Midterm Clinical Outcomes of Reimplantation Versus Remodeling Valve-Sparing Aortic Root Replacement in Patients With Connective Tissue Disorders: A Meta-Analysis

,

Connective tissue disorders (CTDs) are a group of genetic conditions that affect the function of the extracellular matrix, resulting in subsequent pathology of multiple systems, including the cardiovascular system.The most common of which is Marfan syndrome, an autosomal dominant condition associated with the FBN-1 genetic mutation that transcribes the fibrillin-1 extracellular matrix protein.Similar pathophysiologic patterns are observed in Ehlers-Danlos syndrome and Loeys−Dietz syndrome, inheritable conditions caused by collagen and transforming growth factor dysfunction, respectively. 1Patients with CTDs often present with more severe aortopathy-including valvular regurgitation, aneurysmal dilation, and dissection-at a younger age, requiring alternative surgical consideration to patients without CTDs. 2,3A total of 2 previous meta-analyses have compared the clinical outcomes of valve-sparing aortic root replacement (VSARR) versus composite valve grafts in patients with CTDs.In addition to the lack of anticoagulation burden, Flynn et al 4 and Soto et al 5 have demonstrated decreased rates of thromboembolic events and infective endocarditis in patients who underwent a Faculty of Health Sciences, University of Bristol, Bristol, United Kingdom; b Swansea University Medical School, Swansea, United Kingdom; c Faculty of Health, University of Plymouth, Plymouth, United Kingdom; d Wessex Cardiac Centre, Southampton, United Kingdom; and VSARR of this demographic.As a result of earlier diagnoses and improved aortic surveillance, patients with CTDs are more frequently offered prophylactic VSARR surgery, providing improved outcomes compared with nonelective surgery. 6,7There is limited evidence and a lack of randomized data as to whether reimplantation of the native valve 8 or remodeling of the aortic root 9 produces superior clinical outcomes in patients with CTD.Many surgeons advocate for reimplantation techniques because of the perceived increased risk of annular dilation in remodeling VSARR despite the technique producing postoperative structures similar to native anatomy and preserving physiologic movements associated with the aortic root in systole and diastole. 10A 2020 meta-analysis by Zhou et al 11 demonstrated improved rates of reoperation and aortic regurgitation and decreased mortality in reimplantation compared with remodeling VSARR in patients without CTDs. 11However, the etiology of the included study population does not represent the demographic of patients with CTDs nor does it assess the implementation of structural annuloplasty in remodeling surgery.As a result of the recent increase in published data, we aimed to produce the first feasible meta-analysis comparing the clinical outcomes of reimplantation and remodeling VSARR surgery in adult patients with CTDs, with consideration given to developing surgical techniques.

Search strategy and selection criteria
PubMed, Web of Science, and Embase databases were used for preliminary study identification, using the search terms (["connective tissue disorder" OR "Marfan's syndrome"] AND "aortic root replacement") and reviewed in adherence with Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA). 12The PRISMA guidelines and adherence evidence are available to view in the supplementary material.Studies that met the inclusion criteria were selected for review.The inclusion criteria required double-arm studies comparing reimplantation versus remodeling VSARR in patients with CTDs, with a mean age ≥18 years.Abstracts, conference presentations, case reports, editorials, expert opinions, and non-English language studies were excluded from screening.

Study selection and data extraction
Search results were screened by way of assessment of the study title/abstract, and full manuscripts were reviewed by the application of the aforementioned inclusion and exclusion criteria.Uncertainty or variation of study assessment was resolved by discussion.The outcomes were independently assessed for potential patient crossover by the evaluation of study data unit location and study periods.Priority was given to bias-treated data and more recently published in the event that study location and study period presented a high risk of patient crossover.The Newcastle −Ottawa scale was used to assess the quality of all included studies, with scores ≥6 of a potential 9 considered as high quality. 13Funnel plot assessment was used to determine the risk of reporting and publication bias. 14Analysis data were extracted from study text, tables, and figures, with raw figures calculated from percentages where available and necessary.The obtained study data include study period and location, method, surgical techniques, demographic, and patient outcomes.The use of structural annuloplasty in remodeling surgery was extracted from surgical technique data provided in the study methods.All clinical outcomes reported by a significant number of included studies were pooled for analysis.Early and midterm outcomes include in-hospital mortality, stroke, re-exploration for bleeding, long-term mortality, reoperation rates, and postoperative occurrence of aortic regurgitation.Sensitivity analysis was performed, excluding studies and patient subgroups that received suture or ring annuloplasties.

Statistical analysis
Statistical meta-analyses and forest plots were performed on Review Manager 5.3 by way of the Mantel −Haenszel test, producing odds ratios (ORs) for dichotomous data with a 95% confidence interval (CI). 15Fixedeffects models were used where heterogeneity (I 2 ) was calculated to be <50%, with random effects used when I 2 ≥50%.Individual forest plots were calculated where 3 or more included studies reported postoperative clinical outcomes.Propensity score-matched and bias-treated data were prioritized where available.

Search results
Initial study identification using search preliminary search terms (["connective tissue disorder" OR "Marfan's syndrome"] AND "aortic root replacement") produced 1,041 articles from PubMed, Web of Science, and Embase databases, March 2023.After the exclusion of 483 duplicates, 558 abstracts were screened for full manuscript analysis.Of the 35 studies examined for inclusion, 17 were excluded on the grounds of including an inappropriate patient population, either a mean patient age of <18 years or the absence of patients with CTDs.A total of 9 studies met the inclusion criteria for pooling; however, the assessment of the unit location and study period identified possible patient crossover.A subsequent prioritization of biastreated and more recently published data resulted in the final inclusion of 7 observational studies (Figure 1).During subsequent research, a newly published study by David et al 16 was later identified, which provided more patient information from the same institution and study period as the included 2022 study by Elbatarny et al. 17 After discussion, we concluded that we would prioritize data from the previously published 2022 study because of the use of propensity score matching and subsequent reduction in selection bias.Quality assessment of the included studies evidenced all data to be high quality.The Newcastle−Ottawa scale assessment is visible in the attached Supplementary Material.Table 1

Outcomes
Five included studies reported long-term mortality in 435 patients with an estimated mean follow-up (EMFU) of Review/Reimplantation Versus Remodeling VSARR in CTDs 10.5 years.The pooled reimplantation group demonstrated a mortality incidence of 5.9% compared with 6.5% of patients who received remodeling VSARR (OR 0.66, 95% CI 0.30 to 1.48, I 2 = 30%, p = 0.32).Postoperative mortality did not achieve statistical significance between the patient intervention groups.However, in the sensitivity analysis, the exclusion of patients who received annuloplasty in addition to remodeling surgery (EMFU 10.9 years, 257 patients) led to an incidence of 7.5% mortality (OR 0.19, 95% CI 0.06 to 0.64, I 2 = 0%, p = 0.007).Pooling demonstrated postoperative survival in the reimplantation group to be superior to remodeling surgery without additional structural annuloplasty.
The rates of reoperation, with an EMFU of 10.3 years, were reported in an accumulative 449 patients with an incidence of 8.1% in the implantation group and 8.9% in the remodeling intervention group (OR 0.35, 95% CI 0.04 to 3.30, I 2 = 81%, p = 0.36).The sensitivity analysis excluding studies and subgroups of patients that received remodeling with annuloplasty VSARR (EMFU 10.6 years) demonstrated a reoperation incidence of 10.0% in the reimplantation group and 13.0% in the remodeling group (OR 0.43, 95% CI 0.05 to 3.53, I 2 = 71%, p = 0.43) (Figure 2).The meta-analysis of included studies did not reach statistical significance in the reoperation rate between reimplantation  and remodeling VSARR, regardless of the addition of structural annuloplasty in remodeling surgery.
The occurrence of moderate to severe postoperative aortic regurgitation in 402 patients with an EMFU of 10.9 years was recorded as 8.8% and 16.3% in patients who received reimplantation and remodeling VSARR (OR 0.56, 95% CI 0.31 to 1.02, I 2 = 47%, p = 0.06, respectively).The sensitivity analysis excluding patients who received annuloplasty in addition to remodeling surgery with an EMFU of 11.4 years demonstrated an occurrence of 10.1% in the reimplantation and 25.3% in the remodeling group (OR 0.23, 95% CI 0.10 to 0.53, I 2 = 47%, p = 0.0005, respectively) (Figure 3).Therefore, the sensitivity analysis demonstrates that patients with CTDs who received remodeling surgery in the absence of annuloplasty are significantly more likely to develop aortic insufficiency.

Discussion
Patients with CTDs are diagnosed earlier in life as a result of improved genetic testing and undergo repeated multi-imaging surveillance of aortic function and diameter.
In addition to medical management of blood pressure, offering prophylactic aortic root replacement is well established within the practice. 24,25Although the indication for prophylactic surgery alters between associated CTDs, the broad movement toward valve-sparing surgery considers clinical evidence and the burden of anticoagulation/reoperation in mechanical/bioprosthetic valves on the patient's quality of life. 26,27The use of prophylactic surgery in concordance with the young age of presenting patients with CTD provides a long-expected postoperative life expectancy.Therefore, this begs the question of whether preserving the native valve in VSARR provides the long-term durability necessary in patients with CTDs.In a 2022 study, Svensson et al 28 propensity matched and compared the long-term outcomes of reimplantation-only VSARR in 214 patients with CTDs against 645 patients without.The 10year follow-up reported no difference in reoperation, mortality, or aortic regurgitation rates between the study groups, demonstrating the structural integrity of the valves of patients with CTDs.The study excluded remodeling techniques because of perceived postoperative failure; however, with the starting study year of 1980, it is noted that Figure 2. Rates of reoperation forest plot.Meta-analysis of data from included studies on rates of reoperation and sensitivity analysis, excluding studies and patient subgroups that received annuloplasties in addition to remodeling. Figure 3.Rates of aortic regurgitation forest plot.Meta-analysis data from included studies on the occurrence of aortic regurgitation and sensitivity analysis, excluding studies and patient subgroups that received annuloplasties in addition to remodeling.remodeling developments with structural annuloplasty had not yet been implemented. 28Given the durability of the aortic valve after surgery, postoperative aortic regurgitation is attributed to annular dilation secondary to CTD vasculopathy; intraoperative repair of the aortic cusps is, at times, required to rectify pathology resulting from prolonged preoperative annular dilation. 29uring reimplantation, the native valve is sutured into the Dacron graft, protecting the reimplanted valve from dilation, as opposed to remodeling surgery in which the vascular graft is sutured into the remaining aortic wall above the insertion of valve leaflets, thereby providing minimal structural support against later annular dilation.Subsequently, the development of remodeling techniques used suture and ring extra-aortic structural annuloplasty to avoid redilation of the aortic annulus and postoperative aortic regurgitation by way of the stabilization of the atrioventricular junction. 25,30The clinical significance of additional structural support in remodeling surgery is evidenced in our meta-analysis; although the midterm rates of postoperative aortic regurgitation and mortality were not significant between the study groups, after sensitivity analysis and the exclusion of patients who underwent remodeling and annuloplasty, the outcomes favored reimplantation surgery.Although this meta-analysis does not assess the patient's cause of death, there may be an association of postoperative aortic regurgitation (in the absence of basal annuloplasty) and mortality because of the development of heart failure.The current body of knowledge does not allow the subgrouping of remodeling with annuloplasty alone versus reimplantation clinical outcomes.Previous studies suggest that surgeon and institution experience influences the occurrence of postoperative aortic regurgitation and requirement of reoperation, with high-volume centers reporting 10-year freedom from reoperation >90% in patients with Marfan syndrome. 31This may be explained by the more rigorous correction of the aortic valve, measuring the effective height and optimal cusp free margin alignment to avoid prolapse and later aortic regurgitation secondary to the reduction in aortic root diameter. 32,33revious techniques and clinical outcomes, primarily annular dilation, have led to an increased favoring of reimplantation VSARR in patients with CTDs, despite the consideration of postoperative hemodynamics.A Swiss study compared the fluid dynamic composition of postoperative structures after reimplantation and remodeling surgery in porcine research subjects.The results demonstrated an increased duration of high pressure (>150 mm Hg) and low shear stress in reimplantation surgery, as opposed to remodeling surgery, that produced hemodynamic models more comparable with the native root structures. 34The poor hemodynamic results may be associated with the use of straight tube grafts in reimplantation VSARR, whereas remodeling techniques preserve the sinuses of Valsalva.The introduction of bulb-shaped neo-Valsalva grafts in reimplantation surgery has produced satisfactory Valsalva to aortoventricular junction ratios, generating geometry more akin to natural physiology and improved valve durability. 35,36The assessment of reimplantation with neo-Valsalva grafts and remodeling with extra-aortic annuloplasty demonstrated similar hemodynamic results in multiple studies.The heterogeneity of human and porcine anatomic structures should be considered in view of the aforementioned studies. 37Real-world long-term patient data continue to report excellent results of reimplantation surgery using neo-Valsalva grafts. 38lthough patients with CTDs continue to be at a high risk of distal aortic dissection after root surgery and require life-long follow-up and assessment of aortic function, prophylactic VSARR continues to evidence improved longterm clinical outcomes. 39Given the complexity of VSARR surgery, the impact of surgeon preference and experience on patient outcomes should not be underestimated. 40Our analysis has demonstrated that remodeling without basal annuloplasty is not advisable in patients with CTDs.We believe this analysis provides the highest quality available data specific to patients of this demographic and warrants further research of CTD-specific VSARR outcomes given refined surgical techniques.Randomization, control of covariables, and prospective studies are essential to the improved validity of future research endeavors in this field.

Limitations
This meta-analysis has several limitations, including the limited study size included in the analysis.The occurrence of thromboembolic events and early outcomes, including the incidence of in-hospital mortality, stroke, and returning to theater for active bleeding, were reported by a significant number of studies to allow pooling.However, the infrequency of events in the context of low patient quantities did not allow meaningful analysis.In addition, the limited number of included studies resulted in a low-quality assessment of funnel plots and the possibility of reporting/publication bias.
Because of the nature of VSARR surgery in patients of this limited etiology, there is a lack of randomized data and statistical adjustment to treat for bias selection, with only 1 of the included studies using propensity score matching.The sensitivity analysis by way of nonbias-treated data from the study by David et al 16 (excluding the study by Elbatarny et al 17 ) only altered the statistical significance of long-term aortic regurgitation.
Several of the included studies produced subgroup data comparing reimplantation and remodeling surgery under the larger study group of VSARR.Subsequently, not all demographic data are available specific to our study groups (Table 2).

Conclusions
Remodeling techniques have demonstrated comparable midterm clinical outcomes in patients with CTDs compared with reimplantation VSARR surgery.The exclusion of patients who received structural annuloplasty produced superior rates of mortality and aortic regurgitation in the reimplantation study group.There may be postoperative physiologic advantages provided by remodeling surgery, and more research is required to assess the long-term effects of structural annuloplasty and remodeling surgery in patients of this demographic.
17−23lists the characteristics of the included studies.