Left Atrial Structure and Function in Atrial Fibrillation
Left Atrial Structure and Function in Atrial Fibrillation
In a pre-specified echocardiographic substudy of the ENGAGE AF-TIMI 48 trial, we identified strong correlations between increasing abnormalities of LA structure and function with greater burdens of AF and higher CHADS2 score, an estimate of stroke risk. While the majority of AF subjects had LA enlargement, we also demonstrated impaired LA function among a sizeable number of subjects with normal LA size. Furthermore, we found that LA contractile function was abnormal even in subjects in sinus rhythm. These findings suggest that the assessment of LA function may add important new information in the evaluation of the AF patient.
In an effort to improve stroke risk stratification beyond that achieved with conventional clinical characteristics, others have explored the importance of LA size, but with mixed results in predicting stroke. While the majority of AF patients have enlarged LA, 36% of AF subjects in this study had normal LA size, suggesting that a substantial number of AF patients are at moderate-to-high risk for stroke based on the CHADS2 score despite normal LA size. Given the limitations of LA size, increasing attention has been given to LA function in AF. We identified LA dysfunction in approximately half of subjects with normal LAVI, suggesting not only that LA mechanical dysfunction may precede LA enlargement, but also that those with LA dysfunction may be at a higher risk for stroke despite normal LA size. This raises the possibility that the evaluation of LA function may be helpful in assessing stroke risk and in the identification of those who may benefit from anticoagulation strategies. Importantly, we also found that 17% of AF subjects had both normal LA size and function, which may not only represent a subgroup at a lower risk for stroke, but also patients most likely to respond to rhythm control strategies, although, these hypotheses remain to be formally tested in outcomes based studies.
Several studies demonstrate a similar risk of stroke regardless of intermittent or sustained forms of AF, hence, current guidelines recommend anticoagulation regardless of whether AF is paroxysmal, persistent, or permanent. However, these studies did not include evaluations of cardiac structure and function to potentially explain the clinical findings. In this study, we identified LA contractile dysfunction in the subset of subjects in sinus rhythm at the time of echocardiography, including those with paroxysmal AF, suggesting electro-mechanical dissociation. While previously described in the post-cardioversion period, our findings suggest that electro-mechanical dissociation may be common in the broader AF population, including those with paroxysmal AF. Left atrial contractile dysfunction among paroxysmal AF patients may not only help to explain the similar risk of stroke between intermittent and sustained forms of AF, but also raises the possibility that patients with abnormal LA contractile function may have a higher risk of stroke than those with normal contractile function. Additionally, we found that features of cardiac structure and function associated with a higher CHADS2 score, including greater LV filling pressures and abnormal LV geometry, were similar across paroxysmal, persistent, and permanent AF. These results are consistent with proposed mechanisms for thrombus formation in AF, including stasis and endothelial dysfunction, and are congruent with existing data from smaller studies. Together, these findings may partially explain the previously reported similarities in stroke risk despite differing burdens of AF. However, whether these factors are of prognostic importance in the stratification of stroke risk needs to be prospectively tested.
While we evaluated a relatively large contemporary AF population, limitations should be noted. The cross-sectional design precludes the assessment of causal relationships between cardiac structure and function and AF type. The irregular rhythm of AF may lead to beat-to-beat variability in echocardiographic measurements; however, our assessments were averaged over multiple cardiac cycles. Cardiac structure and function may differ between those in AF and those not in AF at the time of echocardiography, although we performed sensitivity analyses with stratification by rhythm at the time of echocardiography and the results did not substantially change (Supplementary material online, Table S2). Assessing LA deformation and function with strain imaging may provide additional information and is a future direction. While we cannot exclude misclassification bias with respect to AF type, as paroxysmal, persistent, and permanent designations were determined by site investigators, our results are largely consistent with registries of AF patients. By design, ENGAGE AF-TIMI 48 enrolled an AF population at moderate-to-high risk for stroke (CHADS2 ≥2), therefore our assessment does not include the entire spectrum of CHADS2 scores. Furthermore, we utilized CHADS2 as a surrogate for stroke risk and therefore our findings need prospective evaluation in relation to clinical thrombo-embolic events.
In conclusion, we identified increasing abnormalities of LA structure and function with greater burdens of AF and stroke risk estimated by the CHADS2 score. Moreover, we found that impairments in LA function were present even among subjects with normal LA size and sinus rhythm. These findings suggest that the assessment of LA function may add important incremental information to the evaluation of AF patients.
Discussion
In a pre-specified echocardiographic substudy of the ENGAGE AF-TIMI 48 trial, we identified strong correlations between increasing abnormalities of LA structure and function with greater burdens of AF and higher CHADS2 score, an estimate of stroke risk. While the majority of AF subjects had LA enlargement, we also demonstrated impaired LA function among a sizeable number of subjects with normal LA size. Furthermore, we found that LA contractile function was abnormal even in subjects in sinus rhythm. These findings suggest that the assessment of LA function may add important new information in the evaluation of the AF patient.
In an effort to improve stroke risk stratification beyond that achieved with conventional clinical characteristics, others have explored the importance of LA size, but with mixed results in predicting stroke. While the majority of AF patients have enlarged LA, 36% of AF subjects in this study had normal LA size, suggesting that a substantial number of AF patients are at moderate-to-high risk for stroke based on the CHADS2 score despite normal LA size. Given the limitations of LA size, increasing attention has been given to LA function in AF. We identified LA dysfunction in approximately half of subjects with normal LAVI, suggesting not only that LA mechanical dysfunction may precede LA enlargement, but also that those with LA dysfunction may be at a higher risk for stroke despite normal LA size. This raises the possibility that the evaluation of LA function may be helpful in assessing stroke risk and in the identification of those who may benefit from anticoagulation strategies. Importantly, we also found that 17% of AF subjects had both normal LA size and function, which may not only represent a subgroup at a lower risk for stroke, but also patients most likely to respond to rhythm control strategies, although, these hypotheses remain to be formally tested in outcomes based studies.
Several studies demonstrate a similar risk of stroke regardless of intermittent or sustained forms of AF, hence, current guidelines recommend anticoagulation regardless of whether AF is paroxysmal, persistent, or permanent. However, these studies did not include evaluations of cardiac structure and function to potentially explain the clinical findings. In this study, we identified LA contractile dysfunction in the subset of subjects in sinus rhythm at the time of echocardiography, including those with paroxysmal AF, suggesting electro-mechanical dissociation. While previously described in the post-cardioversion period, our findings suggest that electro-mechanical dissociation may be common in the broader AF population, including those with paroxysmal AF. Left atrial contractile dysfunction among paroxysmal AF patients may not only help to explain the similar risk of stroke between intermittent and sustained forms of AF, but also raises the possibility that patients with abnormal LA contractile function may have a higher risk of stroke than those with normal contractile function. Additionally, we found that features of cardiac structure and function associated with a higher CHADS2 score, including greater LV filling pressures and abnormal LV geometry, were similar across paroxysmal, persistent, and permanent AF. These results are consistent with proposed mechanisms for thrombus formation in AF, including stasis and endothelial dysfunction, and are congruent with existing data from smaller studies. Together, these findings may partially explain the previously reported similarities in stroke risk despite differing burdens of AF. However, whether these factors are of prognostic importance in the stratification of stroke risk needs to be prospectively tested.
While we evaluated a relatively large contemporary AF population, limitations should be noted. The cross-sectional design precludes the assessment of causal relationships between cardiac structure and function and AF type. The irregular rhythm of AF may lead to beat-to-beat variability in echocardiographic measurements; however, our assessments were averaged over multiple cardiac cycles. Cardiac structure and function may differ between those in AF and those not in AF at the time of echocardiography, although we performed sensitivity analyses with stratification by rhythm at the time of echocardiography and the results did not substantially change (Supplementary material online, Table S2). Assessing LA deformation and function with strain imaging may provide additional information and is a future direction. While we cannot exclude misclassification bias with respect to AF type, as paroxysmal, persistent, and permanent designations were determined by site investigators, our results are largely consistent with registries of AF patients. By design, ENGAGE AF-TIMI 48 enrolled an AF population at moderate-to-high risk for stroke (CHADS2 ≥2), therefore our assessment does not include the entire spectrum of CHADS2 scores. Furthermore, we utilized CHADS2 as a surrogate for stroke risk and therefore our findings need prospective evaluation in relation to clinical thrombo-embolic events.
In conclusion, we identified increasing abnormalities of LA structure and function with greater burdens of AF and stroke risk estimated by the CHADS2 score. Moreover, we found that impairments in LA function were present even among subjects with normal LA size and sinus rhythm. These findings suggest that the assessment of LA function may add important incremental information to the evaluation of AF patients.
Source...