This study was a single-center, retrospective clinical study. Patients with PFO complicated with CS or migraine who received treatment by transcatheter closure of PFO in the First Affiliated Hospital of Nanjing Medical University from September 2019 to March 2021 were selected. Inclusion criteria: (1) Contrast transcranial Doppler (c-TCD) screening was grade 3; (2) PFO with stroke/transient ischemic attack (TIA) or history of migraine not easily controlled by drugs; (3) patients who received transcatheter closure of PFO and regular follow-up after occlusion. Exclusion criteria: (1) stroke etiology determined after investigation (excluded based on TOAST classification (5); (2) definite secondary headache; (3) age <16 years old or >70 years old; (4) other organic heart diseases requiring surgical intervention; (5) abnormal coagulation mechanism, autoimmune diseases, serious heart, liver, kidney failure, etc., who are unable to complete transthoracic echocardiography (TTE), transesophageal echocardiography (TEE), c-TCD examination and interventional occlusion; (6) unwillingness to receive closure treatment and follow-up. The general information of patients and relevant examination results, including sex, age, height, weight, blood pressure, heart rate, and results of TTE, TEE, and c-TCD, were recorded. The study design was approved by the Ethics Committee of the First Affiliated Hospital of Nanjing Medical University.

According to the occluder selected for interventional occlusion, the selected patients were divided into a symmetrical group (occluder size: 24/24 mm; 28/28 mm; 30/30 mm) and an asymmetric group (occluder size: 18/24 mm; 18/25 mm; 22/28 mm; 25/35 mm). The choice of occluder model was determined by the PFO separation width, tunnel length, and shape, as well as ultrasound findings. According to current national and international guidelines, a moderate-sized 18/25 mm occluder is typically selected as the first choice for most PFO. However, for complex PFO, such as those with atrial septal aneurysm (ASA), long tubular PFO, or secondary septal hypertrophy, there is concern that the occluder disc may abrade the aorta, in which case a 25/35 mm or 30/30 mm PFO occluder may be directly chosen.

The patients were subjected to a comprehensive follow-up schedule that included visits at 3 days, 1 month, 3 months, 6 months, and 12 months post-occlusion. TTE, c-TCD, dynamic electrocardiogram and other examinations were to be performed in each follow-up to determine whether there was pericardial effusion or tamponade, occluder embolization or displacement, atrial fibrillation and other related complications. Magnetic resonance imaging (MRI) was performed at 6 months and 12 months after the occlusion, in migraine patients, the MIDAS test was repeated to evaluate the efficacy of occlusion. The second recheck was performed to confirm complete blockage of PFO without residual RLS. Effectiveness was defined as no recurrence of stroke/remission of migraine symptoms and a negative foaming test after occlusion. Safety events were noted as the comprehensive results of serious adverse events related to the occlusion of occlusion or devices (such as pericardial effusion or tamponade, occlusion or displacement of occluder, arrhythmia, etc.).

SPSS 24.0 software was used to process the data. The counting data are expressed as the number and percentage of cases, and the results of continuous variables with a normal distribution of measurement data are expressed as the mean ± standard deviation. Independent sample t tests were used for intergroup comparisons. The data of categorical variables were compared by the chi square test. When the expected frequency of each cell was ≥5 and the total number of cases was ≥40, the Pearson chi square test was used. If the expected frequency of any cell was ≥1 but <5, the continuity correction chi square test was used. When the expected frequency of each cell is less than 1, or the total number of cases is less than 40, Fisher exact probability method is used. P < 0.05 was considered statistically significant.

The type of occluder was chosen based on the admission time, with symmetric occluders applied for patients admitted from September 2019 to June 2020 and asymmetric occluders for those admitted from July 2020 to March 2021. By March 2021, 92 patients had been enrolled, including 45 patients in the symmetrical group and 47 patients in the asymmetric group. Among the 92 patients, 24 were male and 68 were female, and ages ranged from 16 to 70 (41.70 ± 13.67) years. There were 53 patients of migraine (57.61%) and 39 patients of cryptogenic stroke (42.39%). Two sets of baseline characteristics are described in Table 1.

The transthoracic echocardiography results of the two groups were as follows (Table 2): There was no significant difference in aortic root diameter, left atrial diameter, left ventricular diastolic diameter, left ventricular systolic diameter, ventricular septal thickness, left ventricular posterior wall thickness, right atrial diameter, right ventricular diastolic diameter, left ventricular shortening fraction or left ventricular ejection fraction between the symmetrical group and the asymmetric group (P > 0.05). There were 21 cases (46.67%) of atrial septal aneurysm in the symmetrical group and 18 cases (38.30%) in the asymmetrical group. There was no significant difference in the results (P > 0.05).

Table 3 shows the results of PFO morphology under transesophageal echocardiography in both groups, with no significant differences in PFO separation width, tunnel length, ASA, proportion of secondary septal hypertrophy, or complex PFO between symmetric and asymmetric patients (P > 0.05). Occluder model selection was made based on ultrasound results, with 25/35 mm or 30/30 mm occluders selected for complex PFO (4).

In the symmetrical group, 13 cases (28.89%) were 24/24 mm, 11 cases (24.44%) were 28/28 mm, and 21 cases (46.67%) were 30/30 mm. In the asymmetric group, occluder devices were selected as follows: 18/24 mm in 5 cases (10.64%), 18/25 mm in 19 cases (48.94%), 22/28 mm in 5 cases (10.64%), and 25/35 mm in 18 cases (38.30%). There was no significant difference in the diameter of the right disc between the symmetrical group and the asymmetrical group [(27.78 ± 2.57 mm) vs. (29.07 ± 4.90 mm), P = 0.121], and there was a significant difference in the diameter of the left disc [(27.78 ± 2.57 mm) vs. (21.09 ± 3.36 mm), P < 0.001].

There was no significant difference in the operation x-ray exposure time or hospitalization time between the symmetrical group and the asymmetric group (exposure times 3.51 ± 1.70 min and 3.48 ± 3.13 min, respectively, P = 0.966; and hospitalization times 4.02 ± 0.54 d and 4.07 ± 0.44 d, respectively, P = 0.680). No occlusive complications, such as pericardial effusion or tamponade, occluder embolism or displacement, during or 24 h after occlusion in the symmetrical group or the asymmetric group. No new serious arrhythmia was found on the dynamic electrocardiogram after occlusion. Reexamination with c-TCD within 3 days after the occlusion showed that there were still large RLS (in 21 patients from the symmetrical group and 11 patients from the asymmetric group), statistical analysis showed that the difference between the two groups was statistically significant (χ² = 5.484, P < 0.05) (Supplementary Table S1); there were 11 patients who were negative for RLS in the symmetric group and 20 such patients in the asymmetric group, statistical analysis showed that there was no significant difference between the two groups (χ² = 1.056, P > 0.05) (Table 4).

The patients were followed up until March 2022, with a median follow-up of 12 months. The occluder position of all patients was good, and there were no complications, such as occluder displacement, hematogenous infection, thrombosis, and pericardial effusion.

One month after occlusion, 88 patients underwent c-TCD reexamination in the outpatient department, and 4 patients were lost to follow-up. Fourteen patients in the symmetrical group still had large RLS, and 6 patients in the asymmetric group still had large RLS. Statistical analysis showed that the difference between the two groups was statistically significant (χ² = 5.146, P < 0.05) (Supplementary Table S1). There were 16 patients who were negative for RLS in the symmetric group and 30 such patients in the asymmetric group. Statistical analysis showed that the difference between the two groups was statistically significant (χ² = 6.473, P < 0.05) (Table 4).

Three months after occlusion, 82 patients completed c-TCD follow-up, and 6 patients were lost to follow-up. Three patients in the symmetrical group still had large RLS, and no patient in the asymmetrical group had large RLS (P > 0.05) (Supplementary Table S1). There were 18 patients who were negative for RLS in the symmetric group and 32 such in the asymmetric group. Statistical analysis showed that the difference between the two groups was statistically significant (χ² = 4.305, P < 0.05) (Table 4).

Six months after occlusion, 76 patients completed c-TCD follow-up, 6 patients were lost to follow-up. There were 17 patients who were negative for RLS in the symmetric group and 30 such patients in the asymmetric group. Statistical analysis showed that the difference between the two groups was statistically significant (χ² = 4.842, P < 0.05) (Table 4). At 12 months after occlusion, 74 patients completed c-TCD follow-up, 2 patients were lost to follow-up. There were 25 patients who were negative for RLS in the symmetric group and 35 such patients in the asymmetric group, the remaining 14 patients had moderate/mild shunts. Statistical analysis showed that the difference between the two groups was statistically significant (χ² = 4.034, P < 0.05) (Table 4).

During the follow-up period, no safety events occurred in either group, and there was no symptom recurrence in either group. There was no significant difference in the MIDAS scores between the symmetric group and the asymmetric group before occlusion (P = 0.528) (Table 5). The MIDAS scores of the symmetric group were significantly higher than those of the asymmetric group during 6 and 12 months follow-up after occlusion (P = 0.038; P = 0.049) (Table 5).