LAUSR

Direct oral anticoagulants (DOACs) have predictable pharmacokinetics and are licensed for stroke prevention in non-valvular atrial fibrillation (SPAF) and the treatment and secondary prevention of venous thromboembolism (VTE) without the need for monitoring. However there are occasions when it is useful to measure their effect and it is important that laboratories know the sensitivity of the screening reagents they use (Kitchen et al, 2014) when estimating the DOAC effect, as this may help with management (Baglin, 2013). Rivaroxaban concentrations can be calculated by measuring the anti-Xa effect of the drug in an assay calibrated using appropriate standards. Only 68 participants of the United Kingdom National External Quality Assessment Service (UKNEQAS) for Blood Coagulation external quality assurance scheme returned results for anti-Xa assays for rivaroxaban in September 2015, compared to 866 for prothrombin time (PT) and 889 for activated partial thromboplastin time (APTT) (UKNEQAS for Blood Coagulation, unpublished data), suggesting that these assays are not widely available. The effects of rivaroxaban on laboratory tests of haemostasis have been published, but most of the data are from spiking studies where drug is added to plasma/blood in vitro (Samama et al, 2010; Asmis et al, 2012), and some are from studies that measured ex vivo samples from healthy volunteers who have been given the drug (Asmis et al, 2012; Douxfils et al, 2013). The International Society on Thrombosis and Haemostasis (Baglin et al, 2013) and the British Committee for Standards in Haematology (Baglin et al, 2012) recommend that laboratories establish the sensitivity of screening reagents using commercially available calibrators, but this practice may underestimate drug concentration (Gosselin et al, 2015). In such studies, PT is considered to be more sensitive to rivaroxaban than APTT (Baglin, 2013). Innovin (Siemens Dade, Marburg, Germany) is the second most commonly used reagent by UKNEQAS participants, and appears to be insensitive to rivaroxaban when compared to other thromboplastins (Samama et al, 2010; Douxfils et al, 2013), emphasising the importance of studying the effect of rivaroxaban using local reagents. We therefore looked at the correlation between routine coagulation tests, using Innovin for PT and Siemens Actin FS (Siemens, Marburg, Germany) for APTT, and rivaroxaban concentrations in patients taking this DOAC for either SPAF or treatment/secondary prevention of VTE. A total of 124 consecutive plasma samples received in the laboratory between June 2012 and December 2015 from 91 patients taking rivaroxaban were frozen at 80°C and rapidly thawed at 37°C before analysis. All tests were performed on a Sysmex CS2100i coagulation analyser (Sysmex UK, Milton Keynes, UK). PT was measured in 123 samples and APTT was measured in 122 samples. Rivaroxaban concentration was measured using the BIOPHEN DiXal assay (Hyphen Biomed, Neuville-sur-Oise, France). Patients on rivaroxaban 10 mg od are predicted to have peak concentrations of 125 ng/ml (95% confidence interval [CI]: 91–195), and those on 20 mg od are predicted to have peak concentrations of 223 ng/ml (95%CI: 160–360) (Mueck et al, 2008) (Table I). In patients with rivaroxaban below 125 ng/ml, 65/67 (97%) had normal PT, 61/66 had normal APTT and 57/64 (89%) had normal PT and APTT. In patients with rivaroxaban below 223 ng/ml, 79/88 (90%) had normal PT, 73/87 (84%) had normal APTT, and 68/86 (79%) had normal PT