Computer simulations of transapical mitral valve repair with neochordae implantation: Clinical implications.

Objectives: Transapical beating heart neochordae implantation is an innovative mitral valve repair technique that has demonstrated promising clinical results in patients with primary mitral regurgitation. However, as clinical experience continues to increase, neochordae implantation criteria have not been fully standardized. The aim of this study was to investigate the biomechanical effects of selecting an antero-lateral apical access site compared with a postero-lateral site, and suboptimal neochordae length compared with optimal suture length, on restoring physiologic left heart dynamics.
Methods: Transapical neochordae implantation using 3 and 4 sutures was computer simulated under 3 posterior mitral leaflet prolapse conditions: isolated P2, multiscallop P2/P3 and multiscallop P2/P1. Physiologic, pre- and postrepair left heart dynamics were evaluated using a fluid-structure interaction modeling framework.
Results: Despite the absence of residual mitral regurgitation in all postrepair models with optimal neochordae length, selecting an antero-lateral apical entry site for the treatment of P2/P3 prolapse generated a significant increase (>80%) in neochordae tension and P2 peak stress, with respect to a postero-lateral entry site. During isolated P2 prolapse repair, although neochordae overtension by 5% led to minimal hemodynamic changes in the regurgitant volume compared with using an optimal suture length, a significant increase in systolic and diastolic neochordae tension (>300%) and posterior leaflet average stress (70%-460%) was quantified. On the other hand, neochordae undertension by 5% led to worsening of regurgitation severity. Conclusions: This parametric computer study represents a further step toward an improved understanding of the biomechanical outcomes of transapical neochordae technologies.

Physiologic, pre- and postrepair simulations of transapical neochordae repair.

First-in-human experience with FSI computer simulations of transapical neochordae repair under various mitral valve prolapse conditions, left ventricular apical entry sites, and neochordae number and length.

A deeper dive into which primary MR patients benefit from transapical neochorde implantation (and which do not) is necessary and requires a more integrated and individualized biomechanical approach. We report original quantitative data regarding left heart hemodynamics, mitral apparatus deformation, and suture mechanics throughout the cardiac cycle from physiologic, pre- to postrepair states.

See Commentaries on pages 45 and 46.

Significant mitral regurgitation (MR) is present in 1.7% of the general population, increasing to 9.3% in patients older than age 75 years. The most frequent cause of MR in the western world is primary or degenerative mitral valve (MV) disease, in which there is an abnormality of 1 or more components of the mitral apparatus that often leads to elongation or rupture of the native chordae tendineae. After the success of transcatheter aortic valve (AV) replacement in treating aortic stenosis, great effort has been made in recent years toward the development of minimally invasive technologies for MV repair. In particular, transapical beating heart implantation of expanded polytetrafluoroethylene (ePTFE) sutures (ie, neochordae) has demonstrated potential effectiveness and safety in treating a subset of patients with primary MR.

Transapical neochordae implantation is an emerging technology with several medical devices being actively developed and limited experience in humans. Some aspects of this technique, such as determining the optimal left ventricular (LV) apical access site, ideal neochordae number, length, and leaflet placement, have not been fully standardized and can only be assumed at the moment., A recent feasibility study that evaluated safety and performance of transapical neochordae implantation reported 10% technical failure at 30 days, with patients requiring conversion to open MV surgery. At 6-month follow-up, 15% of the remaining patients had suboptimal outcomes with moderate to severe MR. Moreover, it is important to highlight that the growing experience with this technique has been mainly in patients with isolated central prolapse. One-year freedom from composite end points has been reported as high as 94% in simple lesions, but lower in more complex cases. Clearly, limited knowledge about clinical outcomes in patients with multiscallop or para-commissural prolapse exists. Overall, there is a need to refine the procedure and better understand the biomechanical implications of transapical neochordae placement.

Computer simulation studies have been performed in the past with the goal to investigate MV function under pathologic and repaired states,9, 10, 11, 12, 13, 14, 15, 16 including traditional neochordae MV repair, where the ePTFE sutures are surgically anchored to the papillary muscle (PM) tips.17, 18, 19, 20, 21 Yet, only a few numerical studies have evaluated the novel transapical neochordae implantation procedure.22, 23, 24 These studies have failed to accurately simulate full valve dynamics and directly quantify the regurgitant flow across the MV. Recently, Caballero and colleagues examined the effects of neochordae number (3 vs 4), and complexity of posterior mitral leaflet (PML) prolapse (isolated vs multiscallop) on postprocedure left heart fluid and tissue mechanics when using a postero-lateral LV entry site. The present study aims to extend the previous biomechanical analysis by investigating the effects of selecting an antero-lateral LV access site, the second-most common apical site for this minimally invasive procedure, and suboptimal neochordae length on restoring physiologic left heart dynamics throughout the cardiac cycle. An improved understanding of these aspects through a detailed and objective engineering analysis is an important step toward the success of novel transcatheter MV interventions in the short- and long-term.

Methods

Modeling of Physiologic and Prerepair Left Heart Dynamics

In a number of recently published studies, we used a validated fluid-structure interaction (FSI) computer framework to investigate cardiac hemodynamics and valve mechanics in the sequential transition from a physiologic (ie, healthy) state, to severe acute MR due to chordae rupture, to MV repair using postero-lateral transapical neochordae implantation. Briefly, a subject-specific healthy left heart model was developed and validated using full phase cardiac multislice computed tomography images and echocardiography data from a 72-year-old woman at Hartford Hospital (Hartford, Conn)., End-diastolic volume, end-systolic volume, stroke volume, and LV ejection fraction were 113 mL, 47 mL, 66 mL, and 58%, respectively. The use of de-identified clinical data for this study was approved by the institutional review board. As seen in Figure 1, A, the subject-specific left heart model comprises all major cardiac structures, including the aortic root, AV, MV, and LV and left atrial endocardial walls.

Figure 1

A, Representative isolated P2 prolapse left heart model showing antero-lateral apical access site with NC3. B, Neochordae number and leaflet attachment location for the 6 postrepair models. C, Epicardial and endocardial wall motion between peak diastole (green) and peak systole (yellow). AV, Aortic valve; MV, mitral valve; LV, left ventricle; LA, left atrium; APM, antero-lateral papillary muscle; PPM, postero-medial papillary muscle; AML, anterior mitral leaflet; PML, posterior mitral leaflet divided into P1, P2, and P3 scallops; NC3, 3 neochordae; NC4, 4 neochordae.

The MV model was defined by the anterior mitral leaflet (AML) and the division of the PML in 3 scallops (ie, P1, P2, and P3). Native chordae were classified into 5 groups according to the leaflet insertion zone: AML strut, AML marginal, PML marginal, PML intermediate, and PML basal. Following the approach by Caballero and colleagues, 3 PML prolapse models that reproduce common clinical scenarios of chordae rupture were selected and analyzed in this study: isolated P2, multiscallop P2/P3, and multiscallop P2/P1. Further details on left heart model reconstruction, FSI simulation setup, and validation studies can be found in Appendix 1.

Antero-Lateral Transapical Neochordae Modeling

In this study, transapical neochordae repair was simulated following the manufacturer's recommendations., Clinically, 3 or 4 neochordae are implanted on average to avoid any unbalanced configuration of the leaflets that can lead to excessive and localized mechanical stresses. Thus, 3 (NC3) and 4 (NC4) neochordae were simulated in this study for each PML prolapse model, as seen in Figure 1, B. Neochordae were uniformly distributed along the prolapsed scallop(s) and fixed 4 mm away from the leaflet free edge, as clinically recommended.

Neochordae were modeled as CV-4 ePTFE sutures (Gore-Tex Suture, W.L. Gore & Associates Inc, Flagstaff, Ariz) with an initial uniform cross-sectional area of 0.074 mm2. Mechanical properties were obtained from literature., As shown in Figure 1, A, an antero-lateral LV entry site approximately 3 cm from the LV apex was selected for the neochordae origin. In addition to modeling ideal neochordae length, 2 suboptimal neochordae length configurations were simulated for the isolated P2 prolapse model when using NC4. Optimal suture length was altered by ± 4 mm (5% of total neochordae length), simulating apparently correct transapical neochordae techniques with a slight undertension and overtension. More details on neochordae modeling can be found in Appendix 1.

Data Analysis

A total of 12 FSI simulations were performed in this parametric biomechanical study. One for the physiologic healthy left heart model, 3 for the prerepair left heart models with acute MR, 6 for the postrepair left heart models with optimal neochordae length by using NC3 and NC4, and 2 for the isolated P2 NC4 postrepair left heart models with suboptimal neochordae length (optimal length ± 4 mm).

Physiologic, pre- and postrepair left heart dynamics were analyzed in terms of the following hemodynamic and structural parameters:

Stroke volume (SV) in the AV (SVAV) and MV (SVMV), regurgitant volume (RV) in the AV (RV) and MV (RV), regurgitant fraction in the MV, RF= RV/LVSV, where LVSV is the total SV of the LV (SV + RV), and MR severity;

Native chordae tension, defined as the resultant force carried by the different chordae groups. The resultant force experienced by a particular chordae group was calculated as the sum of forces exerted by each native chorda attached to that chordae group;

PM force, defined as the resultant reaction force exerted on the antero-lateral PM and postero-medial PM to bear the tension of the native chordae;

Neochordae tension, defined as the resultant force exerted on the ePTFE sutures;

Percentage of the force carried by each neochorda relative to the total neochordae tension;

Peak value of the maximum principal stress on the neochordae;

Average maximum principal stress in the MV leaflets. Leaflet free edge and annular regions were not included in the average stress calculation); and

Peak value of the maximum principal stress on the mitral leaflets with location.

Additionally, we compared the postrepair left heart dynamics obtained in this work with the results obtained when using a postero-lateral LV entry site. In the text, AL-NC refers to the antero-lateral neochordae configuration implemented in this study, whereas PL-NC refers to the postero-lateral neochordae configuration used in the study by Caballero and colleagues.

Results

Left Heart Hemodynamics

Cardiac hemodynamics for the physiologic, pre- and postrepair left heart models with optimal neochordae length are shown in Table 1. Acute MR correction was achieved in all postrepair models, with MV function being restored to physiologic levels. By using the RF as a quantitative parameter to grade the MR severity, all postrepair models can be classified as having trivial MR. The MV flow rate curves throughout the cardiac cycle for all LH models are shown in Figure 2.

Table 1

Hemodynamic parameters for the physiologic, pre- and postrepair (optimal neochordae length) left heart models

Variable	Physiologic	Isolated P2	Isolated P2 NC3	Isolated P2 NC4	P2/P3	P2/P3 NC3	P2/P3 NC4	P2/P1	P2/P1 NC3	P2/P1 NC4
SVAV (mL)	58.22	41.94	58.98	57.26	20.46	57.78	58.14	16.16	59.39	59.11
RVAV (mL)	4.27	4.61	4.53	4.62	4.62	4.29	4.36	4.72	4.23	4.56
SVMV (mL)	62.54	61.36	63.55	62.99	63.60	62.87	62.88	63.42	64.02	63.59
RVMV (mL)	9.27	24.32	9.81	11.02	47.19	10.13	9.85	51.53	9.58	9.74
RFMV (%)	13.74	36.70	14.27	16.14	69.76	14.92	14.49	76.13	13.90	14.15
MR severity (RFMV)	Trivial	Moderate	Trivial	Trivial	Severe	Trivial	Trivial	Severe	Trivial	Trivial

SV, Stroke volume in the AV; RV, regurgitant volume in the AV; SV, stroke volume in MV; RV regurgitant volume in the MV; RF, regurgitant fraction in the MV.

Figure 2

Flow rate (mL/sec) curves across the mitral valve (MV) throughout the cardiac cycle. The negative systolic mitral flow indicates backflow of blood into the left atrium due to valve closing and regurgitation. APM, Antero-lateral papillary muscle; PPM, postero-medial papillary muscle; NC3, 3 neochordae, NC4, 4 neochordae.

Dynamic Neochordae and PM Tension During the Cardiac Cycle

Figure 3 shows the dynamic PM tension curves for the physiologic model, and the neochordae tension curves for the postrepair models with optimal neochordae length throughout the cardiac cycle. During peak systole, the isolated P2, P2/P3, and P2/P1 AL-NC postrepair models had a total suture tension between 2 and 2.7 N, 5.5 and 6 N, and 3 and 3.3 N, respectively (Table 2). Moreover, the highest and lowest individual neochorda tension values between all models were 2.05 and 0.39 N, respectively. When comparing the use of NC3 and NC4, implanting NC4 for the repair of isolated P2 prolapse showed a 30% increase in neochordae tension. For the multiscallop postrepair models, negligible differences in total suture tension were reported. Native mitral chordae tension results are presented in Appendix 1.

Figure 3

Papillary muscle (PM) and neochordae tension (N) curves for the antero-lateral (AL-NC) and postero-lateral (PL-NC) neochordae implantation configurations throughout the cardiac cycle. NC3, 3 Neochordae, NC4, 4 neochordae.

Table 2

Structural parameters for the physiologic, pre- and postrepair (optimal neochordae length) left heart models at peak systole

Variable	Physiologic	Isolated P2	Isolated P2 NC3	Isolated P2 NC4	P2/P3	P2/P3 NC3	P2/P3 NC4	P2/P1	P2/P1 NC3	P2/P1 NC4
Fchordae (newton)
AML marginal	1.61	1.71 (6.3)	2.17 (34.8)	1.71 (5.7)	1.95 (20.6)	2.30 (42.7)	2.28 (41.4)	1.99 (23.6)	2.38 (47.7)	2.53 (56.9)
AML strut	3.49	3.15 (–9.9)	3.63 (4)	3.55 (1.6)	2.68 (–23.3)	3.27 (–6.4)	3.31 (–5.3)	2.56 (–26.7)	3.73 (6.7)	3.78 (8.1)
PML marginal	0.70	0.52 (–25.9)	0.33 (–53.4)	0.38 (–44.9)	1.03 (47.1)	0.27 (–61.7)	0.30 (–56.7)	0.28 (–60.5)	0.45 (–36)	0.39 (–43.9)
PML intermediate	1.11	0.43 (–60.8)	0.54 (–51.5)	0.52 (–53.4)	0.58 (–47.3)	0.65 (–40.8)	0.63 (–43)	1.36 (23.4)	0.09 (–91.6)	0.10 (–90.9)
PML basal	4.97	4.35 (–12.3)	3.51 (–29.3)	3.10 (–37.6)	1.71 (–65.6)	1.64 (–67)	1.65 (–66.8)	1.74 (–65.1)	2.22 (–55.4)	2.29 (–54)
FPM (newton)
APM	6.15	5.39 (–12.3)	5.18 (–15.8)	4.84 (–21.3)	5.70 (–7.3)	5.47 (–11)	5.48 (–10.9)	2.45 (–60.1)	3.28 (–46.7)	3.36 (–45.4)
PPM	5.73	4.78 (–16.6)	5.00 (–12.7)	4.42 (–22.9)	2.25 (–60.8)	2.67 (–53.5)	2.69 (–53)	5.48 (–4.4)	5.59 (–2.5)	5.73 (0)
FNC (newton)			2.06	2.68		5.50	5.95		2.97	3.29
Fneochorda (%)			35.3, 34.9, 29.8	29.8, 14.5, 33.3, 22.5		32, 37.5, 30.5	30, 30, 18.5, 21.5		26.2, 42.2, 31.6	18.1, 27.5, 27.8, 26.6
SɪMAXNC (MPa)			9.8	12.07		27.87	24.13		16.91	12.35
SɪAVRGMV (kPa)
AML	135.0	107 (–20.7)	139.1 (3)	132.7 (–1.7)	105.5 (–21.9)	136.6 (1.2)	141.6 (4.9)	96 (–28.9)	140.4 (4)	145.1 (7.4)
P1	59.2	51.5 (–13)	63 (6.4)	63.3 (7)	44.3 (–25.2)	62.9 (6.2)	63.2 (6.8)	100.4 (69.6)	132.4 (123.8)	120.4 (103.5)
P2	118.8	61.3 (–48.4)	179.5 (51.1)	160.1 (34.8)	128.7 (8.4)	217.1 (82.7)	220.9 (85.9)	182.6 (53.7)	209.1 (76)	204.4 (72.1)
P3	52.6	48.5 (–7.8)	56.5 (7.3)	55.5 (5.6)	119 (126.1)	163.2 (210.1)	153.3 (191.4)	43.6 (–17.1)	59 (12.1)	57.9 (9.9)
SɪMAXMV (MPa)	0.43-AML	0.52-AML	0.66-P2	0.66-P2	2.80-P2	1.85-P2	1.97-P2	2.25-P2	1-P2	1.41-P2

Percentage variations with respect to physiologic left heart model are reported in parenthesis. Fneochorda percentage values from left to right correspond to neochorda located from P1 to P3 scallops. F, Native chordae tension; AML, anterior mitral leaflet; PML, posterior mitral leaflet; F, PM force; APM, antero-lateral papillary muscle; PPM, postero-medial papillary muscle; F, neochordae tension; F, neochorda percentage distribution of total suture tension; S, peak value of the maximum principal stress on the neochordae; S, average maximum principal stress in the MV leaflets; S, peak value of the maximum principal stress on the MV leaflets.

Mitral Leaflet Stress

The average leaflet stress at peak systole calculated for each of the four mitral segments (Figure 1, A) is presented in Figure 4 and Table 2. After transapical neochordae implantation, AML stress returned to physiologic values in all postrepair models. For the isolated P2 postrepair models, the nonprolapsing PML scallops (ie, P1 and P3) maintained physiologic values, whereas the average stress in the repaired P2 scallop increased >50% and >30% for the NC3 and NC4 models, respectively, when compared with the physiologic model.

Figure 4

Average stress (kPa) in the mitral leaflets at peak systole. Circles highlight a marked increase (>50%) in leaflet stress with respect to the physiologic left heart model (blue). NC3, 3 Neochordae; NC4, 4 neochordae; AML, anterior mitral leaflet.

Similarly, for the P2/P3 and P2/P1 postrepair models, the nonprolapsing scallops (ie, P1 and P3, respectively) maintained physiologic values, whereas the average stress in the P2 scallop increased ∼80%. There was also an important increase in the average stress of ∼200% and >100% in the lateral prolapsing scallops (ie, P3 and P1, respectively). No significant differences in leaflet stress were observed when implanting NC3 or NC4. Figure 5 shows the stress distribution across the mitral leaflets at peak systole. More information can be found in Appendix 1.

Figure 5

Stress (MPa) distribution in the mitral leaflets at peak systole. A stress value threshold of 0.5 MPa was applied such that relatively large stress values were displayed in grey, facilitating comparison between models. Native chordae and neochordae not shown for clarity. PML, Posterior mitral leaflet; AML, anterior mitral leaflet; NC3, 3 neochordae; NC4, 4 neochordae.

AL-NC Versus PL-NC Implantation Biomechanics

Table 3 compares AL-NC and PL-NC biomechanical outcomes by presenting the percentage variations of the main hemodynamic and structural parameters of the AL-NC postrepair models with respect to the PL-NC models. In regard to hemodynamic variables, no significant differences in MR reduction were found between AL-NC and PL-NC implantation configurations. When comparing dynamic neochordae tension measurements, a significant increase in peak systolic suture tension (∼80%) for the P2/P3 AL-NC postrepair models was quantified, as pointed by the blue arrow in Figure 3. The other 2 clinical scenarios of MV prolapse did not show noticeable changes in neochordae tension between the 2 apical implantation configurations.

Table 3

Percentage variations of the hemodynamic and structural parameters for the antero-lateral neochordae (AL-NC) postrepair left heart models with respect to the postero-lateral neochordae (PL-NC) postrepair left heart models

Variable (%)	Isolated P2 NC3	Isolated P2 NC4	P2/P3 NC3	P2/P3 NC4	P2/P1 NC3	P2/P1 NC4
RVMV	–2.39	8.57	–11.06	–6.72	–5.80	–7.59
RFMV	–1.79	10.10	–10.12	–5.85	–5.51	–8.35
Fchordae
AML marginal	–2.82	–27.44	5.86	2.32	–0.90	7.13
AML strut	1.46	–1.43	1.20	0.45	1.72	4.90
PML marginal	1.04	–13.69	–4.35	24.57	8.96	–9.61
PML intermediate	–5.78	–0.56	–14.09	–14.07	1267.75	403.79
PML basal	1.03	–0.85	–12.48	–15.12	5.89	8.79
FPM
APM	–0.03	–6.29	–5.12	–5.80	2.88	5.94
PPM	–0.06	–9.47	4.25	2.54	3.63	7.11
FNC	12.43	27.06	81.99	78.73	–1.70	5.91
SɪMAXNC	7.93	23.92	70.25	83.36	7.91	6.93
SɪAVRGMV
AML	–1.18	–6.68	–2.32	1.11	–0.90	4.07
P1	–2.51	–0.61	–0.26	0.88	–3.40	–9.20
P2	0.74	–16.51	31.74	25.05	9.11	11.12
P3	–3.32	–10.00	–0.07	–0.30	–1.01	3.86
SɪMAXMV	–7.69	–0.96	109.68	140.56	–3.47	9.18

RV, Regurgitant volume in the MV; RF, regurgitant fraction in the MV; F, native chordae tension; AML, anterior mitral leaflet; PML, posterior mitral leaflet; F, PM force; APM, antero-lateral papillary muscle; PPM, postero-medial papillary muscle; F, neochordae tension; S, peak value of the maximum principal stress on the neochordae; S, average maximum principal stress in the MV leaflets; S, peak value of the maximum principal stress on the MV leaflets.

Similarly, and from a leaflet stress perspective, the P2 scallop in the P2/P3 AL-NC postrepair models showed a higher leaflet average stress (∼30%) than the PL-NC postrepair model. In contrast, no marked differences in the stress were found in the other leaflet segments (Table 3). Finally, when comparing peak leaflet stress for the P2/P3 postrepair models, the P2 peak stress increased >100% for the AL-NC configuration.

Influence of Suboptimal Neochordae Length

Table 4 and Figure 6 compare the biomechanical parameters for the isolated P2 NC4 postrepair models with optimal and suboptimal neochordae lengths. From Figure 6, A, it is evident that whereas suture overtension (shown in green) resulted in a similar RVMV as the optimal suture length model, neochordae undertension (shown in red) led to a significant increase (∼143%) in the RVMV, leading to moderate MR (as the prerepair model). Figure 6, D, shows the regurgitant jet at peak systole for the isolated P2 NC4 model with neochordae undertension. Additionally, Video 1, Video 2, Video 3, Video 4, Video 5 show the coupled blood flow dynamics and valve kinematics throughout the cardiac cycle for the physiologic, pre- and postrepair left heart models with optimal and suboptimal neochordae lengths when treating isolated P2 prolapse.

Table 4

Hemodynamic and structural parameters for the isolated P2 NC4 postrepair left heart models with optimal and suboptimal neochordae lengths

Variable	Optimal	Undertension	Overtension
SVAV (mL)	57.26	41.06	55.55
RVAV (mL)	4.62	4.54	5.28
SVMV (mL)	62.99	63.76	62.36
RVMV (mL)	11.02	26.71 (143)	12.33 (13)
RFMV (%)	16.14	39.41	18.16
MR severity (RFMV)	Trivial	Moderate	Trivial
Fchordae (newton)
AML marginal	1.71	1.60 (–6.4)	3.56 (108.8)
AML strut	3.55	3.22 (–9.3)	4.27 (20.1)
PML marginal	0.38	0.43 (12.1)	0.54 (39.4)
PML intermediate	0.52	0.40 (–21.7)	0.39 (–24.8)
PML basal	3.10	4.27 (37.6)	2.16 (–30.4)
FPM (newton)
APM	4.84	5.27 (9.0)	5.47 (13.1)
PPM	4.42	4.65 (5.1)	5.44 (23.0)
FNC (newton)	2.68	0.41 (–84.5)	12.22 (356)
Fneochorda (%)	29.8, 14.5, 33.3, 22.5	29.0, 25.6, 17.9, 27.5	25.7, 9.0, 31.1, 34.2
SɪMAXNC (MPa)	12.07	1.63 (–86.5)	56.47 (368)
SɪAVRGMV (kPa)
AML	132.7	109.7 (–17.3)	189.5 (42.8)
P1	63.3	54.2 (–14.3)	105.9 (67.3)
P2	160.1	65.7 (–58.9)	898.7 (461.2)
P3	55.5	48.7 (–12.3)	96.1 (73.0)

Percentage variations with respect to isolated P2 NC4 model with optimal neochordae length are reported in parenthesis. Fneochorda percentage values from left to right correspond to neochorda located from P1 to P3 scallops. SV, Stroke volume in the AV; RV, regurgitant volume in the AV; SV, stroke volume in the MV; RV, regurgitant volume in the MV; RF, regurgitant fraction in the MV; F, native chordae tension; AML, anterior mitral leaflet; PML, posterior mitral leaflet; F, PM force; APM, antero-lateral papillary muscle; PPM, postero-medial papillary muscle; F, neochordae tension; F, neochorda percentage distribution of total suture tension; Sɪ, peak value of the maximum principal stress on the neochordae; Sɪ, average maximum principal stress in the MV leaflets.

Figure 6

Isolated P2 NC4 postrepair left heart models with optimal and suboptimal neochordae lengths. A, Mitral valve (MV) flow rate (mL/sec) curves throughout the cardiac cycle. The negative systolic flow indicates the backflow into the left atrium due to valve closing and mitral regurgitation (MR). B, Neochordae tension (N) curves throughout the cardiac cycle. C, Average mitral leaflet stress (kPa) at peak systole. D, Velocity (mm/sec) vectors showing MR jet at peak systole with neochordae undertension. Circles highlight a marked reduction/increase (>50%) with respect to the left heart model with optimal neochordae length (orange). Video 1, Video 2, Video 3, Video 4, Video 5 linked to this figure show the left heart dynamics for the physiologic, pre- and postrepair LH models with optimal and suboptimal neochordae lengths when treating isolated P2 prolapse. MV, Mitral valve; MR, mitral regurgitation; AML, anterior mitral leaflet.

Concerning dynamic neochordae tension measurements (Figure 6, B), suture undertension by 5% resulted in a markedly lower (∼85%) total suture tension than the optimal length configuration. On the contrary, neochordae overtension by 5% caused a significant increase (>300%) in total suture tension throughout the cardiac cycle, including the diastolic phase, as pointed out by the blue arrows in Figure 6, B. Finally, a significant increase in PML systolic stress was quantified for the overtension model (Figure 6, C). Specifically, the central P2 scallop gave the highest stress variations between optimal and suboptimal neochordae configurations (460%), as seen in Table 4.

Discussion

We investigated the influence of ventricular apical entry site, neochordae number, and neochordae length on restoring physiologic left heart dynamics after transapical neochordae repair. The main findings, as shown in Figure 7, were the following:

Figure 7

Graphical abstract summarizing study methodology, main findings and clinical implications.

When using optimal neochordae length, an AL-NC implantation configuration resulted in successful MV repair from a hemodynamics perspective regardless of the complexity of the prolapse lesion. From a tissue mechanics standpoint, PML stress and native chordae tension had important changes between physiologic and postrepair states. No marked differences in intraventricular hemodynamics, native chordae tension, and leaflet average stress were found when using NC3 or NC4 for each clinical scenario of MV prolapse.

For this patient, selecting an AL-NC entry site for the treatment of multiscallop P2/P3 prolapse generated a significant increase in neochordae tension (∼80%), P2 average stress (∼30%), and P2 peak stress (>100%) with respect to using a PL-NC access site. No differences in MR reduction were found between AL-NC and PL-NC implantation configurations.

Neochordae length tuning is critical and needs careful assessment. During isolated P2 prolapse repair, although neochordae overtension by 5% led to minimal hemodynamic changes in the RVMV compared with the optimal model, a significant increase in neochordae tension (>300%), and PML average systolic stress (70%-460%) was quantified. Additionally, neochordae undertension by 5% led to MR severity worsening from trivial to moderate.

Effect of LV Entry Site on MV Loading

As shown in Table 3 and Figure 3, a marked increase in P2 peak stress (>100%) and systolic suture tension (∼80%) was quantified for the P2/P3 AL-NC postrepair models compared with using a PL-NC access site, although a similar RVMV was quantified for both implantation configurations. We hypothesized that this increase in leaflet stress and neochordae tension during AL-NC repair can be related to a less physiological anterior trajectory of the implanted sutures, together with a longer suture length compared with a more posterior insertion that provides a more natural orientation for the neochordae, with the LV anchoring point close to the base of the postero-medial PM., It has been shown that neochordae insertion from the anterior aspect of the LV imparts both apical as well as anterior forces, the latter of which serves to decrease and stabilize the antero-posterior dimension of the mitral annulus. A slightly more anterior entry site also modifies the working angle of the PML, stretches it below the AML, and thereby increases the potential leaflet coaptation. On the other hand, some studies do not recommend an excessively anterior apical access site for neochordae implanted on the PML, due to the unphysiologic crossing of the LV outflow tract and the significant risk of interference with the native chordae that may lead to AML rupture.

It is important to underscore that current clinical experience is per protocol, mainly limited to isolated P2/A2 prolapse cases, not supporting the evidence that AL-NC or PL-NC procedures will be effective or ineffective on the wide spectrum of primary MR, such as multiscallop, bileaflet, or paracommissural diseases., Procedure refinements, especially the revision of the LV entry site will be critical when attempting to restore physiologic LV-MV dynamics. It is possible that this field will split in the near future between transapical technologies allowing ease of precise implantation and transseptal technologies (eg, Pipeline Medical Technologies Inc, Wilmington, Del) allowing true minimal invasiveness. Nonetheless, an optimal LV access site should not interfere with the MV subvalvular apparatus, preserve LV-valve physiology, and mantain the best symmetry and force distribution between the neochordae. A more in-depth discussion about physiologic versus AL-NC implantation biomechanics can be found in Appendix 1.

Effect of LV Entry Site on Neochordae Mechanics

During transapical neochordae repair is important to know how forces are distributed along the sutures when tightened to the LV apex (anterior and posterior anchoring), and how they compared with native chordae tension curves (PM anchoring). Our computer framework allowed for a direct quantification and comparison of these forces in a simulated human beating heart for the first time (Figure 3). In an in vivo animal study by Jensen and colleagues, force measurements on 1 neochorda attached to the antero-lateral PM tip and to the postero-lateral LV wall were compared in 8 pigs with AML prolapse. Although no significant differences in neochorda tension measurements were found, the rate of loading (calculated as dFneochordae/dtmax), which reflects the tension fluctuations in the sutures, was significantly higher in the transapical fixation group. It was suggested that such abnormal and faster suture tension fluctuations might be explained by the absence of the shock absorbing effect of the PM, or by the increased transapical neochorda length relative to the PM implantation length., Table 5 shows the maximum loading rate of the AL-NC and PL-NC tension curves shown in Figure 3. A marked increase (>30%) in the slope from PL-NC to AL-NC implantation configurations was found for the isolated P2 and P2/P3 postrepair models. This means that under our simulated patient conditions, suture tension increased more rapidly in the AL-NC implantation group than in the PL-NC group, resulting in a more abnormal and dynamic loading profile.

Table 5

Maximum slope (dFneochordae/dtmax) of the neochordae tension curves for the postrepair left heart models with optimal neochordae length

Isolated P2
AL-NC3	178
AL-NC4	233
PL-NC3	138
PL-NC4	166
P2/P3
AL-NC3	372
AL-NC4	378
PL-NC3	242
PL-NC4	286
P2/P1
AL-NC3	289
AL-NC4	286
PL-NC3	260
PL-NC4	264

AL-NC, Antero-lateral neochordae; PL-NC, postero-lateral neochordae.

This abnormal fluttering and increased rate of loading of the transapical neochordae might predispose to leaflet tears, native and neochordae rupture, and repair failure. Durability of ePTFE sutures is well established, with only a few cases of neochordae rupture reported in conventional MV repair surgery in the past 25 years. However, recent cases of late suture rupture after transapical neochordae implantation have been reported.,, The length of neochordae implanted transapically is often twice the length of those anchored to the PM tips under conventional repair. Biomechanical studies have shown that an increase in neochordae length is accompanied by a change in suture performance, especially an increase in the stiffness that can compromise the long-term resistance of the sutures. Careful evaluation of neochordae failure and possible complications should be performed on a patient-specific basis. If the suture rupture occurs at the MV leaflets, the repair of the valve can be completed with new neochordae or converted to conventional MV surgery. In the case of rupture from the apex site, surgeons must concern themselves with myocardial fragility.

Effect of Suboptimal Neochordae Length on Procedure Outcome

Previous studies have suggested that suboptimal neochordae tensioning during conventional MV repair can lead to leaflet stress hot spots, excessive localized increase of suture tension, and an unbalanced mitral apparatus configuration. During treatment of isolated P2 prolapse, our analysis showed that whereas transapical sutures with an optimal length carried a net peak systolic load of 2.7 N, overtensioned neochordae carried a significant higher total load of 12.2 N (Table 4). Thus, the force exerted by the shortened sutures was 360% higher than for sutures with an optimal length, although hemodynamic outcomes were similar. Our study also found that the highest tension experienced by a single neochorda was 4.18 N, which corresponded to the overtension model. In all postrepair models, the peak neochorda tension was well below the failure load of a ePTFE CV-4 suture, which is about 16 N.

This failure load or ultimate tensile strength is considered a safety factor to ensure the material is not going to fail. In this regard, most materials are overdesigned in term of this requirement. In the short-term, CV-4 ePTFE suture tensile strength is unlikely to be approached under the immediate postrepair loading conditions modeled in this study. Rather, the relationship between material elongation and applied load, known as stiffness, as well as the loading rate may be more important in governing the material response over time, as previously discussed. Moreover, it was found that the stress on the leaflets is directly related to the neochordae tensile force (Table 4). Although the measured leaflet stress under suture overtension is below the mitral leaflet's ultimate strength, it can be speculated that the local stress overload induced by the sutures on the leaflet could be a key factor in triggering local biological mechanisms such us tissue growth and remodeling, affecting valve tissue integrity, which could then have a large effect on the durability of the procedure over time.

A key finding of this study is that during suture overtension there was not only a significant increase in neochordae tension during systole but also during diastole (Figure 6, B). During the cardiac cycle there is a synchronized reciprocal behavior between the mitral annulus and the PM. Under physiologic conditions, this phenomenon probably induces small elongations/shortenings of the native mitral chordae. Under transapical neochordae implantation, much larger changes in neochordae length can occur because the distance between the mitral annulus and the LV apex decreases in systole and increases in diastole. Thus, this phenomenon probably induces important elongations of the neochordae that can be exacerbated if initial suture overtension is performed. These variations can generate an unwanted high neochordae deformation during diastole as shown in Figure 6, B, that translates to increased stress on the mitral leaflets (Table 4). Furthermore, a slight overtension is usually clinically applied during neochordae final fixation as a result of LV remodeling and volume reduction after the procedure. Overall, careful estimation of initial neochordae length and mitral apparatus stress reduction are paramount to ensure mid- to long-term durability of the repair procedure.

Limitations

The following limitations should be taken into account when interpreting the results of this study. First, this work is based on a previously validated subject-specific LH model that was modified to simulate various transapical neochordae implantation procedures. Thus, our results cannot be assumed to represent the entire population and caution should be taken to extrapolate our findings into the clinical setting. Nevertheless, such well-controlled side-by-side comparisons under the same patient and working conditions are difficult to obtain clinically. Developing a large cohort of patient-specific LH models will provide improved correlations between biomechanical parameters and clinical outcomes. Second, we used healthy MV tissue properties for the physiologic, pre-, and postrepair LH models. Primary MR can be associated with the alterations of tissue characteristics involving myxomatous degeneration. Therefore, simulation results may differ if diseased valvular data are incorporated.

Third, although image-based prescribed motion was used for the endocardial wall and mitral annulus, the aortic annulus was kept fixed during the simulations. Due to their anatomic linkage, AV and MV function are coupled with reciprocal behavior during the cardiac cycle. Thus, aortic-mitral coupling motion simplification is likely to affect valve and neochordae deformation results. Fourth, this study aimed to evaluate the changes in left heart biomechanics immediately after transapical neochordae implantation for acute MR correction. Thus, the same endocardial wall motion was maintained after the repair procedures, without considering any compensatory cardiac remodeling mechanisms. Similarly, the pressure boundary conditions employed did not take into account the relationship of pressure and flow rate at the upstream and downstream vasculature of the patient. The use of lumped-parameter models at the inlet and outflow boundaries would allow to implement more realistic pressure waveforms and consider the entire circulation of the patient under different conditions.

Conclusions

In this work, we quantified the complex biomechanical interaction between transapical neochordae and the left heart complex under various PML prolapse conditions, LV access sites, and neochordae number and length. We report original quantitative data that assess the magnitude and time course of the force transfer between the neochordae, the mitral apparatus, and the blood flow at physiologic, pre- and postrepair states using an FSI computer modeling framework. Despite apparent similarities between conventional and catheter-based neochordae implantation procedures, some major differences such as suture axis and length can cause differences in the short- to long-term results. Thus, a direct comparison of the LV-valve biomechanical environment between physiologic and postrepair states is imperative to evaluate the role of transapical neochordae technologies in restoring physiologic-like cardiac dynamics and demonstrate noninferiority to surgical MV repair.

Conflict of Interest Statement

Dr Sun is a cofounder and serves as chief scientific advisor of Dura Biotech. He receives compensation and owns equity in the company. All other authors reported no conflicts of interest.

The Journal policy requires editors and reviewers to disclose conflicts of interest and to decline handling or reviewing manuscripts for which they may have a conflict of interest. The editors and reviewers of this article have no conflicts of interest.

Table 6

Material parameters of cardiac tissues

Model	Parameter
MHGO model	C10(kPa)	C01	k1 (kPa)	k2	θ (°)	κ	D (kPa−1)
AV leaflets	1.738	11.368	2159.4	1158.9	4.59	0.2359	1.0e-5
AML	0.1245	13.665	11.007	84.84	13.09	0.0800	1.0e-5
PML	0.0502	15.004	3.021	144.48	25.51	0.0534	1.0e-5
Ogden model	μ1 (kPa)	a1	μ2 (kPa)	a2	μ3 (kPa)	a3
Basal/intermediate chordae	10256.1	16.579	10653.8	16.554	10671.3	16.554
Strut chordae	24341.7	11.338	10331.9	11.167	14913.6	11.188
Marginal chordae	12995.5	15.651	13082.9	15.683	12869.7	15.662

MHGO, Anisotropic hyperelastic material model; AV, aortic valve; AML, anterior mitral leaflet; PML, posterior mitral leaflet.