Mitral valve disease is the most common valve pathology in the general population and the second most common in the elderly population, with an anticipated increase in prevalence over the next decade [1]. Due to its deep and inferior location, exposure and accessibility of the mitral valve is more challenging than other valves. In the society of thoracic surgery (STS) 2019 database, the mortality rate associated with mitral valve replacement is 5%, which is higher than aortic valve replacement [2]. Although the most commonly preferred incision in mitral valve surgery is a left atriotomy anterior to the left pulmonary veins, a Sondergaard’s Groove incision with an interatrial sulcus dissection is also preferred. However, in patients with a deep chest and a long anteroposterior axis, a history of previous aortic valve surgery and a small left atrium, the exposure is challenging [3, 4]. Reports from the early 2000s concluded that transseptal atriotomies, which provide better exposure for the mitral valve, were associated with complete loss of sinus rhythm with a need for transient pacemaker insertion up to 66% and permanent pacemaker insertion of 10% in the postoperative period [5, 6, 7]. As a result, transseptal atriotomy was infrequently performed in most surgical practices.

Minimally invasive techniques are gaining popularity in cardiac surgery but the conventional left atriotomy poses significant limitations in patients who require concomitant procedures. In the last decade, there have been successful reports of concomitant cardiac operations including multiple valve procedures and minimally invasive direct coronary artery bypass (MIDCAB) plus mitral valve procedures using minimally invasive approaches, which have contributed to a resurgence in the popularity of transseptal atriotomies [8, 9]. In previous studies that reporting unfavourable results, the incision was extended to the dome of the left atrium and concomitant valve procedures were performed more frequently in the Group TS. Meta-analyses have demonstrated that these factors are significant risk factors for postoperative arrhythmias. In addition, studies comparing transseptal atriotomy with conventional left atriotomy have not excluded surgical selection bias and have not conducted analyses on diverse patient Groups [7, 10, 11, 12].

We aimed to investigate the long-term outcomes of limited transseptal atriotomy during mitral valve surgery in patients with isolated small left atria.

The study was designed as a retrospective cohort. Patient data undergoing mitral valve surgery at the SBÜ Kartal Kosuyolu High Specialization Training and Research Hospital between January 2010 and February 2014 were analyzed. 1214 patients were extracted from the hospital database. Informed consents was obtained from all patients. The ethical approval number is E-10840098-202.3.02-596. Pre-operative echocardiographic evaluations were performed routinely in all patients. Left atrial (LA) diameter was derived from the parasternal long-axis B-mode view on 2-D echocardiography. The operations were performed by 4 different surgical teams, and two of them performed a left atriotomy in all cases. One surgical team decided the incision intra-operatively and performed transseptal incisions in patients with challenging mitral valve exposure. A 4th surgical team routinely performed transseptal incisions in cases with an LA diameter below 4.5 cm. A left atrial diameter measured at 4.5 cm or below on two-dimensional echocardiography was classified as small left atrium. Emergency indications, infective endocarditis, critical preoperative conditions, pre-operatively implanted permanent pacemaker, and concomitant rhythm ablation procedures were excluded. Additionally, patients with a history of cancer, cirrhosis, or those receiving immunosuppressive medication, which could influence long-term survival, were excluded from the study to ensure accurate survival analysis. 119 patients with a small left atrium and meeting the inclusion criteria were enrolled in the study. Of these, 57 patients (47.9%) underwent a transseptal atriotomy (Group TS), while 62 patients (52.1%) underwent a left atriotomy (Group LA).

In all patients a post-operative electrocardiography (ECG) was performed shortly after transfer to the intensive care unit (ICU). An ECG was recorded in all patients and was repeated daily until discharge. A-V block other than Type 1, junctional rhythm, atrial fibrillation-flutter, amiodarone or electro-cardioversion therapies were defined as arrhythmias. In case of new-onset atrial fibrillation/flutter an electro-cardioversion was performed as the first stage therapy and treatment was continued with iv-amiodarone. An anticoagulation regimen was started on postoperative day 1 with low molecule weight heparins (LMWHs) and warfarin after the removal of all drainage catheters. The international normalized ratio (INR) goal was 2–3 in the repair group and 2.5–3.5 in the mitral mechanical prosthesis group. The anticoagulation therapy was decided based on the patients CHA₂DS₂-VASc score after 3 months; and warfarin was preferred over non-vitamine K oral anti-coagulants (NOACs). Post-operative evaluations were performed routinely at 1st week, 1st month, 1st year and 2nd year after discharge. 3 points major adverse cardiac events (MACE) rates at 2 years were obtained from the hospital database. However, further evaluation of MACE could not be performed due to inadequate contact addresses and telephones. Survival datas for all patients were obtained from the National population administration system (MERNIS).

The distribution of continuous variables were evaluated both visually (using histograms) and statistically (using the Kolmogorov-Smirnov or Shapiro-Wilk tests). For numerical variables, statistical comparisons were performed using the independent samples t-test if the data were normally distributed, and the Mann-Whitney U test if the data were not normally distributed. For categorical variables, the Chi-square and Fisher’s exact tests were used. Survival analyses were conducted using the Kaplan-Meier method and Cox regression tests.

The significance level was set at p 0.05. In comparison analysis when the difference can not reach to 0.05, it was noted as non-significant (NS).

The mean age was 52.8 $\pm$ 14 years, and 37.8% of the patients were male. A history of previous cardiac surgery was present in 13 cases (10.9%). 28 patients (23.5%) have left ventricular systolic dysfunction and the left ventricular ejection fraction (LVEF) was below 35% in 7 patients (5.8%). The mean left atrial diameter was 4.2 $\pm$ 0.3 cm. The prevalence of secondary mitral regurgitation was significantly higher in the mitral valve repair (MVr) group compared to the replacement (MVR) group (44.4% in MVr vs. 4.8% in MVR, p 0.001). Normal sinus rhythm was present in 88 patients (73.9%) during the preoperative period. No significant differences were observed between the Groups in terms of demographic characteristics (Table 2).

Postoperative variables are summarized in Table 3, and both Groups have similar outcomes. Surgical mortality occurred in 4 patients (3.4%). One patient who underwent isolated MVR died due to an atrioventricular groove rupture, two patients who underwent aortic valve replacement (AVR) plus MVR died in the early postoperative period due to low cardiac output syndrome (LCOS), and one patient who underwent CABG plus MVr died due to mediastinitis and sepsis. Permanent pacemaker implantation was required in 6 patients (5.2%) and no statistically significant differences were observed between the Groups (p $>$ 0.05). Post-operatively permanent pacemaker implantation rates for isolated mitral surgery, concomitant tricuspid valve (TV) repair, and concomitant AVR were 2.1%, 9.5% and 8% respectively. Respiratory insufficiency was defined as reintubation and found to be similar in both Groups (p $>$ 0.05).

The mean follow-up duration was 10.7 $\pm$ 4.2 years (0–14.9 years). At the end of the follow-up period, the overall survival rate in the cohort was 64.7%. 1, 5, 10, and 14-year survival rates were similar in both Groups (Table 4). MACE free survival rates at 1st year and 2nd year after surgery are comparable (91.2% and 86% in Group TS vs. 90.3% and 88.7% in Group LA respectively, p $>$ 0.05).

Survival analyses revealed no significant differences between the Groups (Group TS: 11.7 $\pm$ 0.6 years, 95% CI: 10.5–12.9; Group LA: 11.8 $\pm$ 0.6 years, 95% CI: 10.6–12.9; p $>$ 0.05) (Fig. 1). In the survival analysis based on surgical procedures, the survival rate in the concomitant CABG group was significantly lower compared to the concomitant AVR and isolated mitral groups (12.5 $\pm$ 0.5 years for mitral, 11.4 $\pm$ 1 years for concomitant AVR, and 8.2 $\pm$ 1.2 years for concomitant CABG; p 0.01) (Fig. 2). Mitral valve repair did not provide superior long-term survival compared to mitral valve replacement (10.9 $\pm$ 0.9 years for MVr vs. 12.1 $\pm$ 0.4 years for MVR; p $>$ 0.05) (Fig. 3).

In univariate analyses, several variables were identified as risk factors for long-term survival: age, obesity, chronic obstructive pulmonary disease (COPD), left ventricular systolic dysfunction, procedure type, and EuroSCORE II (p 0.05). However, in the multivariate Cox regression analysis, only age, obesity, procedure type, and left ventricular dysfunction were found to be independent risk factors for late mortality (Table 5).

Transseptal atriotomies preferred for its better exposure in challenging mitral valve cases. However, it has been associated with a high incidence of postoperative sinus node dysfunction in studies conducted between 1990 and 2000 and an incidence of permanent pacemaker insertion [5, 6, 13]. This pathology was primarily attributed to the transection of the sinus node artery during the extension of the transatrial incision to the left atrial dome. In a previous report, the superior vena cava was also transected to enhance exposure, resulting in a ten-fold increase rate of permanent pacemaker implantation in the Group TS [14]. Large-scale meta-analyses have shown that concomitant valve procedures increase the need for pacemaker implantation by up to 2.7-fold independent of the incision type [11, 12]. Studies reporting higher rates of postoperative transient and permanent pacemaker requirements in the Group TS also noted a significantly higher prevalence of redo cases and concomitant tricuspid and aortic valve interventions in this Group [7, 8]. In our study, concomitant procedures were similar in both Groups (61.4% in Group TS vs. 53.5% in Group LA; p $>$ 0.05). New-onset arrhythmias, transient pacemaker requirements and permanent pacemaker implantation were comparable in Group TS vs. Group LA (26.3% vs. 19.4%, 21.1% vs. 11.3%, 5.3% vs. 4.8% respectively; p $>$ 0.05). In studies where the incision was extended to the left atrial dome and the superior vena cava was transected, ischemic time and total perfusion time were found to be longer compared to the left atriotomy approach [5, 7, 14]. This may be caused by difficulties in the closure of the atriotomy. In our surgical technique, since the incision was not extended to the left atrial dome, the risk of injury to the sinus node artery was lower compared to other studies. Additionally, closure was easier than for extended transseptal atriotomies. Thus ischemic time and total perfusion time were similar in both Groups in our study (87.6 $\pm$ 33.5 vs. 77.4 $\pm$ 27.8 and 117.2 $\pm$ 38.4 vs. 112.3 $\pm$ 33.8 in Group TS vs. Group LA, respectively; p $>$ 0.05).

The mitral valve is more difficult to access and expose compared to the other valves. Studies reported that isolated mitral valve procedures have an operative mortality rate between 1.1% and 1.4%. However, real-world population data show that this rate can rise to 5%, and the mortality rate for mitral valve replacement (MVR) is twice as high as that for aortic valve replacement (AVR) [2, 15, 16]. In our study, the overall operative mortality rate was 3.4%, and comparable in both Groups (3.5% in Group TS vs. 3.2% in Group LA; p $>$ 0.05). Meta-analyses have shown that concomitant tricuspid valve (TV) repair does not increase mortality in mitral valve surgery [17]. Due to the limited number of patients in our study, mortality analyses were conducted independently of TV repair to avoid statistical distortion in cross-tabulation analyses. Operative mortality rates were found as 1.3% for isolated mitral valve procedures, 7.4% for concomitant AVR, and 6.7% for concomitant CABG. The median EuroSCORE II value for our patients was 2.35 $\pm$ 2.35 (range: 0.8–23.5) and our observed operative mortality rate was higher. Recent publications report operative mortality rates for concomitant AVR plus MVR ranging between 4.2% and 13.5% [8, 18], while real-world population data for concomitant CABG plus MVR indicate a rate of 9% [2]. Our mortality rates are consistent with the literature.

In mitral valve surgery, mitral valve repair has demonstrated a significant improvement in both early postoperative outcomes and long-term survival, reducing mortality by 50% [2, 18]. However, the use of biological valves in the mitral position negatively impacts survival in patients under 70 years of age [19]. In our study, mechanical valves were chosen as the prosthetic valve, and mitral repair did not show a long-term survival advantage over replacement. This may be attributed to the significantly higher prevalence of secondary mitral regurgitation (MR) in the repair group (44.4% in MVr vs. 4.8% in MVR; p 0.001). Sannino et al. [20] reported that the risk of cardiac mortality is 2.62 times higher in secondary MR, and the higher rate of secondary MR in the repair group in our study may have adversely affected the long-term survival of these patients. Concomitant CABG had worse survival rates in our study. Tsao et al. [21] reported that 1, 5, 10-year survival rates are 92%, 74% and 39% in CABG plus MVr in patients with severe secondary MY. We found similar results as the 10-year survival rate was 40% in this group. In our study, the survival rates for isolated mitral valve surgery were 95.9% at 1 year, 93.9% at 5 years, and 87.8% at 10 years. These findings are consistent with large-scale analyses, which reported similar survival rates of 95% at 5 years and 87.3% at 8 years [14, 15].

Transseptal atriotomy, when performed without extending the incision to the left atrial dome, yields similar early and long-term outcomes compared to conventional left atriotomy. In our study, no adverse effects on postoperative arrhythmias were observed with this approach.

The datasets analyzed during the current study are not publicly available because they were extracted from the hospital database. But they are available from the corresponding author on reasonable request.

DG designed and drafted the research study. MET and DG made reviewing the study critically for important intellectual content. The interpretation of the data for the work was conducted by DG and MET. Both authors made the final approval of the version to be published. Both authors agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

The Institutional Ethical Committee of Istanbul Medipol University approved the study protocol (date: 17 January 2025/reference number: E-10840098-202.3.02-596. Preoperative informed consent was obtained from all patients who underwent surgery. The study was conducted in accordance with the Declaration of Helsinki.

This research received no external funding.

The authors declare no conflict of interest.