Patient and accessory pathway characteristics
During the study period (2015-2019), a total of 304 accessory AP ablation were performed. 281 of these procedures were initial ablations (Figure 1). 22 patients underwent repeat ablation which were all successful without long term recurrence except for one, for whom a third procedure was performed successfully.
The distribution of pathway locations was not uniform (Figure 2). Two groups together accounted for 83.2% of cases (LFW: 61.6% and PS: 24.6%). LFW pathways were less likely to be manifested (31.8%) compared to pathways from other locations (p=0.004; Table 1). RFW and AS pathways were relatively uncommon comprising 9.6%, and 4.3% of all cases respectively.
In our population, 80.8% of patients undergoing RF ablation had demonstrated atrioventricular reciprocating tachycardia (AVRT). Multiple accessory pathways were reported in 3.6% of cases (Table 1).
Patient characteristics such as age, sex, and left ventricular function were not statistically different among the four AP groups (Table 1).
Known major complications associated with ablations such as cerebral vascular events, cardiac perforation, tamponade, myocardial infarction or death did not occur. Two cases were complicated by AV block requiring pacemaker implants (one from LPL group and one from RPS group). One patient had a femoral arterial pseudoaneurysm (LPL group).
Initial procedure acute outcome
For left atrial and ventricular access, our initial approach was generally retrograde aortic (196 of 199 cases, 98.5%), with 173 of these cases being LFW pathway ablations. Overall use of open-irrigated catheters and 3D mapping during initial ablation was 11% and 3.9 %, which was much lower than during repeat ablations (36.4% and 31.8%, respectively). More open-irrigated catheters were used in the PS group (21.7%) with 7/15 of these cases being inside the coronary sinus system (Table 1).
Acute success was achieved in 94.7% of all initial procedures (LFW: 97.1%; PS: 95.7%; RFW: 87%; and AS: 75%) (Table 1). Median fluoroscopy time was highest in AS group (33 ± 30 min) and lowest in LFW group (15 ± 19 min). Longest ablation duration times were in AS (422 ± 318s) and RFW (519 ± 539s) groups (Table 1 and Figure 3). Total case time, as defined by time from patient arrival to exit from the EP laboratory, was longest in AS group (203 ± 77 min) and shortest in LFW group (144 ± 68 min) (Table 1 and Figure 3). Significant number of outliers were seen in each group (Figure 3).
The number and proportion of cases with prolonged procedure time parameters (fluoroscopy, ablation, and case times) for all four groups and specific locations within the groups were tabulated (Table 3). The following times were defined as prolonged: procedure time ³ 200 min, fluoroscopy time ³ 30 min, and ablation time ³ 400 sec. Anatomic locations that had a greater than 50% of cases with prolonged time parameters were highlighted. These locations consisted of AS, LAL, Epi-CS, and RAL. Common reasons cited for prolonged times were, (i) limited mapping capability, (ii) limited lesion formation, (iii) difficult ablation catheter navigation, (iv) multiple chambers mapped, and (v) incidental secondary arrhythmias requiring ablation.
Long term outcome of initial ablation procedures
With median follow up of 931 days post ablation in patients who had initial successful ablations, 93.4% remained free from recurrence of SVT or ventricular pre-excitation (Figure 4a). Long term success was lower at 75.0% for AS pathways compared with 98.2% for LFW pathways (p=0.002 log-rank test; Figure 4b). RFW and PS pathways have similar long-term outcome (90.0% and 90.8% respectively, Figure 4b). There were no statistical differences by age, sex, or ventricular pre-excitation in acute or long-term outcome (data not shown).
ECG predictors for specific pathway locations of interest
An ECG algorithm was developed based on the following rationale. Sets of leads represent planes of axis through the heart. Anterior-posterior axis is represented by the precordial leads V1-V6. Pathways which are located most anteriorly such as RAL, transitions the latest, generally with R/S = 1 at or after V3; whereas pathway farthest away from the chest wall such as LFW leads display R>S by V1 (Figure 5a). The superior-inferior axis is revealed by inferior leads II, aVF, and III. As the pathway location transitions from an inferior to a superior location, delta wave in lead II first becomes positive followed by lead aVF and then by lead III (Figure 5b). Left-right axis is represented by lateral leads I and aVL. As the pathway location transitions from right to left, lead aVL first becomes negative, then lead I (Figure 5c).
This ECG algorithm was tested on 116 ECGs with ventricular pre-excitation representing all four groups. Of special interest were specific locations within groups (AS, LAL, Epi-CS, RAL) characterized as being more difficult to ablate based on prolonged procedure time parameters (Table 3). The algorithm was validated at each pathway location for sensitivity, specificity, and negative predictive value (Figure 5d). The positive predictive value (PPV) was not utilized as it is subject to wide variations at locations with only a small number of pathways relative to adjacent locations that may have overlap in ECG characteristics. There is significant overlap with AS and RAL pathways, as precordial transition pattern does not differentiate clearly in the anterior portion of the RA shared by both pathways.
Sensitivity and specificity varied but were generally high at most individual locations. Of note, both the sensitivity and negative predictive value for AS, LAL, Epi-CS, and RAL locations were all equal to 1 (Figure 5d). There were no false negatives, suggesting that if an ECG did not display characteristics of AS, LAL, Epi-CS, and RAL pathways as described, there is extremely high likelihood (100%) that the pathway would not be localized to these specific sites. ECGs with baseline preexcitation patterns from AS, LAL, Epi-CS, and RAL pathways demonstrate characteristic patterns (Figure 6).