The expected minimal effective orifice area (eEOA) required to avoid PPM was calculated preoperatively for all patients [eEOA = body surface area (BSA) × 0.85]. Following the analysis of the hemodynamic profiles provided by the manufacturers, prosthetic valves that provided similar or greater eEOA values were selected. ARE was performed in cases where the native aortic annulus diameter was smaller than that of the selected prosthetic valve.

Following aortic cross-clamping, a cardioplegic solution was administered to the aortic root or selectively to the coronary ostia. A “hockey stick” aortotomy and complete resection of the aortic leaflets were performed, followed by annular sizing, to assess the adequacy of the selected prosthetic valve. When indicated, ARE was performed first. Our preferred technique was Nick's technique (NT) (10), using a heterologous pericardial patch through an incision in the non-coronary cusp, extending to approximately 1.5 cm into the anterior mitral leaflet (Figure 1). The annular diameter was measured after ARE to ensure proper enlargement prior to prosthetic replacement. The mitral valve was then accessed through a standard left atriotomy and replaced using single interrupted sutures reinforced with pledgets. Particular attention was paid to maintaining the annulo-ventricular continuity by preserving the posterior leaflet or implanting artificial polytetrafluoroethylene (PTFE) chords. The aortic valve was successively implanted in a supra-annular position using single interrupted sutures reinforced with pledgets.

Statistical analysis was conducted using SPSS version 26.0 and Microsoft Excel 2016. The variables for descriptive analysis were expressed as proportions and mean ± standard deviation, while comparative analyses were performed using the chi-squared test. The correlation between variables was established by determining the p-value, which was considered statistically significant at p < 0.05.

The aortic cross-clamping duration was greater in the ARE-DVR group (177.6 ± 37.9 min vs. 148.3 ± 66.3 min, p = 0.047). The operative mortality rate was 5.6% for the entire cohort, and no statistically significant difference was found between the groups (ARE-DVR: 8% vs. DVR: 4.5%, p = 0.46) (Table 3). Table 4 summarizes the clinical data of the deceased patients. The postoperative mean aortic gradient was 19.68 ± 7.20 mmHg (range: 6.14–24 mmHg), with no statistically significant differences between the groups (ARE-DVR: 16.2 ± 9.9 mmHg vs. DVR: 17.3 ± 6.6 mmHg, p = 0.62). Table 5 reports the post-operative events according to age. At a mean follow-up of 6.9 ± 3.9 years, the estimated 5-year survival rates for ARE-DVR and DVR were 86.5 ± 7.2% and 89.9 ± 4.8%, respectively (p = 0.52) (Figure 2). Only one patient in the ARE-DVR group had undergone reoperation at follow-up for prosthetic valve endocarditis (n = 1/22, 4.5%).

PPM is caused by an inadequacy between the prosthesis orifice area and the body size of the patient. When severe, PPM is associated with increased cardiac events such as poor regression of the left ventricular mass and reduced survival (3). Although PPM occurs in all cardiac valves, it is more common after AVR, with a reported prevalence rate between 20% and 70% (3). Smaller aortic annular size, increased BSA, and younger age at surgery were found to be associated with a major risk of PPM (11, 12).

Although other preventive measures, such as the use of aortic sutureless valves, have progressively gained interest, ARE techniques have historically been advocated to avoid PPM. Castro et al. (13) reported a reduction in the incidence rate of PPM to 2.5% (compared with 17%) in 657 patients who underwent combined ARE and AVR, with no increase in operative mortality. In the largest comparative study by Rocha et al. (14), ARE-AVR and AVR patients had similar postoperative mortality outcomes despite the longer cardiopulmonary bypass (CPB) and X-clamping times in the ARE-AVR group. However, higher in-hospital mortality was found in patients who underwent ARE-AVR when other procedures (coronary bypass or other valve surgeries) were performed. This was corroborated by a recent meta-analysis that reported increased mortality in patients who underwent ARE-AVR in combination with other cardiac procedures (8, 15). In a cohort of 13,174 patients, Sà et al. (8) reported an increase in perioperative mortality in patients who underwent ARE-AVR associated with other procedures compared with those who underwent AVR and associated procedures alone. A successive meta-analysis of 40,447 AVR cases by the same authors confirmed previous findings, with a higher mortality rate in the ARE-AVR and concomitant procedures groups (p < 0.001) (15). Although this increased mortality is potentially related to patient factors rather than to ARE, the reluctance to associate ARE techniques with complex surgeries is understandable. Although ARE has been used for more than four decades, experience with such techniques remains poor globally and has been potentially associated with adverse procedural events such as bleeding, patch rupture, and death. Indeed, ARE has been reported in only 5.7%–26.3% of patients undergoing AVR in meta-analyses (8). Moreover, the controversial benefits of ARE in some groups and the growing interest in other techniques, such as transcatheter aortic valve replacement (TAVI) and sutureless valves, have contributed to limiting the need for ARE. In fact, TAVI and sutureless prostheses might be preferable in elderly patients with comorbidities undergoing concomitant procedures rather than a more time-consuming ARE-AVR (16, 17). This re-emphasizes the need to tailor PPM preventive strategies to specific cases, considering the physician's skills, device availability, patient clinical characteristics, and local context.

ARE is commonly indicated in sub-Saharan African (SSA) patients undergoing AVR (36.2% in our series). In fact, a large number of patients, including those with multiple valve diseases, present at surgery with a hypoplastic annulus due to their younger age and small BSA. In this subgroup, biological options, such as stentless or sutureless valves, are limited by early structural deterioration and prohibitive costs. Although technically challenging, the Ross procedure remains a valuable tool because it allows for the growth potential of the neo-aortic valve. However, it is less suitable for cases with multiple valvular lesions, and a high long-term failure rate has been reported in patients with rheumatic diseases (18–20).

To our knowledge, this is the first study to report ARE and double DVR in SSA. We routinely perform aggressive ARE in young patients requiring AVR or DVR by implanting adult-sized prostheses whenever possible, as the potential for growth in these patients remains a determining factor for the late recurrence of PPM. In addition to ARE, high hemodynamic profile prostheses (21) (Medtronic Advance Performance, SJM Hemodynamic Plus, or Regent) were preferred in our aortic patients, representing 42% of the implanted valves. NT has been our preferred ARE technique. When used with the appropriate technique, NT is associated with excellent root stability in the long term, even with an autologous pericardial patch (22). All patients received a larger aortic prosthesis, and no technical difficulties regarding prosthesis implantation were reported despite concomitant adult-size mitral prostheses. We believe that NT is less time-consuming than Manougian's technique. The latter might require a deeper extension in the anterior mitral leaflet and dome of the left atrium, with extensive patch reconstruction time. Despite the higher procedural duration in ARE-DVR patients, no significant differences were found in operative mortality between the two groups. Similar findings were reported by Okuyama et al. (23) and Zhong et al. (24), suggesting that ARE is not an independent factor for mortality in DVR despite the increased operative time. This contrasts with some meta-analyses that reported increased perioperative mortality in patients undergoing ARE-AVR and associated cardiac procedures (15, 15). It is possible that differences in patient demographics and clinics between the current and previous studies led to the observed heterogeneous outcomes. While our study was performed in an SSA environment with younger patients with Rheumatic Heart Diseases (RHD), the patients in the meta-analyses were from Western countries, were older, and presented with degenerative etiologies, including coronary diseases, and high comorbidity rates. As stated earlier, the strategy for PPM prevention, including the indication for ARE, should be tailored to the specific case following an exhaustive assessment of patient characteristics, disease patterns, and team expertise.

The small number of cohorts that have potentially impacted the statistical power of our analysis is a limitation of this study.

In conclusion, the association of the ARE technique with DVR does not significantly affect operative mortality in young SSA with RHD. ARE can be safely used whenever indications arise to reduce the occurrence of PPM and the risk of reoperation, especially in patients with growth potential.