OUTCOMES AND DETERMINANT FACTORS OF PERCUTANEOUS TRANSCATHETER OCCLUSION AMONG CHILDREN WITH PATENT DUCTUS ARTERIOSUS IN THE EAST COAST MALAYSIA: A 10 YEAR EXPERIENCE

survival analysis was applied to determine median time for diagnosis and transcatheter closure of PDA. Multiple Cox regression analysis was used to identify predictors for early intervention of PDA closure. Results: A total of 123 patients were recruited with 65.9% of them were predominantly female. The age at the time of diagnosis was between 4-day to 14.5-year-old and median age for undergoing the procedure was 2.6-year-old (range:3-month to 15.8-year-old). The PDA size ranged between 0.2-12mm with 59% of the patients had large size PDA. Successful closure rate was achieved in 96.2 % with 3.8 % having very minimal residual shunt at 2 years post procedure. Complications were noted in 17 patients (4.0% major, 9.8% minor). Overall, the median time to diagnose and to close PDA were 6 months and 16.9 months respectively. Earlier intervention was likely to be performed in younger children (p<0.001) with weight of less than 10 kg (p<0.001). Conclusions : Transcatheter device closure was safe and effective in children with PDA with excellent outcome and minimal complication. Age and weight were the determinant factors for early PDA closure.


Introduction
Patent ductus arteriosus (PDA) is a common congenital heart disease, and it represents about 10-15% of all congenital heart lesions with the incidence of 1 in 2500-5000 live births [1,2,3]. It is an important foetal structure that closes spontaneously in about 90% of full-term infants during the initial 48 hours of life. However, the persistent patency of the ductus arteriosus can manifest as an asymptomatic murmur or present with broad spectrum of clinical heart failure symptoms depending on the size, magnitude of the shunt and pulmonary vascular resistance [2,3,4]. In premature infants, untreated PDA is a significant precursor for mortality and morbidity whilst in bigger children and adult populations, it can lead to congestive heart failure [5]. PDA can be categorized according to the angiographic illustration using Krichenko In Malaysia, the first reported transcatheter closure of PDA was in performed in 1994 at National Heart Institute (IJN), Kuala Lumpur using Gianturco coil [13]. Later, in 1997, the Amplatzer ductal occluder was used in the transcatheter approach [14]. Since then, multiple type of devices was introduced and the other paediatric cardiac centres nationally had started to perform the procedure. Currently, besides IJN, the procedure is available in 5 other centres such as Penang, Selangor, Sabah, Sarawak and Kelantan.
Even though the transcatheter approach had shown favourable outcomes, the major pitfalls were the complexity of the delivery systems especially in infancy group, the higher incidence of residual leakage post-procedure, and the risk of device embolization and protrusion [9,12,15]. However, with the evolving modern techniques, the efficacy and safety profile of this approach has been documented in older infants, children, and adults [4,11,16].
The aim of this study was to describe the clinical characteristics and determine the outcomes of transcatheter PDAs device closure. This includes the median time for PDAs closure from the time of initial diagnosis and the predictive factors for early intervention among PDA patients in our institution. Patients who were registered from January 2007 to January 2017, age from 0 to 18-year-old, underwent transcatheter PDA device occlusion at Hospital USM were included. Children who had had complex congenital heart lesions and underwent additional procedures beyond PDA closure on the same settings, were excluded. Any records with more than 30% of missing data were also excluded. Information on demographic data (age, gender, race), clinical data such as age of diagnosis, type of procedures, angiographic findings during the procedure, hemodynamic data as well as the outcomes during follow-up were analysed.

Methods
A successful transcatheter PDAs device closure is defined as a state when a patient left the cardiac catheterization laboratory with a coil or an occluder, with minimal residual flow detected by the echocardiography. Whereas failure of closure is defined as a state when children who undergone the attempt of transcatheter closure but failed to occlude the defect due to larger defect size, inappropriateness of patient clinical status for device closure, unavailability of a bigger device size, or device dislodgement during the procedure. The size of PDA was confirmed using the measurement at the narrowest diameter of the pulmonary end on the angiogram during the procedure. This was then further classified as small (diameter <1.5mm), moderate (diameter 1.5-3mm) and large (diameter > 3mm) [17]. Residual flow was measured immediately after device closure and during the regular follow-up using echocardiography performed by a single cardiologist.
This residual flow was further subdivided into trivial (residual shunt <1mm), small (residual shunt 1-2mm) and large (residual shunt > 2 mm) [18]. Whereas pulmonary hypertension is defined as mean pulmonary arterial pressure above 20mmHg measured during the procedure [19]. implantation, embolization of device, significant residual shunt, significant obstruction of aortic arch or left pulmonary artery, infective endocarditis, and massive blood loss [8]. Whereas minor complications ae defined as complications that spontaneously resolved without requiring any urgent intervention or treatment (Level I-II) e.g., mild narrowing of left pulmonary artery or aorta and transient weak arterial pulse [8].
The data were analysed using the Statistical Package for the Social Science (SPSS) version 26 (SPSS, Inc, Chicago, IL). Descriptive statistics were employed for sociodemographic variables in which the categorical data were reported as number and percentage while numerical data were reported as mean and standard deviation (SD) or median and interquartile range (IQR). Outcome groups were compared in terms of success rate and residual shunts using Pearson Chisquare test and were presented in cross-tabulation tables. All documented complications were tabulated, and frequency and percentage were calculated by attributability. The median survival time to close the PDAs was measured from time of diagnosis to time of PDAs closure, analysed using Kaplan-Meier survival analysis and predictive factors were compared using simple and multiple Cox regression analysis. A p-value <0.05 was considered statistically significant.

Results
A total of 123 patients with PDA were recruited with the majority of them were female (65.9%), median weight of 10.8 (IQR 13) kg and the lowest weight of 3.0 kg during the procedure. The age at the diagnosis were between 4-day-old to 14.5-year-old with the median age of 6-month-old. The youngest patient underwent the procedure was 3-month-old and the eldest in this group was 15.8-year-old with the median age at procedure was 2.6 (IQR 6.5) -yearold as illustrated in Table 1 Devices were successfully deployed in 95.1% of PDA patients and all the failure patients were in occluder group, there was statistically difference as compared to the coil group (p=0.035). The total occlusion rate increased from 79.3% at 24 hours post procedure (prior to discharged) to 89.6% and 93.0% after 3 to 6month follow-up then achieved 96.3% total occlusion rate by 2 years follow up. The remaining 3.8% of them having only trivial residual shunt which did not require further intervention. The analysis showed there was more residual seen in coil group compared to occluder group throughout the followup, but this was statistically not significant ( Table 2). Seventeen patients developed complications during or after the procedure, attributed by 5 major and 12 minor complications. There were no statistically significance between the two devices selection (p= 0.593).
Major complications were mainly due to device embolization and displacement (n=3), thrombosis (n=1) and anaphylactic shock towards contrast (n=1). For minor complications, 2 patients developed haematoma at puncture site, 3 had mild narrowing pulmonary artery and 3 had mild coarctation of aorta and the remaining were due to bleeding (n=1) and mild allergic reaction (n=2).     [9,20]. In our study, majority had moderate to large sized PDA and hence the frequency of occluder device used was high. Nevertheless, based on our experience, the other factors such as cost differences, PDA's type and suitability of different anatomy were also the determining factors in the choice of devices used. The final outcome for both coil and occluder, used in PDA, were illustrated in Figure 2, which showed aortogram pre and post placement of the devices.
In our study, the success occlusion rate of 96.3% was comparable to the international figures [1,2,9,12,21]. This was not surprising since transcatheter approach has been an established procedure for the past 50 years with the evolution of various devices and techniques, as well as cumulative experiences among the practicing interventional cardiologists. There were only 4 patients who had residual shunt of less than 1 mm by the end of our follow-up. These had occurred mostly among those who used coil as compared to the occluder. These findings were again similar to previous studies which demonstrated that occluder devices showed less residual shunt and better occlusion rate of 99.7% to 100% compared to coil devices which ranged between 80.5% to 96.2%, at one year follow up [1,22]. These differences were because of most of our patients had moderate to large PDA where occluder device was preferred compared to coil, but there were no statistically differences on further analysis.
Our major and minor complications rate for transcatheter closure of PDA was comparable to both local and international studies [1,2,9,12,21], and coil type device was responsible for the majority of the major complications (60%) [11,20]. One of our patients had coil embolization which was retrieved by snaring with any adverse clinical and haemodynamic effects. This could be construed as PDA coil delivery technique was more challenging and required a longer learning curve for the users to acquire the skills as compared to the closure with an occluder device [3,21]. The other reason probably because of the underestimation of the PDA size. One patient developed anaphylaxis reaction to the contrast given in our study which required a short duration of ventilatory support while another patient developed short segment thrombosis of the right femoral artery, and this was sufficiently treated medically. The frequency of minor complications was low and commonly occurred in younger and smaller children. These complications usually were associated with access related injury where smaller sized children less than 6 kg were more prone to this sequalae [11,26]. However, Carl et al conducted the largest study in infants less than 6 kg and concluded that transcatheter PDA occlusion was technically feasible with the success rate of 94.3% but it has the risks of major and minor adverse events [23]. Hence, proper case selection is vital. According to our experience, the PDA closure could be performed at later age if patient is clinically asymptomatic with the absence of clinical signs of heart failure.  [24,25]. Although transcatheter closure of very large PDA with pulmonary hypertension had been reported, scrupulous attention need to be advocated in patient's selection prior to the procedure. As mentioned, one of our patients had concomitant double chamber right ventricle which was then referred to cardiothoracic surgeon for open-heart surgery.
Presently, the optimal timing and thresholds for transcatheter PDA closure remain debatable, especially in the infancy group, which has becoming an on-going clinical dilemma due to age, weight, general condition as well as different institutions' practice variation [12,19]. Likewise, in bigger children, there was no homogeneity with regards to the timing for PDA closure, although some cardiologist would advocate when the diagnosis has been established. In our study, median survival time to PDA closure among paediatric age group was 16.9 months from the time of diagnosis. Following the diagnosis of PDA, patients will be enlisted, planned for cardiac catheterization, and proceeded with device closure if indicated and deemed fit for the procedure. Generally, our patients were given an elective procedure date within 1 to 2 years after the diagnosis, if clinically not in failure, and in smaller children the procedure would prevent the risks of access related vascular injury and failure of device placement [4]. Nevertheless, there was a delay in the intervention among a few of our patients when the PDA closure was indicated. The reasons for this delay were attributed by multiple factors such as logistic concern and financial constraint. A group of our patients had geographical barriers to access to our hospital facilities and hence postponement in the decision process for the procedure. Some of the children came from low socio-economic background and could not afford to pay for the device. Thus, they would require a referral for financial assistance which could have influenced the interventional timing.
Our study also revealed that those with weight less than 10 kg and younger age had a shorter median survival time to close the PDA. This could be explained by majority of them had moderate to large PDA which often presented with failure symptoms and failure to thrive. Hence, diagnosis and targeted intervention should be performed at an early stage. Oppositely, school-goers and teenagers typically presented with asymptomatic PDA. The procedure could be performed at a later stage after a thorough cardiac evaluation. Interestingly, our study showed that the size of PDA had no impact on the decision for earlier intervention. Generally, moderate to large size PDA often is associated with hemodynamic consequences which has warranted for an early intervention, while the management of small PDA is often less clear [15]. Nonetheless, the routine closure of small defect had been advocated to eliminate the risk of infective endocarditis, but a more prudent approach was recommended. There were several limitations in this study. This was a retrospective record review in a homogenous population with no definite PDA protocol. Therefore, biases in data collection and missing of information were expected with some records unavailability due to of the loss in the subsequent follow-up. Furthermore, the relatively small sample size also precludes from drawing a larger conclusion from the data obtained. Finally, further clinical evaluations and longer follow up were needed to assess any potential limitation related to the use of this device especially in younger children. However, the results of this study could be used as a guide when explaining the procedure to the patients or guardians regarding the outcome and possible complications that could have occurred during or after the procedure.

Conclusion
The findings from our study showed that transcatheter closure of PDA was very efficient and safe with excellent outcomes and had a comparable level of safety and efficacy with other cardiac centres. The minimal incidence of complications and residual shunts makes this approach as an ideal for the closure of PDA in paediatric population. This study also found that age and weight were important determinant factors for early PDA closure intervention.

Conflict of Interest
No possible conflict of interest to be declared.