THE GREAT DEBATE: Thrombocardiology post-COMPASS
Frans van de Werf (Leuven, Belgium) and Deepak Bhatt (Newton, MA, USA)
Before the topic was introduced, two statements were voted on to take a snapshot of the baseline position of the audience. The first was ‘for secondary prevention in high-risk stable coronary artery disease (CAD), a low dose of a novel oral anticoagulant (NOAC) should be added to aspirin’, to which 52% replied yes, and 48% said no. The second was ‘for secondary prevention in high-risk stable CAD, a P2Y12 antagonist should be added to aspirin’, which was a little more divided with 59% saying yes and 41% saying no.
Positioning of the topic - The concept of thrombocardiology
Eugene BRAUNWALD (Boston, MA, USA)
Professor Braunwald likes to think of thrombin as the villain and centerpiece of clotting. It has three major roles: it causes platelet aggregation, platelet adhesion, and it catalyzes the conversion of fibrinogen to fibrin. In the long history of studies into atherothrombotic therapy in CAD, the first results were obtained with aspirin alone: it has moderate efficacy in acute CAD and it causes a minor degree of bleeding.
For many years, people have wanted to improve results as compared with what could be obtained on aspirin, thus an anticoagulant was added. The combination of aspirin with warfarin was found to be more efficacious post-ACS (acute coronary syndrome), but also to cause more bleeding. The lesson was that combination therapy was more powerful than aspirin alone. The CURE data next showed similar efficacy post-ACS, and a bit less bleeding with dual antiplatelet therapy (DAPT). The ATLAS ACS 2-TIMI 51 trial then showed a 34% reduction in CV death when 2.5 mg rivaroxaban twice daily was added to aspirin and clopidogrel. A re-analysis focusing on fatal or irreversible bleedings, revealed that this therapy with an ultra-low dose non-vitamin K antagonist OAC resulted in a net reduction of these type of events: 95 non-bleeding CV deaths were prevented with rivaroxaban 2.5 mg vs placebo, at the cost of 10 extra fatal bleedings. It is well known that in the context of anticoagulants, there is no such thing as a free lunch, but here costs were low, said Braunwald.
Then came the COMPASS trial, which evaluated the same low-dose rivaroxaban with or without aspirin in over 27,000 stable CVD patients. Rivaroxaban 2.5 mg plus aspirin 100 mg showed greatest efficacy and only modest bleeding, as compared to aspirin alone. Subsequently, the applicability of the COMPASS regime was evaluated in the large international REACH registry, which indeed suggested good external applicability, as COMPASS-eligible patients represented a substantial fraction of stable CAD/PAD patients encountered in clinical practice.
It was then up to the debaters to convince the audience of their positions on whether or not anticoagulation should replace antiplatelets in CAD prevention.
Anticoagulation should replace antiplatelets in CAD prevention – PRO
Felicita Andreotti (Rome, Italy)
Dr. Andreotti was asked to defend the PRO-position, which she did after slightly rephrasing the statement, to ‘Should anticoagulation replace at least one antiplatelet to prevent major adverse CV events (MACE)?’. She started her argumentation by noting that thrombin is the most powerful platelet agonist, that all thrombi contain fibrin or fibrinogen, with or without platelets. Fissured-plaque thrombosis is initiated by tissue factor-coagulation and finally, she stated that anticoagulation-fibrinolysis is a natural vasoprotective system. Thus, she set out to argue that fibrin and anticoagulation may play a more important role than platelets.
Enhanced coagulation and impaired fibrinolysis are integral parts of metabolic syndrome, inflammation and smoking. Hemostatic predictors of MACE are mainly coagulation-, rather than platelet-related. Indeed, prolonged endogenous fibrinolysis has been shown to predict CV death in a PLATO substudy, in which ex vivo clot lysis time was measured. The longer it took, the worse was prognosis. Genetic studies have also suggested that gene variants associated with coagulation, more so than those associated with platelets, predispose to arterial thromboembolism. Andreotti was involved in a study in which the risk associated with a prothrombin gene variant (the G20210A allele) was evaluated in presence or absence of metabolic or acquired risk factors. In absence of other risk factors, which reflects the future since we are getting better at lowering for instance cholesterol and blood pressure, the risk attributed to this allele was substantial, as compared with the situation in which risk factors are present.
She showed some data that support the hypothesis that anticoagulation may be more relevant than antiplatelet therapy. In the Dutch 60+ trial, treatment with OAC was strikingly superior to placebo with regard to survival after myocardial infarction (MI) (92.4% vs. 86.6% after 700 days of follow-up). Another study evaluating DAPT in comparison with OAC plus single antiplatelet (SAP) agent, demonstrated benefit of DAPT only early (<30 days) after percutaneous coronary intervention (PCI). These data were published already 20 years ago.
All this evidence available in the pre-COMPASS era has been considered for the 2018 ESC/EACTS Guidelines on myocardial revascularization, which were published during this year’s ESC congress. They recommend SAP long-term, and DAPT for 6 months after stable PCI and for 12 months after ACS, both with 1A recommendations. Additional recommendations advice to consider DAPT longer than 12 months in those with low bleeding risk (IIb A), and for 1-3 months in stable patients with a high risk of bleeding (IIb C – IIa A). Thus, the strongest recommendation is long-term SAP and shorter-term DAPT.
The COMPASS results now change this situation. Andreotti raised the question ‘Should anticoagulation plus SAP replace DAPT to prevent MACE in stable CAD or post-MI?’. The answer to the question as stated here, is no, as there is no data. It should be noted that in COMPASS, some patients were even asymptomatic. PEGASUS-TIMI 54, which evaluated ticagrelor 60 mg bid plus aspirin vs aspirin alone, included a higher risk population of patients after acute MI. Here, the absolute reduction of the primary endpoint with addition of ticagrelor to aspirin was low. COMPASS even achieved a reduction in overall mortality, but at the cost of a greater increase in bleedings. No head-to-head comparison is available, but the results are in favor of strategies involving anticoagulation.
The next question Andreotti raised was ‘should low-dose anticoagulation replace or be added to an antiplatelet agent in high-risk individuals or patients?' In this context it is interesting to look at the not so well-known Thrombosis Prevention Trial. This trial showed a striking effect of anticoagulation with warfarin on death, in the primary prevention of high-risk men. This raises the question whether the COMPASS-strategy should also be considered for primary prevention.
To answer this question, important temporal trends should be taken into account. Firstly, control of traditional risk factors has improved over time, which may allow an innate hypercoagulability state with impaired fibrinolysis to emerge increasingly. As outlined in the beginning of her presentation, this is a risk factor for MACE. Additionally, Andreotti noted that the ageing global population is prone to atrial fibrillation, venous thromboembolism and atherothrombotic stroke, thus they may derive additional benefit from anticoagulation.
In summary, in support of replacing antiplatelets for anticoagulants, she stated that targeting coagulation to prevent MACE has a rational basis, that low-dose anticoagulation on top of SAP is beneficial in a COMPASS-like population, and low-dose anticoagulation, with or without antiplatelet therapy may play a role in primary and post-ACS prevention of MACE. Still, head-to-head comparison of low-dose anticoagulation plus SAP vs.SAP therapy (vs. nothing?) for primary prevention in subjects deemed at high risk of MACE, is needed.
Anticoagulation should replace antiplatelets in CAD prevention – CON
Jean-Philippe Collet (Paris, France)
Dr. Collet set out to defend the opposite standpoint. He showed a time line of trials in this field, evaluating various anticoagulation and antiplatelet strategies. Over 30 years ago, OAC was taken out of the evaluated regimens, and now it is the topic of discussion to put anticoagulation back in.
In stable CAD, treatment with SAP is well-established; both the European and US guidelines recommend monotherapy with aspirin in this setting. In the 2017 ESC Focused Update on DAPT in Coronary Artery Disease, DAPT is recommended for 6 months after PCI in stable CAD and for 12 months after PCI after ACS, after CABG in case of ACS, and in case of medical treatment alone, but after an ACS. These recommendations apply in absence of a high bleeding risk; if the bleeding risk is high, DAPT is recommended for shorter periods. In all these situation without high bleeding risk, DAPT may be extended after 12 months (IIb recommendation).
Most patients stabilize after one year. Therefore, DAPT beyond one year after ACS has been evaluated. Over a mean duration of 31 months, DAPT was shown to give various CV benefits, including on the risk of MACE, CV death, MI, stroke and stent thrombosis. But this is not a very long period, Collet argued, and if you would look at 10 years, extra benefit would likely be seen. The effect on colon cancer should also be taken into account here.
Safety of treatment strategies should also be considered. Collet showed data of the ATLAS ACS 2-TIMI 51 trial, in which rivaroxaban 2.5 mg and 5.0 mg were compared with placebo. Patients with previous stroke were excluded. Intracranial hemorrhage (ICH) was seen more often with rivaroxaban than with placebo (placebo: 0.2% per 2 years, rivaroxaban 2.5 mg: 0.4% per 2 years and 5 mg: 0.7% per 2 years, rivaroxaban vs. placebo: P=0.009). No statistically significant differences between rivaroxaban and placebo were seen in the occurrence of fatal bleeding events (0.2%, 0.1% and 0.4% respectively for placebo, 2.5 mg and 5 mg) or fatal ICH (0.1%, 0.1% and 0.2%). There was no independent indication of who is at risk of stroke.
Collet illustrated his position with a patient case, who would have been eligible for the COMPASS trial. The 38-year old male had anterior NSTEMI, with double vessel disease and risk factors. The question is whether adding a factor Xa-inhibitor would be appropriate here. He would probably not do that, because there is no labelling for this situation, because it is not reimbursed, and because it gives an unpredictable risk of ICH. Moreover, there are established alternatives with fewer safety hazards. But how about a year later? The risk of ischemic events is high, but the patient is asymptomatic and has no ischemia/MI scar. In a vote, the largest part of the audience chose ‘switch from P2Y12 inhibitor to low-dose factor Xa-inhibitor’ among four options (other options: stop P2Y12 inhibitor, maintain the same DAPT regimen and de-escalation from prasugrel to clopidogrel). Collet would choose the same.
Previously, safety events were related to drug adherence and logistic issues associated with vitamin K antagonist therapy, as suggested by the WARIS-2 study, in which episodes of major, nonfatal bleeding were more frequent in those treated with warfarin plus aspirin, as compared with aspirin. The COMPASS results did not show a significant reduction of stroke, MI and CV death with rivaroxaban alone compared with aspirin. Thus, efficacy with regard to the ischemic endpoint with anti-factor Xa has not been demonstrated, while a significant increase in intracranial bleeding was seen. Moreover, Collet noted a trend towards more fatal bleeds.
Collet sees some pending issues with the obtained results on rivaroxaban alone, namely that early termination of the COMPASS trial may have overestimated the therapeutic effect. Furthermore, in light of the bleedings requiring blood transfusion and/or hospitalization, he wonders about the net clinical benefit. Patients should be informed that they may face more intracranial bleeds. Moreover, add-on therapy induces the challenge of treatment adherence and cost-issues.
Considering all this, he concluded that aspirin should remain standard of care for stable CAD (class I recommendation), and DAPT may be used when the bleeding risk is low (class IIb). Dual therapy with anti-factor Xa may be an option (IIb), but not without stressing the following: only in addition to aspirin, without the possibility of titration of the treatment intensity, no early initiation after PCI, and it has a lower net clinical benefit than DAPT.
In her rebuttal, Andreotti pointed out that Collet mentioned that anticoagulation was a disaster before 2012. Yet she had shown the striking superiority of warfarin over aspirin in the 60+ trial, while SAP had no effect on mortality. Moreover, ticagrelor tended to loose efficacy in PEGASUS, after 33 months. “What are 30 months compared to a lifetime?”, she wondered. She advocated to go beyond PCI, otherwise we are not good doctors.
She ended by summing up her major points, namely a survival benefit with anticoagulation has repeatedly emerged from the Prevention Trial in England, the Dutch 60+ trial, and COMPASS. To her, these are not coincidental findings, she believes they are mechanistic. Fatal bleeds were small in number. “We now have the great advancement of NOACs: they are more practical. Measuring tools are available, they have rapid onset and offset of action, anti-dotes are available, and the pharmacokinetics are known." Finally, to prove her point, she mentioned that the European Medicine’s Agency approved the COMPASS strategy two days before the debate, for treatment of COMPASS-like patients.
Collet, in his rebuttal, went back to his case. For his patient, he liked to stick to DAPT: he switched to clopidogrel, 8 months after PCI, which was interrupted a year later on the request of the primary care physician. Four years later, the patient suffered from acute occlusion of the posterolateral from RCA. Collet noted that the event was related to disease progression, not to the original event.
Ongoing trials are eagerly awaited, including the GLOBAL LEADERS that would be presented at the ESC congress the day after the debate. Collet argued that if we show that more potent P2Y12 inhibitors are safe, why would we switch to anticoagulation and wait for more bleeds? Also the COMMANDER-HF, presented as ESC, will be informative: this type of trial should drive our decisions.
If trials turn out not to show good results, then we should rethink the options. Until then, he stays with his viewpoint for practical reasons and because of the scientific evidence. His major reason is the risk of intracranial bleeds.
Now, how did this debate affect the opinion of the audience? When the starting questions were asked again, the first (for secondary prevention in high-risk stable coronary artery disease (CAD), a low dose of a NOAC should be added to aspirin) had shifted from a balanced vote to 73% saying yes and 27% voting no. On the second question (for secondary prevention in high-risk stable CAD, a P2Y12 antagonist should be added to aspirin), the audience was even a little less divided than before the debate, with 54% voting yes.
In his concluding remarks, Deepak Bhatt noted that a large part of the audience chose the COMPASS strategy, but also DAPT, and both answers are right. Lots of data has come in, and mostly use of DAPT has been validated in various settings. Research is still ongoing and lots of things remain to be accomplished in the field. Lots of answers will still follow. But Bhatt concluded on a positive note, by saying that in the field of thrombocardiology, we have terrific options already, which are backed up by data, for approved indications.