Rivaroxaban With or Without Aspirin for the Secondary Prevention of Cardiovascular Disease

Clinical Implications of the COMPASS Trial

Stephen J. Nicholls; Adam J. Nelson


Am J Cardiovasc Drugs. 2019;19(4):343-348. 

In This Article

How to Translate Compass to Clinical Practice?

The findings from COMPASS provide compelling evidence that addition of a DOAC reduces cardiovascular risk in patients with stable atherosclerotic vascular disease. The clinical benefits were consistent across all prespecified subgroups, namely sex, geographic region, race, body weight, renal function, and traditional cardiovascular risk factors, and across all age strata. This was accompanied by a statistically significant reduction in death, which further highlights the importance of these findings in the contemporary era. While bleeding risk was predictably greater in both rivaroxaban groups, this did not negate the net clinical benefit observed with treatment with low-dose rivaroxaban and aspirin compared with aspirin alone. This suggests a potential new treatment paradigm for patients with established atherosclerotic disease, in which the cardiovascular benefits are likely to outweigh any increase in clinically relevant bleeding.

The mechanism underlying the benefit of low-dose rivaroxaban in combination with aspirin is not certain. While this could reflect reductions in thromboembolism in the setting of subclinical AF, the reduction in a range of clinical events, not simply stroke, suggests a more widespread effect. Rather, it is more likely to reflect an atherosclerosis-mediated benefit, largely due to its effect on the extrinsic, hemostatic (thrombin forming) coagulation pathway—a process that is complementary to platelet activation and is mediated through the release of tissue factor during plaque rupture and erosion.[34]

The benefits in COMPASS were observed in a very large, multinational cohort. While a high rate of guideline-based treatments were used, patients were not required to have the strictest risk factor control possible. Statin use was not recorded in all patients, and high-intensity statin therapy was not a prerequisite. While the mean baseline total cholesterol was 162 mg/dL, not all patients are likely to have low-density lipoprotein (LDL) cholesterol levels below currently accepted treatment targets at either aggressive (70 mg/dL) or modest (100 mg/dL) levels. This is potentially of major importance, given the direct association between LDL cholesterol levels and progression of atherosclerotic disease.[35] With consistent evidence from imaging that patients typically require LDL cholesterol levels < 70 mg/dL to achieve plaque regression,[35–38] it is likely that many patients in COMPASS have an ongoing lipid stimulus for disease progression, rupture, and erosion. Whether such clinical benefits of low-dose rivaroxaban would be demonstrated in patients achieving very low LDL cholesterol levels remains unknown. This may be particularly important given that increasingly effective approaches now exist to achieve low LDL cholesterol levels in a greater proportion of patients with atherosclerotic disease than previously observed.[39–41]

In a similar fashion, more than one-third of patients in COMPASS had diabetes mellitus. The therapeutic landscape for treatment of patients with established cardiovascular disease and diabetes mellitus has changed, with evidence of clinical benefit with multiple glucose-lowering agents in addition to metformin.[42–45] While the mechanistic factors underlying these benefits remain uncertain, it is likely that their use, which has become commonplace in clinical practice, will influence event rates and potential benefits from the use of rivaroxaban in this group. With low-dose rivaroxaban now available in the USA and Europe, real-world and registry-based studies may allow greater insight into which groups are deriving the greatest risk reduction.

The findings from the peripheral arterial disease cohort warrant specific comment. It is well-recognized that the presence of peripheral arterial disease is a major risk factor for future cardiovascular events in patients with and without established coronary disease.[46] The finding of potentially greater proportional reductions in cardiovascular events reflects the modifiability of risk in these patients and further suggests that patients with peripheral arterial disease require aggressive secondary prevention and targeting of metabolic risk factors.[47,48] In addition, benefit was observed in terms of less MALE, including the need for revascularization. When combined with the robust limb benefits observed with intensive lipid lowering with the proprotein convertase subtilisin/kexin type 9 (PCSK9) inhibitor, evolocumab,[49] this suggests the importance of systemic medical therapy in targeting events due to atherosclerotic disease in a range of vascular territories. For patients with peripheral arterial disease, the need for revascularization or amputation are major clinical events with considerable implications for their health and wellbeing. The benefits observed in the peripheral arterial disease population of COMPASS come at the end of a long line of underwhelming trials of both antiplatelets[1,50,51] and antithrombotics[33] in this space. It is therefore not surprising that the prevention of MALE, an event that carries a 2-year risk of death of 14% and subsequent amputation of 27%,[52] has been favorably evaluated by the regulatory bodies and ought to translate into improved outcomes for this cohort of patients.

Additional information from COMPASS would help clarify how to best position the net clinical benefit findings of the study. The number of patients with prior stroke but not clinical disease in other vascular territories is relatively small. Therefore, information gained from study of this cohort is unlikely to be meaningful for these patients, who remain understudied. Accordingly, whether this strategy will be commonly used in patients with prior non-hemorrhagic stroke remains uncertain. The optimal duration of treatment with combination therapy similarly requires further investigation. Is this a lifelong commitment or does the degree of clinical benefit, relative to bleeding risk, change over time? Cost-effectiveness analysis will be important to determine how to optimally use this strategy in clinical practice.

The final factor to consider will be whether COMPASS provides compelling evidence for widespread use of low-dose rivaroxaban or whether we need a more personalized approach to use. In the setting of AF, parallel clinical algorithms permit stratification of risk of both thromboembolism and bleeding in an attempt to identify net clinical benefit of thromboprophylaxis prescription. COMPASS is a large clinical cohort, which provides a great opportunity to develop clinically useful tools to appropriately guide the use of potentially beneficial therapies. Whether this manifests in the clinic with "risk scores" similar to the "dual antiplatelet therapy (DAPT) score"[53] (aimed at helping clinicians identify those likely to benefit most from prolonged DAPT) remains to be seen. In an emerging world of precision medicine, such an approach would seem to make sense. In an environment of constrained health funding, ongoing research will need to delineate the balance between net clinical benefit and cost effectiveness at both an individual (consumer out-of-pocket expense, or insurer) and a population level (health system, economy). While this will have geographic relevance based on local costs and payment systems, linkage of large clinical and administrative datasets and modeling will provide much needed insight.[54,55]