Here is a message about the dilute Russell viper venom time assay (DRVVT) for lupus anticoagulant (LA) from international expert Thomas Exner, HAEMATEX, Sydney, Australia:
It was a pleasure to meet you at the International Society on Thrombosis and Haemostasis (ISTH) meeting recently. I would like to respond an interesting question raised on July 16, 2013 in your Fritsma Factor website. Best wishes.
The problem of dRVVT “confirm” results being longer than those obtained with “screen” reagent is not uncommon and occurs almost as often as the converse, even among plasma samples from normal healthy donors. However it can occur more frequently among hospital patients samples.
One reason is that residual platelets can shorten screen results because they release phospholipids that are in low concentration in screen reagents but at much higher levels in confirm reagents. LA screening tests are consequently more affected by phospholipids than confirm type tests. Even platelets at currently acceptably low levels of <10,000/ul can shorten screen results, particularly after freezing and thawing plasmas. See Paul Froom’s recent paper, “Testing for lupus anticoagulants–fresh or frozen? in Clin Chem Lab Med. 2012; 50; 1607–9. This can cause some low (weakly) positive LA plasmas that are tested fresh to appear to become LA-negative after being frozen and thawed. That is, unless the reference interval selected is appropriate for frozen samples.
There could be other reasons as well, including low levels of factor V. This effect has been noticed recently, mainly among Coumadin (warfarin) and heparin patient plasmas where it might cause “false negative” LA results if LA happened to be present. However neither Coumadin nor heparin plasmas are deemed fully acceptable for LA testing under current ISTH or tentative Clinical and Laboratory Standards Institute (CLSI) LA testing guidelines. Conversely, Coumadin plasmas are more widely noted to occasionally give “false positive” LA results. The laboratory diagnosis of LA may be strengthened by the use of a mixing test, 1:1 with normal plasma, which generally improves cut-off values.
Here is an August 7, 2013 follow-up from Dr. Exner: Please substitute factor II (prothrombin) instead of factor V deficiency as a more common cause of “confirm” results being longer than “screen” dRVVT results. Just to check, we ran some tests with 10% factor levels and the confirm result was 18% longer than the screen result with prothrombin deficient plasma but only 9% longer with FV deficiency.
I’m posting this response for friend and colleague Ali Sadeg
I’m posting this response for friend and colleague Ali Sadeghi-Khomami, PhD, Lead Researcher at Precision BioLogic Inc:
I welcome discussion on the determination of a meaningful cut-off by the users of DRVVT reagents. How often do you find patients with a lab diagnosis of a weak positive LA but with manifestation of clinical symptoms such as thrombosis or pregnancy complication?
What is the best practice in determining a meaningful cut-off? I am skeptical about simply computing mean +2 or 3 SD since this provides differentiation of normal from abnormal but not necessarily LA positive. This is especially true when you realize that that majority of data generated in the lab do not following normal distribution! I truly believe that a challenge using a set of known LA specimens, or perhaps a receiver-operating characteristic (ROC) analysis is more meaningful.