Author + information
- Received September 21, 2016
- Revision received January 23, 2017
- Accepted January 27, 2017
- Published online April 3, 2017.
- John Jose, MD, DMa,b,
- Dmitriy S. Sulimov, MDa,
- Mohamed El-Mawardy, MDa,
- Takao Sato, MDa,c,
- Abdelhakim Allali, MDa,
- Erik W. Holy, MDa,
- Björn Becker, MDa,
- Martin Landt, MDa,
- Julia Kebernik, MDa,
- Bettina Schwarz, MDa,
- Gert Richardt, MDa and
- Mohamed Abdel-Wahab, MDa,∗ ()
- aHeart Center, Segeberger Kliniken (Academic Teaching Hospital of the Universities of Kiel, Lübeck, and Hamburg), Bad Segeberg, Germany
- bChristian Medical College Hospital, Vellore, Tamil Nadu, India
- cTachikawa General Hospital, Nagaoka, Japan
- ↵∗Address for correspondence:
Dr. Mohamed Abdel-Wahab, Heart Center, Segeberger Kliniken, Am Kurpark 1, 23795 Bad Segeberg, Germany.
Objectives The aim of this study was to determine the incidence, characteristics, and treatment outcomes of patients diagnosed with clinical transcatheter heart valve thrombosis.
Background Limited data exists on clinical or manifest transcatheter heart valve thrombosis. Prior studies have focused on subclinical thrombosis.
Methods A retrospective analysis was conducted of prospectively collected data from a single-center registry that included 642 consecutive patients who underwent transcatheter aortic valve replacement between 2007 and 2015 (305 patients had self-expanding valves; balloon-expandable, n = 281; mechanically expanding, n = 56). Long-term oral anticoagulation (OAC) was indicated in 261 patients, while 377 patients received dual-antiplatelet therapy post-procedure. All patients underwent scheduled clinical and echocardiographic follow-up.
Results The overall incidence of clinical valve thrombosis was 2.8% (n = 18). No patient on OAC developed thrombosis. Of the detected thrombosis cases, 13 patients had balloon-expandable, 3 had self-expanding, and 2 had mechanically expanding valves. Thrombosis occurred significantly more often with balloon-expandable valves (odds ratio: 3.45; 95% confidence interval: 1.22 to 9.81; p = 0.01) and following valve-in-valve procedures (odds ratio: 5.93; 95% confidence interval: 2.01 to 17.51; p = 0.005). Median time to diagnosis of valve thrombosis was 181 days. The median N-terminal pro–brain natriuretic peptide level was 1,318 pg/ml (interquartile range: 606 to 1,676 pg/ml). The mean transvalvular gradient and valve area were 34 ± 14 mm Hg and 1.0 ± 0.46 cm2, respectively. Computed tomography showed hypoattenuating areas with reduced leaflet motion. Initiation of OAC resulted in significant reduction of transvalvular gradient and clinical improvement. No deaths were related to valve thrombosis.
Conclusions Clinical transcatheter heart valve thrombosis is more common than previously considered, characterized by imaging abnormalities and increased gradients and N-terminal pro–brain natriuretic peptide levels. It occurred more commonly after balloon-expandable transcatheter aortic valve replacement and valve-in-valve procedures. OAC appeared to be effective in the prevention and treatment of valve thrombosis. Randomized control trials are needed to define optimal antithrombotic therapy after transcatheter aortic valve replacement.
- aortic stenosis
- clinical thrombosis
- oral anticoagulation
- transcatheter aortic valve replacement
- transcatheter heart valve thrombosis
Transcatheter aortic valve replacement (TAVR) is an innovative, minimally invasive procedure that has become the standard of care for patients with severe aortic stenosis who are either inoperable or at high risk. With short-term efficacy and safety established in several trials in the past decade (1–4), recent research efforts have focused on durability and long-term outcomes after TAVR. There are few studies on modes of transcatheter heart valve (THV) failure, such as endocarditis or thrombosis, and consequently the current state of knowledge on this important aspect is fragmented. Although more than 200,000 TAVRs have been performed worldwide in the past 10 years (5), THV thrombosis remains a poorly characterized phenomenon.
Several sporadic cases of THV thrombosis were recently reported and have generated significant research interest on this entity (6–12). The most data so far on THV thrombosis were recently published by Latib et al. (13), reporting an overall incidence of 0.61% in a multicenter experience. However, design limitations of this registry precluded any detailed analysis regarding antithrombotic therapy, device type, or comparison with surgical rates. It was also believed that the risk for thrombosis was significantly underestimated in this registry analysis. In another recent report, Makkar et al. (14) analyzed computed tomographic (CT) data from patients in a clinical trial and 2 single-center registries and provided data on possible subclinical thrombosis. More recently, a 10% incidence of subclinical THV thrombosis was identified in patients undergoing implantation of balloon-expandable valves (SAPIEN 3, Edwards Lifesciences, Irvine, California) (15). Both these latter studies focused on subclinical THV thrombosis and relied mainly on CT examination performed relatively early after TAVR, for diagnosing thrombosis. However, risk for THV thrombosis is likely to be a continuous hazard that cannot be wholly assessed with a single CT examination. Well-designed longitudinal evaluation studies are needed to truly identify the magnitude and natural history of this problem.
The Heart Center at Segeberger Kliniken has had a well-established TAVR program since 2007, with a pre-defined clinical and echocardiographic follow-up schedule for patients after TAVR. In a retrospective analysis of prospective data from the center’s database, we aimed primarily to estimate the incidence of clinical or manifest THV thrombosis in a general TAVR population. Secondary objectives of our study were to describe the timing, characteristics, predictors, and treatment outcomes of patients diagnosed with clinical THV thrombosis.
Study design and participants
For this retrospective analysis, we identified a total of 649 consecutive TAVR procedures (n = 642) performed between September 2007 and August 2015 at the Heart Center at Segeberger Kliniken. Although baseline characteristics, procedural data, treatment details, and clinical outcomes, including follow-up data, for all patients were collected prospectively for the institute’s TAVR database, the analysis was retrospective. The institutional database is approved by the local ethics committee, and informed consent was obtained from all patients. The study was conducted in accordance with principles of good clinical practice. All procedures followed were in accordance with ethical standards of the responsible committees on human experimentation (institutional and national) and with the Declaration of Helsinki of 1964, as revised in 2013. Flow of patients in this analysis is shown in Figure 1.
Device and procedure types
TAVR was performed using 1 of the following devices: balloon-expandable (SAPIEN XT and SAPIEN 3, both Edwards Lifesciences), self-expanding (CoreValve and Evolut R, both Medtronic, Minneapolis, Minnesota; Biovalve, Biotronik, Bülach, Switzerland; JenaValve, Jena Valve Technology, Munich, Germany; and Symetis, Symetis, Geneva, Switzerland), or mechanically expanding (Lotus, Boston Scientific, Marlborough, Massachusetts). TAVR was performed through transfemoral, transsubclavian, transapical, or transaortic routes. In most patients, it was done under conscious sedation without transesophageal echocardiographic guidance. Pre-dilatation was done at the operator’s discretion. Valve-in-valve procedures were not excluded from this analysis.
Post-TAVR antithrombotic treatment regimen
In accordance with current guidelines, TAVR patients receive oral aspirin and clopidogrel for 3 months after the procedure, followed by lifelong aspirin therapy. Patients who are concomitantly treated with percutaneous coronary intervention using drug-eluting stents are prescribed clopidogrel for a longer duration (6 months). Patients with indications for oral anticoagulation receive a combination of phenprocoumon (a warfarin analog) and clopidogrel for 3 months, followed by lifelong phenprocoumon with an international normalized ratio in the therapeutic range. Patients with warfarin intolerance and those who are unable to remain in the therapeutic window of international normalized ratio are treated with novel oral anticoagulant agents.
Transthoracic echocardiographic assessments were done by experienced cardiologists using a Vivid 7 ultrasound machine and an M4S matrix-array sector transducer (GE Medical Systems, Milwaukee, Wisconsin). Assessments were done immediately post-procedure, 24 to 48 h after implantation, and pre-discharge. The mean transaortic pressure gradient was calculated using the Bernoulli formula. The aortic valve area was calculated using the continuity equation. Left ventricular ejection fraction was measured using Simpson’s method. Aortic regurgitation was graded as 0 (none or trace), 1 (mild), 2 (moderate), or 3 (severe). Echocardiographic assessments were carried out in accordance with the Valve Academic Research Consortium recommendations (16). N-terminal pro–brain natriuretic peptide (NT-proBNP) plasma levels were checked for patients who were symptomatic or had elevated gradients during follow-up.
All patients in the registry undergo a pre-defined echocardiographic and clinical follow-up schedule (30 days, 6 months, 1 year, and yearly thereafter).
Transesophageal echocardiographic and CT evaluation
All patients who reported worsening of their symptoms or had elevated transaortic mean gradients underwent transesophageal echocardiographic assessment using a Vivid 7 ultrasound machine and a 6T transesophageal ultrasound transducer probe. Initial cases of suspected valve thrombosis did not undergo CT examination, but this was made routine later as more cases of valve thrombosis were recognized on transesophageal echocardiography.
CT examination was done using a second-generation dual-source CT scanner (Somatom Definition Flash, Siemens Healthcare, Forchheim, Germany). Contrast-enhanced electrocardiographically gated acquisition of the aortic root was performed after injection of iodinated contrast agent, with the region of interest placed in the ascending aorta and axial sections of 0.6 mm. All data were transferred to a dedicated post-processing workstation (Syngo Multimodality Workplace, Siemens Healthcare) for analysis. Images were evaluated for hypoattenuating areas with or without reduced mobility of 1 or more leaflets identifiable in 2 different projections. The location of the affected cusp was defined with regard to the native cusp position, as right, left, or noncoronary. Assessments were also made for stent frame underexpansion and malapposition.
The primary outcome of this study was clinical THV thrombosis, defined using the criteria for thrombosis of Latib et al. (13). Briefly, they are as follows: 1) valve dysfunction (mean transvalvular gradient >20 mm Hg, reduction of aortic valve area to <1.2 cm2, or new-onset more than mild transvalvular regurgitation) secondary to thrombosis diagnosed on the basis of response to anticoagulation therapy or typical findings on imaging modality (echocardiography or computed tomography); or 2) mobile mass suspicious of thrombus detected on the valve, irrespective of dysfunction and in the absence of infection.
THV thrombosis was classified on the basis of the timing of diagnosis after TAVR as acute (up to 10 days), subacute (10 days to 1 month), or late (more than 1 month).
Secondary outcomes were timing, clinical presentation, imaging findings, and treatment outcomes of patients diagnosed with THV thrombosis.
Data are presented as mean ± SD or as median (interquartile range) for normally and non-normally distributed continuous variables, respectively and as count (percentage) for categorical variables. Standardized differences were calculated for both continuous and categorical baseline variables. Comparisons between categorical variables were made using the chi-square or Fisher exact test as appropriate. Continuous variables with nonparametric distributions were compared using the Mann-Whitney U test. All comparative analyses were patient based. Changes in pressure gradients and valve area were calculated and analyzed using repeated-measures analysis of variance. Post hoc comparisons were performed using the Bonferroni test. Because NT-proBNP values were not normally distributed, variations in levels before and after treatment of valve thrombosis were compared using Wilcoxon signed rank tests. For more than 2 repeated measurements of NT-proBNP levels, comparison was done using the Friedman test. Post hoc comparisons were performed using Wilcoxon signed rank tests with Bonferroni adjustment. Box-and-whisker plots were also generated to evaluate temporal variation of gradients, valve area, and NT-proBNP levels. Penalized logistic regression using the Firth method was used to identify independent predictors of THV thrombosis.
All tests were 2-tailed, and a p value <0.05 was considered to indicate statistical significance for all tests. All analyses were performed using Stata version 13.1 (StataCorp, College Station, Texas) and SPSS release 21 (SPSS, Chicago, Illinois).
Of the 649 TAVRs (n = 642) performed between September 2007 and August 2015, 309 (47.6%) were with self-expanding valves (the majority [97.1%] with the CoreValve or Evolut R), 284 (43.8%) with balloon-expandable valves, and 56 (8.6%) with mechanically expanding valves. Six hundred thirty-three TAVRs were performed through the transfemoral route, 11 were transapical, 3 were transsubclavian, and 2 were transaortic. Valve-in-valve procedures, including TAVR in a previous THV, accounted for 50 procedures (n = 43; 7 TAVRs in a previous THV). Total follow-up duration for all TAVR procedures was 1,221 patient-years (average 1.9 years per procedure), with more than 5 years’ follow-up available for 70 patients.
Incidence of valve thrombosis
A total of 18 patients (2.8%) were diagnosed with clinical THV thrombosis, for an event rate of 1.47 per 100 patient-years of follow-up. Thirteen cases (72.2%) involved SAPIEN valves; CoreValve and Lotus valve thrombosis occurred in 3 (16.6%) and 2 (11.1%) patients, respectively (Figure 2). None of the 7 cases of TAVR in a previous THV developed thrombosis. Post-interventional oral anticoagulation was indicated in 261 patients, while 377 patients received antiplatelet therapy for at least 3 months post-procedure. All 18 patients with THV thrombosis (4.8%) belonged to the antiplatelet group (p < 0.001 vs. oral anticoagulation group); 10 patients were on dual-antiplatelet therapy with aspirin and clopidogrel at the time of diagnosis. The incidence of valve thrombosis was significantly higher for balloon-expandable valves compared with other valve types (odds ratio: 3.45; 95% confidence interval: 1.22 to 9.81; p = 0.01) and valve-in-valve procedures compared with native valve procedures (11.6% vs. 2.2%; odds ratio: 5.93; 95% confidence interval: 2.01 to 17.51; p = 0.005) (Figure 2).
Clinical and procedural characteristics
Timing and modes of clinical presentation
Median time to diagnosis of clinical THV thrombosis was 181 days (interquartile range: 25 to 297 days). Thrombosis was acute in 3 (16.7%), subacute in 2 (11.1%), and late in 13 (72.2%) patients (Table 2). Three patients presented with thrombosis more than 1 year after TAVR. Both cases of thrombosis involving Lotus valves were acute, occurring within 1 week after the procedure.
Most patients with THV thrombosis reported no worsening of their symptoms. Seven patients (38.9%) reported worsening dyspnea and functional class. One patient presented with stroke 21 months after TAVR. There were no other cases of peripheral embolism or myocardial infarction. Serum NT-proBNP levels were significantly elevated.
Elevated transvalvular gradients were observed in 16 patients (88.9%) on transthoracic echocardiography. The peak and mean transaortic gradients at the time of diagnosis of THV thrombosis were 53 ± 20 and 34 ± 14 mm Hg, respectively. The mean aortic valve area was 1.0 ± 0.46 cm2, and the mean left ventricular ejection fraction was 56 ± 14%. One patient presented with moderate central aortic regurgitation, and 9 had new-onset mild aortic regurgitation. Findings on transesophageal echocardiography included immobile leaflets or restricted leaflet mobility (n = 7), thrombotic mass (n = 5), and thickened leaflets (n = 4) (Figure 3). Transesophageal echocardiography ruled out left atrial appendage thrombus in all patients.
CT examination was available for 10 patients, 9 of whom had interpretable images showing findings of hypoattenuated lesions that appeared to involve the bases of the leaflets and extend to the center (Figure 3). The right cusp was involved in 7, the noncoronary cusp, in 6 and the left cusp in 6. Three patients had involvement of all 3 cusps. There were no cases of THV malpositioning, malapposition, or underexpansion. Four patients underwent follow-up CT imaging after 1 month of therapy, and hypoattenuated lesions had completely disappeared in 2 of them; there was a significant reduction of the lesions in the other 2 (Figure 3).
Treatment and clinical outcomes
All patients who were diagnosed with THV thrombosis were treated with therapeutic doses of phenprocoumon. With oral anticoagulant therapy, the transvalvular pressure gradient reduced to baseline in all patients except 1, in whom the diagnosis of THV thrombosis was made at 3 years post-TAVR (gradient reduced but not to baseline). Mean aortic valve area also increased in response to treatment (Figure 4). The median time to reduction of transvalvular gradients was 14 days. Serum NT-proBNP levels also reduced significantly with therapy (Figure 4). No recurrence of thrombosis occurred while on oral anticoagulation, while 2 patients had transient elevations of gradients after temporary cessation of anticoagulation. In 4 patients, phenprocoumon was subsequently changed to a novel oral anticoagulant agent with no recurrence of thrombosis. There were no deaths directly related to valve thrombosis.
THV thrombosis after valve-in-valve TAVR
All CoreValve-related thrombosis was in the setting of valve-in-valve procedures. Surgical valves involved in patients with thrombosis after valve-in-valve procedures were either Hancock II (n = 2) or the closely related Mosaic valve (n = 3). Median duration to diagnosis was identical to that after native valve TAVR (i.e., 181 days). Only 1 patient reported worsening of symptoms, but all patients had significantly elevated transvalvular mean gradients (34 ± 3 mm Hg) and reduced aortic valve areas (0.88 ± 0.34 cm2) on transthoracic echocardiography. All patients responded to treatment with oral anticoagulation.
Predictors of THV thrombosis
Penalized logistic regression using the Firth method was used to identify important predictors of clinical THV thrombosis. Firth regression was performed without univariate pre-filtering. Because there was a null count for oral anticoagulation in the thrombotic group (quasi-complete separation), this parameter was constrained to zero and left in the model to allow its contributing to the penalization. A likelihood ratio test performed to compare the log likelihoods of the 2 models was statistically significant (test statistic = 17.38; p < 0.001), suggesting that unconstrained model offered significantly better goodness of fit. Use of a balloon-expandable valve, antiplatelet therapy alone, obesity, and valve-in-valve procedures predicted THV thrombosis (Table 3).
The key findings of this observational retrospective study are as follows: 1) the overall rate of clinical valve thrombosis after TAVR was 2.8%, with incidence as high as 4.8% in patients on antiplatelet drugs, and no patient on oral anticoagulation developed thrombosis; 2) valve thrombosis was more common with balloon-expandable valves and after valve-in-valve procedures; and 3) the median time to presentation was 181 days (∼6 months) (Central Illustration).
Valve thrombosis appears to be a generally underestimated issue of biological valves, including surgical ones. The incidence of clinical THV thrombosis in this study of 2.8% concurs with the rates of thrombosis observed in previous surgical bioprosthetic series (17–19). Interestingly, the incidence of surgical bioprosthetic valve thrombosis reported in the most recent retrospective analysis by Egbe et al. (19) was as high as 11.6%, of which 63% involved the aortic valves. Several factors may be considered responsible for surgical and transcatheter bioprosthetic valve thrombosis: inadequate antithrombotic therapy, prothrombotic states, reduced antiplatelet effect or high platelet reactivity, poor left ventricular systolic function, atrial fibrillation, and so on (20). Additionally, for THVs, localized areas of stagnant flow due to underexpansion, poor endothelialization of the stent frame resulting from malapposition, and native valve leaflet fissuring and endothelial disruption induced by balloon pre-dilatation could create a prothrombotic milieu (13,21). Precise mechanisms underlying increased thrombosis risk observed with balloon-expandable valves in current as well as previous studies are not known yet, but it is likely that complex patient- and device-related factors, including those mentioned here, are involved.
Of interest, all cases of thrombosis after valve-in-valve TAVR involved either Hancock II or the closely related Mosaic valve. These are also the valves at high risk for thrombosis after surgical valve replacement (17). Brown et al. (17) postulated that the “rail” design in the sinus portions of these porcine valves can create pockets of stasis leading to valve thrombosis. It is likely that recesses may form between the TAVR and surgical valve that extend into these rail pockets during a valve-in-valve procedure.
Thrombosis following surgical valve replacement has been shown to be associated with significant morbidity and mortality (22–24), while THV thrombosis is not considered a medical emergency and rarely requires thrombolysis. Reports of THV thrombosis prior to this study have been largely asymptomatic or subclinical cases (14,15). Our study focused on manifest THV thrombosis. The difference in thrombosis rates among the studies could be explained by the variations in methodology, diagnostic methods, and definitions of THV thrombosis adopted by the earlier studies. Three recent studies based their diagnoses on CT morphological findings, all of which relied on CT examination at a relatively early post-TAVR phase (Pache et al. , median of 5 days; Leetma et al. , 91 days; Makkar et al. , Portico IDE cohort, median of 32 days, pooled RESOLVE and SAVORY cohorts, mean of 30 ± 10 days after the procedure in 42 patients [32%] and within 3 months in 73 patients [55%]). More recently, Hansson et al. (26) published their experience of THV thrombosis diagnosed by multidetector computed tomography 1 to 3 months post-TAVR with Edwards SAPIEN XT and SAPIEN 3 valves. The incidence of multidetector computed tomography–verified thrombosis was 7%, whereas 5 of 405 patients (1.2%) had evidence of clinically overt obstructive thrombus. It must be emphasized here again that the risk for valve thrombosis is an ongoing hazard, and serial follow-up is ideally needed to precisely characterize varying morphological manifestations and natural history. Varied manifestations of THV thrombosis may be placed into 2 categories: obstructive/clinical or nonobstructive/subclinical. The development of obstructive THV thrombosis is more than likely dependent on the amount of thrombus, the number of leaflets involved, and time (14). From the findings of the present study, it may be speculated that there is an early phase of THV thrombosis when the predominant finding is an imaging abnormality with or without elevated gradients. As time progresses or with extensive thrombus, more leaflets may become involved, with resultant elevation of gradients and manifestation of symptoms (Figure 5). In our study, median time to diagnosis and/or symptoms were 6 months. A little over one-half of our patients did not report any worsening of their symptoms despite having elevated transvalvular gradients. However, it is well known that older patients often have restricted life-styles or even subconsciously limit their activities to avoid discomfort. Only 1 patient in this study had an embolic manifestation, an embolic stroke at 2 years. Another patient presented at 3 years with symptoms and elevated gradients with typical morphological imaging findings. The latter patient did not completely respond to oral anticoagulant therapy, possibly because some thrombus had become organized and adherent with time, and ultimately needed a repeat TAVR procedure. Considering the existence of an early subclinical phase of THV thrombosis, the following questions arise: 1) should all TAVR patients undergo routine CT imaging at specified intervals starting early after TAVR, and should those determined to have subclinical thrombosis be treated with oral anticoagulant agents? 2) For those who are negative for thrombosis on early scans, at what time points and until what time frame should the scans be performed? With the current available knowledge, CT examinations and transesophageal echocardiography are warranted if the patient reports worsening of symptoms or if increasing gradients, reduction of valve area, or new valvular regurgitation is observed on transthoracic echocardiography (Central Illustration). Transthoracic echocardiography alone may not be sensitive enough to pick up imaging abnormalities diagnostic of thrombus.
Lifelong aspirin and 3 to 6 months of concomitant clopidogrel or vitamin K antagonists alone, if indicated, are the recommended antithrombotic regimen after TAVR according to current multisociety guidelines (27). Patients with atrial fibrillation with high stroke risk, those with previous thromboembolic events, and those with mechanical valves require oral anticoagulation. Atrial fibrillation and subtherapeutic anticoagulation were also identified as clinical predictors of surgical bioprosthetic valve thrombosis (19). Whether the THV thrombosis rate of 4.8% in patients not on anticoagulant agents in our study justifies their routine use after TAVR is difficult to decide, as benefits can be offset by anticoagulation-related bleeding, especially in high-risk older subjects. Randomized controlled trials and multicenter prospective longitudinal follow-up studies are needed to identify potential predictors and high-risk groups who may benefit from oral anticoagulant therapy. Of interest, short-term oral anticoagulant therapy after surgical bioprosthetic valve replacement has been shown to reduce risk for thromboembolic events (28,29).
The optimal duration of anticoagulant therapy for THV thrombosis is unknown. In our experience, 2 patients had transient elevation of gradients after temporary cessation of anticoagulant agents. Currently, patients with THV thrombosis in our practice are treated with long-term anticoagulation.
In this study, transesophageal echocardiographic and CT examinations were performed only in patients with elevated gradients on transthoracic echocardiography or new or worsening symptoms. It is possible that routine transesophageal echocardiographic and CT examinations might identify additional subclinical cases of valve thrombosis. It is also possible that some of the patients who died and did not undergo formal diagnostic evaluation could have had valve thrombosis. The incidence rate of THV thrombosis reported in this study is hence likely to be an underestimated value when considering the entire spectrum of subclinical and clinical THV thrombosis. Nevertheless, the study design and methodology allow a “near accurate” estimation of the magnitude of clinical THV thrombosis.
Our study suggests that clinically relevant THV thrombosis is more common than anticipated. As the volume of TAVRs performed increases, this entity is likely to be encountered more frequently. Cardiologists and physicians involved in management of post-TAVR patients should have a heightened awareness of this possible complication. Patients with progressive dyspnea and rising gradients should be referred early to experienced centers with CT facility to rule out valve thrombosis. A consensus criteria for identifying subclinical and clinical THV thrombosis needs to be established. Currently there exists uncertainty regarding ideal antithrombotic regimen and routine CT evaluation post-TAVR. The results of this study provide some valuable insights on follow-up strategies after TAVR and invite further debate on the role of oral anticoagulant therapy after TAVR. Longitudinal, prospective observational studies and randomized controlled trials are needed to verify the results of our study and identify groups that may benefit from prophylactic oral anticoagulant therapy.
WHAT IS KNOWN? Recent observational studies have reported on subclinical THV thrombosis, diagnosed by the presence of impaired leaflet motion on computed tomography. However, data on clinical THV thrombosis are still limited.
WHAT IS NEW? In this retrospective analysis, clinical THV thrombosis was identified in 2.8% of TAVR patients and was characterized by imaging abnormalities, increased transvalvular gradients, and elevated NT-proBNP levels. Post-interventional anticoagulation appears to be effective in the prevention of valve thrombosis. Normal valve function was restored in patients with THV thrombosis following treatment with anticoagulation. Thrombosis occurred significantly more often with balloon-expandable valves and valve-in-valve procedures.
WHAT IS NEXT? The findings of this study call for further discussion of the optimal antithrombotic regimen and the role of routine CT evaluation after TAVR. Careful follow-up is especially needed after valve-in-valve procedures.
Drs. Abdel-Wahab and Richardt have received institutional research grants from St. Jude Medical and Biotronik. Dr. Abdel-Wahab is a proctor for Boston Scientific. Dr. Richardt has received lecture fees from Edwards Lifesciences and Boston Scientific. Drs. Jose and Holy were supported by European Association of Percutaneous Cardiovascular Interventions grants in interventional cardiology, which were partially sponsored by Medtronic and Edwards Lifesciences, respectively. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- computed tomographic
- N-terminal pro–brain natriuretic peptide
- transcatheter aortic valve replacement
- transcatheter heart valve
- Received September 21, 2016.
- Revision received January 23, 2017.
- Accepted January 27, 2017.
- 2017 American College of Cardiology Foundation
- Popma J.J.,
- Adams D.H.,
- Reardon M.J.,
- et al.,
- for the CoreValve United States Clinical Investigators
- Pingpoh C.,
- Pache G.,
- Nawras D.,
- et al.
- Pache G.,
- Blanke P.,
- Zeh W.,
- et al.
- Lancellotti P.,
- Radermecker M.A.,
- Weisz S.H.,
- et al.
- Cota L.,
- Stabile E.,
- Agrusta M.,
- et al.
- Latib A.,
- Messika-Zeitoun D.,
- Maisano F.,
- et al.
- Latib A.,
- Naganuma T.,
- Abdel-Wahab M.,
- et al.
- Pache G.,
- Schoechlin S.,
- Blanke P.,
- et al.
- Kappetein A.P.,
- Head S.J.,
- Généreux P.,
- et al.
- Egbe A.C.,
- Pislaru S.V.,
- Pellikka P.A.,
- et al.
- Pibarot P.,
- Dumesnil J.G.
- Mylotte D.,
- Piazza N.
- Leetmaa T.,
- Hansson N.C.,
- Leipsic J.,
- et al.
- Hansson N.C.,
- Grove E.L.,
- Andersen H.R.,
- et al.
- Holmes D.R. Jr..,
- Mack M.J.,
- Kaul S.,
- et al.
- Brennan J.M.,
- Edwards F.H.,
- Zhao Y.,
- et al.,
- for the DEcIDE AVR Research Team
- Nishimura R.A.,
- Otto C.M.,
- Bonow R.O.,
- et al.,
- for the American College of Cardiology/American Heart Association Task Force on Practice Guidelines