November 28th, 2011
Roundtable: Two Trials, Two Anticoagulants, Many Implications
Samuel Goldhaber, MD, Elaine Marie Hylek, MD, MPH and Christian Thomas Ruff, MD, MPH
After findings from the industry-funded ATLAS ACS 2–TIMI 51 and ADOPT trials were presented at AHA and published in the New England Journal of Medicine, CardioExchange asked three experts to weigh in.
Background
The ATLAS ACS 2–TIMI 51 investigators randomized 15,526 patients within 7 days after hospital admission for an ACS to standard therapy plus either placebo or rivaroxaban in one of two doses (2.5 or 5 mg) twice daily. At a mean follow-up of 13 months, the rate of the composite primary endpoint (death, MI, and stroke) was significantly lower among rivaroxaban recipients (both dose groups combined) than among placebo recipients (8.9% vs. 10.7%). Compared with placebo, rivaroxaban at either dose was associated with significant increases in major bleeding (2.1% vs. 0.6%) and intracranial hemorrhage (0.6% vs. 0.2%). Rivaroxaban has been approved by the FDA, but (like other anticoagulants) not for ACS.
In ADOPT, 4495 hospitalized medical patients at risk for thromboembolism were randomized to receive apixaban for 30 days or enoxaparin for 6 to 14 days. After 30 days, the groups’ rates of thromboembolic events were similar, but major bleeds occurred twice as often with apixaban (0.47%) as with enoxaparin (0.19%). Apixaban is awaiting FDA approval.
Question to the Experts
Given the evidence to date, how quickly should rivaroxaban and, if it’s approved, apixaban be used in clinical practice — and for what patient populations?
The Experts Respond
Samuel Goldhaber, MD (principal investigator of ADOPT, focusing here on ATLAS)
The landmark ATLAS ACS 2 trial changes forever the way we’ll think about the pathophysiology of ACS and the way we’ll manage patients with STEMI, NSTEMI, or unstable angina. Now we can discard the notion that ACS depends solely upon quelling the platelet and preventing “white clot.” After all, rivaroxaban is the same anticoagulant used to prevent extension of “red thrombus” hiding in the left-atrial appendage or common femoral vein. (Conversely, pulmonary embolism pathophysiology abounds with activated platelets, but we’ll save that discussion for another day.) By my count, many ACS patients will be prescribed six core medications to prevent recurrent events: 1) aspirin; 2) clopidogrel/prasugrel/ticagrelor; 3) a statin; 4) a beta-blocker; 5) an ACE inhibitor or ARB; and 6) rivaroxaban 2.5 mg twice daily.
The TIMI investigators were courageous to test the tiny dose of rivaroxaban 2.5 mg twice daily. It’s only one quarter the dose used to prevent stroke in atrial fibrillation. And it’s only half the dose used to prevent DVT and pulmonary embolism in patients undergoing total-hip or -knee replacement. As cardiovascular clinicians, we’re used to prescribing ultra-low doses of some medications, such as baby aspirin at 81 mg once daily. And some have advised (with a bit of a smile) that the aspirin dose can be lowered even further so that we can instruct our patients to simply “lick an aspirin tablet once a day.” We might place rivaroxaban in the same ultra-low-dose category.
The concept of ultra-low-dose warfarin to manage coronary artery disease has been floating around for generations. Several decades ago, one of my patients who had undergone CABG was having recurrent angina but wasn’t a candidate for reoperation CABG or for PCI. I asked his cardiac surgeon for advice. This wise surgeon (now deceased) gave me a recommendation without any explanation or evidence: “Try warfarin 1 mg a day.” Sure enough, the angina abated and my patient lived another 12 years. Is rivaroxaban the modern-day version of licking a warfarin tablet?
I have some questions for the TIMI investigators who designed and executed ATLAS ACS 2, as well as for clinicians who blog on CardioExchange:
1) What insights can we gain about understanding the differences between ACS and atrial fibrillation, given the 4-fold difference in rivaroxaban dose to manage these two diseases?
2) If rivaroxaban 2.5 mg twice daily is superior to rivaroxaban 5 mg twice daily for preventing cardiovascular and all-cause mortality, might rivaroxaban 2.5 mg once daily yield even more striking results than rivaroxaban 2.5 mg twice daily?
3) Does having one novel anticoagulant suffice for secondary prevention of ACS? Should dabigatran and apixaban be tested again with lower doses than previously used?
4) Should rivaroxaban be tested for lowering the rate of stent thrombosis outside the setting of ACS?
5) Will hospital formularies be tempted to stock only a single novel anticoagulant, one that can be used for SPAF, prevention and treatment of VTE, and ACS?
What an exciting time to be a practitioner of cardiovascular medicine! We’re gaining new insights and discovering new life-saving therapies at an impressively rapid pace.
Elaine Marie Hylek, MD, MPH (focusing on ATLAS)
ATLAS ACS 2 excluded patients with prior stroke or clinically significant gastrointestinal bleeding in the previous 12 months. The mean age was 62 years, 9% of participants were age 75 or older, and 75% of participants had a creatinine clearance of ≥68 mL/minute. Given these baseline characteristics, it would be difficult to extrapolate these trial results to older patients or to patients with renal impairment. Younger patients who have high risk for recurrent ischemic events and low baseline bleeding risk are likely to be those with the most favorable benefit-to-risk ratio.
On ATLAS
The findings from ATLAS ACS 2 were perhaps the biggest surprise at AHA. The most interesting result was that the lower dose of rivaroxaban, 2.5 mg twice daily, demonstrated substantial reductions in CV and all-cause mortality that were not seen with the higher dose. One of the biggest questions fluttering around the convention hall was why this trial was positive when similar trials with novel anticoagulants, APPRIASE-2 (apixaban) and RUBY-1 (darexaban), had failed. The answer is dose.
After a robust phase 2 dose-ranging study, the investigators chose doses of rivaroxaban that were half (5 mg twice daily) and one quarter (2.5 mg twice daily) of the full anticoagulant dose used in AF (20 mg once daily). It may be that with respect to anticoagulant therapy in ACS, less is more. More-intensive antithrombotic therapy may attenuate the ischemic benefits by directly increasing severe and fatal bleeding or bleeding that results in discontinuation of important concomitant medical therapy. This may also help to explain the apparent paradoxical reduction in mortality with the “very low” dose of rivaroxaban, compared with the “low” dose. Some important questions remain:
1) What was the mechanism of the mortality reduction, as the substantial benefit was not accounted for by reductions in MI or stroke?
2) How do we reconcile these data with the impressive results in ACS with the potent antiplatelet drugs ticagrelor and prasugrel?
What’s clear is that ATLAS changes the landscape in ACS: Chronic treatment with an anticoagulant, at least at a very lose dose, offers the potential of substantial benefit. Nevertheless, the ideal combination of antiplatelet and anticoagulant regimens is yet to be determined, as is clinicians’ comfort with the higher rates of major bleeding, particularly intracranial hemorrhage.
On ADOPT
ADOPT addressed the last frontier in VTE prophylaxis: hospitalized medically ill patients. Although a negative study, it offers two key lessons:
1. VTE prophylaxis trials should no longer include the soft endpoint of asymptomatic incidental DVTs picked up by ultrasounds, as they are not performed as part of routine care and distort the natural history of DVT. ADOPT showed a trend toward a significant reduction in the hard endpoint of symptomatic VTE and VTE-related deaths in the apixaban group, but this secondary endpoint was underpowered.
2. More important: ADOPT, along with complementary data from MAGELLAN with rivaroxaban in a similar population, illustrates that our entire paradigm of VTE prophylaxis in medically ill patients may need to be rethought. The event curves reveal that the rates of VTE continued to increase after the cessation of enoxaparin after 6 to 14 days, compared to extended treatment with apixaban. The risk for VTE appears to continue unabated in medically ill patients well after discharge, and patients may accrue risk after 30 days.
While factor Xa inhibitors’ role is clear for DVT prophylaxis in orthopedic patients and for stroke prevention in AF, further trials of extended prophylaxis are needed to further define the optimal regimen and treatment duration for VTE prophylaxis in medically ill patients. At least we’re beginning to ask the right questions.
How do you think factor Xa inhibitors will change clinical practice? Offer your perspective on ATLAS and ADOPT — and on our experts’ opinions.
6 Responses to “Roundtable: Two Trials, Two Anticoagulants, Many Implications”
NEJM Journal Watch — Recent Cardiology Articles- Anticoagulation in Patients with Cancer-Associated Pulmonary Embolism: How Long is Long Enough?
- Spironolactone Use After Percutaneous Intervention for Acute Myocardial Infarction
- CRISPR/Cas9 Gene Editing for Transthyretin Cardiomyopathy
- Does Colchicine Reduce Cardiac Events After Myocardial Infarction?
- First-Line Therapy for Ventricular Tachycardia After MI: What's the Best Approach?
Fascinating discussion. Thank you! Dr Goldhaber mentions the 6 requisite drugs that will be needed in all (or most) ACS patients. Speaking of this, in The ATLAS ACS 2–TIMI 51, what was the average statin dose? We know from PROVE-IT TIMI-22, A-to-Z Z phase, SAGE, TNT, etc. that high statin dosing improves outcomes.
Secondly, were beta blockers appropriately titrated to target HR? Were ACE inhibitors used in the most optimal doses, proven by clinical trials such as HOPE and EUROPA?
These are not the sorts of questions answered by Table 1’s, unfortunately, which tend to produce only binary data for medication (patient received therapy or didn’t receive therapy). Incidentally, I’ve included those data below:
ASA 99%, thienopyridine 93%, beta blocker 66%, ACE or ARB 39%, statin 84%.
Of course, these are the baseline data, and do not include initiation of additional therapies during follow-up.
My overall point is that before adding a sixth agent, perhaps we should maximize titration of the ones we should already be using. One should note, though, that optimal medical therapy in the clinical trial world is much, much better than in most clinical settings. It is still probably not good enough, however.
As always very astute observations from the experts and Dr. Goldhaber raises extremely interesting questions. I’m going to give a shot simply as a mental exercise:
1) What insights can we gain about understanding the differences between ACS and atrial fibrillation, given the 4-fold difference in rivaroxaban dose to manage these two diseases?
It does seem like an atherosclerotic cerebral bed (an extrapolation from the fact fact that the patient has suffered an ACS) is “vulnerable” to bleeding with these agents versus the subjects typically enrolled in Afib trials. What are these differences at a cellular level clearly deserves investigation. Maybe it is that the permeability of the drug is higher across the blood-brain barrier in folks with atherosclerotic events?
2) If rivaroxaban 2.5 mg twice daily is superior to rivaroxaban 5 mg twice daily for preventing cardiovascular and all-cause mortality, might rivaroxaban 2.5 mg once daily yield even more striking results than rivaroxaban 2.5 mg twice daily?
This is obviously anyone’s guess.
3) Does having one novel anticoagulant suffice for secondary prevention of ACS? Should dabigatran and apixaban be tested again with lower doses than previously used?
It’ll be extremely difficult to get investigators to test dabigatran in the ACS indication given it’s background in RE-LY and similar trends with ximelagatran. Would be great to have apixaban or edoxaban as options for the ACS indication. However, given the extent of the phase II evaluation that was required to get to that “sweet spot” dose with rivaroxaban-the end of the phase 2 and then 3 experiments would easily be in the 5-7 year time frame.
4) Should rivaroxaban be tested for lowering the rate of stent thrombosis outside the setting of ACS?
The stent thrombosis numbers with the newer generation everolimus stents are quite negligible. The potential to cause harm seems might be higher than a real benefit.
5) Will hospital formularies be tempted to stock only a single novel anticoagulant, one that can be used for SPAF, prevention and treatment of VTE, and ACS?
Very likely. Based on ATLAS, rivaroxaban was really started 4 days after the ACS-in the U.S that would mean initiation in an outpatient setting typically-the related question si will practitioners find it easier to simply gain familiarity with one drug and use it it various scenarios or take the effort to choose between DTIs, and multiple Xa inhibitor options.
Competing interests pertaining specifically to this post, comment, or both:
Fellow with the TIMI Study group.
What with all this discussion of novel (i.e. expensive) anticoagulants, has anyone else picked up on Dr. Goldhaber’s passing allusion to the possible benefits of low-dose warfarin? Is it not possible that low-dose warfarin (INR around 1.5 or less) would achieve the same results as rivaroxaban at a fraction of the cost?
Competing interests pertaining specifically to this post, comment, or both:
None
The plural of anecdote is not data.
This has been attempted in a couple randomized trials in the setting of 1) chronic coronary artery disease (Post-MI patients) and 2) atrial fibrillation. In neither case did it work well. These trials were published in the mid 1990’s to the mid 2000’s and I could dig them out if anyone wants them.
One of Dr. Goldhaber’s questions for the TIMI investigators who designed and executed ATLAS ACS 2 was “What insights can we gain about understanding the differences between ACS and atrial fibrillation, given the 4-fold difference in rivaroxaban dose to manage these two diseases?” One hypothesis that may suggest at least a partial answer to that question is: Controlling the thrombogenicity of the fibrillating atrial endocardium, whose expression of thrombomodulin and tissue factor pathway inhibitor are reduced, might require a higher level of systemic anticoagulation than is required to control the thrombogenicity of a ruptured coronary plaque.
Competing interests pertaining specifically to this post, comment, or both:
Competing interests: none.
For ACS, I think Xa inhibitors may help prevent the extension of clots and further complications whereas antiplatelets are helpful from early phases until later times. In line with that, whereas in recent antiplatelet trials, such as PLATO, patients were included within 24 hours of symptom onset, ATLAS-ACS2-TIMI-51 patients were enrolled within 7 days after hospital admission. Therefore, the concept of “white clot” might be true for early initiation of ACS but beyond that period it might actually be a “pink clot”! That is why rivaroxaban worked in the ATLAS-ACS2-TIMI-51. APPRAISE-2 was a negative trial, probably because the dose of apixaban was too high (it used the same dose as used for SPAF in ARISTOTLE).
As to why a very low dose of these novel anticoagulants worked in ACS but may not work well for SPAF, I think it comes back to the disease pathophysiology. Whereas for SPAF clot formation should be prevented, in ACS we need to prevent full development of an evolving clot in the bed of ruptured plaque(s). Moreover, the ACS patients already receive several other therapies that affect clot formation or propagation.
A few questions:
1. Where does heparin (or LMWHs) fit in the contemporary therapy of ACS? Some old studies suggested certain benefits (Théroux P et al. N Engl J Med. 1988; 319: 1105-11. And Telford AM, Wilson C. Lancet. 1981; 1225-8.) but some others such as ISIS-3 (Lancet. 1992; 339: 753-70) suggested no benefit from heparin therapy. Could it be that a very lose dose heparin can similarly play a helpful role?
2. I am also very keen to know whether novel anticoagulants can help the ACS patients who are resistant to clopidogrel. Despite mechanistic justification, ticagrelor is not the solution to this problem (Wallentin L, Lancet. 2010; 376: 1320-8). A recent study has shown some promise with tripling the dose but there is still a long way to prove that this strategy works for such patients (Mega JL, et al. JAMA. 2011; 306: 2221-8.).
3. Irrespective of the costs, findings of the PLATO trial suggested that ticagrelor is superior to clopidogrel, albeit with increased rate of major bleeding. Similarly, prasugrel is more effective than clopidogrel in reducing ischemic events but increase the rates of bleeding (based on TRITON-TIMI-38, awatingin TRILOGY ACS). Until clinical trial data will be available on combination of novel antiplatelets and anticoagulants, would it be helpful to add rivaroxaban (or another Xa-inhibitor or DTI) on top of ticagrelor/prasugrel in patients receiving standard ACS therapy?
4. Last, it would be nice to look at the opposite direction and see how novel antiplatelets work for prevention of VTE, compared to anticoagulants. The PEP trial had suggested a role for antiplatelents in VTE prevention but other antiplatelets have not been tested for thromboprophylaxis.
Competing interests pertaining specifically to this post, comment, or both:
None