Here is a challenging technical question about the platelet neutralization procedure (PNP) for identification of lupus anticoagulant (LA) from Charlene Bierl MD, PhD Director of the Clinical Laboratory. Cooper University Hospital, Camden, NJ. Dr. Bieri’s question was originally addressed to Drs. Marisa B. Marques and Jill Adamski and to specialty technologists Laura Taylor and Patti Tichenor at the University of Alabama at Birmingham Hospital Department of Laboratory Medicine. Following is a lengthy dialogue:
Original question: Marisa, we are in the process of validating a number of additional coagulation tests, one of which in the PNP. During our reference range study, we were expecting our normal limit to fall within 3–5s of their baseline partial thromboplastin time (aPTT, PTT) based upon our read of the literature. However, everything fell < 2s. We even increased the sample size in case our initial study was too narrow. Our fear with setting a reference range of >2s is our false positive rate and we were wondering if we should consider making it >3 based upon our findings and the literature review. Any words of wisdom that you could provide?
Here is George’s somewhat simplistic response: Foremost, we recommend that you substitute an integrated kit for the PNP, for example, Instrumentation Laboratory’s LAC Screen and LAC Confirm. The most common approach is to team a PTT-based integrated kit, with a DRVVT kit, for example, Precision BioLogic’s CryoCheck LA-Check and LA-Sure to create a lupus anticoagulant profile, as recommended in Pengo V, Tripodi A, Reber G, et al. Update of the guidelines for lupus anticoagulant detection. J Thromb Haemostas 2009;7:1737–40.
However, if you wish to continue with the PNP, please ensure that you are using an LA-sensitive PTT reagent such as Stago’s PTT–LA, and that you are using a commercially prepared PNP reagent such as Precision BioLogic’s CryoCheck Platelet Lysate. Compare the results of the PNP mix tube to a tris buffered saline (TBS) control tube:
Control: 0.1 mL plasma + 0.1 mL TBS + 0.1 mL aPTT reagent, incubate, add 0.1 mL CaCl2 and time to clot formation
Test: 0.1 mL plasma + 0.1 mL PNP + 0.1 mL aPTT reagent, incubate, add 0.1 mL CaCl2 and time to clot formation, Compute the difference.
To establish your limit, run at least 25 plasma aliquots known to be negative for LA and compute the mean and standard deviation (SD) of the results in seconds, then set the limit at mean +3 SD. The typical mean difference is 2–3 seconds with an SD of 1.0, yielding a cut point of 5–6 seconds. Any result whose difference exceeds the +3 SD limit is considered LA-positive. Finally, confirm your limit by running some known LA-positive specimens.
I should have known it wasn’t as easy as this, here is Dr. Bieri’s subsequent reply (with permission):
George, Thank you for your thoughtful comments. Based upon your comments I suspect I did not provide enough detail in my initial question. First, I would would like to update my statement: our cutoff numbers are in reference to the TBS tube, not the original PTT.
We have IL-TOP instruments, and we are using Precision reagents. Our plan is to offer both the DRVVT Check and Sure and confirm as well as a lupus sensitive PTT followed by the PNP. We are using exactly the method you describe, but we are testing a lot more than 25 negatives!
We ran 62 samples, got a mean difference of 0.25 seconds, a SD of 0.42, making a 3 SD cutoff anything >1.5. We repeated the 62 samples with another lot of PNP reagent, and got a tighter set of values, with a mean of 0.15, SD of 0.31, for a 3 SD cutoff of >1.1. We actually then got an additional 41 samples, changed the PTTSP lot but kept the PNP lot the same, and ran the reference range analysis again and got similar results. A specialty lab in the field then suggested that we use the highest delta from our group as the cutoff, but even using this, our highest value from all of these tests (165 runs in total) was a delta of 1.9.
So our question is whether we should feel confident with our calculated 3 SD cutoff, given that our values are much lower than the cutoff from what you describe, which matches what we have seen others reporting and the textbooks. Thanks for your attention and assistance, Charlene.
Before attempting any additional analysis, I decided (with Dr. Bieri’s consent) to post the entire question and look for expert advice from the field. So, if you have any further answers, please provide help in the comment space below. I will follow up with a summary and further information. Geo.
No comments here.