Comparing low-molecular-weight heparin dosing for treatment of venous thromboembolism in patients with obesity (RIETE registry)

Obesity is a growing epidemic affecting up to 11% of adults worldwide and 38% of Americans.¹  Obese people are at an increased risk of venous thromboembolism (VTE), compared with individuals of normal weight.²  Putative mechanisms include stasis secondary to decreased mobility and compression of intra-abdominal and femoral vessels,³  inflammation secondary to adipokines,⁴  and endothelial damage secondary to associated metabolic diseases, including diabetes and hypertension.⁵  With predictable pharmacokinetics and near 100% bioavailability, low-molecular-weight heparins (LMWHs) have been routinely used in the initial treatment of VTE, before transition to an oral agent.^6-8  However, the most appropriate LMWH dosing in patients with obesity remains unclear.

LMWHs are hydrophilic and therefore largely remain in the intravascular compartment. In obese individuals with disproportionately more adipose tissue, there is a concern about overdose and bleeding when treatment with standard dosing by actual body weight is applied.⁷  The alternative strategy, capping at a maximum dose regardless of actual body weight, may cause an underdose.⁹  There are no rigorous trials or observational studies comparing therapeutic LMWH dosing strategies in obese patients with VTE that examine clinical end points.¹⁰  The Canadian drug product monographs recommend capped dosing, given that LMWH preferentially distributes to plasma^11-13 ; however, practice patterns and the 2018 American Society of Hematology guidelines conditionally recommend uncapped LMWH doses¹⁴  because of concerns about underdosing and the danger of treatment failure. The 2018 American Society of Hematology guidelines conditionally recommend capped LMWH based on evidence of very low certainty.¹⁴  The evidence is based on a meta-analysis of 5 studies, none of which directly compared dosing strategies, revealing no difference in benefit or harm between indirect comparison of dosing strategies. Concerns about underdosing and the serious consequences of therapeutic failure are cited as justification for the recommendation.

The RIETE (Registro Informatizado Enfermedad TromboEmbólica) database is an ongoing, multinational, observational registry of consecutive patients with objectively confirmed, acute VTE (registered at www.clinicaltrials.gov as #NCT02832245). The database was founded in Spain in 2001 and has expanded to >15 countries since then. The rationale and methodology of RIETE has been published.¹⁵  Patients are excluded if they are currently participating in a therapeutic clinical trial with a blinded therapy. All patients (or their relatives) provide written or oral consent for participation in the registry, in accordance with local ethics committee requirements.

The study population included patients registered in the RIETE registry who met 4 criteria: age ≥18 years; diagnosis of acute, objectively confirmed deep vein thrombosis (DVT) or pulmonary embolism (PE); obesity; and initial therapy with LMWH for any treatment duration. Our concern in specifying a minimum for treatment duration was the risk of excluding patients who experienced adverse events and therefore had LMWH discontinued. Obesity was defined as body weight >100 kg, a value used by other studies¹⁰  as a pragmatic proxy, because height and body mass index (BMI) are not always recorded in clinical practice. For individual patients, the longest treatment phase on LMWH following the qualifying VTE was chosen for analysis. Patients treated with thrombolytics were excluded from analysis.

We compared capped dosing (chosen as a dose of 18 000 IU/d, the typical suggested capped dose for LMWH) to noncapped LMWH dosing (any dose equivalent to >18 000 IU/d). Only patients treated with enoxaparin, dalteparin, or tinzaparin were included, given that the recommended capped dose is 18 000 IU for all 3 drugs in the Canadian product monographs.^11-13 

The following parameters were collected when the qualifying episode of VTE was diagnosed: sex, age, body weight, and, in most cases, height; risk factors for VTE: recent immobility (ie, total bed rest with bathroom privileges for ≥4 days in the 2-month period before VTE diagnosis), recent surgery (in the 2 months before VTE), active cancer (defined as newly diagnosed cancer, metastatic cancer, or cancer that was receiving treatment), hormonal therapy, pregnancy, recent birth prior VTE, and recent travel; and presence of coexisting conditions: chronic heart or lung disease, concomitant therapies, recent (<30 days before enrollment in RIETE) major bleeding, and laboratory data (also at baseline), including complete blood counts and serum creatinine levels.

Patients were treated according to the clinical practice of each participating hospital (ie, there was no standardized therapy). The drug, dose, and duration of therapy were recorded. The decision on the type and duration of anticoagulant therapy was left to the attending physicians. Patients were followed up in the outpatient clinic (or by telephone interview if they could not visit the clinic). During each visit, any signs or symptoms suggesting VTE recurrences or major bleeding were noted.

The primary outcome was the composite of recurrent VTE, major bleeding, or all-cause death attributable to LMWH treatment. The secondary outcome was the composite outcome at 15 days, whether the patient had transitioned from LMWH to another anticoagulant or not.

Major bleeding was defined as overt bleeding requiring a transfusion of 2 or more units of blood or retroperitoneal, spinal, intracranial, or fatal bleeding episodes. Bleeding was attributed to LMWH therapy when it occurred during LMWH treatment or within 3 days of discontinuation of LMWH. Our decision to include up to 3 days after discontinuation was based on 3 factors: (1) there may be a delay in the diagnosis of bleeding; (2) the half-life of LMWH is increased when the glomerular filtration rate is reduced; (3) doses in this population are at and above the upper limits of the studied doses and therefore may be associated with altered kinetics and dynamics.

Each episode of suspected recurrent VTE was investigated by repeat compression ultrasonography, lung scan, computed tomographic pulmonary angiography or pulmonary angiography, as chosen by the treating physician. Most outcomes were classified as reported by the clinical centers; there was no central adjudication of outcome events. VTE recurrences were attributed to LMWH if they occurred while the patient was receiving therapy or during or within 7 days of discontinuation of treatment. Attributing clots within 7 days of discontinuation encompasses cases in which there is delayed diagnosis.

Similarly, deaths were considered attributable to LMWH therapy if they occurred during therapy or within 7 days of treatment discontinuation.

All calculations were performed with SPSS Statistics (IBM). Differences in patient and treatment variables were assessed with the χ² test for categorical variables; the independent Student t test for normally distributed, continuous variables; and the Mann-Whitney U test for continuous variables without normal distribution.

Unadjusted rates of the primary and secondary outcomes were calculated during LMWH therapy. Binary multivariable logistic regression was used to calculate adjusted odds ratios [ORs] for the association of treatment (capped vs uncapped) with the primary composite outcome during LMWH therapy and also at 15 days from initiation of therapy (regardless of transition to another agent) as a secondary outcome. Potential confounders were entered into the multivariable model based on associations from the literature.^16,17  Variables were kept in the model if they were associated with the outcome (P < .1). The following variables were entered as potential confounders: age, sex, history of chronic heart or lung disease, recent immobility, cancer, metastatic cancer, recent major bleeding, anemia, thrombocytopenia, renal insufficiency, initial presentation of VTE, type of LMWH used for initial therapy, concomitant therapy with antiplatelets, and long-term therapy with vitamin K antagonists. We planned to compare the effect of dosing strategy in patients with and without metastatic cancer. Of the variables included, 3 had missing data: concomitant antiplatelets (208 missing values) and platelet and hemoglobin count (3 missing values for both counts). The missing data were imputed by the mode value.

A total of 2846 patients weighing >100 kg were included: 454 (16%) received capped doses (18 000 IU/d), and the remaining 2392 received uncapped doses (mean, 20 865 IU/d). Mean treatment duration was 14.8 ± 20.6 and 14.3 ± 32.3 days, respectively. Patients were followed up for a median of 300 days and a mean of 501 days, respectively. Patients who received capped doses were slightly older (56.8 ± 15.8 and 54.8 ± 15.0 years, respectively), weighed less (106.9 ± 8.6 and 111.3 ± 13.5 kg), and were more likely to have active cancer and to initially present with DVT (vs PE) than those receiving uncapped doses (Table 1). In addition, patients who received capped doses of LMWH were more likely to be prescribed tinzaparin or dalteparin, but less likely to receive enoxaparin, and were also less likely to have concomitant therapy with antiplatelets (Table 2). For long-term therapy, most patients switched to vitamin K antagonists (70% vs 74%), but a higher proportion of those with capped doses (24% vs 16%) continued to receive LMWH for long-term therapy. The breakdown of weights between groups is reported in Table 3.

Forty-nine total events were attributable to LMWH therapy, according to the prespecified rule of counting events if they occurred within 3 or 7 days of discontinuation (Table 4). Two events were counted in the capped-dose group: 2 PEs (0.44%) and 1 death attributable to PE (0.22%). There were no bleeding events. (Only the first event in a single individual was counted toward the composite, therefore the sum of individual events do not match the total.) In the uncapped dose group, there were 47 events: 3 PEs (0.13%), 3 DVTs (0.13%), 24 major bleeding incidents (1%), and 21 deaths (0.88%).

During the first 15 days of anticoagulant therapy (including events not attributable to LMWH by our rule), 11 patients (0.39%) developed VTE recurrences (7 recurrent PE, 4 recurrent DVT), 26 (0.91%) had major bleeding, 22 (0.77%) died (10 died of PE, 1 of bleeding), and 54 (1.9%) had at least 1 of the composite outcomes (Table 5).

We prespecified that patients with metastatic cancer would have a higher risk of death, thrombosis, and bleeding. However, we did not have sufficient power to test statistically for interaction of dosing strategy with metastatic cancer. The composite outcomes of the capped vs uncapped strategy in metastatic cancer at 15 days were 0 vs 11 events (0% vs 10.4%; P = .051). Table 6 includes detailed outcomes for this subgroup.

Multivariable binary logistic regression revealed that patients receiving capped doses of LMWH were at a lower risk for composite outcomes, both during LMWH therapy (OR, 0.16; 95% confidence interval [CI], 0.04-0.68) and at 15 days (OR, 0.17; 95% CI, 0.04-0.71; Table 7). This effect was mainly due to a trend toward lower rates of unadjusted major bleeding: no episodes in the capped-dose group vs 24 in the uncapped-dose group (Figure 1).

The optimal dose of LMWH therapy for obese patients with acute VTE is guided by limited pharmacokinetic data with no substantiating clinical data. To our knowledge, there are no data in the literature that compare rates of clinical events in obese patients with VTE treated with capped vs uncapped doses. Ours is the first study to compare these 2 therapeutic LMWH dosing strategies on the basis of patient-important clinical outcomes.

Our data, obtained from RIETE, a large registry of patients with VTE, reveal that 1 in every 6 patients (19%) weighing >100 kg was prescribed capped doses of LMWH, whereas the remainder received an uncapped dose. The concern regarding uncapped doses is the subtherapeutic anticoagulant effect resulting in otherwise avoidable thromboembolisms.⁹  In our study, 1.5% of patients experienced major bleeding and 1.2% had a VTE when exposed to uncapped dosing, compared with 0.22% and 0.66%, respectively, of those receiving a capped dose at 90 days after therapy initiation (results are not statistically significant for all individual comparisons). After adjustment for multiple potential confounders, patients with obesity who were receiving capped doses were at a lower risk of having the composite outcome of VTE recurrences, major bleeding, or all-cause death at 15 and 90 days.

This study has several significant limitations, broadly categorized as risk of significant bias and problems relating to the analysis. First, data were derived from the RIETE registry, which is a large, prospective case series of patients with VTE and may be subject to selection bias. Second, there were several prognostic differences at baseline between the 2 groups: patients receiving capped doses were significantly more likely to have active and metastatic cancer, to weigh less, and to have DVT (vs PE) and were significantly less likely to have received concomitant antiplatelet therapy. Despite our attempt to control for these potential confounders in multivariable analysis, residual confounding may partially explain the difference in outcomes between capped and uncapped dosing. Moreover, some baseline differences point toward confounding by indication: for example, those who received capped doses typically weighed less and were more likely to receive enoxaparin than those who received capped doses (91% vs 28%). Further, dosing regimens were different for tinzaparin 175 μg/kg per day vs dalteparin 200 IU/kg per day vs enoxaparin 1 mg (∼100 IU/kg) twice daily or 1.5 mg daily. A maximum unit threshold that is the same across each drug and dose strategy means that the weights at which a dose should be defined as capped vs not capped can vary, as would the pharmacologic effects. Third, outcome events were not adjudicated centrally. As physicians were aware of treatment allocation, it is possible that their knowledge of dosage influenced their reporting of clinical events. In terms of the multivariable analysis, there was a relatively low number of events, especially in the capped-dose group, which may have led to overfitting in multivariable analysis. In addition, the mean actual dose difference between the 2 groups was small (18 000 vs 20 865 IU/d). Finally, our composite analysis assumed that an episode of VTE, a major bleeding episode, and all-cause death had equal value. Higher doses could cause both more bleeding and decreased VTE, reducing the value of this composite end point.

Given the higher percentage of patients with active and metastatic cancer in the capped-dose group and the lower unadjusted event rates in the group, it seems improbable that capped dosing is harmful in either subgroup.

Our results are potentially valuable, however, as they suggest that capped dosing may have a role in obese patients with VTE who are at high risk for bleeding. Indeed, our findings are consistent with a meta-analysis of the bariatric surgery literature suggesting higher rates of bleeding episodes, without reduction in rates of VTE, using weight-adjusted heparin prophylaxis dosing.¹⁸  We emphasize that our findings are at high risk of bias and should not change current practice, but should spur further investigation in the form of an appropriately powered randomized controlled trial.

The RIETE investigators express their gratitude to Sanofi Spain for supporting this registry with an unrestricted educational grant; the RIETE Registry Coordinating Center, S&H Medical Science Service, for quality control data, logistic, and administrative support; and Salvador Ortiz, (Universidad Autónoma Madrid) and Silvia Galindo, Statistical Advisors with S&H Medical Science Service, for providing ongoing statistical analysis and assistance with respect to the data.

The RIETE registry receives funding from Sanofi Spain.

Contribution: R.M. conceived and coordinated the study, designed the study, conducted the initial analysis, and wrote the initial draft; R.N. played a central role in formulating the question, conducting statistical analyses, and editing the manuscript. J.J.L.-N., R.B., A.A., C.F., R.I., and E.G. contributed significantly to the interpretation of the findings and the writing of the manuscript, in addition to contributing to the database itself; M.C. helped recruit experts to choose and interpret the outcomes and made major contributions to the text, leading to substantial revisions; and M.M. provided the data, recruited a team that assisted throughout the study, and assisted at each stage of the process.

Conflict-of-interest disclosure: R.I. has participated on an advisory board once with both Apsen Pharmacare and LEOPharma. M.C. is receiving or has received monetary compensation from Bayer, Pfizer, CSL Behring, Servier Canada, Diagnostica Stago, Asahi Kasei, and Alnylam. The remaining authors declare no competing financial interests.

A complete list of RIETE (Registro Informatizado de Enfermedad TromboEmbólica) investigators appears in the supplemental appendix.

Correspondence: Reza Mirza, Center for Prognosis in Rheumatic Diseases, Toronto Western Hospital, 399 Bathurst St, Toronto, ON M5T 2S8, Canada; e-mail: reza.mirza@mail.utoronto.ca.