High-resolution mapping of reentrant atrial tachycardias: Relevance of low bipolar voltage

BACKGROUND Bipolar voltage is widely used to characterize the atrial substrate but has been poorly validated, particularly during clinical tachycardias.OBJECTIVE The purpose of this study was to evaluate the diag-nostic performance of voltage thresholds for identifying regions of slow conduction during reentrant atrial tachycardias (ATs).METHODS Thirty bipolar voltage and activation maps created dur -ing reentrant ATs were analyzed to (1) examine the relationship be-tween voltage amplitude and conduction velocity (CV), (2) measure the diagnostic ability of voltage thresholds to predict CV, and (3) identify determinants of AT circuit dimensions. Voltage amplitude was categorized as "normal" (.0.50 mV), "abnormal" (0.05-0.50 mV), or "scar" (<0.05 mV); slow conduction was defined as <30 cm/s.RESULTS A total of 266,457 corresponding voltage and CV data points were included for analysis. Voltage and CV were moderately correlated (r 5 0.407; P < .001). Bipolar voltage predicted regions of slow conduction with an area under the receiver operating char- acteristic curve of 0.733 (95% confidence interval 0.731-0.735). A threshold of 0.50 mV had 91% sensitivity and 35% specificity for identifying slow conduction, whereas 0.05 mV had 36% sensitivity and 87% specificity, with an optimal voltage threshold of 0.15 mV. Analyses restricted to the AT circuits identified weaker associations between voltage and CV and an optimal voltage threshold of 0.25 mV.CONCLUSION Widely used bipolar voltage amplitude thresholds to define "abnormal" and "scar" tissue in the atria are, respectively, sensitive and specific for identifying regions of slow conduction during reentrant ATs. However, overall, the association of voltage with CV is modest. No clinical predictors of AT circuit dimensions were identified.