Original Contribution

Percutaneous Paravalvular Leak Closure for Balloon-Expandable Transcatheter Aortic Valve Replacement: A Comparison With Surgical Aortic Valve Replacement Paravalvular Leak Closure

Kazuaki Okuyama, MD;  Hasan Jilaihawi, MD;  Mohammad Kashif, MD;  Vikas Soni, MD;  Takashi Matsumoto, MD; Wen Loong Yeow, MD;  Mamoo Nakamura, MD;  Wen Cheng, MD;  Saibal Kar, MD;  Raj R. Makkar, MD

Kazuaki Okuyama, MD;  Hasan Jilaihawi, MD;  Mohammad Kashif, MD;  Vikas Soni, MD;  Takashi Matsumoto, MD; Wen Loong Yeow, MD;  Mamoo Nakamura, MD;  Wen Cheng, MD;  Saibal Kar, MD;  Raj R. Makkar, MD

Abstract: Aims. Paravalvular leak (PVL) is frequently observed after transcatheter aortic valve replacement (TAVR) and is related to increased mortality. Percutaneous PVL closure, which is a viable option for this complication, has been performed following surgical aortic valve replacement (SAVR); however, the experience in TAVR remains limited. We sought to compare this technique between post-TAVR and post-SAVR cases. Methods and Results. A single-center series of patients consecutively undergoing percutaneous PVL closure was reviewed. Each group had 10 cases and procedural/imaging variables were assessed. Although there was no severe complication during the procedures, procedural success rate was lower in the post-TAVR group (60% vs 100%; P=.04). There was resistance in all 4 unsuccessful cases, and we were unable to advance the delivery sheath over the wire. Computed tomography revealed that unsuccessful cases had higher calcification volume in the corresponding leaflet (351.4 ± 205.1 mm3 vs 121.8 ± 111.7 mm3; P=.049). This finding can explain the mechanism of difficulty; the higher volume of calcification increases the resistance while advancing the delivery sheath. Conclusion. This is the first study revealing the difficulty of percutaneous PVL closure following TAVR compared with SAVR. However, it is a preferred approach in TAVR patients given their high risk for a surgical procedure. Computed tomography assessment of calcification volume provides important information for preprocedural planning.   

J INVASIVE CARDIOL 2015;27(6):284-290

Key words: transcatheter aortic valve replacement, cardiac computed tomography, vascular closure devices


Aortic regurgitation (AR) due to paravalvular leak (PVL) after transcatheter aortic valve replacement (TAVR) is a common and important adverse event that is linked to poor prognosis.1 It has been reported that nearly all patients after TAVR have some degree of PVL. Although most cases are mild, 6.9%-21% may have significant (moderate to severe) PVL.1-3 Significant PVL is usually treated acutely by postdilatation or transcatheter valve in transcatheter valve (TV-in-TV) implantation during the index procedure.4-7 However, progressive worsening of PVL after TAVR can also occur; PVL worsened in 22.4% of cohort A and 16.4% of cohort B in the PARTNER 2-year follow-up trial.1,2 The clinical management of this phenomenon is not yet well established. Percutaneous PVL closure after TAVR could be a reasonable strategy in these high-risk patients, but the experience in performing this intervention is limited.6,8-10 Conversely, this technique for aortic and mitral surgical prosthetic valves is an established approach with a reasonable success rate; technical success rates of 77%-86% and clinical success rates of 67%-77% have been reported.11 Percutaneous PVL closure methodology is similar after either surgical aortic valve replacement (SAVR) or TAVR. In this article, we compare the acute outcomes of percutaneous PVL closure in post-TAVR and post-SAVR groups, and the imaging assessments that were performed.


Patient selection. Percutaneous PVL closures consecutively performed for PVL following SAVR or TAVR at a single center between March 2009 and June 2013 were identified retrospectively. Thereafter, they were divided into two groups of post-TAVR and post-SAVR patients. Index TAVR procedures were performed with a balloon-expandable prosthetic valve (Edwards Sapien or Sapien XT; Edwards Lifesciences) at our institute. All patients had symptomatic PVL (moderate or severe) despite optimal medical therapy, and received treatment by closure at a date after the index procedure. Written informed consent was obtained prior to all procedures. 

Procedure. All PVL closure procedures were approached in a retrograde manner via the femoral artery and performed as previously described.12 The defect site was crossed with an angled or straight hydrophilic guidewire, supported by an appropriate catheter that directed the wire toward the defect. Once the defect was crossed, the support catheter was advanced across the defect and an Amplatz Extra-Stiff wire (Cook Medical, Inc) was advanced into the left ventricle. Once the stiff wire was in place, an appropriately sized delivery sheath was introduced into the left ventricle to deliver the occluder device. The distal end of the device was deployed below the aortic valve. The device was then pulled back into the defect (Figure 1). All occluder devices deployed in this study were Amplatzer Vascular Plug II (AGA Medical), except one Amplatzer PDA and one Amplatzer muscular VSD used in the SAVR group. Single-antiplatelet therapy with aspirin was administrated to all patients after the procedure, and clopidogrel (1 patient) or warfarin (1 patient) was also prescribed based on the comorbidities. 

Echocardiography. TEE was used during PVL closure. Real-time two-dimensional and three-dimensional reconstructions were used to guide each procedure. The location and severity of PVL was assessed using two-dimensional color Doppler echocardiography. The severity was classified as none, mild, moderate, or severe based on the Valve Academic Research Consortium-2 definition.13 The position of the TAVR valve was defined as the depth of the prosthetic valve in relation to the aortic annulus in the left ventricular outflow tract in the longitudinal view. Also, the length and width of the PVL and circumferential PVL extents were assessed in the cross-sectional view of the valve with color Doppler echo in TAVR patients retrospectively to further understand the etiology. 

Computed tomography. Electrocardiography (ECG)-gated multislice computed tomography (CT) was performed prior to TAVR with a Siemens Somatom Cardiac 64 scanner (Siemens Healthcare USA), using a collimation of 0.6 mm for contrast CT and 1.2 mm for non-contrast CT. A dedicated protocol was formulated, with 120 kV and tube current modified according to patient size. A standard convolution kernel of B35f was applied, with a gantry rotation time of 330 ms. The ECG at the time of acquisition was reviewed before reconstruction to select out ectopy. Reconstruction was performed with slice thickness of 1 mm at 0.5 mm intervals for contrast CT and slice thickness of 3 mm at 2 mm intervals for non-contrast CT. 

CT images were analyzed using a dedicated TAVR software package (3mensio Structural Heart; 3mensio Medical Imaging BV). Non-contrast CT was used in order to assess calcium volume, except for 2 patients who had only contrast CT. Calcium volume of the native aortic valve or each leaflet was measured from the level of annulus to the tip of leaflet. A threshold of 450 Hounsfield units was applied both to contrast CT and non-contrast CT to standardize the influence of contrast.14 Aortic root angulation was measured in a coronal view with an angle between the axis of the aortic root and a vertical line.

Procedural success and clinical success. Procedural success was defined as a reduction in the degree of PVL to mild or none following deployment of the occluder device as assessed by TEE.6 Clinical success was defined as improvement in PVL severity as determined by TEE or fluoroscopy with no periprocedural/in-hospital mortality. 

Statistical analysis. Quantitative variables are expressed as mean ± standard deviation for normally distributed variables and median (interquartile range) for variables not distributed normally. Qualitative variables were expressed as numbers with percentages. An independent sample t-test was used for comparison of independent, normally distributed qualitative variables; Mann-Whitney U-test was used for quantitative variables not distributed normally; and Fisher’s exact test was used for qualitative variables. A P-value ≤.05 was considered statistically significant. The data were analyzed with SPSS software (PASW v18, SPSS, Inc) 


From March 2009 to June 2013, a total of 20 consecutive patients underwent percutaneous PVL closure: 10 post-TAVR cases out of the 657 TAVR patients (1.5%) and 10 post-SAVR cases. Age, Society of Thoracic Surgeons (STS) score, and days after the index procedure were similar in post-TAVR and post-SAVR patients. The post-TAVR group had more cerebrovascular disease (CVD; 40% vs 0%; P=.04) and pulmonary disease (40% vs 0%; P=.04). There were no significant differences in the other variables between the two groups (Table 1).

Contrast volume and number of deployed devices during the procedure were significantly greater in the post-SAVR group (30.0 ± 75.0 mL vs 145.0 ± 118.8 mL [P=.01] and 1.0 ± 1.0 vs 1.5 ± 2.0 [P=.01], respectively). This was probably because the PVL closure procedure was more likely to be completed in the SAVR group vs the TAVR group (procedural success rate, 100% vs 60%; P=.04) (Table 2). All successful cases of both groups had immediate PVL improvement during the procedure. On the other hand, all 4 unsuccessful cases in the post-TAVR group had problems with an inability to advance the delivery sheath over the wire and were aborted after multiple attempts (Table 3). Consequently, the post-TAVR group had more PVL ≥moderate after closure (40% vs 0%; P=.04). However, in all 6 procedure-completed cases in this group, postclosure PVL was less than mild and New York Heart Association classes were improved by >1 grade except in 1 patient who expired. 

In 1 post-TAVR case, the patient received a transcatheter valve in a 23 mm Hancock II surgical prosthetic valve (Medtronic). The percutaneous PVL closure for this patient was successfully performed in the space between the surgical valve and TAVR valve (Figure 2). Another patient underwent TV-in-TV at 149 days after the index TAVR procedure because of worsening PVL; however, the PVL remained significant and the patient remained symptomatic. Percutaneous PVL closure was attempted 369 days after the index TAVR, but it was unsuccessful due to difficulty advancing the delivery sheath (Figure 3). 

Complications. In the post-TAVR group, 1 of the successful PVL closure cases was discharged in a stable condition on postprocedure day 4, but the patient unfortunately expired 38 days later due to septic shock secondary to pneumonia/urinary tract infection at another institution. Another patient who had a failed PVL closure attempt underwent SAVR, and he too passed away after 38 days due to poor recovery post surgery (Table 3). 

In the post-SAVR group, 1 patient had congestive heart failure due to PVL with a suspicion of endocarditis. He received empiric antibiotics and did well and his cultures came back negative. Percutaneous PVL closure was performed successfully, but the patient expired due to respiratory insufficiency. The cause of death was formally documented as sepsis secondary to endocarditis. In another patient, the procedure was complicated by a non-flow-limiting femoral artery dissection. In a third patient, a small hematoma developed at the access site. 

Anatomical features of successful and unsuccessful cases (post-TAVR group). Although there were no significant differences in the echocardiographic assessment between successful and unsuccessful closure cases, computed tomography (CT) assessment showed that there was a trend of greater calcium volume in the aortic valve of unsuccessful cases (937.1 ± 775.0 mm3 vs 313.9 ± 303.0 mm3; P=.14). Further evaluation of calcification in each leaflet corresponding to the PVL revealed a significant difference of calcium volume between unsuccessful vs successful cases (351.4 ± 205.1 mm3 vs 121.8 ± 111.7 mm3, respectively; P=.049). Aortic root angles, which can affect the ability to back up an advanced catheter, were almost identical in both groups (46.0 ± 2.9° vs 36.2 ± 12.0°; P=.15) (Table 4).


Although the malignancy of PVL is well recognized, the management of PVL following TAVR still remains controversial. Mild PVL is considered relatively benign, is not progressive in the majority of patients, and is usually only clinically followed. Ussia et al reported that rates of mild and moderate PVL had been reduced at 3 years post TAVR (from 53% to 47% in mild PVR cases and from 15% to 10% in moderate PVR cases).15 On the other hand, Kodali et al reported that PVL at 2 years had remained unchanged in 46.2% and improved by 1 grade in 31.5% of patients, whereas it increased by ≥1 grade in 22.4% of patients.1 There can be progressive PVL in a minority of patients. In our study, all 10 PVL cases began as mild post TAVR, but worsened to moderate or severe over time and an intervention was required. 

The etiology of PVL contributes to the procedure selection. There are several factors that have been considered as the etiology contributing to PVL after TAVR, ie, prosthesis underexpansion, undersizing, impingement of calcium nodules interfering with stent expansion, or incorrect positioning (too high or too low or not coaxial) leading to incomplete apposition of the valve skirt to the aortic annulus.10,16,17 Usually, balloon postdilatation using a slightly oversized balloon is the first option for PVL after TAVR. The use of this technique, however, is limited due to the possibility of damage to the prosthetic valve, prosthesis migration, aortic annulus rupture, or worsening central aortic regurgitation if present. TV-in-TV to treat this complication has also been described,4-6 but this technique may be less effective if the leak is produced by the presence of focal bulky calcium at the annulus, as was seen in 1 case of the present study wherein PVL was caused by calcification at the left coronary cusp; this patient underwent TV-in-TV at a later date post TAVR, but it was an unsuccessful procedure as previously described. The stent frame of the second valve might have been deformed by the focal calcification and contributed to the residual PVL (Figure 4). Percutaneous PVL closure should be considered for this focal PVL caused by calcification. Martinez et al proposed an algorithm for the percutaneous management of PVL in which repeat ballooning was the first option; if PVL persists, then another procedure should be considered based on the PVL shape (either TV-in-TV for circumferential PVL or percutaneous PVL closure with a closure device for focal PVL).6 

Although percutaneous PVL closure definitely has a role in the TAVR era, it seems to be more challenging compared with the post-SAVR patients in our study. During TAVR, the irregular surface of the native calcified aortic valve leaflets are compressed between the wall of the aortic root and the expanded prosthetic valve stent frame. This creates resistance, and the occluder sheath and device are not able to cross an eccentric space between the calcified native leaflet and the stent of the valve. This hypothesis is supported by our observation that we had problems in advancing delivery catheters in all unsuccessful cases, although wires initially did cross. Moreover, the calcium burden of the corresponding leaflets was greater in these cases (351.4 ± 205.1 mm3 in unsuccessful cases vs 121.8 ± 111.7 mm3 in successful cases; P=.049), as demonstrated in Figure 4. This is in contrast to the post-SAVR group, in which all percutaneous PVL closures were performed successfully. The SAVR procedure removes the native leaflet, and the PVL site is therefore a much smoother surface. 

There is a paradox in that bulky calcium is both a major cause of PVL and a significant resistance to catheter advancement, which is a fundamental technique for successful percutaneous closure. Ingenuity during the procedure may provide a possible solution. There is 1 reported case of a percutaneous PVL closure following TAVR in which there was difficulty in advancing the catheter by retrograde approach. However, the leak was successfully crossed and treated by an antegrade transseptal approach.8 The investigators speculated that the antegrade access of a difficult-to-cross PVL can be achieved with less resistance due to the anatomical differences. Indeed, the aortic valve apparatus from the aortic side is crowded compared to the smooth left ventricular outflow tract. So far, all reported percutaneous PVL closures after TAVR have been performed electively.6,8-10 Detailed imaging assessment prior to the procedure provides important information in understanding the etiology of the PVL and the anatomical findings the contribute to the difficulties encountered during closure. Based on this information, modifications in the procedure (including approach) may improve the procedural success rate.

In this study, all devices used in the post-TAVR group were Amplatzer Vascular Plug II occluders. This device is compatible with 4-7 Fr sheaths and 5-9 Fr guide catheters. The sheaths and catheters were chosen depending on the device size. We used either 5 Fr or 6 Fr shuttle sheaths (Cook Medical, Inc) for delivery. Progressive device iterations will allow lower-profile catheters with the Amplatzer Vascular Plug 4, which can be delivered through a 0.038˝ diagnostic catheter. The application of this low-profile occluder device may also reduce unsuccessful cases.

There are possible severe adverse events with this technique, eg, embolization of the occluder device, cerebrovascular disease by device manipulation, entrapment of a wire into the stent struts while crossing the defect, and prosthetic valve dislodgment during device delivery.18 We experienced 1 case in which bioprosthetic valve movement was observed under fluoroscopy during advancement of the catheter after several attempts; given concerns of damaging the valve with more forceful maneuvers, the procedure was aborted. All cases in this study, however, were performed without serious complications caused directly by the procedure, showing it to be safe despite limited efficacy. 

Study limitations. This was a small, single-center study and a type-1 error could have been introduced because of this sample size. The data were collected retrospectively and TAVR was performed only with a balloon-expandable device. Further analysis in larger multicenter series is warranted to definitively assess the value of this approach. For now, a dedicated closure device is not available for this procedure, and the ones used in this study were off-label. 


Our data show that percutaneous PVL closure for post-TAVR patients is a safe procedure, but has limited efficacy compared with post-SAVR cases. However, PVL improvement was feasible, particularly in the absence of high calcium burden in the native leaflet anatomically associated with PVL foci. A carefully considered strategy, including access approach based on the preprocedural imaging, could achieve higher success rates. Additional research with larger series and long-term follow-up is required for definite evaluation. 


  1. Kodali SK, Williams MR, Smith CR, et al. Two-year outcomes after transcatheter or surgical aortic-valve replacement. N Engl J Med. 2012;366:1686-1695.
  2. Makkar RR, Fontana GP, Jilaihawi H, et al. Transcatheter aortic valve replacement for inoperable severe aortic stenosis. N Engl J Med. 2012;366:1696-1704.
  3. Gilard M, Eltchaninoff H, Iung B, et al. Registry of transcatheter aortic-valve implantation in high-risk patients. N Engl J Med. 2012;366:1705-1715.
  4. Makkar RR, Jilaihawi H, Chakravarty T, et al. Determinants and outcomes of acute transcatheter valve-in-valve therapy or embolization: A study of multiple valve implants in the US PARTNER trial (Placement of AoRTic TraNscathetER Valve Trial Edwards SAPIEN Transcatheter Heart Valve). J Am Coll Cardiol. 2013;62:418-430.
  5. Toggweiler S, Wood DA, Rodes-Cabau J, et al. Transcatheter valve-in-valve implantation for failed balloon-expandable transcatheter aortic valves. JACC Cardiovasc Interv. 2012;5:571-577.
  6. Martinez CA, Singh V, O’Neill BP, et al. Management of paravalvular regurgitation after Edwards Sapien transcatheter aortic valve replacement: management of paravalvular regurgitation after TAVR. Catheter Cardiovasc Interv. 2013;82:300-311.
  7. Nombela-Franco L, Rodes-Cabau J, DeLarochelliere R, et al. Predictive factors, efficacy, and safety of balloon postdilation after transcatheter aortic valve implantation with a balloon-expandable valve. JACC Cardiovasc Interv. 2012;5:499-512.
  8. Martinez CA, O’Neill B, Singh V, O’Neill WW. Transseptal approach for the management of paravalvular regurgitation after transcatheter aortic valve replacement. Catheter Cardiovasc Interv. 2013;82:E587-E591. Epub 2013 Mar 25.
  9. Whisenant B, Jones K, Horton KD, Horton S. Device closure of paravalvular defects following transcatheter aortic valve replacement with the Edwards Sapien valve. Catheter Cardiovasc Interv. 2013;81:901-905.
  10. Luu J, Ali O, Feldman TE, Price MJ. Percutaneous closure of paravalvular leak after transcatheter aortic valve replacement. JACC Cardiovasc Interv. 2013;6:e6-e8.
  11. Kliger C, Eiros R, Isasti G, et al. Review of surgical prosthetic paravalvular leaks: diagnosis and catheter-based closure. Eur Heart J. 2013;34:638-649.
  12. Ruiz CE, Cohen H, Del Valle-Fernandez R, Perk VJG, Kronzon I. Closure of prosthetic paravalvular leaks: a long way to go. Eur Heart J 2010;12(Suppl E):E52-E62.
  13. Kappetein AP, Head SJ, Généreux P, et al. Updated standardized endpoint definitions for transcatheter aortic valve implantation: the Valve Academic Research Consortium-2 consensus document. J Am Coll Cardiol. 2012;60:1438-1454.
  14. Schultz C, Rossi A, van Mieghem N, et al. Aortic annulus dimensions and leaflet calcification from contrast MSCT predict the need for balloon postdilatation after TAVI with the Medtronic CoreValve prosthesis. EuroIntervention. 2011;7:564-572.
  15. Ussia GP, Barbanti M, Petronio AS, et al. Transcatheter aortic valve implantation: 3-year outcomes of self-expanding CoreValve prosthesis. Eur Heart J. 2012;33:969-976.
  16. Detaint D, Lepage L, Himbert D, et al. Determinants of significant paravalvular regurgitation after transcatheter aortic valve: implantation impact of device and annulus discongruence. JACC Cardiovasc Interv. 2009;2:821-827.
  17. Koos R, Mahnken AH, Dohmen G, et al. Association of aortic valve calcification severity with the degree of aortic regurgitation after transcatheter aortic valve implantation. Int J Cardiol. 2011;150:142-145. Epub 2010 Mar 28.
  18. Estevez-Loureiro R, Salgado-Fernandez J, Vazquez-Gonzalez N. Percutaneous closure of paravalvular leaks after transcatheter aortic valve implantation with Edwards Sapien prosthesis: a report of two cases. J Invasive Cardiol. 2013;25:92-95.


From the Heart Institute, Cedars-Sinai Medical Center, Los Angeles, California.

Disclosure: The authors have completed and returned the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Makkar reports grants from St. Jude Medical and Edwards Lifesciences; consulting and lecture fees from Medtronic. Dr Jilaihawi reports personal fees from Edwards Lifesciences and Venus Medtech; non-financial support from St. Jude Medical. The remaining authors report no conflicts of interest regarding the content herein.

Manuscript submitted May 29, 2014, provisional acceptance given August 18, 2014, final version accepted October 1, 2014.

Address for correspondence: Raj R. Makkar, MD, 8631 W. Third Street, Suite 410 E, Los Angeles, CA 90048. Email: Raj.Makkar@cshs.org