Cancer is not a single disease: is it safe to extrapolate evidence from trials of direct oral anticoagulants in cancer-associated venous thromboembolism to patients with haematological malignancies?