Perventricular device closure is a common treatment for VSD. The first real off-pump perventricular device closure of VSD was conducted in animal experiments in 1997 under TEE guidance and then applied in an infant with muscular VSD [28]. Subsequently, perventricular device closure of pmVSD was first reported in 2004 [29]. Recently, this technology has been widely applied in China. However, perventricular device closure of pmVSD is not applied worldwide due to safety concerns, especially concerns of heart block. Through this systematic review, we have attempted to evaluate the efficacy and safety of this technology.
The included studies were 8 case series and 7 case-control studies with a high quality and acceptable publication bias. The invasive intervention limited the blinding of the participants and personnel, and this contributed to the lack of RCTs. We contributed the bias to the following factors: first, the different study designs; second, the lack of multicenter studies in this analysis and the different patient selection criteria among the single centers; and third, the increased likelihood of studies with promising results being accepted and published.
We defined operational success as patients without fatal or severe early-term or late-term complications requiring reoperation. The pooled success rate of perventricular device closure was 0.95 (95% CI: 0.92–0.97, I2 = 86.2%, P = 0.000), including 15 studies with 1368 patients. The subgroup analysis suggested that the study type may be a source of heterogeneity. Furthermore, no uniform patient inclusion criteria were applied in all medical centers. However, only patients with isolated pmVSD were included, and patients with other coexisting cardiac anomalies, severe pulmonary hypertension, or significant aortic prolapse and newborns or young infants with a large VSD were excluded. The subaortic rim was required to be greater than 1–2 mm. The VSD size ranged from 4 to 12 mm. There was no correlation between the operational success rate and the following factors: publication year, sample size, study type, mean age, mean VSD size and TTE/TEE guidance, which indicates the short learning curve and easy promotion of this technology. Compared with conventional surgical repair, there is no need for cardiopulmonary bypass (CPB) in perventricular device closure. Compared with the transcatheter approach, the perventricular approach provides direct access and facilitates manipulation of the device position and orientation during device deployment. We attributed this to the shorter delivery path. A shorter delivery path also minimizes the risk of intracardiac structural damage due to catheter friction and rubbing. Thus, for experienced cardiac surgeons, the learning curve is short, and the promising prospects of this technology are easily promoted.
The pooled rate of postoperative residual shunting was 0.02 (95% CI: 0.01–0.03, I2 = 87.3%, P = 0.00). However, most of the shunts disappeared during the follow-up period, and the pooled follow-up rate of residual shunting was 0.00 (95% CI: 0.000–0.000, I2 = 30.5%, P = 0.00). Only 1 case of mild residual shunting during the follow-up period was observed. This change means that most residual shunts disappeared naturally during the follow-up period. Endothelialization finished several weeks after the operation, covering the surface of the device and forming neointima, thereby fully closing the residual shunt [30].
The pooled rate of severe intraoperative complications was 0.050 (95% CI: 0.028–0.071, I2 = 71.0%, P = 0.000). Patients with severe intraoperative complications, including significant residual shunting, mild to significant aortic regurgitation, severe arrhythmia, failure to establish a path and mild to significant tricuspid regurgitation, were all converted to conventional surgical repair under CPB. Significant residual shunting and mild to significant aortic regurgitation were the most common reasons for conversion. The incidence rates of severe arrhythmia, failure to establishing a path and mild to significant tricuspid regurgitation were low in perventricular device closure of pmVSD. Hu and his coworkers contributed approximately 10% of transthoracic device closure (TTDC) conversion to conventional surgical repair to unsuitable occluders, as all complications were resolved by removing the occluder [22]. A lack of multiple attempts with different types and sizes of occluders may also be a reason for conversion. Thus, among selected studies, the rate of conversion to surgical repair may also be identical. Upon the occurrence of complete atrioventricular block (cAVB), significant residual shunting (> 2 mm), new aortic regurgitation, or mild to significant tricuspid regurgitation, the procedure was converted to conventional surgical repair with CPB. Most complications disappeared after removal of the occluder, suggesting the importance of choosing a suitable occluder type and size. Asymmetrical and symmetrical occluders were the most widely used occluders and were selected for TTE/TTE-measured defect-to-aortic valve rims < 2 mm and ≥ 2 mm. The occluder size was selected according to the pmVSD diameter and was larger than the pmVSD by 1–2 mm. Failure to establish a path was reported in 5 studies, with a pooled rate of 0.000 (95% CI: − 0.000-0.000, I2 = 0.0%, P = 0.901). The precondition of establishing a path is finding a suitable puncture site perpendicular to the plane of the VSD. Surgeons mostly determine the puncture site by depressing the right ventricular free wall with an index finger to find the strongest pulsatory site under continuous TEE/TTE guidance. Unsuitable puncture results in the failure to establish a path.
The pooled rate of severe postoperative complications was 0.000 (95% CI: 0.000–0.000, I2 = 71.0%, P = 0.000). A total of 4 patients required a second operation, including 1 for occluder dislodgement and 3 for cAVB. Another patient with cAVB recovered a sinus rhythm and did not undergo a second operation or permanent pacemaker. Occluder dislodgement may be a procedure-related complication caused by a lack of experience with TTDC. In other cases of postoperative arrhythmia mentioned in the enrolled studies, a sinus rhythm was recovered within 48–72 h after surgery. This finding may be attributable to early procedure-related inflammation or the limited number of cases [18]. Only one patient experienced new mild tricuspid regurgitation, which disappeared during the follow-up period. No cases of new mild or significant aortic regurgitation were observed. The pooled rates of aortic regurgitation and tricuspid regurgitation were both 0.000 (95% CI: 0.000–0.000, I2 = 0.0%, P = 1.0). This promising result may be attributable to suitable occluder selection or the limited number of cases in this meta-analysis.
The pooled rate of severe complications in the follow-up period was 0.000 (95% CI: − 0.000-0.000, I2 = 0.0%, P = 0.487), including 3 cases of late cAVB, 5 cases of mild aortic regurgitation, and 1 case of residual shunting (> 2 mm). The above 3 patients with late cAVB recovered a sinus rhythm spontaneously or after steroid therapy. However, previous reports have emphasized that once late-onset cAVB occurs, a permanent pacemaker is the only cure for cAVB, which is in contrast to the above findings [31, 32]. One possible explanation may be that the conduction system was recently affected by the device-related inflammatory response or scar formation and the patients came to hospital for therapy immediately.
The mechanism of cAVB remains unclear. It is possible that occluder devices may cause an initial inflammatory response with subsequent formation and fibrosis in the conduction system [14]. Progressive device flattening may also be a mechanism for the development of cAVB, according to Butera G’s hypothesis [33]. Compared with the transcatheter approach, perventricular device closure involves a shorter path and thus avoids friction and rubbing of the conduction system and the subsequent inflammation. Meta-regression analysis indicated no significant correlation between early/late cAVB and the following factors: publication year, sample size, study type, mean age, mean VSD size, male prevalence, occluder-VSD size difference and TTE/TEE guidance (all P > 0.05). It is still a challenge to completely avoid cAVB given the surrounding anatomical structures in pmVSD; thus, precautions with suitable device selection (both type and size) are paramount. Certain devices have already been approved in some countries for use in pmVSD closure. No one type of occluder is suitable in all cases of VSD; thus, progressive improvements of these devices are also necessary.
Aortic regurgitation is another severe complication of perventricular device closure due to the short subaortic rim of pmVSD and the use of unsuitable occluders. Only 5 cases of mild aortic regurgitation were observed during the follow-up period. The pooled rate of aortic regurgitation in the follow-up period was 0.000 (95% CI: − 0.000-0.000, I2 = 0.0%, P = 0.982). This result show that the incidence of aortic regurgitation in the follow-up period was low, emphasizing the importance of accurately evaluating the subaortic rim and choosing a suitable occluder.