The impact of leaflet resection versus leaflet preservation on left ventricular reverse remodelling after mitral valve repair: insights from the UK mini mitral trial Authors: Eteesha Rao, Christopher D. Bayliss, Janelle M. Wagnild, Richard Graham, Rebecca Maier, Enoch Akowuah Published: 2026 ABSTRACT ObjectivesMitral valve repair (MVr) promotes left ventricular (LV) reverse remodelling, and subsequent reduction in LV diameters and volumes. The effectiveness of different MVr techniques, Resect (leaflet resection) and Respect (leaflet preservation) remains debated. This study aimed to investigate the impact of ‘Resect’ and ‘Respect’ techniques on LV reverse remodelling, primarily assessed by changes in indexed LV end-systolic diameter (ILVESD) post-MVr.MethodsData were drawn from the UK Mini Mitral Trial, which randomised 330 patients to MVr via mini-thoracotomy or median-sternotomy. This subanalysis included patients with isolated posterior leaflet prolapse, who had undergone MVr via Resect or Respect techniques. Transthoracic echocardiography was performed pre-operatively and at 12 and 52 weeks post-operatively, with all studies analysed by a core laboratory. Linear mixed-effects models compared changes in indexed LV measurements between groups.ResultsThis subanalysis included 175 patients (36 Resect and 139 Respect). Annuloplasty was performed in all but one repair. We found no significant difference between groups in recurrent MR severity at 52 weeks postoperatively (p = 0.31). Both groups had significant improvement in ILVESD by 52 weeks (mean change from baseline − 5.6%, p < 0.001), with no difference between groups (p = 0.24). Both techniques led to significant improvements in indexed end-diastolic LV volumes and diameters, decreasing from baseline to 12 and 52 weeks. However, LV end-systolic volumes and diameters increased at 12 weeks before decreasing by 52 weeks postoperatively.ConclusionNo significant differences in LV reverse remodelling between Resect and Respect techniques. Both can be recommended to surgeons. Longer follow-up is needed.Graphical Abstract Introduction Mitral valve repair (MVr) is the preferred surgical approach for treating patients with severe mitral regurgitation (MR) [1] Globally, MR is the second most common valvular abnormality with a prevalence of 2–10%, which increases with age [2–4]. Severe MR can reduce normal life expectancy without intervention, with an average mortality rate of ~ 5% per year and 70% at 8 years [2]. In chronic progressive MR, left ventricular (LV) remodelling occurs to compensate and allow for the cardiac output to be maintained. Initially, there may be an observed increase in LV ejection fraction (LVEF), however the effective ejection fraction may be reduced, dependent on the volume of the regurgitant fraction [2, 3]. As the disease progresses, a positive feedback loop may establish due to the volume overload, causing LV dilatation and eccentric remodelling [2, 3]. After patients with MR undergo MVr, the LV normally undergoes reverse remodelling, reverting towards its normal morphology over time. The exact sequence for these changes has been debated with studies reporting different patterns. Some studies, such as Shafii et al. 2012 have reported a significant decline in LV end-systolic diameters (LVESD) and LV end-diastolic diameters (LVEDD) at one year following MVr, whilst Suri et al., 2009 found that only the LVEDD had a significant decline with LVESD not changing appreciably postoperatively until three to five years [5, 6]. Similar discrepancies exist regarding postoperative changes in LV volumes, with Le Tourneau’s two-stage phased model being the preferred hypothesis [7, 8]. The two-stage phased model proposed by Le Tourneau et al. (2019), based on their single centre retrospective study, compared non-indexed LV dimensions and volumes after MVr during the early postoperative period (within 45 days of MVr surgery) with the late postoperative period (mean 428 days). In the first phase, during the immediate postoperative period, they found a reduction in LVEDV with no change in left ventricular end-systolic volume (LVESV). This was followed by a second stage of reverse LV remodelling, featuring a decrease in LVEDV and LVESV [8]. In most western populations, posterior leaflet mitral valve (MV) prolapse due to degenerative MV disease remains the most common MV pathology. This can be attributed to the ageing population. Posterior leaflet MV pathologies were traditionally repaired using the ‘Resect’ approach, first described and evidenced in 1983, as “the French correction” by Alain Carpentier [9]. This approach involves resecting away the leaflet tissue and became the gold standard technique until more recent wider adoption of the ‘Respect’ approach, which has increasingly been reported to be an equally effective, if not arguably superior approach [10–12]. The Respect technique involves the placement of one or more expanded polytetrafluoroethylene neo-chords, without resecting any leaflet tissue. This was later proposed as an alternative approach that may allow for preserved leaflet mobility with minimal changes to annular geometry and the ability to implant larger annuloplasty rings [13, 14]. Thus, creating the recent shift in MVr surgical practice to ‘Respect rather than Resect’ [10]. Over the past three decades, several studies have compared Resect and Respect MVr techniques using various outcomes, commonly: residual MR severity, recurrent MR, reoperation rates and LV changes [15]. Some studies have found no significant difference between Resect and Respect in the aforementioned outcomes [7, 15–18]. Others found Resect to be superior [19–20], whilst some found Respect to outperform Resect [2, 10–12, 21, 22]. A major limitation with the current studies is that they predominantly have a small sample size and/or are single centre observational studies, with the potential to be influenced by selection bias [17, 20, 21, 23]. Thus, it is difficult to definitively state whether one repair technique is superior to the other. To our knowledge, there are currently only three published studies assessing the impact of MVr technique on LV reverse remodelling: two single centre observational studies and one small randomised controlled trial (RCT) [7, 23, 24]. This study aims to use high-quality data from the UK Mini Mitral Trial to address previously reported discrepancies in the literature and ascertain if there are any differences in LV reverse remodelling between Resect and Respect approaches at 52 weeks post-MVr. Patients and methods Ethical statement The UK Mini Mitral Trial was granted ethical approval by the NHS Wales Ethics Committee 6 (16/WA/0156). This sub-analysis is covered by the same ethical approval. Study population: the UK mini mitral trial The UK Mini Mitral Trial was a multicentre, superiority RCT which randomised 330 patients with degenerative MR between November 2016 and January 2021. The trial protocol has been previously published [25, 26]. This subanalysis only included patients with isolated posterior leaflet pathology that had undergone MVr through Resect or Respect approaches. Patients were excluded if they had: anterior or bileaflet pathologies, received MV replacement, any other MVr techniques (excluding ring annuloplasty) or not completed the 52-week echocardiography core laboratory follow-up. Surgical techniques MVr was carried out through mini-thoracotomy or conventional median-sternotomy. Patients undergoing MVr via the Resect approach had the prolapsing segment of the posterior leaflet excised via either triangular or quadrangular resection (+/- sliding plasty). All patients undergoing the Respect approach underwent chordal reconstruction using artificial expanded polytetrafluoroethylene (ePTFE). A ring annuloplasty was also performed for all but one repair. Concomitant procedures for atrial fibrillation or tricuspid valve repair were allowed in the trial. Patients requiring other additional procedures or emergency surgery were excluded. All operations were performed by experienced cardiac surgeons who had performed at least 50 MVr surgeries previously. Outcomes The primary outcome of this study was the difference in indexed LVESD, measured by transthoracic echocardiography (TTE) assessed by a core laboratory, compared between Resect and Respect groups from baseline to 52 weeks after MVr surgery. Secondary outcomes were early differences in indexed LV end-systolic and end-diastolic measurements from baseline to 12 weeks postoperatively; late differences in indexed LV end-systolic and end-diastolic measurements from baseline to 52 weeks postoperatively and differences in severity of residual MR at 52 weeks postoperatively between Resect and Respect groups. Echocardiography All patients underwent TTE at three specified time points. These were at baseline (prior to surgery), early post-operative period (within 12 weeks after surgery) and late post-operative period (52 weeks after surgery). A central blinded independent core laboratory adjudicated echocardiographic data from all patients at all three time points to obtain data on LV diameters and volumes at end-systole and end-diastole. Statistical analysis Statistical analyses were conducted using IBM SPSS version 29.0.2.0 [27]. Comparisons of categorical variables were conducted using chi-squared or Fisher’s exact test, depending on sample size. Continuous variables were assessed using independent t-tests if normally distributed or the Mann-Whitney U test if non-normally distributed. Primary and secondary outcomes were analysed using linear mixed-effects models adjusted for surgical approach (mini-thoracotomy versus sternotomy) and the variables used as minimisation factors in the UK-Mini’s randomisation scheme (presence of AF (present/absent), severity of tricuspid regurgitation (none/mild/moderate/unknown), and the patient’s baseline Physical Function Score based on the SF-36v2 (low (0–33), moderate (> 33–66), high (> 66)). Intra-patient correlation was taken into account using a random intercept model. Time was included as a fixed effect to estimate changes in indexed LV diameters and volumes from baseline to 52 weeks post-MVr, regardless of Resect or Respect technique. To determine whether changes in indices over time differed by MVr technique, tests for interaction between time and technique were included in the model. Statistical significance was set at p < 0.05 with mean differences and 95% confidence intervals reported. Results After assessing the data from all 330 patients in the UK-Mini, 175 patients met the inclusion criteria with 36 Resect and 139 Respect patients for the comparative analyses, refer CONSORT diagram in Fig. 1. Fig. 1CONSORT diagram CONSORT diagram Patient demographics Baseline patient demographics were compared between Resect and Respect groups and are displayed in Table 1. Patients in the Respect group were more likely to have a higher NYHA classification than those in the Resect group. Table 1Baseline patient characteristicsTotaln = 175Respectn = 139Resectn = 36 Age, years, Mean ± SD 66 ± 1066 ± 1066 ± 10Male, n (%)131 (75)103 (74)28 (78)NYHA class, n (%) 123 (13)19 (14)4 (11) 259 (34)39 (28)20 (56) 370 (40)62 (45)8 (22) 423 (13)19 (14)4 (11)NYHA class > 2, n (%)93 (53)81 (58)12 (33)BMI, mean ±SD26.4 ± 4.226.5 ± 4.025.9 ± 4.7EuroSCORE II, Mean ± SDn = 1741.58 ± 1.31.62 ± 1.31.42 ± 1.2Clinical History, n (%) Atrial fibrillation65 (37)53 (38)12 (33) Tricuspid Regurgitationn = 3524 (69)17 (12)7 (19) Heart failure51 (29)41 (29)10 (28) Diabetes13 (7)10 (7)3 (8) Pulmonary hypertension32 (18)27 (19)5 (14) Previous myocardial infarction4 (2)4 (3)0 (0) Stroke9 (5)6 (4)3 (8) Chronic obstructive pulmonary disease9 (5)8 (6)1 (3) Asthma12 (7)8 (6)4 (11) Renal failure6 (3)6 (4)0 (0) Peripheral vascular disease1 (<1)0 (0)1 (3)Urgency, n (%) Elective153 (87)121 (87)32 (89) Urgent22 (13)18 (13)4 (11)Carpentier class, n (%) Type I8 (5)7 (5)1 (3) Type II155 (89)121 (87)34 (94) Type I + II12 (7)11 (8)1 (3)SF-36v2-PF, n (%) Low (<33)37 (21)31 (22)6 (17) Moderate (33-66)61 (35)49 (35)12 (33) High (>66)77 (44)59 (42)18 (50) Table 1Baseline patient characteristicsTotaln = 175Respectn = 139Resectn = 36 Age, years, Mean ± SD 66 ± 1066 ± 1066 ± 10Male, n (%)131 (75)103 (74)28 (78)NYHA class, n (%) 123 (13)19 (14)4 (11) 259 (34)39 (28)20 (56) 370 (40)62 (45)8 (22) 423 (13)19 (14)4 (11)NYHA class > 2, n (%)93 (53)81 (58)12 (33)BMI, mean ±SD26.4 ± 4.226.5 ± 4.025.9 ± 4.7EuroSCORE II, Mean ± SDn = 1741.58 ± 1.31.62 ± 1.31.42 ± 1.2Clinical History, n (%) Atrial fibrillation65 (37)53 (38)12 (33) Tricuspid Regurgitationn = 3524 (69)17 (12)7 (19) Heart failure51 (29)41 (29)10 (28) Diabetes13 (7)10 (7)3 (8) Pulmonary hypertension32 (18)27 (19)5 (14) Previous myocardial infarction4 (2)4 (3)0 (0) Stroke9 (5)6 (4)3 (8) Chronic obstructive pulmonary disease9 (5)8 (6)1 (3) Asthma12 (7)8 (6)4 (11) Renal failure6 (3)6 (4)0 (0) Peripheral vascular disease1 (<1)0 (0)1 (3)Urgency, n (%) Elective153 (87)121 (87)32 (89) Urgent22 (13)18 (13)4 (11)Carpentier class, n (%) Type I8 (5)7 (5)1 (3) Type II155 (89)121 (87)34 (94) Type I + II12 (7)11 (8)1 (3)SF-36v2-PF, n (%) Low (<33)37 (21)31 (22)6 (17) Moderate (33-66)61 (35)49 (35)12 (33) High (>66)77 (44)59 (42)18 (50) Baseline patient characteristics Intraoperative details Intraoperative details were compared between Resect and Respect groups, as summarised in Table 2. Triangular (67%) and quadrangular (33%) resection were the two Resect techniques used. All Respect patients underwent chordal reconstruction with artificial chordae. Table 2Intraoperative detailsTotal population n = 175Respect group n = 139Resect group n = 36Risk ratio (respect vs. resect: 95% CI)p valueApproach, n (%)<0.001 Minimally invasive89 (51)80 (58)9 (25)1.31 (1.12, 1.54) Sternotomy86 (49)59 (42)27 (75)MVr technique, n (%)Leaflet resection Triangular resection0 (0)24 (67) Quadrangular resection0 (0)12 (33) Neochord reconstruction139 (100)0 (0)Ring annuloplasty,n (%)174 (99)139 (100)35 (97) Complete ring150 (86)122 (88)28 (80) Incomplete ring24 (14)17 (12)7 (20) Mean difference (respect-resect) [95% CI] Ring size, mmMean ± SDn= 17432 ± 2.531 ± 2.333 ± 2.5-2.24 [-3.12,-1.36]<0.001 Perfusion details Mean ± SDCardiopulmonary bypass time, min114 ± 34117 ± 33101 ± 3615.9 [3.62, 28.2]0.01Aortic cross-clamp time, min82 ± 7484 ± 8373 ± 2011.5 [-16.0, 38.9]0.41 Table 2Intraoperative detailsTotal population n = 175Respect group n = 139Resect group n = 36Risk ratio (respect vs. resect: 95% CI)p valueApproach, n (%)<0.001 Minimally invasive89 (51)80 (58)9 (25)1.31 (1.12, 1.54) Sternotomy86 (49)59 (42)27 (75)MVr technique, n (%)Leaflet resection Triangular resection0 (0)24 (67) Quadrangular resection0 (0)12 (33) Neochord reconstruction139 (100)0 (0)Ring annuloplasty,n (%)174 (99)139 (100)35 (97) Complete ring150 (86)122 (88)28 (80) Incomplete ring24 (14)17 (12)7 (20) Mean difference (respect-resect) [95% CI] Ring size, mmMean ± SDn= 17432 ± 2.531 ± 2.333 ± 2.5-2.24 [-3.12,-1.36]<0.001 Perfusion details Mean ± SDCardiopulmonary bypass time, min114 ± 34117 ± 33101 ± 3615.9 [3.62, 28.2]0.01Aortic cross-clamp time, min82 ± 7484 ± 8373 ± 2011.5 [-16.0, 38.9]0.41 Intraoperative details Ring annuloplasty, n (%) Ring size, mm Mean ± SD Perfusion details Mean ± SD The surgical approach used for the MVr was significantly different (p < 0.001) between groups, with the Respect technique being the preferred option in patients undergoing the minimally invasive approach (89.9% Respect versus 10.1% Resect). The Respect patients also had longer bypass times (p = 0.01), likely due to the higher uptake of the minimally invasive approach with the Respect technique. All patients except one in the Resect group received an annuloplasty ring. There was a significant difference in annuloplasty ring type and size between groups (Resect group mean 33 mm versus Respect group mean 31 mm, p < 0.001), refer Fig. 2. Fig. 2Left ventricular changes from baseline to 52 weeks: Resect vs. Respect MVr techniques Left ventricular changes from baseline to 52 weeks: Resect vs. Respect MVr techniques MR severity at 52 weeks: resect versus respect We observed no significant difference in qualitative MR severity between Resect and Respect groups at 52 weeks postoperatively (p = 0.31). Primary outcome: results We found no significant difference in change in ILVESD (mm/m²), measured by TTE, between Resect and Respect groups from baseline to 52 weeks (p = 0.24), refer Fig. 2. LV remodelling post-MVr: total population Across the total patient population in this study (n = 175), there was a significant reduction in ILVEDD and ILVEDV post-MVr regardless of repair technique at 12 weeks post-MVr as compared to baseline (see Table 3). There was a significant increase (p = 0.03) in ILVESV and a non-significant increase in ILVESD (p = 0.18) in this early postoperative period. However, by the late postoperative period, at 52 weeks, there was a significant decrease in all indexed LV end-systolic and end-diastolic volumes and diameters post-MVr regardless of repair technique as compared to baseline. Table 3Comparison of indexed LV diameters and volumes from baseline to 12 and 52 weeksTotal Patients (n=175)Baseline12 weeks52 weeksBaseline to 12 weeks differenceBaseline to 52 weeks differenceMean ± SDMean ± SDMean ± SDMean difference (12 weeks-baseline:[95% CI])p valueMean difference (52 weeks-baseline:[95% CI])p valueILVEDD mm/m228.7 ± 3.7(n=171)25.6 ± 3.2(n=170)24.6 ± 3.2(n=175)-3.04[-3.47, -2.60](n=166)<0.001-3.99[-4.46, -3.53](n=171)<0.001ILVESD, mm/m217.9 ± 3.3(n=171)18.2 ± 3.4(n=170)16.8 ± 2.8(n=175)0.31[-0.15, 0.76](n=166)0.18-1.01[-1.53, -0.61](n=171)<0.001ILVEDV, mL/m379.0 ± 21(n=170)58.7 ± 17(n=170)56.0 ± 12(n=171)-20.1[-22.7, -17.5](n=166)<0.001-23.0[-26.1, -20.0](n=166)<0.001ILVESV, mL/m325.0 ± 9.4(n=170)26.6 ± 13(n=170)22.1 ± 7.3(n=171)1.62[0.14, 3.09](n=166)0.03-2.95[-4.23, -1.6](n=166)<0.001 Table 3Comparison of indexed LV diameters and volumes from baseline to 12 and 52 weeksTotal Patients (n=175)Baseline12 weeks52 weeksBaseline to 12 weeks differenceBaseline to 52 weeks differenceMean ± SDMean ± SDMean ± SDMean difference (12 weeks-baseline:[95% CI])p valueMean difference (52 weeks-baseline:[95% CI])p valueILVEDD mm/m228.7 ± 3.7(n=171)25.6 ± 3.2(n=170)24.6 ± 3.2(n=175)-3.04[-3.47, -2.60](n=166)<0.001-3.99[-4.46, -3.53](n=171)<0.001ILVESD, mm/m217.9 ± 3.3(n=171)18.2 ± 3.4(n=170)16.8 ± 2.8(n=175)0.31[-0.15, 0.76](n=166)0.18-1.01[-1.53, -0.61](n=171)<0.001ILVEDV, mL/m379.0 ± 21(n=170)58.7 ± 17(n=170)56.0 ± 12(n=171)-20.1[-22.7, -17.5](n=166)<0.001-23.0[-26.1, -20.0](n=166)<0.001ILVESV, mL/m325.0 ± 9.4(n=170)26.6 ± 13(n=170)22.1 ± 7.3(n=171)1.62[0.14, 3.09](n=166)0.03-2.95[-4.23, -1.6](n=166)<0.001 Comparison of indexed LV diameters and volumes from baseline to 12 and 52 weeks 28.7 ± 3.7 (n=171) 25.6 ± 3.2 (n=170) 24.6 ± 3.2 (n=175) -3.04 [-3.47, -2.60] (n=166) -3.99 [-4.46, -3.53] (n=171) 17.9 ± 3.3 (n=171) 18.2 ± 3.4 (n=170) 16.8 ± 2.8 (n=175) 0.31 [-0.15, 0.76] (n=166) -1.01 [-1.53, -0.61] (n=171) 79.0 ± 21 (n=170) 58.7 ± 17 (n=170) 56.0 ± 12 (n=171) -20.1 [-22.7, -17.5] (n=166) -23.0 [-26.1, -20.0] (n=166) 25.0 ± 9.4 (n=170) 26.6 ± 13 (n=170) 22.1 ± 7.3 (n=171) 1.62 [0.14, 3.09] (n=166) -2.95 [-4.23, -1.6] (n=166) Early LV changes post-MVr: resect versus respect In the early postoperative period at 12 weeks, there was no significant difference in ILVESD (p = 0.35), ILVEDD (p = 0.56), ILVEDV (p = 0.06) and ILVESV (p = 0.24) post-MVr between Resect and Respect techniques (refer Table 4). There was also no significant difference in mean transmitral gradient (p = 0.34) and coaptation length (p = 0.25) between groups. Table 4Post-MVr echocardiographic findings from baseline to2 and 52 weeks with differences betweenResect and Respect groupsRespect group (n=139)Resect group (n=36)Difference at baselineResect vs. Respect: difference of changesBaseline12 weeks52 weeksBaseline12 weeks52 weeksBaseline to 12 weeksBaseline to 52 weeksMean ± SDMean ± SDMean ± SDMean ± SDMean ± SDMean ± SDp value[95% CI]Mean difference (respect-resect:[95% CI]) p valueMean difference (respect-resect:[95% CI])p valueILVEDD, mm/m228.5 ± 3.4(n=136)25.5 ± 3.2(n=135)24.5 ± 3.3(n=139)29.5 ± 4.4(n=35)25.9 ± 3.3(n=35)25.0 ± 3.0(n=36)0.13[-2.42, 0.32]0.42[-0.66,1.50]0.450.60[-0.55, 1.76]0.31ILVESD, mm/m217.8 ± 3.2(n=136)18.0 ± 3.3(n=135)16.6 ± 2.8(n=139)18.1 ± 3.7(n=35)18.6 ± 3.5(n=35)17.5 ± 2.8(n=36)0.64[-1.54,0.95]-0.49[-1.60,0.63]0.39-0.69[-1.82,0.45]0.24ILVEDV, mL/m378.2 ± 20(n=137)57.5 ± 17(n=135)55.6 ± 12(n=136)82.2 ± 23(n=33)63.5 ± 16(n=35)57.6 ± 12(n=35)0.32[-12.0, 3.94]-1.77[-8.28, 4.85]0.602.15[-5.46,9.94]0.58ILVESV, mL/m325.0 ± 9.6(n=137)26.0 ± 14(n=135)21.7 ± 7.4(n=136)25.1 ± 8.4(n=33)28.9 ± 12(n=35)23.6 ± 6.9(n=35)0.96[-3.70, 3.52]-2.59[-6.27,1.10]0.17-1.62[-4.85,1.62]0.33 Table 4Post-MVr echocardiographic findings from baseline to2 and 52 weeks with differences betweenResect and Respect groupsRespect group (n=139)Resect group (n=36)Difference at baselineResect vs. Respect: difference of changesBaseline12 weeks52 weeksBaseline12 weeks52 weeksBaseline to 12 weeksBaseline to 52 weeksMean ± SDMean ± SDMean ± SDMean ± SDMean ± SDMean ± SDp value[95% CI]Mean difference (respect-resect:[95% CI]) p valueMean difference (respect-resect:[95% CI])p valueILVEDD, mm/m228.5 ± 3.4(n=136)25.5 ± 3.2(n=135)24.5 ± 3.3(n=139)29.5 ± 4.4(n=35)25.9 ± 3.3(n=35)25.0 ± 3.0(n=36)0.13[-2.42, 0.32]0.42[-0.66,1.50]0.450.60[-0.55, 1.76]0.31ILVESD, mm/m217.8 ± 3.2(n=136)18.0 ± 3.3(n=135)16.6 ± 2.8(n=139)18.1 ± 3.7(n=35)18.6 ± 3.5(n=35)17.5 ± 2.8(n=36)0.64[-1.54,0.95]-0.49[-1.60,0.63]0.39-0.69[-1.82,0.45]0.24ILVEDV, mL/m378.2 ± 20(n=137)57.5 ± 17(n=135)55.6 ± 12(n=136)82.2 ± 23(n=33)63.5 ± 16(n=35)57.6 ± 12(n=35)0.32[-12.0, 3.94]-1.77[-8.28, 4.85]0.602.15[-5.46,9.94]0.58ILVESV, mL/m325.0 ± 9.6(n=137)26.0 ± 14(n=135)21.7 ± 7.4(n=136)25.1 ± 8.4(n=33)28.9 ± 12(n=35)23.6 ± 6.9(n=35)0.96[-3.70, 3.52]-2.59[-6.27,1.10]0.17-1.62[-4.85,1.62]0.33 Post-MVr echocardiographic findings from baseline to2 and 52 weeks with differences betweenResect and Respect groups 28.5 ± 3.4 (n=136) 25.5 ± 3.2 (n=135) 24.5 ± 3.3 (n=139) 29.5 ± 4.4 (n=35) 25.9 ± 3.3 (n=35) 25.0 ± 3.0 (n=36) 0.13 [-2.42, 0.32] 0.42 [-0.66,1.50] 0.60 [-0.55, 1.76] 17.8 ± 3.2 (n=136) 18.0 ± 3.3 (n=135) 16.6 ± 2.8 (n=139) 18.1 ± 3.7 (n=35) 18.6 ± 3.5 (n=35) 17.5 ± 2.8 (n=36) 0.64 [-1.54,0.95] -0.49 [-1.60,0.63] -0.69 [-1.82,0.45] 78.2 ± 20 (n=137) 57.5 ± 17 (n=135) 55.6 ± 12 (n=136) 82.2 ± 23 (n=33) 63.5 ± 16 (n=35) 57.6 ± 12 (n=35) 0.32 [-12.0, 3.94] -1.77 [-8.28, 4.85] 2.15 [-5.46,9.94] 25.0 ± 9.6 (n=137) 26.0 ± 14 (n=135) 21.7 ± 7.4 (n=136) 25.1 ± 8.4 (n=33) 28.9 ± 12 (n=35) 23.6 ± 6.9 (n=35) 0.96 [-3.70, 3.52] -2.59 [-6.27,1.10] -1.62 [-4.85,1.62] Late LV changes post-MVr: resect versus respect In the late postoperative period at 52 weeks, there was no significant difference in ILVEDD (p = 0.31), ILVEDV (p = 0.57) and ILVESV (p = 0.33) between Resect and Respect techniques as compared to baseline, refer Table 4. Discussion This observational, multicentre, cohort study assessed postoperative outcomes for 175 MR patients with degenerative posterior leaflet pathology, who had undergone MVr via either the Resect or Respect technique. The main finding of this study was that there is no significant difference in LV reverse remodelling between Resect and Respect approaches at 52 weeks post-MVr surgery. Our findings are in line with the only published RCT but not in keeping with previous single centre, observational studies by [7, 23, 24]conducted the only RCT comparing Resect and Respect MVr techniques and their effect on LV reverse remodelling. 104 patients from seven Canadian centres were included. 54 patients were randomised to the Resect group and 50 patients to the Respect group. They measured indexed LV end-systolic and end-diastolic measurements in the early postoperative period and at one year, finding no significant difference in LV reverse remodelling between techniques. The time period for their early postoperative period was immediately before discharge and did not match with our study, which was within 12 weeks of MVr surgery [23]. The propensity-matched retrospective study Del Forno et al. (2023) matched 170 patients to Resect and Respect groups in a 1:1 ratio. In contrast to our results, they found that the Respect group had improved reverse LV remodelling as compared to the Resect group [24]. They reported a higher reduction in LVEDD in the Respect group at last follow-up (p < 0.001) with average last follow-up being at 5.97 years (4.49–9.61 years). It is important to note that the Del Forno study did not use indexed LV measurements and included patients with severe degenerative MR and LV dilatation [24]. Thus, they looked at a slightly different patient group with more severe pathology and lacked structured follow-up, which is present in our study. Similarly, a single centre observational study by van Wijngaarden et al. [7], analysed 125 patients, finding a significant reduction in ILVESV in the Respect group (p = 0.043) but not the Resect group when comparing from baseline to one year postoperatively [7] The observational nature and limited geographical distribution of the Wijngaarden study has the potential for bias, undermining its validity and generalisability to other populations as compared to our findings [7, 23]. Furthermore, unlike our study, the limited group size and lack of data between discharge (5 days; IQR 4–6 days) and one year follow up may limit Hibino et al. and van Wijngaarden et al.’s ability to sufficiently identify group differences and track the progression of LV reverse remodelling during the first year of recovery post-MVr [7, 23]. In the early postoperative period, at 12 weeks, we found no significant difference between Resect and Respect techniques in ILVESD, ILVESV, ILVEDV and ILVEDD. According to the current literature, Respect techniques allow for a more ‘physiological repair’ through preservation of the valve geometry and leaflet mobility. Previous studies by Imasaka et al. (2015) and Mazine et al. (2018) have reported improved postoperative LV efficiency in patients undergoing Respect techniques due to the technique’s maintenance of the ventriculo-annular continuity [28, 29]. Despite the borderline difference in ILVEDV at 12 weeks (p = 0.06) between Resect and Respect groups, the difference from baseline to 12 weeks (p = 0.60) and 52 weeks (p = 0.58) was not significant. This suggests that the MVr technique used did not have clinically or statistically significant impact on the overall recovery of the LV. Another key finding from this sub-analysis is the confirmation of a two-stage phased process model for LV reverse remodelling progression post-MVr, but the phases did not align with Le Tourneau’s model [8]. Although we found a significant decrease in ILVEDV (p < 0.001) in the first stage, we observed an increase in ILVESV (p = 0.03). This was contrary to the model, which expected the ILVESV to stay the same during the first phase. In the second phase, we observed a significant decrease from baseline in both ILVESV and ILVEDV, in line with the proposed model [8]. Our study’s ability to compare Resect and Respect techniques and demonstrate the early and late changes in LV end-systolic/end-diastolic volumes and diameters post-MVr provides further clarity, as compared to previous studies [7, 23, 24]. This study has several strengths. Firstly, we have used high quality data from a large-scale multicentre RCT. Secondly, highly skilled cardiac surgeons, who met the trial’s required case experience criteria performed all the MVr operations in this trial. This ensured that there was similar relative expertise of the surgeons and effectively controlled for variability due to the learning curve. Furthermore, the decision to keep the early postoperative period at 12 weeks has ensured our study provides a better insight into the progression of LV reverse remodelling from index surgery to one-year post-MVr. Finally, all echocardiographic data was reported by a core laboratory, reducing risk of reporting bias. Limitations This study is likely to be underpowered, particularly given the discrepancy between group size, hence these results should be deemed as hypothesis generating only. Secondly, this was an observational study and is at risk of bias by confounding factors. For example, we found there to be a significant difference between ring sizes between Resect and Respect groups, with the Resect group having a larger ring size than the latter (31 vs. 33), which was unexpected and could not be theoretically or practically explained. Thirdly, the use of three-dimensional strain echocardiography measurements, as in the van Wijngaarden study, may have provided a more holistic picture of LV changes and function [7]. Finally, we recommend a longer follow-up period in future studies to assess the long-term impact of Resect and Respect approaches on MR and indexed LV diameters and volumes over time. Conclusion This study shows that in patients with MR, the MVr technique, Resect versus Respect, has no significant impact on LV reverse remodelling progression at one year postoperatively. Both techniques led to similar significant improvements in ILVESD, ILVEDD, ILVEDV and ILVESV after MVr at 52 weeks.