More on normal prothrombin times in the presence of therapeutic levels of rivaroxaban – early experience from King's College Hospital
We read with interest the letter by van Veen et al (2013) describing their experience on the performance of the Innovin® and Thromborel S® prothrombin time assays in a 74-year-old patient with renal impairment who was commenced on rivaroxaban for the treatment of deep vein thrombosis. It was interesting to observe that with both assays, the prothrombin time (PT) was not prolonged, despite therapeutic concentrations of rivaroxaban, and we thank van Veen and colleagues for highlighting this important observation to the wider clinical community. To complement their report, we would like to share our experience of how a different PT assay to those utilized by van Veen et al (2013) performed in a group of patients attending King's College Hospital who were prescribed rivaroxaban.
During the past 6 months, we have been prescribing novel oral anticoagulants (NOAC) for selected patients, if they fit the local South London Cardiac and Stroke network criteria for NOAC use (http://www.slcsn.nhs.uk/noacs.html). For these patients, when clinically indicated (e.g. extremes of body weight, renal dysfunction, or presence of interacting drugs), we measured the activity of the rivaroxaban, in order to satisfy ourselves that the desired response was being achieved. Our laboratory uses the STA-Neoplastine® (Diagnostica Stago, Asnières-sur-Seine, France) CI Plus PT assay for detection of the presence of the drug (rivaroxaban), and the STA-liquid anti-Xa assay® (Diagnostica Stago, Asnières-sur-Seine, France), with appropriate rivaroxaban calibrators and controls for quantification of the actual rivaroxaban concentration in the patients plasma.
Table 1 describes the results of the anti-Xa rivaroxaban concentrations, along with the respective PT assay results from 18 patients in whom we measured these activities. In all cases (except Patient 10), the patients had been prescribed rivaroxaban for a minimum of 1 month prior to the sample being drawn.
| Patient | Age (years) | Daily dose of rivaroxaban (mg) | Time after dose (h:min) | Rivaroxaban concentration (ng/ml) | PT (s) (PT ratio)a |
|---|---|---|---|---|---|
| 1 | 34 | 20 | 15:52 | <20b | 12·7 (0·94) |
| 2 | 50 | 20 | 17:00 | <20b | 13·4 (1·00) |
| 3 | 45 | 20 | 17:26 | 48 | 13·7 (1·02) |
| 4 | 56 | 20 | 07:55 | 59 | 15·5 (1·16) |
| 5 | 56 | 20 | 17:35 | 72 | 18·5 (1·38) |
| 6 | 57 | 20 | 18:45 | 101 | 19·5 (1·45) |
| 7 | 32 | 20 | 12:18 | 115 | 18·5 (1·38) |
| 8 | 57 | 20 | 03:38 | 176 | 19·7 (1·47) |
| 9 | 76 | 20 | 05:00 | 184 | 22·1 (1·65) |
| 10 | 19 | 30c | 07:00 | 185 | 22·1 (1·65) |
| 11 | 87 | 20 | 02:45 | 188 | 21·0 (1·57) |
| 12 | 45 | 20 | 04:00 | 196 | 22·0 (1·64) |
| 13 | 72 | 20 | 03:21 | 268 | 19·2 (1·43) |
| 14 | 53 | 20 | 06:51 | 312 | 25·6 (1·91) |
| 15 | 73 | 20 | 15:22 | 334 | 26·0 (1·94) |
| 16 | 86 | 15 | 05:15 | 354 | 25·3 (1·88) |
| 17 | 67 | 20 | 04:15 | 363 | 22·4 (1·67) |
| 18 | 73 | 20 | 02:37 | 466 | 26·3 (1·96) |
- PT, prothrombin time.
- a PT ratio reflects the patient PT divided by the control PT, which in all cases was 13·4 s.
- b Below the limit of quantification of the assay.
- c Prescribed as 15 mg twice a day.
Acknowledging that the patients from whom the samples were drawn were diverse in terms of their past medical history, the results demonstrate that, in contrast to the PT assays evaluated by van Veen et al (2013), the STA-Neoplastine® CI Plus assay was found to be sensitive across a wide therapeutic concentration range for rivaroxaban in vivo (<20–466 ng/ml), and our results endorse the findings from earlier in vitro work (Samama et al, 2010; Hillarp et al, 2011). Clearly the Neoplastine assay has a greater sensitivity for rivaroxaban than other thromboplastins, and would certainly explain the differences in results observed in our patients to those reported by van Veen et al (2013).
The observations of van Veen et al (2013) coupled with the results from our report are clinically important; at King's, the current guidelines (http://www.kingsthrombosiscentre.org.uk/index.php/anticoagulation) for the management of bleeding for those patients on rivaroxaban suggests a PT screening test, in order to determine whether the presence of rivaroxaban may be contributing to bleeding. Our real-world experience using the STA-Neoplastine® CI Plus assay suggests such a strategy is appropriate, as this PT assay should detect clinically relevant concentrations of rivaroxaban. However, if another less sensitive PT assay is used, then this screening test strategy may be ineffective.
In light of this, we endorse the recent recommendations made by the recent British Society of Haematology guidance on monitoring of NOAC therapy (Baglin et al, 2012), where they stipulate that laboratories should be aware of the sensitivity of their own assays (PT and others) to each NOAC. This will become even more important, as more patients are prescribed these agents and more agents become available in clinical practice in coming months.
Author contributions
JP analysed the data and co-wrote the paper; LNR co-wrote the paper; PBC analysed the samples in the laboratory and co-wrote the paper; RKP co-wrote the paper; RA co-wrote the paper.
Conflicts of interest
The authors have no relevant conflicts of interest to declare.
