Supported via arm's length funding from Bayer

Highlights from

European Society of Cardiology

Congress 2018

Munich 25-29 August 2018

COMPASS trial shows that GI or GU bleeds may signal cancer in patients with vascular disease treated with antithrombic therapy

Take-home messages

Among patients with vascular disease on long-term antithrombic therapy enrolled on the COMPASS trial:

  • More than 1 in 5 new diagnoses of cancer are preceded by bleeding
  • GI bleeding and GU bleeding are powerful predictors of new cancer diagnosis; more than 75% are diagnosed within 6 months of the bleed
  • The early increase in GI bleeding with rivaroxaban appears to unmask cancer at an earlier time point
“The COMPASS data not only highlights the importance of bleeding as a possible sign of underlying cancer, but also demonstrate that increased GI bleeding with rivaroxaban and aspirin leads to an earlier diagnosis of GI cancer.”

Dr John Eikelboom
McMaster University, Hamilton, Canada

More than one fifth of new cancer diagnoses, in particular new gastrointestinal (GI) and genitourinary (GU) cancers, are preceded by bleeding among patients with vascular disease on long-term antithrombotic therapy, as shown by the latest data analyses from the Cardiovascular OutcoMes for People Using Anticoagulation StrategieS (COMPASS) trial (NCT03022630).

Specifically, the analyses show that GI and GU bleeding are powerful predictors of new cancer diagnoses, and over 75% of cancers diagnosed after GI or GU bleeding are detected within 6 months of the bleeding event.

“The COMPASS data not only highlights the importance of bleeding as a possible sign of underlying cancer, but also demonstrate that increased GI bleeding with rivaroxaban and aspirin leads to an earlier diagnosis of GI cancer”, commented presenter and COMPASS Co-investigator Dr John Eikelboom of McMaster University, Hamilton, Canada.

The COMPASS trial evaluated 27,395 patients who were randomly assigned to a combination of rivaroxaban 2.5 mg BID and aspirin 100 mg OD, or rivaroxaban 5 mg BID alone, or aspirin 100 mg OD alone.

The primary efficacy outcome data (composite of cardiovascular death, stroke, or myocardial infarction) has previously been published and the study was stopped for superiority of the rivaroxaban-plus-aspirin group after a mean follow-up of 23 months. At ESC 2018, Dr Eikelboom presented the analysis of information collected on bleeding and new cancer diagnoses.

“Previously published results from COMPASS demonstrated that the increase in bleeding with rivaroxaban-based treatments compared with aspirin was mainly GI and occurred mainly in the first year after starting treatment”, noted Dr Eikelboom.

However, the data presented at ESC 2018 indicates that this early increase in GI bleeding is associated with an early increase diagnosis of GI cancer, which could benefit patients by improving cancer outcomes.

The proportions of new cancers diagnosed before and after bleeding were determined, as was the association between bleeding and new cancer diagnosis, and the rates of cancer diagnosis according to randomised treatment.

Of the 1,082 cancers found during the trial, 257 (23.8%) were diagnosed after bleeding. Of the 307 GI cancers, 70 (22.8%) were diagnosed after GI bleeding, and of the 138 GU cancers, 62 (44.9%) were diagnosed after GU bleeding.

Additional analyses showed a strong and specific association between GI bleeding and GI cancer, and between GU bleeding and GU cancer. “GI bleeding was associated with a 12-fold (hazard ratio [HR]:12.9, p<0.0001) increased hazard of being diagnosed with GI cancer, and GU bleeding with an 80-fold (HR: 83.4; p<0.0001) hazard of being diagnosed with GU cancer,” reported Dr Eikelboom.

Rates of cancer diagnosis were analysed according to timing and randomised treatment. “We found most of the GI bleeding was in the first year, and there was a clear excess of GI cancer diagnoses in the first year,” reported Dr Eikelboom. Specifically, 77.1% of GI cancers were diagnosed within the first six months after bleeding, and 88.7% of GU cancers were also diagnosed within this time frame.

“Effectively, both GI and GU bleeds are very powerful predictors of GI and GU cancer. If the patient has a non-GI bleed then it is a weak predictor of GI cancer so it is relatively specific,” he added.

“The message for clinicians is simple”, concluded Dr Eikelboom. “If a patient receiving antithrombotic drugs has GI or GU bleeding, then look for cancer in the same organ system. We don’t yet know if these patients have better outcomes but you’d hope that diagnosing early will improve the chances of a cure.”

Based on Eikelboom J W, et al. COMPASS trial - bleeding and cancer risk in patients with vascular disease treated with rivaroxaban (1321). Presented Sunday, 26 August 2018

Top image: © Mutlu Kurtbas

Article image: © yodiyim

The content and interpretation of these conference highlights are the views and comments of the speakers/authors.

Supported by