Valve-sparing aortic root replacement for valves with paracommissural fenestrations: which valve will fail?

Abstract

OBJECTIVES

To define morphologic risk constellations during valve-sparing aortic root replacement (VSARR) for aortic valves with paracommissural fenestrations.

METHODS

Patients from the multicentre prospective intention-to-treat VSARR-registry German Aortic Root Repair Registry (GEARR) were screened for paracommissural cusp fenestrations. We studied a combined end-point of residual aortic regurgitation (rAR) on post-cardiopulmonary bypass (CPB) transesophageal echocardiography (TEE), mid-term progress of rAR (transthoracic echocardiography) and aortic valve replacement for AR.

RESULTS

Of a total of 762 registry patients (operated 2016–2024), 145 were identified with ≥1 paracommissural cusp fenestration. Eighteen patients (12%) were not treated as planned but underwent composite valved graft (CVG) implantation. Mean follow-up time was 3 years. Upon post-CPB TEE, rAR grade 1 or 2 were present in 44 (33%) and 3 (2%) patients. 50% of patients with fenestrations in more than one cusp showed early progression of rAR. At 3 years, freedom from the combined end-point was 78% (99% CI 74–79%) for the complete study cohort. Patients with a maximum free margin length difference of <5 mm, reflecting balanced root anatomy with respect to sinus and cusp sizes, had a significantly better outcome than those with ≥5 mm: at 3 years, freedom from the combined end-point was 86% (99% CI 80–91%) vs 41% (99% CI 38–46%, P = 0.011). Outcome was worst for patients with cusp prolapse and a free margin length difference of >5 mm (30% vs 70%, P = 0.018).

CONCLUSIONS

Fenestrations in more than one cusp, inhomogeneities of cusp-free margin lengths, and additional prolapse are associated with inferior outcome after VSARR for valves with paracommissural fenestrations.

Clinical trial registration number

DRKS00007872

Open in new tabDownload slide

INTRODUCTION

Valve-sparing aortic root replacement (VSARR) is a durable alternative to prosthetic aortic valve and root replacement for patients with aortic root aneurysms [1]. Residual aortic regurgitation (rAR) has been shown to have the potential to progress in the first years after VSARR and is linked to reoperations [2, 3]. The impact of the presence of cusp fenestrations on outcome after VSARR is not fully understood. Our objective was to define morphologic risk constellations during VSARR for aortic valves with paracommissural fenestrations.

PATIENTS AND METHODS

Patients from a multicentre prospective intention-to-treat VSARR registry (German Aortic Root Repair Registry [GEARR]) were screened for paracommissural cusp fenestrations. This intention-to-treat design registry focuses on full (meaning all three sinuses replaced) VSARR procedures with the main inclusion criterion being that, after evaluation of preoperative imaging, the patient had been scheduled for a VSARR as surgical ‘plan A’ in the respective centre. Root ‘Remodeling’ operations are not included in the registry. The VSARR types included are full reimplantation operations with a straight, two different sized grafts or Valsalva graft, according to the respective surgeon’s usual practice. Structured intraoperative analysis and quantification of valve and root parameters, including the number and distribution of cusp fenestrations, are routinely performed [4]. Study end-points included the presence of rAR on post-cardiopulmonary bypass (CPB) transesophageal echocardiography (TEE), mid-term progress of rAR (transthoracic echocardiography [TTE]), newly detected aortic valve regurgitation during follow-up, and aortic valve replacement (AVR) for AR (combined end-point).

Statistical analysis

All statistical analyses were performed using SPSS Statistics 26 and SigmaPlot 12.0 software packages. Kaplan–Meier and LogRank analyses were calculated for time-dependent end-points. Kaplan–Meier graphs are given with 95% confidence bands. Logistic multiple regression analysis was performed to identify factors associated with adverse outcome. Data are provided as mean ± SD if normally distributed and as median (range and interquartile range) if not normally distributed. There was no prespecified plan to adjust for multiple comparisons.

Intraoperative and TTE measurements and definitions

Under direct vision, surgeons performed measurements of geometric cusp heights, which is the distance from a cusp-free margin to its insertion line at the lowest point (‘nadir’) in mm. The cusp-free margin lengths were measured using a ruler after sinus tissue resection and before reimplantation of the valve into the graft. On TEE after CPB, cusp coaptation height and effective cusp height [5] were measured. Details of the number and localization of cusp fenestrations were documented.

Ethical statement and IRB approval

The study was approved by the Institutional Review Board (IRB) and Ethics Committees (EC) of all participating institutions where the work was carried out (Primary investigator institution FR 546/14). Written informed consent was obtained by all patients included in this report.

RESULTS

Between October 2016 and February 2024, a total of 762 adult patients were recruited into the GEARR (DRKS-ID DRKS00007872) in six different major cardiac and aortic surgical centres (Munich, Leipzig, Freiburg-Bad Krozingen, Hamburg, Lübeck and Vienna). N = 145 total patients were identified in the registry with at least one cusp fenestration documented (for examples from this cohort, see Fig. 1). This cohort was predominantly male (89%), with a mean weight of 90 ± 18 kg, and a mean body length of 181 ± 9 cm (mean age 50 ± 6 years). For further details of basic clinical characteristics and surgical details, see Table 1.

Number and distribution patterns of fenestrations

Fenestrations of all sizes were documented, and a total of 241 fenestrations were identified in 145 patients. In 70 patients (48%), fenestrations were not limited to one cusp only, but at least two or more cusps were affected. More than one fenestration per cusp (at least one cusp with multiple fenestrations) was found in 14 patients (10%). All of the fenestrations were described as having a paracommissural location, and none of them were closed surgically. See Fig. 2 for details of localization and number of identified fenestrations. N = 23 (16%) of the patients were categorized as having a clearly identified cusp prolapse. None of the patients with bicuspid aortic valve were categorized as having a clearly identified cusp prolapse.

Conversion rates

Of the 145 patients identified with at least one cusp fenestration, in n = 12, the surgeon decided not to perform a VSARR after inspection of the valve, and a conversion to Bentall procedure was performed (conversion rate 9%). A study group of n = 133 patients remained with at least one cusp fenestration and received David procedure according to the plan. The rate of immediate failure of David VSARR leading to conversion to Bentall or prosthetic AVR during the same operation using a second CPB run was 5% (6 of 133), all of these cases because of moderate or more aortic valve insufficiency (AI) with excentric jets. There were no re-repairs in the second pump run.

Valve measurements

The heterogeneity of free margin lengths and geometric heights was studied as parameters indicating unbalanced root anatomy. The median of within-patient difference of geometrical heights between 2 cusps was 3 mm, and the mean value was 2.8 ± 2.1 mm. The median within-patient difference of any two free margin lengths was 3 mm, the mean 4 ± 4 mm. See Fig. 3 for details of commissure heights, free margin lengths and geometric heights, and respective histograms.

Aortic valve function

The preoperative annulus size was 26 ± 6 mm, aortic sinus diameter 39 ± 10 mm and ascending aortic diameter 37 ± 9 mm, suggesting that patients were treated primarily for AI rather than large aneurysms. 91 patients (62%) had moderate or high-grade AR preoperatively (37 with severe, 53 with moderate-severe AR).

Excluding the n = 6 patients who received a Bentall for rAR during the index procedure, n = 3 patients (2%) left the operating room with rAR grade 2 (moderate), n = 44 (33%) left the operating room with rAR grade 1, and all other patients had grade 0 (trace or no AR) immediately after the index procedure. Mean coaptation heights in patients with rAR was 9 ± 3 (9.4) mm, and in patients without rAR was 9 ± 3 (8.8) mm.

The incidence of dynamic AR and AVR in the first 5 postoperative years was higher than in other reported series. Over the first 5 years of follow-up, n = 16 (12%) patients showed progress of rAR and n = 9 (7%) showed recurrent AR on one of the follow-up echocardiograms. N = 5 patients (4%) received an AVR at 1.5 ± 1.5 years, within a range of 3 months to 4.4 years after the index procedure. At 3 years, the freedom from a combined end-point of progress of rAR, detection of recurrent AR or AVR for AR was 78% (74–79%) for the complete study cohort (Fig. 4). The freedom from this combined end-point was not different when bicuspid (76% at 4 years) was compared to tricuspid valves (80% at 4 years). The presence of a clear cusp prolapse negatively affected the outcome; the freedom from combined end-point was 50% vs 86% at 4 years.

Morphologic risk constellations for rAR, progress of AR, AVR

In the 16 patients with early progress of rAR, inhomogeneities of free margin lengths were found: 39 ± 6 mm (left) vs 45 ± 6 mm (right) vs 45 ± 7 mm (none), with the cusp carrying a fenestration being significantly shorter (P = 0.001). In those patients with stable aortic valve function, free margin lengths were equal: 42 ± 5 mm (left), 44 ± 7 (right), 44 ± 6 (none). Logistic regression analysis revealed that differences in free margin lengths were linked to a higher likelihood of the combined end-point progression of rAR, new AR and/or AVR (P < 0.001). Patients with a maximum free margin length difference of <5 mm, reflecting balanced root anatomy with respect to sinus and cusp sizes, had a significantly better outcome than those with ≥5 mm. At 3 years, freedom from the combined end-point was 86% (99% CI 80–91%) vs 41% (38–46%), P < 0.011. See Fig. 5 for comparative freedom from the combined end-point according to heterogeneity of free margin lengths and presence of cusp prolapse in patients with cusp fenestrations. When tricuspid aortic valve (TAV) patients were studied as an isolated cohort, the effect of having one or more fenestrations together with a free margin cusp length (FML) difference of more than 5 mm was the same: the freedom from combined end-point was 46% vs 83% at 4 years. Patients with a documented cusp prolapse (n = 25) had significantly inferior outcome: freedom from the combined end-point at 3 years was 50% (40–60%) if cusp prolapse was present vs 78% (74–80%) in regular cusps. In addition, 50% of patients with fenestrations in more than one cusp showed early progression of rAR. Mid-term follow-up until 5 years revealed 5 deaths; one died during the first 30 days.

DISCUSSION

VSARR procedures are a well-established treatment for aortic root aneurysms [1, 6–10]. There is a growing body of evidence indicating that VSARR procedures tend to be superior to the Bentall procedure with regards to reoperation and survival rates in different recently published propensity-matched analyses, which will likely further increase the popularity of these procedures in the surgical community [10–13]. Svensson et al. have shown an 8-year freedom from severe AR of 95%, a 10-year freedom from reintervention of 98% after VSARR, and 10-year survival was reported to be 95% [10]. Arabkhani et al. showed in a propensity score-matched analysis from the AVIATOR registry better survival (95% vs 85%) and fewer valve-related events after VSARR [11] compared to prosthetic replacement.

The importance of paracommissural fenestrations for aortic valve function after VSARR, however, has been studied in only a limited number of reports [14]. This is a subject worth studying in further detail, as cusp abnormalities and their impact on residual AR, progress of rAR and reoperations is not fully understood. A report from the Homburg group from 2023 shows that cusp pathologies are a significant surgical challenge, as over 90% of 582 patients with tricuspid aortic valves and aortic root aneurysm in their cohort showed cusp pathologies, and only 9% showed completely normal three cusps [15]. In a subcohort of n = 281 patients who were included in a detailed TEE study by the group, only n = 12 fenestrations (4%) were documented in the Homburg cohort. GEARR data, on the basis of meticulous documentation of all cusp fenestrations, showed fenestrations to be more frequent: our n = 145 patients with definite diagnosis of a fenestration account for 19% of the whole GEARR cohort. These patients seemed to represent a specific subpopulation, being predominantly male and of large body size, with a relatively large preoperative annular diameter (mean of 28 ± 2.5 mm).

The presented data should be interpreted according to three separate sets of end-points: (1) conversion rates (after inspection of the valve and after having the patient scheduled for a VSARR as Plan A) or immediate conversion after failed VSARR attempt; (2) incidence or any grade of residual AR; (3) the dynamism of rAR or new AR, including AVR for AR. There is no official nomenclature for a ‘failed’ VSARR, but it should include immediate conversion after first CPB run, dynamic AR (any type) and AVR. Conversion upon inspection and mild stable rAR should certainly not be included, and on the basis of this rationale, we defined our combined end-point for the reported study.

The presented data show that in the presence of cusp fenestrations, one should be quite selective in deciding to perform a VSARR. An important aspect of the data is the conversion rates upon inspection of the aortic valve or after a failed VSARR attempt necessitating a second CPB run. With 9% and 5%, these rates seem to be higher than in patients with no cusp fenestrations and higher than in the general VSARR experience, but this has to remain speculative as data for direct comparisons are not readily available. This information is rarely reported in publications on other series and registries, and it is likely that these end-points are generally underreported. Recent work indicates that testing the aortic valve after reimplantation into the graft decreases the need for second bypass runs [16, 17].

With respect to the incidence and severity of rAR, with which the patients left the operating room, with more than trace AR seen in n = 44 (33%, Grade 1), this was relatively comparable to other reported series. In the overall GEARR cohort, the rate of rAR is about 23%, so the group with fenestrations did seem to have a slightly higher risk of rAR. But the main and most profound difference in outcome could be found in the third set of adverse events, all aspects of dynamic AR over the first five postoperative years: freedom from the combined end-point of only 78% at 3 years is certainly inferior to the general VSARR results.

As recently reported by T. David in 111 patients with fenestrations and a median follow-up of 8 years, running a GORE-TEX suture along the complete free margin of a cusp with fenestration results in ingrowth into the cusp and cusp stabilization, and this might be a solution in some selected patients [14]. In this work, comparable aortic valve function was described in patients treated this way to patients without any fenestrations. The GEARR data, showing inferior outcomes for those patients with VSARR but no further surgical steps to address the problem of fenestrations, support the hypothesis that this free margin stabilization is an effective method.

Analysing GEARR data so far, there had been no signal indicating inferior outcome for valves with differences in free margin lengths, except for those with fenestrations, where the difference in outcome was relatively profound. If one is not sure to perform a VSARR or not after having identified fenestrations in one or more cusps, the presence of additional prolapse or difference of more than 5 mm of free margin lengths between any of the cusps should lead the surgeon to be hesitant to follow a valve-sparing approach.

Objectives and future goals of GEARR

GEARR is a prospective multicentre registry focusing on root and valve intraoperative quantitative analysis as well as long-term results after full VSARR procedures. The registry, with commencement of patient recruitment in the fall of 2016, was implemented with an intention-to-treat design, which means that the main inclusion criterion is the respective surgeon’s plan A for surgery, with all imaging available, being a full VSARR. GEARR is thus a registry with a highly selected and homogenous patient substrate, operated on in recent years (2016–2024) and within a relative short period of time to avoid era effects.

CONCLUSION

Fenestrations in more than one cusp and inhomogeneities of cusp free margin lengths might be associated with inferior outcome in the setting of aortic valves with paracommissural fenestrations and should have an influence on the individual surgical decision-making.

Limitations

All limitations of a registry study apply. The included patients do not reflect the complete number of patients treated with a VSARR per centre; inclusion of a patient according to predefined inclusion criteria was at the discretion of the respective surgeon. Bicuspid and tricuspid aortic valves were analysed together in this study, and specific morphological details typical for bicuspid aortic valves might interfere with the results.

FUNDING

The German Aortic Root Repair Registry has been continuously funded since 2016 by the German Heart Research Foundation (Deutsche Herzstiftung).

Conflict of interest: Martin Andreas is proctor/consultant/speaker (Edwards, Medtronic, Abbott, Boston, Zoll, AbbVie, Braun) and received institutional research grants (Edwards, Abbott, Medtronic, LSI). The other authors have declared no conflict of interest.

ACKNOWLEDGEMENTS

The authors acknowledge the important support by the Center for Clinical Studies (University of Freiburg) in Freiburg, Germany, for conducting this study.

DATA AVAILABILITY

Data are available on request.

ETHICAL APPROVAL

The study was approved by the Institutional Review Board (IRB) and Ethics Committee (EC) of the institutions (EK-FR 546/14, dated 6 February 2015). Written informed consent was obtained by all included patients.

Author contributions

Fabian A. Kari, MD: Conceptualization; Data curation; Formal analysis; Funding acquisition; Investigation; Methodology; Project administration; Resources; Supervision; Validation; Writing—original draft; Writing—review & editing. Martin Czerny: Conceptualization. Michael Borger: Conceptualization. Martin Misfeld: Conceptualization. Bartosz Rylski: Conceptualization; Data curation. Emmanuel Zimmer: Investigation. Matthias Siepe: Investigation. Christian Hagl: Investigation; Validation. Christian Detter: Investigation. Johannes Petersen: Investigation. Doreen Richardt: Investigation. Stephan Ensminger: Investigation. Paul Werner: Investigation. Martin Andreas: Investigation. Sven Peterss: Investigation; Visualization. Maximilian Pichlmaier: Conceptualization; Investigation. Christoph S. Mueller: Conceptualization; Formal analysis; Investigation

Reviewer information

European Journal of Cardio-Thoracic Surgery thanks Tomas Holubec, George T. Stavridis and the other anonymous reviewers for their contribution to the peer review process of this article.

REFERENCES

ABBREVIATIONS

ABBREVIATIONS
 
• AR

  Aortic regurgitation

 
• AVR

  Aortic valve replacement

 
• CPB

  Cardiopulmonary bypass

 
• CVG

  Composite valved graft (Bentall operation)

 
• GEARR

  German Aortic Root Repair Registry

 
• rAR

  Residual aortic valve regurgitation after valve-sparing surgery

 
• TEE

  Transesophageal echocardiography

 
• TTE

  Transthoracic echocardiography

 
• VSARR

  Valve-sparing aortic root replacement

Author notes