Abstract
Patients in cardiogenic shock due to severe failure of a mitral bioprosthesis have extremely high risk for repeat standard cardiac surgery. This patient population may benefit the most form the least invasive intervention. We report a successful transseptal mitral valve-in-valve implant performed in a patient in cardiogenic shock due to a flail mitral bioprosthesis requiring bridging intravenous inotropic and intra-aortic balloon pump support.
1
Introduction
Over the years, immense progress has been made in the transcatheter treatment of valvular heart disease. Although transcatheter aotic valve replacement (TAVR) is an established treatment option for patients with symptomatic severe aortic stenosis who have intermediate surgical risk or higher, transcatheter mitral valve replacement (TMVR) is at a much earlier phase of development. Patients with severe dysfunction of a degenerated mitral bioprosthesis usually have high surgical risk for a repeat standard cardiac surgery. TMVR is an attractive option in this high risk patient population. Since transcatheter heart valves (THV) designed for the mitral position are not available for commercial use yet, hundreds of patients with severe dysfunction of a degenerated mitral bioprosthesis have been treated with the use of transcatheter aortic valves worldwide [ , ]. Most of these transcatheter mitral valve-in-valve (V-in-V) procedures have been performed in hemodynamically stable patients and using a transapical approach. Although the transseptal approach may be more technically challenging compared with the transapical approach, it is less invasive and has been associated with better outcomes than the transapical approach [ ]. Accordingly more operators are adopting transseptal technique in current practice. Patients with hemodynamic compromise may benefit more from the utilization of the least invasive approach. We describe a patient with severe mitral regurgitation (MR) due to a failed bioprostethic leaflet who presented with cardiogenic shock and was successfully treated with a transseptal transcatheter mitral V-in-V implantation of a SAPIEN 3 valve (Edwards Lifesciences, Irvine, CA).
1.1
Case description
An 80-year-old female presented with past medical history of atrial fibrillation, remote pacemaker placement, asthma, pulmonary fibrosis, and MR for which she underwent mitral valve replacement with a 27 mm Biocor-Epic valve (St. Jude Medical, St. Paul, MN) 7 years prior. During the same prior surgical procedure, she had left atrial appendage (LAA) excision, MAZE procedure and tricuspid valve repair with a 27 mm St. Jude flexible ring. She presented to our institution with sudden onset of progressive dyspnea of 12 h duration. On arrival to the emergency room, she was in pulmonary edema with cardiogenic shock requiring intubation and mechanical ventilation. The pro-BNP level was 5221 pg/ml, and she had acute kidney injury with a creatinine of 2.3 mg/dL. The echocardiogram showed a left ventricular ejection fraction of 78%, severe MR with flail prosthetic mitral leaflet and a mean mitral valve gradient of 11 mm Hg, probably due to increased flow. The pulmonary artery systolic pressure was 72 mm Hg. The findings were confirmed with transesophageal echocardiogram (TEE) ( Fig. 1 ). She was initially treated with intravenous dopamine, norepinephrine, and intra-aortic balloon pump (IABP) support. Blood cultures were negative. She was evaluated by the cardiothoracic surgery team and was deemed to have extremely high surgical risk with STS predicted risk of mortality score of 42% with combined mortality and morbidity of 93%. Due to extreme surgical risk in the setting of cardiogenic shock, the multidisciplinary structural heart disease team considered a transseptal TMVR mitral V-in-V the best treatment option. After a careful discussion with the patient and her spouse about the options as well as risks and benefits, the decision was to proceed with transseptal mitral V-in-V.
1.2
Procedure description
Due to hemodynamic instability and acute renal failure, computed tomography was not performed for pre-procedural planning. The dimensions of a 27 mm Epic valve reported by the manufacturer and the Valve-in-Valve application [ ] were used for valve size selection. The internal diameter of a 27 mm Epic valve is 22.5 mm so we selected a 26 mm size SAPIEN 3 valve. An overexpanded 23 mm SAPIEN valve might have also worked but we considered the risk of embolization in the setting of severe MR would be lower with the 26 mm valve. The procedure was performed under general anesthesia in a hydrid operating room. IABP support was continued during the procedure. After right and left heart catheterization and selective coronary angiography, transseptal puncture was performed via the right femoral vein with an 8 Fr Mullins catheter and a BRK 1 extra sharp needle under fluoroscopic and TEE guidance. The Mullins sheath was advanced in the left atrium (LA) over a 0.032 J wire. A 0.025 × 230 cm Toray (Toray America, Houston, TX) wire was introduced in the LA, the Mullins sheath was removed, and upsized to a 14 Fr Edwards expandable sheath over the Toray wire. An 8.5 Fr Agilis sheath (St Jude, Minneapolis, MN) was then introduced in the LA over the Toray wire, and the Toray wire removed. Simultaneous LA and left ventricular pressures were recorded with the Agilis sheath in the LA and the pigtail in the left ventricle (LV). We connected a pressure transducer to the long 6 Fr sheath in abdominal aorta and the 5 Fr pigtail catheter in the left ventricle for continuous simultaneous LV and aortic pressures to detect left ventricle outflow tract obstruction (LVOT) obstruction after TMVR. The mitral valve was crossed from the LA with a 6 Fr pigtail and J wire. The J wire was removed and a 0.035 Extra-small Safari wire (Boston Scientific, Minneapolis, MN) was introduced in the LV. The pigtail catheter and Agilis sheath were removed. A septostomy was performed with a 14 mm diameter balloon. We then introduced a 26 mm SAPIEN 3 valve, prepared with 3 ml of additional contrast for balloon inflation. The SAPIEN 3 valve was deployed under fluoroscopic guidance using rapid pacing at 160 bpm ( Fig. 2 A ). TEE showed adequate MV function ( Fig. 2 B). The delivery system was removed. A 6 Fr multi-purpose guide catheter was introduced into the LA over the Safari wire. Simultaneous LA and LV pressures were recorded. The mean hemodynamic mitral valve gradient was 5 mm Hg, and 6 mm Hg by TEE, while supported by norepinephrine and IABP. The multipurpose catheter was introduced from the LA in LV to retrieve the Safari wire. The catheters and e-sheath were removed, and hemostasis was achieved with two previously placed Perclose sutures in the right femoral vein and also a figure of 8 temporary subcutaneous suture. There were no complications and the patient tolerated the procedure well.
1.3
Hospital course and follow-up
The patient had significant clinical improvement and tolerated weaning off of pressors and IABP over the following 2 days. Transthoracic echocardiogram done on post-operative day one showed preserved left ventricular function with ejection fraction of 55%, a SAPIEN 3 valve in proper mitral position with a mean mitral gradient of 5 mm Hg, no paravalvular or central regurgitation and no significant LVOT gradient. The length of stay in the hospital was 16 days. At 1-month follow-up, she was in New York Heart Association class II with mild dyspnea on exertion while walking and tolerating physical therapy at home with mild dyspnea. Echocardiogram at one month showed mean mitral valve gradient of 6.8 mm Hg, and no MR or LVOT gradient ( Fig. 3 ).
2
Discussion
Patients with failing surgical mitral bioprostheses usually have high risk for a repeat cardiac surgery. Hundreds of patients worldwide have been treated successfully with the use of aortic THV devices. The Valve In Valve International Database (VIVID) was the first large international registry collecting outcomes data of these procedures. In this registry, the most frequently used THV is the Edwards SAPIEN family of valves. In an analysis of 349 patients in the VIVID registry, the 30-day survival free from significant MR (moderate or more) or clinically-evident LVOT obstruction was 88.8%. Approximately 80% of those patients were treated with a transapical approach. Although the transseptal approach may be more technically challenging, it is less invasive and may be better tolerated by patients [ ]. The VIVID registry reported that LV function in patients with ejection fraction <50% at baseline had greater improvement with transseptal versus transapical mitral V-in-V [ ]. A recent report from a single center found that among 24 patients who underwent TMVR V-in-V, those patients treated with transseptal approach had significant increase in cardiac output and improved survival compared with patients treated with transapical approach [ ]. This has not been systemically evaluated in a prospective trial. The MITRAL Trial (Mitral Implantation of TRAnscatheter vaLves) is a prospective FDA approved physician-sponsored trial evaluating the safety and feasibility of TMVR in three patient populations: native valves with severe mitral annular calcification (MAC), failed surgical rings, and failed surgical bioprosthesis (NCT 02370511). The transvenous transseptal approach is the delivery method of choice in this trial. Enrollment was recently completed and it is expected that this trial will provide important information about the safety and outcomes of transseptal TMVR.
Cardiac CT evaluation as part of the pre-procedural planning is useful to confirm the size of the bioprosthesis, estimate the risk of TMVR-induced LVOT obstruction, to plan the landing zone and valve deployment fluoroscopic angles as well as to plan transseptal or transapical access [ ]. Although cardiac CT has become an important tool in the patient selection process in our practice, a CT scan was not obtained in this patient due acute renal failure in the setting of cardiogenic shock. When a CT is not available, transthoracic and transeshopaegal echocardiogram can provide useful information about factors associated with increased risk of LVOT obstruction including aortomitral angle, size of left ventricular cavity, interventricular septum thickness, size of LVOT space and distance from the ventricular edge of the mitral bioprosthesis to the septum.
In addition to preprocedural planning with cardiac imaging, standardizing the implantation technique has been helpful to obtain reproducible outcomes. The use of pre-shaped wires in the LV and left atrium are important to reduce perforation risk. After transseptal access, we use a 0.025 Toray wire to exchange for the Edwards and Agilis sheaths. This wire provides adequate support to introduce the large bore Edwards e-sheath in the femoral vein. It is important to not introduce the Edwards sheath too early in the procedure when full anticoagulation has not been given and to avoid long indwelling times to decrease the risk of thrombus forming in the sheath. We typically introduce the large sheath after the transseptal puncture when full anticoagulation has been given. Alternatively, this step can be done after the septostomy just prior to valve insertion. Valve deployment should be slow to allow operators to adjust the position of the THV at the landing zone if needed. The use of additional contrast during initial valve deployment as was done in this case (3 ml), facilitates flaring the THV in the ventricular side to decrease the risk of embolization. A step-by-step description of the transseptal mival valve-in-valve technique used at our center has been reported [ ].
Although the risk of LVOT obstruction during mitral valve-in-valve procedures is lower than valve-in-ring procedures or during TMVR in a native valve, it is a potential complication which can lead to significant hemodynamic compromise or even death. Therefore, it is wise to continuously monitor LVOT gradient to be able to detect obstruction immediately when it occurs to expedite treatment. The risk of LVOT obstruction was suspected to be low in this case based on the lack of high risk features on echocardiogram and the fact that the Biocor Epic valve has short porcine leaflets.
Patients in cardiogenic shock may benefit from the least invasive approach for TMVR. There are only a handful of reports of transseptal mitral V-inV procedures performed in patients with cardiogenic shock. The first procedures reported used a transapical approach. VanGarsse et al., described the first case of a transapical mitral V-in-V implantation of an Edwards Sapien valve in a critically-ill patient with acute prosthesis failure with good results in 2011 [ ]. Two more cases of successful mitral V-in-V implantation with transapical approach in patients with cardiogenic shock have since been described [ , ].
The first transseptal transcatheter mitral V-in-V procedure in cardiogenic shock was reported by Fassa et al., in a patient not suitable for transapical approach due to high haemorrhagic risk. A 26 mm Edwards SAPIEN XT was implanted in a 31 mm Edwards Perimount Magna bioprothesis. Due to the unavailability of a 29 mm SAPIEN XT at that time, a size 26 mm was chosen. The intervention was complicated by valve migration and embolization, requiring a second implantation of a SAPIEN XT 26 mm, using more contrast medium to achieve larger inflated delivery balloon diameter. The procedure was performed under venoarterial extracorporeal membrane oxygenation (ECMO) support [ ]. The same group subsequently reported a series of 17 patients who underwent transseptal mitral V-in-V or valve-in-ring. Three patients in this series were in cardiogenic shock, the one mentioned above and 2 additional patients. In one of them, the procedure was unsuccessful due to inability to cross the mitral bioprosthesis. The maneuvers to cross the bioprosthesis resulted in inferior vena cava rupture and fatal bleeding. In the other patient, a 29 mm SAPIEN XT was implanted successfully on a severely stenosed 31 mm Edwards Perimount bioprosthesis [ ]. Another case was recently reported by Alli et al. In this patient with coagulopathy secondary to liver failure, a transapical approach was not considered a safe option. A 29 mm SAPIEN XT was implanted by transseptal approach in a severely stenosed 29 mm Edwards Perimount bioprosthesis with good immediate and short term results [ ]. A summary of cases reported in the literature of mitral V-in-V procedures performed in patients with cardiogenic shock is listed in Table 1 .
