The impact of advances in percutaneous catheter interventions on redo cardiac surgery

Abstract

Toward the end of the twentieth century, redo cardiac surgery accounted for approximately 15–20% of total cardiac surgical volume. Major risk factors for redo cardiac surgery include young age at time of the first operation, progression of native coronary artery disease (CAD), vein graft atherosclerosis, bioprosthetic valve failure and endocarditis, and transplantation for end stage heart failure. Historically, redo coronary artery bypass grafting (CABG) alone carried a mortality risk of around 4%. Factors such as older age, female sex, comorbidities, combined procedures, hemodynamic instability, and emergency procedures contributed to even higher mortality and morbidity. These poor outcomes made it necessary to look for less invasive alternate methods of treatment. Advances in catheter-based interventions have made a major impact on redo cardiac surgeries, making it no longer the first option in a majority of cases. Percutaneous interventions for recurrence following CABG, transcutaneous aortic valve replacement (TAVR) for calcific aortic stenosis, valve in valve (VIV) implantations, device closure of paravalvular leaks (PVL), and thoracic endovascular aortic repair (TEVAR) for residual and recurrent aneurysms and mitral clip to correct mitral regurgitation (MR) in heart failure are rapidly developing or developed, obviating the need for redo cardiac surgery. Our intent is to review these advances and their impact on redo cardiac surgery.

Introduction: Role of catheters in cardiac surgery

Cardiac catheterization which began by measuring the intracardiac pressures and later progressed to delineating the anatomy has paved way for the advancement in cardiac surgery as we see it today [1]. Use of a Swan-Ganz catheter became a part of standard of care during cardiac surgical procedures. Diagnostic coronary angiogram, first performed by Mason Sones in the 1960s, later evolved over decades to complex coronary interventions. Rashkind first described a catheter-based approach for creating an atrial septostomy in 1966 as a palliative approach to treat transposition of the great vessels [2]. Subsequently, several percutaneous techniques have evolved, which have either delayed the first surgical intervention like percutaneous coronary intervention (PCI) for coronary artery disease (CAD), or balloon valvuloplasty for mitral stenosis [3], or completely eliminated the need for surgery (percutaneous closure of atrial septal defect (ASD), patent ductus arteriosus (PDA), and radiofrequency ablation (RFA) for Wolf-Parkinson-White syndrome (WPW) [4]. Numerous catheter-based procedures have evolved to emulate the Cox maze procedure, considered to be the gold standard in the surgical treatment of atrial fibrillation [5].

History of cardiac surgery

Cardiac surgery started in several forefronts by performing palliative operations like the Blalock Taussig operations for congenital heart disease and closed mitral commissurotomy for rheumatic mitral stenosis. The Vineberg operation, along with several innovative procedures, was performed to treat CAD. Then, cross circulation was considered a way to empty the heart from blood to perform intracardiac operations [6]. After the first successful cardiac operation performed by John Gibbon in May of 1953, with the help of the newly developed heart lung machine, a new era started for cardiac surgery [7]. As the specialty advanced, cardiac surgeons began to realize that in many cases, the “state-of-the-art” surgical procedures which they performed did not pan out as expected, and many required repeat cardiac intervention. Graft occlusions, progression of native CAD, mechanical valve failure, and bioprosthetic valve degeneration were some of the few causes of redo surgery. Redo surgery was the first choice for a majority of cases until the end of the nineteenth century. The higher mortality and morbidity associated with redo surgeries and disruptive innovation in cardiac interventions significantly contributed to the drastic decline in the number of redo surgeries. Patients were considered high surgical risk for redo surgery if they had (1) Society of Thoracic Surgeons (STS) predicted operative mortality of 6 to 8%, which increases depending upon whether they need an isolated procedure or multiple procedures; (2) porcelain aorta indicating high risk for intra-operative stroke; (3) patient frailty; (4) previous mediastinal infections; (5) liver failure; (6) pulmonary hypertension; (7) active malignancy; or (8) prior thoracic radiation [8].

Redo cardiac surgery

Reoperation, which represents a significant proportion of modern cardiac surgical procedures, is often associated with significantly higher mortality than first-time operations. By contrast, major medical centers with surgical excellence did not classify first-time redo cardiac surgeries as carrying higher surgical risk, unless the patient had previous radiation to the chest or a sternal wound infection [9, 10]. Older age, multiple prior sternotomies, preoperative renal failure, and peripheral vascular disease are all considered preoperative risk factors for higher mortality and morbidity [11]. A patent and a well-functioning left internal mammary artery (LIMA) graft close to the midline acted as a deterrent for redo cardiac surgery by many surgeons [12]. Emergency redo cardiac surgery carries higher mortality, even at centers with surgical excellence. This compelled surgeons and cardiologists to look for alternative approaches. Improvements in the medical management of cardiovascular diseases [13], and rapid advances in catheter technology, have significantly reduced the need for redo cardiac surgery, except in cases of active infection.

Impact of angioplasty

In 1978, Andreas Grüntzig introduced PCI, which provided an alternative treatment strategy for symptomatic CAD. Major innovations in interventional cardiology in the form of rota ablation, intravascular ultrasound, and high-quality drug-eluting stents (DES) opened the flood gates and PCI rates started to grow exponentially [14]. By the turn of the century, coronary artery bypass grafting (CABG) numbers across the globe started to decline and the twenty-first century saw a continuous increase in PCIs, while the number of CABG procedures continued to decline. Redo CABG, once a common procedure in many centers, is now becoming an obsolete procedure.

Initially, bypassing for other than left anterior descending artery (LAD) disease, when 1 or 2 vessels were stenotic involving a large portion of the myocardium, was considered as class IIb indication for CABG as per American College of Cardiology(ACC)/American Heart Association(AHC) guidelines [15]. As such, redo CABG was performed when right coronary artery (RCA) and left circumflex (LCX) vessel stenoses were identified in spite of a patent LIMA graft to LAD. As cardiologists started to master PCI of native coronary vessels, they started to address more complex lesions with severe stenosis (> 75%) including vein graft stenosis. Subramanian and Sabik [16] published data of 4640 patients who underwent prior CABG surgery and had patent LIMA graft to LAD and had at least 50% stenosis in the non-LAD coronary arteries. Of these patients, 731 had redo CABG procedure, 994 had PCI, and 2782 had medical management only. They concluded that patients who had patent LIMA graft to LAD had no survival benefit at 1 year over the other groups and that patients who underwent invasive therapeutic intervention had improvement in their symptoms. However, they required more repeat therapeutic interventions in their follow-up years. The remaining 133 patients, though initially classified under medical management, crossed over to PCI or surgery because of increasing anginal symptoms. PCI became a good option for patients who had primary CABG done at a very young age and delayed the major redo cardiac surgery to a later date to avoid the inherent complications of the surgical procedure. As a less invasive option than a redo CABG, hybrid procedures were adopted as accepted approaches to treat CAD in a select group of patients, using a combination of non-sternotomy surgical approaches and angioplasty of stenotic vessels [17]. The timing of these combined procedures varied depending upon the coronary anatomy, patient’s anginal symptoms, and individual institutional practices.

Left main coronary (LM) artery stenosis, identified for the first time during cardiac catheterization or in patients with prior CABG, was not amenable to PCI for fear that acute occlusion during the procedure would result in ventricular fibrillation, cardiac arrest, or cardiogenic shock. Instead, patients were generally sent for primary or redo CABG surgery. Depending upon preoperative risk factors, the morbidity and mortality were assessed and for lack of other alternatives, both patient and physicians have accepted these high-risk surgical procedures. Later, PCI for left main trunk (LMT) stenosis was adopted as a bailout procedure in patients who were very high surgical risk or patients who had a protected left main trunk stenosis with a patent LIMA graft to LAD.

Once cardiologists developed confidence to tackle LM stenosis by intravascular ultrasound (IVUS)-guided PCI, even low-risk patients were subjected to this procedure, initially supported by intra-aortic balloon pump (IABP) assist and later with the Impella device that has been used in patients with severely depressed cardiac function. The PROTECT I trial showed that the Impella 2.5 (ABIOMED, U.S. Danvers, MA, USA) provided excellent hemodynamic support during high-risk PCI [18]. The EXCEL (Everolimus-Eluting stents or Bypass surgery for left main trunk coronary artery disease) study randomized low- to intermediate-risk patients to both LM angioplasty and CABG surgery and found that PCI was not inferior to surgery in regard to predicted risk of mortality (PROM) at 3 years [19]. Park and associates published a multicenter study comparing PCI versus CABG for left main coronary artery stenosis and concluded that at 5-year follow-up, patients that received CABG did better than those that received PCI and there were higher deaths with DES [20]. Currently with dual antiplatelet therapy for at least 1 year, the incidence of in-stent thrombosis has decreased. While ostial and left main coronary artery stenosis can be stented, stenosis of the left main at the bifurcation into LAD and LCX was initially considered a surgical disease. With increased experience, even those lesions are also subjected to PCI, mainly depending upon the institution. As long as the surgeons continue to use arterial grafts for coronary revascularization, the pendulum will swing back to surgery again for LM stenosis, as the durability of the arterial grafts is proven beyond doubt. If surgeons continue to use all arterial grafts, preferably in young patients, long-term freedom from further interventions will occur in this subgroup [21]. Then, PCI will only be indicated in elderly patients because of their increased comorbidities and risk for operative intervention.

Conduit failures

Vein graft atherosclerosis is the primary cause of vein graft stenosis that requires further intervention. Endothelial damage, intimal hyperplasia, and accelerated atherosclerosis cause vein graft stenosis or failure [22]. Medical management with antiplatelet agents and lipid-lowering therapy help to maintain vein graft patency, but when patients develop recurring anginal symptoms, further intervention is required. With improved endovascular interventional technologies, the need for isolated redo CABG due to vein graft stenosis has gone down. Stenotic bypass grafts are addressed surgically only in those patients who needed valvular surgery and had prior CABG. About 6% of all PCIs are to treat vein graft failure. Both drug-eluting and bare metal stents are used to treat vein graft stenosis. Though there was an initial preference for DES to treat vein graft pathology, several studies have not shown any benefit over bare metal stents when followed for 12 months, including all-cause mortality, major adverse cardiac events (MACE), and stent thrombosis [23]. Both distal embolization and restenosis of the vein graft were the predominant complications in early studies.

Anastomotic stenosis due to technical complications or from purse string effect of the suture does occur at both arterial and vein graft anastomosis. Spasm of the arterial conduits can occur during post-operative use of vasopressors and generally occur at the anastomotic site, as this is the part of the vessel that is surgically manipulated the most. Earlier, prior to the angioplasty era, reoperation was performed for the culprit lesions involving the LAD vessel at the LIMA and LAD anastomotic site. Redo cardiac surgery within the first month of the prior surgery has increased risk because of the acute inflammation all around the heart making surgical dissection challenging. Angioplasty of the LAD through LIMA graft or internal mammary artery stenosis at the proximal or distal end was performed in 327 patients with 87% success rate as reported by Marx [24]. This shows the advances in catheter-based technology. As such, it was proposed that angioplasty of the stenotic lesions, either in the native vessels or in the bypass conduits, should be performed first before surgical intervention.

Aortic valve

Redo aortic valve surgery may be required following valve repair for congenital abnormalities during childhood, thrombosed mechanical valves, degenerated bio-prosthetic valves, and endocarditis affecting the aortic valve, aortic annulus, and previously implanted Dacron grafts. Aortic stenosis and mitral regurgitation, either combined or as an isolated pathology, are the common long-term sequelae that develop in surviving patients with CABG. Aortic valve replacement after previous CABG is a more common procedure [25]. The survival benefit of a well-functioning LIMA graft to LAD is clear and patients with a longer life span do develop calcific aortic stenosis as part of the aging process. Redo cardiac surgery for valve replacement carried a 10–30% mortality depending upon the patient’s age, other risk factors, and experience of the institution [26]. Medical management of symptomatic patients with critical aortic stenosis provides only minor symptomatic relief and does not alter the long-term mortality. The presence of a patent and well-functioning LIMA graft to LAD created a surgical debate as to whether the LIMA should be dissected and temporarily occluded to achieve uniform myocardial protection, or it can be left alone, perfusing the myocardium, while the patient is cooled to lower systemic temperatures to decrease the global oxygen demand to the body [27]. Damaging a patent and well-functioning LIMA graft during surgery increases the complexity of the operation and increases the surgical mortality. The presence of patent vein grafts on the aorta also alters the aortotomy incision, thus making the surgical aortic valve replacement (SAVR) difficult. Due to the aforementioned surgical risks, we believe that the following approach is more optimal. Prior to treatment, patients should be seen by a Heart Team that includes both cardiovascular surgeons and interventional cardiologists, and the patient’s procedure-related risk calculated using the STS criteria. For patients with previous coronary surgery and a predicted operative mortality of 8% or greater with SAVR (the high-risk group in Partner-1), transcatheter aortic valve replacement (TAVR) would seem a logical approach as long as serious coronary stenosis is not present in the arteries. The presence of life-threatening coronary disease warrants at least a consideration for SAVR, if the lesions cannot be safely approached with PCI. For patients in the moderate risk (4–8%), TAVR would seem to be indicated in the absence of severe three-vessel native coronary and graft disease. SAVR being used for patients with aortic stenosis combined with serious coronary or graft disease usually means that no LIMA graft to LAD is functioning. For patients in the lower range of operative risk (< 4%), who commonly have a predicted life expectancy of greater than 10 years, SAVR would seem the most prudent approach, particularly in the face of severe coronary stenoses involving major coronary vessels that lend themselves to bypass grafting. Those reoperations should be carried out by surgeons with substantial experience with coronary and aortic valve reoperations and at institutions where such procedures are common.

In spite of the fact that TAVR is now an accepted procedure in high-risk patients, it is also becoming a choice for moderate as well as low-risk patients. Coupled with the advancements in the device, as well as valve durability, it will likely be the preferred choice for the patients in the future. In a study, 1660 patients who were intermediate surgical risk (STS predicted risk of mortality of 4.5 ± 1.6%) and having severe aortic stenosis underwent randomization for TAVR or SAVR. These were followed for a 24-month period and it was concluded that TAVR was noninferior to surgery [28].

Vascular access initially started as a trans-femoral approach. Physicians have tried several alternative arterial sites such as trans-axillary, trans-carotid, and trans-apical approaches through the left ventricle. A mini right thoracotomy to directly access the aorta was also tried to deploy the valve. Aortic regurgitation from a paravalvular leak (PVL), conduction disturbances requiring a permanent pacemaker implantation, vascular access complications involving injury to femoral or iliac vessels, acute kidney injury, and intraoperative stroke are some of the serious complications of TAVR [29]. PVL has significant negative implications in the prognosis of the patients, as such several technical improvements in the valve design like skirted valves and also various plugging devices to seal the PVL have been improvised [30]. In the PARTNER trial, preoperative right bundle branch block and left anterior hemiblock patients have a higher incidence of postprocedural permanent pacemaker implantation—up to 17% in 11,210 patients studied [31]. A trans-caval approach from the femoral vein has been used successfully in several patients and entering the aorta below the level of the renal arteries and closing the punctured sites with nitinol occluders. Currently, this is limited to patients with poor arterial access, but with increased experience it may become one of the popular approaches [32]. Acute ventricular rupture and pericardial tamponade required urgent surgical resuscitation, resulting in very high peri-operative mortality.

TAVR is also applied to bioprosthetic valve failures which otherwise would have needed a redo aortic valve replacement. Use of bioprosthetic valves in the aortic position has increased over the years as patients are unwilling to take life-long anticoagulation for mechanical heart valves. Redo aortic valve replacement for a bioprosthetic valve failure is a more difficult operation as the time required to remove the original valve and re-implant the new bio-prosthesis results in prolonged aortic cross clamping. Aortic root enlargement procedures also add more complexity to an already difficult operation. ViV TAVR is also becoming popular as most of the patients that had bio-prosthesis were elderly patients to start with and have multiple comorbid conditions. Bioprosthetic valve failures are recognized as both aortic insufficiency and calcific aortic stenosis. This procedure is applied to patients who had both stentless and stented aortic valves in their prior surgical procedure. Higher gradients are noted in the ViV TAVR procedures in the beginning because of patient-prosthesis mismatch but improvement in the valve design and procedural advances like bioprosthetic valve fracture (BVF) will help to insert larger size valves [33, 34]. Modification of a stretchable mounting ring for bioprosthetic valves will facilitate ViV TAVR as the common practice for failed bio-prosthesis in the future, totally avoiding redo valve replacement. Small aortic annulus (SAA) always presents a clinical challenge with aortic stenosis. Prosthesis-patient mismatch (PPM) always resulted in higher perioperative morbidity and mortality. Clavel et al. showed, in a matched comparison of balloon-expandable transcatheter valves with surgical valves (stented and stentless), lower trans-prosthetic valve gradients and higher indexed effective orifice area (EOA) values and reduced PPM rates in TAVR recipients [35]. As experience with TAVR has grown, newer generations of valves have smaller delivery systems and have been designed to reduce PVL while continuing to offer increased EOA and lower gradients. Long-term data however is not available as to the durability and ability of the TAVR valves to maintain their gradients. Preliminary short-term data looks promising.

Mitral valve

Rheumatic mitral valve surgery was very prevalent in the early stages of cardiac surgery, and even now in some parts of the world, with varied incidence involving young female patients. Closed commissurotomy and, later, open commissurotomy were the initial procedures performed to provide symptomatic relief. In 1984, Inoue first described percutaneous balloon valvuloplasty [3]. Like any new procedure, it had complications, including ASD, rupture of the anterior leaflet of the mitral valve, and inability to obtain satisfactory results. As the catheters improved and as operator experience grew, the procedure certainly helped to delay surgical intervention in many patients for several years. The need for mitral valve surgery is increasing in the Western world as patients continue to live longer with poor left ventricular function and dilated cardiomyopathy. Redo mitral valve surgery accounts for nearly 10% of all mitral valve operations [36] and is associated with increased surgical risk as well as increased resource utilization. Ischemic mitral regurgitation (IMR) is a common complication of coronary artery disease as well as poor left ventricular (LV) function. Combined procedures at the initial operation, like mitral and aortic valve surgery or mitral and tricuspid valve surgery or CABG with valvular procedures, tend to result in recurring problems that may increase the need for repeat valvular surgery. Underlying medical conditions such as uncontrolled hypertension and diabetes can produce left ventricular dysfunction and eventually functional mitral regurgitation (MR). Some patients who had aortic valve replacement for bicuspid aortic valve pathology, without aortic root replacement, developed mitral valve regurgitation when followed up to 10 years [37]. Residual mitral regurgitation after mitral valve repair gets progressively worse in the post-operative period, requiring repeat intervention to correct the MR. Kron et al. followed 110 patients for 10 years following mitral valve repair and residual MR at the completion of the repair. In that group, 76 patients developed MR and died of recurrent MR and heart failure [38]. Redo mitral valve surgery in those conditions, even today, carries very high morbidity and mortality. Restrictive annuloplasty, Alfieri stitch, or mitral valve replacement carry a mortality risk of roughly 7–10% at experienced centers [39]. The presence of aortic valve prosthesis gives poor exposure to the mitral valve, resulting in technical complications such as perivalvular leak following mitral valve (MV) replacement or persistent mitral regurgitation following annuloplasty. The other causes for poor exposure of the MV are emphysematous deep chest, presence of patent bypass grafts, and extensive cardiac adhesions. While a right thoracotomy approach would give better exposure of the mitral valve, not all surgeons prefer the approach for fear of not being able to dissect out the aorta for cross clamping. Beating heart mitral valve surgery is being performed by some surgeons and air emboli need to be prevented to eliminate neurological complications [40]. Given the high-risk situation prior to surgery, percutaneous insertion of a mitral clip, while not the perfect solution, became a better alternative and acts as a temporizing procedure to limit the degree of MR. The MitraClip (Abbott Vascular, Menlo Park, CA, USA) has been approved for percutaneous repair and has been successfully used by cardiologists for high-risk surgical patients, who are in heart failure or are symptomatic secondary to IMR. The EVEREST (Endovascular Valve Edge-to-Edge Repair Study) phase I and phase II trials were established to assess the safety, durability, morbidity, and recurrence rates between surgical repair or replacement and mitral clip repair [41]. The 1-year follow-up mortality after MitraClip remained around 6%, granted that these are high-risk surgical patients. Along with MitraClip, several percutaneous annuloplasty techniques also have been approved for clinical trials. Recently published Percutaneous Repair with the MitraClip Device for Severe Functional and Secondary Mitral Regurgitation (MITRA-FR) versus Cardiovascular Outcomes Assessment of the MitraClip Percutaneous Therapy for Heart Failure with Functional Mitral regurgitation (COAPT) study shows which patients benefitted depending upon the effective regurgitant orifice area (EROA), regurgitant volume, and left ventricular ejection fraction (LVEF), along with optimal guideline directed medical therapy. COAPT patients with worse LV function and more regurgitant volume fared better when combined medical therapy and MitraClip were applied. The end point considered is repeat hospital admissions for heart failure and survival for a period of 2 years and the quality of life [42].

Anwer et al. compared 75 conventional redo cardiac surgery and MV repair patients to 56 patients that had MitraClip placed through the transcatheter technique. In their study, the patients that had MitraClip had higher surgical STS risk scores compared with the patients that had surgery [43]. Rogers et al. reported performing surgery and repairing the mitral valves up to 5 years after MitraClip implantation, as patients continued to have symptoms from recurrent or persistent MR [44]. With improved experience as well as visualization of the valve, the technical failure rate continues to decrease significantly. However, as long as there is LV dilatation and poor LV function, the presence of MR is an inevitable sequela of the failing heart and will require some kind of palliation.

Except in the hands of very few experienced surgeons, mitral annular calcification (MAC) has been a surgical challenge for most cardiac surgeons resulting in perioperative complications like bleeding secondary to atrioventricular groove disruption, inability to implant a decent size prosthetic valve leading to PPM, and occurrence of delayed PVL leading to further complications. In a redo cardiac surgical setting, the presence of MAC further increases the surgical risk of mitral valve surgery. Transcatheter mitral valve replacement is becoming an emerging alternative in high surgical risk patients using trans-venous or trans-atrial approaches [45]. Unlike the aortic valve, the mitral valve has both the valve and mitral apparatus to deal with and the annulus is fairly large in size and bulky valve design can lead to left ventricular outflow obstruction. The D-shaped configuration of the mitral annulus does not consent to achieve radial force, increasing the risk of PVL and prosthesis migration [46].

Partial or complete leaflet entrapment of mechanical mitral valves has been reported secondary to inadequate anticoagulation therapy and subsequent thrombus formation. This is more prevalent in developing countries. Patients develop heart failure symptoms of mitral stenosis. Patients presenting with New York Heart Association (NYHA) class IV status are at very high surgical risk compared with the patients in NYHA I-III. Roudaut et al. presented 263 patients with such mechanical valve thrombosis, of which 136 were treated surgically with valve replacement and 127 by removal of the pannus, with a 10.3% mortality in NYHA class III–IV patients [47]. Patients who had successful surgery experienced higher rates of complete hemodynamic improvement and lower rates of thromboembolic events, compared with the patients who had fibrinolysis. Immediate mortality in both groups was similar, around 10%. Thrombolytic therapy has been administered through percutaneous catheters passed via trans-septal approach into the left atrium. Hariram reported percutaneous catheter manipulation of thrombosed mitral valve leaflets in 5 patients successfully [48].

Paravalvular leaks

In both aortic and mitral valve replacements, despite being frequently performed procedures, PVLs have been reported for various reasons. Out of 60,000 valve implantations a year in the USA, it is estimated that 12% of patients have some degree of paravalvular leaks [49, 50]. This could be due to nonpliable calcific tissues, very fragile tissues secondary to infection, or technical difficulties with placing sutures because of difficult surgical exposure. While small leaks are benign, others can lead to serious problems such as hemolysis, anemia, heart failure, and even bacterial endocarditis. In those patients who require redo cardiac surgery, there is no assurance that their PVLs will not recur or can be fixed. Surgical mortality is generally around 20–30%. Higher incidence of PVL is reported with mitral valves (60%) compared with aortic valves and with bio-prosthesis (51%) than mechanical valves. Proper identification of the PVL requires 3D echocardiography and even cardiac magnetic resonance imaging (MRI), if percutaneous approach is planned to fix these leaks [51]. Different types of Amplatzer devices (Abbott Vascular, Menlo Park, CA, USA) have been designed to address percutaneous repair of PVL. The optimal approach depends upon valve type (tissue, mechanical) and valve location (mitral, aortic). Trans-femoral, trans-septal, or trans-apical approaches can be used, depending upon the location of PVL. Retrograde trans-femoral approach is optimal to repair aortic PVL. The first successful closure of PVL by a transcatheter technique was reported by Hourihan in 1992. The double umbrella technique has been used in 7 out of 8 patients in aortic position and were successful in all 7 patients [52]. While this catheter approach still needs to be perfected, it is considered a choice for high-risk patients. Only a few medical centers in any given region can generate enough volume and experience to give satisfactory results. Sorajja et al. in a series of 243 patients from Mayo Clinic had 74% mitral and 26% aortic paravalvular leaks [53]. They had 92% success rate in deploying the device, with 7% major adverse events in the 30-day perioperative period. Device deployment success is considered when there were no embolic complications or leaflet impingement.

Ventricular septal defect

Acute ventricular septal defect (VSD) secondary to septal muscle rupture following acute myocardial infarction is becoming less frequent because rapid catheter intervention of acute coronary occlusion limits myocardial necrosis. However, when this complication occurs in patients with prior cardiac surgery, they are definitely of high surgical risk and immediate surgical mortality ranges from 30 to 40%, depending upon surgeon experience and patient’s preoperative status. These are classified as Becker types I, II, and III depending upon whether they formed early, immediately after the acute myocardial infarction, or delayed when thinning and aneurysmal malformation of the infarcted muscle has taken place [54]. While an antero-apical VSD is localized and is through-and-through, inferior VSD is more serpiginous across the septum and is very close to the tricuspid valve and is more challenging to treat. Schlotter et al. reviewed catheter-based approaches in 13 published series totaling 273 patients and found an 89% success rate to prevent left to right shunt across the VSD. However, the overall 30-day mortality remained high at 32%, which goes along with the pathology. Major complications included device embolization, ventricular perforation, and arrhythmias. The Amplatzer post-infarction (PI) septal occluder device (AGA Medical corporation, Plymouth, MN, USA) has been the most common device implanted via the trans-venous and trans-femoral arterial approaches [55]. These devices come in different sizes starting from 10 to 35 mm and different shapes, some cribriform and multi-fenestrated to fit into the different VSD types.

Tricuspid and pulmonary valves

Tricuspid valve regurgitation (TR) occurs from left heart disease as well as pulmonary hypertension. Primary valvular pathology is also a known entity and in the adult population, it is mostly secondary to endocarditis from drug abuse. Patients who had prior mitral valve surgery also develop tricuspid valve regurgitation over a period of time. Both aortic and mitral valvular diseases do lead to gradual TR and many a time, the tricuspid regurgitation persists in spite of the treatment for the left-sided valvular disease. From that knowledge, concomitant repair of tricuspid valve (TV) for moderate TR, along with left-sided valvular surgery, is becoming more popular among surgeons [56]. Untreated TR leads to slow right heart failure resulting in shortness of breath and peripheral edema, and sometimes these symptoms do not improve in spite of aggressive medical therapy. Surgery for the TV in patients who had prior cardiac surgery, or replacement of the TV in patients who had prior repair or replacement of the tricuspid valve, carries a high mortality because of right heart failure and subsequent multiorgan failure [57]. Injury to the right ventricle during sternal re-entry can occur as the right ventricle is enlarged and stuck behind the sternum and also perioperative bleeding secondary to hepatic dysfunction can add to increased morbidity and mortality. Tricuspid annulus is less fibrous and very large in diameter, especially when patients have severe TR. The percutaneous prosthetic valve devices have a larger profile and can only be inserted through the jugular vein easily [58]. Several devices like Triclip (Abbott Lab), GATE System (NaviGate Cardiac Structures, Inc.), Trisol prosthesis (Trisol Medical Ltd., Inc., Yokneam, Israel), LUX-Valve (Jenscare Biotechnology), and many other percutaneous tricuspid valves are being developed and studied for implantability and durability.

Pulmonary valve insufficiency occurs generally after previous congenital cardiac surgery for tetralogy of Fallot or Ross procedures performed in both children and adults. As the insufficiency gets worse, patients do develop right heart failure symptoms. Redo cardiac surgery was done to implant a prosthetic valve, but with advent of percutaneous techniques, pulmonary valves are deployed more safely. Being on the right side of the heart, it is relatively easier to insert and chances of systemic emboli do not exist, as seen in the left side of the heart procedures. Melody (Medtronic) transcatheter pulmonary valve (TPV) therapy is a nonsurgical option to restore pulmonary valve function in patients having ventricular outflow obstruction or pulmonary regurgitation from bioprosthetic valve failures [59]. A total of 845 patients were followed for a period of 5.9 years after Melody valve implantation and post-implantation endocarditis of the deployed valve is a great concern resulting in high morbidity and mortality in the subgroup, when it occurs.

Conclusions

Technological advances in the field of medicine are occurring at a rapid pace. What was considered to be an impossible idea is now a reality. When there were no other options, both physicians and patients were forced to accept redo cardiac surgery at the expense of high cost, morbidity, and mortality. Since the first reported case of endovascular intervention in 1929, when Werner Forssmann introduced a ureteral catheter through his antecubital vein, use of catheter-based technologies has skyrocketed, and in less than 100 years, the field of cardiac surgery has changed well beyond what we could have imagined. Site-specific delivery of thrombotic agents, as well as thrombolytic agents in general, has helped to treat major conditions that previously could only be treated with open surgical intervention. If only site-specific delivery of high-dose antibiotics is shown to control infections, we will eliminate some more high-risk redo surgical procedures. While there is always going to be a role for redo cardiac surgery, it is prudent for younger surgeons to get equally trained and proficient in catheter-based procedures, failing which they will fall behind the rapidly advancing surgical trends.

Funding

None.

Compliance with ethical standards

Conflict of interest

The authors declare that they have no conflict of interest.

Ethical approval and consent to participate

Not applicable.

Statement of human and animal rights

Not applicable.