The  Scottish Medicines Consortium issued on 7th November, 2008 an advice on rivaroxaban (Xarelto):

ADVICE: following a full submissionrivaroxaban (Xarelto®) is accepted for use  within NHS Scotland for the prevention of venous thromboembolism (VTE) in adult patients undergoing elective hip or knee replacement surgery.In three large phase III studies in patients undergoing either total knee or total hip replacement surgery, rivaroxaban was superior to low molecular weight heparin in reducing the incidence of VTE and all cause mortality with patients while having a similar incidence of major bleeding events.

This is an excerpt of the detailed advice on rivaroxaban (Xarelto®)

For the prevention of venous thromboembolism (VTE) in adult patients undergoing elective hip or knee replacement surgery.

Dosing information
Rivaroxaban 10mg orally once daily, starting 6  to 10 hours after surgery, provided that haemostasis has been established.  For patients undergoing major hip surgery, a treatment duration of five weeks is recommended.  For patients undergoing major knee surgery, a treatment duration of two weeks is recommended.

Product availability date
01 October 2008

Summary of evidence on comparative efficacy

Rivaroxaban is a highly selective direct factor Xa inhibitor which interrupts the intrinsic and extrinsic pathway of the blood coagulation cascade inhibiting thrombin formation and the development of thrombi.

Three large phase III randomised, double-blind, double-dummy studies have compared rivaroxaban with enoxaparin for the prevention of VTE after hip or knee replacement surgery.

In each study, patients were randomised to receive either rivaroxaban (10mg orally once daily, starting 6 to 8 hours after wound closure) plus placebo injections or enoxaparin (40mg subcutaneously (SC) starting 12 hours pre-operatively and restarted 6 to 8 hours after wound closure) plus placebo tablets. The duration of active study medication varied among studies: in the first study in patients undergoing total hip replacement (THR), both rivaroxaban and enoxaparin were continued for 35±4 days; in the second study, again in patients undergoing THR, rivaroxaban was continued for 35±4 days and enoxaparin for 12±2 days; and in the third study, in patients undergoing total knee replacement (TKR), both rivaroxaban and enoxaparin were continued for 12±2 days. All patients underwent mandatory bilateral venography on the day after the last dose of study medication. The primary endpoint in all studies was the composite of any deep vein thrombosis (DVT) (symptomatic or asymptomatic detected venography), non-fatal pulmonary embolism (PE), or death from any cause.  The major secondary endpoint was major VTE defined as the composite of proximal DVT, non-fatal PE, or VTE-related death.  All three studies were designed to test for non-
inferiority which if demonstrated in the per-protocol population, resulted in a test for superiority being performed in the modified intention to treat (mITT) population (patients who had undergone surgery, taken a study drug, and undergone adequate assessment for VTE).

Rivaroxaban was superior to enoxaparin in all studies as assessed by the primary and major secondary endpoint. The key results are presented in the table below.
Summary of evidence on comparative safety

The main safety outcome, in all three studies, was the incidence of major bleeding defined as bleeding that was fatal, occurred in a critical organ (e.g. retroperitoneal, intracranial, intraocular, and intraspinal bleeding) or required re-operation or extrasurgical-site bleeding that was clinically overt and was associated with a fall in haemoglobin of  ≥ 2g/dl or that required transfusion of  ≥ 2 units of whole blood or packed cells. The incidence of major
bleeding across all studies was low (<0.1% to 0.6%) and did not differ significantly between rivaroxaban and exoxaparin. The incidences of on-treatment bleeding (4.8% to 6.6%) and non-major bleeding (4.3% to 6.5%) were also similar between the two treatments.  No differences were reported between rivaroxaban and enoxaparin in the liver function test results.  In the second study (described above), there was an excess of cardiovascular events after discontinuation of rivaroxaban; five patients (0.4%) including two cardiovascular deaths (although one occurred in a patient who had received placebo injection only) and no
cardiovascular events after discontinuing enoxaparin.

Summary of clinical effectiveness issues

All three studies described above have demonstrated superior efficacy in terms of the primary endpoint with rivaroxaban compared to enoxaparin.  This difference was mainly driven by a reduced rate of DVT (both symptomatic and asymptomatic) in the rivaroxaban group. In the second study described above, the relative efficacy of rivaroxaban and enoxaparin is less clear due to the unequal treatment durations used (35±4 days for rivaroxaban and 12±2 days for enoxaparin).

Improved efficacy was achieved with no increase in the risk of major bleeding which was relatively low in both treatment groups compared to other studies.  However the definition of major bleeding excluded patients with surgical site bleeding unless it led to re-operation or death.  Non-major bleeding, which included surgical site bleeding, was similar in both groups. Factors such as age >75 years, extremes of weight, renal insufficiency and previous DVT are
known to increase both the risk of VTE and bleeding.  The number of patients included in the three studies with these risk factors was limited.  The studies also excluded patients with a high risk of bleeding.  This patient population is reflective of most clinical studies in this therapeutic area but it may not truly represent those treated in clinical practice.

In one of the studies, there was an excess of cardiovascular events after discontinuing rivaroxaban.  The authors note that this could have been due to chance and that no trend was evident across all three studies; the numbers are too low to draw any firm conclusions.
However this does raise concerns about rebound activation of coagulation and further research is warranted.

A relatively high number of patients were excluded from the efficacy analysis (6,588/9,581 (69%) randomised patients were available for mITT analysis) and this was mainly due to inadequate assessment of thromboembolism.  All three studies did pre-empt that there would be non-evaluable patients and included a non-evaluable rate of 25% in the sample size calculation to account for this.  However, the actual non-evaluable rate in all three studies was slightly higher than 25% (26% to 30%).  Two of the studies recruited extra patients to compensate and maintain statistical power.  In one study (the second described above),
there was no increase in recruitment, which may have potentially limited the results. In all studies, sensitivity analyses were performed to confirm that missing data did not affect the power of the study or bias the outcome.

Like dabigatran, rivaroxaban offers the advantage of an oral preparation that does not require regular monitoring and dose adjustment or daily subcutaneous injections.  When thromboprophylaxis is continued for up to five weeks this may be an advantage in the community setting. There are no direct comparative data between dabigatran and rivaroxaban so their relative efficacy and safety are unknown.

Cost of relevant comparators

Source: Scottish Medicines Consortium Drug Advice: rivaroxaban (Xarelto®). Full text PDF