Noninvasive Strategies in Identifying Patients at Risk for Life-Threatening Arrhythmic Events

Takanori Ikeda
Toho University Ohashi Hospital, Tokyo, Japan

The precise identification of patients at risk for sudden cardiac death (SCD) is crucial for the cost-effective application of prophylactic implantation of implanted cardioverter defibrillator therapy (ICD) for improving survival in post-infarction patients. Various noninvasive markers have been proposed as risk stratifiers for SCD. These include left ventricular ejection fraction (LVEF), Holter ECG 9PVC ≥ 10/hr, nonsustained ventricular tachycardia (NSVT), late potentials by signaling ECG,, QT dispersion, autonomic activity, and µV-level T-wave alternans (TWA).

A prospective study to assess the predictive value of 11 noninvasive prognostic variable for arrhythmic events in patients with acute myocardial infarction (MI) was performed by Ikeda and colleagues.

The mean age of the 372 patients was 63 years, 82% male. The infarction site was the anterior wall in 254 patients. Coronary intervention (PCI) was performed in 335 patients and CABG in 17 patients. 41 patients were taking antiarrhythmic drugs (20 pts amiodarone). B blockers were used in 51 patients. Mean LVEF was 51.

Prognostic variables studies were: LVEF < 40% measured by left ventriculography; NSVT on Holter (≥3 consecutive ventricular beats); late potential on signal averaging ECG (LP by SAECG; 2of 3 criteria present: f-QRS, RMS₄₀, LAS₄₀); TWA (µV), age > 70 years, gender, anterior wall MI, PCI, CABG, antiarrhythmic drugs, and beta blockers.

TWA was assessed using a CH2000 system based on the power spectral analysis during supine bicycle exercise. TWA was defined as positive, negative, or indeterminate (during heart rate of 105-10 bpm, sustained alternans ≥1 min with alternans voltage ≥1.9 µV and alternans ratio ≥3.0). All patients underwent noninvasive tests when hemodynamically stable. The mean time of noninvasive sampling was 24 days after acute MI. Noninvasive testing was done in 91% of patients between 2 and 5 weeks after acute MI and in 9 patients before 2 weeks post-MI or 2 months post-MI.

Clinical follow-up was obtained at 2-week and 1-month intervals. During follow-up, restudy of coronary angiography (usually 3-6 months post-MI) was performed in 335 patients who had undergone revascularization. PCI or CABG was performed in patients with significant coronary restenosis.  

The study endpoint was SCD, ventricular fibrillation (VF), and SVT. Nine patients died of non-arrhythmic causes such as pump failure and re-infarction during follow-up and were excluded from data analysis. Therefore, 363 patients were assessed.

Figure 1 shows the incidence of the noninvasive markers in the study population. LVEF < 40% was present in 15% of patients, NSVT positive in 17%, LP 14%, and TWA 35% of patients.

During the mean follow-up of 30 months, 9% of patients (33/363 patients) had an arrhythmic event (SCD 9 patients, VF 5 patients, SVT 19 patients) (Figure 2).

The association between risk factors and endpoints is shown in Figure 3. LVEF, NSVT, LP, TWA, bypass surgery, and use of beta blockers were significantly related to the occurrence of an endpoint on univariate analysis and are independent markers of risk. Multivariate analysis revealed reduction in LVEF and TWA as statistically significant markers for an event. Event-free survival was significantly reduced during follow-up in patients with an LVEF < 40% compared to > 40%, and in patients with TWA, as shown in Figure 4. In patients with an LVEF < 40% and TWA, compared to the absence of both, event-free survival is significantly worse (Figure 5).

Analysis for the predictive value of LVEF and TWA shows that LVEF has a high specificity (90%) and that TWA has a high sensitivity (91%). The combination of LVEF and TWA has a 61% sensitivity and 96% specificity (Figure 6). The positive predictive value for LVEF and TWA is 65%.

Ikeda and colleagues conclude that their data suggest that both a reduced LVEF and TWA are strong risk stratifiers for arrhythmic events after acute MI. Further, TWA could enhance the predictive value of a reduced LVEF, which was introduced by the MADIT II trial as a useful indicator for the prophylactic implantation of ICD in post-MI patients, and hence could contribute to its cost-effective use.

Substrate Mapping and Ablation of Ventricular Tachycardia Late after Myocardial Infarction by Electroanatomical Mapping

Shigeru Ikeguchi
Takeda General Hospital, Kyoto, Japan

Treatment for life-threatening ventricular arrhythmias after myocardial infarction (MI) includes ICD implantation as a first-choice treatment and medical therapy with beta blockers (BB) and amiodarone. However, ablation may be needed to treat frequent arrhythmias.

The recurrence rate of ventricular tachycardia (VT) and ventricular fibrillation (VF) after ICD implantation was 60% rehospitalization at 1 year in the AVID study. At 2 years, the addition of drug therapy, particularly amiodarone, to ICD therapy was needed in 38% of patients with an ejection fraction (EF) < 0.20, 18% of patients with an EF > 0.34, and 30% of patients with an EF of 0.20-0.34. In patients with stable monomorphic VT, mapping for ablation is done during VT pacing. If the post pacing interval (PPI) equals the VT cycle length, the ablation site lies within the reentry circuit.

Soejima and colleagues showed that the incidence of unstable VT in 40 patients referred for catheter ablation after MI was 33%, stable VT was 18%, and both unstable and stable VT was 49%. So, 82% of patients had some degree of unstable VT. The LVEF was 0.29 and the mean number of VT was 3.5 per patient. Unstable VT is characterized as being hemodynamically unstable, where the blood pressure drops suddenly; unstable reentrant circuit, meaning spontaneous change in VT morphology; and non-induciblity during electrophysiologic study.

Non-contact endocardial mapping (ESI) is one mapping approach. Schilling and colleagues reported 24 patients with VT after MI in whom 81 of 97 VTs were mapped with ESI. In 99% of the patients, exits were identified, but presystolic activity was identified in 67%. Ablation was performed for 47% of the VTs, with a success rate about 64%.

Concepts of substrate mapping

The size of the abnormal myocardium is estimated using voltage mapping, with entrainment mapping and pace mapping used to further identify the focus in detail. Isolated delayed potential during sinus rhythm is monitored and then the site is ablated. However, isolated delayed potential is usually within the abnormal myocardium, so voltage mapping may be applied to both methods.

Linear ablation lesions are used for control of unmappable VT using the electroanatomical mapping system. In abnormal myocardium, the substrate map is done using pace mapping and linear ablation lesions created.

Marchlinski and colleagues used a bipolar 4 mm tip electrode to differentiate ischemic and nonischemic myocardium for VT ablation. 95% of the normal electrograms had an amplitude > 1.55 mV. However, in the Carto system, the measurement is independent of the QRS timing. This means that the measurement of maximum amplitude of Egm is performed.  The classification developed by Marchlinski and colleagues is shown in Figure 1.  

Principles for guiding linear lesions can be taken from the work of Marchlinski and colleagues. First, they identified abnormal myocardium and then created linear lesions across the border of the endocardium with abnormal electrogram amplitude. For dense scar and normal myocardium, the lesion was extended from the lowest amplitude signal area (<0.5 mV) to a distinctly normal signal area (>1.5-2.0 mV) or valve continuity. Also, ablation lines were created crossing border zones at sites where pace mapping approximated the QRS morphology of VT.

Limitations of pace mapping include the fact that during sinus rhythm pace mapping does not always correspond to VT QRS, particularly when proximal to the central channel is paced, at which time the potential might be captured in the opposite direction of the outer loop (Figure 2).

Stevenson and Soejima at Brigham and Women's Hospital studied catheter ablation of multiple and unstable VT after MI. For stable VT, entrainment mapping during VT was performed to identify concealed fusion and postpacing interval identical to VT cycle length (reentrant circuit isthmus). For unstable VT, voltage substrate mapping was performed and then pace mapping to identify the lesions and conductive delay.

Soejima developed several principles for guiding linear lines. Select the initial ablation site at the reentrant circuit isthmus or by the pace mapping finding (Figure 3). The linear lesions should be parallel with the border zone with a margin of 1-2 cm. If the mitral valve is within 2-3 cm of the expected lesion, a submitral isthmus ablation line was selected. The identification of the reentry isthmus was associated with better VT control in this study.

Another approach has been described by researchers at Oklahoma University. Usually local ventricular potential in sinus rhythm is recorded corresponding to QRS complex.. But sometimes in MI cases, the potential is delayed, and this may be related to reentry circuit. Their ablation strategy is based on this concept. The sites with delayed potential in sinus rhythm are selected and pace mapping in sinus rhythm or entrainment mapping during VT is not required in their approach. In 16 patients with prior MI and frequent unmappable VT, there were 2-8 (median 4) localized areas or lines of isolated late potential (ILP) per patient. For the ILP areas, ablation was performed 1-12 times (median 5). Per patient, 8-50 ablations (median 25 ) were performed. In 8 patients, irrigation tip was used. The follow-up was 3-62 months, with a median of 11.5 months, and 12 of 16 patients were free of VT, 3 of 4 patients with recurrence were well controlled with antiarrhythmic therapy.

In summary, substrate mapping and ablation are effective to reduce the frequency of unstable VT episodes. The challenge for future research is to determine the most effective approach. What is the best method to guide mapping? Pace mapping during sinus rhythm, entrainment mapping during VT and pace mapping in sinus rhythm, or isolated late potential mapping? Limitations of substrate mapping and ablation of the endocardium include 1) the potential reentry circuit being too deep in the endocardium and thus unmappable or ablatable. 2) For such cases, saline cooled or irrigation tip ablation system can be used. The epicardial approach may be required for epicardial substrate mapping, in which the epicardium is punctured, and ablation in selected patients.

Molecular and Genetic Basis for the Treatment of Long QT-Related Lethal Ventricular Arrhythmias

Minoru Horie
Kyoto University Graduate School of Medicine, Kyoto, Japan

Genetic screening techniques are useful, but time-consuming and costly. EKGs must be used to diagnose these patients, rather than depending on the more expensive genetic diagnostic method. The triggers of the 3 major genotypes of LQTS have been identified: LQT1, exercise and emotional stress; LQT2, auditory stimuli, sleep, bradycardia, and hypokalemia; LQT3, infrequent in Japan, bradycardia and sleep. Patients with LQT2 and LQT3 have a high recurrence rate with beta blocker therapy, and annual mortality for LQT3 is reported to be 58% with beta blocker therapy. Thus, it is critical to identify the genotype of LQTS on clinical findings.

LQTS is caused by distinct mutations in different genes, so the phenotype differs depending on the genotype. In the experimental model, it has been shown that the interval between the peak and the end of the T wave (Tpe) on transmural ECG reflects transmural dispersion of repolarization (TDR), which is amplified by ޏ-adrenergic stimulation in the LQT1 model. Cardiac events are more frequently associated with enhanced adrenergic factors in LQT1.

Horie and colleagues sought to identify the genotype-specific changes in body surface 12-lead ECGs, to determine whether Tpe in 12-leads ECG reflect TDR and whether exercise stress testing can help to differentially diagnose LQT1 and LQT2.

They studied 51 patients with LQT1 and 31 patients with LQT2 and 35 patients in the control group. The parameters of repolarization studied were T wave morphology, QT, QTc, Tpe, and Tpec. The clinical characteristics of the study population are shown in Figure 1. In LQT1, syncope was induced by exercise, and 30 of the patients were symptomatic. In LQT2, 17 of the patients were symptomatic and syncope was induced by sleep, auditory stimulation, and bradycardia. Baseline ECG data showed that QTc and Tpec were longer in the LQT1 and LQT2 groups compared to control (510 ms, 520 ms, 402 ms, respectively; and 142 ms, 195 ms, 99ms, respectively).

The three patterns of T wave in LQT1 are shown in Figure 2. The broad-based T pattern defined as a single and smooth T was seen in 43% of patients, the normal appearing T pattern of a small QT prolongation was seen in 28% of patients, and the late-onset T pattern with a prolonged ST segment was seen in 25% of patients. The T wave patterns in LQT2 are shown in Figure 3.  The broad-based T pattern was seen in 34% of patients, the Bifid T with a small notch in 33%, and the Bifid T with a large notch in 255 of patients.

On exercise, in LQT1 the QTc is lengthened to 590 ms from 452 ms at baseline, and the Tpec is lengthened to 258 ms from 108 ms at baseline. The end of the Q wave is not clearly distinguishable because of this lengthening. In LTQ2, the bifud pattern is clearly seen on exercise. Figure 4 illustrates the change of T wave pattern during exercise in the LQT1 and LQT2 groups.

The ECG data before and during exercise is shown in Figure 5. In LQT1, the QTC is longer on exercise compared to baseline (599 ms vs 511 ms, respectively), whereas there is no change in LQT2. Tpec is longer in LQT1 on exercise compared to baseline 9215 ms vs 142 ms), whereas it is decreased on exercise in LQT2 (163 ms vs 197 ms).

In summary, at baseline, LQT1 and LQT2 have three types of T wave patterns. Broad-based T pattern is seen in 40-50% of LQT1 and 30% of LQT2 patients, so differentiation is not possible. However, exercise stress does differentiate the two genotypes. LQT1 has morphologic changes of the T wave into a broad-based T and there is significant QTx prolongation mainly due to Tpec prolongation with negative Tpe/RR slope. In LQT2, exercise produced a prominent notch on T wave with no significant change in QTc and Tpec.

Horie and colleagues conclude that exercise testing is useful to facilitate genotyping of most common variants of the LQTS. Exercise induces genotype-specific changes in the T wave pattern. Exaggerated prolongation of the QT interval in LQT1 was primarily due to an increase in Tpe, presumably reflecting TDR.