Article Text
PDF
Statistics from Altmetric.com
Cardiologists’ perception of the risk of stroke in rheumatic heart disease (RHD) is influenced more by anecdote and opinion than by the evidence.1 Despite data indicating a similar, if not a lower risk of stroke, even among the potentially highest-risk patients with RHD (those with mitral stenosis (MS) and atrial fibrillation (AF)),1 there is a pervasive belief that they may be at the ‘highest risk of thromboembolism’.2 As a consequence, in current practice, any patient with significant valvular heart disease and AF (valvular AF) becomes a candidate for long-term oral anticoagulation, without further risk stratification. The perception of very high stroke risk in RHD is also reinforced by estimates of stroke burden, which are based on extrapolation from the proportion of stroke admissions who have RHD (3%–8%),3 rather than the proportion of strokes attributable to RHD in the community (1%–2%).4 Contemporary data on stroke risk in RHD are therefore needed to realign risk perceptions to reality, to aid in risk stratification and to help make decisions on long-term anticoagulation. The study by Vasconcelos and colleagues in this issue of Heart is therefore a welcome addition to the literature in this area.5
Incidence, predictors and risk stratification
Patients with valvular AF have a risk of stroke similar to that of those with non-valvular AF. Data from the Framingham study showed that the absolute risk of stroke was 4.5 per 100 patient-years among patients with MS and AF, similar to the risk in those with non-valvular AF (4.2 per 100 patient-years).6 In passing, it is worth noting that most references to this study in the literature (including the one by Vasconcelos et al) have chosen to highlight the relative risk in patients with MS and AF compared with an age-adjusted normal population (a ‘17-fold’ increase in risk), thereby reinforcing the misperception of very high risk. In another study, among patients with mitral valve disease and AF not receiving any anticoagulation, the risk of stroke was 5.1 per 100 patient-years, compared with 5.9 per 100 patient-years among those with non-valvular AF.7
However, stroke risk is not uniform among patients with valvular AF (table 1). Data from prospective studies suggest that risk varies between 0.4% and 4.2% among those with MS and AF who, from pathophysiological considerations, are believed to be at the highest risk.1 In general, older patients with a high prevalence of hypertension, diabetes, coronary disease and prior stroke have a higher risk8 (table 1). However, stroke risk also varies widely among younger patients with RHD who have a low prevalence of risk factors for atherosclerosis. The large difference in stroke rates between the Rheumatic Heart Disease Global Registry (REMEDY) and the study by Vasconcelos and colleagues (table 1) suggests that further risk stratification may be needed in order to optimise the risk–benefit trade-off inherent to long-term anticoagulation.5 9 Limited data support the notion that, in addition to older age and AF, the presence and severity of untreated MS, left atrial spontaneous echo-contrast or thrombus, aortic regurgitation,and heart failure may also increase stroke risk10 (figure 1). The study by Vasconcelos et al unfortunately did not evaluate heart failure as a risk factor and does not suggest that MS increased stroke risk. This could be because 80% of the patients in the study had significant MS, thereby reducing the utility of mitral valve area as a predictive variable.5 The study is also limited by the small number of strokes, in relation to the number of potential predictive factors, thereby greatly reducing statistical power. Given the aforementioned considerations, blanket recommendations for the use of anticoagulation for all patients with valvular AF may not be appropriate. However, reliable data to guide risk stratification are lacking. The results of larger prospective studies of unselected patients with RHD, such as the Investigation of Rheumatic Atrial Fibrillation Treatment Using VKAs and Rivaroxaban Registry, may provide more information in this regard.11
Differences in patient characteristics, and risk of stroke and death among patients with AF with significant RHD and other valve disease
Pathophysiology and potential risk factors for stroke in RHD. figure showing the complex interaction between multiple potential risk factors for stroke in RHD. Risk factors are enclosed in red boxes. Boxes in deeper shades of red indicate a possibly stronger pathophysiological association with stroke. Green boxes enclose factors which may protect from stroke by altering a particular risk factor (indicated by the green arrows). These risk factors are also associated with a higher prevalence of prior stroke, which in turn increases future risk. AF, atrial fibrillation; ASCVD, atherosclerotic cardiovascular disease; DM, diabetes mellitus; HF, heart failure; HTN, hypertension; LA, left atrial; LAA, left atrial appendage; RHD, rheumatic heart disease; SEC, spontaneous echo contrast.
High competing risk of death in RHD and AF
Patients with AF have a higher risk of death compared with those in sinus rhythm, irrespective of the presence of significant valve disease. While the risk of death (which is largely due to heart failure) is about two to three times the risk of stroke in patients with non-valvular AF,12 it may be many times higher in patients with RHD. The risk of death in REMEDY was an order of magnitude greater than the risk of stroke (table 1).13 However, this was not observed in the study by Vasconcelos et al, despite enrolling a substantial proportion of patients in New York Heart Association class III or IV. The reasons for this discrepancy are not entirely clear. The study population may have been biased by selectively enrolling a population at high stroke risk (which was the primary objective of the study). Moreover, patients were censored at the time of undergoing valve surgery or intervention, thereby ignoring deaths which may have occurred beyond this period.5 Nevertheless, the authors present an analysis adjusted for the competing risk of death, which showed an attenuation of the predictive value of AF for stroke. AF in patients with RHD is largely restricted to patients who have severe mitral valve disease (particularly MS) and left atrial hypertension and dilatation, all of which may contribute to the risk of stroke. Therefore, AF in RHD may be more a marker of severe disease and a less robust predictor of stroke (after adjustment for the other metrics of disease severity). The high competing risk of death in these patients has another major implication. Most deaths in patients with RHD are due to heart failure,13 and it is unlikely that the magnitude of benefit seen with anticoagulation in non-valvular AF will be replicated in patients with RHD. The high risk of heart failure-related death points to the need to prioritise treatment of heart failure (by timely surgical and percutaneous interventions) over simply anticoagulating patients with RHD and AF. As seen in some studies, these interventions may also work synergistically with anticoagulation in reducing stroke risk.10
In summary, there is wide variation in the risk of stroke among patients with RHD. Even among patients with AF, stroke risk may be low because of young age and the low prevalence of conventional risk factors for stroke. Further risk stratification based on metrics of valve disease severity (such as mitral valve area, left atrial size and heart failure) may be needed to identify patients who may benefit from long-term anticoagulation. The presence of AF alone may not be a robust enough marker of stroke risk in patients with RHD and significant valve disease.
References
Footnotes
Twitter @drkarthik2010
Contributors GK is the sole author of this work.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Patient and public involvement Patients and/or the public were not involved in the design, conduct, reporting or dissemination plans of this research.
Provenance and peer review Not commissioned; externally peer reviewed.
Linked Articles
- Valvular heart disease