- Email: [email protected]
Reduced-dose warfarin or interrupted warfarin with heparin bridging for pacemaker or defibrillator implantation: A randomized trial
Thrombosis Research 134 (2014) 814–818
Contents lists available at ScienceDirect
Thrombosis Research journal homepage: www.elsevier.com/locate/thromres
Regular Article
Reduced-dose warfarin or interrupted warfarin with heparin bridging for pacemaker or defibrillator implantation: A randomized trial S. Schulman a,b,e,⁎, J.S. Healey b,c,d, J.D. Douketis a,b, J. Delaney b, C.A. Morillo b,c,d a
Thrombosis and Atherosclerosis Research Institute, McMaster University, Hamilton, ON, Canada Department of Medicine, Thrombosis Service, McMaster University, Hamilton, ON, Canada Arrhythmia Services, Cardiology Division, McMaster University, Hamilton, ON, Canada d Population Health Research Institute, McMaster University, Hamilton, ON, Canada e Karolinska Institutet, Stockholm, Sweden b c
a r t i c l e
i n f o
Article history: Received 1 May 2014 Received in revised form 9 July 2014 Accepted 21 July 2014 Available online 1 August 2014 Keywords: Atrial fibrillation Anticoagulation Pacemaker Defibrillator Surgery
a b s t r a c t Background: Perioperative management with reduced-dose warfarin is of potential interest by eliminating the need for bridging while still maintaining a degree of anticoagulation. The outcomes of this regimen have not been well determined. Methods: In a randomized controlled trial we compared two regimens for management of anticoagulation with warfarin in patients with implantation of a pacemaker or defibrillator. Half dose of warfarin for 3–6 days, depending on the baseline international normalized ratio (INR), before surgery aiming at an INR of ≤1.7 was compared with interrupted warfarin for 5 days with preoperative bridging with low-molecular-weight heparin (LMWH) at therapeutic dose for 2.5 days. Main safety outcome was pocket hematoma. Secondary outcomes were major bleeding, thromboembolism – all within 1 month, days of hospitalization and number of patients requiring correction of INR with vitamin K. Results: The study was planned for 450 patients but it was discontinued prematurely due to a change in practice. Pocket hematoma occurred in 4 of 85 patients (5%) randomized to the bridged regimen and in 3 of 86 patients (3%) randomized to reduced-dose warfarin. One pocket hematoma in each group was severe. There were no major hemorrhages or thromboembolism within the 1-month window. Duration of hospitalization was similar in the two groups. Correction of INR the day before surgery with vitamin K had to be used for significantly more patients in the reduced-dose warfarin group (41%) than in the bridged regimen group (6%). Conclusion: The reduced-dose warfarin regimen appeared to have similar safety after device implantation as interrupted warfarin with preoperative LMWH bridging. Due to premature discontinuation no firm conclusion can be drawn. The reduced-dose warfarin regimen often failed to achieve the intended preoperative INR. ClinicalTrials.gov Identifier: NCT 02094157 © 2014 Elsevier Ltd. All rights reserved.
Introduction Among patients who have implantation of a pacemaker or implantable cardioverter defibrillator (ICD) device, 35-40% are receiving oral anticoagulant therapy [1,2], mainly for stroke prophylaxis in atrial fibrillation. This creates a challenge in the balance between minimizing the risk for anticoagulation-associated perioperative bleeding complications and the risk for thromboembolism by withholding anticoagulation. The most common hemorrhagic complication is pocket hematoma, reported in 4.9% of patients [2]. This may require exploration and evacuation of the hematoma and prolonged hospitalization [2,3]. The risk of ⁎ Corresponding author at: Thrombosis Service, HHS-General Hospital, 237 Barton Street East, Hamilton, ON, L8L 2X2, Canada. Tel.: +1 905 5270271x44479; fax: +1 905 5211551. E-mail address: [email protected] (S. Schulman).
http://dx.doi.org/10.1016/j.thromres.2014.07.028 0049-3848/© 2014 Elsevier Ltd. All rights reserved.
thromboembolic complications is 0-2% [4–6] but it might be higher in patients with risk factors when anticoagulation is interrupted without bridging with heparin. In a systematic review of peri-procedure anticoagulation strategies the incidence of pocket hematoma was 1220% for bridging anticoagulation with therapeutic-dose heparin and 1.9-6.6% with perioperative continuation of vitamin K antagonists (VKA) [5]. We showed in a previous study that bridging with therapeutic-dose low-molecular-weight heparin (LMWH) before and after implantation compared to only before surgery reduced the incidence of pocket hematoma from 23% to 8% [7]. A perioperative strategy with reduced-dose VKA aims at lowering the international normalized ratio (INR) to 1.5-2.0 at the time of the procedure [7–9]. The rationale for reduced-dose VKA is to provide thromboembolic protection without an increase in the risk of bleeding and to avoid the inconvenience and expense of heparin injections. This regimen was evaluated in one previous randomized trial in 131
S. Schulman et al. / Thrombosis Research 134 (2014) 814–818
patients with dental surgery [10]. The reduced dose of VKA aimed at an INR of 1.8 but the information on the regimen was limited in this brief report. Against this background, we compared in this randomized trial reduced-dose warfarin versus interrupted warfarin combined with preoperative therapeutic-dose LMWH in patients at increased risk of thromboembolism and who had implantation of a pacemaker or ICD. Our main objective was to establish whether the reduced-dose regimen is safe in terms of risk of pocket hematoma. Methods The study had a randomized, parallel, open-label design. Patients planned for device implantation and on treatment with warfarin were routinely referred from the Arrhythmia Service to the Thrombosis Service for perioperative management of anticoagulation. Patients were eligible for inclusion if they had been on warfarin for more than one week and were planned to remain on this drug for at least one month after elective implantation of a pacemaker or ICD. Exclusion criteria were age b 18 years, previous stroke, a CHADS2 score of 0 or 1 for patients with atrial fibrillation as the only indication for warfarin, calculated creatinine clearance of b30 mL/min using the CockroftGault formula [11], surgery scheduled for the first day after a weekend or holiday, patient assessed by physician as unsuitable for the study (history of poor compliance, psychiatric disorder, cognitive impairment, language barrier), participation in another clinical trial or failure to obtain informed consent. Study Treatments Patients were seen at baseline when they had the appointment at our Perioperative Anticoagulation Management Clinic, and this was scheduled for 6–8 days before surgery. Eligible patients were randomized using a web-based computer-generated random number table with random block sizes. Patients randomized to warfarin interruption with preoperative bridging received LMWH syringes (enoxaparin, Lovenox®, Sanofi-Aventis) from the center, or by prescription if they had coverage for the expense. The dose was 1 mg/kg rounded off to the nearest lower multiple of 20 (corresponding to the marketed prefilled syringes) twice daily on day 3 and 2 before surgery and once in the morning on day 1 before surgery. For patients with creatinine clearance of 30–59 mL/min the LMWH was given only once daily for the 3 days. Warfarin was held for 5 days before surgery (6 days if the INR at the preoperative visit = baseline was N3.5). Warfarin was resumed in the evening of the implantation at twice the maintenance dose for 2 days, followed by the usual maintenance dose. Patients randomized to reduced-dose warfarin were instructed to take before the implantation half their usual maintenance dose for 3 days if baseline INR was 1.7-1.8, 4 days if INR was 1.9-2.4, 5 days if INR was 2.5-2.9, 6 days if INR was 3.0-3.4 and 7 days if INR was 3.6-4.4. This regimen was based on modelling of data from 70 previous patients receiving reduceddose warfarin before different surgeries at our hospital [12] and aimed at achieving an INR of ≤1.7 for the surgery. Warfarin was resumed in the evening of the implantation at twice the maintenance dose, followed by the usual maintenance dose. The study nurse gave patients explicit oral and written instructions on the warfarin regimen and for the bridged patients also on LMWH dosing and injection technique. They also received one ampoule of vitamin K, 1 mg, for provisional use orally on the day before surgery if the INR that day was N1.7. Antiplatelet therapy was permitted if indicated. For patient on longterm treatment with aspirin without active coronary artery disease, aspirin was held for 3 days (aspirin 81 mg)[13,14] or 7 days (aspirin 325 mg) before surgery. Patients on clopidogrel but without a recent stent had the drug held for 5 days before surgery.
815
Follow-up We obtained a “same-day” INR the day before surgery and then monitored the INR and dosed the warfarin postoperatively until the therapeutic range was reached, after which the pre-study management was resumed. Before discharge from the hospital, a physician responsible for the implantation or a nurse assessed the patient regarding signs or symptoms related to possible pocket hematoma or major bleeding outcomes. The physician/nurse was not involved with the preoperative bridging and was unaware of the regimen the patient was allocated to. After 3 ± 1 weeks and at 3 months the patient returned for a visit at the Pacemaker/ICD Clinic with inspection of the incision and the generator site by a nurse, unaware of the regimen the patient was allocated to. These visits coincided with routine clinical follow-up of the device. End Point Evaluation Events qualifying for end points were those with onset from the first day after stopping or reducing the dose of warfarin until the visit at the Pacemaker/ICD clinic 3 ± 1 weeks after the implantation. A hematoma or a thromboembolic complication with onset after that time point was deemed highly unlikely to be associated with the study medication period. The primary outcome was pocket hematoma, defined and identified by the blinded Pacemaker/ICD clinic staff as palpable swelling of the pacemaker/ICD pocket exceeding the size of the generator [2]. Severe pocket hematoma was defined when there was in addition one or more of pain, prolonged hospitalization or requirement for re-operation. Secondary outcomes were major bleeding as previously defined [15], arterial or venous thromboembolism that was symptomatic and objectively verified, and death. In addition, we collected data on length of hospitalization and proportion of patients requiring reversal with vitamin K on the day before surgery to achieve INR of 1.7 or lower. Statistical Analyses Our sample size estimate was predicted on demonstrating non-inferiority for the outcome of pocket hematoma, which is clinically important because of its potentially serious consequences, e.g., infected device pocket or wires, evacuation, prolonged hospitalization. The estimated incidence of pocket hematoma for our standard regimen (stopping warfarin for 5 days, preoperative bridging with LMWH) was 4.8% (95% CI: 1.1-8.5), based on our own data from 125 consecutive patients during one year. Our null hypothesis was that the ‘reduced-dose warfarin' strategy is in fact worse than the 'warfarin interruption' strategy (i.e. leads to a higher rate of pocket hematoma and is therefore inferior). If the risk of bleeding does not exceed 9.8%, resulting in a δ of 5% or essentially a doubling of the event rate compared to the conventional 'warfarin interruption' approach, the experimental treatment would still be acceptable in view of lower costs, improved convenience and the modest harm caused by the majority of pocket hematomas. With a one-sided α of 5% and a power of 80%, we needed 225 evaluable patients per group. We discontinued the study prematurely when the Arrhythmia Service changed their routine and allowed therapeutic anticoagulation with warfarin during device implantation, based on the results of the BRUISE CONTROL study [16]. Interim analyses for safety were planned for every 100 patients randomized to review the incidence of serious events (stroke, major bleeding or death), applying pre-defined stopping guidelines. Binomial data were analyzed with chi square test or in case of less than 10 in any numerator with Fisher’s exact test. The Mann–Whitney U-test was used for quantitative values with skewed distribution and ttest for normal distribution. Results of continuous variables are expressed as mean ± standard deviation (SD) in case of normal distribution and as median and interquartile range (IQR) for skewed distributions.
816
S. Schulman et al. / Thrombosis Research 134 (2014) 814–818
The Research Ethics Boards of McMaster University and Hamilton Health Sciences/St Joseph’s Healthcare approved the study. No external funding was obtained but study enoxaparin was provided by SanofiAventis, Canada. This company had no involvement in the design, conduct or analysis of the study. The trial was registered at ClinicalTrials.gov NCT 02094157.
Results Between December 2007 and July 2013 we randomized 172 patients, 169 at Hamilton Health Sciences-General Hospital and 3 at St. Joseph’s Healthcare, both in Hamilton, Ontario. The flow of the patients is shown in Fig. 1. One patient allocated to reduced-dose warfarin had her ICD implantation postponed 10 days and in the meantime her cardiologist switched the anticoagulant therapy to dabigatran, excluding her from the outcome analyses. Baseline characteristics were not significantly different between the two groups with the exception for concomitant treatment with clopidogrel (Table 1). All 5 patients with clopidogrel in the bridged regimen group had this drug stopped for 5 days before surgery. The surgical characteristics, i.e. pacemakers vs. ICDs and first vs. replacement surgery, did not differ between the groups (Table 2). A mean of 5 doses LMWH (range 2–7) were given during the days before surgery according to the bridged regimen and 4 of the patients required home visits by a nurse for these injections. Although warfarin was stopped for 5 days in the group with bridged regimen, the INR on the day before surgery was still mildly elevated with a mean of 1.4. Four patients still had INR ≥2.0, the highest one being 2.4. With the reduceddose regimen 22 patients still had INR ≥2.0 on the day before surgery. In fact, INR increased from the baseline visit to the day before surgery by 0.1 to 0.6 in 14 patients in the reduced warfarin group versus 1 patient (by 0.4) in the bridged group. One mg of vitamin K was given on that day as per protocol to patients with INR ≥1.8 – 5 in the bridged regimen and 35 patients in the reduced-dose warfarin group. On the day of surgery the INR had decreased to b 1.8 in all these patients. One patient (reduced-dose warfarin) with INR 2.2 the day before surgery, treated with vitamin K and reduced to 1.7 on the day of surgery was given prothrombin complex concentrate 1000 units because minithoracotomy was required. Another 2 patients in the reduced-dose warfarin group had an INR the day before surgery of 2.0-2.5 and the procedure was postponed – twice for one of the patients– and the dose reduction was repeated with finally acceptable INR of 1.5.
Not randomized Did not meet criteria: 120 Paent declined: 49
Table 1 Characteristics of the patients.
Age, years, mean (±SD) Female sex, n (%) Indication for warfarin, n (%) AF alone AF & mechanical valve AF and bioprosthetic valve AF and left ventricle thrombus AF & cardiomyopathy AF & venous thromboembolism Mechanical valve Left ventricle thrombus Cardiomyopathy Venous thromboembolism CHADS2 score, median (IQR)⁎ Creatinine clearance, mL/min, mean (±SD) Hemoglobin, g/L, mean (±SD) History of major bleeding, n (%) Years on warfarin, median (IQR) Dose of warfarin, mg/week, mean (±SD) Therapeutic range, n (%) INR 2.0-3.0 INR 2.5-3.5 INR at preoperative visit, mean (±SD) Aspirin therapy, n (%) Continued perioperatively, n (%) Clopidogrel therapy, n (%)
Bridged regimen
Reduced warfarin
N = 85
N = 87
74 (9.6) 21 (25)
74 (10.9) 22 (25)
0.95 0.93
69 (81) 6 (7) 1 (1) 2 (2) 2 (2) 1(1) 3 (4) 1 (1) 0 0 3.0 (2–3) 69 (27)
69 (79) 5 (6) 0 0 1 (1) 1 (1) 3 (3) 2 (2) 4 (5) 2 (2) 3.0 (2–3) 68 (25)
0.78
0.49 0.83
137 (16) 5 (6) 5.0 (1–10) 31 (14)
135 (15) 4 (5) 5.0 (1–10) 32 (14)
0.46 0.74 0.93 0.53
80 (94) 5 (6) 2.2 (0.5) 32 (37) 19 (22) 5 (6)
84 (97) 3 (3) 2.2 (0.5) 34 (39) 19 (22) 0
0.49
The primary outcome, pocket hematoma occurred in 4 patients (5%) with bridged regimen and in 3 (3%) with reduced-dose warfarin (odds ratio 1.37; 95% confidence interval, 0.22-9.60) (Table 3). For one patient in each group the pocket hematoma was classified as severe due to pain but no hospitalization or revision was required. There was only one major bleed in the study, a subdural and subarachnoid hemorrhage 55 days after surgery and considered unrelated. One patient had ischemic stroke 48 days after surgery (INR 2.0 on admission) and with fatal
Randomized 172 paents
Reduced warfarin 87 paents Switched to dabigatran preop: 1 Incomplete follow-up: 1
Incomplete follow-up: 0
Analyzed 85 paents
0.65 0.82 0.93 0.03
SD – standard deviation; AF – atrial fibrillation; CHADS2 – a score giving 1 point each for presence of congestive heart failure, hypertension, age N75 and diabetes and 2 points for stroke or transient ischemic attack; IQR – interquartile range; INR – international normalized ratio ⁎For the patients with atrial fibrillation
Screened 341 paents
Bridged regimen 85 paents
P-value
Analyzed 86 paents
Fig. 1. Flow of patients in the study.
S. Schulman et al. / Thrombosis Research 134 (2014) 814–818 Table 2 Surgery and perioperative anticoagulation.
Device implanted, n (%) Pacemaker ICD First implantation, n (%) Days off vs days reduced-dose warfarin, mean (±SD) Doses of LMWH, median (IQR) INR Day −1, mean (±SD) Range Vitamin K 1 mg Day −1, n (%) Per protocol given Actually given Days to first therapeutic INR, mean (±SD) 1st postoperative INR, mean (±SD) 1st postoperative INR ≥2.0, n (%) Days until 1st INR, mean (±SD)
Bridged regimen
Reduced warfarin
P-value
N = 85
N = 86
44 (52) 41 (48) 42 (49) 5.0 (0.7)
42 (49) 44 (51) 40 (47) 4.3 (1.2)
5.0 (3–5) 1.4 (0.3) 1.0-2.4
Not given 1.7 (0.3) 1.1-2.6
b0.001
5 (5.9) 5 (5.9) 12 (8.5) 1.9 (0.4) 29 (34) 5.2 (2.0)
35 (41) 40 (47) 11 (8.4) 2.0 (0.5) 38 (48) 6.1 (3.6)
b0.001 b0.001 0.81 0.15 0.07 0.19
0.70
0.70 b0.001
ICD – implantable cardioverter defibrillator; SD – standard deviation; LMWH – low-molecular-weight heparin; IQR – interquartile range; INR – international normalized ratio
outcome after 2 days but also considered unrelated. There were 2 nonfatal serious adverse events in each group. Time to first therapeutic INR after surgery and the mean INR at the first sampling occasion (median 5 days after surgery) were similar in the two groups although there was a trend to a higher proportion of INRs ≥2.0 in the reduced-dose warfarin group at the first sampling. Discussion Preoperative reduction of warfarin has previously to our knowledge and a MEDLINE search only been explored in 2 cohort studies [8,9] and in one randomized trial that was briefly reported in a letter to the editor [10]. We found that reduced-dose warfarin, aiming for an INR of 1.5, was associated with a similar, low risk of pocket hematoma after implantation of pacemaker or ICD as a regimen with stopped warfarin and only preoperative bridging with LMWH. Unfortunately, the study had to be prematurely discontinued due to a change in practice and it therefore lacks power to allow for conclusions on non-inferiority. The event rate of pocket hematoma was low in both treatment groups (3% and 5%). Reduced-dose warfarin avoids the inconvenience of subcutaneous injections with heparin but appears to be a regimen that is difficult to comprehend by some patients, as reflected by the significantly higher
817
proportion of patients with increase in the INR compared to baseline and/or with INR still in the therapeutic range the day before surgery and requiring correction with 1 mg of vitamin K. We did, however, not ask the patients to keep diaries of the warfarin dosing, which might have contributed to illustrate this problem. The recent publication of the BRUISE CONTROL trial [16] resulted in a change of paradigm since pacemaker or ICD implantation at therapeutic INR, i.e. without any interruption of warfarin, was safe with a rate of pocket hematoma of 3.5%. This is identical to the rate with reduceddose warfarin in our study. Other studies on maintained warfarin before device implantation reported a hematoma rate of 1.9-24% [17–24]. When only studies with at least 100 patients are included hematoma rate ranged between 1.9% [19] and 7.7% [23]. Conversely, bridging with heparin or LMWH before and after surgery is associated with an unacceptable high rate of pocket hematoma of 16-23% [7,16] and should be abandoned. Bridging only preoperatively was previously reported by us to reduce this risk to 8% [7] but subsequent data from our in-house registry generated a rate of 4.8%. This was the basis for our sample size calculation and indeed, the rate in the bridged regimen group in this study was 4.7%. For a comparison, in patients not on any anticoagulation we have a mean pocket hematoma rate of 2.6%. The higher rate of pocket hematoma with postoperative parenteral anticoagulation than with therapeutic warfarin throughout the procedures might be due to peak levels of heparin resulting in a stronger anticoagulant effect than the even influence of warfarin. It is unclear whether the therapeutic INR results the day before surgery are explained only by misunderstanding of the reduced warfarin dosing. In the group with interrupted warfarin and preoperative LMWH bridging there were also patients with such INRs, albeit significantly fewer. Another possibility is slow metabolism of VKA due to microsomal enzyme polymorphisms (mainly CYP2C9). Yet another reason could be that reduced-dose warfarin should be done with less than half the maintenance dose or for more days than in our regimen, which was based on our previous experience from 70 patients with various procedures [12]. Since we obtained the preoperative INR on the day before surgery it is possible that a higher proportion of patients with reduced-dose warfarin would have had INRs below 1.7 if measured in the morning of the surgery. Notwithstanding the mechanism, a simplification by maintaining warfarin therapy unchanged for device implantation is probably the best solution. If reduced-dose warfarin is to be used for other types of invasive procedures, the regimen should include a day-before-surgery INR with ability to administer 1 mg of vitamin K orally when INR results below the therapeutic range are considered necessary for safety. Future research should focus on simple solutions by identifying more surgeries that do not require interruption of warfarin and by establishing whether bridging really provides any net benefit
Table 3 Clinically important outcomes. Bridged regimen N = 85
Reduced warfarin N = 86
OR (95% CI)
Pocket hematoma, n (%)
4 (5)
3 (3)
1.4 (0.22-9.6)
Severe pocket hematoma, n (%) Major bleeding, n (%) Thromboembolic event, n (%) Serious adverse events within 1 month, n (%)
1 (1) 1⁎ 0 2†
1 (1) 0 1⁎ 2†
Death Re-hospitalization Discharge from hospital same day, n (%)‡ Nights of postop hospitalization, mean (±SD)§
0 2 72 (85) 0.2 (0.4)
0 2 67 (78) 0.3 (0.8)
1.0 (0.7-14)
OR – odds ratio; CI – confidence interval; SD – standard deviation ⁎ These events occurred more than 1 month after surgery † The serious adverse events were in the bridged regimen group: Ventricular lead dislodgement and heart failure, and in the reduced-dose warfarin group: odontoid fracture and pleural effusion. The last one started after randomization but 5 days before the initiation of study-related treatment. ‡ Chi-square test, P = 0.29 § Mantel-Haenszel test comparing the ranks of the observations, P = 0.22
818
S. Schulman et al. / Thrombosis Research 134 (2014) 814–818
for patients having major surgery that cannot be performed on full anticoagulation, as currently explored in the BRIDGE trial (ClinicalTrials.gov NCT00786474). The main limitation of this study is the lack of power due to its premature discontinuation. It was also essentially a single-center study with 98% of patients recruited at one site. Another limitation is the open-label design. The nurses assessing the presence of pocket hematoma at the Pacemaker- or ICD Clinics were, however, unaware of the perioperative anticoagulation regimen. In conclusion, the reduced-dose warfarin regimen, aiming for an INR of ≤1.7 on the day of device implantation had similar safety with regards to pocket hematoma as interrupted warfarin with preoperative LMWH bridging. The reduced-dose regimen failed to achieve INR of ≤1.7 the day before surgery in 41% of the patients. This illustrates the imprecision of manipulating warfarin dosing to achieve non-therapeutic targets. Reduced-dose warfarin might still be an alternative for a suitable subset of patients having procedures that carry too high risk of bleeding on full anticoagulation, provided that they can check the INR the day before surgery and correct with vitamin K. Conflict of Interest Statement The authors have no conflicts of interest in relation to this article Appendix A. Supplementary Data PACEBRIDGE Study Protocol can be found online at http://dx.doi.org/ 10.1016/j.thromres.2014.07.028. References [1] McComb JM, Gribbin GM. Chronic atrial fibrillation in patients with paroxysmal atrial fibrillation, atrioventricular node ablation and pacemakers: determinants and treatment. Europace 1999;1:30–4. [2] Wiegand UK, LeJeune D, Boguschewski F, Bonnemeier H, Eberhardt F, Schunkert H, et al. Pocket hematoma after pacemaker or implantable cardioverter defibrillator surgery: influence of patient morbidity, operation strategy, and perioperative antiplatelet/anticoagulation therapy. Chest 2004;126:1177–86. [3] Kiviniemi MS, Pirnes MA, Eranen HJ, Kettunen RV, Hartikainen JE. Complications related to permanent pacemaker therapy. Pacing Clin Electrophysiol 1999;22:711–20. [4] Aggarwal RK, Connelly DT, Ray SG, Ball J, Charles RG. Early complications of permanent pacemaker implantation: no difference between dual and single chamber systems. Br Heart J 1995;73:571–5. [5] Jamula E, Douketis JD, Schulman S. Perioperative anticoagulation in patients having implantation of a cardiac pacemaker or defibrillator: a systematic review and practical management guide. J Thromb Haemost 2008;6:1615–21.
[6] Phibbs B, Marriott HJ. Complications of permanent transvenous pacing. N Engl J Med 1985;312:1428–32. [7] Robinson M, Healey JS, Eikelboom J, Schulman S, Morillo CA, Nair GM, et al. Postoperative low-molecular-weight heparin bridging is associated with an increase in wound hematoma following surgery for pacemakers and implantable defibrillators. Pacing Clin Electrophysiol 2009;32:378–82. [8] Larson BJ, Zumberg MS, Kitchens CS. A feasibility study of continuing dose-reduced warfarin for invasive procedures in patients with high thromboembolic risk. Chest 2005;127:922–7. [9] Marietta M, Bertesi M, Simoni L, Pozzi S, Castelli I, Cappi C, et al. A simple and safe nomogram for the management of oral anticoagulation prior to minor surgery. Clin Lab Haematol 2003;25:127–30. [10] Sacco R, Sacco M, Carpenedo M, Moia M. Oral surgery in patients on oral anticoagulant therapy: a randomized comparison of different INR targets. J Thromb Haemost 2006;4:688–9. [11] Cockcroft DW, Gault MH. Prediction of creatinine clearance from serum creatinine. Nephron 1976;16:31–41. [12] Schulman S, Earl B, Robinson M. Perioperative management of anticoagulation with tapered dose warfarin. Haematologica 2008;93(s1):147–8. [13] Zisman E, Erport A, Kohanovsky E, Ballagulah M, Cassel A, Quitt M, et al. Platelet function recovery after cessation of aspirin: preliminary study of volunteers and surgical patients. Eur J Anaesthesiol 2010;27:617–23. [14] Li C, Hirsh J, Xie C, Johnston MA, Eikelboom JW. Reversal of the anti-platelet effects of aspirin and clopidogrel. J Thromb Haemost 2012;10:521–8. [15] Schulman S, Angeras U, Bergqvist D, Eriksson B, Lassen MR, Fisher W. Definition of major bleeding in clinical investigations of antihemostatic medicinal products in surgical patients. J Thromb Haemost 2010;8:202–4. [16] Birnie DH, Healey JS, Wells GA, Verma A, Tang AS, Krahn AD, et al. Pacemaker or defibrillator surgery without interruption of anticoagulation. N Engl J Med 2013;368: 2084–93. [17] al-Khadra AS. Implantation of pacemakers and implantable cardioverter defibrillators in orally anticoagulated patients. Pacing Clin Electrophysiol 2003;26:511–4. [18] Denman RA, Walters DL, Twidale N, Masterson ML. The implantation of permanent pacemakers and automatic implantable cardiac defibrillators in patients on warfarin. CSANZ 47 Annual Scientific Meeting; 1999 [www.csanz.edu.au/abstracts/ 47abstracts/041.doc.htm]. [19] Giudici MC, Barold SS, Paul DL, Bontu P. Pacemaker and implantable cardioverter defibrillator implantation without reversal of warfarin therapy. Pacing Clin Electrophysiol 2004;27:358–60. [20] Goldstein DJ, Losquadro W, Spotnitz HM. Outpatient pacemaker procedures in orally anticoagulated patients. Pacing Clin Electrophysiol 1998;21:1730–4. [21] Michaud GF, Pelosi Jr F, Noble MD, Knight BP, Morady F, Strickberger SA. A randomized trial comparing heparin initiation 6 h or 24 h after pacemaker or defibrillator implantation. J Am Coll Cardiol 2000;35:1915–8. [22] Milic DJ, Perisic ZD, Zivic SS, Stanojkovic ZA, Stojkovic AM, Karanovic ND, et al. Prevention of pocket related complications with fibrin sealant in patients undergoing pacemaker implantation who are receiving anticoagulant treatment. Europace 2005;7:374–9. [23] Tischenko A, Gula LJ, Yee R, Klein GJ, Skanes AC, Krahn AD. Implantation of cardiac rhythm devices without interruption of oral anticoagulation compared with perioperative bridging with low-molecular weight heparin. Am Heart J 2009;158:252–6. [24] Tolosana JM, Berne P, Mont L, Heras M, Berruezo A, Monteagudo J, et al. Preparation for pacemaker or implantable cardiac defibrillator implants in patients with high risk of thrombo-embolic events: oral anticoagulation or bridging with intravenous heparin? A prospective randomized trial. Eur Heart J 2009;30:1880–4.
Contents lists available at ScienceDirect
Thrombosis Research journal homepage: www.elsevier.com/locate/thromres
Regular Article
Reduced-dose warfarin or interrupted warfarin with heparin bridging for pacemaker or defibrillator implantation: A randomized trial S. Schulman a,b,e,⁎, J.S. Healey b,c,d, J.D. Douketis a,b, J. Delaney b, C.A. Morillo b,c,d a
Thrombosis and Atherosclerosis Research Institute, McMaster University, Hamilton, ON, Canada Department of Medicine, Thrombosis Service, McMaster University, Hamilton, ON, Canada Arrhythmia Services, Cardiology Division, McMaster University, Hamilton, ON, Canada d Population Health Research Institute, McMaster University, Hamilton, ON, Canada e Karolinska Institutet, Stockholm, Sweden b c
a r t i c l e
i n f o
Article history: Received 1 May 2014 Received in revised form 9 July 2014 Accepted 21 July 2014 Available online 1 August 2014 Keywords: Atrial fibrillation Anticoagulation Pacemaker Defibrillator Surgery
a b s t r a c t Background: Perioperative management with reduced-dose warfarin is of potential interest by eliminating the need for bridging while still maintaining a degree of anticoagulation. The outcomes of this regimen have not been well determined. Methods: In a randomized controlled trial we compared two regimens for management of anticoagulation with warfarin in patients with implantation of a pacemaker or defibrillator. Half dose of warfarin for 3–6 days, depending on the baseline international normalized ratio (INR), before surgery aiming at an INR of ≤1.7 was compared with interrupted warfarin for 5 days with preoperative bridging with low-molecular-weight heparin (LMWH) at therapeutic dose for 2.5 days. Main safety outcome was pocket hematoma. Secondary outcomes were major bleeding, thromboembolism – all within 1 month, days of hospitalization and number of patients requiring correction of INR with vitamin K. Results: The study was planned for 450 patients but it was discontinued prematurely due to a change in practice. Pocket hematoma occurred in 4 of 85 patients (5%) randomized to the bridged regimen and in 3 of 86 patients (3%) randomized to reduced-dose warfarin. One pocket hematoma in each group was severe. There were no major hemorrhages or thromboembolism within the 1-month window. Duration of hospitalization was similar in the two groups. Correction of INR the day before surgery with vitamin K had to be used for significantly more patients in the reduced-dose warfarin group (41%) than in the bridged regimen group (6%). Conclusion: The reduced-dose warfarin regimen appeared to have similar safety after device implantation as interrupted warfarin with preoperative LMWH bridging. Due to premature discontinuation no firm conclusion can be drawn. The reduced-dose warfarin regimen often failed to achieve the intended preoperative INR. ClinicalTrials.gov Identifier: NCT 02094157 © 2014 Elsevier Ltd. All rights reserved.
Introduction Among patients who have implantation of a pacemaker or implantable cardioverter defibrillator (ICD) device, 35-40% are receiving oral anticoagulant therapy [1,2], mainly for stroke prophylaxis in atrial fibrillation. This creates a challenge in the balance between minimizing the risk for anticoagulation-associated perioperative bleeding complications and the risk for thromboembolism by withholding anticoagulation. The most common hemorrhagic complication is pocket hematoma, reported in 4.9% of patients [2]. This may require exploration and evacuation of the hematoma and prolonged hospitalization [2,3]. The risk of ⁎ Corresponding author at: Thrombosis Service, HHS-General Hospital, 237 Barton Street East, Hamilton, ON, L8L 2X2, Canada. Tel.: +1 905 5270271x44479; fax: +1 905 5211551. E-mail address: [email protected] (S. Schulman).
http://dx.doi.org/10.1016/j.thromres.2014.07.028 0049-3848/© 2014 Elsevier Ltd. All rights reserved.
thromboembolic complications is 0-2% [4–6] but it might be higher in patients with risk factors when anticoagulation is interrupted without bridging with heparin. In a systematic review of peri-procedure anticoagulation strategies the incidence of pocket hematoma was 1220% for bridging anticoagulation with therapeutic-dose heparin and 1.9-6.6% with perioperative continuation of vitamin K antagonists (VKA) [5]. We showed in a previous study that bridging with therapeutic-dose low-molecular-weight heparin (LMWH) before and after implantation compared to only before surgery reduced the incidence of pocket hematoma from 23% to 8% [7]. A perioperative strategy with reduced-dose VKA aims at lowering the international normalized ratio (INR) to 1.5-2.0 at the time of the procedure [7–9]. The rationale for reduced-dose VKA is to provide thromboembolic protection without an increase in the risk of bleeding and to avoid the inconvenience and expense of heparin injections. This regimen was evaluated in one previous randomized trial in 131
S. Schulman et al. / Thrombosis Research 134 (2014) 814–818
patients with dental surgery [10]. The reduced dose of VKA aimed at an INR of 1.8 but the information on the regimen was limited in this brief report. Against this background, we compared in this randomized trial reduced-dose warfarin versus interrupted warfarin combined with preoperative therapeutic-dose LMWH in patients at increased risk of thromboembolism and who had implantation of a pacemaker or ICD. Our main objective was to establish whether the reduced-dose regimen is safe in terms of risk of pocket hematoma. Methods The study had a randomized, parallel, open-label design. Patients planned for device implantation and on treatment with warfarin were routinely referred from the Arrhythmia Service to the Thrombosis Service for perioperative management of anticoagulation. Patients were eligible for inclusion if they had been on warfarin for more than one week and were planned to remain on this drug for at least one month after elective implantation of a pacemaker or ICD. Exclusion criteria were age b 18 years, previous stroke, a CHADS2 score of 0 or 1 for patients with atrial fibrillation as the only indication for warfarin, calculated creatinine clearance of b30 mL/min using the CockroftGault formula [11], surgery scheduled for the first day after a weekend or holiday, patient assessed by physician as unsuitable for the study (history of poor compliance, psychiatric disorder, cognitive impairment, language barrier), participation in another clinical trial or failure to obtain informed consent. Study Treatments Patients were seen at baseline when they had the appointment at our Perioperative Anticoagulation Management Clinic, and this was scheduled for 6–8 days before surgery. Eligible patients were randomized using a web-based computer-generated random number table with random block sizes. Patients randomized to warfarin interruption with preoperative bridging received LMWH syringes (enoxaparin, Lovenox®, Sanofi-Aventis) from the center, or by prescription if they had coverage for the expense. The dose was 1 mg/kg rounded off to the nearest lower multiple of 20 (corresponding to the marketed prefilled syringes) twice daily on day 3 and 2 before surgery and once in the morning on day 1 before surgery. For patients with creatinine clearance of 30–59 mL/min the LMWH was given only once daily for the 3 days. Warfarin was held for 5 days before surgery (6 days if the INR at the preoperative visit = baseline was N3.5). Warfarin was resumed in the evening of the implantation at twice the maintenance dose for 2 days, followed by the usual maintenance dose. Patients randomized to reduced-dose warfarin were instructed to take before the implantation half their usual maintenance dose for 3 days if baseline INR was 1.7-1.8, 4 days if INR was 1.9-2.4, 5 days if INR was 2.5-2.9, 6 days if INR was 3.0-3.4 and 7 days if INR was 3.6-4.4. This regimen was based on modelling of data from 70 previous patients receiving reduceddose warfarin before different surgeries at our hospital [12] and aimed at achieving an INR of ≤1.7 for the surgery. Warfarin was resumed in the evening of the implantation at twice the maintenance dose, followed by the usual maintenance dose. The study nurse gave patients explicit oral and written instructions on the warfarin regimen and for the bridged patients also on LMWH dosing and injection technique. They also received one ampoule of vitamin K, 1 mg, for provisional use orally on the day before surgery if the INR that day was N1.7. Antiplatelet therapy was permitted if indicated. For patient on longterm treatment with aspirin without active coronary artery disease, aspirin was held for 3 days (aspirin 81 mg)[13,14] or 7 days (aspirin 325 mg) before surgery. Patients on clopidogrel but without a recent stent had the drug held for 5 days before surgery.
815
Follow-up We obtained a “same-day” INR the day before surgery and then monitored the INR and dosed the warfarin postoperatively until the therapeutic range was reached, after which the pre-study management was resumed. Before discharge from the hospital, a physician responsible for the implantation or a nurse assessed the patient regarding signs or symptoms related to possible pocket hematoma or major bleeding outcomes. The physician/nurse was not involved with the preoperative bridging and was unaware of the regimen the patient was allocated to. After 3 ± 1 weeks and at 3 months the patient returned for a visit at the Pacemaker/ICD Clinic with inspection of the incision and the generator site by a nurse, unaware of the regimen the patient was allocated to. These visits coincided with routine clinical follow-up of the device. End Point Evaluation Events qualifying for end points were those with onset from the first day after stopping or reducing the dose of warfarin until the visit at the Pacemaker/ICD clinic 3 ± 1 weeks after the implantation. A hematoma or a thromboembolic complication with onset after that time point was deemed highly unlikely to be associated with the study medication period. The primary outcome was pocket hematoma, defined and identified by the blinded Pacemaker/ICD clinic staff as palpable swelling of the pacemaker/ICD pocket exceeding the size of the generator [2]. Severe pocket hematoma was defined when there was in addition one or more of pain, prolonged hospitalization or requirement for re-operation. Secondary outcomes were major bleeding as previously defined [15], arterial or venous thromboembolism that was symptomatic and objectively verified, and death. In addition, we collected data on length of hospitalization and proportion of patients requiring reversal with vitamin K on the day before surgery to achieve INR of 1.7 or lower. Statistical Analyses Our sample size estimate was predicted on demonstrating non-inferiority for the outcome of pocket hematoma, which is clinically important because of its potentially serious consequences, e.g., infected device pocket or wires, evacuation, prolonged hospitalization. The estimated incidence of pocket hematoma for our standard regimen (stopping warfarin for 5 days, preoperative bridging with LMWH) was 4.8% (95% CI: 1.1-8.5), based on our own data from 125 consecutive patients during one year. Our null hypothesis was that the ‘reduced-dose warfarin' strategy is in fact worse than the 'warfarin interruption' strategy (i.e. leads to a higher rate of pocket hematoma and is therefore inferior). If the risk of bleeding does not exceed 9.8%, resulting in a δ of 5% or essentially a doubling of the event rate compared to the conventional 'warfarin interruption' approach, the experimental treatment would still be acceptable in view of lower costs, improved convenience and the modest harm caused by the majority of pocket hematomas. With a one-sided α of 5% and a power of 80%, we needed 225 evaluable patients per group. We discontinued the study prematurely when the Arrhythmia Service changed their routine and allowed therapeutic anticoagulation with warfarin during device implantation, based on the results of the BRUISE CONTROL study [16]. Interim analyses for safety were planned for every 100 patients randomized to review the incidence of serious events (stroke, major bleeding or death), applying pre-defined stopping guidelines. Binomial data were analyzed with chi square test or in case of less than 10 in any numerator with Fisher’s exact test. The Mann–Whitney U-test was used for quantitative values with skewed distribution and ttest for normal distribution. Results of continuous variables are expressed as mean ± standard deviation (SD) in case of normal distribution and as median and interquartile range (IQR) for skewed distributions.
816
S. Schulman et al. / Thrombosis Research 134 (2014) 814–818
The Research Ethics Boards of McMaster University and Hamilton Health Sciences/St Joseph’s Healthcare approved the study. No external funding was obtained but study enoxaparin was provided by SanofiAventis, Canada. This company had no involvement in the design, conduct or analysis of the study. The trial was registered at ClinicalTrials.gov NCT 02094157.
Results Between December 2007 and July 2013 we randomized 172 patients, 169 at Hamilton Health Sciences-General Hospital and 3 at St. Joseph’s Healthcare, both in Hamilton, Ontario. The flow of the patients is shown in Fig. 1. One patient allocated to reduced-dose warfarin had her ICD implantation postponed 10 days and in the meantime her cardiologist switched the anticoagulant therapy to dabigatran, excluding her from the outcome analyses. Baseline characteristics were not significantly different between the two groups with the exception for concomitant treatment with clopidogrel (Table 1). All 5 patients with clopidogrel in the bridged regimen group had this drug stopped for 5 days before surgery. The surgical characteristics, i.e. pacemakers vs. ICDs and first vs. replacement surgery, did not differ between the groups (Table 2). A mean of 5 doses LMWH (range 2–7) were given during the days before surgery according to the bridged regimen and 4 of the patients required home visits by a nurse for these injections. Although warfarin was stopped for 5 days in the group with bridged regimen, the INR on the day before surgery was still mildly elevated with a mean of 1.4. Four patients still had INR ≥2.0, the highest one being 2.4. With the reduceddose regimen 22 patients still had INR ≥2.0 on the day before surgery. In fact, INR increased from the baseline visit to the day before surgery by 0.1 to 0.6 in 14 patients in the reduced warfarin group versus 1 patient (by 0.4) in the bridged group. One mg of vitamin K was given on that day as per protocol to patients with INR ≥1.8 – 5 in the bridged regimen and 35 patients in the reduced-dose warfarin group. On the day of surgery the INR had decreased to b 1.8 in all these patients. One patient (reduced-dose warfarin) with INR 2.2 the day before surgery, treated with vitamin K and reduced to 1.7 on the day of surgery was given prothrombin complex concentrate 1000 units because minithoracotomy was required. Another 2 patients in the reduced-dose warfarin group had an INR the day before surgery of 2.0-2.5 and the procedure was postponed – twice for one of the patients– and the dose reduction was repeated with finally acceptable INR of 1.5.
Not randomized Did not meet criteria: 120 Paent declined: 49
Table 1 Characteristics of the patients.
Age, years, mean (±SD) Female sex, n (%) Indication for warfarin, n (%) AF alone AF & mechanical valve AF and bioprosthetic valve AF and left ventricle thrombus AF & cardiomyopathy AF & venous thromboembolism Mechanical valve Left ventricle thrombus Cardiomyopathy Venous thromboembolism CHADS2 score, median (IQR)⁎ Creatinine clearance, mL/min, mean (±SD) Hemoglobin, g/L, mean (±SD) History of major bleeding, n (%) Years on warfarin, median (IQR) Dose of warfarin, mg/week, mean (±SD) Therapeutic range, n (%) INR 2.0-3.0 INR 2.5-3.5 INR at preoperative visit, mean (±SD) Aspirin therapy, n (%) Continued perioperatively, n (%) Clopidogrel therapy, n (%)
Bridged regimen
Reduced warfarin
N = 85
N = 87
74 (9.6) 21 (25)
74 (10.9) 22 (25)
0.95 0.93
69 (81) 6 (7) 1 (1) 2 (2) 2 (2) 1(1) 3 (4) 1 (1) 0 0 3.0 (2–3) 69 (27)
69 (79) 5 (6) 0 0 1 (1) 1 (1) 3 (3) 2 (2) 4 (5) 2 (2) 3.0 (2–3) 68 (25)
0.78
0.49 0.83
137 (16) 5 (6) 5.0 (1–10) 31 (14)
135 (15) 4 (5) 5.0 (1–10) 32 (14)
0.46 0.74 0.93 0.53
80 (94) 5 (6) 2.2 (0.5) 32 (37) 19 (22) 5 (6)
84 (97) 3 (3) 2.2 (0.5) 34 (39) 19 (22) 0
0.49
The primary outcome, pocket hematoma occurred in 4 patients (5%) with bridged regimen and in 3 (3%) with reduced-dose warfarin (odds ratio 1.37; 95% confidence interval, 0.22-9.60) (Table 3). For one patient in each group the pocket hematoma was classified as severe due to pain but no hospitalization or revision was required. There was only one major bleed in the study, a subdural and subarachnoid hemorrhage 55 days after surgery and considered unrelated. One patient had ischemic stroke 48 days after surgery (INR 2.0 on admission) and with fatal
Randomized 172 paents
Reduced warfarin 87 paents Switched to dabigatran preop: 1 Incomplete follow-up: 1
Incomplete follow-up: 0
Analyzed 85 paents
0.65 0.82 0.93 0.03
SD – standard deviation; AF – atrial fibrillation; CHADS2 – a score giving 1 point each for presence of congestive heart failure, hypertension, age N75 and diabetes and 2 points for stroke or transient ischemic attack; IQR – interquartile range; INR – international normalized ratio ⁎For the patients with atrial fibrillation
Screened 341 paents
Bridged regimen 85 paents
P-value
Analyzed 86 paents
Fig. 1. Flow of patients in the study.
S. Schulman et al. / Thrombosis Research 134 (2014) 814–818 Table 2 Surgery and perioperative anticoagulation.
Device implanted, n (%) Pacemaker ICD First implantation, n (%) Days off vs days reduced-dose warfarin, mean (±SD) Doses of LMWH, median (IQR) INR Day −1, mean (±SD) Range Vitamin K 1 mg Day −1, n (%) Per protocol given Actually given Days to first therapeutic INR, mean (±SD) 1st postoperative INR, mean (±SD) 1st postoperative INR ≥2.0, n (%) Days until 1st INR, mean (±SD)
Bridged regimen
Reduced warfarin
P-value
N = 85
N = 86
44 (52) 41 (48) 42 (49) 5.0 (0.7)
42 (49) 44 (51) 40 (47) 4.3 (1.2)
5.0 (3–5) 1.4 (0.3) 1.0-2.4
Not given 1.7 (0.3) 1.1-2.6
b0.001
5 (5.9) 5 (5.9) 12 (8.5) 1.9 (0.4) 29 (34) 5.2 (2.0)
35 (41) 40 (47) 11 (8.4) 2.0 (0.5) 38 (48) 6.1 (3.6)
b0.001 b0.001 0.81 0.15 0.07 0.19
0.70
0.70 b0.001
ICD – implantable cardioverter defibrillator; SD – standard deviation; LMWH – low-molecular-weight heparin; IQR – interquartile range; INR – international normalized ratio
outcome after 2 days but also considered unrelated. There were 2 nonfatal serious adverse events in each group. Time to first therapeutic INR after surgery and the mean INR at the first sampling occasion (median 5 days after surgery) were similar in the two groups although there was a trend to a higher proportion of INRs ≥2.0 in the reduced-dose warfarin group at the first sampling. Discussion Preoperative reduction of warfarin has previously to our knowledge and a MEDLINE search only been explored in 2 cohort studies [8,9] and in one randomized trial that was briefly reported in a letter to the editor [10]. We found that reduced-dose warfarin, aiming for an INR of 1.5, was associated with a similar, low risk of pocket hematoma after implantation of pacemaker or ICD as a regimen with stopped warfarin and only preoperative bridging with LMWH. Unfortunately, the study had to be prematurely discontinued due to a change in practice and it therefore lacks power to allow for conclusions on non-inferiority. The event rate of pocket hematoma was low in both treatment groups (3% and 5%). Reduced-dose warfarin avoids the inconvenience of subcutaneous injections with heparin but appears to be a regimen that is difficult to comprehend by some patients, as reflected by the significantly higher
817
proportion of patients with increase in the INR compared to baseline and/or with INR still in the therapeutic range the day before surgery and requiring correction with 1 mg of vitamin K. We did, however, not ask the patients to keep diaries of the warfarin dosing, which might have contributed to illustrate this problem. The recent publication of the BRUISE CONTROL trial [16] resulted in a change of paradigm since pacemaker or ICD implantation at therapeutic INR, i.e. without any interruption of warfarin, was safe with a rate of pocket hematoma of 3.5%. This is identical to the rate with reduceddose warfarin in our study. Other studies on maintained warfarin before device implantation reported a hematoma rate of 1.9-24% [17–24]. When only studies with at least 100 patients are included hematoma rate ranged between 1.9% [19] and 7.7% [23]. Conversely, bridging with heparin or LMWH before and after surgery is associated with an unacceptable high rate of pocket hematoma of 16-23% [7,16] and should be abandoned. Bridging only preoperatively was previously reported by us to reduce this risk to 8% [7] but subsequent data from our in-house registry generated a rate of 4.8%. This was the basis for our sample size calculation and indeed, the rate in the bridged regimen group in this study was 4.7%. For a comparison, in patients not on any anticoagulation we have a mean pocket hematoma rate of 2.6%. The higher rate of pocket hematoma with postoperative parenteral anticoagulation than with therapeutic warfarin throughout the procedures might be due to peak levels of heparin resulting in a stronger anticoagulant effect than the even influence of warfarin. It is unclear whether the therapeutic INR results the day before surgery are explained only by misunderstanding of the reduced warfarin dosing. In the group with interrupted warfarin and preoperative LMWH bridging there were also patients with such INRs, albeit significantly fewer. Another possibility is slow metabolism of VKA due to microsomal enzyme polymorphisms (mainly CYP2C9). Yet another reason could be that reduced-dose warfarin should be done with less than half the maintenance dose or for more days than in our regimen, which was based on our previous experience from 70 patients with various procedures [12]. Since we obtained the preoperative INR on the day before surgery it is possible that a higher proportion of patients with reduced-dose warfarin would have had INRs below 1.7 if measured in the morning of the surgery. Notwithstanding the mechanism, a simplification by maintaining warfarin therapy unchanged for device implantation is probably the best solution. If reduced-dose warfarin is to be used for other types of invasive procedures, the regimen should include a day-before-surgery INR with ability to administer 1 mg of vitamin K orally when INR results below the therapeutic range are considered necessary for safety. Future research should focus on simple solutions by identifying more surgeries that do not require interruption of warfarin and by establishing whether bridging really provides any net benefit
Table 3 Clinically important outcomes. Bridged regimen N = 85
Reduced warfarin N = 86
OR (95% CI)
Pocket hematoma, n (%)
4 (5)
3 (3)
1.4 (0.22-9.6)
Severe pocket hematoma, n (%) Major bleeding, n (%) Thromboembolic event, n (%) Serious adverse events within 1 month, n (%)
1 (1) 1⁎ 0 2†
1 (1) 0 1⁎ 2†
Death Re-hospitalization Discharge from hospital same day, n (%)‡ Nights of postop hospitalization, mean (±SD)§
0 2 72 (85) 0.2 (0.4)
0 2 67 (78) 0.3 (0.8)
1.0 (0.7-14)
OR – odds ratio; CI – confidence interval; SD – standard deviation ⁎ These events occurred more than 1 month after surgery † The serious adverse events were in the bridged regimen group: Ventricular lead dislodgement and heart failure, and in the reduced-dose warfarin group: odontoid fracture and pleural effusion. The last one started after randomization but 5 days before the initiation of study-related treatment. ‡ Chi-square test, P = 0.29 § Mantel-Haenszel test comparing the ranks of the observations, P = 0.22
818
S. Schulman et al. / Thrombosis Research 134 (2014) 814–818
for patients having major surgery that cannot be performed on full anticoagulation, as currently explored in the BRIDGE trial (ClinicalTrials.gov NCT00786474). The main limitation of this study is the lack of power due to its premature discontinuation. It was also essentially a single-center study with 98% of patients recruited at one site. Another limitation is the open-label design. The nurses assessing the presence of pocket hematoma at the Pacemaker- or ICD Clinics were, however, unaware of the perioperative anticoagulation regimen. In conclusion, the reduced-dose warfarin regimen, aiming for an INR of ≤1.7 on the day of device implantation had similar safety with regards to pocket hematoma as interrupted warfarin with preoperative LMWH bridging. The reduced-dose regimen failed to achieve INR of ≤1.7 the day before surgery in 41% of the patients. This illustrates the imprecision of manipulating warfarin dosing to achieve non-therapeutic targets. Reduced-dose warfarin might still be an alternative for a suitable subset of patients having procedures that carry too high risk of bleeding on full anticoagulation, provided that they can check the INR the day before surgery and correct with vitamin K. Conflict of Interest Statement The authors have no conflicts of interest in relation to this article Appendix A. Supplementary Data PACEBRIDGE Study Protocol can be found online at http://dx.doi.org/ 10.1016/j.thromres.2014.07.028. References [1] McComb JM, Gribbin GM. Chronic atrial fibrillation in patients with paroxysmal atrial fibrillation, atrioventricular node ablation and pacemakers: determinants and treatment. Europace 1999;1:30–4. [2] Wiegand UK, LeJeune D, Boguschewski F, Bonnemeier H, Eberhardt F, Schunkert H, et al. Pocket hematoma after pacemaker or implantable cardioverter defibrillator surgery: influence of patient morbidity, operation strategy, and perioperative antiplatelet/anticoagulation therapy. Chest 2004;126:1177–86. [3] Kiviniemi MS, Pirnes MA, Eranen HJ, Kettunen RV, Hartikainen JE. Complications related to permanent pacemaker therapy. Pacing Clin Electrophysiol 1999;22:711–20. [4] Aggarwal RK, Connelly DT, Ray SG, Ball J, Charles RG. Early complications of permanent pacemaker implantation: no difference between dual and single chamber systems. Br Heart J 1995;73:571–5. [5] Jamula E, Douketis JD, Schulman S. Perioperative anticoagulation in patients having implantation of a cardiac pacemaker or defibrillator: a systematic review and practical management guide. J Thromb Haemost 2008;6:1615–21.
[6] Phibbs B, Marriott HJ. Complications of permanent transvenous pacing. N Engl J Med 1985;312:1428–32. [7] Robinson M, Healey JS, Eikelboom J, Schulman S, Morillo CA, Nair GM, et al. Postoperative low-molecular-weight heparin bridging is associated with an increase in wound hematoma following surgery for pacemakers and implantable defibrillators. Pacing Clin Electrophysiol 2009;32:378–82. [8] Larson BJ, Zumberg MS, Kitchens CS. A feasibility study of continuing dose-reduced warfarin for invasive procedures in patients with high thromboembolic risk. Chest 2005;127:922–7. [9] Marietta M, Bertesi M, Simoni L, Pozzi S, Castelli I, Cappi C, et al. A simple and safe nomogram for the management of oral anticoagulation prior to minor surgery. Clin Lab Haematol 2003;25:127–30. [10] Sacco R, Sacco M, Carpenedo M, Moia M. Oral surgery in patients on oral anticoagulant therapy: a randomized comparison of different INR targets. J Thromb Haemost 2006;4:688–9. [11] Cockcroft DW, Gault MH. Prediction of creatinine clearance from serum creatinine. Nephron 1976;16:31–41. [12] Schulman S, Earl B, Robinson M. Perioperative management of anticoagulation with tapered dose warfarin. Haematologica 2008;93(s1):147–8. [13] Zisman E, Erport A, Kohanovsky E, Ballagulah M, Cassel A, Quitt M, et al. Platelet function recovery after cessation of aspirin: preliminary study of volunteers and surgical patients. Eur J Anaesthesiol 2010;27:617–23. [14] Li C, Hirsh J, Xie C, Johnston MA, Eikelboom JW. Reversal of the anti-platelet effects of aspirin and clopidogrel. J Thromb Haemost 2012;10:521–8. [15] Schulman S, Angeras U, Bergqvist D, Eriksson B, Lassen MR, Fisher W. Definition of major bleeding in clinical investigations of antihemostatic medicinal products in surgical patients. J Thromb Haemost 2010;8:202–4. [16] Birnie DH, Healey JS, Wells GA, Verma A, Tang AS, Krahn AD, et al. Pacemaker or defibrillator surgery without interruption of anticoagulation. N Engl J Med 2013;368: 2084–93. [17] al-Khadra AS. Implantation of pacemakers and implantable cardioverter defibrillators in orally anticoagulated patients. Pacing Clin Electrophysiol 2003;26:511–4. [18] Denman RA, Walters DL, Twidale N, Masterson ML. The implantation of permanent pacemakers and automatic implantable cardiac defibrillators in patients on warfarin. CSANZ 47 Annual Scientific Meeting; 1999 [www.csanz.edu.au/abstracts/ 47abstracts/041.doc.htm]. [19] Giudici MC, Barold SS, Paul DL, Bontu P. Pacemaker and implantable cardioverter defibrillator implantation without reversal of warfarin therapy. Pacing Clin Electrophysiol 2004;27:358–60. [20] Goldstein DJ, Losquadro W, Spotnitz HM. Outpatient pacemaker procedures in orally anticoagulated patients. Pacing Clin Electrophysiol 1998;21:1730–4. [21] Michaud GF, Pelosi Jr F, Noble MD, Knight BP, Morady F, Strickberger SA. A randomized trial comparing heparin initiation 6 h or 24 h after pacemaker or defibrillator implantation. J Am Coll Cardiol 2000;35:1915–8. [22] Milic DJ, Perisic ZD, Zivic SS, Stanojkovic ZA, Stojkovic AM, Karanovic ND, et al. Prevention of pocket related complications with fibrin sealant in patients undergoing pacemaker implantation who are receiving anticoagulant treatment. Europace 2005;7:374–9. [23] Tischenko A, Gula LJ, Yee R, Klein GJ, Skanes AC, Krahn AD. Implantation of cardiac rhythm devices without interruption of oral anticoagulation compared with perioperative bridging with low-molecular weight heparin. Am Heart J 2009;158:252–6. [24] Tolosana JM, Berne P, Mont L, Heras M, Berruezo A, Monteagudo J, et al. Preparation for pacemaker or implantable cardiac defibrillator implants in patients with high risk of thrombo-embolic events: oral anticoagulation or bridging with intravenous heparin? A prospective randomized trial. Eur Heart J 2009;30:1880–4.