Portal de Periódicos da CAPES

Reversal of warfarin-induced over-anticoagulation with individualized dosing of oral vitamin K: a pilot study

2010; Elsevier BV; Volume: 8; Issue: 5 Linguagem: Inglês

10.1111/j.1538-7836.2010.03784.x

1538-7933

Montserrat Briz, Kate Talks, John Hanley, Patrick Kesteven, Ann K. Daly, Peter Avery, Farhad Kamali,

Blood Pressure and Hypertension Studies

Evidence from randomized controlled trials supports the use of low-dose oral vitamin K for the reversal of International Normalized Ratio (INR) in asymptomatic patients [1Lubetsky A. Yonath H. Olchovsky D. Loebstein R. Halkin H. Ezra D. Comparison of oral vs intravenous phytonadione in patients with excessive anticoagulation.Arch Intern Med. 2003; 163: 2469-73Crossref PubMed Scopus (191) Google Scholar, 2Crowther M.A. Julian J. McCarty D. Douketis J. Kovacs M. Biagoni L. Schnurr T. McGinnis J. Gent M. Hirsh J. Ginsberg J. Treatment of warfarin-associated coagulopathy with oral vitamin K: a randomised controlled trial.Lancet. 2000; 356: 1551-3Abstract Full Text Full Text PDF PubMed Scopus (164) Google Scholar, 3Ageno W. Garcia D. Silingardi M. Galli M. Crowther M. A randomized trial comparing 1 mg of oral vitamin K with no treatment in the management of warfarin-associated coagulopathy in patients with mechanical heart valves.J Am Coll Cardiol. 2005; 46: 732-3Crossref PubMed Scopus (33) Google Scholar, 4Dezee K.J. Shimeall W.T. Douglas K.M. Shumway N.M. O’Malley P.G. Treatment of excessive anticoagulation with phytonadione (vitamin K): a meta-analysis.Arch Intern Med. 2006; 166: 391-7PubMed Google Scholar]. However, there is no clear consensus about the optimal dose that should be given to patients. The majority of the reported studies have used fixed oral doses of vitamin K (ranging from 1.0 to 5.0 mg) based upon assigning a fixed vitamin K dose (according to a dichotomy of elevated INR values) to reverse excessive anticoagulation, with a high percentage of patients at 24 h after vitamin K dosing being either undercoagulated or over-anticoagulated, leaving them at risk of either thromboembolism or hemorrhage, respectively [5Sconce E.A. Kamali F. Appraisal of current vitamin K dosing algorithms for the reversal of over-anticoagulation with warfarin: the need for a more tailored dosing regimen.Eur J Haematol. 2006; 77: 457-62Crossref PubMed Scopus (39) Google Scholar, 6Denas G. Marzot F. Offelli P. Stendardo A. Cucchini U. Russo R. Nante G. Iliceto S. Pengo V. Effectiveness and safety of a management protocol to correct over-anticoagulation with oral vitamin K: a retrospective study of 1,043 cases.J Thromb Thrombolysis. 2009; 27: 340-7Crossref PubMed Scopus (31) Google Scholar]. The anticoagulation response to warfarin is influenced by a number of clinical, environmental and genetic factors [7Sconce E.A. Khan T.I. Wynne H.A. Avery P. Monkhouse L. King B.P. Wood P. Kesteven P.J. Daly A.K. Kamali F. The impact of CYP2C9 and VKORC1 polymorphisms and patient characteristics upon warfarin dose requirements: proposal for a new dosing regimen.Blood. 2005; 106: 2329-33Crossref PubMed Scopus (845) Google Scholar] that could influence the vitamin K dose requirement. This pilot study evaluated the effectiveness of a novel ‘tailored’ oral vitamin K dosing regimen for the reversal of non-urgent over-anticoagulation based on the individual patient’s INR. The influences of patient age, sex, body weight, height, target INR, warfarin daily dose and CYP2C9 and VKORC1 polymorphisms on the extent of INR reversal were also retrospectively determined. The ‘tailored’ vitamin K dosing algorithm was developed on the basis of a best fit bivariate regression model containing data on venous INR values at presentation on day 0 and on day 1 for 84 patients after oral vitamin K administration (1 or 2 mg), according to current anticoagulation reversal guidelines in place in the Northern Region of England [8Hanley J.P. Warfarin reversal.J Clin Pathol. 2004; 57: 1132-9Crossref PubMed Scopus (220) Google Scholar]. The regression model produced the following equation for estimation of vitamin K dose: 1 where: 2 Change in INR = initial venous INR at entry – venous INR at 24 h (for vitamin K dose calculation, the INR after 24 h was set as that of the target value). Patients on warfarin therapy with a confirmed venous index INR ≥ 6.0 and either asymptomatic or with minor bleeds, according to the classification of Fihn et al. [9Fihn S.D. Callahan C.M. Martin D.C. McDonell M.B. Henikoff J.G. White R.H. The risk for and severity of bleeding complications in elderly patients treated with warfarin. The National Consortium of Anticoagulation Clinics.Ann Intern Med. 1996; 124: 970-9Crossref PubMed Scopus (545) Google Scholar], were eligible to take part (day 0). The vitamin K1 (phytomenadione; Konakion; Roche, Milan, Italy) dose was calculated on the basis of the dosing algorithm, using a simple conversion chart. Unless already taken, the day’s warfarin dose was withheld and the patient was discharged. Venous INR was determined at 24 h after vitamin K administration (day 1). Statistical analyses were conducted using Minitab version 15 (Minitab Inc., Coventry, UK). Data were compared using anova (having been log transformed to obtain approximate normality) or chi-square tests as applicable. The change in INR value was used as the dependent variable, and age, gender, weight, height, body surface area (BSA), warfarin maintenance dose, diagnosis, target INR and CYP2C9 and VKORC1 polymorphisms as possible explanatory variables, in a multiple regression analysis. Some patients were studied more than once, and this generated a small amount of dependence in the data. Therefore, for comparisons, a P-value of < 0.01 was chosen as statistically significant to ensure that any effect was not due to this small amount of dependence. Sixty-nine patients (33 males) were recruited. Some of these presented with further episode(s) of over-anticoagulation at a later date (outside the period of study monitoring) and were studied again. There were 87 episodes in the 69 patients, with 55 having only one, 11 having two, two having three and one having four episodes of over-anticoagulation that required oral vitamin K administration. The patients’ baseline characteristics are shown in Table 1A. Seventeen patients (19.5%) had already taken their warfarin daily dose on day 0 prior to vitamin K administration.Table 1The patients’ baseline characteristics(A) VariableAge (years), mean (± SD) (range)65.2 (16.2) (32–92)Sex, n (%)Female49 (56.3)Male38 (43.7)Indication for anticoagulation, n (%)Atrial fibrillation32 (36.8)Venous thromboembolism33 (37.9)Valvular prosthesis20 (23)Other2 (2.3)Therapeutic INR target, n (%)2.0–3.054 (62.1)2.5–3.514 (16.1)3.0–4.019 (21.8)Bleeding symptoms at presentation, n (%)None71 (81.6)Minor16 (18.4)Warfarin dose (mg), mean (SD)5.6 (3.8)Warfarin withheld on day 0, n (%)Yes70 (80.5)No17 (19.5)Initial INR, mean (range)8.07 (6–16)(B) Changes in INR values over timeDay 0(n = 87)Day 1(n = 87)Mean INR (± SD) (range)8.02 (2.24)2.62 (1.01)(6.0–16.0)(1.3–6.4)INR < 1.8, n (%)16 (18.4)INR 1.8–4.0, n (%)62 (71.3)INR 4.1–5.0, n (%)8 (9.2)INR 5.1–6.0, n (%)0INR > 6.0, n (%)1 (1.1)INR in target range, n (%)33 (37.9)INR above target range, n (%)17 (19.5)INR below target range, n (%)37 (42.5)INR, International Normalized Ratio; SD, standard deviation. Open table in a new tab INR, International Normalized Ratio; SD, standard deviation. The mean initial INR (day 0) was 8.02 ± 2.2 (range 6.0–16.0). The mean vitamin K dose administered was 1.7 ± 0.4 mg (range 1.1–3.4 mg). The median time for INR determination on day 1 was 23.5 h (range 20.5–28.0 h). The mean INR on day 1 was 2.6 ± 1.0 (median 2.4; range 1.3–6.4). INR values on entry (day 0) and after vitamin K administration (day 1) are shown in Table 1B. Among the 87 episodes analyzed, INR was within range on day 1 in 33 (37.9%), above range in 17 (19.5%), and below range in 37 (42.5%). For 62 episodes (71.3%), INR was between 1.8 and 4.0, and for 16 (18.4%) it was overcorrected to < 1.8. For nine episodes (10.3%), INR was above 4 (for eight it was between 4.0 and 6.0, and for one it was > 6.0). There were no significant differences in INR values on day 0 between those patients whose INR was above range, within range or below range on day 1. In 84 over-anticoagulated patients who had previously been given a fixed oral dose of vitamin K (1 or 2 mg), INR was within range on day 1 in 21%, above range in 74%, and below range in 5%. For 39%, INR was between 1.8 and 4.0, and for 4% it was < 1.8. For 57% of the patients, INR was above 4.0 (for 43% it was between 4.0 and 6.0, and for 13% it was > 6.0). The proportion of patients receiving the tailored vitamin K dose whose INR was within 1.8–4.0 on day 1 was significantly higher than those who received the fixed-dose vitamin K (P < 0.001). Six patients considered to be at high risk of recurrent thrombosis (INR < 1.8; range 1.5–1.7) on day 1 received treatment with low molecular weight heparin (LMWH) until their INR reached the therapeutic range. In four patients (three with venous thromboembolism, and one with atrial fibrillation), warfarin was stopped either immediately or sometime within the next 2 weeks after vitamin K administration, as it was considered by the responsible clinician that the risk of treatment with warfarin outweighed its benefits; one of these patients subsequently received LMWH. Four patients were admitted to hospital within 24 h after vitamin K administration; two because of symptoms related to underlying malignancy, one with major bleeding (traumatic hemorrhagic pleural effusion), and one because of heart and renal failure. The change in INR (INR on day 0 - INR on day 1) was positively associated with vitamin K dose, as expected, and initial INR (P < 0.001 for both). The change in INR was negatively associated with BSA (P < 0.001), after allowance for vitamin K dose and initial INR. This concurs with an earlier study [10Kelly K.C. Raj G. Weideman R.A. Influence of body weight on response to subcutaneous vitamin K administration in over-anticoagulated patients.Am J Med. 2001; 110: 623-7Abstract Full Text Full Text PDF PubMed Scopus (6) Google Scholar]. Being female is independently associated with INR overcorrection [6Denas G. Marzot F. Offelli P. Stendardo A. Cucchini U. Russo R. Nante G. Iliceto S. Pengo V. Effectiveness and safety of a management protocol to correct over-anticoagulation with oral vitamin K: a retrospective study of 1,043 cases.J Thromb Thrombolysis. 2009; 27: 340-7Crossref PubMed Scopus (31) Google Scholar]. One possible reason for such an association is that women usually have a smaller stature (being lighter and shorter) and therefore have a smaller BSA than men. In our study, we found that overcorrection of INR was more frequent in women than in men (23 vs. 14), although the differences were not statistically significant, probably because of the small number of patients evaluated. Although it has been reported that elderly patients have a lower spontaneous decay in INR when warfarin dose is simply withheld [11Hylek E.M. Regan S. Go A.S. Hughes R.A. Singer D.E. Skates S.J. Clinical predictors of prolonged delay in return of the international normalized ratio to within the therapeutic range after excessive anticoagulation with warfarin.Ann Intern Med. 2001; 135: 393-400Crossref PubMed Scopus (167) Google Scholar], and that warfarin daily maintenance dose is inversely correlated with the fall in logarithm of the INR [12White R.H. Minton S.M. Andya M.D. Hutchinson R. Temporary reversal of anticoagulation using oral vitamin K.J Thromb Thrombolysis. 2000; 10: 149-53Crossref PubMed Scopus (14) Google Scholar], we found no association between either patient age or warfarin dose and the change in INR subsequent to vitamin K administration, unadjusted or after allowance for vitamin K dose, initial INR, and BSA. There was no association between the change in INR, unadjusted or adjusted, and patient sex, weight or height. Neither the change in INR (adjusted or unadjusted) nor its outcome at 24 h was associated with CYP2C9 or VKORC1 genotype. However, given that CYP2C9 and VKORC1 affect patient sensitivity to warfarin, it is worth investigating further the possible influence of these genes on vitamin K dose requirement in a larger patient population. The data generated were used to obtain a new formula for predicting vitamin K dose, as follows: Vitamin K dose = [0.247 (initial INR on entry) – (target INR-0.1320) + (1.417 × BSA)]/1.0135. This formula predicts, on average, essentially the same vitamin K dose for the group of patients for whom the INR was restored to within their target range (median dose change, 0.0; range, − 0.8 to + 0.8 mg), but a marked reduction in the predicted dose in the overcorrected group (median dose change, − 0.5; range, − 1.0 to + 0.7 mg), and a small rise in predicted dose in the undercorrected group (median dose change, + 0.1; range, − 0.9 to + 1.0 mg). In conclusion, we have shown in this pilot study that a tailored approach to vitamin K dosing can lead to a more predictable reversal of elevated INR than a fixed-dose regimen. A recent randomized controlled study involving 712 asymptomatic over-anticoagulated patients (INR range of 4.0–10.0) showed that whereas a fixed dose of 1.25 mg of oral vitamin K was more effective in reversing elevated INR than placebo, there were no significant differences in the incidence of bleeding between the two treatments [13Crowther M.A. Ageno W. Garcia D. Wang L. Witt D.M. Clark N.P. Blostein M.D. Kahn S.R. Vesely S.K. Schulman S. Kovacs M.J. Rodger M.A. Wells P. Anderson D. Ginsberg J. Selby R. Siragusa S. Silingardi M. Dowd M.B. Kearon C. Oral vitamin K versus placebo to correct excessive anticoagulation in patients receiving warfarin: a randomized trial.Ann Intern Med. 2009; 150: 293-300Crossref PubMed Scopus (136) Google Scholar]. It remains to be seen whether a tailored approach to vitamin K dosing through a more controlled reversal of INR can reduce the incidence of bleeding in over-anticoagulated patients. The present tailored dosing algorithm should be refined to further increase its accuracy for vitamin K dose prediction. On the basis of the data generated, we have devised a new dosing algorithm that calculates vitamin K dose according to BSA, index INR, and target therapeutic INR. Further studies are required to evaluate the performance of the new dosing algorithm in reversing excessive anticoagulation. The authors are grateful to the nurses and secretaries of the Thrombosis and Haemophilia Centre at the Royal Victoria Infirmary for their help with patient management. This work was supported by the UK NIHR Biomedical Research Centre for Ageing and Age-related disease award to the Newcastle upon Tyne Hospitals NHS Foundation Trust. M. Briz was supported by the IFIMAV Institute of the Fundación Pública Marqués de Valdecilla.