Pulmonary Embolism Four Days After Interruption of Therapy with Rivaroxaban

210 Case Report

Pulmonary embolism four days after interruption of therapy with rivaroxaban

Gabor Göndör1; Claudia Stöllberger1 1Krankenanstalt Rudolfstiftung, Wien, Austria

Keywords The relevance of a rivaroxaban rebound phenom- Daten über dieses Phänomen nach einer The- Non-vitamin K-antagonist, oral anticoagu- enon, manifesting as arterial embolism, stroke or rapie mit nicht Vitamin K-Antagonisten ora- lants, rivaroxaban rebound phenomenon, venous thromboembolism should be clarified. It len Antikoagulantien (NOAKs). Ein 58-jähri- atrial fibrillation, pulmonary embolism should be assessed if rebound-phenomena also ger Mann mit paroxysmalem Vorhofflimmern

exist for the NOACs dabigatran, apixaban and (AF) und einem CHA2DS2VASC Score von 4 Summary edoxaban. Thus, the randomized trials and regis- erlitt eine zentrale Pulmonalarterienembolie Thrombosis after cessation of anticoagulation, tries investigating patients with AF or venous vier Tage nach Unterbrechung der Therapie also named rebound thrombosis, is a matter of thromboembolism should be re-analysed and, mit Rivaroxaban wegen eines chirurgischen concern and controversy. There are only few based on these data, recommendations should Eingriffs an der Parotis. Zum Zeitpunkt des Er- published data about occurrence of rebound be developed for situations in which NOAC-ther- eignisses stand er unter einer Therapie mit thrombosis associated with non-vitamin K-an- apy has to be interrupted or ceased. 40 mg Enoxaparin/Tag. Die Parotis wurde in tagonist oral anticoagulant drugs (NOACs). We einer 6-stündigen Operation ohne relevanten report on a 58-year-old male with paroxysmal Blutverlust partiell reseziert. Am ersten post-

atrial fibrillation (AF) with a CHA2DS2VASC operativen Tag traten eine Pulmonalembolie score of 4 who developed central pulmonary Schlüsselwörter und tachykardes AF auf. Er erholte sich unter embolism four days after interruption of rivar- Nicht Vitamin K-Antagonisten, orale Antiko- einer Therapie mit niedermolekularem Hepa- oxaban because of parotid surgery. He had re- agulantien, Rivaroxaban, Rebound-Phäno- rin in therapeutischer Dosis und Amiodaron ceived 40 mg enoxaparin/d. The parotid gland men, Vorhofflimmern, Pulmonalembolie und wurde mit einer Therapie mit Phenpro- was partially resected within 6 hours without coumon entlassen. blood loss. Pulmonary embolism and AF oc- Zusammenfassung Die Relevanz eines “Rivaroxaban-Rebound- curred on the first postoperative day. He re- Ob es nach Beendigung einer Antikoagulati- Phänomens”, das als arterielle Embolie, covered with low-molecular-weight heparin in onstherapie zu einer neuerlichen Thromboem- Schlaganfall oder venöse Thromboembolie therapeutic dosages and amiodarone and was bolie, auch „Rebound“ genannt, kommen auftreten kann, sollte genauer erforscht wer- discharged with phenprocoumon. kann, ist umstritten und es gibt nur wenige den. Es sollte festgestellt werden, ob ähnliche Phänomene auch bei den NOAKs Dabigatran, Apixaban und Edoxaban auftreten. Randomi- Korrespondenzadresse Pulmonalembolie vier Tage nach Unterbrechung sierte Studien und Register, die Patienten mit Univ. Prof. Dr. Claudia Stöllberger der Therapie mit Rivaroxaban Steingasse 31/18 Phlebologie 2018; 47: 210–214 AF oder venöser Thromboembolie eingeschlos- A-1030 Wien, Österreich Nachdruck aus und zu zitieren als: sen haben, sollten unter diesem Gesichtspunkt Tel.: 0043 676 403 11 87 Hämostaseologie 2017; 37: 302–306 ausgewertet werden. Es sollten Empfehlungen Fax: +43 1 71165 2209 https://doi.org/10.5482/HAMO-17-01-0005 ausgearbeitet werden für die Antikoagulation Email: [email protected] received: January 23, 2017 accepted in revised form: August 10, 2017 in Situationen, in denen die Therapie mit NO- epub ahead of print: August 24, 2017 AKs unterbrochen oder beendet wird. Funding No support for the work in the form of grants, equip- ment, drugs or any combination of these

Introduction ACs), which are increasingly used in patients nous thromboembolism based on the results with atrial fibrillation for prevention of stroke of two randomized trials (2, 3). Rivaroxaban is a direct anti-Xa oral anticoa- or embolism, based on the results of a large In comparison to the traditionally used gulant and belongs to the group of non-Vita- randomized trial (1). Furthermore, rivaroxa- vitamin-K-antagonists, the NOACs do not min K-antagonist oral anticoagulants (NO- ban is also approved for the treatment of ve- require any routine monitoring. On the ot-

Phlebologie 4/2018 © Georg Thieme Verlag KG 2018 Göndör G, Stöllberger C: Pulmonary embolism after rivaroxaban 211 her hand, in emergency situations it is mo- histological result showed a partial resecti- Concomitantly, the veins of the pelvis and re difficult to estimate their plasma con- on of the parotid gland with two benign lower extremities were investigated by CT centration and there are no recommended cystadenolymphomas. and did not show any thrombosis. A bed- specific antidotes so far. Postoperatively the patient complained side-echocardiography showed a slightly Furthermore, rivaroxaban and other of pain in the right popliteal fossa and 15 reduced left ventricular systolic function, a NOACs carry a black box warning regar- hours after the end of the surgery, he start- normally sized right ventricle with normal ding the risk of rebound thrombosis after ed to suffer from sudden onset of severe function, no pericardial effusion and no discontinuation (4, 5). dyspnea why he was transferred to the me- thrombus. Tachycardious atrial fibrillation dical department. At admission, the blood was converted into sinus rhythm after the pressure was 140/85 mmHg, the heart rate application of amiodarone, potassium and Case report 150/min and the O2-saturation 88 %. After magnesium. Thrombolysis was considered supplemental oxygen therapy with 5 l/min, but refused because of the recent surgery. A 58-year-old male patient was admitted to the oxygen saturation increased to 92 %. Thus, enoxaparin sodium 100 mg was in- our department because of an acute pul- The D-dimer was 7.60 mg/l (▶ Table 1). jected as a bolus and prescribed two times monary embolism. He had a history of pa- The electrocardiogram showed tachycar- daily. An oral anticoagulation therapy with roxysmal atrial fibrillation since 3 years, dious atrial fibrillation. phenprocoumon was initiated after 7 days currently under therapy with rivaroxaban The CT-angiography of the pulmonary and the patient was discharged without any 20 mg, arterial hypertension since many ye- arteries showed a central pulmonary em- complaints after 12 days. ars, peripheral arterial disease stage IIb sin- bolism on both sides of the lung reaching ce 5 years, fatty liver disease, prostatic hy- into the segmental arteries (▶ Figure 1). perplasia, cholecystectomy 5 years before, lateral parotidectomy of the left parotid gland 13 years before, substituted hypothy- Tab. 1 Laboratory findings roidism since the thyroidectomy at the age Parameter Day 1 Day 3 Day 6 Day 8 Day 9 of 41years and appendectomy at the age of (normal range) (16:10) (11:16) (9:02) (09:04) (9:22) 14 years. His CHA2DS2VASc score was 4 and the HAS-BLED score was 2. He was PT % (70–140) 111 95 NM NM 94 scheduled to an elective parotidectomy be- INR 0,91 1,01 NM NM 1,01 cause of cystadenolymphoma of the right D-Dimer mg/l (-0,50) NM 7.60 NM NM NM parotid gland. Leukocytes G/l (4–9) NM 17,6 13,6 12,5 NM Originally, he took phenprocoumon since 3 years, but the oral anticoagulation Thrombocytes G/l (150–400) NM 221 277 345 NM was changed to rivaroxaban six weeks be- Haemoglobin g/dl (14–17) NM 13,5 13,9 14,3 NM fore the actual surgery by his general CRP mg/l (-5) NM 39,5 28 13,9 NM practitioner. The cause for the readjust- Sodium mmol/l (136–145) NM 141 141 NM 140 ment was a transient ischemic attack of the right medial cerebral artery with sensory Potassium mmol/l (3,4–4,5) NM 3,8 3,9 NM 4,5 abnormality of the left arm for 15 minutes. Creatinin mg/dl (0,70–1,20) NM 0,75 0,80 NM 0,82 The International Normalized Ratio at the CK U/l (-190) NM 2075 422 NM NM time of the TIA was 2.75. CK-MB U/l (-25) NM 66 12 NM NM Preoperatively, the Caprini DVT risk score (a risk stratification score for perio- Troponin T ng/l (-14) NM NM 17 NM NM perative venous thromboembolism) was 4, Haemoglobin A1c % (4–6) NM NM 6,7 NM NM which corresponds to high risk and, in case Folic acid ng/ml (4,60–18,70) NM NM 3,67 NM NM of inclusion of TIA, the score would add up Vitamin D3 nmol/l (75-) NM NM 10,34 NM NM to 9 (6). He stopped taking rivaroxaban four days before the surgery and received PSA mcg/l (-3,200) NM NM NM 7,270 NM enoxaparin sodium 40 mg once daily. Besi- ALT U/l (10–50) NM NM 42 NM NM des, he took amlodipine/valsartan/hydro- GGT U/l (-60) NM NM 43 NM NM chlorothiazide 5/160/25 mg once, bisopro- ALP U/l (40–129) NM NM 66 NM NM lol 5 mg twice and levothyroxine sodium 100 µg once daily. The surgery took 6 hours Bilirubin mg/dl (-1,20) NM NM 0,53 NM NM and succeeded without any complication. ALP = alkaline phosphatase; ALT = alanine-aminotransferase; CK = creatine phosphokinase; CRP = The patient had no blood loss and he did C-reactive protein; GGT = gamma-glutamyl transferase; INR = international normalized ratio; NM not need any erythrocyte concentrates. The = not measured; PSA = prostate-specific antigene; PT = prothrombin time

that dabigatran does not cause rebound thrombosis (13), other studies suggested the opposite (14). Thorne et al reported three arterial or venous thromboembolism cases within one month after cessation of dabigatran (15). Data from a recently published non-in- dustry-sponsored registry show that the NOACs rivaroxaban and dabigatran offer good protection against thromboembolic a b events in real-life patients with nonvalvular atrial fibrillation, but that interruptions of Fig. 1 Thrombus in the CT-angiography of the pulmonary arteries treatment represent a vulnerable period (16). In that registry, 866 patients with non-valvular atrial fibrillation with an ave- Discussion dian time to reach a therapeutic INR value rage CHA2DS2-VASc-Score of 2.1, who we- was 13 days in the rivaroxaban-group com- re started on dabigatran or rivaroxaban, Rivaroxaban is a NOAC, which takes its ef- pared to 3 days in the warfarin-group (5, were analysed for thromboembolic events fect by inhibiting the prothrombinase- 9). Unfortunately, there are no data about and survival. Patients who had temporary complex and the free factor Xa directly. venous thromboembolic events occurring or permanent discontinuation of NOACs The prothrombinase-complex is an enzy- after the ROCKET AF study since only were compared to patients on continuous matic complex of the blood coagulation stroke and arterial embolism were registe- NOAC treatment. The incidence of throm- cascade consisting of the activated factors red. boembolic events during treatment inter- X and V, calcium and phospholipids (7). In our case, we considered three possi- ruptions was more than 6-times higher The advantage of rivaroxaban compared to ble triggers that could have led to the than the “natural course” incidence pre- the vitamin-K-antagonists is that it does thrombus formation and the subsequent dicted by the CHA2DS2-VASc-score. On not require any routine monitoring. Based pulmonary embolism. the other hand, the incidence of thrombo- on randomized trials in patients with ve- First, the phenomenon of rebound embolism was significantly lower in pa- nous thromboembolism (2, 3), the guideli- thrombosis could have been a reason. Riva- tients with uninterrupted NOAC treatment nes of acute pulmonary embolism mention roxaban carries a black box warning regar- than the predicted incidence for untreated rivaroxaban 20 mg once daily as an alterna- ding the risk of thrombosis after disconti- patients (16). Also in that registry, only ar- tive to vitamin-K-antagonists if extended nuation: „discontinuing rivaroxaban places terial embolism and strokes, and no venous anticoagulation treatment is necessary as a patients at an increased risk of thrombotic thromboembolic events, were counted. class IIa, level B recommendation (8). events” and „an increased risk of stroke was The possibility of rebound thrombosis Rivaroxaban was also found to be non- observed following rivaroxaban discontinua- after cessation of therapy with vitamin inferior to warfarin for the prevention of tion in clinical trials in atrial fibrillation pa- K-antagonists has been subject of discussi- stroke and systemic embolism in patients tients” (4, 5). There are only limited data on for a long time, e.g. by Genewein et al.in with moderate- to high-risk nonvalvular about the prevalence and clinical relevance 1996 (17). atrial fibrillation and concerns about a po- of a rivaroxaban rebound phenomenon. The pathophysiologic mechanism of a tential rebound effect after cessation of ri- Bakhit et al. reported a case with rebound rivaroxaban rebound phenomenon is un- varoxaban have been raised in association thrombosis. Twenty-four hours after the ces- certain. Haynes et al., based on an in vitro with the ROCKET AF (Rivaroxaban Once sation of rivaroxaban, an elective radiofre- study, suggested that the discontinuation of Daily Oral Direct Factor Xa Inhibition quency catheter ablation was performed but rivaroxaban leads to prothrombotic activi- Compared with Vitamin K Antagonism for it had to be interrupted because a large septal ty. The reason of this phenomenon may be Prevention of Stroke and Embolism Trial in thrombus in the left atrium was found (10). that when rivaroxaban plasma concentrati- Atrial Fibrillation) study, a multicenter, The black box warning, however, exists ons decrease after cessation of therapy, the- randomized, double-blind trial with the for other oral anticoagulants as well. “Pre- re is an unmasking of thrombus-associated participation of 14,264 patients (1). After mature discontinuation of any oral anticoa- prothrombinase (18). the ROCKET AF study, an end of study gulant in the absence of adequate alternati- A second pathogenetic possibility for (EOS) visit was performed and the partici- ve anticoagulation increases the risk of the thrombotic event in our case are the 6 pants were transitioned to open-label vita- thrombotic events” (4). The direct oral anti hour-long surgery and the following hospi- min-K-antagonist-therapy. Between days 3 Xa-inhibitors apixaban and edoxaban car- talization which are well known risk factors and 30, an excess of stroke and systemic ry the same black box warning (11, 12). Re- for thrombosis (19). However, the inci- embolic events were observed in partici- garding the direct oral thrombin-inhibitor dence of thrombosis and pulmonary em- pants assigned to rivaroxaban, and the me- dabigatran, the RE-MODEL trial suggested bolism are not as common in patients after

Phlebologie 4/2018 © Georg Thieme Verlag KG 2018 Göndör G, Stöllberger C: Pulmonary embolism after rivaroxaban 213 ear- nose- and throat-surgery than after ot- European Heart Rhythm Association (EH- References her surgical specialties (20, 21). Obviously, RA) practical guide on the use of new oral enoxaparin 40mg was not sufficient to pre- anticoagulants in patients with non-valvular 1. Patel MR et al. Rivaroxaban versus warfarin in vent venous thromboembolism in our pa- atrial fibrillation (26). If we apply these re- nonvalvular atrial fibrillation. N Engl J Med 2011; 365(10): 883-891. tient. The relatively low dosage is debatable commendations to our patient, the last dose 2. Bauersachs R et al. Oral rivaroxaban for sympto- regarding the high risk in the Caprini of rivaroxaban should have been given three matic venous thromboembolism. N Engl J Med score. Since the patient was transferred to days before the surgery and accordingly to 2010; 363(26): 2499-2510. 3. Buller HR et al. Oral rivaroxaban for the treatment the medical department only after expe- the EHRA, more than 48 hours preoperati- of symptomatic pulmonary embolism. N Engl J riencing the postoperative dyspnea, we we- vely, assuming that we identify the procedu- Med 2012; 366(14): 1287-1297. re not involved in the decision-making re- re as a high bleeding risk surgery. In case of 4. United States Food and Drug Administration. XA- garding the perioperative bridging of his emergency procedures, Levy et al.’s publica- RELTO (Rivaroxaban) tablets, for oral use. https://www.accessdata.fda.gov/drugsatfda_docs/ anticoagulation therapy. tion is a helpful tool in the management and label/2017/202439s021lbl.pdf. Initial U.S. Appro- Most of the recommendations regarding discontinuation of NOACs (27). val: 2011. bridging of anticoagulant therapy during A third possibility for the thromboem- 5. Patel MR et al. Outcomes of discontinuing rivarox- surgery are derived from patients treated bolism could be heparin-induced throm- aban compared with warfarin in patients with nonvalvular atrial fibrillation: analysis from the with vitamin-K-antagonists. The American bocytopenia, a complication of heparin- ROCKET AF trial (Rivaroxaban Once-Daily, Oral, College of Chest Physicians (ACCP) inter- therapy, associated with a prothrombotic Direct Factor Xa Inhibition Compared With Vit- national guidelines on anticoagulant perio- state. Despite the definition “thrombocyto- amin K Antagonism for Prevention of Stroke and perative bridging in patients with atrial fi- penia”, this condition may rarely occur wit- Embolism Trial in Atrial Fibrillation). J Am Coll Cardiol 2013; 61(6): 651-658. brillation summarize the recommendation hout a reduction of the number of platelets 6. Caprini JA. Risk assessment as a guide to throm- only for vitamin K-antagonists. After distin- (28). Since the platelet count in our patient bosis prophylaxis. Curr Opin Pulm Med 2010; guishing between high risk and low risk for remained within the normal range at re- 16(5): 448-452. thromboembolism, a bridging therapy is re- peated measurements, heparin-induced 7. Perzborn E et al. In vitro and in vivo studies of the novel antithrombotic agent BAY 59–7939 – an commended or not (22). New considerati- thrombocytopenia (HIT) was considered oral, direct Factor Xa inhibitor. J Thromb Haemost ons to avoid the increased risk of thrombo- as unlikely and no testing for HIT-antibo- 2005; 3(3): 514-521. tic events after discontinuation suggest an dies was carried out. 8. Konstantinides SV et al. 2014 ESC Guidelines on the diagnosis and management of acute pulmon- uninterrupted approach in anticoagulation It remains uncertain which of the three ary embolism The Task Force for the Diagnosis management during procedures (23). Ko- described mechanisms were responsible for and Management of Acute Pulmonary Embolism walewski et al. performed a meta-analysis the venous thromboembolism of our patient, of the European Society of Cardiology (ESC) En- and came to the conclusion that in patients and probably a coincidence of these different dorsed by the European Respiratory Society (ERS). European Heart Journal 2014; 35: 3033–3080. undergoing coronary angiography, the un- mechanisms may have played a role. 9. Mahaffey KW et al. End of study transition from interrupted oral anticoagulation is safer We conclude from this observation that study drug to open-label vitamin K antagonist than the discontinuation with bridging. there is an urgent need to clarify the rate of therapy: the ROCKET AF experience. Circ Car- Further studies, investigating other proce- a rivaroxaban rebound phenomenon, ma- diovasc Qual Outcomes 2013; 6(4): 470-478. 10. Bakhit A et al. Rebound thrombosis within dures are required to assess whether or not nifesting as arterial embolism and stroke as 24hours after interruption of therapy with rivarox- this approach would have been justifiable in well as venous thromboembolism. Further- aban. Int J Cardiol 2016; 212: 235-236. our case. The BRIDGE trial was not inclu- more, it has to be assessed if rebound phe- 11. United States Food and Drug Administration. ded in the ACCP guidelines, and it reported nomena also exist for other NOACs like ELIQUIS (Apixaban) tablets, for oral use. https://www.accessdata.fda.gov/drugsatfda_docs/ that forgoing bridging anticoagulation was dabigatran, apixaban and edoxaban. For label/2012/202155s000lbl.pdf. Initial U.S. Appro- noninferior to perioperative bridging with that purpose, the data of the large rando- val: 2012. low-molecular-weight heparin for the pre- mized trials and registries investigating pa- 12. United States Food and Drug Administration. SA- vention of arterial thromboembolism and tients with atrial fibrillation or venous VAYSA (Edoxaban) tablets, for oral use. https://www.accessdata.fda.gov/drugsatfda_docs/ even decreased the risk of major bleeding thromboembolism should be analysed. label/2015/206316lbl.pdf. Initial U.S. Approval: (24). There was no significant difference There is a need to develop recommendati- 2015. between the bridging and non-bridging ons for situations in which NOAC-therapy 13. Eriksson BI et al. Oral dabigatran etexilate vs. sub- groups in the rates of deep vein thrombosis has to be interrupted or ceased. At present cutaneous enoxaparin for the prevention of venous thromboembolism after total knee replacement: and pulmonary embolism either. However, it remains uncertain if low-molecular- the RE-MODEL randomized trial. J Thromb Hae- this trial did not include any NOACs. weight heparin in therapeutic doses has to most 2007; 5(11): 2178-2185. Spyropoulos et al. elaborated recommen- be administered for several days until the 14. Hermans C, Claeys D. Review of the rebound phenomenon in new anticoagulant treatments. Curr dations for bridging therapy both for vita- hypercoagulable state has resolved. Med Res Opin 2006; 22(3): 471-481. min K-antagonists and for direct oral anti- 15. Thorne KM, Dee S, Jayathissa S. Thrombotic events after discontinuing dabigatran: rebound or coagulants depending on patient- and pro- Conflicts of interest cedure-related risks of thromboembolism resumption? BMJ 2012; 345: e4469. 16. Vene N et al. Risk of Thromboembolic Events in and bleeding, and the patient’s renal functi- The authors declare that they have no con- Patients with Non-Valvular Atrial Fibrillation on (25), which are slightly different from the flicts of interest.

After Dabigatran or Rivaroxaban Discontinuation of current practice. J Laryngol Otol 2015; 129(2): 24. Douketis JD et al. Perioperative Bridging Antico- – Data from the Ljubljana Registry. PLoS One 164-167. agulation in Patients with Atrial Fibrillation. N 2016; 11(6): e0156943. 21. Chiesa Estomba C et al. The risk of venous throm- Engl J Med 2015; 373(9): 823-833. 17. Genewein U et al. Rebound after cessation of oral boembolism in ENT and head & neck surgery. 25. Spyropoulos AC, Douketis JD. How I treat antico- anticoagulant therapy: the biochemical evidence. Otolaryngol Pol 2015; 69(3): 31-36. agulated patients undergoing an elective pro- Br J Haematol 1996; 92(2): 479-485. 22. Guyatt GH et al. Executive summary: Antithrom- cedure or surgery. Blood 2012; 120(15): 18. Haynes LM, Orfeo T, Mann KG. Rivaroxaban de- botic Therapy and Prevention of Thrombosis, 9th 2954-2962. livery and reversal at a venous flow rate. Arte- ed: American College of Chest Physicians Evi- 26. Heidbuchel H et al. EHRA practical guide on the rioscler Thromb Vasc Biol 2012; 32(12): dence-Based Clinical Practice Guidelines. Chest use of new oral anticoagulants in patients with 2877-2883. 2012; 141(2 Suppl.): 7S-47S. non-valvular atrial fibrillation: executive sum- 19. Caprini JA, Merli GJ, Bhatt DL. Update on Risk 23. Kowalewski M et al. Meta-analysis of uninterrupt- mary. Eur Heart J 2013; 34(27): 2094-2106. Factors for Venous Thromboembolism ed as compared to interrupted oral anticoagulation 27. Levy JH. Discontinuation and Management of Di- http://www.venousdisease.com/Publications/Up with or without bridging in patients undergoing rect-Acting Anticoagulants for Emergency Pro- date%20on%20risk%20factors-caprini.pdf. Am J coronary angiography with or without percut- cedures. Am J Med 2016; 129(11S): S47-S53. Med 2005. aneous coronary intervention. Int J Cardiol 2016; 28. Alvarez GF, Bihari D, Collins D. Heparin-induced 20. Nash R, Randhawa N, Saeed SR. Venous throm- 223: 186–194. thrombosis with a normal platelet count. Crit Care boembolism prophylaxis in ENT surgery: a survey Resusc 2007; 9(1): 51-53.