Efficacy and Safety of Rivaroxaban in Patients with CAT: A Pooled Meta-Analysis of 3 Randomized Controlled Trials

Introduction: CALLISTO, a comprehensive programme of research on cancer-associated thrombosis (CAT) , included 3 randomized trials of rivaroxaban versus low molecular weight heparin (LMWH) for the treatment of venous thrombosis in patients with solid and haematological cancers (SELECT-D, CASTA-DIVA and CONKO-11). A meta-analysis of these studies was conducted to improve the precision of current estimates of the efficacy and safety of rivaroxaban in this patient group and investigate how patient characteristics impact the treatment effects.

Methods: The primary endpoint was the cumulative incidence of venous thromboembolism (VTE) recurrence at the end of the treatment period (≥3 months). Other endpoints included major bleeding (MB), a composite of MB or clinically relevant non-major bleeding (clinically relevant bleeding [CRB]) and deaths from any cause. All endpoints were assessed by 8 pre-defined subgroup analyses: age, gender, creatinine clearance, type of index VTE, index VTE localization, cancer localization, performance status and presence of metastases (Prospero submission 266227). Patient-level data were used in this analysis.

The cumulative incidences of VTE recurrence, CRB were estimated using the Kalbfleisch and Prentice model, while the Kaplan-Meier model was used to estimate the incidence of death. Comparisons between rivaroxaban and LMWH for VTE recurrence, MB and CRB were assessed by sub-distribution hazard ratios (SubHR) and 95% confidence intervals (CI), whereas hazard ratios and 95% CIs were used for the all-cause death endpoint. The pooled treatment effect size of each study was estimated using fixed-effect and random-effects models.

Results: When considering the prevention of VTE recurrence in the 3 randomized trials (N=804), an overall reduction of 48% was observed with rivaroxaban compared with LMWH (SubHR = 0.52, 95% CI 0.28─0.98). The estimation appeared to be homogeneous across subgroups of patients. In comparison with LMWH, rivaroxaban was associated with an increased risk of CRB (SubHR = 2.03, 95% CI 1.34─3.09), without significant difference in MB (SubHR = 1.24, 95% CI 0.60─2.57), and no difference was observed for death.

Conclusions: This pooled analysis suggests that rivaroxaban may be an alternative treatment option for the prevention of VTE recurrence in cancer patients with VTE. The gain in statistical power has shown significant benefit, as well as some risk associated with rivaroxaban treatment in this complex patient population. The impact of patient characteristics on these treatment effects will be presented at the meeting.