Figure. Inclusion and exclusion criteria of patients. CrCl indicates creatinine clearance; PI, package insert. To convert CrCl to milliliters per second, multiply by 0.0167.
DeCarolis DD, Thorson JG, Clairmont MA, Leuthner AM, Rector TS, Johnson GJ. Enoxaparin Outcomes in Patients With Moderate Renal Impairment. Arch Intern Med. 2012;172(22):1713-1718. doi:10.1001/2013.jamainternmed.369
Author Affiliations: Pharmacy Service (Drs DeCarolis, Thorson, Clairmont, and Leuthner), Center for Chronic Disease Outcomes Research (Dr Rector), and Hematology/Oncology Section (Dr Johnson), Minneapolis Veterans Affairs Health Care System, Minneapolis, Minnesota.
Background Enoxaparin sodium has predictable pharmacokinetics that allow for simplified dosing without laboratory monitoring. Reliance on renal function for excretion may lead to accumulation of enoxaparin in patients with moderate renal impairment. However, there is no dose adjustment recommended for these patients. We conducted a review to compare bleeding events in patients with moderate renal impairment compared with those with normal renal function.
Methods Patients received enoxaparin sodium, 1 mg/kg, every 12 hours or 1.5 mg/kg once daily between June 1 and November 30, 2009. Moderate renal impairment was defined as creatinine clearance (CrCl) of 30 to 50 mL/min. Normal renal function was defined as CrCl greater than 80 mL/min. The primary outcome was major bleeding, defined as any bleeding resulting in death, hospital admission, lengthened hospital stay, or an emergency department visit. The secondary outcome was thromboembolism.
Results A total of 164 patients met the inclusion criteria: 105 with normal renal function and 59 with moderate renal impairment. The primary outcome occurred in 6 of 105 patients (5.7%) with normal renal function vs 13 of 59 patients (22.0%) with moderate renal impairment, representing an unadjusted odds ratio of 4.7 (95% CI, 1.7-13.0; P = .002). The odds ratio using multivariable logistic regression adjusting for differences in risk was 3.9 (95% CI, 0.97-15.6; P = .055). There was no recurrent thromboembolism in either group.
Conclusions Our results suggest an increased risk of major bleeding in patients with moderate renal impairment who receive enoxaparin. Because enoxaparin is frequently used and outcomes can be life saving or life threatening, we encourage further study of the appropriate dose in patients with moderate renal impairment.
Enoxaparin sodium, a low-molecular-weight heparin, has a predictable pharmacokinetic profile and dose-response curve, allowing simplified dosing without the need for vigilant monitoring through laboratory tests.1 A disadvantage of enoxaparin is reliance on kidney function for excretion and potential accumulation of its anticoagulant effect in patients with declining renal function. Quiz Ref IDThe Food and Drug Administration–approved dosing recommends decreasing the daily dose by one-half in patients with severe renal impairment, defined as creatinine clearance (CrCl) of less than 30 mL/min (to convert to milliliters per second, multiply by 0.0167).2 The manufacturer's package insert2 and pharmacokinetic studies3- 8 have consistently described a linear relationship between enoxaparin clearance measured by antifactor Xa activity and renal function measured by CrCl. Despite this linear relationship, there is no recommendation for dose adjustment for patients with moderate renal impairment (CrCl 30-50 mL/min). The package insert2 recommends that these patients be observed carefully for signs and symptoms of bleeding but otherwise provides no guidance regarding dose adjustment or laboratory monitoring. In support of these concerns are reports9- 14 of increased bleeding in patients with moderate renal impairment who receive the recommended enoxaparin dose.
Enoxaparin's risk-benefit profile balances treatment of thrombosis with risk of bleeding, both of which can be of major significance. Enoxaparin use in patients with moderate renal impairment results in a higher concentration of the anticoagulant; however, the recommended dose remains the same as with normal renal function. Ensuring an accurate dose may have significant influence on relevant outcomes. Therefore, appropriate dose modification would appear necessary to maintain a proper balance of efficacy and safety in patients with reduced renal function. The large-scale clinical studies instrumental in the approval of current dosing enrolled healthier patient populations than those seen in everyday clinical practice and often excluded patients with renal dysfunction. Smaller studies7,15- 17 have made dose modifications to target an antifactor Xa activity therapeutic range, despite a lack of strong evidence correlating antifactor Xa activity target range to outcomes.18,19 Thus, we conducted a review of medical records to compare the outcomes of bleeding and recurrent thrombotic events in patients with normal renal function vs those with moderate renal impairment who were receiving the manufacturer's recommended therapeutic dose of enoxaparin.
This study was performed at the Minneapolis Veterans Affairs Health Care System (MVAHCS), a tertiary care teaching facility that provides inpatient and primary and subspecialty ambulatory care. An electronic medical record documents dispensing of outpatient medications, administration of inpatient medications, laboratory test results, and clinical progress notes. The electronic medical record provides access to data from other VA medical facilities, allowing follow-up of clinical status when primary care is transferred. Because the VA serves as the primary care provider of most patients, admissions to other hospitals or emergency department visits are documented in the electronic medical record when this information is made available or telephone contact is made. The MVAHCS uses an enoxaparin protocol that requires baseline laboratory data and body weight to be documented before treatment. These data permit clinical pharmacists to calculate the dose on the basis of current product labeling and evidence-based guidelines.1,2 Deviation from this dose requires approval via the hematology consult service.
Patients who received enoxaparin between June 1 and November 30, 2009, during a hospital inpatient stay or as an outpatient prescription were identified via dispensing records. Patients were excluded if they received enoxaparin for venous thromboembolism prophylaxis, dosing differed from that in the manufacturer's package insert, or long-term enoxaparin therapy was prescribed. Long-term therapy was defined as a regimen in which enoxaparin is the anticoagulant to be used for the full duration of treatment. Quiz Ref IDThus, our study concentrated on the use of enoxaparin for initiation of parenteral anticoagulation as well as for bridge therapy, both with planned transition to oral anticoagulation therapy (warfarin sodium). Initiation of anticoagulation was defined as treatment for a new thrombotic indication with eventual transitioning to oral anticoagulation. Bridge therapy was defined as transitioning from a therapeutic level of oral anticoagulation followed by interruption and replacement with therapeutic doses of enoxaparin, with a planned transition back to oral anticoagulation. For patients who received more than 1 course of enoxaparin during this period, only the first course was included in the analysis.
The main objective of our study was to compare major bleeding outcomes in patients with normal renal function vs those with moderate renal impairment receiving the product-labeled enoxaparin dose. Normal renal function was defined as CrCl greater than 80 mL/min, and moderate renal impairment was defined as CrCl OF 30 to 50 mL/min.2 Our final study population consisted of patients who received enoxaparin sodium at a therapeutic dosage of 1 mg/kg of body weight every 12 hours or 1.5 mg/kg every 24 hours and estimated CrCl as specified. Creatinine clearance was calculated using the Cockcroft-Gault formula with lean body weight.20,21 The actual body weight was used to calculate the enoxaparin dose.
The following data were extracted from the electronic medical record: enoxaparin indication, dose, and duration of therapy; age; sex; race; weight; levels of serum creatinine, alanine aminotransferase, aspartate aminotransferase, and hemoglobin; platelet count; baseline international normalized ratio; and highest international normalized ratio during enoxaparin therapy. Baseline risk factors for bleeding were collected, including history of uncontrolled hypertension, cancer, cerebrovascular accidents, bleeding, falls, or recent trauma or surgery and the World Health Organization performance score.22,23 Concurrent medications 7 days before or during enoxaparin therapy, including antiplatelet agents, nonsteroidal anti-inflammatory drugs, fish oil, vitamin E, and warfarin, were recorded as risk factors.22
An extensive medical record review was conducted to assess outcome data that included the 14 days following completion of enoxaparin therapy. The primary outcome of the study was major bleeding, defined as any bleeding resulting in hospital admission or death, lengthened hospital stay, or an emergency department visit. The secondary outcome was any evidence of new, progressing, or recurring thromboembolism during the same period. The protocol was approved by the MVAHCS institutional review board.
Characteristics of the patients categorized according to CrCl were described using means (SDs) or proportions (percentiles). An odds ratio for a bleeding episode was estimated with 95% CIs, using logistic regression to adjust for differences at baseline other than renal function that might have represented a substantial difference in the risk of bleeding. Commercial software (Stata, version 10.1; StataCorp) was used for all analyses.
A total of 605 enoxaparin courses were identified during the 6-month study period. Inclusion and exclusion data are illustrated in the Figure. The final sample consisted of 164 patients: 105 with normal renal function and 59 with moderate renal impairment.
Baseline characteristics are summarized in Table 1. In addition to the selected differences in renal function, the group with moderate renal impairment was older and had a greater percentage with platelet counts less than 100 × 103/μL (1:1 conversion to × 109/L). The normal renal function group had a greater percentage with elevated liver enzyme levels. The mean baseline international normalized ratio was similar between the groups, as were baseline risk factors and use of bridge therapy. The mean duration of enoxaparin treatment was slightly longer in the normal renal function group (12.8 days vs 11.9 days). A greater percentage of patients with normal renal function were taking a nonsteroidal anti-inflammatory drug.
Quiz Ref IDThe primary outcome of major bleeding occurred in 6 of 105 patients (5.7%) with normal renal function vs 13 of 59 patients (22.0%) with moderate renal impairment. There were no deaths. The unadjusted odds ratio for major bleeding was 4.7 (95% CI, 1.7-13.0; P = .002). With adjustment for differences in risk factors using multivariable logistic regression, the odds ratio was 3.9 (95% CI, 0.97-15.6; P = .055). There was no recurrent thromboembolism.
Indications for enoxaparin anticoagulation therapy are presented in Table 2. More bleeding occurred in patients undergoing bridge therapy vs new anticoagulation (13.7% vs 8.1%). This trend was similar in both renal function groups. Reasons for bridging therapy and bleeding outcomes are reported in Table 3. A high frequency of major bleeding (5 of 17 patients [29.4%]) was noted in patients undergoing cardiac ablation procedures receiving bridge therapy with standard doses of enoxaparin. When these procedures were excluded, the overall incidence of major bleeding remained substantial (18.9%) in the group with moderate renal impairment.
This study evaluated the safety of enoxaparin administered in recommended doses to patients with moderate renal impairment. The results reveal a disturbing rate of major bleeding in 1 of 4.5 patients with moderate renal impairment, constituting an odds ratio of 4.7, compared with patients with normal renal function. Bleeding severe enough to require hospital admission occurred in 1 of 7.5 patients in the group with moderate renal impairment. Quiz Ref IDThis rate of major bleeding is substantially higher than that in the large trials24- 31 of enoxaparin. Of note, these large trials included homogeneous patient groups (ie, 1 indication), multiple exclusions, and short durations of enoxaparin therapy. In contrast, our study was longer and is more representative of everyday, real-world practice of using enoxaparin for mixed indications and longer duration, with concomitant antiplatelet or anticoagulant agents, varied renal function, and bridging therapy.
Our results are consistent with those of other studies that showed increased bleeding in patients with moderate renal impairment. A pharmacokinetic-pharmacodynamic analysis of the Thrombolysis in Myocardial Infarction (TIMI) 11A trial9 revealed the risk of hemorrhage to be strongly associated with moderate renal impairment. Enoxaparin clearance was reduced by 21% in this subgroup and the risk of a major hemorrhagic event was increased 1.4 to 2.8 times. Thorevska et al10 reported an 11% incidence of major bleeding with use of manufacturer-recommended doses in patients with moderate renal impairment. Barras et al12 demonstrated that 23% of patients with moderate renal impairment receiving the manufacturer-recommended therapeutic dose of enoxaparin experienced major bleeding, a result strikingly similar to ours. The Fifth Organization to Assess Strategies in Ischemic Syndromes (OASIS-5) trial32 documented increased bleeding events as renal function progressively declined below an estimated glomerular filtration rate of 58 mL/min/1.73 m2 and revealed renal function to be an important independent determinant of major bleeding. Each 10-mL/min decrement of CrCl was associated with increased bleeding risk when enoxaparin was used with thrombolytics for the treatment of ST-segment elevation myocardial infarction.14 In a multivariate regression model, CrCl remained an independent risk factor for intracranial and major bleeding, leading these authors to suggest an empirical dose adjustment to maintain the best risk-benefit ratio.14
The mechanism responsible for increased bleeding in patients with reduced renal function appears to be increased anticoagulant effect secondary to drug accumulation. Studies5- 7,9,15,17,33 evaluating enoxaparin pharmacokinetics via antifactor Xa activity have consistently shown increased drug half-life and accumulation of antifactor Xa activity in patients with moderate renal impairment. The consensus of these studies is that dose adjustment is necessary to avoid drug accumulation. However, a limitation of pharmacokinetic studies is the unclear relationship of antifactor Xa activity to thrombosis resolution or bleeding.18,19,34- 37 Thus, the practice of monitoring antifactor Xa activity to adjust doses has not been widely adopted. The results of our study combined with those cited in the first section of this paragraph suggest the potential usefulness of measuring antifactor Xa activity in patients with moderate renal function to assess anticoagulant accumulation and further stratify bleeding risk.
In an effort to bridge the gap between pharmacokinetic data and pharmacodynamic outcomes (bleeding), Barras et al38 studied a model to quantify the probability of bleeding based on enoxaparin “exposure.” They determined that bleeding as a function of total enoxaparin exposure was best described as the cumulative area under the concentration-time curve of antifactor Xa activity. Consequently, this analysis entails not only antifactor Xa activity at one point but also duration of therapy. Notably, the early major trials studied durations of therapy on average of only 4 days, which is markedly less exposure than in our study. The findings of Barras et al are important because they determined that cumulative exposure to anticoagulant levels over time was more directly related to bleeding than were antifactor Xa levels. Thorevska et al10 also showed duration of therapy of significance for occurrence of bleeding. We did not measure antifactor Xa; however, our study used recommended therapeutic doses in patients with moderate renal impairment that pharmacokinetic studies have consistently shown to provide elevated antifactor Xa activity. Our results support the concept that increased exposure to enoxaparin via duration of therapy and possibly antifactor Xa accumulation is correlated with increased bleeding (Table 4). Studies of shorter duration would be less apt to discover bleeding or correlation of an antifactor Xa level determined early in therapy to predict bleeding.
When our results are added to previous evidence and the pharmacokinetics of enoxaparin, we believe that there are ample data confirming enoxaparin accumulation and an increased risk of major bleeding in patients with moderate renal impairment. Considering the impact of major bleeding as an adverse drug event that may include life-threatening or fatal hemorrhage, we suggest that regulatory agencies, the manufacturer, and/or guideline-writing organizations explore all available data, published and unpublished, and consider revising dosing and/or monitoring guidelines in patients with moderate renal impairment. Others7,15,17 have offered alternative dosing strategies and, although use has been investigated in a limited number of patients, these regimens merit consideration (Table 5).
Limitations of this study include the retrospective data collection, therefore relying only on evaluation of clinical progress notes, laboratory test results, and other documentation. Our patient population consisted predominately of men and reflects patient selection from a population representative of a VA medical center. Quiz Ref IDThe decision to use enoxaparin vs other anticoagulants was not controlled, and bias may have been introduced without the proper randomization associated with prospective clinical trials. There also may be some residual confounding despite our efforts to control for differences in risk factors via multivariable regression. Several bleeding episodes followed cardiac ablation procedures (Table 3). It is possible that these bleeding events could have been avoided by using reduced doses of enoxaparin as recommended by some39 or continuation of warfarin without bridging therapy as suggested by others.40,41 The frequency of bleeding remained high even if these procedures were excluded. Despite these limitations, the comparative groups were similar in many demographics, including concurrent risk factors for bleeding. The demographics associated with reduced renal function (age, serum creatinine) differed as expected. It has been shown10 that patients with reduced renal function may be at higher risk for bleeding with unfractionated heparin therapy as well as enoxaparin therapy; therefore, renal impairment itself and/or comorbidities associated with chronic kidney disease may contribute to increased bleeding. Even if this has an effect on risk separate from our concerns of drug accumulation, a dose adjustment to offset this risk should be investigated.
In conclusion, our results indicate a significantly increased risk of major bleeding in patients with moderate renal impairment who receive the recommended therapeutic dose of enoxaparin. This caution includes using these doses for bridge therapy. Further investigation of alternative dosing regimens in patients with moderate renal impairment to maintain efficacy with reduced risks in this patient population is warranted, as are studies of the role of monitoring antifactor Xa activity to guide dosing of enoxaparin.
Correspondence: Douglas D. DeCarolis, PharmD, Pharmacy Service (119), Minneapolis Veterans Affairs Health Care System, One Veterans Drive, Minneapolis, MN 55417 (firstname.lastname@example.org).
Accepted for Publication: July 26, 2012.
Published Online: November 5, 2012. doi:10.1001/2013.jamainternmed.369
Author Contributions: Dr DeCarolis had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: DeCarolis, Thorson, Clairmont, and Johnson. Acquisition of data: DeCarolis, Thorson, Clairmont, and Leuthner. Analysis and interpretation of data: DeCarolis, Thorson, Clairmont, and Rector. Drafting of the manuscript: DeCarolis, Thorson, and Johnson. Critical revision of the manuscript for important intellectual content: DeCarolis, Thorson, Clairmont, Leuthner, and Rector. Statistical analysis: Rector. Administrative, technical, and material support: DeCarolis. Study supervision: DeCarolis and Johnson.
Conflict of Interest Disclosures: None reported.