cutoffs have not been established.
From Maria E. Martinez, Orlando: Our laboratory would like to start offering the Bethesda titer. We currently do special coagulation tests for several hospitals and clinics. I was wondering if there is a procedure you are willing to share.
George replies there is a generalized protocol in the Esoterix Coagulation Handbook, 2002. Here’s their protocol:
“Serial dilutions are made of patient plasma with veronal buffered saline, then mixed 1:1 with normal plasma containing 100% factor VIII activity and are then incubated for 2 hours. A partial thromboplastin time (PTT)-based factor assay using factor-depleted plasma is then performed on the incubated mixtures. Results are compared to those of incubated normal plasma. One Bethesda unit is defined as the inverse of the dilution that neutralizes 0.5 (50%) of the factor being assayed.”
For a specific step-wise protocol, you may wish to contact the company whose coagulometer you are using, as the methods are instrument-specific. They are likely to provide the method in their manual or accompanying materials, and will provide assistance in setting up your assay.
In a July 10 comment, Herb Crown, St Louis U, adds: The factor VIII Bethesda assay is certainly doable in your laboratory. What is needed is attention to detail and strong pipetting skills. I would spend some time looking over this link: http://www.practical-haemostasis.com/Factor%20Assays/inhibitor_assays.html. Once you have a handle on the principle, you may wish to look this link over: http://www.slm-hematology.com/uploads/media/Factor_VIII_Inhibitor_Assays_Methodology__Shortcomings.pdf.
From “Cristina:” Why should lipemic and hemolyzed coagulation sample for D-dimer be rejected? Is it only the D-dimer that is susceptible?
George suggests that lipemia is likely to cloud the final reaction solution, interfering with the results. On all instruments, the quantitative D-dimer is a photo-optical assay, thus you cannot trust the results when lipemia is present. Visible hemolysis not only may interfere in a photo-optical reading, it also implies the ex vivo activation of platelets and coagulation factors, rendering coagulation results invalid in electro-mechanical instruments as well as in photo-optical instruments.
In a July 12 comment, “Jlow” writes, “Plus red cells contain procoagulant phospholipid as well.”
July 18, Pam Owens: I also recommend the article George cited in an earlier post about heme binding factor VIII.
August 1, “Göran”: It is the combination of a photo-optical instrument with the relatively low concentration of D-dimer that creates a problem. In a normal sample, the D-dimer concentration is typically 50 ng/mL. For the immunoturbidimetric method that most D-dimer assays rely on, this is relatively low. For this reason, to get a response, a fairly large volume of sample is needed in the cuvette, so a large amount of lipids, hemoglobin or bilirubin will also be introduced if present. These substances absorb light at, for instance 405 nm, which is the wavelength traditionally used in coagulation instruments. Newer instrument typically also have higher wavelengths, 700 or even 800 nm. By using the newer D-dimer methods that are adapted for these higher wavelengths, many of the problems that come with lipids are avoided. The ex vivo activation of platelets and coagulation factors does not affect the D-dimer results, as such, as George has pointed out in a different thread.
From “japgi.” If a patient with an uncharacterized bleeding disorder has been transfused plasma or CRYO in an emergency setting, how soon afterwards can one do a coagulation screen to avoid the results from getting compromised by the transfused factors?
George answers that it depends upon the half-lives of each of the coagulation factors. For instance, factor VII deteriorates to half its original activity in 6 hours, factor VIII in 12, and factor IX’s half-life is 24 hours. Of course, you also need to know the starting point, which involves some guesswork based upon original factor activities available within the product and time of infusion. The most you can achieve with plasma or CRYO, without creating transfusion-associated circulatory overload (TACO), is 30% activity, so that could be considered your starting point. If your goal is to simply establish which of the factors is deficient, this maximum makes the task relatively easy.
July 25, “Jlow” asks, is it to see if the transfused product is correcting the abnormality in vitro? Do the coags straightway. Or is to determine what the patient’s baseline factor levels are, in which case you would assay when patient has recovered from the current bleeding crisis and whatever might have precipitated it. The latter would be performed much later than the time suggested by the factors’ half lives.
Our recent series of discussions and comments on the efficacy of D-dimer results in superficial thrombophlebitis for FVL patients has pointed us to a discussion of the value of screening unselected populations for APCR. Dr. Emmanual Favaloro (referenced) provides the American College of Medical Genetics current guidelines for directed FVL screens, and reminds readers that the guidelines do not recommend generalized screening. His article also records a trend towards increased FVL testing with a parallel drop in the percentage of positive results, suggesting that ordering patterns are becoming indiscriminate. His article is framed as a debate in Blood Transfusion, and will appear in the August 2012 issue (references).
Additionally, Dr. F employs a DRVVT-based assay for APCR, Siemens ProC Ac R, whose clinical characteristic he describes favorably (reference). Is anyone else using the DRVVT-based APCR assay?
Michael Suter monitored Plavix therapy using the VerifyNow P2Y12 assay kit. Accumetrics is in the process of changing form the reporting of percent inhibition to PRUs.
George spoke with Debra Feinberg, Stan Arachniewicz, and Dr. Jackie Coleman, of Accumetrics. They confirmed that since most recent clinical outcomes trials that employed the Accumetrics VerifyNow P2Y12 cartridge based their results on PRUs, it is necessary to change the reporting system. They have notified all their clients that they will support the current software only through August 27, 2012, and recommend local institutional make their changes as soon as possible. The cartridge design remains unchanged for the present, and the company recommends an upper PRU reference limit of 208 units.
A puzzler from Pam Owens: We recently encountered a patient on warfarin whose INR is 2.22, protein C and S are in the 30–40% range, but her factors VII and IX are normal. What could cause this?
George posed this to colleagues while attending the ASCLS/AACC annual meeting. Nobody had a really good answer, so he’s taking a “scientific wild guess.” Assuming factors II and X were reduced, and that V and VIII were normal, the patient may have recently discontinued warfarin, so that VII and IX, with half-lives of 6 hours and 24 hours, respectively, had recovered whereas II and X, half-lives of 60 hours and 48–52 hours had not. The low II would account for the INR, which should return to normal as their activity levels reach normal. The weakness in George’s argument is that protein C, whose half-life is also around 6 hours, should also have returned to normal!
A July 25 comment from Pam: The protein S antigen was decreased as expected, but her protein C antigen was normal. Makes one think she may have a type 2 deficiency (that would explain that pesky half life issue), but we are recommending the clinician draw no conclusions while she is on warfarin.
July 26: George thinks these results help to strengthen his guess that the warfarin had been recently discontinued, as you would anticipate that protein C would recover at nearly the same rate as factor VII.
From Teresa Lovejoy, Children’s of Alabama: Why is it not important to double-spin D-dimer specimens for PPP when there’s a delay in testing? Is this because D-dimers are not affected by labile factors?
George mentioned that Children’s of Alabama is moving into a magnificent new building that now dominates the Birmingham’s medical center skyline. The purpose for PPP is to ensure no platelet secretions affect coagulation test results. Platelets may become activated in vitro and secrete, for example platelet factor 4, which neutralizes heparin and artifactually shortens the PTT. Platelets also secrete factor V, VIII, and VWF, and release plasma membrane phospholipids that interfere with LA testing. Platelet materials become especially critical when plasma is frozen, as cells become ruptured. The D-dimer concentration seems to be unaffected by platelets, as D-dimer is a product of fibrin crosslinking and fibrinolysis. Consequently, it is not necessary to prepare PPP for a D-dimer assay; however, there is no harm in it, either. Since most labs prepare PPP for all coagulation testing, most D-dimer assays probably are performed using PPP.
From Thomas Exner, Owner, HAEMATEX RESEARCH, Sydney, Australia: “I read your response to a question about LA mixing tests. I’m a little disappointed that you still recommend prolonged incubation times. Most recommendations now are that LA mixes should be tested immediately. There is no logical reason for time-dependent LA testing since they work by interfering with the phospholipid added in the reagent at the time of testing. Yes, you may see prolongation of an inhibitor effect on PTTs with some test plasmas but it is an artifact due to pH increase. The same problem was noted in FVIII Bethesda assays a while ago and fixed by the Nijmejen people by including a pH buffer. Uncovered unbuffered citrated plasmas drift to pH above 8.5 over an hour at 37ºC and this knocks out a lot of the factor V. Time dependent inhibitors are more typical for FVIII antibodies and should be looked for for that purpose, not for LA. Sorry to disagree with your recommendations in this case.
George thanks Tom for this important update. Several laboratories incubate their mixing studies for the purpose of detecting LAs as well as specific inhibitors such as factor VIII inhibitors, this is an opportunity to disseminate this information.
July 27 from Herb Crown, “SLU.” The PTT mixing study suffers from a number of weaknesses. It seems simple enough, but it does not have a universally recognized standardized procedure and the method of interpretation varies widely. Few laboratories have reference ranges and generally use some type of interpretation that refers back to the patient’s PTT. The PTT mixing study has four uses; a) to aid in the detection of LA, b) to aid in the detection of a factor inhibitor c) to aid in the detection of a factor deficiency and d) to provide direction when working up a patient sample that unexpectedly presents a prolonged PTT.
For a known LA, the PTT mixing study is of limited utility to further define the abnormality. The PTT mixing study is not necessarily sensitive to LA. In our laboratory, about 50% of our LA patients show a normal PTT and a normal PTT mixing study. There are better tests to determine whether or not a patient has an LA.
For a known factor deficiency, a mixing study incorporating an incubation step should be performed to rule out an inhibitor. Most often, when a sample shows a normal mixing study at immediate mix and prolongation after a 60 minute incubation at 37º a factor inhibitor is generally suspected. A normal PTT mix before and after incubation would indicate a factor deficiency.
More often, we see the greatest value of the PTT mixing study in the context of helping to explain a prolonged PTT. After ruling out the presence of heparin we will perform the PTT mixing study with a 60 minute incubation. This will drive the direction of the rest of the work up. If the mixing study is normal before and after incubation, we would work up factors. If the mixing study is abnormally prolonged before and after incubation, we would work up the specimen for LA. If the mixing study is normal before incubation and prolonged after incubation we would look for a factor inhibitor followed with a Bethesda titer.
The PTT mixing study is not 100% specific or 100% sensitive. It is one part of a complete set of tools we use. It is not the “end all” of all things coag. As an isolated test, on a patient with a known disorder, the PTT mixing study is of limited utility. As part of a prolonged PTT workup, the PTT mixing study, with incubation, is a valuable tool to help guide the workup.
This is a basic understanding of the PTT mixing study. There are certainly exceptions to the above discussion. We see those exceptions on a routine basis. One thing predictable about coag is its unpredictability.
From “Maria:” Literature states that the Rosner Index must be validated with each new lot of PTT or PT reagent. Do we also need to validate each new lot of CaCl2? Also, what would be the way to validate the Rosner Index?
George knows of no direct method for validating the Rosner index or any parallel means for establishing correction, instead, you need simply validate the underlying PT and PTT methods. NASCOLA provides a proficiency testing module, and mixing study guidelines are available from the ISTH Subcommittee on Lupus Anticoagulant and Phospholipid-Dependent Antibodies, chaired by Thomas Ortel, MD, of Duke University Medical Center. These may help you to develop your mixing study protocols.
The July issue of Clinical Chemistry presented a well-written case of combined factor V and VIII deficiency (referenced). George noted the recommendation for plasma to replace factor V and asked: As an alternative to FFP, would you recommend platelet concentrate as a means for delivering adequate doses of factor V, in view of the high concentration of factor V in platelets?
From the author: “To our knowledge, platelet transfusions are a standard of care for those who have factor V inhibitors because indeed they do carry a lot of FV in the alpha granules and have an added benefit of protecting it from plasma-based inhibitors. In theory, inhibitors have much less time to react once platelets release factor V at the location of clotting. The importance of platelet factor V is demonstrated by bleeding in patients with Quebec abnormality where platelet factor V gets proteolyzed in the alpha granules. As for treating FV deficiency with platelets, one has to balance the risk of developing HLA antibodies, cost, and efficacy in comparison with FFP. It appears that FFP is still standard of care for FV deficiency without inhibitors. It is also likely the combined FVIII and FV deficiency patients would be less likely to develop inhibitors as they have been exposed to autologous normal FVIII and FV.”
From Donna Lawler, University of Wisconsin: Do you have any references concerning PT/INR reagents? We use a low ISI thromboplastin (close to 1.0) with good factor sensitivities but we are getting pressure from our lab director to switch to a different reagent because the peer group in the CAP proficiency testing is small. What are your thoughts?
George references a recent publication describing thromboplastin standardization. This article describes lot-to-lot reagent validation using a set of reference plasmas, an approach that is mostly unavailable in the US because of FDA restrictions. He suggests that if the reagent performs well, matches well with instrumentation, and meets internal validation requirements, there is no reason to change, just because it is not the reagent with the biggest market share. If there is any reason to be concerned about you reagent’s accuracy, you can also check it by comparing to the chromogenic factor X assay (reference). In fact, there are a few people around who advocate for routine use of the chromogenic factor X in place of the PT, as it is less prone to interference.