haematology and oncology Oral anticoagulants and cancer-associated thrombosis Randomised clinical trials comparing prolonged treatment with low molecular weight heparins and non-vitamin K oral anticaogulants in patients with cancer-associated thrombosis are discussed Guy Meyer MD Université Paris Descartes, Sorbonne Paris Cité; Assistance Publique Hôpitaux de Paris, Hôpital Européen Georges Pompidou, Paris, France Venous thromboembolism (VTE) is a frequent complication in patients with cancer. 1–3 About 8% of cancer patients will develop VTE during their disease. 4 VTE is associated with higher risks of bleeding and recurrence during anticoagulant treatment in patients with cancer than in other patients. 5 Low molecular weight heparins A series of trials have shown that prolonged treatment with low molecular weight heparin (LMWH) given for at least three to six months is associated with a reduction in the risk of recurrent VTE as compared with the traditional treatment regimen with LMWH overlapped and followed by vitamin K antagonists (VKA) in patients with cancer-associated thrombosis (CAT). 6 Even with LMWH, about 7% of patients with CAT experience recurrent VTE during the first six months of treatment. 7 The prolonged use of LMWH is associated with the need of daily subcutaneous injections, bruising at the injection site and a higher cost than VKA. In clinical practice, a significant proportion of patients with CAT continue receiving VKA. 8 Most of the patients with CAT have a significant risk of recurrent VTE after the initial six-month treatment and anticoagulation needs to be prolonged for more than six months in a significant number of these patients making the use of subcutaneous treatment even more difficult. 9 NOACs Non-vitamin K oral anticoagulants (NOACs) are direct inhibitors of factor IIa or factor Xa. These drugs have a predictable anticoagulant effect and do not need monitoring; they have an early onset of action with peak plasma concentrations obtained about 3–4 hours after ingestion. Rivaroxaban and apixaban can be administered in the acute phase of VTE without any preceding parenteral anticoagulation, whereas dabigatran and edoxaban have been tested after at least five days of treatment with LMWH. 10–13 In a series of 17 HHE 2018 | hospitalhealthcare.com six randomised controlled trials on approximately 26,000 patients with VTE, NOACs were effective (relative risk (RR) 0.90; 95% CI 0.77–1.06 for first recurrent VTE or VTE-related death) and safer than VKA (RR 0.61; 95% CI 0.45–0.83 for major bleeding) in patients with VTE. 14 A minority of patients included in these trials had underlying cancer. Meta-analyses of this subgroup of patients with CAT show that DOACs are associated with a non-significant reduction in the risk of recurrent VTE (RR 0.65, 95% CI 0.38–1.09) and bleeding (RR, 0.67, 95% CI 0.31–1.46) as compared with VKA. 6 Of note, cancer patients in these trials had less advanced cancer, a smaller proportion received anticancer treatment and the mortality was lower than in the trials comparing LMWH and VKA in patients with CAT. The reference treatment (VKA) in these trials was not the most appropriate for patients with CAT. Several observational cohort studies have been published describing initial experiences with NOACs in patients with CAT and have been summarised in a systematic review. 15 Most studies used rivaroxaban and enoxaparin. The on-treatment duration of NOACs was longer than that of LMWH. All studies except one reported lower rates of recurrent VTE for patients using NOAC as compared with those on LMWH, but these results may be attributable to patient selection. Because of the observational design, patients were not randomised to receive one of the two treatments and the treatment groups are not comparable. In most studies, the outcomes were not assessed independently. The bleeding outcomes were heterogeneous across different studies with two studies that only included gastrointestinal and gynaecological cancers reporting higher rates of major bleeds in patients receiving a NOAC. 16,17 Clinical trials Two randomised clinical trials comparing prolonged treatment with LMWH and a NOAC in patients with CAT have been reported.