Aortic valve reconstruction with CardioCel: Midterm Results

Abstract

OBJECTIVE

The aim of the study was to evaluate the outcomes of aortic valve repair techniques using cusp patch-plasty with CardioCel.

METHODS

Between September 2014 and June 2021, a total of 167 patients underwent aortic valve reconstruction using cusp repair. In all patients, CardioCel patch was used exclusively. An isolated cusp repair was performed in 117 patients (70%), while 50 patients with concomitant aortopathy needed a combined valve and root repair. Seventy-two patients (43%) presented with tricuspid valve. The mean age of the entire cohort was 54.3 ± 12.3 years, with 143 patients being males.

RESULTS

Early (30-day/in-hospital) mortality was 0.6%. The survival at 2, 4, and 6 years was 98.8, 96.8, and 95.7%, respectively. During the mean follow-up of 4.2 ± 1.7 years (resulting in 697 patient-years), a relevant aortic insufficiency occurred in 10 patients (8 of them presenting with bicuspid valve). All the patients underwent a valve replacement, resulting in a 7.8 ± 2.5% cumulative risk of aortic valve reoperation and/or insufficiency ≥3+ at 6 years. The causes of reoperation were cusp tear at the suture line, progressive valve pathology, endocarditis, and unknown in 4, 4, 1, and 1, respectively. Degeneration and/or calcification of the CardioCel has not been observed.

CONCLUSIONS

The intermediate results of aortic cusp repair using CardioCel are good. Anatomo-pathology of the aortic valve and quality of the cusps seem to be the main reasons for repair failure. Further investigations are needed to assess the long-term durability of CardioCel patch-plasty as an alternative to biological valve replacement in specific aortic valve pathologies.

Open in new tabDownload slide

INTRODUCTION

Aortic insufficiency (AI) is most frequently valve-related, and therefore a valve replacement is the usual form of surgical therapy [1]. The use of artificial valves (mechanical or biological) is still a surgical standard, yet it is associated with several, well-known complications or limitations [2]. On the other hand, valve-sparing root replacement (VSRR) techniques, especially if performed in patients with viable aortic cusps, revealed to offer excellent long-term outcomes with fewer valve-related events and very good quality of life [3–5]. The latter can also regard isolated aortic valve repair (AVR) methods; however, the suitability, and consequently the experience with those surgeries, is limited to small numbers of patients without extensive cusp pathologies [6, 7]. Nevertheless, there is a growing interest in AVRs. The recent report of the German Aortic Valve Registry revealed a 29% rate of repair in pure AI, of which at least 40% was isolated (no root involvement) AVR [8].

The aim of the study was to evaluate the surgical and midterm results after consecutive AVR procedures in which cusp patch-plasty (CP-P) with bovine pericardium (CardioCel, Admedus Regen Pty Ltd, Malaga, WA, Australia) was used exclusively.

PATIENTS AND METHODS

Between September 2014 and June 2021, a total of 167 consecutive patients underwent AVR. The cusp repair with CP-P rather than any narrowing of the root/annulus was used exclusively.

Aortic valve (AV) diagnostics were based on transthoracic and transoesophageal echocardiography, heart catheterization, and computed tomography angiography (angio-CT), if appropriate.

Indication for AVR was considered in patients with AI or mixed defect without extensive calcifications, taking into account the surgeon’s experience and the patient’s individual valve patho-morphology as assessed during surgery.

Surgical technique

The most frequent cusp pathology necessitating surgery was cusp restriction with consecutive pseudo-prolapse resulting from the different height of the free margins of the adjoining cusps (Table 1). Corresponding to the underlying pathology, several CP-Ps were used. The most frequent technique in BAVs was a central patch at the end of the raphe, followed by commissural patch and a combination of both (Figs 1 and 2, Supplementary Material, Fig. S1). A strip along the entire free margin (Fig. 2D) was used occasionally in symmetrical BAVs [9]. In 7 BAV patients, a tricuspidation of the valve was performed by implantation of 2 single neo-cusps in place of the non-separated one.

Implantation of a neo-cusp (Video 1) was the most frequent repair technique in TAV patients (49). In 38 of them, it was the right coronary cusp that was replaced. The suture line between the neo-cusp and annulus was performed with 5-0, while the suture lines between the patch and native cusp tissue in all other techniques mentioned above were performed with 6-0 or 7-0 polypropylene sutures.

In cases with coexisting aortopathy, a VSRR procedure with selective sinus replacement, as described in detail previously [3, 9], was performed in 50 patients, 21 of them having BAV. In 38 BAV patients, an additional, modified remodelling of the aortic root (that is not regarded as a root repair due to aortopathy) was performed for transformation of the asymmetrical commissure orientation to the symmetrical one. This technique was described in detail previously [10].

After repair, the depth of the sinuses and the level of coaptation were assessed visually with reversed forceps, as demonstrated in Video 2.

Data collection and statistical analysis

Informed consent was obtained from all patients, and all perioperative data were collected prospectively; however, data analysis was performed retrospectively. The institutional review board approval has been waived for this study because of the retrospective and completely anonymous character of the analysis.

The patients were followed up in our outpatient clinic or by their cardiologist, from whom written documents and images, if available, were requested and reviewed. Deaths for any reason, relevant aortic defect, and any reintervention on the AV and/or aortic root for any reason were assigned as the primary endpoints.

Early mortality was considered for the postoperative time of 30 days as well as for the in-hospital stay. Categorical variables were expressed in the tables and text as frequency (percentage) and continuous variables as mean (standard deviations) and as median [interquartile range (IQR)]. If appropriate, the variables were compared using the χ² test for categorical and Student’s t-test for continuous variables.

The last follow-up (FU) was completed within 6 months (between March and October 2022) before analysis. Survival estimation and freedom from AV and/or root reintervention for any reason were assessed by the Kaplan–Meier method. Furthermore, a competing risk analysis was used to assess a cumulative risk of the events mentioned above. To compare the curves of freedom from AV and/or root reintervention for any reason in TAV and non-TAV patients, a statistical hypothesis (log-rank) test was used. The statistical analysis was performed with the IBM SPSS software (version 27.0; IBM Corp., Armonk, NY, USA).

RESULTS

Seventy-two patients (43.1%) presented with tricuspid aortic valve (TAV) and 95 (56.9%) with non-TAVs, including bicuspid aortic valve (BAV) in 91 (54.5%), unicuspid in 3 (1.8%), and quadricuspid in 1 (0.6%). One hundred and seventeen patients underwent an isolated AVR, while a concomitant VSRR was performed in 50 patients.

The detailed patient characteristics are presented in Table 1.

Surgical results

The mean size of the aortic annulus was 28 ± 3 mm (range 19–37) and differed significantly between 95 non-TAV and 72 TAV patients (29 ± 2 vs 27 ± 3, P < 0.001). Similarly, in patients with isolated AVR (normal root) and in those with concomitant root repair (root aneurysm), the mean size of the aortic annulus was 28 ± 3 vs 29 ± 2, P = 0.001, respectively.

The transoesophageal echocardiography performed after surgery showed no AI in 133 patients (79.6%) and mild AI (1+) in 34 patients (20.4%). The mean gradient across the repaired AV, true coaptation height, and AV orifice area were 6.3 ± 3.8 mmHg, 6.7 ± 1.8 mm, and 3.9 ± 1.0 cm², respectively. In TAV versus non-TAV patients, the values were 7.1 ± 4.1 vs 4.6 ± 1.9 mmHg, P < 0.001; 6.6 ± 1.8 vs 6.9 ± 1.7 mm, P < 0.001; and 3.8 ± 0.9 vs 3.9 ± 0.9 cm², P = 0.97; and in AVRs versus combined root and valve repairs, 6.7 ± 3.7 vs 5.7 ± 1.9 mmHg, P < 0.001; 6.6 ± 1.8 vs 6.9 ± 1.7 mm, P = 0.173; and 3.8 ± 0.9 cm² versus 4.1 ± 1.0 cm², P = 0.03; respectively.

The detailed surgical data are presented in Table 2.

Early mortality and morbidity

One patient died after discharge but during the 30-day postoperative period due to cardiac tamponade.

The neurological morbidity could be assessed after surgery in all patients. The rate of permanent neurological deficit was 0.6% (one minor stroke). Temporary focal neurological deficits were reported in 1 patient, while other temporary neurologic dysfunctions comprising confusion, delirium, agitation, etc., without evidence in computed tomography or magnetic resonance imaging, were documented in 3 patients (1.8%).

Four patients (2.4%) required a rethoracotomy due to bleeding; however, there were 6 (3.6%) further cases of pericardial effusion treated by a subxiphoidal puncture. Temporary atrial fibrillation was the most frequent postoperative complication that occurred in 54 patients (32.3%), and 2 patients (1.2%) needed a pacemaker implantation during the hospital stay. One patient (0.6) required tracheotomy due to long-term mechanical ventilation. No patient in the cohort required dialysis.

Survival

The mean FU for all patients was 4.17 ± 1.73; range 0.02–7.64 years, resulting in 697 patient-years. The median age at the last FU or at death was 58.9 years (IQR: 50.8–67.1), while the median age at late death was 72.2 years (IQR: 59.9–78.0). The FU was complete for all but one patient, who was lost from the FU after 17 months.

A total of 4 late deaths occurred after surgery. All causes of death are listed in Supplementary Material, Table S1. The estimated survival at 2, 4, and 6 years was 98.8 ± 0.85%, 96.8 ± 1.64%, and 95.7 ± 1.96%, respectively (Supplementary Material, Fig. S2).

Late functional data, reoperations and valve-related morbidity

The mean time of echocardiographic FU was 3.6 ± 1.7; range 0.02–7.4 years (599 patient-years). The mean transvalvular gradient documented at the last echocardiographic FU was 5.5 ± 3.4 mmHg and was slightly lower than that presented before discharge (6.3 ± 3.8 mmHg). Most patients (117) showed no AI (Video 3), while mild AI (1+) or moderate AI was presented in 38 and 2, respectively. A relevant AI (3+) occurred in 10 patients (8 of them presenting with bicuspid valve) (Supplementary Material, Fig. S1). All of them underwent a valve replacement, resulting in a linearized risk (relevant AI/reoperation) rate of 1.43%/year. A competing risk was 7.8 ± 2.5% at 6 years (Fig. 3), while the freedom from AI ≥3 and/or AV replacement (Kaplan–Meier) is demonstrated in Supplementary Material, Fig. S3.

The causes of reoperation were progressive pathology of the valve cusps, cusp tear at the suture line, endocarditis, and unknown in 4, 4, 1, and 1, respectively (Supplementary Material, Fig. S4). Degeneration and/or calcification of the explanted CardioCel has not been observed. The median time between the AVR and reoperation in these 10 patients was 1.73 years (IQR: 0.39–3.15). All patients underwent AV replacement, survived the surgery and are still alive. The median size of the implanted AV prosthesis was 27 mm (IQR: 27–29).

DISCUSSION

There are two surgical ways to improve the cusp coaptation in AI: by narrowing the annulus or by enlarging the cusps [11]. Both methods can be traced until the 1960s and the pioneering works by Ross and Cabrol. While many annuloplasty modifications and devices for narrowing of the annulus have been reported, cusp repair techniques using pericardium did not change considerably over time and range from circumscribed repair of cusp defects to cusp extensions to complete cusp replacement as described originally in the 1990s by Duran et al. [12]. In some reports, the use of non-valvular tissue for cusp repair has been considered as a factor limiting the repair durability [13], yet the question arose if it is really the pericardium that adversely impacts the long-term results or the progressive cusp pathology. Recent reports from the centres of excellence demonstrated that it is the severity of the cusp pathology that mainly determines the mid- and long-term durability [3, 9, 14–17]. It is an unquestionable fact that in VSRR surgeries, a sufficient amount of viable native material is the very precondition for favourable long-term results. Nevertheless, the cusp repair using pericardial patches allows repair of many pathological cusp defects, which cannot be approached with more conservative cusp repairs (e.g. plication) alone.

Admittedly, a sufficient downsizing of the annulus improves the height of cusp coaptation and is a valuable addendum in valve-sparing surgeries with a sufficient amount of cusp tissue [16, 17]. Yet, the narrowed annulus decreases the orifice area and can result in an increased transvalvular gradient and hinder the implantation of a sufficiently large valve prosthesis in the future—an option that the surgeons should always bear in mind after reconstructive valve surgery. Unfortunately, we were not able to find in the literature any data regarding the size of the valvular prostheses that were implanted after AVRs with annulus narrowing. An implantation of a large prosthesis was not problematic in our cohort. The median size of the implanted AV prosthesis in 10 patients with failed AVR was 27 mm (IQR: 27–29), which was not unimportant, considering the mean body surface area of 2.1 ± 0.15 m² (range: 1.87–2.36) in those patients.

The use of autologous pericardium for AVR is certainly a drawback, not only because of application of a non-valvular tissue in cusp position. Harvesting the pericardium and the intraoperative treatment increase the time of surgery, and the subsequent lack of pericardium increases the risk of redo surgery.

Unfortunately, not much is known about the predictability of the long-term performance of autologous pericardium in AV position. The process of degeneration and calcification frequently depends on individual quality of the pericardium. In addition, glutaraldehyde, which is generally used for stabilization, makes the tissue prone to calcification [18]. The search for more predictable material has moved many surgeons towards commercially available xenologous pericardium, especially when larger amount of substitute tissue is necessary [19]. As compared with other bioscaffolds used in cardiac surgery, CardioCel demonstrated superior physical and biological properties in experimental settings [20]. However, the current clinical experience with CardioCel as substitute for AV cusp is very limited. Nordmeyer et al. reported about 40 consecutive patients treated using CP-P with CardioCel. The estimated freedom from AV reoperation in their cohort was only 57 ± 12% at 3 years [21]. In contrast, the midterm functional data in our cohort were excellent with 7.8 ± 2.5% cumulative risk of AV reoperation and/or insufficiency ≥3+ at 6 years. There was, however, a clear tendency towards higher incidence of repair failure in BAVs (Supplementary Material, Fig. S5), even if no significant difference between the two subgroups could be demonstrated when the time of occurrence was considered (log-rank test, P = 0.2). In agreement with other observations, we were able to demonstrate that the reason for reintervention was the progression of the underlying valve disease rather than any degeneration of the patch material [14, 15]. We also did observe a relatively high rate of cusp tears after complete raphe excision (Supplementary Material, Fig. S4), and have abandoned this technique subsequently.

The lack of randomization and patient selection, which may have introduced a selection bias, can be considered a limitation of the study. However, the study cohort contains all patients in whom CP-P was performed at our centre, and CardioCel was the only bioscaffold that was used for repair. The very important limitation, which all reports have in common, is the heterogeneity of the AV pathologies and root involvements. Thus, our cohort contains 30% of patients with concomitant valve and root repair; yet, in contrast to other reports [14, 15], the CP-P was never combined with any annuloplasty in our patients.

Furthermore, as the study was not commercially supported, the FU was not systematically programmed, and the control examinations were performed by the patients’ cardiologists. Thus, as usual in real life, the control examinations in this cohort were not performed regularly once a year, as it is generally recommended after reconstructive valve surgery, and therefore a yearly determination of the AV function was not possible. Yet, for the purpose of the study, all but one patient who could not be contacted underwent echocardiographic examinations.

In conclusion, the CP-P allows repair of many pathological defects, which cannot be approached with narrowing of the aorto-ventricular junction alone. The described techniques of CP-P using CardioCel offer very good cumulative freedom of AV reintervention and/or relevant insufficiency at midterm. The anatomo-pathology of the AV and quality of the cusps seem to be the main reasons for repair failure and, especially, the bicuspid phenotype itself seems to limit the repair durability. The use of CP-P without any annulus narrowing enables implantation of the largest valve possible, should a late replacement of the AV become necessary. Further investigations with longer FUs are needed to assess the long-term durability of CardioCel patch-plasty as an alternative to biological valve replacement in specific AV pathologies.

SUPPLEMENTARY MATERIAL

Supplementary material is available at EJCTS online.

FUNDING

No funding was provided for this work.

CONFLICT OF INTEREST

No potential conflicts of interest exist for any author concerning this work.

ACKNOWLEDGEMENTS

The authors would like to thank Ms Melissa Winkens for her assistance in preparing this article.

DATA AVAILABILITY

The data underlying this article will be shared on reasonable request to the corresponding author.

Author contributions

Vadim P. Irimie: Software; Validation; Visualization; Writing—original draft; Writing—review & editing. Wasim Nasra: Data curation; Investigation. Alaa Atieh: Data curation; Investigation. Akram Ahmidou: Formal analysis; Methodology; Software. Lukas Lehmkuhl: Investigation; Software; Visualization. Paul P. Urbanski: Conceptualization; Project administration; Supervision.

Reviewer information

European Journal of Cardio-Thoracic Surgery thanks Nelson Alphonso, Rune Haaverstad, J. Scott Rankin and the other anonymous reviewers for their contribution to the peer review process of this article.

Presented at the Annual Meeting of the European Association for Cardio-Thoracic Surgery, Lisbon, 2024.

REFERENCES

ABBREVIATIONS

ABBREVIATIONS
 
• AI

  Aortic insufficiency

 
• AV

  Aortic valve

 
• AVR

  Aortic valve repair

 
• BAV

  Bicuspid aortic valve

 
• CP-P

  Cusp patch-plasty

 
• CT

  Computed tomography

 
• FU

  Follow-up

 
• TAV

  Tricuspid aortic valve

 
• VSRR

  Valve-sparing root replacement