From Wiki Journal Club
Jump to: navigation, search
Khorana AA, et al. "Rivaroxaban for Thromboprophylaxis in High-Risk Ambulatory Patients with Cancer". The New England Journal of Medicine. 2019. 380(8):720-8.
PubMedFull textPDF

Clinical Question

In ambulatory patients with cancer at high risk for thromboembolism (Khorana score ≥2), is low-dose rivaroxaban more effective than placebo in reducing incidence of venous thromboembolic disease?

Bottom Line

In ambulatory patients with cancer at high risk for thromboembolism (Khorana score ≥2), low-dose rivaroxaban did not result in a statistically significant reduction in incident thromboembolism at 180 days when compared to placebo (6.0% with rivaroxaban versus 8.0% with placebo). In a pre-specified intervention-period analysis (only time on drug considered), however, there was a 3.6% absolute reduction in thromboembolism with rivaroxaban, which should be considered hypothesis generating. There was a marginal 1.0% absolute increase in major bleeding with rivaroxaban.

Major Points

Patients with active cancer have an increased risk of venous thromboembolism (VTE), but whether there are patients who benefit from anticoagulation to prevent VTE (thromboprophylaxis) is not clear. A few small trials have suggested benefit with use of low-molecular-weight heparin (LMWH) products in this setting, but event rates were low and the absolute benefit was modest.[1][2] Furthermore, LMWH products are injectable rather than oral, which has limited widespread use for this indication. More recently, risk stratification for incident VTE through the use of validated scores such as the Khorana score (scaled from 0 to 6, with higher scores indicating a higher risk of venous thromboembolism) now facilitate identification of patients at higher risk for developing VTE, and therefore more likely to benefit from thromboprophylaxis. [3] Furthermore, direct-acting oral anticoagulants (DOACs) have recently shown efficacy in VTE treatment even in patients with active malignancy (see Hokusai VTE Cancer Trial for edoxaban and SELECT-D for rivaroxaban). Whether DOACs may facilitate a more convenient and efficacious strategy for VTE prevention in cancer patients at particularly high risk of VTE remains unclear.

The 2019 Rivaroxaban for Preventing Venous Thromboembolism in High-Risk Ambulatory Patients with Cancer (CASSINI) trial randomized 841 cancer patients at high risk for VTE (Khorana score ≥2) to low-dose rivaroxaban (10MG QD) versus placebo. At 180 days, rivaroxaban was associated with a non-significant 2.8% absolute reduction in VTE. Of note, 47% of enrolled patients prematurely discontinued the trial drug (for a variety of reasons). In a pre-specified intervention-period analysis (only time on drug considered), there was a 3.8% absolute reduction in VTE with rivaroxaban that met statistical significance. Although prespecified, this analysis is subject to bias and is best considered hypothesis generating. The rates of major bleeding were low in both groups (2.0% with rivaroxaban and 1.0% with placebo), with no statistically significant difference.

Particularly when considered in the context of the contemporaneous AVERT trial demonstrating a reduction in VTE with apixaban, the results of CASSINI do provide some corroborative evidence of a modest VTE reduction with DOAC in patients with active malignancy. However, the observed benefit is small, and was not apparent in the primary analysis (likely due to high rates of premature discontinuation of study drug). Furthermore, there was a numerical increase in major bleeding with rivaroxaban use although the absolute rate was modest. On balance, routine use of DOAC to prevent VTE is unlikely to become routine on the strength of these data, although it may be reasonable in patients at highest risk for VTE and lowest risk of bleeding, for whom oral anticoagulation is most likely to be tolerated and continued.


As of March 2019, no guidelines have been published that reflect the results of this trial.


  • Multicenter, randomized, parallel-group, phase 3b trial
  • N=841
    • Rivaroxaban 10MG daily (N=420)
    • Placebo (N=421)
  • Setting: 143 centers in 11 countries
  • Duration of follow-up: 180 days
  • Analysis: Intention-to-treat
  • Primary Outcome: Venous thromboembolism


Inclusion Criteria

  • Age ≥18
  • Ambulatory outpatient with solid tumor or lymphoma
  • Khorana score ≥2
  • Expected survival >6 months with plan to start a new systemic regimen within 1 week before or after initiating the trial regimen

Exclusion Criteria

  • Primary brain tumor or known brain metastasis
  • Eastern Cooperative Oncology Group (ECOG) performance status score ≥3
  • Active bleeding or at risk for bleeding

Baseline Characteristics

From the placebo group.

  • Demographics: age 62 years, male 48.9%, white 82.2%
  • Khorana score: <2 0.7%, 2 70.1%, 3 22.8%, 4 5.9%, ≥5 0.5%
  • Previous VTE: DVT 0.5%, PE 0%
  • Malignancy: pancreatic 32.8%, breast 2.1%, gastric 20.7%, genitourinary 4.0%, lung 17.1%, lymphoma 6.2%, ovarian 7.1%, other 10.0%


  • Patients randomized 1:1 to rivaroxaban 10MG daily versus placebo
    • Randomization stratified according to pancreatic/non-pancreatic tumor site
  • Enrolled patients underwent venous duplex compression ultrasonography of both legs to rule out preexisting proximal DVT
  • Randomized patients followed for 180 days
  • Trial visits occurred at week 8, week 16, and at day 180 or the end of trial intervention
  • Each trial visit included screening compression ultrasonography of both legs
  • All primary outcome events adjudicated by an independent clinical end-point committee whose members were unaware of trial-group assignments
  • Pre-specified subgroup analyses were planned for subgroups according to the stratification factor (pancreas as primary tumor site, age, sex, race or ethnic group, body-mass index, primary tumor type, stage, Khorana score, performance-status score, geographic region, creatinine clearance)


Comparisons are rivaroxaban versus placebo

Primary Outcomes

Venous thromboembolism (180 days)
25 (6.0%) vs. 37 (8.8%); HR 0.66 (95% CI 0.40-1.09); p = 0.10

Secondary Outcomes

Venous thromboembolism (180 days, intervention period only)
11 (2.6%) vs. 27 (6.4%); HR 0.40 (95% CI 0.20-0.80)
84 (20.0%) vs. 100 (23.8%); HR 0.83 (95% CI 0.62-1.11)
Venous thromboembolism or death
97 (23.1%) vs. 124 (29.5%); HR 0.75 (95% CI 0.57-0.97)
Arterial thromboembolism
4 (0.95%) vs. 7 (1.7%); HR 0.58 (95% CI 0.17-1.98)

Safety Outcomes

Major bleeding
8 (2.0%) vs. 4 (1.0%); HR 1.96 (95% CI 0.59-6.49)

Subgroup Analyses

  • No heterogeneity of effect on the primary outcome was detected for any prespecified subgroup (pancreas as primary tumor site, age, sex, race or ethnic group, body-mass index, primary tumor type, stage, Khorana score, performance-status score, geographic region, creatinine clearance)


  • A VTE benefit with rivaroxaban was only seen in a pre-specified intervention-period analysis (only time on drug considered). This analysis is subject to bias (e.g., patients tolerant of anticoagulation contribute disproportionate person-time), and therefore should be considered hypothesis-generating.
  • The rate of premature study drug discontinuation was 47%, biasing towards null effect of rivaroxaban (bias towards null related to study drug discontinuation also supported by positive intervention-period analysis).
  • Secondary endpoints not corrected for multiple testing.


  • Study cosponsored by Janssen and Bayer
  • Sponsors involved in data collection and analysis

Further Reading

  1. Agnelli G et al. Nadroparin for the prevention of thromboembolic events in ambulatory patients with metastatic or locally advanced solid cancer receiving chemotherapy: a randomised, placebo-controlled, double-blind study. Lancet Oncol. 2009. 10:943-9.
  2. Agnelli G et al. Semuloparin for thromboprophylaxis in patients receiving chemotherapy for cancer. N. Engl. J. Med. 2012. 366:601-9.
  3. Khorana AA et al. Development and validation of a predictive model for chemotherapy-associated thrombosis. Blood 2008. 111:4902-7.