Interaction Between Balloon-Expandable Valves and Coronary Ostia: Angiographic Analysis and Impact on Coronary Access

Abstract: Objectives. We sought to evaluate the position of balloon-expandable valves in relation to the coronary ostia using an angiographic- and computed tomography (CT)-based analysis, and to determine the impact of valve position on coronary angiography (CA)/percutaneous coronary intervention (PCI) feasibility and results. Methods. A total of 533 patients who received a Sapien XT or Sapien 3 valve were included in the angiographic analysis. Of these, 49 benefited from an opportunistic electrocardiography-gated CT after transcatheter aortic valve replacement (TAVR) and were included in the CT analysis. Results. Regarding the left coronary artery (LCA) ostium, the top of the transcatheter heart valve (THV) frame was infraostial in 49% of cases, and the valve totally covered the LCA ostium in 27% of patients. The stent frame of the Sapien 3 valve completely covered the LCA ostium more frequently than the Sapien XT valve (43% vs 12%, respectively; P<.001) and the relative implantation depth was significantly less ventricular in the Sapien 3 group than in the Sapien XT group (28.0 ± 12.3% vs 36.8 ± 12.6%, respectively; P<.001). The CT evaluation found similar results to angiographic evaluation. A total of 53 patients (10%) underwent CA (± PCI) following TAVR, and valve position did not influence CA performance/quality and PCI results. Conclusions. The stent frame of balloon-expandable Sapien valves exceeded the coronary ostia in about one-fourth of patients, and this percentage was >40% with the new-generation Sapien 3 valve. However, THV position did not affect the feasibility, quality, and results of CA/PCI post TAVR.

J INVASIVE CARDIOL 2020;32(6):235-242. Epub 2020 May 20. 

Key words: coronary angiogram, percutaneous coronary intervention, transcatheter aortic valve replacement

About half of transcatheter aortic valve replacement (TAVR) candidates have concomitant coronary artery disease,¹ and a recent study reported a coronary event rate of ~10% at 2-year follow-up post TAVR.² In addition, following the positive results of two randomized trials,^3,4 the United States Food and Drug Administration recently expanded the indication of TAVR to patients considered at low surgical risk. Thus, TAVR will likely be implemented in younger patients with a much longer life expectancy compared with their moderate-to-high risk counterparts, which would indeed increase the likelihood of coronary events during their lifetime post TAVR. Therefore, ensuring the possibility of coronary access following TAVR is of high clinical relevance. However, the feasibility of coronary angiography and percutaneous coronary intervention (PCI) in patients previously treated with a transcatheter heart valve (THV) in aortic position remains unclear,¹ and its evaluation needs to integrate the evolution of devices and the optimization of implantation techniques (including implantation depth). Coronary height, valve implantation depth, and bioprosthesis design are all variables that influence the interplay between the THV and coronary ostia. Moreover, the neo-sinus width will also dictate how easily the coronary arteries can be accessed, especially in valve-in-valve cases. All published studies on the subject to date consisted of feasibility studies, mainly including first- and second-generation THVs,¹ with only one study describing the position of older-generation transcatheter bioprostheses in relation to coronary ostia in a small cohort of patients.⁵ The objectives of this study were: (1) to evaluate the position of (older- and newer-generation) balloon-expandable valves in relation to the coronary ostia using an angiographic- and computed tomography (CT)-based analysis; and (2) to determine the impact of valve position on coronary angiography/PCI feasibility and results.  

Methods

A total of 820 patients who had severe symptomatic aortic stenosis and were treated with either the Sapien XT (n = 429) or Sapien 3 THV (n = 391) (Edwards Lifesciences) at our center from October, 2009 to June, 2019 were screened for eligibility in the angiographic and electrocardiography (ECG)-gated CT analysis. The main design characteristics and differences in the Sapien XT and Sapien 3 valves are presented in Table 1 and Figure 1.

A total of 263 patients without final angiographic control or with inadequate angiography quality at the end of the TAVR procedure were excluded. In addition, we excluded 4 patients who had valve embolization or valve misplacement requiring a second THV. The final study group consisted of 553 patients in the angiographic analysis. All angiographies were reviewed by two experienced interventional cardiologists and all measurements were gathered in a dedicated database for analyses. The position of the THV with respect to the left coronary artery (LCA) and right coronary artery (RCA) ostia was assessed and the length of the stent frame above the LCA ostium was measured when appropriate. Valve position in regard to the coronary ostia was classified as either supraostial, at the ostial level, or infraostial, with the upper frame strut as the reference point (Figure 2). Total expanded valve height was also measured and implantation depth and length of the stent frame above the LCA ostium were presented in relative and absolute values (Figure 3). Finally, the orthogonal view of the THV at aortography rarely allowed an acceptable view of the RCA ostium, and RCA opacification was frequently suboptimal to assess the THV position with respect to the RCA ostium, which explained the limited number of measurements available concerning the THV position and RCA ostium (n = 99).     

A total of 49 patients had an opportunistic ECG-gated CT available at a median of 5.8 months (interquartile range [IQR], 1.7-11.8 months) post TAVR. All CTs were reviewed by two experienced radiologists, and all measurements were gathered in a dedicated database for analyses. The native and contrast-enhanced CT datasets were transmitted to a dedicated CT workstation (Aquarius iNtuition, version 4.4.12; TeraRecon). CT measurements included implantation depth, position of the THV with respect to the LCA and RCA ostium, length of the stent frame above the coronary ostium, and the width of the neo-sinuses of Valsalva (Figure 4). 

Patients who underwent coronary angiogram and/or PCI after TAVR were identified. Procedural characteristics, including the type and number of catheters used, injection selectivity, performance indicators, and PCI success rate were recorded and analyzed according to THV position with respect to the LCA ostia (infraostial vs ostial/supraostial).

Statistical analysis. Qualitative variables were expressed as number (percentage), and continuous data as mean (standard deviation) or median (IQR) according to variable distribution. Categorical variables were compared using the Chi-square test or Fisher’s exact test, as appropriate. Numerical variables were compared using the Student’s t-test or Mann-Whitney non-parametric U-test according to their distribution (assessed by the Kolmogorov-Smirnov test). Statistical analyses were performed with SAS, version 9.4 (SAS Institute).

Results

Angiographic analysis. The main patient and procedural characteristics, and angiographic findings are presented in Table 2. The Sapien XT and Sapien 3 valves represented 50.3% and 49.7% of the study population, respectively. Most TAVR procedures (67%) were performed through a transfemoral approach, and a valve-in-valve procedure was performed in 7% of patients. Assessment of the THV position regarding the LCA ostium was available in all patients, and assessments of implantation depth and THV position regarding the RCA ostium were obtained in 65% and 18% of patients, respectively. 

Mean THV implantation depth was 5.9 ± 2.3 mm (32.5 ± 13.1% of the total device length). Regarding the LCA ostium, the top of the THV frame was infraostial in half of cases, while the valve totally covered the LCA ostium in 27% of patients, with a mean length of the valve stent frame above the ostium of 2.0 ± 2.0 mm (10.2 ± 10.1% of the total device length). The THV position regarding the RCA ostium was infraostial, ostial, and supraostial in 49%, 13%, and 38% of patients, respectively.

Comparison between Sapien XT and Sapien 3 valves. The comparison of angiographic findings according to device generation (Sapien XT [n = 278] vs Sapien 3 [n = 275]) is presented in Table 3. Absolute implantation depth did not significantly differ between the two devices (6.0 ± 2.1 mm in the Sapien XT group vs 5.8 ± 2.5 mm in the Sapien 3 group; P=.24), but the relative implantation depth was significantly less ventricular with the Sapien 3 group vs the Sapien XT group (28.0 ± 12.3% vs 36.8 ± 12.6%, respectively; P<.001). The Sapien 3 frame totally covered the LCA ostium more frequently than the Sapien XT frame (43% vs 12%, respectively; P<.001) (Figure 5), but the mean valve stent frame length above the ostium did not significantly differ between the two devices (P=.65 for both). Assessment of the THV position regarding the RCA ostium was feasible in 15 Sapien XT and 84 Sapien 3 patients (5% and 31%, respectively) and did not significantly differ between these two groups (P=.25).

CT analysis. Procedural characteristics and ECG-gated CT findings are presented in Table 4. Sapien XT and Sapien 3 valves represented 72% and 28% of the implanted devices, respectively. Assessment of implantation depth and position of the THV regarding the LCA and the RCA ostia were achieved in 98%, 100%, and 94% of cases, respectively. Left and right neo-sinuses of Valsalva measurements were feasible in 82% and 71% of patients, respectively.

Mean implantation depth was 6.3 ± 2.8 mm (37.1 ± 15.7% of the total device length). LCA and RCA ostia were both totally covered in about 10% of cases, while the THV position regarding the LCA and RCA was infraostial in 47% and 76% of cases, respectively. Mean widths of the left and right neo-sinuses of Valsalva were 4.6 ± 2.0 mm and 4.4 ± 1.6 mm, respectively. 

Coronary angiogram and PCI post TAVR. A total of 53 patients (10%) underwent coronary angiography with or without PCI following TAVR. The THV position was infraostial, ostial, and supraostial with respect to the LCA ostia in 30 patients (57%), 13 patients (24%), and 10 patients (19%), respectively. Procedural characteristics of coronary angiography and PCI post TAVR according to the THV position with respect to coronary ostia are presented in Table 5. There were no significant differences in the total number of catheters used, the need for using guiding catheters to improve imaging during coronary angiography, and the rate of selective (vs non-selective) coronary injection (Figure 6). A PCI of the left main or left anterior descending artery was more frequently performed in the infraostial group (79% vs 22%; P=.02), but no differences in PCI success rate were observed between groups (successful procedures in 88.9% and 92.9% of the ostial/supraostial and infraostial groups, respectively; P>.99).

Discussion

The present study is, to the best of our knowledge, the first to investigate the interplay between balloon-expandable valves and coronary ostia, and its potential impact on coronary access and coronary angiography/PCI results. The main findings can be summarized as follows: (1) balloon-expandable valves totally covered the LCA ostium in about one-fourth of the cases; (2) a supraostial THV position was four times more frequent with the newer-generation Sapien 3 valve; and (3) valve position with respect to the coronary ostia had no significant impact on coronary access, coronary angiography quality, and PCI results. 

The position of the balloon-expandable valve with respect to the LCA ostium was supraostial in 27% of cases and the mean valve stent frame length above the LCA ostium was about 2 mm. These findings are consistent with the Ferreira-Neto et al⁵ study that reported a supraostial THV position in about 20% of patients who received a Sapien or Sapien XT valve and had coronary angiography/PCI post TAVR. Despite the supraostial valve position in a significant percentage of cases, the position of the inner skirt (located in the inferior 50% of the device) makes highly unlikely the possibility of coronary access failure through the struts of this THV system (Figure 1). A frequently non-orthogonal view of the RCA ostium and the insufficient RCA opacification in some cases limited the evaluation of the THV position with respect to the RCA ostia in most cases. However, our findings were comparable with the THV position regarding the LCA ostium, with a supraostial and infraostial position in about 30% and 50% of patients, respectively. Further evaluation of the THV position regarding the RCA ostium is needed in future larger studies with post-TAVR ECG-gated CT.

In those patients who required coronary angiogram and PCI post TAVR, no coronary access issues were observed, along with a lack of significant differences regarding the number and type of catheters used according to the THV position. Likewise, the rate of selective injections did not differ between groups. These findings are in accordance with previous studies^6-8 and provide reassuring information about coronary reaccess through balloon-expandable valves. Future studies are needed for evaluating potential coronary reaccess issues and the interaction between coronary ostia and self-expanding THV systems, particularly those with a taller stent frame/sealing skirt.¹ 

The rate of supraostial THV position was significantly higher with the use of the newer-generation Sapien 3 valve compared with older-generation Sapien XT valves. The combination of a longer stent frame and a higher implantation depth (ie, more aortic and less ventricular) with the Sapien 3 valve would explain such differences, as evidenced by the similar absolute implantation depth (about 6 mm) but the smaller relative implantation depth observed with Sapien 3 valves. Caution would therefore be advisable regarding coronary access with last-generation balloon-expandable THVs. However, no coronary reaccess issues following Sapien 3 valve implantation were observed in the present study, and this was in agreement with the few clinical data available from prior studies about coronary reaccess through a Sapien 3 THV.^5,6 The larger size of stent frame struts of the Sapien 3 system (vs Sapien XT) (Figure 1) may facilitate coronary access and compensate for the increased rate of supraostial valve positioning. 

Conductions disturbances, such as permanent pacemaker implantation and new-onset persistent left bundle-branch block, remain the most frequent complications of TAVR. The current trend is therefore to minimize implantation depth to reduce the burden of conduction disturbances post TAVR,^9-11 but this will inevitably be associated with an increased interaction between the THV system and coronary ostia. TAVR is going to expand toward the treatment of low-risk patients in the near future. In this respect, conduction disturbances and coronary reaccess post TAVR are of high clinical relevance, and the optimal positioning of the THV will have to take into account these two issues.

Post-TAVR ECG-gated CT allows the measurement of neo-sinus of Valsalva widths and an accurate evaluation of THV position regarding the aortic annulus and coronary ostia. Interestingly, CT evaluation found similar results to angiographic evaluation, with a mean implantation depth of 6.3 mm and a mean THV stent frame length above the LCA ostium of 1.5 mm (vs 5.9 mm and 2 mm from the angiographic evaluation, respectively). Conversely, CT-derived analysis reported only 10% of THVs in the supraostial position (vs one-fourth in the angiographic evaluation). This discrepancy could be attributed to the differences between the CT analysis and the angiographic analysis populations, with a larger proportion of Sapien XT valves (76% vs 50%, respectively) and smaller (and shorter) bioprostheses (49% vs 33% of 20/23 mm, respectively) in the CT analysis group. Finally, mean left and right neo-sinus of Valsalva widths were 4.6 and 4.4 mm, respectively. This particular finding provides reassuring information regarding coronary access with 5 and 6 Fr catheters even when coronary ostia are totally covered by the THV stent frame.

Study limitations. The presence of unmeasured confounders cannot be excluded due to the retrospective nature of the study. Although standardized for the purposes of this study, some measurement inaccuracies may have occurred, especially in cases of non-orthogonal projections. However, final angiographies with insufficient contrast opacification or excessive crush views were excluded to ensure an accurate assessment of THV position, and the angiographic findings were consistent with CT findings. Assessment of the interaction between THV tissue and coronary ostium was not feasible. However, the possible interference between valve tissue and coronary ostia was unlikely due to the small THV stent frame length above coronary ostium (2 mm) in supraostial valve position cases. 

Conclusion

The position of balloon-expandable THVs with respect to the coronary ostia was infraostial in most cases, but a much higher rate of supraostial positioning was found with the newer-generation Sapien 3 valve. However, transcatheter valve position did not affect the feasibility, quality, and results of coronary angiography/PCI post TAVR. Future studies are needed to determine the potential interaction between other transcatheter valve types and coronary ostia, and the potential impact on coronary access.

^*Joint first authors.

From the Quebec Heart and Lung Institute, Laval University, Quebec City, Quebec, Canada.

Funding: Dr Rodés-Cabau holds the Research Chair “Fondation Famille Jacques Larivière” for the Development of Structural Heart Disease Interventions. Drs Junquera, del Val, and Muntané-Carol were supported by a grant from the Fundacion Alfonso Martin-Escudero (Madrid, Spain).

Disclosure: The authors have completed and returned the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Faroux reports fellowship support from Institut Servier and the Association Régionale de Cardiologie de Champagne-Ardenne (ARCCA); research grant support from Biotronik, Edwards Lifesciences, and Medtronic. Dr Rodés-Cabau reports institutional research grants from Edwards Lifesciences, Medtronic, and Boston Scientific. The remaining authors report no conflicts of interest regarding the content herein.

The authors report that patient consent was provided for publication of the images used herein.

Manuscript submitted December 10, 2019, final version accepted December 19, 2019.

Address for correspondence: Josep Rodés-Cabau, MD, Quebec Heart & Lung Institute, Laval University, 2725 Chemin Ste-Foy, G1V4G5, Quebec City, Quebec, Canada. Email: josep.rodes@criucpq.ulaval.ca

1. Faroux L, Guimaraes L, Wintzer-Wehekind J, et al. Coronary artery disease and transcatheter aortic valve replacement: JACC state-of-the-art review. J Am Coll Cardiol. 2019;74:362-372.
2. Vilalta V, Asmarats L, Ferreira-Neto NA, et al. Incidence, clinical characteristics, and impact of acute coronary syndrome following transcatheter aortic valve replacement. JACC Cardiovasc Interv. 2018;11:2523-2533.
3. Mack MJ, Leon MB, Thourani VH, et al. Transcatheter aortic-valve replacement with a balloon-expandable valve in low-risk patients. N Engl J Med. 2019;380:1695-1705.
4. Popma JJ, Deeb GM, Yakubov SJ, et al. Transcatheter aortic-valve replacement with a self-expanding valve in low-risk patients. N Engl J Med. 2019;380:1706-1715.
5. Ferreira-Neto AN, Puri R, Asmarats L, et al. Clinical and technical characteristics of coronary angiography and percutaneous coronary interventions performed before and after transcatheter aortic valve replacement with a balloon-expandable valve. J Interv Cardiol. 2019;2019:3579671.
6. Zivelonghi C, Pesarini G, Scarsini R, et al. Coronary catheterization and percutaneous interventions after transcatheter aortic valve implantation. Am J Cardiol. 2017;120:625-631.
7. Blumenstein J, Kim WK, Liebetrau C, et al. Challenges of coronary angiography and intervention in patients previously treated by TAVI. Clin Res Cardiol. 2015;104:632-639.
8. Chakravarty T, Sharma R, Abramowitz Y, et al. Outcomes in patients with transcatheter aortic valve replacement and left main stenting: the TAVR-LM registry. J Am Coll Cardiol. 2016;67:951-960.
9. Mauri V, Reimann A, Stern D, et al. Predictors of permanent pacemaker implantation after transcatheter aortic valve replacement with the Sapien 3. JACC Cardiovasc Interv. 2016;9:2200-2209.
10. Hamdan A, Guetta V, Klempfner R, et al. Inverse relationship between membranous septal length and the risk of atrioventricular block in patients undergoing transcatheter aortic valve implantation. JACC Cardiovasc Interv. 2015;8:1218-1228.
11. Jilaihawi H, Zhao Z, Du R, et al. Minimizing permanent pacemaker following repositionable self-expanding transcatheter aortic valve replacement. JACC Cardiovasc Interv. 2019;12:1796-1807.