Background
Current standard treatment for acute venous thromboembolism (VTE) in patients with active cancer is long-term low-molecular-weight-heparin (LMWH) [
1,
2]. This recommendation is based primarily on the ‘Comparison of Low-Molecular-Weight-Heparin vs Oral Anticoagulant Therapy for the Prevention of Recurrent Venous Thromboembolism in Patients with Cancer (CLOT)’ study which showed an overall 52% relative risk reduction in the rate of symptomatic, recurrent venous thromboembolism (rVTE) over 6 months with the LMWH dalteparin versus oral vitamin K antagonist (VKA) therapy without an increased risk of bleeding [
3]. This approach is effective in most patients, but patients with active cancer are heterogeneous. There are substantial differences in risk for rVTE and hemorrhagic complications because of differences in clinical risk factors [
4]. Individual risk factors for rVTE in cancer cohorts include locally advanced or metastatic cancer, antineoplastic treatment, primary tumor site, and recent diagnosis of malignancy [
2‐
6]. Key risk factors associated with an increased risk of major bleeding include metastatic disease, a recent bleeding event, and renal insufficiency [
4]. Importantly, substantial differences in clinical risk factors between patients with cancer-associated thrombosis may lead to different clinical outcomes despite therapeutic anticoagulation with the same agent. As a result, the relative benefit of LMWH versus VKA in patients with cancer-associated thrombosis at high or low risk of rVTE and bleeding is unclear.
We performed an exploratory, post hoc analysis of the CLOT study to compare the efficacy and safety of the LMWH dalteparin versus VKA in subgroups of patients with active cancer at high risk or low risk of rVTE and bleeding.
Methods
Study design and population
A detailed description of the CLOT study design, population, treatment regimens, and outcome measures has been published previously [
3]. The CLOT study was conducted in compliance with the ethical principles originating in, or derived from, the Declaration of Helsinki and in compliance with all International Conference on Harmonisation Good Clinical Practice Guidelines. All patients provided written, informed consent.
This exploratory post hoc analysis of data from the CLOT study compared the efficacy and safety of subcutaneous LMWH with that of oral VKA therapy in preventing rVTE in patients with active cancer at high or low risk of recurrent thrombosis and bleeding. High-risk patients were empirically defined as those with known metastatic disease and/or were receiving recent antineoplastic treatment (chemotherapy, radiation, and/or surgery 6 months prior to or at randomization) at baseline. These parameters, readily determined in clinical practice, are established independent risk factors for rVTE and bleeding during therapeutic anticoagulation [
4]. Low-risk patients excluded high-risk patients as defined above, and therefore, included patients with neither metastatic disease nor antineoplastic treatment at baseline.
LMWH patients were administered 200 IU of dalteparin per kilogram once daily for 30 days, with a maximum daily dose of 18,000 IU. For the remaining 5 months, dalteparin was administered at a dose of approximately 150 IU per kilogram once daily. If the platelet count fell to 50,000–100,000 per microliter during treatment, the dalteparin dose was temporarily reduced.
Oral VKA patients initially received 200 IU of dalteparin per kilogram (maximal daily dose, 18,000 IU) once daily and, within 24 h, a vitamin K antagonist for 6 months. Dalteparin was discontinued after a minimum of 5 days and once the international normalized ratio (INR) remained between 2.0 and 3.0 for 2 consecutive days. All VKA doses were adjusted to reach a target INR of 2.5 (therapeutic range 2.0–3.0). Target INR was reduced to 2.0 (range 1.5–2.5) if the platelet count fell to 50,000–100,000 per microliter during treatment. In all patients, treatment was to continue for 6 months.
Outcome measures
The primary efficacy outcome was the first episode of objectively diagnosed, symptomatic rVTE, including deep-vein thrombosis, pulmonary embolism, or both, during the 6-month study period in the intention-to-treat population (ITT; all randomized patients in the high- and low-risk patient subgroups).
Secondary outcomes included clinically overt bleeding during the 6-month study period (both any and major) and death. Major bleeding events were defined by an overt bleeding event associated with a fall in hemoglobin ≥ 2.0 g/dL, the need for a transfusion of ≥ 2 units of blood, involvement of a critical site (retroperitoneal, intracranial, intraspinal, intraocular, or pericardial areas), or an event leading to death. Any bleeding event also included all other overt hemorrhagic events that did not meet the criteria for classification as major bleeding. The safety analyses included patients who received at least one treatment dose.
All suspected rVTE and bleeding events were objectively reviewed and verified by an independent Central Adjudication Committee blinded to treatment allocation.
Statistical analysis
Descriptive statistics were provided for baseline characteristics, tumor and VTE history, and risk factors. Time to first event analyses were conducted for the high-risk group using the Kaplan–Meier method and data were compared using the log-rank test with a 5% significance level. Cox proportional hazard regression models were employed to assess the effect of treatment (LMWH versus VKA) and other potential prognostic factors on VTE recurrence, bleeding, and death events in the high-risk group. Time to first event analyses and multivariate analyses were not conducted for the low-risk group due to the small sample size (
n = 44). Competitive risk analysis for death was not performed in either group due to similar mortality rates between treatment arms [
7]. Significance was set at the 5% level. No statistical adjustments for multiple testing were made.
Discussion
To our knowledge, this exploratory, post hoc analysis is the first to study the efficacy and safety of LMWH versus VKA for the treatment of acute VTE in patients with active cancer at high or low risk of rVTE and bleeding complications. High-risk patients, 93% of the total CLOT population, were defined as having metastatic disease and/or receiving recent antineoplastic treatment at baseline, while low-risk patients had neither. These definitions were empirically derived, based on prior reports and clinical experience that metastatic disease and active cancer therapy are risk factors for recurrent thrombosis and bleeding. Our analysis showed that LMWH treatment for up to 6 months in high risk patients was more effective than VKA in reducing the risk of recurrent thromboembolism without increasing the risk of bleeding. At 6 months, the overall risk of recurrent thrombosis and major bleeding was 12% and 5%, respectively. Importantly, the 8% rate of rVTE at 6 months in the high-risk LMWH group of the CLOT study is consistent with the rate reported recently for LMWH in the DALTECAN Study [
9].
Among high-risk patients, there was a higher risk of rVTE in patients with metastatic versus non-metastatic cancer, which supports the empirical use of baseline metastatic cancer to define high-risk patients. Interestingly, the risk of rVTE was not higher in patients who received recent antineoplastic treatment versus those without such treatment. Together, these findings suggest that baseline metastatic disease may contribute more to increased risk of recurrent thrombosis than recent antineoplastic treatment in high-risk patients. Neither metastatic disease nor recent antineoplastic treatment was found to increase the rate of any or major bleeding events in the high-risk patient group. As a result, the risk of hemorrhagic complications under therapeutic anticoagulation may be more dependent on other risk factors than baseline metastatic disease or antineoplastic treatment. Recent findings from randomized trials comparing dalteparin with a direct oral anticoagulant (DOAC) [
10,
11], indicate that choice of anticoagulant is an important predictor of bleeding, and that DOAC therapy may not be the optimal treatment for cancer-associated thrombosis among those patients with risk factors for bleeding.
Treatment with LMWH in patients with active cancer at low risk of rVTE and bleeding complications showed similar rates of rVTE and bleeding with VKA. These results should be interpreted with caution due to the small number of subjects and imbalance of prognostic factors (eg, age, tumor type and site, hospitalization) in the low-risk treatment group. However, our results are consistent with the CATCH trial, which showed that the LMWH tinzaparin was not more effective than VKA in preventing rVTE [
12]. Although the CATCH trial used eligibility criteria similar to CLOT, it had fewer patients with metastatic disease and/or receiving anticancer therapy and better overall survival. Similar findings have also been reported in previous trials comparing LMWH and VKA for secondary prophylaxis of VTE in patients without cancer who are at low risk of recurrent events and hemorrhagic complications [
13,
14]. Consequently, patients with less advanced cancer or better prognosis who are at lower risk of rVTE may not experience the same benefit with LMWH as those with metastatic malignancies.
Data to date show that there are substantial and clinically relevant differences in baseline patient characteristics and prognosis among various clinical trials (Table
2). This heterogeneity is an important consideration for cross-study comparisons and network meta-analyses in determining optimal anticoagulant therapy. For example, compared with high-risk CLOT patients, fewer patients with cancer in the DOAC-VKA [
15‐
17], CATCH [
12], and Hokusai VTE Cancer (edoxaban versus dalteparin) [
10] trials had metastatic disease at baseline. Fewer patients in both the CATCH and Hokusai VTE Cancer trials had recent antineoplastic therapy at baseline or an Eastern Cooperative Oncology Group performance status of 2. Most importantly, all-cause mortality was lower in patients with cancer in DOAC-VKA trials, CATCH, and Hokusai VTE Cancer trials than in high-risk patients enrolled in CLOT. Notably, 6-month overall mortality was 26% in Hokusai VTE Cancer but 41% in high-risk CLOT. Despite a decade gap between these two trials, it is unlikely that differences in cancer therapeutics alone are responsible for the observed difference in mortality. As a result, patients with cancer in the DOAC-VKA, CATCH, and Hokusai VTE Cancer trials had better prognosis, with lower thrombotic and hemorrhagic risks, than overall and high-risk patients in CLOT. These findings are a strong reminder that extrapolation of clinical outcomes from one patient cohort to another may not be valid and should be avoided.
Table 2
Comparison of baseline characteristics and mortality rates in trials/subgroup analyses investigating DOAC or LMWH for the treatment of VTE in patients with cancer
Trial/subgroup analysis |
Patients with cancer at baseline, na | 462 | 221 | 771 | 900 | 1046 | 676 | 406 |
Patients with metastatic disease at baseline, n (%) | 101 (22) | 28 (13) | 46 (6) | 492 (55) | 554 (53) | 455 (67) | 236 (58) |
Patients receiving antineoplastic therapy at baseline, n (%) | NR | NR | NR | 476 (53) | 757 (72) | 525 (78) | 282 (69) |
ECOG PS = 2 | NR | NR | NR | 209 (23) | 247 (24) | 240 (36) | 95 (23) |
Mortality rateb, n (%) | 74 (16)c | 32 (14)d | 80 (10)c | 288 (32)d | 267 (26)d | 266 (40)d | 104 (26)d |
Overall, our results suggest that patients with cancer-associated thrombosis have different levels of risk for rVTE and bleeding which can lead to different clinical outcomes despite therapeutic anticoagulation with the same agent. As a result, clinicians must be mindful of patient and cancer characteristics when considering the choice of anticoagulant therapy. Our results, combined with recent clinical trials, suggest that patients with active cancer at high risk of rVTE and, particularly, bleeding are likely to benefit more from therapeutic anticoagulation with LMWH, whereas lower-risk patients can choose among LMWH, DOAC, and warfarin. Individual patient benefit-risk assessment and preference should be essential components of the therapeutic decision.
Interpretation of this post hoc analysis has limitations. First, CLOT did not stratify patients by specific risk factors, and was not powered to detect between-treatment differences in subgroup analyses [
3]. However, we used a priori documented baseline characteristics that have been established as important risk factors for thrombosis and bleeding. Second, we did not examine other risk factors such as biomarkers. Our approach was to use easily identifiable features for simple risk stratification. Finally, patient numbers in the low-risk non-metastatic disease/non-antineoplastic treatment subgroup were small, thus limiting the validity and value of our findings for this subgroup.
Compliance with ethical standards
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.