RESUMEN
BACKGROUND: Incidence of and risk factors for bleeding in cancer patients with venous thromboembolism (VTE) treated with apixaban are poorly described. METHODS: We analyzed data from the prospective CAP study where 298 cancer patients with any type of VTE received 5 mg apixaban twice daily for 6 months, and then 2.5 mg apixaban twice daily for 30 months. For most analyses, major bleedings and clinically relevant nonmajor bleedings were merged to "clinically relevant bleedings." Risk factors were estimated by odds ratios (OR) and 95% confidence intervals (CIs). RESULTS: The incidence of clinically relevant bleedings was 38% per person-year during the first 6 months of treatment, 21% per person-year from 7 to 12 months, and between 4 and 8% per person-year from 13 to 36 months. Clinically relevant bleedings were associated with age above 74 years (OR: 2.0, 95% CI: 1.0-4.1), body mass index (BMI) below 21.7 (OR: 2.3, 95% CI: 1.1-4.8), and hemoglobin at baseline below 10.5 for females (OR: 2.8, 95% CI: 1.1-7.3) and 11.1 for males (OR: 3.3, 95% CI: 1.3-8.4) during the first 6 months. Gastrointestinal (GI) or urogenital cancer was not associated with clinically relevant bleedings compared with other cancers. Among patients with luminal GI cancer, nonresected cancer had increased risk of bleeding (OR: 3.4, 95% CI: 1.0-11.6) compared with resected GI cancer. CONCLUSION: There were very few bleedings while patients were on low-dose apixaban. Factors associated with bleeding in patients treated with full-dose apixaban were high age, low BMI, and low hemoglobin, and probably nonresected luminal GI cancer.