1407 In Atrial Fibrillation, Size Does Matter DAVID E. HAINES, M.D. From the Cardiovascular Divisioti, University of Virginia Health System, Charlottesville, Virginia Editorial Comment Since early reports of patients undergoing the maze operation,' investigators have attempted to optimize lin- ear atrial ablation procedures for the cure of atrial fibril- lation. It has been hypothesized that linear atrial ablation prevents the propagation of multiple reentrant wavelets because the new linear anatomic barriers connect the natural anatomic barriers and compartmentalize the atria. If the anatomic size of a reentrant wavelet (a function of its wavelength) is larger than the unobstructed geo- graphic region available for impulse propagation, then that wavelet should extinguish. If there is an insufficient number of simultaneously propagating wavelets, then the arrhythmia should terminate.^ New findings in the study of atrial fibrillation mechanisms include the observation from Langendorff-perfused ovine heart preparations that there may be dominant rotors that act as drivers in atrial fibrillation with fibrillatory conduction instead of multi- ple wavelet reentry in the remaining regions of the atria.^ Hence, disruption of driving rotors by linear ablative lesions also would achieve the salutary effect of termi- nation of this arrhythmia. It has been proposed that atrial electrophysiologic remodeling is a major factor leading to persistence of atrial fibrillation.'' Atrial fibrillation leads to shortening of the atrial refractory period that is blocked acutely by calcium channel blockers.^'' In experimental animals, rapid atrial pacing caused an increase in intracellular calcium over a 2-week time period until homeostatic mechanisms predominated and intracellular calcium lev- els ultimately normalized by 16 weeks.^ The molecular mechanisms of this process include a reduction of the transient outward potassium current 1^^ and the L-type calcium current I^aL affected by down-regulation of the pore-forming channel protein subunits.^" As the atrial refractory periods shorten, the wavelengths shorten, the wavelet sizes decrease, and more wavelets are able to exist simultaneously within a fixed geographic region. Conversely, reverse electrophysiologic remodeling is characterized by return of atrial refractory periods to baseline values after cessation of atrial fibrillation (usu- ally within hours) with an associated increase in atrial fibrillation wavelength. Early recurrence of atrial fibril- lation after cardioversion may be attributed, at least in J Cardiovasc Electrophysiol, Vol. 11, pp. 1407-1408, December 2000 Address for correspondence: David E. Haines, M.D., Box 800158, Cardiovascular Division, University of Virginia Health System, Char- lottesville, VA 22908. Fax 804-243-2625; E-mail: dhaines@virginia.edu part, to increased atrial vulnerability due to electrophysi- ologic remodeling.'2 However, electrophysiologic re- modeling and reverse remodeling cannot explain the propensity for atrial fibrillation recurrence long after reverse electrophysiologic remodeling has been com- pleted. Thus, a so-called "second factor" has been pro- posed to predispose susceptible atria to atrial fibrillation recurrence.'^ Anatomic remodeling in atrial fibrillation occurs on all levels. It has long been recognized that gross atrial size increases over time in response to atrial fibrillation, although the mechanism by which this occurs is less clear. Atrial myocytes undergo a dramatic transformation in response to chronic high heart rates, including disso- lution of the myofibrils, increase in intracellular glyco- gen, dispersion of the chromatin pattern, increase in number of mitochondria, and fragmentation of the sar- coplasmic reticulum.'* This cellular process has been described as dedifferentiation, resulting in myocytes that appear to have many of the characteristics of hibernating myocardium. The time course and degree of reversibility of anatomic remodeling has not been fully characterized, but persisting anatomic abnormalities likely account for the "second factor" in the maintenance and late recur- rence of atrial fibrillation. .>In this issue of the Journal, Avitall et al." describe changes in atrial size and function after creation of linear atrial lesions with a catheter-based radiofrequency en- ergy ablation system. Importantly, atrial size was de- creased by 21% in normal canines and 40% in canines with pacing-induced atrial fibrillation after performing linear atrial ablation. Atrial fibrillation was terminated by linear atrial lesions in all animals, and no spontaneous recurrence was observed during the 5-month follow-up period. One might conclude that terminating atrial fibril- lation was entirely dependent on creating continuous and contiguous linear lesions, but the ablation-induced de- crease in atrial size may be an important component of the overall success of this procedure. Another study of linear atrial ablation in a canine model of chronic atrial fibrillation reported successful atrial fibrillation termina- tion and prevention of reinduction despite incomplete creation of linear lesions in >30% of cases.'* The suc- cess of the surgical maze procedure is attributed to cre- ation of linear lesions, but amputation of both atrial appendages and, in some cases, additional atrial size reduction are performed with this procedure as well.''''^ Thus, by reducing the overall size of the atria in which rotors and reentrant wavelets may propagate, atrial de- bulking and size reduction may be important components of the success of linear atrial ablation procedures.