PB1137 - Parallelism in PT-based one-stage extrinsic factor assays; is slope ratio or %CV a better indicator?

Background: Non-parallelism in clotting factor assays can lead to erroneous results. Parallelism of test sample against a calibration curve can be assessed by analysing multiple dilutions of sample and calculating either coefficient of variance (%CV) or slope ratio (SR).

Aims: This study compared %CV and SR as an assessment of parallelism in PT-based one-stage factor assays in different patient groups using the new generation of Sysmex haemostasis instrumentation.

Methods: FII, FV and FX were performed with Innovin and Thromborel S (THS) PT reagents (Siemens) and Siemens deficient plasmas on a Sysmex CN6500 analyser. Citrated plasma from normal donors (n=10), patients with congenital (n=2-5) or acquired (n=14-17) deficiencies and patients on oral anticoagulants including warfarin (n=5, INR 4.0->20), argatroban (n=5, 0.31-1.43 µg/mL) and rivaroxaban (n=6, 155-635 ng/mL) were analysed. %CV of < 15% and SR 0.85-1.15 were considered parallel.

Results: Median normal SR ranged from 0.92-0.97 Innovin and 0.91-1.02 with THS. Median normal %CV ranged from 0.92-5.55 with Innovin and 1.25-4.45 with THS.
SR < 0.85 or >1.15 were calculated for FII in 13/35 Innovin and 9/35 THS; for FV in 6/35 Innovin and 7/35 THS; FX 5/36 Innovin and 6/36 THS (figure 1).
%CV >15% were calculated for FII in 6/35 Innovin and 2/35 THS; for FV 1/35 Innovin and 3/35 THS, for FX 8/36 Innovin and 5/36 THS (figure 2).
Mann-Whitney t-tests calculated statistically significant differences between normal donors and argatroban for FII and FV (SR and %CV); rivaroxaban for FX (SR for Innovin and THS, %CV for THS).

Conclusion(s): There was variation in parallelism between PT reagents; however, the use of SR 0.85-1.15 was a better indicator of parallelism than %CV of < 15%. Non-parallelism was most pronounced in samples containing argatroban or rivaroxaban.