Detecting clinically relevant rivaroxaban or dabigatran levels by routine coagulation tests or thromboelastography in a cohort of patients with atrial fibrillation

For more than 60 years vitamin K antagonists (VKA) were the drugs of choice to prevent thrombosis in patients with NVAF and mechanical heart valves, as well as to treat and prevent recurrence of thrombosis in patients with venous thromboembolism (VTE).

The management of patients on anticoagulant therapy was guided by laboratory testing of prothrombin time (PT), which was internationally harmonized by the ISI factor resulting in (international normalized ratio) INR, an easily interpretable laboratory test [1]. With the introduction of direct oral anticoagulants it was emphasized that no routine laboratory testing would be needed [2–7]. However, in emergency situations the need for testing remains since residual activity of the anticoagulant treatment might introduce e.g. bleeding during surgery. Screening tests for detecting anticoagulants, which can be performed 24 h a day instead of just during working hours should therefore be available to guide treatment decisions. The concentration below 30 ng/mL is proposed as a safe-for surgical treatment threshold [8]. Thereby, this concentration is sufficient to administer antidote against DOAC if needed [9].

Activity-based drug levels in plasma can be determined but are not obtainable in all care settings [10]. Traditional routine coagulation tests are widely available and may be used for first line testing in emergency situations [11–15]. In many countries regional guidelines have been issued upon the introduction of the DOACs indicating the preferred strategies of action for both the prescription and (laboratory) management of these drugs [16–18]. These guidelines do indicate that the thrombin time (TT) is the most sensitive routine coagulation assay for detection of dabigatran and that the PT is the most sensitive routine coagulation assay for detection of rivaroxaban. Guidelines do however not (sufficiently) report on the sensitivity of the different individual TT, PT and activated partial thromboplastin Time (aPTT) tests that are commercially available. This may be due to the fact that most of these guidelines were based on laboratory data using plasma spiked with DOAC instead of samples acquired from patients using the drugs.

We therefore set out to evaluate two proposed screening algorithms [16, 18] and at the same time validate seven different routine anticoagulation tests for screening of plasma drug activity in daily practice, using plasma samples from a cohort of patients on dabigatran and rivaroxaban. In addition we evaluated ROTEM® for the same purpose in whole blood in a subsample of the study population [19].

The main findings of the present study are that in patients with NVAF, the ability of routine coagulation tests to detect the presence of significant levels of dabigatran or rivaroxaban is test and reagent dependent. Therefore, emergency care protocols should ensure that local test reagents are sufficiently accurate for detecting the presence of DOACs. Otherwise, the CT-INTEM and CT-EXTEM of ROTEM® may be good and fast whole blood alternatives that may be readily available on site.

In emergency situations it is crucial that routine coagulation tests are clinically validated and that the test sensitivity for the detection of minimal and clinically important levels of anticoagulation is known. In patients with serious bleeding or requiring urgent intervention with bleeding risk, a drug concentration > 30 ng/mL is proposed as clinical relevant and sufficient to administer antidote against DOAC [9]. Emergency coagulation testing according to suggested algorithms is usually dependent on whether knowledge of the actual intake and the timing of ingestion of DOAC is available or not. If the DOAC taken by the patient is indeed known, one may choose the most appropriate test for the drug: for dabigatran the most appropriate screening test is stated to be the TT¹⁵. If the TT is not prolonged, the presence of dabigatran can be completely excluded. The PT is the most appropriate screening test for rivaroxaban; for most PT reagents a normal PT implies that clinical relevant concentrations of rivaroxaban can be excluded [15]. According to current guidelines, a panel of conventional coagulation tests (aPTT, PT and TT) should be used in case of intake of an unknown DOAC, to thereby exclude relevant drug levels for either dabigatran or rivaroxaban. In addition, for other DOACs such as apixaban or edoxaban, calibrated drug specific anti-Xa levels still need to be measured; screening tests are not yet available [20, 21].

However, whether the suggested algorithms are actually feasible in real life situations has not been tested yet. APTT, PT and TT are both widely and rapidly available in emergency situations, but one should be careful with the interpretation of these tests [22–25]. One needs to realize that these standard coagulation tests with low screening sensitivity might lead to a false reassurance when indicating absent drug levels. In our cohort both TT tests were able to pick up clinically relevant levels of dabigatran (≥30 ng/mL). In case of rivaroxaban however, there was a marked difference in sensitivity between the two PT tests used: PT neoplastin had a sensitivity of 97.5%, PT Innovin expressed a much lower sensitivity of only 67.7%. This may result in a underestimation of drug levels in more than 32% of patients in case PT Innovin is used for screening for active drug levels, while this is much less for PT Neoplastin (<3%). As can be appreciated from the data, most of the before mentioned tests are only useful as in- or exclusion tests (Yes, No). It would be a great advantage if also actual drug levels could be predicted form the result of a test in an emergency situation. Therefore, the correlations between drug levels and the panel of routine coagulation tests were also assessed. For dabigatran APTT Cephascreen and Thrombin were both sensitive and accurate. For rivaroxaban only PT Neoplastin was both sensitive and level predictive. Moreover, rapid measurement of DOAC plasma concentration levels have become available and might therefore also feasible for use in emergency situations [26]. Their role in the algorithms should be evaluated in the near future. For all laboratory tests applies that handling of samples by laboratory technicians is mandatory. This is time consuming and may impact swift and adequate interventions in an emergency setting. A point of care test for which whole blood can be used would improve the decision making process.

The CT-EXTEM showed good sensitivity for both rivaroxaban and dabigatran, its linearity is however moderate. Our results differ from previous studies where ROTEM® was able to show a dose dependent increase in DOAC levels in spiked whole blood samples, but poorly performed at low DOAC levels [27]. Our study showed that the ‘real life’ sensitivity of ROTEM® is better in patients using DOAC and might therefore be a good candidate for emergency testing. The added advantage of ROTEM® testing is that the results are readily available (< 5 min) and there is uniform performance with worldwide standardization. ROTEM® can be used by emergency personnel or as point of care testing in operating theaters, but also with standard situations in most hospitals laboratories. ROTEM® testing may in addition also offer the possibility of monitoring the use of antidotes.

Our study has some weaknesses: we included a relatively small number of patients and the DOACs used were limited to dabigatran and rivaroxaban, as patients in our cohort did not use apixaban or edoxaban. We furthermore did not use the accepted gold standard test, liquid chromatography with mass spectrometry (LC-MS) for quantification of DOACs [28] although the selected coagulation activity assays correlated well with LC-MS techniques.

Strengths of our study are 1) that real life samples to assess the feasibility of current guidelines on DOAC testing and 2) an extensive panel of 7 different anticoagulation tests in addition to ROTEM® testing was used.

We conclude that guideline proposed screening algorithms are feasible in a real life steady state NVAF population. However, the ability of routine coagulation tests to detect the presence of significant levels of dabigatran or rivaroxaban is extremely test and reagent dependent.

Emergency care protocols should ensure that local test reagents are sufficiently accurate for detecting the presence of DOACs. CT-INTEM and CT-EXTEM of ROTEM® can be considered as effective whole blood alternatives that have the advantage to be readily available on site.