|Home | About | Journals | Submit | Contact Us | Français|
Catheter ablation of ventricular tachycardia (VT) is a well-established therapeutic technique. Focal VT in structurally normal hearts can be cured by catheter ablation and meaningful clinical results can be achieved in patients with structural heart disease who have drug refractory arrhythmias resulting in device therapy. When hemodynamically stable and sustained, VTs, can be ablated using activation and entrainment mapping. In the past several years, even hemodynamically unstable ventricular tachycardias have been managed by catheter ablation. This has been achieved by targeting potential arrhythmia circuits that are identified by mapping the myocardial substrate during sinus rhythm (or a paced rhythm) (1,2). The evolution of substrate-based ablation of ventricular tachycardia has been aided by three-dimensional electroanatomical mapping that has allows localization of scars in relation to the ventricle using local electrogram voltage characteristics obtained using point-by-point intracardiac mapping in sinus rhythm.(3) With an accurate real-time representation of scar anatomy, ablation along border zone tissue to mimic surgical encircling subendocardial resection has been demonstrated to be an effective technique when VT is unsuitable for entrainment mapping.(4-8) The recognition of the importance of the electrical characteristics of scars in the genesis of VTs and the characterization of their extent and location has been a major advancement in the field of VT ablation.
The feasibility and efficacy of substrate-based radiofrequency ablation of ventricular tachycardia has been well characterized in patients with post-infarct VT. In an initial experience of 16 patients, linear ablation lesions placed in areas where pacemap matches were seen, across border zones, and from areas of dense scar to normal tissue or valve continuities resulted in a 75% freedom from recurrence at eight months(5). Subsequently, single center experiences and multicenter registries in post-myocardial infarction VT ablation have been published (9-15), as well as the first multi-center prospective randomized SMASH VT trial (16), which demonstrated 65% reduction in ICD therapy in patients undergoing adjunctive substrate-based ablation with ICD implantation.
While ischemia produces a predictable wavefront of necrosis (and subsequent scars) progressing from subendocardium to epicardium usually confined to a specific coronary vascular territory (17,18), scars in nonischemic cardiomyopathy has been shown to have a predilection for the mid myocardium and epicardium.(19) The current understanding of scar formation, location, and extent in nonischemic cardiomyopathy is limited compared to post myocardial infarct scars. Analysis of explanted human hearts with dilated cardiomyopathy has revealed increased fibrosis, myocyte disarray, and membrane abnormalities (20-22). In one of the largest autopsy series published, only 14% of 152 patients with idiopathic dilated cardiomyopathy had evidence of grossly visible left ventricular scar. However, histologic examination demonstrated interstitial or replacement fibrosis in 57% of patients analyzed. (23) As a result, fractionated recorded electrical activity in patients with nonischemic cardiomyopathy has been thought to result from non-uniform anistropic conduction through myocardium separated by fibrous tissue. (24)
Detailed substrate mapping studies on nonischemic patients are also limited. Myocardial scars indistinguishable from infarct scars have been demonstrated in patients with nonischemic cardiomyopathy (15) Hsia et al characterized the endocardial substrate of 19 consecutive patients with cardiomyopathy of nonischemic etiology and monomorphic VT.(25) Using a threshold of <1.8mV for abnormal endocardium, a modest size scar with a predilection for regions near the basal ventricle in the perivalvular region was seen in all patients. 88% of the 57 mapped VTs originated from these areas. Epicardial mapping was also performed in 3 patients via the coronary sinus, where abnormal electrocardiograms were recorded along the veins. Percutaneous epicardial mapping that was pioneered by Sosa and colleagues (26) has now been incorporated as a regular component of VT mapping and ablation.(27) Soejima et al performed endocardial mapping in 28 patients with nonischemic cardiomyopathy referred for ablation of monomorphic VT. (28) In this report, 63 % of endocardial scars were adjacent to the valve annulus. In the seven patients that underwent epicardial mapping, all demonstrated epicardial scar and in the five, where combined mapping was performed, epicardial scar surface area was larger than endocardial.
In this issue of the Journal, Cano et al. add to the working understanding of the arrhythmogenic substrate in patients with ventricular tachycardia and idiopathic dilated cardiomyopathy. By examining endocardial and epicardial electoanatomic voltage maps in 22 patients referred for epicardial VT ablation over a five year period, the current study represents one of the largest published experiences of substrate characterization in nonischemic cardiomyopathy.
The primary finding of this study is that amongst the 22 patients, one or more epicardial circuits could be identified in 18 patients. Although not statistically significant, low voltage areas were seen in 12/22 (54%) endocardial maps compared to 18/22 (82%) of epicardial maps and scar tended to be larger on the epicardial surface (46.4±37.3 vs. 30.2±43.7 cm2). Ten patients demonstrated abnormalities confined only to the epicardium and only one patient was observed to have more endocardial than epicardial scar. Consistent with their previously published experience, patients with abnormal endocardial maps demonstrated confluent scar in the basal LV, whereas 16 of 18 patients with epicardial scar demonstrated a unique basal LV predilection, most often in the inferolateral LV free wall.
A strength of this study is the examination of reference control subjects to define the electrogram characteristics in the AV groove by distinguishing epicardial fat from scar (low voltage without abnormally wide, split, or late electrogram). As hypothesized, epicardial fat could be differentiated from scar by abnormal electrogram characteristics such as fractionation in addition to low voltage.
Overall, radiofrequency ablation rendered 67% non-inducible at the end of the procedure, with a 71% success rate in those who underwent epicardial ablation of targeted VT. In the remaining patients, ablation was limited by phrenic nerve capture or proximity to a major epicardial coronary vessel. At a mean 18 month follow-up, 71% demonstrated freedom from VT recurrence, where 14 of these 15 patients underwent epicardial ablation.
Several limitations must be noted. It must be emphasized that the cohort population described represents a highly selected group of patients who not only have nonischemic cardiomyopathy, but met criteria for ICD (21/22), had sustained monomorphic VT refractory to at least one antiarrhythmic agent, and failed previous endocardial ablation (20/22), leading to a strong bias towards an epicardial VT circuit location. The mean number of previous ablation attempts was 1.8 (with up to 6 prior attempts in one patient). A recent study of 29 patients with nonischemic cardiomyopathy referred for VT ablation showed that 48% had scar demonstrable by delayed enhancement-MRI.(29) In contrast, only two patients required epicardial ablation for epicardially confined scar in this cohort.
However, this selection and referral bias, which is unavoidable in studies of this type, may be of importance in the bigger picture of risk stratification in nonischemic cardiomyopathy. By characterizing the electroanatomic milieu in those high-risk patients who present with monomorphic VT, we may be able to work backwards in determining decreasing arrhythmic risk. While the present study is heavily biased toward the patients with discrete scars further understanding of the other patients with non-ischemic cardiomyopathy without discrete scars and their respective arrhythmic burden is necessary. Although a patient's candidacy for substrate-guided ablation may diminish between patients with and without scars due to progressively decreasing target areas. Further the baseline risk for ventricular arrhythmias may be similarly diminished.(30) In this study, low amplitude signals were classified into three groups: wide (>80 msec), split (20 msec isoelectic segment) and late (outside QRS). Almost 50% of the abnormal electrograms in the epicardial scar were wide, split or late. The relationship of these potentials to conducting channels or isthmuses was not reported and is not well understood in patients with dilated cardiomyopathy.
An additional limitation lies in the exclusionary nature of classifying patients as “nonischemic” or “idiopathic dilated”. In the current study, nonischemic cardiomyopathy was defined as the absence of coronary stenosis>75%, prior infarction, valvular abnormalities, and known causes of dilated cardiomyopathy, including arrhythmogenic dysplasia/cardiomyopathy. In reality, idiopathic nonischemic cardiomyopathy is likely to represent the residual myocardial substrate from a heterogeneous group of etiologies such as post-viral, toxin-associated (cocaine, alcohol), occult sarcoidosis, diabetes mellitus and long-standing hypertensive heart disease. Additionally, many patients may have a mixed etiology of cardiomyopathy, which can include a component of coronary artery disease. Spasm or embolic infarction can never be ruled out with clinical certainty.
The lack of pathologic correlation in the study population is another limitation. In the study methodology, voltage maps serve as the modality being tested as well as the gold standard. Electroanatomic mapping has not been validated with gross and histopathologic correlation in non-ischemic cardiomyopathy. As the authors state that several of the patients went on for a transplant, confirmatory pathologic data would be invaluable to confirm areas of scar and fat. Further, an MRI would be have also been helpful to correlate these causes of low voltage, although the current standard of care still dictates avoidance from MRI in those with implanted devices.
Although there are many unanswered questions regarding this population of patients, the current study by Cano et al. is an important contribution to our understanding of one end of a disease spectrum. Such questions include: What determines the extent of fibrosis in nonischemic cardiomyopathy? Why do some areas of low voltage have late potentials while others do not? Why is there a predilection for scar on the basal inferolateral LV? Is it possible that circumflex infarctions, which tend to be less evident on 12 lead ECG contribute to basal scars? Does the finding of a perivalvular scar support a postviral etiology of idiopathic dilated cardiomyopathy? A recent study involving MRI in acute myocarditis demonstrated epicardial delayed enhancement in 91% (21/23 patients) with inferolateral midventricular predominance.(31)
Further studies in patients with nonischemic cardiomyopathy are needed to improve our understanding of their electrophysiologic substrate. Such studies which will help us interpret the ‘electrical footprint’ left behind by myocardial diseases that result in cardiomyopathy will provide additional insights for improving our risk prediction and therapeutic strategies.
Supported by the NHLBI (R01HL084261 and RO1HL067647) to KS