©Ferrata Storti Foundation

Thrombosis • Decision Making and Problem Solving Management of coumarin-associated coagulopathy in the non-bleeding patient: a systematic review [haematologica] 2004;89:857-862 FRANCESCO DENTALI ABSTRACT WALTER AGENO Background and Objectives. Excessive anticoagulation is a frequent complication of anticoagulant therapy. The risk of hemorrhage approximately doubles for each one point increase in the International Normalized Ratio (INR) above 3.0. Reducing a prolonged INR to within the desired therapeutic range requires that oral anticoagulants be with- held. In addition, vitamin K may be administered. Since this latter treatment can produce rapid reductions in the INR, it must be carefully tailored to meet individual needs, balancing the risk of bleeding against the potential risk of causing thromboembolism. Methods. To review available literature on the management of coumarin-associated coagulopathy in asymptomatic patients, a Medline search was carried out and papers published in English from 1966 and 2003 were identified. All available information on the management of asymptomatic patients presenting with coumarin-associated coagulopathy was analyzed. Results. Following the results of clinical studies that only used an elevated INR as a surrogate end-point for the risk of bleeding, low dose oral vitamin K appears as the preferable strategy for rapidly restoring therapeutic INR levels in asymptomatic patients who present with an excessively prolonged INR due to warfarin therapy. For the treatment of patients with asymptomatic acenocoumarol-induced coagulopathy, vitamin K does not add any benefit to the strategy of simply withholding oral anticoagulant treatment. Interpretation and Conclusions. Large randomized trials using clinical end-points are now required to provide evidence-based treatment recommendations for patients with coumarin-associated coagulopathy. Key words: vitamin K, warfarin, acenocoumarol, coagulopathy. From the Internal Medicine, University of Insubria, Varese, xcessively prolonged International ducing a sub-therapeutic INR which could Italy. Normalized Ratio (INR) values are an be associated with either thrombosis or Correspondence: Dr Walter independent predictor for major subsequent resistance to re-anticoagula- E 1-3 Ageno , U.O. Medicina I bleeding, and the risk of hemorrhage tion. Two treatment strategies are avail- Ospedale di Circolo Viale Borri 57, 21100 Varese, Italy. E-mail: approximately doubles for each one point able: withholding oral anticoagulants and [email protected] increase in the INR above 3.0.4 In a recent allowing the INR level to return to within 5 @2004, Ferrata Storti Foundation study, Hylek and associates reported that the therapeutic range, or withholding oral 5 of 114 (4%) asymptomatic patients with anticoagulants and administering vitamin an INR more than 6.0 who were only man- K.6 A third option, withholding oral antico- aged with temporary warfarin withdrawal agulants, administering vitamin K and developed major or life-threatening bleed- administering fresh frozen plasma or ing within 2 weeks of their elevated INR coagulation factor concentrates, should value being recorded. This evidence sup- not be used because of the risks and costs ports the need to examine the optimal associated with prophylactic administra- method for returning prolonged INR values tion of coagulation factor concentrates towards the desired therapeutic range. except in patients who are at the highest The objectives of treatment for non- risk of hemorrhage. bleeding patients with an excessively pro- Vitamin K is a specific pharmacologic longed INR value are to limit the risk of antagonist to coumarin drugs and one bleeding complications by reducing the form of vitamin K, phytonadione or vita- time the INR is elevated, while not pro- min K1, is available for clinical use. haematologica 2004; 89(7):July 2004 857 F. Dentali et al. Vitamin K can be administered by intravenous, sub- Results cutaneous or oral routes and can produce rapid reductions in the INR. This is why its use requires Withholding treatment careful tailoring to meet individual needs, balancing Simply withholding oral anticoagulants and allow- the risk of bleeding against the potential risk of causing the INR to fall into the therapeutic range is a ing thromboembolism. According to the results of widely used approach. Two large case-series exam- two physician surveys that were carried out to assess ined the safety of this approach9,10 and both conclud- the use of vitamin K in patients with warfarin-asso- ed that conservative treatment of non-bleeding over- ciated coagulopathy,7,8 even in non-bleeding patients anticoagulated patients is safe. In these investiga- with markedly prolonged INR values most physicians tions, a total of 352 INR values above 6.0 occurred in would not administer vitamin K. In the survey carried 299 patients and only 2 patients (0.6%) suffered out in the USA, Libby and Garcia8 reported that as hemorrhage after temporary discontinuation of their many as 86% of contacted physicians would not give anticoagulants. One of these studies suggested that vitamin K to asymptomatic patients presenting with simply withholding warfarin was more cost-effective an INR greater than 6.0 and no evident risk of bleed- than administering vitamin K.10 Concerns about these ing, and that 25% of contacted physicians would not studies include retrospective data collection, lack of administer vitamin K even to a similar patient con- uniform outcome assessment, and lack of prospec- sidered to be at high risk of bleeding. tively defined outcome measures. More recently, To assess available evidence on the efficacy and Hylek and colleagues5 prospectively compared 114 safety of different managements of coumarin-associ- patients with INR levels greater than 6.0 with 268 ated coagulopathy and on the efficacy and safety of patients with INR levels in the target range. All the different routes of administration and doses of patients in the former group were managed by sim- vitamin K used, we performed a systematic review of ply withholding warfarin treatment. After a 2-week the literature. follow up, 10 (8.8%) patients with elevated INR values reported abnormal bleeding, and 5 (4.4%) of them had experienced a major hemorrhage whereas Design and Methods none of the patients with an in-range INR had done so. Of interest, only 33% of patients with an INR A series of Medline searches were conducted in the greater than 6.0 had an INR less than 4.0 within 24 database from 1966 to December 2003. All searches hours, and 55% within 48 hours. were carried out without mapping search terms to subject headings. In the first search, terms included Intravenous vitamin K1 vitamin K, phylloquinone, phytonadione, and exclud- Two studies11,12 have compared different doses of ed infant, newborn, fetus, seizure or chemotherapy. vitamin K administered intravenously in patients pre- The results of this search were combined with the senting with warfarin-associated coagulopathy and results of a second search (using the AND function), concluded that 0.5 mg was the optimal dosage if the in which terms included warfarin, acenocoumarol, therapeutic intent was to return the INR to within vitamin K antagonist, coumarin or oral anticoagulant. the usual therapeutic range. In the study by Shetty The results of the combined search were limited to and colleagues,11 50% of patients with an INR greater human, and English language. This strategy was than 5.0 who were treated with 1 mg of intravenous refined by combining the results, using the AND vitamin K had a subtherapeutic INR (lower than 2.0) function, with the results of a search including the after 24 hours, whereas none of the patients who terms treatment or intervention or correction. Papers received 0.5 mg had INR values lower than 2.0. Only whose titles or abstracts suggested they presented 4 out of 21 patients who received 0.5 mg of intra- either prospective cohort studies, randomized trials, or venous vitamin K had an INR greater than 4.0 at 24 academic reviews were selected for detailed review. hours. Similar results were observed by Hung and col- Additional papers meeting these inclusion criteria were leagues12 in a randomized controlled trial that com- selected for review from the authors’ libraries, and pared 3 doses of intravenous vitamin K: 2.0 mg, 1.0 from consideration of the reference lists of those arti- mg, and 0.5 mg. The two higher doses were effective cles selected for detailed review. Studies conducted on but led to an excessive rate of subtherapeutic INR. In non-bleeding patients presenting with coumarin-asso- a third study, Andersen et al.13 reported that adminis- ciated coagulopathy were selected for the purpose of tering 1.0 mg of intravenous vitamin K without with- this review. This search strategy produced 18 relevant holding warfarin was more effective than simply articles.5,9-17,20-27 withholding warfarin. A potential concern with the use of the intravenous route of administering vitamin K is the risk of ana- 858 haematologica 2004; 89(7):July 2004 Managing coumarin induced coagulopathy phylactoid reactions. Although frequently reported, Oral vitamin K1 and likely more common in patients who receive There is increasing interest in the use of oral vitamin large intravenous doses administered rapidly, the true K for the treatment of coumarin-associated coagu- frequency of this complication is probably very low. lopathy. When used in doses of 1 to 2.5 mg, oral vita- Apparently, micelle preparations of vitamin K are less min K does not appear to produce warfarin resistance, allergenic than castor oil preparations and should, and its use has not been associated with anaphylac- therefore, be preferred when commercially available. toid or skin reactions.17-20 Furthermore, the oral admin- An additional strategy to reduce the risk of anaphy- istration of vitamin K has the potential to simplify the lactoid reactions consists in increasing the time of out-of-hospital management of asymptomatic administration of intravenous vitamin K by giving it coumarin-associated coagulopathy by obviating the in a slow infusion over 15 to 30 minutes.

Load more