Subvalvular Pannus Overgrowth after Mosaic Bioprosthesis Implantation in the Aortic Position

Abstract

Purpose: Although pannus overgrowth by itself was not the pathology of structural valve deterioration (SVD), it might be related to reoperation for SVD of the bioprostheses.

Methods: We retrospectively reviewed patients undergoing reoperation for SVD after implantation of the third-generation Mosaic aortic bioprosthesis and macroscopic appearance of the explanted valves was examined to detect the presence of pannus.

Results: There were 10 patients and the age for the initial aortic valve replacement was 72 ± 10 years old. The duration of durability was 9.9 ± 2.0 years. Deteriorated valve presented stenosis (valvular area of 0.96 ± 0.20 cm²; pressure gradient of 60 ± 23 mmHg). Coexisting regurgitant flow was detected in two cases. Macroscopically, subvalvular pannus overgrowth was detected in 8 cases (80%). The proportion of overgrowth from the annulus was almost even and pannus overgrowth created subvalvular membrane, which restricted the area especially for each commissure. In contrast, opening and mobility of each leaflet was not severely limited and pannus overgrowth would restrict the area, especially for each commissure. In other two cases with regurgitation, tear of the leaflet on the stent strut was detected and mild calcification of each leaflet restricted opening.

Conclusion: In patients with the Mosaic aortic bioprosthesis, pannus overgrowth was the major cause for reoperation.

Introduction

Currently, the bioprosthetic aortic valve consisting of bovine pericardium or porcine aortic valve is one of standard and reliable substitutes for excellent hemodynamic performance and durability.^1,2) Satisfactory clinical outcomes were also achieved by the bioprostheses in patients younger than 70 years of age, although reoperations were required due to various causes such as structural valve deterioration (SVD).^3,4) Possible pathology of SVD included restriction of leaflet mobility by calcium deposition, and tear or perforation of the leaflet by mechanical stress over 10 years. In addition to these pathologies, pannus overgrowth was one of the coexisting alterations to affect valvular function, which was observed in 64% of explanted Hancock II porcine valve.⁵⁾ Although pannus overgrowth by itself was not the pathology of SVD, it might be related to reoperation for SVD despite improvement of anti-calcification treatment for leaflets.

We retrospectively reviewed patients undergoing reoperation for SVD after implantation of the third-generation Mosaic aortic bioprosthesis (Medtronic Inc, Minneapolis, Minnesota, USA). The macroscopic appearance of explanted valves was examined to detect the presence of pannus formation.

Material and Methods

Patients

Between May 2000 and April 2014, 359 redo operations (8.0%) were performed in a total of 4485 cases. Among them, redo aortic valve replacement (AVR) was performed in 43 patients who had been treated with AVR. In a total of 43 patients, bioprosthetic valves were implanted in 27 patients, i.e., the Mosaic porcine valve (n = 11), the Carpentier-Edwards PERIMOUNT valve (Edwards Lifesciences, Irvine CA, USA, n = 10), the Medtronic Freestyle aortic root bioprosthesis (Medtronic Inc., Minneapolis, Minnesota, USA, n = 4), and the Edwards Prima Plus stentless porcine bioprosthesis (n = 2). In patients with the Mosaic valve, surgical indication was SVD in 10 cases and active prosthetic valve endocarditis in one case.

In 10 patients with SVD, there were seven males and three females. The age at redo AVR was 72 ± 10 years old (62–84 y.o.). The preoperative New York Heart Association (NYHA) functional class was class III with exertional dyspnea in all patients. There was no patient presenting acute heart failure and pulmonary edema potentially caused by acute fracture of the bioprosthesis. The age for the initial AVR was 62 ± 10 years old and the reason for redo AVR included aortic stenosis (AS) in 8 and aortic regurgitation (AR) with AS in 2. For the initial operation, concomitant procedures included mitral valve replacement (n = 3), mitral valve plasty (n = 2), tricuspid annuloplasty (n = 1), coronary artery bypass grafting (n = 1), ascending aorta replacement (n = 1) and ventricular septal defect repair (n = 1).

Surgical procedures

In all patients, the chest was re-entered via median sternotomy. Cardiopulmonary bypass was installed via with ascending and single/two caval cannulations except one case with fragile ascending aorta (i.e., right axillar artery). Under tepid blood cardioplegic arrest, new prosthetic valve was seated on the aortic annulus with 2-0 Ti-Cron vertical (non-everted) mattress sutures after explantation of the Mosaic valve, i.e. new bioprosthesis was placed on the supra-annular position. The explanted valves were explored to elucidate pathology of SVD.

Statistical analysis

Results are expressed as means ± standard error of the mean (SEM). An analysis was performed by paired Student’s t test to compare the data. The criterion for statistical significance was set at a value of p <0.05.

Results

The sizes of explanted Mosaic valve were 19 mm (n = 2), 21 mm (n = 1), 23 mm (n = 6) and 25 mm (n = 1) in diameter. The duration of durability was 9.9 ± 2.0 (6.7–12.7) years. Two-dimensional echocardiography demonstrated functional valvular stenosis in all cases (the aortic valve area of 0.96 ± 0.20 cm² and peak pressure gradient of 60 ± 23 mmHg) and coexisting regurgitant flow in two cases.

Redo AVR was performed with bioprosthetic and mechanical valves such as the Carpentier-Edwards PERIMOUNT Magna Ease valve (n = 3), the Carpentier-Edwards PERIMOUNT Magna valve (n = 1), the Carpentier-Edwards PERIMOUNT valve (n = 1), the St Jude Medical Epic stented tissue valve (St Jude Medical, Inc, St Paul, Minnesota, USA) (n = 1), the St Jude Medical Regent valve (n = 2), the Medtronic ATS Open Pivot standard valve (n = 1), and the Medtronic ATS Open Pivot AP valve (n = 1). Concomitant procedures included mitral valve replacement (n = 2), mitral valve plasty (n = 1), tricuspid annuloplasty (n = 2), coronary artery bypass grafting (n = 1), ascending aorta replacement (n = 1), hemi-arch replacement (n = 1) and Morrow’s operation (n = 1). The aortic cross clamp time was 120 ± 54 min and total cardiopulmonary bypass was 162 ± 76 min.

After the operation, the Mosaic valves were macroscopically examined. In a total of 10 cases, severe subvalvular pannus overgrowth was detected in 8 cases (80%) (Fig. 1). The proportion of overgrowth from the annulus was almost even and pannus overgrowth created subvalvular membrane, which restrict the effective orifice area especially for each commissure. In contrast, opening and mobility of each leaflet was not severely limited due to calcification, suggesting that subvalvular pannus overgrowth would restrict the effective orifice area especially for each commissure. In other two cases with AR, tear of the leaflet on the stent strut was detected and mild calcification of each leaflet restricted leaflet opening without pannus formation. Additionally, in 10 cases who required reoperation due to SVD after the Carpentier-Edwards PERIMOUNT valve implantation, the major causes included tear and calcification of the leaflets. Although there was only one case with mild pannus formation, the tear of the leaflets was the definitive reason for reoperation.

Fig. 1.

Explanted Mosaic bioprosthetic aortic valve (23 mm in diameter) after 10-years use. (A) On the left ventricular side, pannus formation created white subvalvular membrane underneath the valve. Actual aortic valve area was reduced to less than 50% of the bioprosthesis by pannus overgrowth. (B) On the aortic side, opening of the porcine valve was not limited by calcification and there was no tear of the leaflet. Although maladaptation of facing leaflets was observed due to deterioration, coexisted regurgitation was not detected by echocardiography.

There was no hospital death in this series. One patient with the St Jude Medical Regent valve required the third AVR operation due to stuck valve caused by hypereosionophilic syndrome.

Discussion

We surgically treated 10 patients requiring redo AVR for SVD after Mosaic bioprosthesis implantation in the aortic position. The most characteristic pathology was subvalvular pannus formation, which was highly detected in our series. The pannus overgrowth created subvalvular membrane, which restricted the effective orifice area especially for each commissure. However, the deterioration of porcine leaflets caused by calcification was mild and frequency of leaflet fracture by mechanical stress was low. Accordingly, pannus overgrowth by itself could be one of the possible causes for the reoperation in patients with the Mosaic aortic bioprosthesis other than SVD.

According to the retrospective studies, durability of the Mosaic valve was excellent and actuarial freedom from reoperation from SVD was higher than 95% at 10 years.^2,4,6) However, the mechanism responsible for valve failure seemed different among the studies. In Canadian multicenter analysis reported by Jamieson et al., leaflet tear was the predominant lesion, but not calcification.²⁾ Interestingly, the most distinguished remark was associated with pannus (1–3 mm of tissue overgrowth) observed in 50% cases of SVD. According to single center study in France, the predominant pathology was associated with leaflet tear and calcification, and there was no comments related to pannus formation.⁴⁾ Similarly, in Italian study, SVD was associated with tear and calcification.⁶⁾ In our Japanese study, leaflet tear and calcification were not the predominant pathology and subvalvular pannus overgrowth was detected in 80% of SVD cases. Other Japanese reports noted that eccentric subvalvular pannus overgrowth was detected in two patients, who required redo AVR within 3 years after the initial operation.^7,8) Both reports speculated that the distortion of the bioprosthetic stent might produce fast turbulence underneath the Mosaic valve, resulting in eccentric pannus formation after the early period from the initial AVR. In contrast, our results speculated that implantation of the Mosaic valve without distortion would result in circumferential and proportional pannus overgrowth during follow-up. Although flow turbulence underneath the Mosaic valve might be related to pannus formation, our results suggested that progression of structural deterioration of leaflets by themselves would be slower than that of subvalvular pannus overgrowth.

Earlier studies demonstrated that possible mechanism for pannus formation included prosthetic valve design, biocompatibility, surgical techniques, smaller annuli, flow turbulence, shear stress and inadequate anti-coagulation.^9–11) Although the mechanism has not been clearly elucidated, the duration of pannus progression until reoperation was 9.9 ± 2.3 years in our series. Other Japanese report demonstrated that reoperation by pannus overgrowth rather than SVD was performed after 6 years from the initial operation.¹²⁾ Accordingly, when pannus formation would have not been progressed, estimated longer durability would be consistent with earlier reports.^2,4,6) However, Lawton and associates, showed four cases with SVD caused by calcified leaflets in the early period after Mosaic valve implantation (3–44 months).¹³⁾ Although pannus was detected in two cases (50%), the pathology of the leaflets was apparently different from our series. Thus, there were various types of progression and pathologies for SVD by reviewing several literatures. However, reoperation caused by pannus overgrowth was safely treated by aggressive resection of pannus and repeated AVR, providing satisfactory clinical outcomes.¹⁴⁾

To obtain better freedom from reoperation after Mosaic valve implantation, prevention of pannus overgrowth would be one of the possible resolutions. An earlier histological and immunological study demonstrated that pannus, consisted of myofibroblasts and an extracellular matrix such as collagen fiber, appeared to originate in the neointima in the periannulus of the left ventricular septum and pannus formation may be associated with a healing process of periannular tissue via the expression of transforming growth factor-beta.¹⁵⁾ Although the eccentricity of subvalvular pannus formation was detected after mechanical valve implantation (St Jude Medical), it might be explained by the difference of prosthetic valve design such as the pivot guard system. However, in our series with bioprosthesis, it would start form the intima of the aortic annulus toward the center, resulting in circumferential and proportional pannus formation during follow-up. Interestingly, vertical mattress (non-everted) sutures were used for the initial AVR in all patients in our series, which might be related to the process of pannus formation. Because the artificial material such as pledgets and spaghettis might create flow turbulence underneath the implanted valves. Similarly, the material and configuration of the sawing cuff also influences pannus formation. However, there was no proof and reasons suggesting that vertical mattress (non-everted) sutures and sawing cuff caused pannus formation. Additionally, prevention of pannus overgrowth would be difficult at this time, because inhibition for healing process was unfeasible during recovery from surgery.

There were several limitations in this study. First, our results were obtained from retrospective analysis with a small number. Further cases would be required to validate the data. Second, the effect of pannus overgrowth on pressure gradient across the aortic valve could not be accurately assessed by preoperative echocardiography because the effect of structure deterioration of the leaflets on pressure gradient should not be ignored. Direct measurement of the orifice area encircled by pannus formation had not been performed, because the design of this study was retrospective. Third, we did not perform histological examination of pannus. Pathological study might provide additional information associated with etiology and resolution.

Conclusion

Pannus overgrowth could be the major cause for reoperation in patients with the Mosaic aortic bioprosthesis in this study. Although inhibition of pannus formation is still unsolved, it might reduce reoperation for SVD within 10 years, contributing to better durability of the Mosaic bioprosthesis.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Disclosure Statement

None declared.