A dose-finding study with the oral platelet thrombin receptor blocker vorapaxar on top of dual antiplatelet therapy in elective PCI found a reduction of ischemic complications without an increase in bleeding when compared to aspirin and clopidogrel alone.1
In the TRACER trial 12,944 patients with acute coronary syndrome (ACS) without ST-segment elevation were randomized to vorapaxar 40 mg loading dose then 2.5 mg daily or placebo on top of standard of care, usually including dual antiplatelet therapy, and were followed for 13 months.2 The primary endpoint (cardiovascular death, MI, stroke, recurrent ischemia with hospitalization or urgent revascularization) was reduced by vorapaxar from 19.9% with placebo to 18.5%, but this failed to reach statistical significance (HR, 0.92; 95% CI 0.85 - 1.01 p = 0.07). The secondary endpoint of cardiovascular death, MI and stroke was significantly reduced by vorapaxar (in most cases meaning triple oral antiplatelet therapy) when compared to placebo (with most taking dual antiplatelet therapy): 14.7% versus 16.4% (HR 0.89, 95% CI 0.81 - 0.98, p = 0.02, fig.1).
Fig 1. Ischemic endpoints in TRACER
Surprisingly, stent thrombosis was not reduced by vorapaxar: 1.7% for vorapaxar and 1.5% for placebo (HR 1.12, 95% CI 0.78 - 1.62, p = 0.54). Moreover, there was a substantial increase of GUSTO moderate or severe bleeding with vorapaxar: 7.2% in the vorapaxar group vs 5.2% in the placebo group (HR 1.35; 95% CI 1.16 - 1.58. p < 0.001, fig. 2). Disturbing was the increase in intracranial bleeding: 1.1% and 0.2% (HR 3.39, 95% CI 1.78 - 6.45, p < 0.001, fig. 2). The rates of total stroke, fatal stroke or fatal bleeding were not significantly increased by vorapaxar.
Thus, triple antiplatelet therapy with vorapaxar, clopidogrel and aspirin in the acute phase of ACS and thereafter produces a moderate reduction in recurrent ischemic events when compared to clopidogrel and aspirin alone, with a considerable increase in moderate to severe bleeding and an unacceptable tripling of intracranial bleeding. Whether this form of novel antiplatelet therapy is efficacious and acceptably safe, relative to standard of care (mostly single antiplatelet treatment) for secondary prevention in chronic atherothrombotic disease is being evaluated in the giant TRA2P TIMI 50 study. The study arm with patients with a recent stroke or TIA had to be discontinued due to excess intracranial bleeding with vorapaxar.
The oral direct Xa blocker rivaroxaban has now been approved for both DVT prophylaxis after orthopedic surgery and for stroke prevention in atrial fibrillation. In ACS it has been tested in a phase II dose finding study, where the 5mg and 10mg daily doses were found to be acceptably safe and effective on top of single or dual antiplatelet therapy3 and were picked out for phase III evaluation in ACS. These doses are a fourth and a half, respectively, of those used for stroke prevention in atrial fibrillation in the ROCKET-AF trial4.
In ATLAS-2 TIMI 51 a total of 15,526 patients with a recent ACS were randomized double-blind to 2.5 mg or 5 mg rivaroxaban bid or placebo on top of dual antiplatelet therapy5. Follow-up was for a mean of 13 months. The primary efficacy end point (cardiovascular death, MI and stroke) was reduced from 10.7% with placebo to 8.9% with rivaroxaban (HR 0.84, 95% CI, 0.74 - 0.96, p = 0.008, fig.3) with significant benefit for both the 2.5 mg bid dose (9.1% vs. 10.7%, p = 0.02) and 5 mg bid dose (8.8% vs. 10.7%, p = 0.03). Rivaroxaban 2.5 mg bid reduced cardiovascular (2.7% vs.4.1%, p = 0.002) and total death (2.9% vs. 4.5%, p = 0.002), which was not observed with the 5 mg bid dose.
Fig 3. Efficacy in ATLAS-2 TIMI 51
Remarkably, both rivaroxaban doses reduced stent thrombosis by 31% from 2.9% to 2.3% (p = 0.008) in comparison to antiplatelet therapy alone (fig. 4).
Fig 4. Stent thrombosis in ATLAS-2 TIMI 51
Rivaroxaban increased the rates of non-CABG TIMI major bleeding (2.1% vs. 0.6%, p < 0.001) and intracranial hemorrhage (0.6% vs. 0.2%, p = 0.009), but without a significant increase in fatal bleeding (0.3% vs. 0.2%, p = 0.66, fig. 5). The 2.5 mg bid dose led to fewer fatal bleeding events than 5 mg bid (0.1% vs. 0.4%, p = 0.04).
Fig 5. Bleeding in ATLAS-2 TIMI 51
Thus, low-dose novel oral anticoagulation with a Xa inhibitor on top of dual antiplatelet reduces ischemic complications after ACS, but at a considerably increased risk of major, but not fatal, bleeding. This is in contrast with the results of a recent large and similar trial with the Xa blocker apixaban (APPRAISE-2) using the same dose as tested in atrial fibrillation against warfarin6. In ATLAS-2 a much lower dose was applied than in the successful ROCKET-AF trial in atrial fibrillation4. Apparently, proper dosing is essential to optimize efficacy at an acceptable bleeding risk, when anticoagulation is applied not only in atrial fibrillation, but also in ACS and thereafter. Examples are warfarin, enoxaparin, fondaparinux7 and recently the novel oral Xa blocker darexaban (see below). Finally, in ATLAS-2, stent thrombosis was reduced with oral anticoagulation. This was also seen with apixaban in APPRAISE-2, but not with triple antiplatelet therapy in TRACER (see above), suggesting that the coagulation cascade plays an important role in the pathogenesis of stent thrombosis.
Besides apixaban and rivaroxaban another oral direct Xa blocker darexaban has been tested in secondary prevention of ACS.
In the RUBY-1 dose finding trial 1,279 patients with a recent ACS were randomized double-blind to one of six darexaban regimens: 5 mg bid, 10 mg qd, 15 mg bid, 30 mg qd, 30 mg bid, 60 mg qd or placebo on top of dual antiplatelet therapy87. Follow-up was for a mean of 22 weeks. Compared with placebo, there was a dose-related increase (p < 0.009) in the rate of major or clinically relevant bleeding (primary endpoint: 6.2, 6.5, and 9.3% for 10, 30, and 60 mg daily, respectively, fig. 6), which was statistically significant for 30 mg bid (p < 0.002).
Fig 6. Bleeding in RUBY-1
Although the study was not powered for efficacy, there seems to be a surprising direct dose relation with ischemic endpoints (fig. 7).
Fig 7. Ischemic endpoints in RUBY-1
In conclusion, like the other oral Xa blockers, darexaban increases bleeding when given together with dual antiplatelet therapy. Only the lowest dose had acceptable safety and carries some promise for ischemic event reduction. Whether the drug will be evaluated in further studies is uncertain.
Antidotes against the novel oral anticoagulants seem needed, but none is available at the momenthas emanated valuable .
In an elegant placebo-controlled crossover study 12 volunteers were given prothrombin complex concentrate (PCC or CofactR) after therapy with either dabigatran (150 mg bid) or rivaroxaban (20 mg bid) for 2.5 days98. Rivaroxaban induced a significant prolongation of the prothrombin time (PT) when compared with baseline, which was immediately and completely reversed by PCC. The endogenous thrombin potential (ETP) was inhibited by rivaroxaban and also normalized with PCC, whereas placebo had no effect (fig. 8).
Fig. 8. Reversal of rivaroxaban by PCC or placebo
Thus, theoretically, PCC may be an effective antidote for rivaroxaban, but not for dabigatran. Whether this works clinically in patients, let alone in patients with active bleeding, remains to be established.
Finally, interruption of antiplatelet therapy is deemed necessary when major surgery is planned. Both aspirin and clopidogrel are long acting agents. A; after their cessation for coronary surgery they may be ‘bridged’ by an antiplatelet agent with a fast onset and offset of action.
In the BRIDGE trial 210 patients with an ACS or treated with a coronary stent and receiving a thienopyridine awaiting CABG were randomized to cangrelor (an intravenous P2Y12 receptor antagonist) or placebo after thienopyridines were stopped109. Study drug was given for at least 48 hours and was discontinued 1 to 6 hours before CABG. The primary efficacy endpoint was platelet reactivity (measured in P2Y12 reaction units [PRUs]) assessed daily. Patients treated with cangrelor had low levels of platelet reactivity throughout the entire treatment period compared with placebo, and platelet function was restored immediately after infusion of study drug was terminated (fig. 10)
Fig 10. Platelet reactivity with cangrelor or placebo after thienopyridine cessation before CABG
European Society of Cardiology
European Heart HouseLes Templiers2035 Route des CollesCS 80179 Biot
06903Sophia Antipolis, FR
Our mission: To reduce the burden of cardiovascular disease
© 2017 European Society of Cardiology. All rights reserved