Post-Infarction Ventricular Septal Defect Closure: The “BASSINET” Concept

Ischemic septal rupture producing an acquired ventricular septal defect (VSD) is a catastrophic complication of acute myocardial infarction. Acute therapeutic options are often futile because the tear in the septum is complex with surrounding edematous and necrotic tissue that evolves over several weeks following infarction. Concomitant left ventricular (LV) and/or right ventricular (RV) dysfunction limit effective surgical repair mostly to survivors of natural selection. “Dedicated” endovascular devices are best suited for the minority of patients with thin septums and small, circular defects during the chronic phase of post-infarct VSD¹. In other VSD anatomies, commercially available devices are too rigid to deliver, have an outward force that expands already necrotic defects and are highly permeable, failing to occlude high-flow defects. There is a need for better therapy.

To overcome these limitations, we adopted the following strategies to close post-infarction VSD (Figure A–D). First, we inspect the lesion using high-spatial resolution imaging and multiplanar reconstruction on cardiac CT or MRI (Figure E) and define an infarct zone. Second, we traverse the defects in a non-anatomic trajectory (crossing the center of the infarcted myocardium irrespective of the pathway of the hole created by the rupture) (Figure F). This allows devices to be positioned and deployed where they may best occlude the defect, not just where the defect allows them to cross. Third, we sandwich the defect using two devices positioned against the inflow and outflow (Figure G). Fourth, we implant the devices and tie them together using a COR-KNOT to promote acute closure. Fifth, we anticipate that acute septal edema will regress and that thickened septa will contract during VSD healing. Finally, we compress the tissue between sandwiching devices by adding anchor devices connected by sutures (Figure H), if necessary during a follow-up procedure. We call the technique “Bridge Across Septum using Sutures In a Non-Endoluminal Trajectory (BASSINET).” The first patient treated is described.

Figure. Novel Concept of a Hybrid Closure of Post-Infarct VSD.

A. Two Cribriform devices tethered to each other. B. Suture secured with a COR-KNOT device via subxiphoid approach. C. Commercially available devices, which are deployed through defect lumen, tend to have difficulty in apposing to the complex lesion contours, resulting in the residual gaps (yellow arrow) and incomplete expansion of the device. D. Direct non-endoluminal device placement allows the seal of the complex VSD. E. Inspect the lesion morphology using high-spatial resolution imaging and multiplanar reconstruction on cardiac CT or MRI. F. Traverse the defect through center of the VSD regardless of the shape of the VSD. G. Sandwich the complex VSD with two devices tethered each other with a polyethylene suture (blue line) and implant relatively impermeable devices. H. Compress the contracted tissue between sandwiching devices by adding anchor devices connected by sutures, if necessary during a follow-on procedure.

An 84-year-old woman was transferred to our hospital for evaluation and treatment of post-infarction VSD. The patient was evaluated at an outside hospital and found to have VSD with an initial troponin of 9 ng/mL. Echocardiography showed a left ventricular ejection fraction of 65% with a regional contraction abnormality in the base/mid ventricular septum, moderate to severe right ventricular (RV) dysfunction with an estimated RV systolic pressure (RVSP) of 57 mmHg, and a large VSD in the middle of the ventricular septum. Coronary angiography showed a 50% stenosis in the mid-left anterior descending artery and the right coronary artery, and 75% stenosis in the proximal diagonal branch. There were no obstructive lesions in left circumflex artery or lesions with suspected plaque rupture. Upon admission to our hospital, she complained of fatigue. Aside from a systolic blood pressure 90 mmHg, she was hemodynamically stable. Inspection of the CT showed a multifenestrated muscular mid-septal VSD with a single LV orifice and 2 different “Y-shaped” tunnels into the RV (Figure E). Our multi-disciplinary heart team discussed the treatment options with the patient and her family. Given her age and her moderate to severe RV dysfunction, we suspected that she had a high risk of mortality with a conventional surgical repair under cardiopulmonary bypass. We also believed the anatomy of the ventricular defect was unsuitable for existing dedicated devices used for percutaneous VSD closure. We then decided to proceed with the BASSINET procedure. The procedure was performed in the hybrid operating room under general anesthesia. We accessed the RV free wall via hybrid subxiphoid surgical exposure to place a 14Fr x 13cm introducer sheath. Using echocardiography and fluoroscopy guidance, we then traversed the center-of-mass of the defective tissue using a micropuncture needle, through septal myocardium in a non-endoluminal trajectory (Figure F), rather than navigating through the defect itself. We advanced the introducer sheath across this trajectory and then sandwiched the defect between 35mm and 30mm Amplatzer Cribriform Septal Occluder (Abbott St Jude, St Paul, MN), one complete device (both discs) on either side of the septum (35mm device on LV side; 30mm device on RV side), connected by nylon sutures that are sutured through devices prior to implantation (Figure G). Because both discs of each device laid flat across the endocardial border of the defect, the dual mesh layers reduced the device permeability to blood flow. The sutures were tensioned to snug the two devices across the septal rupture and tied using a COR-KNOT (LSI Solutions, Victor, NY) through the sheath, without cardiopulmonary bypass. Echocardiogram showed a small residual shunt with ventricular septal thickness of 23 mm. Invasive measurements showed the pulmonary artery pressures (PAP) dropped from 68/31 mmHg to 40/19 mmHg and the left ventricular end-diastolic pressure dropped from 30 mmHg to 20 mmHg. The RV puncture was closed using standard surgical pledgeted purse-string technique. The patient was discharged 5 days post-procedure with minimal symptoms.

At one-month follow-up, the patient had recurrence of her shortness of breath and a MDCT and an echocardiography revealed partial separation of the devices, resulting in a recurrence of the VSD. We believe the edematous septal myocardium began to heal and contract, causing the two devices to separate and the defect to leak. Therefore, we performed a repeat hybrid procedure to augment the BASSINET compression (see VIDEO). We accessed the RV free wall again via hybrid subxiphoid surgical exposure and inserted a 4-Fr sheath through both cribriform devices. We then deployed a 4×4 mm Amplatzer Duct Occluder II (ADO II, St. Jude medical) with premounted suture into the left ventricle. After we pulled the ADO II against the LV device, we advanced the COR-KNOT from the RV side, tensioning all three devices together before deploying the suture lock. We then placed a 2nd ADO II using the same technique. We were able to re-appose the two Amplatzer Cribriform Septal Occluders much closer than after the first procedure (compress, Figure H). Final right heart catheterization showed a Qp:Qs ratio of 1.15, PAP of 25/21 mmHg, and RVSP 22 mmHg. Final echocardiography showed only a mild leak and a ventricular septum diameter of 10 mm. The patient was discharged 5 days after the second procedure. At 30-day follow-up, the patient remained asymptomatic and echocardiography showed only a mild leak with devices in secure position.

A hybrid VSD repair consisting of a subxiphoid surgical approach and VSD closure with transcatheter devices can result in good closure with minimal risk and without cardiopulmonary bypass. Our experience suggests we need to consider the structural evolution of the septum geometry (i.e., resolution of inflammation and edema, and septal infarct remodeling) when determining the appropriate timing for intervention^2–4. In the future, there may be benefit to creating a purpose-built transcatheter device, which seals a VSD from both LV and RV side as demonstrated in this BASSINET procedure.