Evaluation of Bivalirudin’s Effect on International Normalized Ratio to Determine an Appropriate Strategy for Transitioning to Warfarin

Abstract

Background: Direct thrombin inhibitors are recommended in confirmed or suspected heparin-induced thrombocytopenia. False elevation of the international normalized ratio (INR) occurs with these agents making bridging to warfarin challenging. There is limited data regarding bivalirudin’s effect on INR. Objective: To evaluate bivalirudin’s effect on the INR and determine a strategy for transitioning to warfarin. Methods: This was a retrospective observational study. Included patients were >18 years old receiving primary bridging therapy with overlapping bivalirudin and warfarin for at least 72 hours. Patients with administration of alternate anticoagulants during the transition interval or active major bleeding within 48 hours prior to bivalirudin initiation were excluded. The primary endpoint was to determine the effect on INR at first therapeutic activated partial thromboplastin time after bivalirudin initiation and prior to warfarin initiation. Secondary endpoints included change in INR 12 and 24 hours after bivalirudin initiation, change in INR 4 hours after bivalirudin cessation, and incidence of major bleeding or new thrombotic events. Results: Thirty-four patients met study criteria. For the primary endpoint, the change in INR at first therapeutic activated partial thromboplastin time was 0.37 (range = 0.28-0.48), which occurred at 8.4 hours (range = 4.6-14.2; n = 14). INR increased at 12 and 24 hours by a median of 0.55 and 0.5 from baseline, respectively. Median change in INR 4 to 8 hours post-bivalirudin cessation was −0.48. Conclusion: Targeting an INR > 2.5 when bridging to warfarin will account for this false elevation and maintain an INR above 2.0 on bivalirudin discontinuation.

Background

Direct thrombin inhibitors (DTIs) are commonly utilized parenteral anticoagulants in confirmed or suspected heparin-induced thrombocytopenia (HIT) because they have no effect on platelet activation.¹ DTIs can falsely elevate the prothrombin time (PT) and the international normalized ratio (INR), which proves challenging when bridging to vitamin K antagonists (eg, warfarin) with these agents. The effects on INR are dependent on the DTI, drug concentration, and reagent used to measure INR. Elevations in the INR differ in significance with lepirudin, bivalirudin, and argatroban (lowest to highest, respectively). The prolongation of the PT when utilizing DTIs leads to a false elevation in INR because the INR is calculated utilizing PT with the following equation:

$$\mathrm{INR}={(\mathrm{PT\; specimen}/\mathrm{mean\; normal\; PT})}^{\mathrm{ISI}}$$

Laboratory reagents used to test INR (thromboplastins) are affected the most if they have a high international sensitivity index (ISI). The ISI represents the sensitivity to detect clotting factor deficiencies of each thromboplastin. Thromboplastins with lower ISI values (typically ~1) are less sensitive to the effects of DTIs, while those with higher ISI values (typically ~2) are more sensitive.^2,3

Bivalirudin is a DTI derived from the saliva of the Hirudo medicinalis leech.⁴ Bivalirudin is indicated for use in percutaneous coronary intervention (PCI), percutaneous transluminal coronary angioplasty, and for patients undergoing PCI with confirmed HIT or who are at risk of developing HIT. Bivalirudin is a specific and reversible analog of hirudin and inhibits thrombin directly by binding to the catalytic and anion-binding site of thrombin. Bivalirudin is predominantly eliminated by plasma enzymes and proteolytic cleavage, with only 20% cleared renally, making it an excellent option for patients with hepatic or renal dysfunction. Due to renal clearance, bivalirudin dosage may need to be adjusted based on renal impairment. The half-life of bivalirudin is 25 minutes in normal renal function and increases to almost 1 hour in severe renal impairment. No specific reversal agent is available for bivalirudin; however, hemodialysis can be utilized to increase clearance.⁵

Argatroban is another DTI indicated for use in HIT and PCI. Argatroban is metabolized hepatically and should be used with caution or avoided in patients with hepatic dysfunction but can be used safely in patients with renal dysfunction. Argatroban use should also be avoided in patients undergoing PCI with liver function tests elevated to greater than 3 times the upper limit of normal. The manufacturer has recommendations for bridging with argatroban to warfarin. Once warfarin is initiated, overlap with argatroban should continue for at least 5 days. Secondary to the significant and established effects on INR elevation, argatroban therapy should be continued until the INR is > 4.0. On discontinuation of argatroban, the INR is tested 4 hours later to determine the independent effects contributable to warfarin.⁶

Unlike argatroban, there is limited data on bivalirudin’s effect on INR and there are no specific manufacturer recommendations for transitioning to warfarin. A retrospective data analysis of patients with HIT treated with bivalirudin found that the increase in INR ranged from <0.1 to 0.85.⁷ We did not identify any other published literature on this topic. Given the lack of published data to support a particular process for bridging to warfarin from bivalirudin, our aim was to further evaluate bivalirudin’s effect on the INR in order to determine a strategy for a successful transition.

Methods

This single-center, retrospective chart review was approved by the institutional review boards at Memorial Hermann Health System and the University of Texas Health Science Center at Houston. The study was conducted in compliance with the requirements of both institutional review boards. Data was collected through review of electronic medical records. All patients admitted to Memorial Hermann-Texas Medical Center from July 2012 through June 2015 were screened for inclusion regardless of location in the hospital. Patients more than 18 years old who received primary bridging therapy with overlapping bivalirudin and warfarin for at least 72 hours without interruption were evaluated. Exclusion criteria included administration of alternate anticoagulants during the transition interval, active major bleeding within 48 hours prior to bivalirudin initiation, administration of thrombolytics within 48 hours of bivalirudin initiation due to increased risk of bleeding, administration of medications that are major inducers of warfarin metabolism (barbiturates, carbamazepine, dicloxacillin, rifampin, or phenytoin) prior to or during hospitalization, and pregnancy. Active major bleeding was defined as any International Classification of Diseases, Clinical Modification (ICD-9-CM) code for intracranial, intraspinal, retroperitoneal, intraocular, muscle with compartment syndrome, or pericardial bleed, or any ICD-9-CM code for bleeding and a 2 g/dL or greater drop in hemoglobin within a 48-hour period (see the appendix). To eliminate other factors that may confound the INR, patients with acute hepatitis or cirrhosis of the liver, antiphospholipid antibody syndrome, lupus anticoagulant syndrome, INR above 1.5 at the time of bivalirudin initiation, and/or administration of ≥2.5 mg of phytonadione within 5 days of the initiation of bivalirudin were also excluded.

Our institution utilizes a nurse-driven protocol for titration of bivalirudin infusions. The starting dose is 0.15 mg/kg/h with activated partial thromboplastin time (aPTT) samples drawn every 4 hours until therapeutic and 4 hours after a rate change. After 2 consecutive therapeutic aPTT values the samples are decreased to every 12 hours. The goal aPTT is between 50 and 80 seconds.

The primary endpoint was to determine change in INR at first therapeutic aPTT after bivalirudin initiation but prior to warfarin initiation. Secondary endpoints included change in INR at 12 and 24 hours after bivalirudin infusion initiation and prior to warfarin initiation, change in INR from cessation of bivalirudin to more than 4 hours after bivalirudin cessation, percentage of time appropriately anticoagulated, and incidence of major bleeding or new thrombotic events. All INR changes and time points are reported as median with interquartile range. Post hoc analyses included the change in INR at all therapeutic aPTTs (prior to warfarin initiation) and a comparison of the magnitude of bivalirudin’s effect on INR at varying aPTT ranges. Data was analyzed using R version 3.3.3 (R Core Team, Vienna, Austria).

Results

A total of 108 patients who received bivalirudin with overlapping warfarin therapy were screened. Seventy-six patients were eligible for inclusion in the study with >72 hours of overlapping therapy. Of these, 42 patients were excluded for the following reasons: INR > 1.5 at time of bivalirudin initiation (n = 21), thrombolytics within 48 hours of bivalirudin (n = 9), alternate anticoagulant during transition interval (n = 5), bleeding within 48 hours prior to bivalirudin (n = 3), diagnosis of liver disease or hypercoagulable condition (n = 3), and phytonadione within 5 days of bivalirudin initiation (n = 1).

Indications for bivalirudin were fairly evenly distributed between atrial fibrillation, deep vein thrombosis/pulmonary embolism, and mechanical heart valve (Table 1). Twenty-four percent of patients had a laboratory-confirmed diagnosis of HIT. Acute renal failure was diagnosed in 59% of patients. Patients received bivalirudin for a median duration of 8.9 days (range = 6.7-11.8) over which the bivalirudin overlapped with warfarin therapy for a median of 5.2 days (range = 3.7-7.5). The median initial infusion rate of bivalirudin was 0.06 mg/kg/h (range = 0.05-0.15) with a final median infusion rate of 0.03 mg/kg/h (range = 0.01-0.10). The median length of bivalirudin infusion prior to initiation of warfarin was 3.2 days (range = 1.8-5.2).

Table 1.

Baseline Characteristics.

Characteristics	N = 34
Age (years), median [IQR]	61 [54-69]
Male, n (%)	21 (62)
Race, n (%)
White	16 (47)
African American	12 (35)
Other	4 (12)
Unknown	2 (6)
Anticoagulation indicationa, n (%)
Atrial fibrillation	8 (24)
VTEb	12 (35)
PE	8 (24)
DVT	6 (18)
Heart valve	3 (9)
Chronic anticoagulation	4 (12)
Other	14 (41)
Comorbidities, n (%)
Renal impairment
ARF	20 (59)
CKD	7 (21)
ESRD	6 (18)
Cancer	2 (6)
HIT	8 (24)

Abbreviations: IQR, interquartile range; VTE, venous thromboembolism; PE, pulmonary embolism; DVT, deep vein thrombosis; ARF, acute renal failure; CKD, chronic kidney disease; ESRD, end-stage renal disease; HIT, heparin-induced thrombocytopenia.

^aFive patients had multiple indications for anticoagulation.

^bSome patients diagnosed with VTE had both PE and DVT.

Primary and secondary outcomes are reported in Table 2. The baseline INR prior to bivalirudin initiation was 1.22 (range = 1.13-1.36). For the primary endpoint, there were 14 patients who had an INR drawn at the time of the first therapeutic aPTT. In these 14 patients, the median change in INR at first therapeutic aPTT was 0.37 (range = 0.28-0.48; Figure 1), which occurred at a median of 8.4 hours (range = 4.6-14.2) after bivalirudin initiation.

Table 2.

Primary and Secondary Outcomes.

Primary Outcomes	Median [IQR]
INR at baseline	1.22 [1.13 to 1.36]
INR at first therapeutic aPTT	1.61 [1.47 to 1.86]
Change in INR at first therapeutic aPTT	0.37 [0.28 to 0.48]
Secondary Outcomes	Median [IQR]
Change in INR 12 and 24 hours after bivalirudin initiation
INR at 12 hours	1.72 [1.51 to 1.94]
Change in INR at 12 hours	0.55 [0.32 to 0.66]
INR at 24 hours	1.96 [1.60 to 2.03]
Change in INR at 24 hours	0.50 [0.25 to 0.52]
Bivalirudin cessation
INR prior to cessation	2.83 [2.40 to 3.33]
INR at 4 to 8 hours after cessation	2.46 [2.17 to 2.87]
Change in INR at cessation	−0.48 [−0.83 to −0.12]
Percent time aPTT values were in range, %
Within goal aPTT range	80.5 [74.8 to 91.3]
Below goal aPTT range	5.8 [0.0 to 14.8]
Above goal aPTT range	7.9 [2.8 to 14.9]
Safety Endpoints	n (%)
Bleeding and thrombosis events
Major bleed	4 (12)
Minor bleed	1 (3)
New thrombosis	0 (0)

Abbreviations: IQR, interquartile range; INR, international normalized ratio; aPTT, activated partial thromboplastin time.

Figure 1.

Box plot representing change in INR at baseline versus bivalirudin initiation prior to initiation of warfarin, and after bivalirudin cessation with concurrent warfarin administration.

In an evaluation of the secondary endpoints, the INR increased by 0.55 (range = 0.32-0.66) and 0.5 (range = 0.25-0.52) at 12 and 24 hours, respectively. The change in INR at 4 to 8 hours post bivalirudin cessation was −0.48 (range = −0.83 to −0.12; Figure 1). Major and minor bleeds occurred in 4 (12%) and 1 (3%) patients, respectively. There were no new thrombotic events reported.

In a post hoc analysis, the median change in INR at all therapeutic aPTT values was found to be 0.25 (range = 0.09-0.39). A scatter plot was produced to examine the magnitude of the effect bivalirudin exerted on the INR as the aPTT increased (Figure 2). This demonstrates that there is a small, but statistically significant (P < .001), linear relationship between the INR and aPTT.

Figure 2.

A scatter plot examining the magnitude of the effect bivalirudin exerted on the INR as the aPTT increased. There is a small, but statistically significant (P < .001), linear relationship between the INR and aPTT.

Discussion

HIT occurs as a potential adverse effect when unfractionated heparin (UFH) or low-molecular weight heparins (LMWH) activate platelets leading to increased thrombin generation, thrombocytopenia, and potential thrombosis.¹ Thrombotic episodes due to HIT may lead to amputation or mortality in almost half of patients who do not receive additional anticoagulant therapy after heparin discontinuation.⁸ Thrombosis may continue to develop for up to 30 days despite discontinuation of the offending agent. Therefore, the risk or presence of thrombosis necessitates treatment with DTIs followed by a minimum of 4 weeks of oral anticoagulation.

There is data available to aid clinicians when bridging with argatroban to warfarin. The manufacturer provides 2 equations that can be utilized to determine the true INR when bridging which is depicted below.⁶

For labs that utilize thromboplastin A:

$$0.19+0.57(\mathrm{INRWA})=\mathrm{INRW}$$

For labs that utilize thromboplastin B:

$$0.18+0.45(\mathrm{INRWA})=\mathrm{INRW}$$

where INRWA is the INR with warfarin and argatroban and INRW is the INR with warfarin alone. It is important to note that this equation is only validated for argatroban doses of 1 to 2 µg/kg/min and cannot be reliably utilized in doses greater than 2 µg/kg/min.⁶ Unfortunately, there are no manufacturer recommendations to guide clinicians in determining the true INR when bridging with bivalirudin.

The aim of this study was to quantify the degree of false INR elevation, which occurs during bivalirudin administration in order to more accurately predict the true INR value, which is important for general laboratory interpretation as well as when bridging to warfarin. Our results indicate that bivalirudin falsely elevates the INR by 0.37 to 0.55 over several different time points of analysis prior to warfarin initiation, with a corresponding decrease of 0.48 on bivalirudin discontinuation. These findings are consistent with those of Francis et al who reported a change in INR ranging from <0.1 to 0.8.⁷ Additionally, this confirms that bivalirudin has a lower magnitude of effect on the INR than argatroban, which causes a median increase of 1.9 in the INR.⁹ Argatroban recommendations target an INR above 4.0 before checking for the true INR when bridging. However, due to the smaller, but reproducible effect of bivalirudin on the INR, alternative recommendations to account for the drug’s effect on INR should be considered.

When bridging with bivalirudin to warfarin it is reasonable to target an INR value > 2.5 in order to achieve an INR goal between 2.0 to 3.0 Once the INR is above 2.5, the bivalirudin infusion should be held for 4 hours, at which point a repeat INR should reflect the true INR value prior to permanent bivalirudin therapy discontinuation. Holding the bivalirudin infusion for 4 hours may be recommended despite the short half-life that bivalirudin possesses. The anticoagulation effects of bivalirudin typically resolves within 1 hour but this can be prolonged up to 4 hours in renal impairment, waiting 4 hours ensures that the true INR is being measured and that elevation in INR is not due to remaining effects of bivalirudin.⁵ For patients who require higher INR goals of 2.5 to 3.5 due to mechanical heart valves, it is reasonable to target an INR of 3.0 to 4.0 in order to achieve a therapeutic INR.

The primary limitation of this study is the small patient population in a single-center review. A larger, more diverse patient population would improve overall external validity as well as provide the ability to determine the effects of specific comorbidities, such as those with renal impairment, on the findings. The retrospective nature of the evaluation also provided fewer data points overall because time points of INR and/or aPTT analysis were not predesignated. Additionally, our institution utilizes Néoplastine reagent and adjusts the ISI annually due to change in lot number. The ISI as of August 2015 was 1.30 and September 2016 was 1.28. Application of our results to other institutions may vary based on the ISI and thromboplastin reagent being utilized.

Conclusion

Bivalirudin’s effects on elevating the INR range from 0.37 to 0.55 over the course of therapy initiation and decreased the INR by 0.48 on bivalirudin discontinuation. It is reasonable to target a higher INR goal (INR > 2.5 for INR goal 2.0-3.0 or INR > 3.0 for INR goal 2.5-3.5) when bridging to warfarin. Proper application of bivalirudin’s effects on INR to therapeutic decision-making will limit premature discontinuation of bivalirudin, and thus decrease the possibility of inadequate anticoagulation in patients during the bridging interval.

Appendix

ICD-9-CM Codes Utilized for Major and Minor Bleeding

Major Bleeding

• 360.43—Hemophthalmos, except current injury

• 430—Subarachnoid hemorrhage

• 431—Intracerebral hemorrhage

• 432.1—Subdural hemorrhage

• 432.9—Unspecified intracranial hemorrhage

• 568.81—Hemoperitoneum (nontraumatic)

• 719.16—Hemarthrosis, lower leg

Minor Bleeding

• 456—Esophageal varices with bleeding

• 456.2—Esophageal varices in diseases classified elsewhere, with bleeding

• 530.7—Gastroesophageal laceration-hemorrhage syndrome

• 530.82—Esophageal hemorrhage

• 531—Acute gastric ulcer with hemorrhage, without mention of obstruction

• 531.01—Acute gastric ulcer with hemorrhage, with obstruction

• 531.2—Acute gastric ulcer with hemorrhage and perforation, without mention of obstruction

• 531.21—Acute gastric ulcer with hemorrhage and perforation, with obstruction

• 531.4—Chronic or unspecified gastric ulcer with hemorrhage, without mention of obstruction

• 531.41—Chronic or unspecified gastric ulcer with hemorrhage, with obstruction

• 531.6—Chronic or unspecified gastric ulcer with hemorrhage and perforation, without mention of obstruction

• 531.61—Chronic or unspecified gastric ulcer with hemorrhage and perforation, with obstruction

• 532—Acute duodenal ulcer with hemorrhage, without mention of obstruction

• 532.01—Acute duodenal ulcer with hemorrhage, with obstruction

• 532.2—Acute duodenal ulcer with hemorrhage and perforation, without mention of obstruction

• 532.21—Acute duodenal ulcer with hemorrhage and perforation, with obstruction

• 532.4—Chronic or unspecified duodenal ulcer with hemorrhage, without mention of obstruction

• 532.41—Chronic or unspecified duodenal ulcer with hemorrhage, with obstruction

• 532.6—Chronic or unspecified duodenal ulcer with hemorrhage and perforation, without mention of obstruction

• 532.61—Chronic or unspecified duodenal ulcer with hemorrhage and perforation, with obstruction

• 533—Acute peptic ulcer of unspecified site with hemorrhage, without mention of obstruction

• 533.01—Acute peptic ulcer of unspecified site with hemorrhage, with obstruction

• 533.2—Acute peptic ulcer of unspecified site with hemorrhage and perforation, without mention of obstruction

• 533.21—Acute peptic ulcer of unspecified site with hemorrhage and perforation, with obstruction

• 533.4—Chronic or unspecified peptic ulcer of unspecified site with hemorrhage, without mention of obstruction

• 533.41—Chronic or unspecified peptic ulcer of unspecified site with hemorrhage, with obstruction

• 533.6—Chronic or unspecified peptic ulcer of unspecified site with hemorrhage and perforation, without mention of obstruction

• 533.61—Chronic or unspecified peptic ulcer of unspecified site with hemorrhage and perforation, with obstruction

• 534—Acute gastrojejunal ulcer with hemorrhage, without mention of obstruction

• 534.01—Acute gastrojejunal ulcer, with hemorrhage, with obstruction

• 534.2—Acute gastrojejunal ulcer with hemorrhage and perforation, without mention of obstruction

• 534.21—Acute gastrojejunal ulcer with hemorrhage and perforation, with obstruction

• 534.4—Chronic or unspecified gastrojejunal ulcer with hemorrhage, without mention of obstruction

• 534.41—Chronic or unspecified gastrojejunal ulcer, with hemorrhage, with obstruction

• 534.6—Chronic or unspecified gastrojejunal ulcer with hemorrhage and perforation, without mention of obstruction

• 534.61—Chronic or unspecified gastrojejunal ulcer with hemorrhage and perforation, with obstruction

• 569.3—Hemorrhage of rectum and anus

• 578—Hematemesis

• 578.1—Blood in stool

• 578.9—Hemorrhage of gastrointestinal tract, unspecified

• 599.7—Hematuria

• 719.11—Hemarthrosis, shoulder region

• 719.12—Hemarthrosis, upper arm

• 719.13—Hemarthrosis, forearm

• 719.14—Hemarthrosis, hand

• 719.15—Hemarthrosis, pelvic region and thigh

• 719.17—Hemarthrosis, ankle and foot

• 719.18—Hemarthrosis, other specified sites

• 719.19—Hemarthrosis, multiple sites

• 784.7—Epistaxis

• 784.8—Hemorrhage from throat

• 786.3—Hemoptysis