Risk-Based Thromboprophylaxis Reduces Bleeding, Mortality in Lung, GI Cancers

Despite evidence-based guidelines, prior trials, and extensive clinical experience using antithrombotic agents for primary prophylaxis, clinical application remains a major hurdle among individuals receiving anticancer therapies in the ambulatory setting. This issue is attributed to the heterogeneity and dynamic nature of risk profiles and bleeding concerns among individuals.

Randomized clinical trials have demonstrated the benefits and feasibility of risk stratification and risk-directed thromboprophylaxis. The findings, however, may not be applicable to all tumor groups given the use of a risk assessment model (RAM) heavily weighted for tumor type, with limited stratification within tumor types, as well as residual risk masked by the exclusion of low-risk individuals.

According to Marliese Alexnader, PhD, and colleagues, “alternate RAMs are available but remain limited by low sensitivity or specificity, nonroutine tests, and/or lack of ability to stratify risk within tumor groups.” In a new study, the researchers assessed the clinical benefits and safety of biomarker-driven thromboprophylaxis among patients with lung and gastrointestinal (GI) cancers receiving anticancer therapies in ambulatory treatment settings. Results were reported in JAMA Oncology [published online September 21, 2023; doi:10.1001/jamaoncol.2023.3634].

Targeted Thromboprophylaxis in Ambulatory Patients Receiving Anticancer Therapies (TARGET-TP) was on open-label, randomized, phase 3 study conducted between June 2018 and July 2021 (with a 6-month primary follow-up). A total of 328 patients aged 18 years and older with a life expectancy of at least 6 months were enrolled prior to beginning systemic anticancer therapies with or without radiotherapy for lung or GI cancer at five hospitals in Australia. The mean patient age was 65 years, 46% were women, and 54% were men. The fibrinogen and d-dimer RAM stratified patients into low-risk (observational arm) and high-risk (randomized arms) cohorts. A total of 61% of patients had GI cancer, 39% had lung cancer, and 40% had metastatic disease; 200 were high risk (100 in each group), and 128 were low risk. Baseline characteristics were comparable between the randomized arms, although those in the low-risk group had higher rates of GI cancer diagnoses (73%) and nonmetastatic disease (72%). High-risk patients were randomly assigned 1:1 to receive enoxaparin 40 mg subcutaneously daily for up to 90 or 180 days according to ongoing risk or no thromboprophylaxis (control group).

The primary outcome measure was the rate of thromboembolism (deep vein thrombosis, pulmonary embolism, and arterial thromboembolism) at 180 days. Secondary outcome measures included bleeding, survival, and risk model validation.

At 180-day follow-up, thromboembolism had occurred in 8% of patients in the high-risk group and 23% in the no thromboprophylaxis group (hazard ratio [HR], 0.31; 95% CI, 0.15-0.70; P=.0005). This result equated to an absolute risk reduction of 15% and relative risk reduction of 65%. The number needed to treat (NTT) to prevent a thromboembolic episode was 6.7. Additionally, 8% of low-risk patients in the control group experienced thromboembolism (high-risk control vs low-risk control: HR, 3.33; 95% CI, 1.58-6.99; P=.002).

Risk model sensitivity was 70%, and specificity was 61%. Researchers recognized that ideally those numbers should have been higher, but compared with other available RAMs, “it is potent and ready for scaled implementation with a routine test available in skin diagnostic and pathology laboratories.”

A post hoc intervention period analysis showed that thromboembolism occurred in 2% and 23% of patients randomized to enoxaparin and the control group, respectively (HR, 0.10; 95% CI, 0.02-0.44; P=.002; NNT, 4.8). Thromboembolism was reported in 28% of patients with lung cancer and 19% of patients with GI cancer. Among the low-risk cohort, the rate of thromboembolism was 5% when excluding catheter-related and arterial events.

There was no significant difference between the groups in the incidence of major bleeding, which occurred with 1% of the enoxaparin group, 2% of the high-risk no-treatment group, and 2% of the low-risk group. Nonbleeding events occurred in 13% of patients in the enoxaparin arm and included bruising (n=12), hematoma (n=2), and thrombocytopenia (n=1).

Treatment with enoxaparin correlated with a significant decrease in 6-month all-cause mortality (13% in the enoxaparin group vs 26% in the high-risk control group; HR, 0.48; 95% CI, 0.24-0.93; P=.03) but not in cancer progression (42% vs 47%, respectively; HR, 0.78; 95% CI, 0.51-1.19; P=.25). The 6-month all-cause mortality rate in the low-risk cohort was 7% (high-risk control vs low-risk control: HR, 2.14; 95% CI, 1.38-3.12; P=.001).

“The improved magnitude of thrombosis risk reduction and new finding of mortality risk reduction are likely explained by the selection of different treatment groups (better risk stratification) compared with prior trials,” wrote the study authors.

The study had limitations. The authors noted the study was inferior to a double-blind, placebo-controlled design with objective end points of radiologically confirmed thromboembolism and all-cause mortality. The risk assessment model validation was limited to lung and GI cancers.

“The fibrinogen and d-dimer RAM identified individuals with lung and gastrointestinal cancers who would most benefit from thromboprophylaxis and those who could avoid intervention,” concluded the researchers. “Future investigations in expanded tumor groups that incorporate clinician and consumer preference for choice of anticoagulant are warranted.”