Revisiting ascending aortic resection in the elective valve-sparing root replacement: assessing the benefits and necessity of hemiarch replacement at three centres^†

Abstract

OBJECTIVE

The aim of this study was to determine the indication and optimal timing for performing a hemiarch procedure in patients undergoing valve-sparing root replacement (VSRR).

METHODS

We conducted a retrospective study on 986 patients undergoing VSRR at three tertiary care centres. Inclusion criteria were all patients undergoing elective VSRR. Exclusion criteria were age <18 years, Stanford type A dissection, dissection in the arch, total aortic arch replacement or previous aortic arch replacement. We performed propensity score matching in a 1:1 ratio. The primary end-point is a composite outcome that includes mortality, aortic arch reintervention, new aortic dissection during follow-up and cerebrovascular incidents within the first 30 days.

RESULTS

A total of 401 patients (41%) had a hemiarch replacement, while 585 (59%) did not. Root phenotype was present in 565 (57%). The mean follow-up time was 4.7 years (SD ± 4.6). In the matched population, there was no significant difference in the 10-year freedom from the composite outcome between the non-hemiarch and hemiarch groups (87.3% vs 85.0%, P > 0.999). Similarly, no difference was found for aortic reinterventions (P = 0.13) or survival (P = 0.5). This was also true for patients with heritable thoracic aortic disease. However, in patients with a bicuspid aortic valve, the intervention rate was significantly higher in the hemiarch group (10.8% vs 0%, P = 0.016). There was no significant difference in the 30-day incidence of cerebrovascular accidents between the groups (5% vs 2.7% in the hemiarch group, P = 0.117). Only the distal ascending diameter showed a tendency with better outcome over 45 mm for the hemiarch procedure; otherwise, we found no reliable cut-off values based on ascending length, diameter-to-height index or ascending length-to-height index.

CONCLUSIONS

Our findings conclusively demonstrate that concomitant hemiarch replacement does not increase the perioperative risk in young patients undergoing VSRR. However, concomitant replacement does not seem to protect from aortic reinterventions during medium-term follow-up.

Open in new tabDownload slide

INTRODUCTION

Long-standing controversy remains in the extent of aortic resection required in patients undergoing elective valve-sparing root replacement (VSRR). The main debate revolves around whether to perform a hemiarch replacement or not [1–3]. The hemiarch approach carries additional risks due to the necessary circulatory arrest time, while performing the anastomosis on the clamped aorta may leave the patient at risk for future aortic events due to residual aortic tissue [4–6].

The most recent aortic guidelines recommend a hemiarch procedure for proximal aortic arch diameters exceeding 45 mm [7, 8]. The current literature indicates that the additional hemiarch procedure does not increase the risk of operative complications or compromise safety [9–13]. However, it also suggests that there is no additional benefit to perform a hemiarch in patients without an arch aneurysm [4, 9–12, 14]. Nevertheless, none of the mentioned studies focused specifically on patients undergoing VSRR, a population that is often younger or has heritable thoracic aortic disease (HTAD).

Our hypothesis was that in a dedicated aortic centre, the risk of circulatory arrest is not significantly increased, justifying a lower threshold for prophylactic hemiarch procedures in this population, ultimately reducing the likelihood of secondary aortic reinterventions.

The aim was to guide surgeons in deciding whether to perform a hemiarch procedure in this young population. To achieve this, we established a transatlantic registry across three tertiary care centres, including patients who have undergone elective VSRR. In this registry, we retrospectively compared the outcomes of patients who underwent the hemiarch procedure with those who did not.

PATIENTS AND METHODS

Ethics statement

This retrospective study was approved by the University of Pennsylvania institutional review board (Protocol #855137), by the institutional review committee of the University of Freiburg (number 24-1247-S1-retro) and by the Swiss Association of Research Ethics Committees (number: 01534). Informed consent was waived due to the retrospective nature of this chart-based review.

Study population

All patients who underwent VSRR (David reimplantation technique) between January 2004 and March 2024 were identified from the three participating centres: the University of Pennsylvania, USA; the University of Freiburg, Germany; and the Inselspital, University of Bern, Switzerland. After anonymization, a transatlantic registry was created using the Research Electronic Data Capture (REDCap) system. Exclusion criteria were: patient age <18 years, Stanford Type A dissection, dissection in the arch, total aortic arch replacement or previous aortic arch replacement.

The observational design adhered to the STROBE statement and STROCSS 2021 guidelines [15–17].

Variables and outcomes

The aim of this study was to determine the indication and optimal timing for performing a hemiarch procedure in patients undergoing VSRR. Due to the small number of postoperative adverse events in previous studies, we decided to use a composite end-point as the primary outcome [4, 9–12, 18, 19].

The composite end-point consists of mortality, aortic reintervention, new aortic dissection during follow-up and cerebrovascular incidents within the first 30 days.

Aortic reintervention was defined as any intervention on the ascending aorta and aortic arch but without aortic valve reintervention.

Mortality and neurologic events were defined according to the Valve Academic Research Consortium-3 (VARC-3) [20]. Cerebrovascular incidents were defined as stroke, symptomatic hypoxic-ischaemic injury and covert central nervous system injury, excluding transient ischemic attack (TIA) and delirium.

The secondary outcomes were defined as mortality and aortic reintervention competing against death.

The same analyses were performed for the subgroups consisting of patients with HTAD and patients with a bicuspid aortic valve. Baseline demographics, comorbidities, concomitant procedures, complications and survival data, as well as operative details, were recorded. Aortic dimensions were measured either from CT scans or from MRI. In accordance with consensus guidelines, measurements of the aorta were performed perpendicular (in multiplanar reconstruction) to the long axis of the aorta at specific segments (annulus, aortic root, ascending aorta, ascending aorta 2 cm proximal to the origin of the brachiocephalic trunk and the aortic arch between the brachiocephalic trunk and left common carotid artery) [7]. The measurements were performed using the outer-to-outer edge method (including the aortic wall) [7].

Core temperature was categorized into two groups: 24°C to 28°C and below 24°C. In all three centres, core temperatures were measured, defined as either bladder or rectal temperatures.

Hemiarch replacement was defined as an open anastomosis at the level of the brachiocephalic trunk. Non-hemiarch procedure was defined as a distal anastomosis performed with the aortic cross-clamp in place and without circulatory arrest. ‘Root phenotype’ was defined as having a greater root diameter than ascending diameter [7].

Statistical analysis

Continuous variables are presented as mean with standard deviation (SD) or median and interquartile range (IQR) depending on whether the variable is normally distributed. The differences in the distribution of preoperative, operative and postoperative characteristics in the hemiarch and the non-hemiarch groups were tested with χ² analysis, or Fisher’s exact test when necessary, for the categorical variables. t-Test or the Wilcoxon test was applied for continuous variables. A P-value <0.05 was considered statistically significant.

In addition to descriptive statistics, Kaplan–Meier survival analysis was conducted for the composite end-point and mortality, with comparisons made using the log-rank test. Cumulative incidence rate of reintervention with competing risk of death is estimated by the proportional subdistribution hazards model and compared by Gray’s test. Time for follow-up began on the surgery date and ended on the date of an event or the most recent clinical and radiological follow-up. Propensity score matching utilized the nearest neighbour method with 0.2 as calliper. Kaplan–Meier survival and cumulative incidence curves in matched sets were performed. To assess balance between the groups, the standardized mean difference was used, with a value of ≤0.1 set as the acceptable threshold. The only exception was the diameter of the ascending aorta, for which a standardized mean difference of 0.128 was accepted due to large distribution discrepancy between the original hemiarch and non-hemiarch groups.

The null hypothesis of our study was that there is no significant difference between patients with and without the hemiarch procedure.

Analysis was performed with R software (version 4.3.2, R Foundation for Statistical Computing).

Variables for propensity score matching were selected based on commonly used parameters in the aortic surgery literature [4, 10, 12, 21]. These included all parameters from Table 1, excluding radiographic features, as well as the ascending aortic diameter measured 2 cm proximal to the origin of the brachiocephalic trunk, categorized as <45 mm or ≥45 mm.

RESULTS

Overall population

Out of the 986 patients included in this study, 401 (41%) underwent a hemiarch procedure, leaving 585 (59%) without hemiarch replacement. Only 54 (5.5%) patients had a distal ascending aortic diameter of ≥45 mm. The baseline and perioperative characteristics are detailed in Table 1 and Supplementary Material, Table S1, while the postoperative outcomes are presented in Table 2. Follow-up was complete for 88.2% of patients. Mean follow-up time was 4.7 years (SD ± 4.6).

Freedom from composite outcome at 10 years was 87.8% (95% CI 83.8–92.0) for the non-hemiarch group and 83.6% (95% CI 78.7–88.9) for the hemiarch group (P = 0.8, Fig. 1A).

We found no significant difference in the 30-day mortality between the groups [4 (0.7%) vs 1 (0.2%) hemiarch, P = 0.654]. Survival at 10 years was 96.3% (95% CI 93.4–99.3) for the non-hemiarch group and 90.8% (95% CI 86.4–95.3) for the hemiarch group (P = 0.09, Fig. 2A).

Cumulative incidence of aortic reintervention with competing risk of death at 10 years was significantly higher in the hemiarch group [1.4% (95% CI 0.3–4.1) vs 4.5% (95% CI 2.2–8.0), P = 0.047, Fig. 3A].

There were no differences between the two groups regarding cerebrovascular accident, new aortic dissections during follow-up or their corresponding 30-day outcomes (Table 2 and Supplementary Material, Table S5).

Matched population

After 1:1 propensity score matching, the original cohort was reduced to 730 patients, with 365 patients in each group.

The baseline and perioperative characteristics are detailed in Table 1 and Supplementary Material, Table S1, while the postoperative outcomes are presented in Table 2.

Freedom from composite outcome at 10 years was 87.3% (95% CI 82.2–92.8) for the non-hemiarch group and 85.0% (95% CI 80.0–90.3) for the hemiarch group (P > 0.999, Fig. 1B).

The cumulative incidence of aortic reintervention did not reach statistical significance [0.9% (95% CI 0.3–2.6)] for the non-hemiarch group vs 4.7% (95% CI 2.2–8.5) for the hemiarch group, P = 0.13, Fig. 3B and Table 2).

There was no significant difference in mortality, cerebrovascular accidents, new aortic dissection during follow-up or their corresponding 30-day outcomes (Table 2 and Supplementary Material, Table S5). Freedom from death at 10 years: 95.2 (95% CI 91.0–99.5) for the non-hemiarch group and 92.6 (95% CI 88.5–96.9) for the hemiarch group, P = 0.5 (Fig. 2B).

Patients with heritable thoracic aortic disease

In total, 213 patients had a proven HTAD. This included 145 patients with Marfan syndrome (68.1%), 30 patients with Loeys-Dietz syndrome (14.1%) and 38 patients with other HTAD (including non-syndromic HTAD; Supplementary Material, Table S2).

At 10 years, the hemiarch and non-hemiarch groups showed no significant difference in the composite end-point (P > 0.999, Supplementary Material, Fig. S1) or the cumulative incidence of aortic reinterventions (P = 0.15, Supplementary Material, Fig. S2). Survival was 100% in both groups.

No significant differences were observed between groups for cerebrovascular accidents, new aortic dissections or 30-day outcomes (Supplementary Material, Table S2).

Patients with bicuspid aortic valve

Out of 986 patients with VSRR, 266 patients (27.0%) had a bicuspid aortic valve (Supplementary Material, Table S3). Freedom from composite end-point at 10 years was 90.2% (95% CI 84.8–95.9) for the non-hemiarch group and 83.6% (95% CI 74.9–93.4) for the hemiarch group (P = 0.7).

Cumulative incidence of aortic reintervention with competing risk of death at 10 years was 0% in the non-hemiarch group vs 10.8% (95% CI 4.1–21.2) in the hemiarch group (P = 0.016, Supplementary Material, Fig. S3). Infection was responsible for 71% (5 out of 7) of the re-interventions. Of the remaining two, one patient experienced an aortic dissection during follow-up, and the other had progressive aortic diameter and aortic valve regurgitation. The multivariate Cox proportional hazards model revealed that patients with a bicuspid aortic valve had a significantly higher risk of aortic arch reintervention due to infection (HR: 2.2, P = 0.013).

There were no significant differences between the two groups regarding mortality, cerebrovascular incidents, new aortic dissections during follow-up or 30-day outcomes.

Long-term mortality and composite end-point predictors

In Cox regression multivariable analysis in the unmatched cohort, only age [HR: 1.06 (95% CI 1.02–1.09), P ≤ 0.001] was significantly associated with mortality over time.

For the composite end-point, root phenotype [HR: 0.50 (95% CI 0.32–0.80), P = 0.003], age [HR: 1.02 (95% CI 1.0–1.04), P = 0.024] and previous cardiac surgery [HR: 2.61 (95% CI 1.04–6.54), P = 0.041] were significantly associated with the outcome over time.

Outcomes based on aortic measurements

A distal ascending aortic diameter >45 mm was associated with significantly better composite outcomes in the hemiarch group [4 (57%) vs 5 (14%), P = 0.028]. However, this difference did not remain significant in the regression analysis (HR 3.67, P = 0.059).

Otherwise, we found no significant difference in the composite end-point when analysed across the different groups based on ascending aortic length, ascending aortic diameter–height index or ascending aortic length–height index (Fig. 4).

There was no significant difference in survival when analysed across the different groups based on ascending aortic diameter (2 cm proximal to the brachiocephalic trunk), ascending aortic length, ascending aortic diameter–height index or ascending aortic length–height index (Supplementary Material, Fig. S4).

Root phenotype

Among the 986 patients, 565 (57%) exhibited a root phenotype, 319 (32%) did not, and the phenotype status was unknown for 102 patients (10%). 183 with a root phenotype and 162 without a root phenotype underwent a hemiarch procedure. In the root phenotype group, 171 patients (30.3%) had HTAD, compared to just 26 patients (8.2%) in the non-root phenotype group. Patients with a root phenotype were significantly younger (age 47.9 years vs 56.9 years, P < 0.001). Among patients who underwent a hemiarch procedure, the distal ascending aortic diameter significantly differed between the root phenotype and non-root phenotype groups (37.0 mm vs 40.6 mm, P < 0.001).

There was no significant difference in composite outcomes between root phenotype and hemiarch procedure groups (Fig. 5).

However, as previously noted, the root phenotype was associated with a significantly lower risk for the composite end-point, even after adjusting for other variables such as age and HTAD, with a hazard ratio of 0.50 (95% CI 0.32–0.80, P = 0.003).

Neurological outcomes

In the unmatched cohort, 40 patients (4.1%) experienced cerebrovascular accidents (29 in the non-hemiarch group and 11 in the hemiarch group, P = 0.117) during the 30-day follow-up. Among these patients, six (0.6%) suffered severe disability (Rankin scale > 2, assessed between 30 and 90 days), and one patient died as a result of the cerebrovascular accident (Supplementary Material, Table S4). By the end of follow-up, a total of 42 patients in the non-hemiarch group and 23 patients in the hemiarch group had experienced a cerebrovascular accident (P = 0.443, Supplementary Material, Table S5).

The anatomy of the circle of Willis was available for 173 patients (17.5%). Out of 173 patients, 22 suffered a cerebrovascular accident. We found no significant difference in the incidence of cerebrovascular accidents between patients with a normal circle of Willis anatomy and those with anatomical variations (18 patients with a normal anatomy and 4 patients with a variation, P = 0.855).

DISCUSSION

Analysing one of the largest multicentre cohorts in current literature, we identified the following key findings: (1) adding a hemiarch replacement does not increase operative risk, particularly concerning cerebrovascular accidents; (2) patients with a root phenotype are not at substantially increased risk of more distal aortic disease progression; and (3) measurements such as ascending aortic diameter, ascending aortic diameter–height index, ascending aortic length and ascending aortic length–height index are not sufficient to guide the decision on whether a hemiarch replacement is necessary.

The most recent aortic guidelines recommend a concomitant hemiarch for proximal aortic arch diameters exceeding 45 mm [7, 8]. This recommendation is primarily based on findings from a retrospective review, which analysed 1132 patients without aortic arch aneurysm undergoing proximal aortic aneurysm repair. This study found that hemiarch repair may be associated with higher short-term mortality compared to non-hemiarch procedures [4]. Greason et al. reviewed 702 patients with bicuspid aortic valve replacement with/without hemiarch procedure and found no difference in survival [9]. Smaller studies also did not find a difference in mortality [10–12].

One reason why this comparison is rather difficult to interpret is due to the fact that patients with a hemiarch procedure have a more severe aortic disease, which could explain a higher mortality or even aortic reintervention rate. However, our findings, along with most other studies, do not indicate a difference in survival.

A recent retrospective study involving 781 patients who underwent David V VSRR demonstrated a low incidence of aortic reintervention. Only 31 patients (4.0%) required additional aortic-related procedures, with 23 of these cases (74%) involving distal aortic reinterventions [5]. Notably, Pearsall et al. found no significant difference in aortic arch reintervention rates at 10 years between patients who underwent hemiarch repair (1%) and those who did not (1.3%). Similarly, Greason et al. did not find a difference (zero aortic arch reintervention). In our study, the unmatched population showed higher aortic reintervention rates at 10 years for patients who underwent hemiarch repair. However, in the matched population, this difference was not statistically significant. This discrepancy is likely attributable to the more severe aortic phenotype observed in the unmatched hemiarch group, as reflected by the significantly larger aortic diameters (Table 1).

Remarkably, none of the patients with a bicuspid aortic valve who underwent clamped aortic repair required a subsequent aortic reintervention. Infection accounted for 71% of reinterventions in patients with a bicuspid aortic valve who underwent hemiarch procedures, with a hazard ratio of 2.2 for reintervention due to infection. All affected patients had undergone leaflet repair. One possible explanation is that patients with a bicuspid aortic valve, leaflet repair and hemiarch procedure may be at an increased risk of infection due to the presence of more foreign material.

Due to the low rates of aortic reintervention and previously reported cerebrovascular accidents, we decided to focus on a composite end-point [4, 9–12]. In our cohort, there was no significant difference in the composite end-point, further highlighting the low risk associated with adding a hemiarch procedure.

Surgeons who avoid the hemiarch procedure often argue that it increases the risk of cerebrovascular accidents. However, our findings indicate that in dedicated aortic centres, this risk is not elevated.

Although the risk associated with a hemiarch procedure does not appear to be elevated, we were unable to demonstrate a clear benefit from this intervention. This may be due to a follow-up period that was too short or because, despite propensity score matching, the severity of aortic disease may not have been fully accounted for.

Patients with the same ascending aortic diameter have sometimes received different treatments. Additionally, a significant portion of patients received a hemiarch procedure despite having a distal ascending aortic diameter of less than 45 mm. Surgeons are generally reluctant to leave pathological tissue behind, and we believe that this variability in treatment is largely due to individual surgical decisions influenced by factors beyond our full assessment. These factors may include patient frailty, the quality of the aortic tissue, as well as the surgeon’s experience and clinical judgement. While these are essential considerations in determining the best course of action, our goal is to further support and refine the decision-making process. To align with current guidelines, which recommend using additional dimensional criteria and indices, we aimed to incorporate these recommendations to guide the decision on whether to perform a hemiarch procedure [7, 8]. We analysed ascending aortic diameter, aortic length and indexed parameters. Although patients with a distal ascending aortic aneurysm >45 mm who underwent a hemiarch procedure showed better composite outcomes, this finding was not significant in the regression analysis, likely due to the limited sample size.

Unfortunately, none of the other metrics proved effective in guiding the decision for a hemiarch procedure.

This is an intriguing finding, as Kari et al. demonstrated that maximum aneurysm size at the time of surgery was a predictor of mid-term mortality, as identified through multiple regression analysis [22].

We also investigated the aortic root phenotype. Although there was no difference in freedom from the composite end-point, patients with an aortic root phenotype had a significantly lower risk for the composite end-point. This suggests that patients with a root phenotype may have issues confined to the root alone.

After matching and analysing composite end-points, mortality, aortic arch reoperations and the diameter and length of the ascending aorta, we were unable to identify a definitive cut-off value to guide the decision for a hemiarch procedure. Although a trend was observed for patients with a distal ascending aortic diameter >45 mm to undergo a hemiarch procedure, this finding did not reach statistical significance. The take-home message is that routine prophylactic hemiarch replacement may not be necessary for all patients undergoing VSRR, as it does not seem to confer long-term protective benefits against reintervention. However, it is reassuring that when hemiarch replacement is performed, it does not increase perioperative risk, leaving room for its selective use in patients with more severe aortic pathology or other high-risk features. In daily clinical practice, we advocate for a patient-specific approach. Surgeons should weigh the anatomical characteristics, including the diameter and extent of the ascending and proximal arch aneurysm, alongside their clinical judgement. Our findings underline the variability in decision-making and the importance of tailoring the surgical strategy to the individual patient rather than adopting a blanket preventive strategy.

Limitations and strengths

This study presents a retrospective observational analysis and is therefore subject to all limitations of such a study design. We could not completely account for patient selection bias due to the variability in surgeon discretion for selecting patients for hemiarch replacement. However, we believe that our careful propensity score matching process helped to reduce potential confounders.

The fact that a significant proportion of patients have a distal ascending aortic diameter of less than 45 mm can be seen as both a limitation and a strength. On one hand, this could introduce a selection bias, even though it reflects real-world data. On the other hand, the differences we observed may hold greater significance, as one would expect a lower incidence of findings with smaller diameters.

The strength of this study is the large sample size and the granularity of the data. Moreover, this is the first study to incorporate aortic diameter into propensity score matching to compare outcomes of the hemiarch procedure.

We provide new insights into the long-standing discussion on the extent of ascending aortic resection required during VSRR.

CONCLUSION

Our findings conclusively demonstrate that concomitant hemiarch replacement does not increase the perioperative risk in young patients undergoing VSRR. However, concomitant replacement does not seem to protect from aortic reinterventions during medium-term follow-up.

SUPPLEMENTARY MATERIAL

Supplementary material is available at EJCTS online.

FUNDING

This study received no external funding.

Conflict of interest: the authors declare that there are no conflicts of interest related to this manuscript.

DATA AVAILABILITY

The authors confirm that the data supporting the findings of this study are available within the article and/or its supplementary materials.

Author contributions

Murat Yildiz: Conceptualization; Data curation; Formal analysis; Investigation; Methodology; Project administration; Supervision; Validation; Visualization; Writing—original draft; Writing—review & editing. Florian Schoenhoff: Conceptualization; Data curation; Formal analysis; Investigation; Methodology; Validation; Visualization; Writing—review & editing. Victoria Werdecker: Conceptualization; Data curation; Formal analysis; Investigation; Validation. Maria Nucera: Conceptualization; Data curation; Formal analysis; Investigation; Validation. Selim Mosbahi: Conceptualization; Data curation; Investigation. Yu Zhao: Data curation; Formal analysis; Investigation; Methodology; Validation; Visualization; Writing—review & editing. Nicholas Goel: Conceptualization; Data curation; Investigation; Methodology; Writing—review & editing. Mikolaj Berezowski: Conceptualization; Data curation; Formal analysis; Investigation; Methodology. Kendall Lawrence: Conceptualization; Data curation; Investigation. Sankrit Kapoor: Data curation; Investigation; Project administration; Validation. Maximillian Kreibich: Conceptualization; Data curation; Investigation; Validation; Visualization. Tim Berger: Conceptualization; Data curation; Investigation. Joseph Kletzer: Conceptualization; Data curation; Methodology. Joseph Bavaria: Data curation; Investigation; Methodology; Visualization; Writing—review & editing. Wilson Szeto: Conceptualization; Methodology; Validation; Visualization; Writing—review & editing. Matthias Siepe: Conceptualization; Methodology; Validation; Visualization; Writing—review & editing. Martin Czerny: Conceptualization; Methodology; Validation; Visualization; Writing—review & editing. Nimesh Desai: Conceptualization; Data curation; Methodology; Validation; Visualization; Writing—review & editing

Reviewer information

European Journal of Cardio-Thoracic Surgery thanks Yutaka Okita, Luca Di Marco, Anthony L. Estrera, David G. Healy and the other, anonymous reviewer(s) for their contribution to the peer review process of this article.

Footnotes

REFERENCES

ABBREVIATIONS

 
• CT

  computed tomography

 
• HTAD

  heritable thoracic aortic disease

 
• MRI

  magnetic resonance imaging

 
• TIA

  transient ischemic attack

 
• VSRR

  valve-sparing root replacement