Background
Among patients with cancer, thrombosis is a major cause of morbidity and mortality.1 Risk of venous thromboembolism (VTE) is predicted to be 4- to 7-fold higher in patients with cancer compared with those without cancer.2,3 The overall risk of VTE in patients with cancer was found to be 13.0 per 1,000 person-years (PY).3 Cancer-associated VTE is associated with significantly greater risks of recurrent VTE and major bleeding (MB) compared with VTE in patients without cancer.4,5 The incidence of recurrent VTE in patients with active cancer was reported as 9.6 per 100 PY, with an incidence of 22.1 per 100 PY in the first 6 months after VTE diagnosis.6 Therefore, patients with VTE and active cancer require longer periods of anticoagulation treatment.7
Results of several clinical trials have shown similar efficacy of direct oral anticoagulants (DOACs) to low-molecular-weight heparins (LMWHs) for recurrent VTE in patients with VTE and active cancer receiving anticoagulation treatment over 6 months.8–12 Similar rates of bleeding have been observed with apixaban and LMWHs during clinical trials in patients with VTE and active cancer, but higher rates of MB and clinically relevant nonmajor bleeding (CRNMB) have been observed with edoxaban and rivaroxaban, respectively.8,9 Based on this evidence, DOACs are included as an option in clinical guidelines for the treatment of VTE in patients with active cancer. For prevention of recurrent VTE in patients with cancer, guidelines generally recommend extended anticoagulation therapy (ie, LMWH, DOAC) for triggered events (eg, postsurgical), and indefinite anticoagulation for patients with active malignancy, a persistent thrombophilic state or an unprovoked thrombotic event, or metastatic disease, or in those receiving chemotherapy.13–16
There is a paucity of real-world evidence examining the comparative safety and effectiveness of apixaban and LMWH for extended treatment after the initial treatment period for VTE among patients with active cancer. Thus, our aim was to assess and compare the risk of MB, CRNMB, and recurrent VTE events in patients with VTE and active cancer who received extended anticoagulant treatment with apixaban or LMWH beyond the initial 3 months.
Methods
Study Design
This was a real-world retrospective cohort study of adult patients (age ≥18 years) diagnosed with VTE and active cancer who had extended VTE treatment beyond the initial 3 months of primary treatment from July 1, 2016, to April 30, 2022. The first evidence of VTE diagnosis in any position (primary or secondary) in any setting (either inpatient or ambulatory) during the identification period (January 1, 2017, to April 30, 2022) was designated as the index VTE event (Figure 1). The time difference between the overall study period and the identification period is due to the 6-month baseline assessment period. Patients with a pharmacy or medical claim for apixaban or an LMWH (not used as a bridge therapy) during the 30-day period following the index VTE event were included; the first observed pharmacy or medical claim date was designated as the initial anticoagulant date. The end of the initial 3 months of continuous anticoagulant treatment (defined as use for ≥3 months without a gap of >30 days or a switch to another anticoagulant) following the index VTE event was designated as the index date. The baseline period was defined as 6 months prior to the initial anticoagulant date until the index date (index date was included in the baseline period). Patient data were assessed from the day after the index date until the earliest of: date of disenrollment, discontinuation of index anticoagulant, switch to another anticoagulant, or end of the study period. Patients were assumed to remain on the index anticoagulant during extended treatment until censored.
Study timeline.
Abbreviation: VTE, venous thromboembolism.
Citation: Journal of the National Comprehensive Cancer Network 22, 6; 10.6004/jnccn.2024.7016
Data Source
Data used for this study were obtained from the Komodo Health Healthcare Map, which included closed claims from >140 million patients enrolled in a US commercial health care plan from 2012 to present. These encounters from a HIPAA-compliant database have census-level representation across patient populations (ie, age, US geography, risk pools), including hospital networks, physician networks, health care claim processing companies (ie, claims clearinghouses), pharmacies, and health insurers. This study used closed claims, which were health care encounters that came directly from the payer, in order to provide a complete patient journey, including full medical and/or prescription benefit information, insurance eligibility, and insurer-reported costs.
Study Population and Study Cohorts
Active cancer was defined by ≥2 medical claims for cancer diagnosis on different days or 1 claim for cancer diagnosis plus ≥1 claim for cancer treatment (eg, chemotherapy, radiation, cancer-related surgery) within 6 months before or 30 days after the index VTE event (Figure 2). Patients were also required to have continuous health plan enrollment for ≥6 months prior to and on the initial anticoagulant date and for ≥3 months enrollment after the initial anticoagulant date. Patients with evidence of atrial fibrillation/flutter, mechanical heart valve, inferior vena cava filter, antiphospholipid syndrome, or pregnancy during the study period were excluded. Patients were also excluded if they had evidence of VTE during the 6-month period preceding the index VTE event, unless one of the following conditions was met: the index VTE event occurred in the inpatient setting and was preceded by an outpatient VTE event within 7 days, or there was evidence of any oral or parenteral anticoagulant use that was not considered prophylactic in the 6 months prior to the index VTE event.
Patient attrition.
Abbreviations: AF, atrial fibrillation; APS, antiphospholipid syndrome; IPTW, inverse probability treatment weighting; IVC, inferior vena cava; LMWH, low-molecular-weight heparin; OAC, oral anticoagulant; PAC, parenteral anticoagulant; VTE, venous thromboembolism.
Citation: Journal of the National Comprehensive Cancer Network 22, 6; 10.6004/jnccn.2024.7016
The patients with active cancer and VTE were further classified into the following cohorts based on their initial anticoagulant treatment. The LMWH cohort was defined as patients with a claim for LMWH within 30 days after the index VTE event and with a treatment duration of ≥14 days and no other anticoagulation between the index VTE event and 14 days after LMWH initiation. The apixaban cohort was defined as patients who initiated apixaban within 30 days after the index VTE event and did not have a claim for another anticoagulant between the index VTE event and apixaban initiation. The cohorts were further defined to have extended treatment based on their continuous treatment with the index anticoagulant (apixaban or LMWH) after initiating treatment. Continuous treatment was defined as utilization of initiated anticoagulant for ≥3 months without a gap of >30 days between the treatment or switch to another anticoagulant.
Study Outcomes
Patient demographics (age, sex, geographic region), clinical characteristics (baseline comorbidities, medications), and VTE-related characteristics were evaluated during the baseline period. Cancer-related clinical characteristics were evaluated 6 months on or before the initial anticoagulant date and included cancer site, cancer type, cancer risk, cancer-related treatment, and metastases.
MB, CRNMB, and recurrent VTE were the safety and effectiveness outcomes that were evaluated 1 day after the index date during the follow-up period.
An MB event was identified using hospital records with an MB diagnosis listed as the primary diagnosis (by ICD-9-Clinical Modification or ICD-10 diagnosis or procedure code). MB was further stratified as gastrointestinal (GI) bleeding, intracranial hemorrhage (ICH), and other MB (genitourinary bleeding, respiratory tract bleeding, ocular bleeding, joint bleeding/hemarthrosis, transfusion of blood and blood components, other bleeding, or no bleeding site specified). A CRNMB was either an inpatient admission for a bleeding event that did not qualify as an MB event (ie, had a bleed diagnosis in the secondary position and the bleeding was for a noncritical site) or an ambulatory care visit for noncritical bleeding. CRNMB was further stratified as GI bleeding and other bleeding. A recurrent VTE event was identified as an acute care inpatient admission with a corresponding primary diagnosis.
Statistical Analysis
Patient demographic and clinical characteristics were analyzed descriptively and reported as means and standard deviations for continuous measures and frequency distributions for categorical variables. Incidence of the outcomes was calculated as the number of events per 100 PY.
For comparisons between the apixaban and LMWH treatment, standardized mean differences (SMDs) were calculated for each variable. SMDs ≤0.1 were considered balanced, whereas SMDs >0.1 were significantly different.
The inverse probability treatment weighting (IPTW) method was used to balance patient characteristics between the apixaban and LMWH cohorts in the baseline period. Propensity scores were used to obtain estimates of the average treatment effect using a logistic regression model with the LMWH cohort as the reference. Model covariates included age, sex, geographic region, type of VTE diagnosis, cancer characteristics (cancer type, risk, and metastases), Charlson comorbidity index score, baseline bleeding event, comorbidities, and medication use. After generating the propensity scores, the distribution of the propensity scores was reviewed. Each patient was weighted by the inverse of the probability of their treatment (weight=1/propensity score). The weights were stabilized by multiplying the original weights with a constant that was equal to the expected value of being in the apixaban or LMWH cohort, which reduced variance of the weighting and treatment effect estimates.
Median time to MB, CRNMB, and recurrent VTE was calculated from the day after the index date to the date of the assessed outcome using Kaplan-Meier analysis. The log-rank test was used to compare differences between treatment groups. Cox proportional hazard models were used to estimate the hazard of MB, CRNMB, and recurrent VTE between the treatment groups. Hazard ratios (HRs) and 95% confidence intervals were reported using adjusted Cox regression models when no significant departure from proportionality was indicated.
All statistical tests were performed as 2-sided hypotheses of no difference between the cohorts at a significance level of 0.05. Snowflake (version 6.7.1; Snowflake Inc.) and R (version 3.6.3; R Foundation for Statistical Computing) were used for statistical analyses.
Results
Study Cohort and Patient Characteristics
A total of 16,441 patients with active cancer and VTE with ≥3 months of primary treatment with apixaban (n=13,527) or LMWH (n=2,914) were included pre-IPTW (Figure 2). Overall, disenrollment rates between the 2 treatment groups were similar (29.6% and 31.0% for the LMWH and apixaban groups, respectively). In the total population (n=16,441), the mean [SD] follow-up time was 185 [227] days for the apixaban group and 124 [178] days for the LMWH group (P<.0001; SMD=0.24). After IPTW, all SMDs from each pairwise comparison were <0.1, indicating that patient demographic and clinical characteristics were balanced between the groups (Table 1, Supplementary Table S1 in the supplementary materials, available online with this article). Post-IPTW, the mean [SD] age in the apixaban (weighted, n=13,564) and LMWH (weighted, n=2,808) groups was 63.8 [12.8] and 63.5 [12.0] years, respectively. The majority of patients in the apixaban and LMWH cohorts had nonhematologic cancers (78.9% and 80.9%, respectively), and approximately 25% from both cohorts had very high-risk cancer. The index VTE event occurred in the inpatient setting in approximately 50% of patients in both groups. Pre-IPTW patient demographic and clinical characteristics are provided in Supplementary Table S2.
Post-IPTW Baseline Demographic and Clinical Characteristics
Outcomes
Pre-IPTW, the unadjusted incidence rate (IR) for MB was significantly lower in the apixaban group versus the LMWH group (5.9 vs 13.8 per 100 PY, respectively) (P<.0001). Likewise, the unadjusted IR for CRNMB was lower in the apixaban group versus the LMWH group (25.6 vs 38.4 per 100 PY, respectively) and for recurrent VTE (3.8 vs 10.4, respectively; P<.0001 for all comparisons) (Supplementary Table S3).
In the follow-up period, the adjusted IRs per 100 PY for recurrent VTE (4.1 vs 9.6), MB (6.3 vs 12.6), and CRNMB (26.1 vs 36.0) were lower for apixaban than LMWH (P<.0001 for all comparisons). The extended use of apixaban was associated with a lower adjusted risk of recurrent VTE (HR, 0.42; 95% CI, 0.34–0.53; P<.0001), MB (HR, 0.50; 95% CI, 0.41–0.61; P<.0001), and CRNMB (HR, 0.76; 95% CI, 0.68–0.85; P<.0001) versus LMWH (Figure 3). Lower risks of subcategories of MB (GI, intracranial, other) and CRNMB (GI, other) were also observed with apixaban (Supplementary Figures S1–5).
Adjusted outcomes for patients with VTE and active cancer after the first 3 months of treatment, receiving extended anticoagulation treatment with apixaban or LMWH.
Outcomes were assessed during the entire follow-up period. The denominator for calculating IR considered the person-years for all patients after weighting until either event is observed or they are censored. The numerator was the weighted cohort following IPTW (apixaban: post-IPTW, weighted n=13,564; LMWH: post-IPTW, weighted n=2,808).
Abbreviations: CRNM, clinically relevant nonmajor; GI, gastrointestinal; HR, hazard ratio; IPTW, inverse probability treatment weighting; IR, incidence rate; LMWH, low-molecular-weight heparin; VTE, venous thromboembolism.
Citation: Journal of the National Comprehensive Cancer Network 22, 6; 10.6004/jnccn.2024.7016
Discussion
In the extended treatment setting of patients with cancer-associated VTE, the findings of this real-world study revealed that apixaban was associated with significantly lower risks of recurrent VTE, MB, and CRNMB versus LMWH. These results complement findings from clinical trials10,11 and a real-world database analysis,17 which have shown apixaban to be noninferior to or to have better efficacy/effectiveness and safety profiles than LMWH during the 6 months of treatment. In the ADAM-VTE trial, which included 300 patients with cancer and VTE, apixaban was associated with lower rates of recurrent VTE at 6 months (0.7% vs 6.3%; HR, 0.099 [95% CI, 0.013–0.78]; P=.0281), whereas both apixaban and dalteparin, an LMWH, were associated with low rates of MB (0% vs 1.4%; HR, not estimable; P=.138).10 Results of the CARAVAGGIO trial showed that apixaban was noninferior to dalteparin with respect to recurrent VTE at 6 months (HR, 0.63 [95% CI, 0.37–1.07]; P<.001 for noninferiority), and similar rates of MB were observed between the 2 study groups.11
Real-world evidence for apixaban in patients with VTE and active cancer has shown a lower risk of recurrent VTE, MB, and CRNMB compared with LMWH during the 6 months of treatment. A retrospective claims database study comparing apixaban (n=3,393; mean follow-up of 3.5 months) and LMWH (n=6,108; mean follow-up of 2.9 months) among patients with VTE and active cancer showed apixaban was associated with lower risks of recurrent VTE (HR, 0.61; 95% CI, 0.47–0.81), MB (HR, 0.63; 95% CI, 0.47–0.86), and CRNMB (HR, 0.81; 95% CI, 0.70–0.94) versus LMWH at 6 months.17 The current study observed similar results in patients with VTE and active cancer who received extended anticoagulant treatment with apixaban beyond an initial 3 months.
Although multiple guidelines suggest indefinite anticoagulation treatment for patients with active cancer at risk for VTE, limited evidence exists to support this recommendation. The AMPLIFY-EXT trial evaluated the extended use of apixaban versus placebo for an additional 12 months in patients with VTE who had previously received anticoagulation for 6 to 12 months. The findings indicated that extended anticoagulation with apixaban reduced the risk of recurrent VTE without increasing the rate of MB at 12 months; however, only a small number of patients with active cancer (24 receiving apixaban and 18 receiving placebo) were included in this study.18 Other health care claims analyses evaluating extended anticoagulation for VTE using apixaban or warfarin also support the extended use of apixaban.19,20 The findings of the current analysis provide additional support for the guideline recommendations of indefinite anticoagulation treatment in patients with active cancer to manage thromboembolic risks. One strength of this analysis includes using LMWH as a comparison with apixaban for extended anticoagulation, because LMWH is more often used in patients with active cancer.21 Additionally, use of the Komodo Health Healthcare Map allowed for observation of a large number of patients to address the research question despite the overall small percentage of patients who receive extended anticoagulation.22 Evidence from claims analyses along with clinical trials may help inform the decision-making process for extended anticoagulation treatment among patients with active cancer at risk for VTE.
Although this analysis provides additional real-world data to support use of apixaban as extended VTE treatment in patients with active cancer, limitations inherent to health care claims analyses must be considered. First, retrospective observational studies can only demonstrate association and not causality. Second, the presence of a claim for a filled prescription does not necessarily indicate whether the medication was consumed or taken as prescribed. Third, the presence of a diagnosis code on a medical claim does not definitively confirm the presence of disease, because the diagnosis code may be incorrectly coded or included as rule-out criteria rather than actual disease. Similarly, the definition used to identify active cancer (ie, evidence of ≥2 medical claims for cancer diagnosis on different days or 1 claim for cancer diagnosis plus ≥1 claim for cancer treatment within 6 months before or 30 days after the index VTE event [first VTE diagnosis]) in the study may not account for all the patients being actively treated for cancer in the defined period. However, this definition is consistent with previously recommended algorithms for identifying chronic conditions in administrative claims databases and these algorithms have demonstrated a high degree of validity.23 In addition, certain information that is not readily available in claims data (eg, laboratory test results) could influence study outcomes. Although the Komodo Health Healthcare Map allows for longitudinal analysis, this study only evaluated extended VTE treatment after the first 3 months of initial therapy; future investigators should consider evaluating extended use after 6 months or longer periods of initial therapy. Likewise, the study did not evaluate or compare mortality as an outcome between the 2 treatment groups because mortality data were not available in the data source used. Results of this study may also not be generalizable to the entire US VTE population with active cancer because this study only evaluated commercially insured patients and did not consider other populations (eg, uninsured; self-insured; those with Veterans Affairs, Medicare, or Medicaid insurance). Finally, claims data may not fully capture factors at physician, system, and patient levels that may influence treatment choice and resource use.
Conclusions
In this analysis of adult patients with VTE and active cancer from a large US administrative claims database, extended use of apixaban following the first 3 months of therapy was associated with lower rates of MB and recurrent VTE compared with LMWH. Further research is needed to validate these findings in the clinical trial settings.
Acknowledgments
We would like to acknowledge Lori Justice, PharmD; Andrea Gundlach, PharmD, MPH, CMPP; and Kylie Matthews, MS, BS, of Xcenda LLC, a Cencora company, for their aid in medical writing and publication support, and Qisu Zhang, MPH, BS, of Xcenda LLC, a Cencora company, for her contributions to the data acquisition and analysis, which was funded by Bristol Myers Squibb Company.
References
- 2.↑
Becattini C, Di Nisio M, Franco L, et al. Treatment of venous thromboembolism in cancer patients: the dark side of the moon. Cancer Treat Rev 2021;96:102190.
- 3.↑
Horsted F, West J, Grainge MJ. Risk of venous thromboembolism in patients with cancer: a systematic review and meta-analysis. PLoS Med 2012;9:e1001275.
- 4.↑
Elting LS, Escalante CP, Cooksley C, et al. Outcomes and cost of deep venous thrombosis among patients with cancer. Arch Intern Med 2004;164:1653–1661.
- 5.↑
Prandoni P, Lensing AW, Piccioli A, et al. Recurrent venous thromboembolism and bleeding complications during anticoagulant treatment in patients with cancer and venous thrombosis. Blood 2002;100:3484–3488.
- 6.↑
Cohen AT, Katholing A, Rietbrock S, et al. Epidemiology of first and recurrent venous thromboembolism in patients with active cancer. A population-based cohort study. Thromb Haemost 2017;117:57–65.
- 7.↑
Kearon C, Akl EA, Ornelas J, et al. Antithrombotic therapy for VTE disease: CHEST guideline and expert panel report. Chest 2016;149:315–352.
- 8.↑
Raskob GE, van Es N, Verhamme P, et al. Edoxaban for the treatment of cancer-associated venous thromboembolism. N Engl J Med 2018;378:615–624.
- 9.↑
Young AM, Marshall A, Thirlwall J, et al. Comparison of an oral factor Xa inhibitor with low molecular weight heparin in patients with cancer with venous thromboembolism: results of a randomized trial (SELECT-D). J Clin Oncol 2018;36:2017–2023.
- 10.↑
McBane RD II, Wysokinski WE, Le-Rademacher JG, et al. Apixaban and dalteparin in active malignancy-associated venous thromboembolism: the ADAM VTE trial. J Thromb Haemost 2020;18:411–421.
- 11.↑
Agnelli G, Becattini C, Meyer G, et al. Apixaban for the treatment of venous thromboembolism associated with cancer. N Engl J Med 2020;382:1599–1607.
- 12.↑
Mokadem ME, Hassan A, Algaby AZ. Efficacy and safety of apixaban in patients with active malignancy and acute deep venous thrombosis. Vascular 2021;29:745–750.
- 13.↑
Key NS, Khorana AA, Kuderer NM, et al. Venous thromboembolism prophylaxis and treatment in patients with cancer: ASCO Guideline update. J Clin Oncol 2023;41:3063–3071.
- 14.↑
Lyman GH, Carrier M, Ay C, et al. American Society of Hematology 2021 guidelines for management of venous thromboembolism: prevention and treatment in patients with cancer. Blood Adv 2021;5:927–974.
- 15.↑
Lyon AR, López-Fernández T, Couch LS, et al. 2022 ESC guidelines on cardio-oncology developed in collaboration with the European Hematology Association (EHA), the European Society for Therapeutic Radiology and Oncology (ESTRO) and the International Cardio-Oncology Society (IC-OS). Eur Heart J 2022;43:4229–4361.
- 16.↑
Streiff MB, Holmstrom B, Ashrani A, et al. Clinical Practice Guidelines in Oncology: Cancer-Associated Venous Thromboembolic Disease. Version 2.2023. Accessed December 11, 2023. To view the most recent version, visit https://www.nccn.org
- 17.↑
Cohen A, Keshishian A, Lee T, et al. Effectiveness and safety of apixaban, low-molecular-weight heparin, and warfarin among venous thromboembolism patients with active cancer: a U.S. claims data analysis. Thromb Haemost 2021;121:383–395.
- 18.↑
Agnelli G, Buller HR, Cohen A, et al. Apixaban for extended treatment of venous thromboembolism. N Engl J Med 2013;368:699–708.
- 19.↑
Kang HR, Lo-Ciganic WH, DeRemer CE, et al. Effectiveness and safety of extended oral anticoagulant therapy in patients with venous thromboembolism: a retrospective cohort study. Clin Pharmacol Ther 2022;112:133–145.
- 20.↑
Pawar A, Gagne JJ, Gopalakrishnan C, et al. Association of type of oral anticoagulant dispensed with adverse clinical outcomes in patients extending anticoagulation therapy beyond 90 days after hospitalization for venous thromboembolism. JAMA 2022;327:1051–1060.
- 21.↑
Xiong W. Current status of treatment of cancer-associated venous thromboembolism. Thromb J 2021;19:21.
- 22.↑
Komodo Health. Healthcare map: the most complete view of the patient. Accessed June 12, 2023. Available at: https://www.komodohealth.com/technology/healthcare-map
- 23.↑
Rector TS, Wickstrom SL, Shah M, et al. Specificity and sensitivity of claims-based algorithms for identifying members of Medicare+Choice health plans that have chronic medical conditions. Health Serv Res 2004;39:1839–1857.
