This supplementary information is presented as submitted by the corresponding author. It has not been copy-edited by NTvG.
Use of the new oral anticoagulant drugs (NOACs) is increasing. The main advantage of these drugs compared to vitamin K –antagonists (VKA) is the fact that routine monitoring of the anticoagulant effects does not seem necessary. However, in this case report we would like to illustrate that monitoring of anticoagulant effects can also have important benefits in case of drug-drug interactions.
A 53 year old man who underwent a partial knee arthroplasty 4 days earlier, was readmitted to the hospital with shortness of breath and respiratory chest pain. The symptoms arose the day after thromboprophylaxis was switched from dalteparin 5000 IE QD, which was used during hospital admittance, to rivaroxaban 10 mg QD at discharge. The patient also used carbamazepine 600 mg BID for epilepsy.
At physical examination the man was neither ill nor pale, and there were no signs of cyanosis. Laboratory analysis showed an elevated D-dimer of 1367 µg/L and a C-reactive protein of 63 mg/L. Arterial blood gas was normal. The duplex scan showed no signs of a deep vein thrombosis and the chest X-ray was also normal. A CT scan showed right basal filling defects. Based on the CT scan the patient was diagnosed with pulmonary embolisms and treated with dalteparin 18.000 QD (200IE/kg), followed by acenocoumarol for a total duration of 6 months.
Although the risk of a pulmonary embolism after knee replacement during treatment with rivaroxaban is small (0,5%), pulmonary embolisms after this type of surgery will be seen regularly in clinical practice based on the sheer number of knee replacements nationwide.
However, in our patient the use of carbamazepine was probably an important complicating factor.
Carbamazepine is a strong inducer of the CYP3A4 enzyme, which is an important enzyme for the degradation of rivaroxaban. Although the actual impact of carbamazepine on the pharmacokinetics of rivaroxaban is unclear, use of another CYP3A4 inducing drug rifampicin reduced the average AUC-value for rivaroxaban with 50%. Given the similarity in CYP3A4 inducing properties of rifampicin and carbamazepine, it is likely that carbamazepine also substantially increased the clearance of rivaroxaban in our patient, which led to underdosing of rivaroxaban. Although the concentration of rivaroxaban in our patient was not measured, the fact that the pulmonary embolism occurred the day after the patient started using rivaroxaban instead of dalteparin is indicative for an important impact of carbamazepine on the efficacy of rivaroxaban.
We encourage monitoring of the anticoagulant effects of the NOACs in case of drug-drug interactions by using anti-Xa assays, especially when NOACs are given in higher doses for a long duration, since this could prevent treatment complications. However, it could also be preferable to choose a VKA or Low Molecular Weight Heparin in those situations.