We collected 30 patients diagnosed with PFO and admitted to the Department of Congenital Heart Disease Center of the Wuhan Asia Heart Hospital of Wuhan University of Science and Technology from August 2019 to December 2020. The enrollment criteria were as follows: (1) PFO diagnosed with a medium-large right-to-left shunt (RLS) by clinical imaging; (2) age ≥ 16 years; (3) voluntarily participate in human clinical trials with signed informed consent. (4) occult ischemic stroke, migraine, dizziness, or other recurrent neurological and psychiatric symptoms. The exclusion criteria were as follows: (1) cerebral embolism of any cause can be found; (2) contraindications to antiplatelet or anticoagulant therapy; (3) inferior vena cava or pelvic vein thrombosis, systemic or local infection, sepsis, intracardiac thrombosis; (4) pregnancy; (5) combined pulmonary hypertension or PFO as a survival channel; (6) massive cerebral infarction within four weeks; (7) migraine of other known causes.

Ethical approval was granted by the local ethics committee of Wuhan Asia Heart Hospital of Wuhan University of Science and Technology (2019-YXGCP-012), and all study participants provided informed consent. This study was registered in the Chinese Clinical Trial Registry (Registration date: 23/06/2019; Number: ChiCTR1900024036).

The biodegradable occluder, made from a polydioxanone (PDO) single filament of 0.298 mm diameter and a polyethylene terephthalate matrix membrane, consists of a blocking umbrella disc, interlayer membrane, and suture (Fig. 1). This device exhibited great structural stability owning to many cross points to resist external force [9] and can be fully compressed and collected into the catheter owning to good shape memory ability. By minimally invasive interventional technology, the biodegradable occluder is transferred to the PFO through the delivery device under the guidance of ultrasound and digital subtraction angiography imaging equipment. Once planted, the occluder will close the PFO tunnel and provide a temporary scaffold for the tissue around to cover the device and the defect underneath, which eventually degrade into nontoxic ingredients (water and carbon dioxide) and be fully absorbed, leaving “native” tissue behind [10].

The diagnosis of PFO was mainly through transthoracic echocardiography (TTE)/contrast-transthoracic echocardiography (cTTE), transesophageal echocardiography (TEE)/contrast-transesophageal echocardiography (cTEE) and contrast transcranial Doppler (cTCD). To achieve maximal accuracy in PFO diagnosis, combining different techniques is warranted [11]. TTE is usually used for an initial evaluation of patients for a possible PFO. We used two-dimensional (2D) echocardiography to scan the anatomical structure of the atrial septum and color Doppler to observe the left-to-right shunt on the parasternal aortic short-axis view, the apical 4-chamber view, or the apical 5-chamber view. As adults are usually affected by various factors such as obesity, excessive lung gas, etc., the detection rate of PFO by TTE is not very high, and it is difficult to measure the size of PFO accurately. TEE provides a clear visualization of the atrial septal anatomy and is the “gold standard” and preferred method for the diagnosis of PFO [12]. Given that TEE is a semi-invasive test, it is not suitable for screening purposes but is used for accurate reassessment before the release of the occluder during the operation. In our study, RLS was mainly assessed using cTCD, and the number of micro-embolic signals determined its classification observed within 10–20 s: grade 0, no microbubbles, no RLS; grade I, 1–10 microbubbles, a small RLS; grade II, 11–25 microbubbles, a moderate RLS; grade III, >25 microbubbles, a large RLS, and the signals could be in the shape of a rain curtain or shower [13].

The risk of paradoxical embolism (RoPE) score, which can evaluate the role of PFO in the etiology of ischemic stroke, was used to predict the probability of PFO-related stroke [14] and the assessment of PFO occlusion before the operation. The RoPE score was calculated as follows: no history of diabetes (1 point); no history of hypertension (1 point); non-smoking (1 point); no history of stroke or transient ischemic attack (TIA) (1 point); cortical infarct on imaging (1 point); an age of 18–29 (5 points), 30–39 (4 points), 40–49 (3 points), 50–59 (2 points), 60–69 (1 point), and ≥70 years (0 points). Given a higher RoPE score indicating a stronger association between PFO and stroke, the score obtained over 6 (population attributable fraction: 62%, 95% CI: 54%–68%) is more likely to reflect a PFO-related stroke [15], indicating possible benefit from PFO closure. The original questionnaire is provided in Supplementary Table 1.

The Headache Impact Test-6 (HIT-6), consisting of six items: pain, social functioning, role functioning, vitality, cognitive functioning, and psychological distress [16], measures the intensity and frequency of headaches within the previous four weeks to assess the impact of headaches especially migraine on the patients’ life [17]. The patients answered each of the six related questions using one of the following five responses: “never”, “rarely”, “sometimes”, “very often”, or “always”. The HIT-6 score (ranged from 36 to 78 points) was divided into four grades: (1) little-to-no impact (HIT-6 score: 36–49), moderate impact (HIT-6 score: 50–55), substantial impact (HIT-6 score: 56–59), and severe impact (HIT-6 score: 60–78), where a higher score indicates a more significant impact of headache on the daily life of the respondent. The headache impact test scale is provided in Supplementary Table 2.

The implantation of the biodegradable PFO occluders referred to the European Society of Cardiology Guidelines for managing adult congenital heart disease [18]. Prophylactic antibiotics can be given 1 h before the operation. Occluders implantation was guided under local anesthesia for X-ray and TTE monitoring or general anesthesia for TEE monitoring alone. For the blocking process, we routinely punctured the femoral vein as the occluder implantation pathway. After an intravenous injection of heparin (100 u/kg) was given, a right cardiac catheter was passed through the PFO, and an exchange guidewire were advanced and placed in left superior pulmonary vein. After immersing in heparinized saline, the occluder was delivered to the left atrium through a delivery sheath and released under the guidance of X-ray and TTE monitoring or simple TEE. First, we released distal umbrella in the left atrium, then we withdraw the occluder device and sheath until the distal umbrella was opposed to the left atrial septum. The right umbrella disc was then released by further withdrawal of the sheath. A descending traction test was conducted to ascertain the steadiness and location under color doppler ultrasonography to observe the morphology and location of the occluder, after which the delivery system was activated to release the implant.

After the operation, patients received routine heparin anticoagulation for 48 h, oral aspirin 100 mg/d for six months, and clopidogrel 75 mg/d for one month.

Patients diagnosed with PFO received intraoperative delivery and release of biodegradable occluder under the guidance of X-ray and TTE or TEE alone. An immediate efficacy evaluation was conducted after occlusion to secure the occluder localization. We evaluated the morphology and location of the occluder and the presence of a residual shunt by TTE at 48 h, 1, 3, 6, and 12 months after the procedure. Residual shunts were detected postoperatively using cTCD 1, 3, 6, and 12 months.

The endpoint of this study was Clinical follow-up at 12 months after occlusion. The incidence of complications during follow-up at 48 h, 1, 3, 6, and 12 months after occlusion was used as the safety evaluation index. The success rate of device release in the immediate postoperative period, the success rate of blocking at each of the times above points, the frequency of migraine and RLS decreased or not at each of the times above points, and the overall success rate at 12 months after occlusion was used as the evaluation indices for efficacy.

All analyses were performed by using R software (The R Foundation, http://www.r-project.org, version 4.2.0) and EmpowerStats (http://www.empowerstats.com, X&Y Solutions, Inc., Boston, MA, USA). Continuous variables conforming to the non-normal distribution were expressed as Median (IQR) for representation, and pairwise comparisons were made using the Mann–Whitney U test. Categorical data were expressed as frequencies and percentages, and comparisons between groups were performed using the χ2 test. Two-sided p < 0.05 indicated that the difference was statistically significant.

The 30 patients (10 males and 20 females) received biodegradable PFO occluder, aged 41.5 (28.0–49.8) years. Twenty-five patients (83.3%) had an ischemic stroke; among these, 7 patients (23.3%) had both cortical infarcts on imaging. The RoPE scores of the 7 patients were 6.0 (6.0–7.0) points. Eighteen patients (60.0%) had migraine, and 9 (30.0%) patients had dizziness. Under the guidance of TEE during the operation, the PFO size was 1.4 (1.0–2.0) mm, the PFO length was 11.0 (8.0–15.0) mm, the secondary septal thickness of the patients was 4.0 (3.0–4.0) mm, and 3 patients (10.0%) got atrial septal aneurysm (ASA). The maximum basal diameter of ASA in the 3 patients was 22 mm. All patients had no secondary atrial septal defect, long inferior vena cava valve, or secondary hypertrophic septum. Other medical conditions included hypertension (n = 7, 23.3%), diabetes mellitus (n = 8, 26.7%), and smoking history (n = 11, 36.7%). The procedure had a 100% success rate, and null of the patients developed perioperative complications (Table 1).

All those 30 patients took oral aspirin 100 mg/d for six months and clopidogrel 75 mg/d for one month after the operation without any bleeding complications. During follow-up for one year, according to the trial procedures, none of the patients got lost to follow-up. In clinical follow-ups at 1, 3, 6, and 12 months after the operation, there was no occluder displacement or loss, thrombosis of the occluder, new arrhythmias, or recurrent stroke. Compared to pre-occlusion levels, no clinically obvious abnormalities were found in routine blood and urine chemistry panels, hematology, liver and kidney function tests, coagulation function, or Electrocardiograms (ECG).

HIT-6 scores of patients with migraine were 57.0 (54.5–61.5), 57.0 (54.0–60.0), 55.0 (50.0–56.0), 50.0 (50.0–55.0), 46.0 (46.0–50.0) before the operation, 1, 3, 6, and 12 months after the operation separately. The scores measured at 1, 3, 6 and 12 months all indicated significant differences compared to the scores measured before the operation (p < 0.05) (Fig. 2). Headache symptoms in all patients were significantly relieved compared with those before the procedure.

Furthermore, all patients underwent cTCD to screen for RLS at 1, 3, 6, and 12 months postoperatively, both at a resting state and a Valsalva maneuver. The proportion of the follow-up cTCD grading III results at the Valsalva maneuver were 96.7%, 40.0%, 16.7%, 26.7%, and 0% separately as follow-up progressed (Fig. 3), indicating the overall success rate of occlusion 12 months after the operation was 100%. Looking at Fig. 3, it is apparent that the proportion of grade 2 and grade 3 measured at a Valsalva maneuver at 6 months accounted for more than one-third of the cohort. In this subgroup, stratified by the cTCD grading results at six months ≥2 or not, the data of patients’ baseline characteristics were analyzed, and no statistical differences were found (Table 2).

The results of follow-up imaging were as follows. We could easily detect the morphology in the former 3 months. About three months after occlusion, the umbrella disc framework degraded slowly from the edge. The evenness and clarity of the occluders showed in TTE images were obviously less than those at 1 day. The architecture of the occluders were mostly degraded at about six months with the outline of the occluders barely visible. Twelve months after occlusion, the hyperechoic area became significantly smaller in the middle of the atrial septum and an echocardiogram demonstrated an intact atrial septum without residual shunts (Fig. 4).