Minimally Invasive Mitral Valve Repair with Artificial Chordae: Insights from a 6-Year Single-Center Study

Purpose . Minimally invasive mitral valve repair (MIMVR) has been demonstrated to be safe and efective, but technical difculty, outcome variation


Introduction
Mitral valve (MV) repair has better long-term results than MV replacement and has been established as the surgical treatment of choice for mitral regurgitation (MR) when the valve lesions are reparable and the procedure is performed at specialized centers with a high rate of successful repairs [1][2][3].Furthermore, minimally invasive mitral valve repair (MIMVR) has been demonstrated to be safe and efective for MV surgery, providing less morbidity, including stroke, atrial fbrillation, renal failure, and length of stay in carefully selected patients [4][5][6].Since its introduction, there has continued to be a need to keep developing simpler, more efective, and more durable surgical techniques for MV surgery.
Resectional techniques, including insertion of annuloplasty rings and resection of prolapsing segments, were established by Carpentier [7].Since then, excellent longterm results have been reported, but their reproducibility and learning curve have shown difculties [3,7,8].Zussa et al. [9] and Frater et al. [10] introduced the use of artifcial chordae and polytetrafuorethylene (PTFE) cords to replace ruptured or elongated native chordae tendineae.
Presently, the use of artifcial chordae has become a standard technique for MV repair, supported by excellent longterm results.Nevertheless, MIMVR using artifcial chordae carries its challenges, especially the proper length of artifcial chordae, adequate placement, technical learning curve, outcome variation, and lack of standardized protocols [1,8,11].Tis study aimed to analyze the midterm outcomes of MIMVR using artifcial chordae through right mini-thoracotomy in a single center.

Patient Selection.
A single-center, retrospective cohort study of all consecutive patients who underwent MIMVR using artifcial chordae (expanded polytetrafuoroethylene) was conducted via right minithoracotomy in Vietnam between April 2016 and April 2022.
Te operations are performed by a single primary surgeon specializing in cardiovascular surgery, supported by other assistants.Preliminary results and techniques were presented by our group in 2023 [12].Valve repairs were separated into 2 groups based on a previously validated complexity score (Anyanwu score) [13]: simple repair (Group 1) and intermediate-to-complex repair (Group 2).

Surgical Technique.
Te patients were placed in a supine position.To establish peripheral cardiopulmonary bypass, cannulation of the femoral vessels was done using an open Seldinger-guided technique.A 4-cm incision was made parallel to the anterior axillary line and dissected into the 4th right intercostal space.Subsequently, a video camera port was inserted into the 3rd space, followed by the insertion and clamping of the Chitwood cross-clamp.Custodiol HTK (histidine-tryptophan-ketoglutarate) solution was then administered.A left atriotomy was performed to expose the MV, and a left atrial retractor was employed.
Our MVMVR methodology closely follows the principles outlined by Gillinov et al. [14].Te process begins with the strategic placement of annuloplasty sutures to improve visibility and enables a thorough examination of the valve.Te repair technique then closely follows the guidelines established by Carpentier [15], which emphasize the stabilization of the annulus, maintenance of leafet mobility, and the creation of an ample coaptation surface.Consequently, in all cases, chordal replacement and annuloplasty constituted the exclusive treatment modalities employed.To install the neochords, frst, the prolapsing segment was identifed and the appropriate papillary muscle was selected.Te A1-P1 region or the nonprolapsed segment of the native chords was used as a reference point.Two techniques were utilized: Te free-hand chordae technique (Figure 1) involves passing a PTFE CV-5 suture (W.L. Gore & Associates, Inc., Newark, DE, USA) through the fbrous region of the papillary muscle, bringing each end through the prolapse area.Te PTFE length was adjusted based on the reference point before tying the suture.Te premeasured chordal loops technique (Figure 2) was implemented using Chord-X loop (On-X Life Technologies, Inc, Austin, TX, USA).It utilizes a caliper to measure the length of a reference point, choosing the loop length and then anchoring it to the papillary muscle and attaching it to the edge of the leafet-free margin in the prolapse area.During ring annuloplasty, the size of the ring is determined very carefully based on the surface area of the anterior leafet.It is then aligned with pre-positioned annuloplasty sutures.In all cases, the Carpentier Physio II ring (Edwards Lifesciences Corp., Irvine, CA, USA) was used.It is important to note that neochords were only secured after the ring annuloplasty was implemented.Transesophageal echocardiography was utilized to verify the fnal repair result.

Statistical Analysis.
Demographic variables, comorbidities, operative characteristics, postoperative outcomes (intubation time, ICU time, mechanical circulatory support, peripheral vessel complications, stroke, reoperation), postoperative echocardiographic features, and follow-up outcomes (mortality, freedom from recurrent MR) were assessed and analyzed.Data are presented as mean ± standard deviation for normally distributed continuous variables and as medians with interquartile ranges for non-normally distributed continuous variables, while categorical variables are presented as frequencies and percentages.
A learning curve was analyzed by plotting trend lines on a two-dimensional scatter plot that compared the number of procedures with operation time and aorta cross-clamp time.Primary endpoints included survival and freedom of recurrent MR at 48 months (4 years), secondary outcomes included echocardiogram fndings.Te research was approved by the Ethical Board of the University of Medicine and Pharmacy at Ho Chi Minh City, number 166/HDDD-DHYD, on February 26th, 2021.

Results
Ninety patients were identifed, including 41 cases of simple repairs (Group 1) and 49 cases of intermediate-to-complex repairs (Group 2; 45 intermediate repairs and 4 complex repairs) [13].Te mean age including both groups was 50.5 ± 12.9 years and 70 (77.7%)were men.Table 1 summarizes the clinical characteristics stratifed by complexity score, and there were no signifcant diferences between the groups.Preliminary results were presented by our group in 2023 [12].
All MR were categorized as Carpentier regurgitation type II, with fbroelastic defciency being the primary etiology observed in both groups.Te mean annular mitral diameter measured 39.2 ± 4.4 mm.A slight diference was observed between Group 1 and Group 2; Group 1 had a smaller MV annulus size compared to Group 2 (37.9 ± 4.0 mm vs. 40.2± 4.5 mm, P � 0.012); the preoperative echocardiography characteristics and perioperative features are presented in Tables 2 and 3.
Tere were no diferences in the left ventricular measurements between groups before and after repair, as shown in Table 4. Te median follow-up time was 30.3 months (18.2-40.4),and there were two (2.2%) deaths during this time or one in each group.No reoperations related to the MV were observed during the follow-up period.Te midterm outcomes have been demonstrated in Table 5.
Tere was no diference between overall midterm survival or recurrent MR.Groups 1 and 2 Kaplan-Meier estimates of survival at 12 and 48 months were 97% vs. 100%, and 97% vs. 96% (95% CI; 0.05-12.2,P � 0.850), respectively (Figure 3).Te freedom from recurrent MR estimates were 97% vs. 92% and 97% vs. 88% (95% CI; 0.49-12.0,P � 0.260), respectively (Figure 4).In both groups, there were no reoperations.Te analyzed learning curve revealed a negative correlation between the number of operations and the duration, as indicated by the trend lines.Specifcally, as the volume of operations increased, the duration of both operation time and aortic cross-clamp time decreased, resulting in shorter durations (Figures 5 and 6).
A comparison of left ventricular end-diastolic diameter (LVEDD) and left ventricular ejection fraction (LVEF) between the two groups showed that LVEDD signifcantly decreased preoperatively and remained below 50 mm during the 4-year follow-up, with Group 2 consistently having higher levels than Group 1. LVEF also signifcantly decreased preoperatively and postoperatively, remaining relatively stable during the follow-up period, with no signifcant differences between the groups.Tese fndings indicate similar outcomes for both groups regarding LVEDD and LVEF, suggesting that the surgical intervention was equally efective (Figures 7 and 8).

Discussion
Mitral valve repair remains the treatment of choice for MR if the expertise is available, and minimally invasive approaches have been explored over the last two decades due to their potential patient-centered advantages.However, they are also related to technical challenges and require adequate training.Outcomes are associated with the patient's conditions, surgeon's experience, and center volume.Tis study aimed to evaluate and describe the initial experience of a valve surgery program at a single center in a lower middle-income country.Te postoperative mortality and stroke rates were 0.0% and 1.1%, respectively.At two years, the mortality was 2.2% due to cardiac deaths and severe MR rates were 8.9%.
Etiology is a crucial consideration when it comes to MV disease and MV repair, especially in Vietnam.In a recent study, Cazaubiel and Iung [16] analyzed 2734 patients who had MV surgery in Vietnam, fnding that surgeries that occurred from 1995-2000 were more likely to be classifed as rheumatic valvular disease, in contrast to the 15.4% increase in degenerative etiology found in the 2005-2010 era of surgeries.Moreover, the change in etiology could not necessarily be pinpointed in these studies, as there was a signifcant increase in the degenerative etiologies, specifcally within the fbroelastic disease-causing issue.Vietnam, despite being considered resource-limited in healthcare due to fnancial burdens, unequal resource distribution, and insurance coverage challenges [17], demonstrates a signifcant uptake of minimally invasive approaches in surgical procedures, with around half of the mitral valve surgeries in the country employing these techniques [18].Tis adoption can be attributed to factors such as a patient-centered culture, increased patient awareness, and early recognition and acceptance of these procedures during the initial phases of the disease [18].Cultural infuences signifcantly impact the acceptance and implementation of new surgical techniques, and Vietnamese patients exhibit a greater receptiveness towards minimally invasive approaches, resulting in their widespread utilization [17,18].Te rise in degenerative conditions and the strong adoption of minimally invasive approaches have generated a greater demand for minimally invasive mitral valve repair (MIMVR) despite the resource limitations in the country.
To ensure a successful mitral regurgitation (MR) repair, grading methods have been developed to assess the complexity of the repair and to choose appropriate techniques, with expert consultation if needed.One such practical and comprehensive system is the Anyanwu system, which considers echocardiography features, valve tissue calcifcation, location of leafet prolapse, and limited leafet mobility [13].In this study, we analyzed patients who required MV repair and had a valve degeneracy rating of simple, intermediate, or complex based on Anyanwu's classifcation of various MV degeneracies, which has also been used by Nakayama [19].Tese results demonstrate the importance of considering etiology and comprehensive grading systems to ensure successful MR repair in degenerative MV disease cases.

4
Journal of Cardiac Surgery Te current study demonstrated the feasibility, safety, and efectiveness of MIMVR using an artifcial chordae technique.Te postoperative mortality and stroke rates were similar to recent data presented by the Mini-Mitral International Registry (mortality 1.7% and stroke 1.7%) [20].All cases used repair techniques, including ring annuloplasty and artifcial chordae, with the latter being the primary technique.Valve replacement was not required in any instance.According to Rankin's fndings, the artifcial chordae repair method is an efective solution for managing mitral disease, including complex cases [21].Tis technique addresses the fundamental issue of chordae rupture or elongation that is common in mitral valve surgeries.One of the advantages of artifcial chordae is their technical simplicity in minimally invasive procedures, while also preserving the valvular tissue that aligns more anatomically compared to traditional leafet resection methods [21,22].However, challenges such as the potential for hemolytic anemia, and longevity concerns arising from PTFE hyalinization or calcifcation, are noted [8].In addition, post-repair remodeling of the left ventricle could lead to mismatched chordae lengths, potentially causing a recurrence of MR [23].Alongside the artifcial chordae technique (addresses MV leafets), annuloplasty ring usage extends repair durability by restoring the mitral annulus's functional support.Tis, in turn, preserves leafet coaptation and prevents future annular dilatation in degenerative MV cases.Te integration of ring annuloplasty in the minimally invasive approach has proven to be seamless, particularly as it complements the fundamental step of exposing the mitral valve for efective repair.
While artifcial chordae can easily fx simple prolapse, repairing anterior leafet or bi-leafet prolapse poses a more signifcant challenge.Per consensus among surgeons, artifcial chordae is recommended for anterior prolapse repair [5].For bi-leafet prolapse, a two-step repair approach was followed, starting with posterior leafet repair, followed by annuloplasty ring insertion and saline testing to assess the extent of regurgitation.Further repair techniques, such as artifcial chordae, leafet resection, and edge-to-edge stitch, were applied as required, as proposed by Castillo et al. [3].In cases of substantial and even bi-leafet prolapse, annuloplasty ring repair may be appropriate.In contrast, artifcial chordae implantation may be suitable for asymmetrical prolapse, according to Koprivanac [5].Te edge-to-edge technique was primarily utilized in Group 2, as opposed to Group 1 (10.2% vs. 2.4%) due to the higher complexity scores from Anyanwu's system seen in Group 2, indicating the need for multiple repair techniques for successful MV repair.Te technique was employed as a backup solution in cases where repair proved inadequate [24].Its straightforwardness and efciency, particularly in challenging circumstances such as those observed in Group 2,

6
Journal of Cardiac Surgery have made it a preferred choice.In our practice, we employ the edge-to-edge technique to prevent systolic anterior motion (SAM), especially in mucoid degeneration prolapse with signifcant posterior leafet height.We also often create shorter posterior neochordae to limit leafet tissue movement toward the left ventricular outfow tract during systole, a crucial step when using smaller annuloplasty rings to prevent postoperative mitral stenosis [25].Likewise, the commissuroplasty technique was more prominent in Group 2 than in Group 1 (26.5% vs. 9.8%).For simpler MV lesions in Group 1, typically a single MV segment prolapse, one or two pairs of artifcial chordae were often sufcient without requiring supplementary repair techniques.Te present study evaluated the ICU length of stay and survival rates of patients with both simple and complex cases of MV disease, revealing favorable outcomes in terms of the duration of ICU stay.However, postoperative complications, such as pleural efusion were noted in 3 cases (3.3%) within both groups, which required re-exploration via the same thoracic incision.Out of these cases, 2 had an unclear bleeding source, while 1 case experienced bleeding at the thymus site.A case of postoperative low cardiac output was identifed in Group 2, which was caused by low blood pressure and a decline in left ventricular function.An intra-aortic balloon pump was inserted and successfully removed, and no in-hospital deaths were reported in our study.Tere were two patients with femoral access complications; one case of postoperative femoral artery stenosis requiring thrombectomy and primary repair, and the second case had a postoperative lower limb compartment syndrome associated with a prolonged cardiopulmonary bypass duration requiring fasciotomies.Te compartment  syndrome presented 10 hours after the surgery and the main symptoms were severe pain, coldness, tenderness, paresthesia, and weak peripheral pulses [26].
In terms of postoperative complications, a higher incidence of moderate-to-severe MV re-regurgitation was observed in patients classifed as intermediate-to-severe MV repair cases (Group 2) in comparison to the patients classifed as simple cases (Group 1) with a rate of 18.4% vs. 2.4%, respectively.Tis outcome underscores the need to evaluate the complexity of MR prior to surgery as a means to optimize the outcome of valve repair.
Perioperative SAM was detected via transesophageal echocardiography in 7.8% of all cases and was managed with a step-wise approach consisting of medical therapy, followed by ventricular loading and surgery as the fnal option.Inotrope was stopped if patients' hemodynamics permitted or used betablockers to reduce heart rate, IV fuids, stopped diuretics, and transoesophageal echocardiogram monitoring.According to Alferi and Lapenna [27], SAM resolves spontaneously in about 1/3 of cases with cessation of inotropes and increased fuid and improves in 80% of patients reducing heart rate and increasing preload.Surgical repair is rarely needed and encounters anatomical challenges.All cases of perioperative SAM were resolved after medical therapy optimization in our study.Tis "transient" phenomenon was not detected during follow-ups or on echocardiography at discharge, which is consistent with previous studies [25,27].Te only statistically signifcant perioperative complication was the occurrence of transient perioperative SAM, which was more common in Group 1 (14.6%)than in Group 2 (2.0%).Te reasons for this discrepancy remain unclear, hinting at a possible underestimation of SAM risk in Group 1 due to the lack of additional repair techniques to prevent SAM.Te underlying reasons for this discrepancy remain unclear, suggesting a possible underestimation of SAM risk in Group 1, where additional repair techniques were not applied to prevent SAM.Our results diverged from Nakayama's study [19], which demonstrated opposite postoperative characteristics.Specifcally, Nakayama's study reported increased MR in patients classifed as an intermediate-complex.However, our study shared a similar learning curve and lower mortality rates than those reported in the Society of Toracic Surgeons database [28].
Te postoperative transmitral mean gradient was 2.9 ± 1.6 mmHg.Patients who underwent MV repair for degenerative MR with a mean gradient greater than 3 mmHg had reduced exercise capacity [29].Te use of certain valve repair techniques, such as artifcial chordae instead of resection techniques, and the selection of smaller annuloplasty mitral rings, rigid full mitral rings, and excessive utilization of the edge-to-edge technique, were associated with elevated postoperative transmitral gradient and an increased risk of functional mitral stenosis, leading to the long-term hemodynamic deterioration [1,29].
During the follow-up period, no signifcant diference in mortality was observed between the two groups, as one cardiac death was recorded in each group.One patient was readmitted after 4 months post-surgery for acute heart failure, and the second patient presented infective endocarditis around the second month postoperative.Other studies have also reported similar fndings, and in addition, our study exhibited lower mortality rates when compared to the Society of Toracic Surgeons database [19,28].
Te Kaplan-Meier survival estimates at 12 and 48 months for Group 1 and Group 2 were 97% versus 100% and 97% versus 96%, respectively.Pfannmueller et al. reported 2134 cases with 1-, 5-, and 10-years survival rates (artifcial chordae group) of 98%, 95%, and 86%, respectively.Te resection group showed survival rates of 97%, 92%, and 81%, respectively [22].Additionally, Davierwala et al. reported 2829 cases predominantly utilized [30].Figure 3 presents the Kaplan-Meier estimates for Groups 1 and 2 regarding freedom from recurrent MR at 12 and 48 months, showing rates of 97% versus 92% and 97% versus 88%, respectively.Notably, no reoperations specifcally related to the MV were observed during the follow-up period.Tis highlights that the rate of reoperation on the MV does not necessarily provide an accurate refection of the valve's functional status in patients who have not undergone reoperation.Te observed diference in the cumulative incidence of recurrent MR and the need for reoperation suggests that factors such as asymptomatic patients, patient's refusal for reoperation, or individuals with advanced age and other comorbidities may contribute to this disparity.It is crucial to recognize that the rate of reoperation alone does not indicate a signifcant occurrence of recurrent MR in these cases [31].Our study's learning curve analysis revealed that before reaching 50 cases, the durations of aortic cross-clamping and operating times displayed instability.However, after surpassing this cumulative case threshold, both durations demonstrated a stable decrease, indicating that high-volume surgeons tend to achieve improved outcomes.Tese fndings align with previous research by Li et al. [32], where cumulative sum analysis was used to assess the average incidence of adverse events and operative time, revealing that surgeons typically require 50 to 200 operations to overcome the learning curve associated with mitral valve repair rates.Tese results provide valuable insights into the procedural efciency of mitral valve repair using artifcial chordae and highlight the signifcance of surgical experience in achieving favorable patient outcomes.
Tere are limitations to the current study.First, using a retrospective design introduces potential biases and limitations in data collection.Moreover, the absence of a control group hampers our ability to draw defnitive conclusions regarding the efectiveness of the interventions.Additionally, our hospital's adoption of the minimally invasive approach since July 2014 prevents comparison with the total sternotomy approach.Te relatively small sample size and single-center setting raise concerns about the generalizability of the fndings to a broader population.Despite the study's retrospective data over a period of 6 years, the limited number of patients and only nine patients at risk during the 48-month follow-up examination undermine the study's statistical power.Furthermore, the learning curve analysis was conducted by a single surgeon, potentially introducing biases.To address these limitations, future research should include larger prospective studies involving multiple centers and surgeons to enhance the reliability of the evidence.

Conclusion
Based on our initial experience, performing minimally invasive mitral valve repair using artifcial chordae through a right minithoracotomy is a viable, safe, and efective option for patients with MR in resource-limited countries.Tis technique can be used for a range of MR complexities, from simple to intermediate-to-complex MV repairs, and has shown a promising midterm success rate in terms of freedom from MR recurrence.Further studies are needed to evaluate long-term outcomes.

Figure 7 :
Figure 7: Left ventricular end-diastolic diameter comparison.LVEDD signifcantly decreased from 60 to 50 mm preoperatively and postoperatively.It remained below 50 mm during the 4-year follow-up, with the intermediate to complex group consistently showing higher levels than the simple group.

Figure 8 :
Figure 8: Left ventricular ejection fraction comparison.LVEF showed a signifcant decrease from 65 to below 60% both preoperatively and postoperatively.It remained relatively stable during the follow-up period for the majority of the time.No signifcant diferences were observed between the groups.

Table 1 :
Preoperative characteristics.Data are presented as mean ± standard deviation or n (%).BMI, body mass index; BSA, body surface area; COPD, chronic obstructive pulmonary disease; NYHA, New York Heart Association.
TAPSE, tricuspid annular plane systolic excursion.Data are presented with associated p values, with bold values indicating statistical signifcance at p < 0.05.
Data are presented as median (interquartile range) or n (%).