Repair of mitral prolapse: comparison of thoracoscopic minimally invasive and conventional approaches

Abstract

Open in new tabDownload slide

OBJECTIVES

Surgical repair remains the best treatment for severe primary mitral regurgitation (MR). Minimally invasive mitral valve surgery is being increasingly performed, but there is a lack of solid evidence comparing thoracoscopic with conventional surgery. Our objective was to compare outcomes of both approaches for repair of leaflet prolapse.

METHODS

All consecutive patients undergoing surgery for severe MR due to mitral prolapse from 2012 to 2020 were evaluated according to the approach used. Freedom from mortality, reoperation and recurrent severe MR were evaluated by Kaplan–Meier method. Differences in baseline characteristics were adjusted with propensity score-matched analysis (1:1, nearest neighbour).

RESULTS

Three hundred patients met inclusion criteria and were divided into thoracoscopic (N = 188) and conventional (sternotomy; N = 112) groups. Unmatched patients in the thoracoscopic group were younger and had lower body mass index, New York Heart Association class and EuroSCORE II preoperatively. After matching, thoracoscopic group presented significantly shorter mechanical ventilation (9 vs 15 h), shorter intensive care unit stay (41 vs 65 h) and higher postoperative haemoglobin levels (11 vs 10.2 mg/dl) despite longer bypass and cross-clamp times (+30 and +17 min). There were no differences in mortality or MR grade at discharge between groups nor differences in survival, repair failures and reinterventions during follow-up.

CONCLUSIONS

Minimally invasive mitral repair can be performed in the majority of patients with mitral prolapse, without compromising outcomes, repair rate or durability, while providing shorter mechanical ventilation and intensive care unit stay and less blood loss.

INTRODUCTION

Primary mitral regurgitation (MR) is the most frequent type of mitral valve dysfunction in developed countries and, within this group, degenerative MR is the most common aetiology [1]. Valve abnormalities in degenerative MR are in most cases amenable to successful surgical repair in centres with experience [2, 3] with a class I recommendation by European and American Guidelines [4, 5]. Furthermore, current guidelines advocate for surgical repair in patients with severe MR who present with a flail leaflet or left atrial enlargement, even if asymptomatic and left ventricular function remains normal, provided a successful repair could be expected with a low perioperative risk.

There is a wide variety of options for mitral repair, including the approach, the repair techniques or the annuloplasty devices used. Currently, the gold-standard treatment implies performing the operation through a full median sternotomy. However, over the last 20 years, minimally invasive mitral valve surgery (MIMVS) techniques have been developed and increasingly performed [6]. Data from large series have shown that MIMVS provides excellent results, comparable to those of conventional surgery without a trade-off in repair rate or durability, associated with less surgical trauma and a faster recovery [7–9]. However, some authors have reported an increased risk of neurological complications, aortic dissection and other complications with the vascular access used in MIMVS [10], but most of these studies were performed in the earlier experience with systematic use of endoaortic balloon-occlusion [11].

Our aim was to examine the contemporary early and mid-term results of mitral repair for degenerative MR according to the surgical approach used: minimally invasive or median sternotomy.

METHODS

Ethical statement

This study was approved by the Institutional Ethics Committee of the Hospital Clinic of Barcelona, Spain (HCB/2021/0248) on 3 February 2021. The need for individual written informed consent was waived.

Study population

We evaluated all consecutive patients treated in our centre from November 2011 to November 2020 with 4+ primary MR secondary to degenerative disease who underwent surgical intervention, regardless of the planned procedure. We excluded patients with previous interventions and those requiring additional concomitant procedures other than atrial fibrillation (AF) ablation or closure of patent foramen ovale, to facilitate direct comparisons of both approaches. Furthermore, patients with contraindications for MIMVS were excluded: severe aorto-iliac disease (contraindication for femoral cannulation and retrograde perfusion), ascending aorta aneurysm (>45 mm), a severely calcified mitral annulus, >2+ aortic regurgitation or previous surgery on the right chest. The choice of the approach was at the discretion of the attending surgeon.

Surgical techniques

Mitral repair strategy included a large variety of standardized techniques but basically consisted in leaflet resection and/or chordal replacement to treat the prolapse of the posterior leaflet and chordal replacement to treat anterior leaflet prolapse. A small number of patients, with more complex forms of the disease, needed additional techniques such as chordal transposition, chordal shortening, papillary muscle repositioning and commissural plication. The choice of annuloplasty devices was made at the discretion of the attending surgeon and included a variety of flexible bands and complete, rigid or semi-rigid rings. All patients were operated under normothermic extracorporeal circulation and using intermittent antegrade cold blood-cardioplegia.

MIMVS was performed using specifically designed, long-shafted endoscopic instruments inserted through a working port created with a right lateral mini-thoracotomy (3–4 cm) in the 4th or 5th intercostal space, or with a periareolar incision in a minority of selected cases (Fig. 1). Two additional 5-mm ports were placed in the right hemithorax for a 5-mm 30° video camera and continuous CO₂ insufflation (3rd/4th intercostal space, mid-axillary line) and for the left atrial retractor (4th/5th intercostal space, midclavicular line). Finally, a 10-mm port was placed for venting the left chambers and to introduce a pleural drain after the intervention (6th right intercostal space, anterior axillary line). Our routine approach to perform these procedures involves cannulation of the right common femoral artery and vein using single femoral venous cannulation, thus avoiding placing a second jugular cannula. Direct transthoracic aortic clamping and antegrade blood-cardioplegia infusion directly in the aortic root were performed in all cases for cardioplegic arrest and myocardial protection, similar to the conventional approach through median sternotomy. All procedures were performed through a left atriotomy.

Outcomes and sources of data

The primary outcomes of this study were mitral repair rate, mortality and mayor complications (stroke, acute renal failure) rates. Secondary outcomes were all other postoperative complications, transfusion and blood loss and duration of mechanical ventilation and intensive care unit (ICU) and hospital stay. Time-related follow-up outcomes included survival, freedom form reoperation and from mitral replacement and freedom from recurrent MR. Data on patient demographics, baseline characteristics, perioperative outcomes, follow-up and vital status were obtained from our institutional databases. All patients had a comprehensive preoperative transthoracic echocardiography, intraoperative transoesophageal echocardiographic monitoring and a complete transthoracic echocardiography before hospital discharge. Patients were followed in our outpatient clinic with annual echocardiography. MR was graded following current recommendations [12]. Follow-up data gathering was completed (and the dataset closed) on November 2022.

Statistical analysis

Descriptive statistics for categorical variables were reported as frequency and percentage, and comparisons were performed using a chi-squared test or the Fisher’s exact test. Continuous variables were reported as mean/standard deviation or as median/interquartile range depending on normality and comparisons were performed using the Wilcoxon–Mann–Whitney test. The Kaplan–Meier method was used to calculate survival time-related events and comparisons were performed using Cox regression. All comparisons were performed on an intention-to-treat basis.

Propensity score matching

The propensity scores (PS) for undergoing surgery using the MIMVS approach were estimated by means of a logistic regression model. Bias reduction and balance between the groups were assessed with standardized difference of covariates. Based on the logit scores obtained and using the nearest neighbour method (1:1, without replacement, without caliper), a cohort of 103 pairs was selected to perform unbiassed comparisons between both surgical approaches. Supplementary Material, Table S1 and Supplementary Material, Figs. S1 and S2 provide a full description of the variables before and after PS matching and the standardized bias reduction accomplished. Two-sided P values <0.05 were considered significant and standardized mean differences (SMD) were calculated for all comparisons. All survival analyses were stratified by matched pairs and using robust variance-covariance matrix estimation, with the Cox proportional hazards model. All analyses were performed using STATA^® (StataCorp, USA).

RESULTS

Preoperative characteristics

A total of 300 consecutive patients met the selection criteria during the study period and were included. Of those, 188 patients underwent a minimally invasive thoracoscopic mitral repair (MIMVS group; 63%) and 112 had surgery through a median sternotomy (Conventional group; 37%). Detailed data on volume and distribution of cases are shown in Fig. 2. Baseline characteristics are shown in Table 1. MIMVS patients were younger (59 vs 64 years; P < 0.01) and had lower body mass index (25 vs 26; P = 0.02) and EuroSCORE II (1.61% vs 2.28%; P < 0.01), less chronic renal disease (2.5% vs 8%; P < 0.05) and a higher percentage was in New York Heart Association class I or II (75% vs 52.5%; P < 0.01). All preoperative characteristics were balanced after PS adjustment with no significant differences between groups at baseline in the matched cohort. All patients had 4+ degenerative MR and the prevalence of more complex morphologies (anterior, commissural or bileaflet prolapse) was high, accounting for more than one-third of cases, with the same proportion in both groups (Table 1).

Cardiac intervention

A successful mitral repair was achieved in 295 patients (>98%), with only 5 patients requiring mitral replacement, 1 in the MIMVS group and 4 in the Conventional group (99.5% vs 96.5%; P = 0.07). Among the MIMVS group, there was only 1 conversion to median sternotomy (0.5%).

Multiple repair techniques were used on both groups, with no significant differences between them (Table 2). Concomitant AF ablation was performed in 36 patients with similar frequency in both groups (11% vs 13.5%; P = 0.58).

In the overall population, operative times were significantly longer in the MIMVS group compared to the Conventional group (+25 min on cardiopulmonary bypass (CPB) and +15 min on cross-clamp duration; P < 0.01). Similar differences were also found after matching, where patients in the MIMVS group required a longer CPB (+30 min, SMD = –0.70) and cross-clamp (+17 min, SMD = –0.54).

Perioperative clinical outcomes

There were no significant differences in major postoperative complications, including stroke (<1%), permanent pacemaker implantation (3%), postoperative renal failure (2%), reintervention (5.4%) and tamponade (1.3%). Detailed information is presented in Table 3.

In the entire cohort, there were no significant differences in 1st 24 h chest tube output, nor in the amount of blood products transfused. However, haemoglobin levels on the 5th day after surgery were significantly higher in the MIMVS group (11 vs 10.2 mg/dl; P < 0.01). This difference persisted significant in the PS-matched cohort (10.9 vs 10.2 mg/dl; SMD = –0.47). Furthermore, patients on the MIMVS group had less incidence of de novo postoperative AF compared to the open group in the entire cohort (24% vs 37%; P = 0.02). This finding was also seen after PS matching, although not statistically significant (29% vs 39%; SMD = 0.21).

Compared to the Conventional group and despite longer CPB times, the MIMVS group required a significantly shorter median duration of mechanical ventilation (6 vs 14 h; P < 0.01), probably explained by a higher proportion of patients extubated in the operating room (60% vs 17%, P < 0.01). These differences persisted after matching. Median ICU and hospital stay were significantly shorter in the MIMVS group (all patients in the study were discharged home and not to other care facilities). The significant reduction in ICU stay was also seen in the PS-matched cohort (1.7 vs 2.7 days; SMD = 0.32). Five patients (4 MIMVS/1 Conventional) underwent reoperation before hospital discharge due to unsatisfactory mitral valve function at the postoperative echocardiographic control. All patients were successfully re-repaired using the same original approach. Details of these cases are described in Supplementary Material, Table S2.

Finally, there were 4 postoperative deaths during the study period, 2 in each group (1.1% vs 1.8%; P = 0.63). In the conventional group, 1 patient died because of an intraoperative atrioventricular groove rupture and another with severe preoperative tracheal stenosis was unable to wean from invasive respiratory support. In the MIMVS group, 1 patient died due to severe pulmonary distress requiring extracorporeal membrane oxygenation support and another patient died due to intercurrent postoperative severe acute respiratory syndrome Coronavirus 2 infection.

Echocardiographic results at hospital discharge

A total of 98.6% of repaired patients were discharged with ≤1+ MR, without significant differences between approaches (99% vs 98%; P = 0.61). Details are presented in Table 3.

Follow-up results

The median clinical follow-up time was 4.0 years (range: 0.1–8.9 years). Survival at 1, 3 and 5 years after the operation was 99%, 98% and 95%, respectively (Table 4 and Fig. 3). The predominant cause of death in the follow-up period was non-cardiovascular (67%). Follow-up index for patients in both groups [median (interquartile range)] were: MIMVS: 0.90 (0.82–0.94), Conventional: 0.92 (0.85–0.95).

Freedom from recurrent 4+ MR after 1, 3 and 5 years in the entire cohort was 98%, 94% and 93%, respectively (Table 4 and Fig. 4). Freedom from reoperation was 98%, 95% and 94%, whereas freedom from mitral replacement was 98%, 96% and 96%, respectively. None of these outcomes showed significant differences in the PS-matched cohort.

A total of 16 patients (5.3%) required reoperation during follow-up (median: 1 year; range 1 month to 7 years). Of these, 10 had undergone MIMVS (5.3%) and 6 (5.4%) a Conventional repair. There were no cases of mitral stenosis or infective endocarditis. The rate of successful mitral re-repair was 62.5% (80% after MIMVS and 33% after conventional surgery). Supplementary Material, Table S3 provides detailed information on all the reinterventions performed during follow-up.

DISCUSSION

Our study describes and compares the contemporary early and mid-term results of conventional and MIMVS repair for mitral prolapse over the last decade. Using statistical methods to adjust for baseline characteristics, we could demonstrate that MIMVS offers a safe and effective alternative to conventional open surgery, with comparable repair rates and even providing advantages on key clinical outcomes.

Noteworthy, our study included all consecutive patients with MR due to leaflet prolapse referred for surgery, regardless of the planned procedure and operation performed. This unselected population provides the most realistic measure of repair rate, which was very high in both groups, exceeding 98%. The minimally invasive approach did not compromise the possibility of using a wide range of surgical techniques and concomitant AF ablation and did not affect repair quality or durability.

Despite requiring longer operative times, MIMVS was not associated with a negative impact on postoperative outcomes, in accordance with other large series of mitral repair using a similar approach [9, 13–15]. On the contrary, MIMVS was associated with significantly shorter median duration of mechanical ventilation and ICU stay, in line with some previous reports [10, 16, 17]. Patients in the MIMVS group presented significantly higher haemoglobin levels, with similar transfusion requirements in both groups. Serum haemoglobin after cardiac surgery decreases due to blood loss, the need for fluid resuscitation in the early postoperative period and the inflammatory effects of the surgical manipulation and CPB. Haemoglobin after cardiac surgery tends to drift down, reaching a nadir around postoperative day 4 and recovering partially in most patients over the following days [18]. Haemoglobin level on postoperative day 5 provides typically a good indication of total blood loss and the maximal anaemic state after cardiac surgery.

Our goal was to leave the operating room without residual MR, as it has a negative impact on late outcomes [19]. All cases were performed under transesophageal echocardiography control and underwent a comprehensive transthoracic echocardiogram before discharge. In the presence of residual MR greater than the observed at the end of the operation, we were proactive taking back the patient to the operating room, particularly if the mechanism was technical. Noteworthy, the valves were successfully re-repaired in all cases, and all reinterventions were performed using the approach initially used, thus maintaining the benefits of MIMVS.

Finally, mid-term outcomes of our cohort showed no differences between groups in terms of survival, freedom from reoperation, valve replacement or recurrent 4+ MR, which is consistent with some previous reports [16, 20–22]. Compared to previous publications [20, 23, 24], our study shows a higher survival rate at 5 years of follow-up for both groups and a higher repair rate. It should be noted that, unlike these referenced series, our study is focused on 1 type of mitral dysfunction (mitral prolapse) instead of on the procedure performed (‘mitral intervention’ in both registries) and comes from a single centre where MIMVS is the most frequent approach (63% MIMVS cases vs 29% and 27% for the Netherlands and UK registries, respectively). This provides different and valuable information on the outcomes that can be achieved in the most frequent type of patients with MR in a dedicated programme, although these might not be directly extrapolated to other causes of MR and centres. In our study, there were no differences in MR recurrence when comparing the surgical approach used or the complexity of mitral lesions (isolated posterior vs anterior or bileaflet prolapse), reinforcing that MIMVS can be used with excellent results in all the disease spectrum, comparable to other large series using a conventional approach.

Limitations

The main limitation of this study is the limited sample size, its unicentre nature and the absence of randomization, which implies a risk of selection bias. Patients in the MIMVS group were operated only by 2 surgeons, although these same 2 surgeons were also performing the majority of the cases in the conventional group (61%). A second analysis excluding the patients operated by surgeons other than the 2 performing MIMVS showed same conclusions for all our defined outcomes. In our cohort, however, MIMVS was the most frequent approach used and we had been performing MIMVS long before the period of the study, so the learning curve of MIMVS is not included [24]. Nevertheless, to minimize the potential risk of selection bias, we performed a sensitivity analysis on a PS-matched cohort obtained from the complete population, thus balancing all preoperative variables. We had enough patients in all risk strata in both groups to effectively balance all variables. This method provides proven, excellent control of selection bias related with the variables included in the regression model but, unlike randomization, does not control for other factors that were not measured. Finally, we performed all MIMVS procedures using a standardized fully thoracoscopic approach. There are other different surgical approaches to repair the mitral valve that could be considered to be minimally invasive, and thus, the results reported might not be extrapolated to other techniques. Larger series and multicentre studies could provide valuable confirmatory data on these issues in the future.

CONCLUSIONS

MIMVS surgery for mitral prolapse is a safe and effective procedure and can be accomplished with excellent results by expert teams. Under this premise, it can be offered to most of these patients, regardless of the complexity of the lesions, with equivalent quality and advantages in key postoperative results as compared with conventional repair through a median sternotomy.

SUPPLEMENTARY MATERIAL

Supplementary material is available at EJCTS online.

Funding

No source of funding is reported.

Conflict of interest: none declared.

DATA AVAILABILITY

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

Author contributions

María Ascaso: Conceptualization; Data curation; Writing—original draft. Elena Sandoval: Data curation; Writing—review & editing. Anna Muro: Data curation; Writing—review & editing. Clemente Barriuso: Writing—review & editing. Eduard Quintana: Writing—review & editing. Jorge Alcocer: Writing—review & editing. Marta Sitges: Writing—review & editing. Bàrbara Vidal: Writing—review & editing. José-Luis Pomar: Writing—review & editing. Manuel Castellà: Writing—review & editing. Ana García-Álvarez: Supervision; Writing—review & editing. Daniel Pereda: Conceptualization; Supervision; Validation; Writing—original draft.

Reviewer information

Reviewer information European Journal of Cardio-Thoracic Surgery thanks Jules R Olsthoorn, Marco Moscarelli, Paul Modi and the other, anonymous reviewer(s) for their contribution to the peer review process of this article.

REFERENCES

ABBREVIATIONS

ABBREVIATIONS
 
• AF

  Atrial fibrillation

 
• CPB

  Cardiopulmonary bypass

 
• ICU

  Intensive care unit

 
• MIMVS

  Minimally invasive mitral valve surgery

 
• MR

  Mitral regurgitation

 
• NYHA

  New York Heart Association

 
• PS

  Propensity score

 
• SMD

  Standardized mean difference