Results:
Patient Disposition : The study enrolled 404 patients, of which 5 were consent ineligible, 63 were ‘roll-in’ pts, and 336 non roll-in. Of the non roll-in pts there were 10 intention to treat (enrolled after study met treatment goals), 1 attempt to treat, and 325 treatment subjects (Supplemental Figure 1). A PV stenosis sub-study enrolled 55 pts, of which 50 contributed to the PV stenosis analysis. These patients underwent MRI or CTA to measure pulmonary veins size pre procedure and then a repeat study 3 months after ablation. Additionally, an extension arm enrolled 54 pts to examine the performance of the variable size POLARx FIT catheter, 0f these 1 was consent ineligible, 3 were Intention to treat, and 50 received treatment with the 31mm inflation size in at least one PV.
Patient Demographics: Characterization of patient demographics is provided in Table 1. Briefly, patients in the primary cohort were 62 ± 11 yrs old on average, 38.2% (124/325) were females, average BMI was 29. Patients had an average CHADS2-VA2SC score of 1.7 ± 1.3, LVEF of 58.6% ± 5.8%. Patients were enrolled an average of 3.0 yrs post initial diagnosis, and an average of 87 days since their most recent episode. Similar demographics were seen in the FIT extension arm, though patients had a shorter time since AF diagnosis (Mean 1.6yrs; Median 0.4, IQR 0.2, 1.0) and higher CHADS2-VA2SC scores (2.2).
Procedural performance: Recommended dosing strategy was based on TTI timing (See Supplement), strategy was not followed in 206/325 pts. General anesthesia was used in a majority of pts (255/325), conscious sedation was used in (70/325), esophageal temperature monitoring was used in 311/325 cases. 3D mapping was used in 184/325 cases. Average procedure, LA dwell, and Fluoroscopy times were 91, 59, and 13 minutes, respectively. Acute success was achieved in 311 cases (95.7%). Fourteen subjects (14) were acute procedural failures due to not achieving entrance block and/or exit block in one or more pulmonary veins, usage of a non-study catheter or not performing the entrance and exit block testing. In the FIT extension arm general anesthesia was used in all patients, 3D mapping had higher utilization (98%), and fluoroscopy time was shorter (7 min). The 31 mm balloon size was used in approximately 64% of ablations. Additionally, in the FIT arm acute success was achieved in all 50 patients for a rate of 100% with a 92.9% lower confidence interval.
Biophysical data: Biophysics data is reported in Table 2. Complete PVI, confirmed with exit and entrance block, was achieved in 95.7% of patients. On average PVs received 1.8 ± 1.3 ablations (IQR 1-2), with a total duration of 4.46 ± 2.7 min (IQR 3-5 min) per PV. Occlusion score of 3 or 4 was reported in 95.9% of PVs. Overall single shot success was achieved in 55.9% of PVs (726/1299), in cryoablations longer than 60s the single-shot rate was 60.1%. The mean nadir temperature for ablations was 56.4 ± 6.9°C. TTI was available from 1044 ablations and averaged 45.1 ± 25.3 s. Biophysical and performance data by PV are reported in Table 2. Biophysical data from the 50 patient FIT extension arm are also reported in Table 2, however as usage of the larger 31mm balloon configuration was at operator discretion, it is challenging to directly compare data. Though, it is notable that nadir temperatures, and time-to-isolation appear relatively similar. See Figure 1 for visual comparison aggregate freeze curves from all ‘first ablation’ uses of the 28mm and 31mm balloon sizes.
Safety: The primary safety endpoint was a composite of multiple acute and chronic safety events. Safety events occurred in 4.0% of patients, resulting in a KM freedom from occurrence rate of 96.0%, which met the study safety goal of 89.0% (Figure 2). One death occurred in a treatment patient during the course of the study at approximately 6 months post ablation due to a non-healing foot wound but was adjudicated to be unrelated to the study device or the procedure. Primary safety events included: 5 vascular access complications, 3 gastroparesis, 2 tamponade, 1 air embolism, 1 pulmonary edema, and 1 myocardial infarction. In addition to prespecified endpoints, 4 patients had phrenic nerve palsy (temporary phrenic nerve impairment present at the end of procedure), all of which resolved by the end of the study. See Figure 1 for list of study safety endpoint events.
PV Stenosis Sub study: A subset of 50 patients in the primary cohort enrolled in an MRI sub-study to examine occurrence of PV stenosis. Core lab analysis of 3 and 6 month MRI (1.5 or 3T) images revealed mild stenosis in the LIPV of 2 patients, and no instances of moderate or severe stenosis.