Transcatheter Closure of Patent Ductus Arteriosus With Severe Pulmonary Artery Hypertension Through Single Venous Approach: a Propensity Score Match Analysis From Long-term Outcomes

Background: Up to now, transcatheter closure of patent ductus arteriosus(PDA) without severe pulmonary artery hypertension (PAH) has been well established and accepted in clinical practice. However, in PDA patients with severe PAH, transcatheter closure still remains a challenge, regardless of the tremendous advancement in devices and techniques. Methods: We reviewed the records of PDA patients who underwent transcatheter device closure though single femoral venous approach between January 2005 and December 2018. Severe PAH was dened as systolic pulmonary artery pressure (SPAP) measured by a catheter >70mmHg, excluding irreversible severe PAH assessed by trial occlusion. Propensity score matching was used to assemble a cohort of patients with similar baseline characteristics. Results: A total of 29 patients were in severe PAH (SPAH) group after excluding 2 for failing to the trial occlusion, and 468 in non-severe PAH(non-SPAH) group. After matching, 25 patients were in SPAH group, and 39 in non-SPAH group, without any signicant difference in baseline characteristics. The median follow-up periods were 57 and 44 months for groups, respectively. No signi ﬁ cant difference was noted in the rates of cumulative adverse events in the 2 groups (p=0.208). No poor effects in left ventricular function and residual shunt evaluated by transthoracic echocardiography(TTE) in SPAH group were detected in the short- and long-term follow-up. Conclusion: In the patients with PDA associated with reversible severe PAH, compared to non-severe PAH, transcatheter closure through single venous approach has comparable adverse events rates and successful closure rate, with equivalent results of TTE examination at the short- and long-term follow-up.


Introduction
As pioneered transcatheter closure by Portsmann et al. in 1967 1 , the majority of patent ductus arteriosus(PDA) were preferable to accept transcatheter closure for the bene cial outcomes and less invasion 2,3 . The procedure of transcatheter closure usually requires femoral artery puncture, and angiography for duct anatomic and shunting evaluation 4 . However, in several patients without accessible femoral artery, Baykan A et al. had to develop the femoral venous access, with the guidance of transthoracic echocardiography (TTE) and aortogram in the return phase, instead aortic access through femoral artery 5 . Subsequently, accumulating experiences have con rmed that transcatheter closure of PDA though single femoral venous approach is technically feasible, with further simplifying the procedure, and fewer invasions and potential complications 6, 7 .
Up to now, transcatheter PDA closure without severe pulmonary artery hypertension (PAH) has been well established and accepted in clinical practice. However, in PDA patients with severe PAH, transcatheter closure still remains a challenge, regardless of the tremendous advancement in devices and techniques.
Actually, the reversibility of severe PAH effects the management and prognosis. In the short-and intermediate-term of follow-up, several studies had observed partial or complete resolution of PAH after transcatheter closure of PDA patients with severe PAH 8-10 . However, the results of long-term follow-up are unclear. Furthermore previous studies were single cohort of PDA patients with severe-PAH, without the comparison of clinical outcomes between the severe PAH and non-severe PAH in PDA patients treated by transcatheter closure. Thus, we performed a propensity score matched(PSM) analysis to evaluate the safety and e cacy of transcatheter closure of PDA patients with severe PAH though single femoral venous approach from the long-term outcomes.

Patient population
We reviewed the records of PDA patients who had underwent transcatheter device closure though single femoral venous approach at Fujian Medical University Union Hospital, between January 2005 and December 2018. These patients all had clinical and echocardiographic ndings of PDA, and severe PAH was de ned as systolic pulmonary artery pressure (SPAP) measured by a catheter > 70mmHg. The PDA patients with irreversible severe PAH, assessed by trial occlusion, were excluded 10 . Meanwhile the exclusion criteria also contained the unavailable follow-up information(less than one year). The study was approved by the ethics committee of Fujian Medical University Union Hospital and the informed consent was also waived for the retrospective nature(2020KY021).
Patients' records routinely contain demographics, clinical manifestations, patient weight, TTE results, PDA size at its narrowest point, PAH before and immediately after occluder release, device type and device size. The anatomy of PDA was classi ed following Krichenko's classi cation by TTE 11 .
Additionally, TTE was performed to evaluate the degree of the residual shunt (RS) after the device release. A left-to-right shunt without a jet was de ned as trivial RS, with a jet diameter < 2 mm as a small shunt and with a jet > 2 mm in diameter as large shunt 3 . Device protrusion was described as any visible protruding of the device into the descending aorta or pulmonary artery, based on TTE.
Interventional procedure Infants and young children received general anesthesia before operation (usually under 10 years older), and local anesthesia for adults and well-coordinated children. Meanwhile, the procedures were carried out by experienced operators. The TTE was performed to de ne the size and anatomy of PDA before procedure in cardiac catheterization laboratory. 6F radial sheath was placed in the femoral vein before administration of heparin(100U/kg). 6F multipurpose catheter (Cordis) was implanted through the inferior caval vein, right atrium, right ventricle and main pulmonary artery (IVC-RA-RV-PA) to accomplish the complete hemodynamic evaluation. Subsequently, the MP catheter was advanced through the PDA into the descending aorta. With the replacement of guidewire (260cm × 0.035 inch), the MP catheter was exchanged for deliver sheaths. The occluder device in suitable size was selected according to the weight of patients and anatomy of PDA measured by TTE. The occluder device was deployed following the standard protocol but without descending aortography.
For PDA patient with severe PAH, trial occlusion was performed for 30 minutes to record the change of hemodynamic and clinical data after the complete hemodynamic evaluation and anatomy of PDA measured by TTE. The occluder would be released if a fall in pulmonary artery pressure or no elevation, no decrease in aortic pressure, and no worsening of signs and symptoms 10 . Otherwise, the occluding device would be retracted into the delivery sheath for the irreversible PAH. A repeat TTE was executed 10 min after device release to assess the degree of the RS, and to make sure the device in appropriate position without protrusion or iatrogenic aorta coarctation.

Statistical analyses
Continuous variables are expressed as mean ± standard deviation(SD). Chi-square test for categorical variables and t-test for continuous variables were involved in the statistical analyses. P-value < 0.05 was considered statistical signi cance. For the pretty intergroup differences between severe PAH(SPAH) and non-severe PAH(non-SPAH) groups in baseline characteristics (Table 1), propensity similar score matching was performed to identify a cohort patient with comparable baseline characteristics. 1:2 matching protocol without replacement (greedy-matching algorithm) with a caliper width equal to 0.2 of the SD of the logit of the propensity score was used to match. In order to evaluate the discrepancy before and after matching, standardized differences were estimated for all the baseline covariates and < 10% for the given covariate indicate a relatively small imbalance. All datum was analyzed by the SPSS 24.0 software package (SPSS Inc., Chicago, IL, USA).  Table 1) at its narrowest point (Table 1). Hence, we used sex, blood pressure and these above as covariates to match SPAH( severe PAH) and non-SPAH groups. After matching, 25 patients were in SPAH group, and 39 patients in non-SPAH group, without any signi cant difference in baseline characteristics.
Interestingly, the anatomies of PDA between the two groups were comparative no matter before or after matching, including the types of tuber, funnel and window, Meanwhile, the bi-directional shunt in SPAH group was not more prominent before and after matching.
Procedural characteristics SPAP and mean pulmonary artery pressure (MPAP) measured by a catheter before and after occlusion in SPAH group were more higher than non-SPAH group, which was also not balanced by PSM (Table 2). For local anesthesia and occluder size, the SPAH group were more dominant before matching, but after matching, they were equivalent (Table 2). The median follow-up periods were 57 and 44 months for groups, respectively. There was no difference in follow-up time between the SPAH and non-SPAH groups. During this follow-up period, no any device migration, embolization, arrhythmia, hemolysis, vascular access and massive blood loss had occurred. No adverse clinical events between the two groups was detected in the short-term follow-up(less than one month), but with one death respectively between the two groups in the long-term follow-up(more than one year). No signi cant difference was observed in adverse clinical events between the two groups after matching(p = 0.208).
For echocardiograph examination, before matching, SPAH group was more worse than non-SPAH group no matter short-or long-term follow-up. However after matching, no signi cant difference between the two groups was observed, such as LVED, LVES, LA, AO and LVEF in the short-and long-term follow-up.
Before matching, RS was detected in 4 patients in the SPAH group in the short-term follow-up (2 small and 2 large), but at the last follow-up, only 1 trivial RS in the SPAH group. Meanwhile in the non-SPAH group, RS was present in 28 patients in the short-term follow-up (10 trivial, 14 small and 4 large), whereas at the last follow-up, only 2 small RS in the non-SPAH group. After matching, no discrepancy in the rate of RS was detected (Table 3).

Discussion
The propensity score matched analysis was performed to assess the safety and e cacy of transcatheter closure of PDA with severe PAH though single femoral venous approach, and we found that device occlusion in patients with severe PAH (excluding the irreversible severe PAH by temporary occlusion in cath lab) only by femoral venous routine demonstrated comparable adverse events rates and successful closure rate, with equivalent results of TTE examination at the short-and long-term follow-up to patients with non-severe PAH.
Indeed, previous studies had preliminary investigated the e cacy of transcatheter closure of PDA with severe PAH, and indicated that this approach would be safe, During the follow-up period, Zhang et al. also observed left ventricular function usually recovered within a few months. Controversially, ISHWARAPPA et al had found that device embolization in aorta was higher in PDA patients with severe PAH from 6-9 months, which had included adults and infants.
In fact, these previous studies were small size in included patients with limited time of follow-up, and only observed single cohort of PDA patients with severe-PAH from clinical outcomes and TTE results. Moreover, the de nition of severe PAH based upon the mean PAP >45mmHg or >50mmHg. For the inherit detects of previous reports, we included PDA patients with severe PAH upon SPAP > 70mmHg by a catheter, whose mean PAH was 63.73 ± 14.86 mmHg( Table 2). And the time of follow-up in our study including the clinical adverse events and TTE results were pretty longer than previous studies (Table 3). In order to determine the safety and e cacy of PDA patients with severe PAH by transcatheter closure, we compared these severe PAH patients to non-severe PAH patients, which had widely accepted transcatheter closure as the rst-choice treatment of PDA.
What's more, in our present study, we only included PDA patients though single femoral venous approach. As Liu et al. had revealed that replacing the arterial approach by single venous access had the potential to eliminate almost complications associated with arterial puncture, such as occlusion, embolism, dissection, bleeding and pseudoaneurysm formation 4 . If transcatheter closure was an inevitable compromise to PDA, the only we could choose was the comparable effects but with fewer invasions and potential complication, even providing more possibility to helpless patients, just like severe PAH. Hence, we conducted this study to seek the foundation of evidence-based medicine in transcatheter closure of PDA patients with severe PAH only by single venous approach. According to clinical experience, skillful echocardiographer was able to evaluate the blood ow in the aorta and pulmonary arteries and describe the anatomy of PDA. Distinguished with the study by Liu et al. 4 , the single venous approach was completely in the guidance of TTE, without aortogram in the return phase in our study. And we had con rmed the feasible and safety of this method in PDA patients regardless of the level of PAP. Besides the reduction of contrast medium during procedure, an additional positive effect of venous approach was to decrease the time of uoroscopy and hospital stay 4, 12 .
As we all known, the PDA patients with severe PAH were pretty different to non-severe PAH in the baseline characteristics. Severe PAH patients were generally associated with poorer left ventricular function and large PDA, which was also consistent with our study (Table 1). However, after matching with these relevant characteristics between SPAH and non-SPAH groups, they were equivalent. Actually, the severe PAH in our study was selected, only including reversible severe PAH. And the preoperative evaluation criteria of severe PAH reversibility was temporary occlusion by occluder in our medical center, which had been considered more accurate than calculation of pulmonary vascular resistance 10,13 .
In this study, we described the rst comparison between severe PAH and non-severe PAH in PDA patients treated by transcatheter closure. Surprisingly, the effects of transcatheter closure by single venous approach were comparable between the two groups not matter before or after matching, especially with equivalent cardiac function and incidence of death. This would remind us to perform transcatheter assessment rstly, and temporary occlusion if necessary for all PDA patients, which could avoid the reversible severe PAH absent.
Our study had several limitation. First, this research was retrospective, not randomized. Although we had performed PSM to balance the baseline characteristics, the inherent limitation of a retrospective study was still unavoidable. Second, the sample size of severe PAH was limited, and the PAH was not assessed again by catheterization during the period of follow-up. Therefore, further researches with large sample size, prospection and randomization are necessary.

Conclusion
In the patients with PDA associated with reversible severe PAH, compared to non-severe PAH, transcatheter closure through single venous approach has comparable adverse events rates and successful closure rate, with equivalent results of TTE examination at the short-and long-term follow-up.
Transcatheter closure of PDA with severe PAH through single venous approach would be reasonable, and trial occlusion would be necessary to avoid the reversible severe PAH absent. Further research with large sample size is requirement. The study was approved by the ethics committee of Fujian Medical University Union Hospital and the informed consent was also waived for the retrospective nature(2020KY021).

Consent for publication
Not applicable.

Availability data and materials
All data generated or analyzed during this study are included in this published article and in the supplementary information les.

Competing interests
All of the authors declare that there is no con ict of interest in this article.

Funding
This study was mainly supported by Startup Fund for scienti c research, Fujian Medical University(Grant number: 2019QH1054) and partially sponsored by Fujian provincial health technology project(Grant number: 2020QNA035) .

Authors' contributions
Lianglong Chen designed the study and provided administrative support. Danqing Hu and En Chen participated in the data analysis and drafted the manuscript. Wei Cai participated in data collection. All authors read and approved the nal manuscript.