Transcatheter aortic valve implantation: Complications

INTRODUCTION — Aortic valve replacement (AVR) has been the mainstay of treatment of symptomatic severe aortic stenosis (AS). The role of transcatheter aortic valve implantation (TAVI; also known as transcatheter AVR or TAVR) as an alternative to surgical aortic valve replacement (SAVR) is evolving. Through both rapidly increasing clinical experience and progressive improvement in TAVI devices (eg, lower profile systems to reduce vascular complications), TAVI outcomes have improved. Ongoing studies continue to scrutinize the risks of TAVI complications and continuing efforts seek to minimize these risks.

Complications of TAVI will be considered in this topic commencing with immediate or periprocedural complications, which are usually apparent during or shortly after the procedure and moving to longer-term considerations. This topic will deal with periprocedural complications related to vascular access (including injury at the arterial access site, arterial tree trauma, and problems with vascular closure), valve deployment (including improper positioning, coronary compromise and annular rupture), valve function (including paravalvular leak), organ injury (including stroke, myocardial ischemia/injury, and acute kidney injury), and arrhythmic complications (including high degree heart block and atrial fibrillation) and late complications including aortic regurgitation and prosthetic valve thrombosis.

Indications, clinical outcomes, and procedural management for TAVI are discussed separately. (See "Choice of intervention for severe calcific aortic stenosis" and "Transcatheter aortic valve implantation: Periprocedural and postprocedural management".)

PERIPROCEDURAL COMPLICATIONS — Procedural complications include vascular issues (access-related iliofemoral and aortic complications), ventricular wall perforation (free wall or septal), valvular complications (annular rupture, valve malpositioning, mitral dysfunction, and paravalvular aortic regurgitation), arrhythmias (eg, conduction abnormalities and atrial fibrillation), coronary artery occlusion, myocardial infarction (MI), cerebrovascular accident, and death.

Complications that may occur or be detected after the procedure include acute kidney injury, high degree atrioventricular block requiring pacemaker implantation, myocardial injury, vascular (access-site or aortic) complications that went unrecognized during the procedure, tamponade after removal of temporary wire, valve migration or embolization, mitral valve dysfunction, and death.

Mortality — Various registries using different transcatheter heart valves (THV) have reported periprocedural mortality for TAVI ranging from 1.1 to 4.2 percent. In the PARTNER 1 trial [1], 30-day mortality was reported at 3.4 percent, whereas a similar figure (3.9 percent) was seen in patients at intermediate risk from the PARTNER IIA trial [2] and 2.2 percent in the SURTAVI trial [3]. In the large SAPIEN 3 registry, 30-day mortality was only 1.1 percent. The LOTUS cohort reported slightly higher periprocedural mortality at 4.2 percent, whereas the ACURATE neo reported a 3.3 percent mortality in 30 days [4]. In the STS-ACC-TVT registry, 30-day all-cause mortality during 2019 (72,991 TAVI procedures) was 2.9 percent [5].

Comparisons of mortality rates with TAVI for patients with bicuspid aortic stenosis (AS) and tricuspid AS are discussed separately. (See "Bicuspid aortic valve: Intervention for valve disease or aortopathy in adults", section on 'Transcatheter aortic valve implantation'.)

Even though conversion to open heart surgery due to TAVI complications is rare (1.4 percent) [6], it should be noted that overall mortality for cases that undergo emergency surgery due to cardiovascular complications carry a very poor prognosis: 45.8 percent mortality at 30 days [7].

Reduction of the mortality risk of TAVI requires a multifactorial and multidisciplinary approach by an effective heart valve team for appropriate case selection, careful procedural planning including choice of suitable access route, technique and appropriate TAVI valve type and size, effective operator training and experience, meticulous attention to vascular closure, and judicious aftercare including appropriate rehabilitative measures and drug therapy.

Access/vascular complications

Bleeding and access-related complications — Early and late (30 or more days post-procedure) bleeding complications are not infrequent after TAVI [8]. Much of the bleeding risk is access related.

Bleeding and other access-related complications may be detected during or following the procedure. The vast majority of TAVI procedures are performed via the transfemoral route [9], and most of the data on complications is for transfemoral access.

Risk — Reported major bleeding rates at 30 days after TAVI have ranged from 17 to 24 percent in extreme-risk cohorts to 11 percent for high-risk cohorts to 7 to 10 percent in international registries [10-12]. In the TVT registry [6], major bleeding rates were on a decline from 5.5 percent in 2012/2013 to 4.2 percent in 2014 and 2.93 percent in 2019 [5]. A similar pattern was seen with regard to life-threatening bleeding, which reduced from 6.3 percent before 2013 to 1.79 percent in 2019 [5]. In the large SAPIEN 3 registry, 30-day life-threatening or disabling bleeding was reported at 4.6 percent [6]. Bleeding early after TAVI is a risk factor for mortality at 30 days and at one to two years [12].

In a multicenter transfemoral TAVI registry from 2012, percutaneous access and closure was performed in 803 patients (81 percent) and a surgical strategy in 183 (19 percent) [13]. Incidence of major vascular complications, life-threatening/disabling bleeding, and major bleeding was 14.2, 11, and 17.8 percent, respectively. In the patient cohort with a completely percutaneous access strategy, major vascular complications and life-threatening/disabling bleedings were related to closure device failure in 64 and 29 percent, respectively.

Results from 1078 intermediate-risk patients enrolled in the SAPIEN 3 observational study show a significant reduction in the incidence of major vascular complications (6.1 percent) and life-threatening or disabling bleeding (4.6 percent), which is partially due to the decreasing size of the delivery system sheaths and accumulating operator experience. In the TVT registry [5,6], the rate of any vascular complications decreased from 5.6 percent in 2012/2013 to 4.2 percent in 2014 and 1.55 percent in 2019.

Risk factors — Risk factors for access-related or vascular complications include sheath-to-artery ratio, presence of circumferential calcification, severe tortuosity, and percutaneous preclosure device failure. In the multicenter registry study cited above, female sex, use of >19F system, peripheral arterial disease, and early stages of the learning curve were independent predictors for major or life-threatening/disabling bleeding [13]. Mechanical factors that may contribute to the risk of bleeding include the use of large delivery catheters with associated high risk of major vascular complications, which may cause bleeding [12], and the transapical approach with associated risk of myocardial tears [12]. While the first systems required 24 to 26F inner diameter sheaths, new generation TAVI devices with 14F inner diameter are now available. These technological advances, alongside accumulating operator experience, explain the dramatic reduction in major vascular complications from 11 to 14 percent in early studies [1,13] to 4 to 6 percent in large registries [6,14].

However, some patients receiving post-procedural blood transfusions have bleeding from sites not related to the procedure or no apparent source of bleeding.

Prevention and management — The optimal choice of percutaneous closure device is controversial and new devices are under development. Many devices can achieve adequate hemostasis; however, appropriate device selection is important according to the access site anatomy, device caliber, and individual operator experience. At the end of the procedure, contralateral control peripheral angiography to confirm hemostasis is recommended, allowing early intervention for complications.

The majority of bleeding/vascular access complications can be managed percutaneously. For this purpose, it is important to ensure continuous access to the punctured artery. This can be provided either by retrograde access from the contralateral side (often called cross-over) or by maintaining antegrade access with a wire following sheath removal. These options allow for placement of an occlusion balloon or a covered stent (self-expanding nitinol stent graft) [15].

A large propensity-matched study compared the incidence of major vascular complications in 944 TAVI patients (472 matched pairs) whose access was sealed percutaneously using Prostar-XL or Perclose ProGlide, the two most frequently used closure devices. Prostar was associated with higher rates of Valve Academic Research Consortium (VARC) major vascular complications (7.4 versus 1.9 percent) and higher rates of major bleeding (16.7 versus 3.2 percent) [16]. The SAFE MANTA IDE clinical trial, the first pivotal trial for a dedicated large-bore closure device using a novel collagen-based technology, reported VARC-2 major bleeding in 4.2 percent of patients.

Transapical tear/rupture — Severe access complications during transapical procedures are rare albeit challenging to manage surgically. The large SOURCE registry [17] summarizing 1394 transapical cases reported only five (0.4 percent) cases, whereas a similar rate (0.8 percent) was reported in the Italian transapical registry [18].

Aortic dissection/perforation — Aortic dissection/perforation occurs infrequently (0.2 percent) during the TAVI procedure [9] and can be caused by guidewire/delivery system manipulations, valve repositioning, retrieval, or retraction, frequently in the setting of suboptimal aortic measurements/visualization. Aortic dissection accounts for 8.9 percent of cases converting to open heart surgery [9]. The diagnosis and treatment of aortic dissection is discussed separately. (See "Clinical features and diagnosis of acute aortic dissection" and "Management of acute type B aortic dissection" and "Overview of acute aortic dissection and other acute aortic syndromes".)

Ventricular perforation — In general, ventricular perforation occurs in 1 percent of TAVI cases [6]. Right ventricular perforation is not an infrequent occurrence when a non-soft-tip active fixation temporary pacing wire is used and left in situ for 48 hours. In an early series, this complication accounted for 52.9 percent of cardiac tamponade cases post-TAVI [19].

Left ventricular rupture occurs rarely (eg, 0.27 percent of TAVI cases) and accounts for 18.9 percent of cases requiring emergency open-heart surgery [6] and 23.5 percent of post-TAVI tamponade cases [19]. This complication can be reduced by use of pre-shaped TAVI specific guidewires. Ventricular septal defects have also been reported post-TAVI, particularly with balloon expandable valves [20]. Their location is most often at the membranous or perimembranous part of the septum (79 percent) adjacent to the valve landing zone. When causing symptoms or significant RV overload, these should be treated percutaneously where possible.

Valve-related complications

Annular rupture — Aortic annular rupture is a rare (eg, 0.4 percent) life-threatening complication of TAVI [9,21-24]. It involves injury of the aortic root or left ventricular outflow tract (LVOT) may occur during balloon dilation of the native aortic valve, prosthetic valve deployment, or valve postdilatation for paravalvular regurgitation. Four types of annular rupture have been identified, including intra-annular, subannular, supra-annular, and combined. Risk factors include smaller annular size or sinotubular junction, bulky calcification of either valve leaflets or annulus, calcium in the LVOT or sinotubular junction, heavily calcified bicuspid valve, severe asymmetric subaortic LV hypertrophy, implantation of a balloon-expandable device, and aggressive balloon pre- or post-dilation [24]. It should be noted that annular rupture could also cause root hematoma and compress a coronary artery [25]. Management options range from emergent conversion to an open procedure with aortic root and valve replacement (accounting for 12.3 percent of all complications requiring conversion to open heart surgery [6]) to pericardial drainage and autotransfusion of small leaks to comfort care [24]. Annular rupture has not been reported with valve-in-valve procedures.

Meticulous valve sizing, scrutiny of calcium burden and location, and in particular avoiding oversized post-dilatation balloons can reduce the risk of this rare, yet potentially catastrophic complication.

Transcatheter heart valve malpositioning — THV malpositioning can often occur in the setting of inadequate ventricular pacing (eg, interrupted stimulation due to pacing lead dislocation) or inadequate visualization of destination morphology (eg, noncoaxial orientation of the prosthesis towards annulus-parallax). In the large TVT registry (n = 26,414) [6], THV migration occurred in 0.3 percent, THV embolized in the left ventricle in 0.3 percent, embolized in the aorta in 0.3 percent, and was retrieved in 0.4 percent of cases. Overall, valve migration accounted for 25.2 percent of cases requiring emergency open-heart surgery. New generation recapturable and repositionable valves (eg, Evolut R/PRO, Portico) have reduced the incidence of these catastrophic complications, which would often end up requiring urgent surgical intervention or complex percutaneous maneuvers.

Aortic regurgitation — Post-TAVI aortic regurgitation includes paravalvular and central (valvular) jets. These conditions are discussed below.

Mitral valve disruption — In the majority of patients, no significant changes occur in the degree of mitral regurgitation (MR) after TAVI [26]. However, operators should be mindful not to disrupt the papillary muscles by ensuring optimal wire placement in the left ventricle. If a self-expanding valve is positioned low in the LVOT, a worsening in MR can be observed [26]. The potential for worsening mitral stenosis after a deeply implanted TAVI valve should be sought in patients with heavy calcifications of the anterior mitral leaflet (eg, rheumatic valve).

Organ injury — Complications of TAVI include brain injury, kidney injury and myocardial injury.

Stroke and subclinical brain injury

Rates — The reported 30-day risk of stroke following TAVI in observational studies and clinical trials is 2 to 5 percent [21,27,28]. In a registry study of 176,316 patients undergoing TAVI during 2011 to 2019, the overall in-hospital stroke rate decreased slightly from 2.1 percent in the early part of the study period to 1.6 percent in 2019 [5]. Symptoms and signs of stroke are discussed separately. (See "Initial assessment and management of acute stroke" and "Initial evaluation and management of transient ischemic attack and minor ischemic stroke".)

Comparisons of stroke rates with TAVI for patients with bicuspid aortic stenosis (AS) and tricuspid AS are discussed separately. (See "Bicuspid aortic valve: Intervention for valve disease or aortopathy in adults", section on 'Transcatheter aortic valve implantation'.)

In a meta-analysis of 25 multicenter registries and 33 single-center studies [29], the pooled 30-day post-TAVI stroke rates between the transfemoral and transaortic approach were similar, estimated at 2.8 percent in multicentre studies. There was no difference in pooled 30-day stroke post-TAVI between the CoreValve and Edwards Valve in multicenter (2.4 [95% CI 1.9-3.2] versus 3 percent [95% CI 2.4-3.7]) and single-center studies (3.8 [95% CI 2.8-4.9] versus 3.2 percent [95% CI 2.4-4.3]). There was a decline in stroke risk with experience and technological advancement.

Data from the randomized cohort A of the PARTNER balloon-expandable TAVI trial in high risk patients showed that rates of strokes or transient ischemic attack were significantly higher after TAVI than after surgical aortic valve replacement (SAVR) at 30 days (5.5 versus 2.4 percent) and at one year (8.7 versus 4.3 percent), and there was a borderline significant trend at two years (11.2 versus 6.5 percent, p = 0.05) [1,30]. However, at five years, there was no significant difference in rate of stroke in the two groups (10.4 versus 11.3 percent) [31]. Similarly, at five years, the composite outcome of stroke and all-cause mortality was observed at similar rates in both groups (69.8 versus 62.9 percent).

In the large TVT registry capturing 26,414 TAVI procedures from 2012 to 2014 in the United States [9], periprocedural strokes were as low as 2.2 percent. A similar one-year stroke rate of 2.7 percent was observed in the SAPIEN 3 registry [14] of 1077 intermediate-risk patients recruited from 51 sites in the United States and Canada.

Subclinical brain injury may be substantially more frequent than stroke as suggested by the following studies:

●A diffusion-weighted magnetic resonance imaging (MRI) study that compared 32 patients who underwent TAVI with 21 historical controls treated with SAVR [32]. New clinically silent cerebral foci of restricted diffusion were significantly more frequent in patients who had undergone retrograde aortic valve implantation than in those who had undergone SAVR (84 versus 48 percent).

●Similarly, high rates of new cerebral ischemic lesions detected by diffusion-weighted MRI were found in a study comparing transfemoral versus transapical TAVI [33]. Similar frequencies of new lesions were found in 19 of 29 (66 percent) patients in the transfemoral group and in 22 of 31 (71 percent) patients in the transapical group. Of note, the patients in the transapical group had a higher frequency of peripheral vascular disease than in the transfemoral group (45 versus 17 percent), although the frequency of aortic plaque ≥4 mm was similar (29 and 28 percent). Most patients (76 percent) with new lesions had multiple lesions (median 3, range 1 to 31). The lesions were largely clinically silent, though one patient in each group had a clinically evident stroke within 24 hours after the procedure.

Further studies are required to determine the clinical significance of these observations. Although most of the new lesions were clinically silent, there is concern that silent cerebral infarcts are associated with subtle cognitive change and with an increased risk of subsequent dementia. (See "Clinical diagnosis of stroke subtypes", section on 'Silent brain infarcts'.)

Effect of cerebral embolic protection — The effects of the use of cerebral embolic protection systems during TAVI are discussed separately. (See "Transcatheter aortic valve implantation: Periprocedural and postprocedural management", section on 'Role of cerebral embolic protection'.)

Myocardial ischemia/injury

Coronary obstruction — One cause of ischemia and hypotension immediately following TAVI is coronary ostia occlusion, which may be treated by valve retrieval/repositioning, percutaneous coronary intervention, or coronary artery bypass graft surgery.

With native aortic valve procedures — Coronary obstruction is a rare complication of TAVI for native aortic valve stenosis, occurring in approximately 0.7 percent of procedures [10]. In a 2013 systematic review of coronary artery obstruction following TAVI for native aortic valve stenosis, a total of 24 patients (83 percent women) with coronary obstruction were identified from 18 publications, with the majority of complications occurring with the balloon-expandable valve (88 percent) [34]. Percutaneous coronary revascularization was successful in 21 of 23 patients in whom it was attempted, with 30-day mortality of 8 percent (two deaths out of 24 patients) following this complication.

Coronary artery occlusion is more common with low coronary artery ostia heights (<10 mm), small sinuses, and with bulky heavy asymmetric valve calcification. Patient evaluation before TAVI should include measurement of the distance from the aortic annulus to coronary ostium and sinus size by computed tomography imaging or 3-D transesophageal echocardiography, as well as assessment of valve calcification. Coronary angiography at the time of TAVI also allows visualization of the annular-coronary distance. (See "Imaging for transcatheter aortic valve implantation", section on 'Coronary ostia'.)

With valve-in-valve procedures — Coronary obstruction is much more frequent with valve-in-valve procedures than with native aortic valve procedures (3.5 versus 0.7 percent) [35]. The major risk factor for coronary obstruction with valve-in-valve is the proximity of the coronary artery ostia to the bioprosthetic leaflets and posts [35]. Thus, risk factors related to the surgical bioprosthetic valve include supra-annular position, high leaflet profile, internal stent frame (eg, Mitroflow, Trifecta), and no stent frame (homograft, stentless valves); and transcatheter valve factors include an extended sealing cuff and high implantation. Anatomic factors are similar to those for native aortic valve procedures, including low-lying coronary ostia and low or narrow sinuses of Valsalva. The main predisposing factor in the setting of valve in valve procedures is the proximity of the coronary ostia to the anticipated final position of the displaced bioprosthetic leaflets after THV implantation [36]. Meticulous fluoroscopic and cardiac computed tomography assessment may identify most cases at risk. This may become a particularly important consideration for the index (initial) implant and device if TAVI valves suffer late leaflet degeneration and require a repeat procedure. (See "Imaging for transcatheter aortic valve implantation", section on 'Coronary ostia'.)

Delayed coronary obstruction — In a large multicenter TAVI dataset (n = 17,092) [37], the incidence of delayed coronary obstruction (DCO) was 0.22 percent. DCO occurred more commonly after valve-in-valve procedures (0.89 versus 0.18 percent; p<0.001). The most frequent presentation was cardiac arrest (31.6 percent, n = 12), followed by ST-segment elevation MI (23.7 percent, n = 9). The left coronary artery was obstructed in most cases (92.1 percent, n = 35). The overall in-hospital death rate was 50 percent (n = 19).

Subclinical myocardial injury — While clinically apparent MI following TAVI is relatively rare, with reported rates in the range of 1 percent [38,39], myocardial injury that does not meet the universal definition of MI appears to be more common, with reported troponin elevations in as many as 95 to 100 percent of patients [40,41].

●In a single-center cohort of 150 patients (mean age 85 years) who underwent TAVI by either the transfemoral (103 patients) or transapical (47 patients) approach, 95 percent of patients had twofold or greater elevations in troponin I post-TAVI (mean 5.9 ng/mL), with higher elevations in patients who had TAVI via the transapical approach [41]. On multivariate analysis, higher troponin I level was associated with greater risk of 30-day mortality (hazard ratio [HR] 1.15, 95% CI 1.05-1.26), although with minimal prognostic accuracy given its low area under the curve (AUC = 0.57) on receiver-operating curve analysis.

●In a single-center cohort of 62 patients (mean age 80 years) who underwent TAVI and survived for greater than 24 hours without major procedural complications, cardiac troponin I was measured serially post-TAVI (6, 12, 24, 48, and 72 hours) and patients were followed for one year [40]. Troponin I was elevated in all patients (median peak troponin I level 3.8 ng/mL at 12 hours post-TAVI); however, no patient met the definition for MI, and there was no correlation between peak troponin level and one-year mortality post-TAVI.

In a small study of 37 patients undergoing TAVI with a balloon expandable valve [42], cardiac MRI was performed a week prior to the procedure and within 30 days following TAVI. In non-transapical TAVIs (n = 26), there was no new myocardial necrosis detected with cardiac MRI. In the transapical group, there was a new focal myocardial necrosis in the apex, with a median myocardial extent and necrotic mass of 5 percent (2 to 7) and 3.5 g (2.3 to 4.5), respectively.

Acute kidney injury — Acute kidney injury (AKI) frequently occurs following TAVI and has been related to a worse outcome [43]. In the large TVT registry, stage 3 AKI was observed in 3.64 percent of cases before or during 2013, which declined to 1.07 percent of cases in 2019. Similarly, in cases before or during 2013, 2.24 percent required dialysis, which declined to 0.58 percent for cases in 2019 [5]. For patients undergoing TAVI, general recommendations for the prevention of contrast-induced nephropathy apply, as discussed separately. (See "Prevention of contrast-associated acute kidney injury related to angiography".)

Limited data are available specifically addressing interventions to reduce the risk of AKI following TAVI. The following examples are consistent with other data on prevention of contrast-induced nephropathy:

●In the PROTECT-TAVI (PROphylactic effecT of furosEmide-induCed diuresis with matched isotonic intravenous hydraTion in Transcatheter Aortic Valve Implantation) open-label trial, a total of 112 consecutive patients undergoing TAVI were randomly assigned to intravenous hydration with normal saline solution matched to urine output with furosemide-induced diuresis (RenalGuard group) or normal saline solution (control group) [44]. The rate of AKI was significantly lower in the RenalGuard group than in the control group (n = 3 [5.4 percent] versus n = 14 [25 percent], respectively). Limitations of this approach are discussed separately. (See "Prevention of contrast-associated acute kidney injury related to angiography".)

●In an observational study, there was no advantage of iso-osmolar contrast media compared with low-osmolar contrast media in protecting TAVI patients from AKI [45]. Comparisons between contrast media are discussed in detail separately. (See "Prevention of contrast-associated acute kidney injury related to angiography", section on 'Dose and type of contrast agent'.)

Arrhythmic complications

High degree heart block — Risk factors for development of atrioventricular heart block with need for permanent pacemaker implantation include preexisting right bundle branch block and use of a CoreValve (versus SAPIEN valve) [21,46]. In the large TVT registry, a new pacemaker or intracardiac defibrillator was implanted in 10.93 percent of TAVI cases before or during 2013, and a similar rate of 10.84 percent was observed for TAVI cases in 2019 [5]. In patients receiving the SAPIEN valve, the risk of this complication (1.8 to 8.5 percent) is similar to that observed following balloon aortic valvotomy or SAVR [47]. For example, in cohort B of the PARTNER trial, the rate of new pacemaker requirement at 30 days was similar in the balloon-expandable TAVI and standard therapy (including balloon valvotomy) groups (3.4 and 5 percent) [48] and in cohort A was also similar in the TAVI and surgical groups (3.8 versus 3.6 percent) [1]. The risk of requiring a permanent pacemaker was higher (ranging from 17 to 42.5 percent) in patients who received the first-generation, self-expanding CoreValve, the LOTUS, and the Direct Flow valves [21,49-52]. In the new generation SAPIEN 3 valve, however, due to the addition of a skirt in the valve design, aiming to reduce paravalvular leak, the need for new pacemaker has increased to 10.2 percent [14]. In the Medtronic Evolut R UK & Ireland Implanter’s Registry (n = 264) and the Evolut R US Study (n = 241), the pacemaker implantation rate was approximately 17 percent, whereas in the ACURATE neo registry, the pacemaker implantation rate was 10.2 percent. In the ATLAS registry, rates of new pacemaker implantation were 21.1 and 11.9 percent for Evolut R and Evolut PRO, respectively [53].

A systematic review and meta-analysis found that new-onset LBBB post-TAVI was associated with increased need for permanent pacemaker implantation (RR 2.18, 95% CI 1.28-3.70) and increased risk of cardiac death (RR 1.39, 95% CI 1.04-1.86) at one-year follow-up [54].

The frequency of permanent pacemaker implantation after valve-in-valve procedure is lower than that for treatment of native aortic valve disease [35]. Of note, outcomes in patients with and those without a new pacemaker were similar in a retrospective analysis of 883 patients from the FRANCE-2 registry [55].

Management of atrioventricular block is discussed separately. (See "Permanent cardiac pacing: Overview of devices and indications" and "Third-degree (complete) atrioventricular block".)

New atrial fibrillation — New onset atrial fibrillation (NOAF) was seen in 6.3 percent of patients undergoing TAVI in the TVT registry [6]. A similar incidence (7.2 percent) was noted in the SOURCE XT [56] registry of 2706 patients. In the PARTNER IIA trial (balloon expandable valve), NOAF incidence was significantly lower in the TAVI versus surgical group (9.1 versus 26.4 percent). Similar rates were seen in the United States CoreValve trial with TAVI patients developing NOAF at a significantly lower rate of 11.7 percent compared with 30.5 percent in the surgical group [49]. Similar predictors of NOAF post-TAVI include left atrium size and transapical approach [57]. At 30-day follow-up, NOAF was associated with a higher rate of stroke/systemic embolism (13.6 versus 3.2 percent, p = 0.021, p = 0.047 after adjustment for baseline differences between groups), with no differences in mortality rate between groups (NOAF: 9.1 percent, no-NOAF: 6.4 percent, p = 0.57). In the large SOURCE XT registry [56], independent predictors of NOAF were age (HR 1.1), New York Heart Association class III or IV (HR 1.9), nontransfemoral access route (HR 3), and balloon postdilation (HR 1.6). No interaction was observed between any degree of postimplantation paravalvular leak and NOAF.

Management of NOAF is discussed separately. (See "Atrial fibrillation: Overview and management of new-onset atrial fibrillation".)

CAUSES OF PERIPROCEDURAL SHOCK — Shock commonly occurs during or immediately following TAVI and may be triggered by anesthesia, volume depletion (eg, due to bleeding related to access site issues, annular rupture or aortic injury), rapid pacing, tamponade, myocardial ischemia, acute aortic regurgitation, and interruption in cardiac output during valve implantation [21].

Periprocedural hypotension or shock could be caused by the following conditions:

●Volume depletion: Generally secondary to a vascular (usually arterial) complication

•Access complication (primary or secondary)

•Annular rupture

•Aortic dissection

●External compression of right or left ventricle (tamponade)

•Right ventricular perforation by the temporary wire

•Left ventricular perforation or rupture with the stiff wire

•Aortic root rupture with associated tamponade

●Flow disturbances due to intracardiac causes

•Ventricular septal defect (causing left to right shunt).

•Papillary muscle rupture or compromise of mitral valve function with severe mitral regurgitation or mitral stenosis.

•Severe aortic regurgitation after aortic valve predilatation, after TAVI implantation due to low/high implant, and/or large paravalvular leak.

•Intracavity or left ventricular outflow tract gradient secondary to hypertrophied left ventricle with small cavity size.

●Conduction abnormalities/arrhythmias

•Bradycardia: Asystole, high atrioventricular block (eg, complete heart block)

•Tachycardia: Ventricular fibrillation, ventricular tachycardia, supraventricular tachycardias, new-onset atrial fibrillation

●Nervous system

•Reduced sympathetic tone (anaesthesia, sedation)

•Parasympathetic overdrive (eg, vagal response)

Management depends on the cause and includes correction of volume depletion; management either percutaneously or surgically of any vascular access complication, pericardial drain in cases of tamponade; open-heart surgery for the treatment of catastrophic complications (rupture, perforation, dissection); post-dilatation or valve-in-valve procedure for malapposed or misplaced valve; pacing for bradycardia; DC cardioversion and pharmacotherapy for tachyarrhythmias; positive inotropic agents, vasopressors, and mechanical circulatory support or intraaortic balloon circulatory support as needed to counteract anaesthetic drug effects. Elective cardiopulmonary bypass or extracorporeal membrane oxygenation is an option for patients at severe risk for hemodynamic instability. (See "Short-term mechanical circulatory assist devices".)

Rarely, shock may result from left ventricular outflow tract or intracavitary obstruction precipitated by volume depletion and/or positive inotropic agents. While hypotension with left ventricular outflow tract obstruction is observed in patients with hypertrophic cardiomyopathy (HCM), this hemodynamic response is not limited to patients with HCM [58]. This condition should be diagnosed promptly (generally by echocardiography), as it is treated by administration of volume and a beta-adrenergic receptor blocking agent.

LONGER-TERM COMPLICATIONS — Longer-term complications include paravalvular regurgitation, valve thrombosis, valve obstruction, infective endocarditis, bleeding, and death.

Post-TAVI aortic regurgitation

Paravalvular regurgitation — Mild or greater paravalvular aortic regurgitation (PVR) is commonly detected following TAVI. Moderate or severe PVR following TAVI was associated with a threefold increase in 30-day mortality and a 2.3-fold increase in one-year mortality following TAVI [59]. Mild PVR was associated with worse outcomes in some studies [30,31,60], but not in others [61-63]. It remains unclear if paravalvular aortic regurgitation itself leads to adverse outcomes or whether this finding identifies a higher-risk patient subgroup.

The reported incidence of PVR following TAVI varies widely from 7 to 70 percent for mild PVR and 0 to 24 percent for moderate to severe PVR [59,64]. In a meta-analysis of 12,926 patients treated between 2002 and 2012 with first generation Corevalve and SAPIEN valves, the pooled reported estimate for moderate or severe aortic regurgitation (AR) post-TAVI was 11.7 percent, most commonly seen with the Corevalve (16 versus 9.1 percent). In the large TVT registry, the percentage of patients post-TAVI with moderate or severe PVR has declined over the years, with a reported incidence of 7.96 percent for cases before or during 2013, which fell to 1.57 percent in 2019. Significantly lower rates of moderate or severe paravalvular AR have been reported for newer-generation valves, namely 1.9 percent for the LOTUS [65], 1.5 percent for the SAPIEN 3 [14], 7.7 and 5.3 percent with the Evolut R (UK Evolut R and US trials, respectively) [66,67], and 3.8 percent for Evolut PRO [53].

A study from the TVT registry on self-expanding valves spanning three generations showed a decrease in PVR in matched populations from 8.3 percent when the first-generation device was used to 5.4 percent with the second- and 3.4 percent with the third-generation device [68].

Comparison of PVR rates for TAVI in patients with bicuspid AS versus tricuspid AS are discussed separately. (See "Bicuspid aortic valve: Intervention for valve disease or aortopathy in adults", section on 'Transcatheter aortic valve implantation'.)

Variation in observed rates of PVR may be related to differences in populations, procedures, timing of assessment, transcatheter heart valves (THVs) used, and methods of assessment of PVR. PVR is caused by incomplete apposition of the prosthesis with the aortic annulus due to inadequate inflation of the prosthesis or calcific deposits that prevent proper valve seating. Contributing factors include a heavily calcified annulus, an undersized prosthesis, and malpositioning of the THV [69].

Diagnosis of PVR is challenging, and multiple modalities may be used, including Doppler echocardiography, angiography, cardiac magnetic resonance (CMR), and hemodynamics. While operators most commonly use echocardiography, the addition of CMR may be helpful for accurate assessment of AR. (See "Imaging for transcatheter aortic valve implantation", section on 'Aortic regurgitation'.)

●Echocardiography is the main modality used for detection and assessment of PVR. Transesophageal echocardiography is frequently used during TAVI procedure performed with general anesthesia, while transthoracic echocardiography is used if TAVI is performed under conscious sedation. Transthoracic echocardiography is generally used for follow-up. Assessment by echocardiography may be difficult due to presence of multiple, eccentric, irregularly-shaped jets that may be aligned along off-axis imaging planes. Imaging of the jets may be suboptimal due to acoustic shadowing from calcifications or from the THV. In addition, differences in echocardiography grading schemes may contribute to variability in interpretation [64].

●If echocardiogram findings are inconclusive, CMR enables highly reproducible measurement of regurgitation volumes independent of regurgitant jet shape and number [64]. A study of 135 TAVI patients [70] sought to evaluate the effect of AR assessed with CMR on clinical outcomes post-TAVI. Higher regurgitant fraction (RF) post-TAVI was associated with increased mortality (hazard ratio [HR] 1.18 for each 5 percent increase in RF, 95% CI 1.08-1.30) and the combined end point of mortality and rehospitalization for heart failure (HF; HR 1.19 for each 5 percent increase in RF, 95% CI 1.15-1.23). However, CMR is less widely available than the other modalities.

●Angiography is routinely performed at the end of the TAVI procedure; however, it is operator dependent and is difficult to standardize due to variation of dye volume, catheter position, and imaging projection. Also, angiography requires administration of exogenous contrast that may not be well tolerated in patients who have kidney disease.

●A more reliable way of assessing PVR severity is the aortic regurgitation index (ARI), which is the ratio of the end-diastolic aortoventricular pressure difference to systolic blood pressure:

The ARI is inversely correlated to the severity of PVR and an index <25 is a predictor of one-year mortality risk in patient with PVR [71]. Factors that influence the ARI besides regurgitation include ventricular and aortic compliance and heart rate [64].

Diastolic aortic pressure is widely used as a crude marker for the degree of AR.

Central regurgitation — Central regurgitation may be caused by improper valve sizing or deployment [21]. Minor valve displacement may resolve with gentle probing of the leaflets but severe leaks are an indication for valve-in-valve deployment.

Valve thrombosis — Symptomatic or hemodynamically significant valve thrombosis is rare, occurring in <1 percent of patients undergoing TAVI [72,73]. Clinical manifestations include exertional dyspnea and increased transvalvular gradients. In such cases, thrombus may or may not be visualized by transesophageal echocardiography.

Evidence is emerging on the frequency and clinical significance of subclinical valve leaflet thrombosis of bioprosthetic valves following TAVI or surgical aortic valve replacement (SAVR) [74,75]. In the largest study to date, 931 patients with bioprosthetic valves enrolled in two registries were studied by a four-dimensional computed tomography (CT) imaging protocol at varying intervals after TAVI or SAVR (median time 83 days); 890 had interpretable CT scans [75]. Subclinical leaflet thrombosis was defined as the presence of reduced leaflet motion along with hypoattenuating valve lesions on CT. (See "Antithrombotic therapy for surgical bioprosthetic valves and surgical valve repair", section on 'Bioprosthetic valve thrombosis'.)

●Subclinical leaflet thrombosis was identified in five (4 percent) of patients with surgical valves and 101 (13 percent) of 752 patients with transcatheter valves.

●Subclinical leaflet thrombosis was significantly less frequent among patients receiving anticoagulants (4 percent) compared with patients receiving dual antiplatelet therapy (15 percent). Rates were similar with direct oral anticoagulants (DOACs) and warfarin (3 and 4 percent).

●Subclinical leaflet thrombosis resolved in all 36 patients treated with anticoagulants (warfarin or DOAC), and persisted in 20 (91 percent) of 22 patients not treated with anticoagulants.

●Aortic valve gradients of greater than 20 mmHg and increases in aortic valve gradients of more than 10 mmHg were more frequent among patients with subclinical leaflet thrombosis than among those with normal leaflet motion (14 versus 1 percent).

●Subclinical thrombosis was associated with significantly increased rates of transient ischemic attacks (TIAs; 4.18 versus 0.6 per 100 person years) and all strokes or TIAs (7.85 versus 2.36 per 100 person years) but not a significantly increased rate of stroke (4.12 versus 1.92 per 100 person years).

Four-dimensional volume-rendered computed tomography (4DCT) has revealed the presence of subclinical leaflet thrombosis in a significant number of patients who received surgical or transcatheter bioprosthetic aortic valves. Among 890 patients from the RESOLVE and SAVORY registries, it was reported that 5 of 138 patients (3.8 percent) had subclinical leaflet thrombosis after SAVR and 101 of 752 patients (13.4 percent) after TAVI [76]. This valve leaflet thickening and reduced leaflet motion, with reference to their CT appearance, have been referred to as hypoattenuating leaflet thickening (HALT) and the more severe hypoattenuation affecting motion (HAM), respectively. Of interest, these states of subclinical valve thrombosis appear to be dynamic (progressing and regressing over the course of months) [77]. The clinical impact of subclinical valve thrombosis remains unclear.

Further investigation is required to determine the incidence, clinical significance, and appropriate management of subclinical bioprosthetic valve (including transcatheter valve) leaflet thrombosis. Randomized trials are underway to assess the safety and efficacy of anticoagulation compared with the standard of antiplatelet therapy following TAVI. Both vitamin K antagonists and novel oral anticoagulants seem to protect against and be an effective treatment of TAVI valve thrombosis [78]. (See "Bioprosthetic valve thrombosis, thromboembolism, and obstruction: Management".)

Valve obstruction — The incidence of valve hemodynamic deterioration (VHD) after TAVI [79], defined as an absolute increase in mean transprosthetic gradient >10 mmHg between discharge and last follow-up, in a multicenter cohort of 2418 patients was 4.5 percent (overall VHD) and 2.8 percent within the first year (early VHD). The mean transprosthetic gradient in patients with VHD increased from 9.5±5 mmHg at hospital discharge to 26.1±11 mmHg at follow-up. Independent predictors of transcatheter valve VHD at follow-up included absence of anticoagulation at discharge, Valve-in-valve procedure, ≤23 mm THV size and increased body mass index. Marked increases in gradients on transthoracic echocardiography post-TAVI invariably imply the possibility of valve thrombosis necessitating empirical oral anticoagulation, which is often effective at reducing transprosthetic gradients [72]. (See "Bioprosthetic valve thrombosis, thromboembolism, and obstruction: Clinical manifestations and diagnosis" and "Bioprosthetic valve thrombosis, thromboembolism, and obstruction: Management".)

Prosthesis-patient mismatch (PPM) following SAVR or TAVI is a cause of valve obstruction and is associated with adverse outcomes. A study of data from the Society of Thoracic Surgeons/American College of Cardiology Transcatheter Valve Therapy registry found that severe PPM was present in 12 percent and moderate PPM in 25 percent of 62,125 patients undergoing TAVI between 2014 and 2017 [80]. Independent predictors of severe PPM included small (≤23 mm diameter) valve prosthesis, valve-in-valve procedure, larger body surface area, female sex, younger age, lower ejection fraction, atrial fibrillation, severe mitral or tricuspid regurgitation, and being from a non-White/Hispanic population. At one year, rates of mortality as well as HF rehospitalization were significantly higher in patients with severe PPM. Mortality at one year was 17.2, 15.6 and 15.9 percent of patients with severe, moderate, and no PPM, respectively. HF rehospitalization occurred in 14.7, 12.8 and 11.9 percent of patients with severe, moderate, and no PPM.

Late bleeding — Major bleeding 30 or more days after TAVI is frequent (eg, 5.9 percent with median onset at 132 days in PARTNER cohorts and registries) and a strong independent risk factor for mortality between 30 days and one year (adjusted HR 3.91, 95% CI 2.67-5.71) in the PARTNER randomized cohorts and continued access registries [8]. The most frequent types of major late bleeds were gastrointestinal (40.8 percent), neurological (15.5 percent), and traumatic fall-related (7.8 percent).

Prosthetic valve endocarditis — All patients with prosthetic valves, including those who have undergone TAVI, are considered among those at highest risk for infective endocarditis and therefore prophylaxis for bacterial endocarditis is suggested for high-risk procedures. Recommendations for antibiotic prophylaxis for endocarditis are discussed in detail separately. (See "Prevention of endocarditis: Antibiotic prophylaxis and other measures".)

Following TAVI, prosthetic valve infective endocarditis (PVE) affecting the THV may occur during early (within 60 days), intermediate (between 60 and 365 days), or late (after 365 days) time periods [81]. Reported rates of early PVE have ranged from 0.3 to 3.4 percent per patient-year [81-84]. Complications include paravalvular invasion (eg, aortic root or paravalvular abscess), embolic events, valve stenosis, and valve and paravalvular regurgitation. HF is common (occurring in one-third of patients in one multicenter series [81]). Concomitant mitral valve involvement is common (in 14 and 24 percent of cases in two series [81,82]), perhaps at least partly related to contact between a low-seated TAVI and the mitral apparatus. In a multicenter registry [84] of 7944 TAVI patients, 53 suffered PVE following their procedure (0.5 percent at first year). Of those, only 11 percent underwent intervention, either surgical excision or percutaneous valve in valve implantation. In-hospital mortality rate for patients with PVE was 47.2 percent. Most frequently encountered microorganisms were coagulase-negative staphylococci (24 percent), followed by Staphylococcus aureus (21 percent) and enterococci (21 percent) [84].

In another review including 34 cases of TAVI-associated PVE (29 definite, five possible), 20 patients were treated with antibiotic therapy only; of these, 13 were discharged home (with one later dying of sepsis at three months) [82]. Of 14 patients treated surgically (13 with THV removal and SAVR), nine were discharged home.

The limited available data suggest that considerations similar to those operative in the treatment of surgically-placed aortic valve PVE may apply to treatment of TAVI-related PVE, including careful analysis of risk/benefit of surgical intervention. (See "Surgery for prosthetic valve endocarditis".)

COMPARISON BETWEEN TAVI AND SAVR — Randomized trials have shown that the complication profiles of TAVI and surgical aortic valve replacement (SAVR) differ. TAVI is generally associated with lower rates of major bleeding and atrial fibrillation, but higher rates of short-term aortic valve reintervention, pacemaker implantation, and aortic regurgitation compared with SAVR.

As an example, a meta-analysis of trials with patients with predominantly intermediate surgical risk found that transfemoral TAVI was associated with lower mortality risk and acute kidney injury compared with SAVR. TAVI resulted in lower rates of major bleeding and atrial fibrillation compared with SAVR [85]. On the other hand, TAVI was associated with increased rates of short-term aortic valve reintervention, permanent pacemaker insertion, moderate or severe symptoms of HF, and presence of at least moderate aortic regurgitation. This analysis and other outcome comparisons are discussed in detail separately. (See "Choice of intervention for severe calcific aortic stenosis", section on 'In intermediate-risk patients'.)

Fewer randomized trial data are available on the comparative risk of complications for alternate (nontransfemoral) access TAVI versus SAVR. The best available data come from the transthoracic cohort of the PARTNER 2A trial, but this cohort included only 482 patients [2]. Mortality and stroke rates were nominally but not significantly higher in the TAVI group compared with the SAVR group (mortality 25.2 versus 20.7 percent and stroke 12.9 versus 7.9 percent at two years). Major bleeding was more frequent in the SAVR group (54.1 versus 29.6 percent at two years). Aortic valve reintervention was significantly more frequent in the TAVI group (2 versus 0 percent).

COMPLICATIONS OF VALVE-IN-VALVE PROCEDURE — Complications of valve-in-valve implantation for failed bioprosthetic valve are similar to those for TAVI for native aortic valve stenosis. However, rates of paravalvular regurgitation and need for permanent pacemaker implantation are much lower for valve-in-valve procedures and annulus rupture is not a reported complication of valve-in-valve implantation [35]. Coronary artery obstruction is more frequent with valve-in-valve procedures than with native aortic valve procedures, as discussed above. (See 'Coronary obstruction' above.)

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Transcatheter aortic valve implantation".)

SUMMARY AND RECOMMENDATIONS

●Complications of transcatheter aortic valve implantation (TAVI) include periprocedural complications related to vascular access (including injury at the arterial access site, arterial tree trauma, ventricular perforation, and vascular closure device failure), valve deployment (including malpositioning, coronary compromise and annular rupture), valve function (including paravalvular leak), organ injury (including stroke, myocardial ischemia/injury, and acute kidney injury), and arrhythmic complications (including high degree heart block and atrial fibrillation); long-term complications include aortic regurgitation, prosthetic valve thrombosis, late bleeding, and prosthetic valve endocarditis. (See 'Introduction' above and 'Periprocedural complications' above and 'Longer-term complications' above.)

●Periprocedural mortality for TAVI has ranged from 1.1 to 4.2 percent in registry reports. Reduction of the mortality risk of TAVI requires a multifactorial and multidisciplinary approach by an effective heart valve team for appropriate case selection, careful procedural planning including choice of suitable access route and technique and appropriate TAVI valve type and size, effective operator training and experience, meticulous attention to vascular closure, and judicious aftercare including appropriate rehabilitative measures and drug therapy. (See 'Mortality' above.)

●Hypotension may develop during or following a TAVI procedure due to volume depletion, generally secondary to a vascular complication, tamponade of the right or left ventricle, flow disturbances due to intracardiac causes (eg, ventricular septal defect, acute valve dysfunction or left ventricular gradient), conduction abnormality, arrhythmia, or an alteration in sympathetic or parasympathetic tone. (See 'Causes of periprocedural shock' above.)

●At least mild paravalvular aortic regurgitation is commonly detected following TAVI. The presence of moderate or severe paravalvular aortic regurgitation has been associated with increased one-year mortality rates. It remains unclear if paravalvular aortic regurgitation itself leads to adverse outcomes or whether this finding identifies a higher-risk patient subgroup. (See 'Paravalvular regurgitation' above.)

●The complication profiles of TAVI and surgical aortic valve replacement (SAVR) differ. TAVI is generally associated with lower rates of major bleeding and atrial fibrillation, but higher rates of short-term aortic valve reintervention, pacemaker requirement, and paravalvular aortic regurgitation compared to SAVR. (See 'Comparison between TAVI and SAVR' above.)